EPA
EPA/635/R-22/191a
External Review Draft
www.epa.gov/iris
IRIS Toxicological Review of Hexavalent Chromium [Cr(VI)]
[CASRN 18540-29-9]
October 2022
Integrated Risk Information System
Center for Public Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
Toxicological Review ofHexavalent Chromium
DISCLAIMER
This document is an external review draft for review purposes only. This information is
distributed solely for the purpose of public comment It has not been formally disseminated by
EPA. It does not represent and should not be construed to represent any Agency determination or
policy. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
This document is a draft for review purposes only and does not constitute Agency policy.
ii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
CONTENTS
AUTHORS | CONTRIBUTORS | REVIEWERS xii
EXECUTIVE SUMMARY xiv
1. INTRODUCTION 1-1
1.1. OVERVIEW 1-1
1.1.1. Background 1-2
1.1.2. Chemical Properties 1-2
1.1.3. Sources, Production, and Use 1-6
1.1.4. Environmental Occurrence 1-8
1.1.5. Potential for Human Exposure 1-10
1.2.SUMMARY OF ASSESSMENT METHODS 1-13
1.2.1. Literature Search and Screening 1-13
1.2.2. Evaluation of Individual Studies 1-15
1.2.3. Data Extraction 1-16
1.2.4. Evidence Synthesis and Integration 1-17
1.2.5. Dose-Response Analysis 1-21
2. LITERATURE SEARCH AND STUDY EVALUATION RESULTS 2-1
2.1. LITERATURE SEARCH AND SCREENING RESULTS 2-1
2.2. STUDY EVALUATION RESULTS 2-3
3. HAZARD IDENTIFICATION 3-1
3.1. OVERVIEW OF PHARMACOKINETICS 3-1
3.1.1. Pharmacokinetics 3-1
3.1.2. Description of Pharmacokinetic Models 3-15
3.2. SYNTHESIS AND INTEGRATION OF HEALTH HAZARD EVIDENCE BY ORGAN/SYSTEM 3-19
3.2.1. Respiratory Tract Effects Other Than Cancer 3-19
3.2.2. Gastrointestinal Tract Effects Other Than Cancer 3-46
3.2.3. Cancer 3-62
3.2.4. Hepatic effects 3-161
3.2.5. Hematologic effects 3-187
3.2.6. Immune effects 3-202
This document is a draft for review purposes only and does not constitute Agency policy.
iii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
3.2.7. Male Reproductive Effects 3-236
3.2.8. Female Reproductive Effects 3-263
3.2.9. Developmental Effects 3-283
3.3.SUMMARY OF HAZARD IDENTIFICATION AND CONSIDERATIONS FOR DOSE-RESPONSE
ANALYSIS 3-307
3.3.1. Susceptible Populations and Life Stages 3-307
3.3.2. Effects Other Than Cancer 3-312
3.3.3. Cancer 3-318
4. DOSE-RESPONSE ANALYSIS 4-1
4.1. ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER 4-1
4.1.1. Identification of Studies for Dose-Response Analysis of Selected Effects 4-1
4.1.2. Methods of Analysis 4-8
4.1.3. Derivation of Candidate Values 4-12
4.1.4. Derivation of Organ/System-Specific Reference Doses 4-18
4.1.5. Selection of the Overall Reference Dose 4-20
4.1.6. Uncertainties in the Derivation of Reference Dose 4-20
4.1.7. Confidence Statement 4-24
4.1.8. Previous IRIS Assessment: Oral Reference Dose 4-25
4.2. INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER THAN CANCER 4-25
4.2.1. Identification of Studies for Dose-Response Analysis of Selected Effects 4-26
4.2.2. Methods of Analysis 4-36
4.2.3. Derivation of Candidate Values 4-41
4.2.4. Derivation of Organ/System-Specific Reference Concentrations 4-45
4.2.5. Selection of the Overall Reference Concentration 4-48
4.2.6. Uncertainties in the Derivation of Reference Concentration 4-48
4.2.7. Confidence Statement 4-49
4.2.8. Previous IRIS Assessment: Inhalation Reference Concentration 4-49
4.3. ORAL SLOPE FACTOR FOR CANCER 4-50
4.3.1. Analysis of Carcinogenicity Data 4-51
4.3.2. Dose-Response Analysis—Adjustments and Extrapolations Methods 4-51
4.3.3. Derivation of the Oral Slope Factor 4-53
4.3.4. Application of Age-Dependent Adjustment Factors 4-54
4.3.5. Uncertainties in the Derivation of the Oral Slope Factor 4-55
This document is a draft for review purposes only and does not constitute Agency policy.
iv DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
4.3.6. Previous IRIS Assessment: Oral Slope Factor 4-58
4.4. INHALATION UNIT RISK FOR CANCER 4-58
4.4.1. Analysis of Carcinogenicity Data 4-58
4.4.2. Dose-Response Analysis—Adjustments and Extrapolations Methods 4-66
4.4.3. Inhalation Unit Risk Derivation 4-70
4.4.4. Application of Age-Dependent Adjustment Factors 4-73
4.4.5. Uncertainties in the Derivation of the Inhalation Unit Risk 4-75
4.4.6. Previous IRIS Assessment: Inhalation Unit Risk 4-79
REFERENCES R-l
SUPPLEMENTAL INFORMATION (see Volume 2)
TABLES
Table ES-1. Organ/system-specific RfDs and overall RfD for Cr(VI) xvi
Table ES-2. Summary of reference dose (RfD) derivation xvii
Table ES-3. Organ/system-specific RfCs and overall RfCfor Cr(VI) xix
Table ES-4. Summary of reference concentration (RfC) derivation xx
Table ES-5. Summary of oral slope factor (OSF) derivation xxii
Table ES-6. Summary of inhalation unit risk (IUR) derivation xxiii
Table 1-1. Chemical identity and physicochemical properties of Cr(VI) 1-4
Table 1-2. Major anthropogenic sources of atmospheric chromium in the United States 1-8
Table 1-3. Industries and occupations that may be sources of chromium exposure 1-13
Table 1-4. Endpoint grouping categories 1-18
Table 3-1. Overall findings by system and implications for the toxicological assessment 3-3
Table 3-2. Pharmacokinetic models for Cr(VI) 3-16
Table 3-3. Summary of human studies for Cr(VI) lower respiratory effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome.3 3-21
Table 3-4. Summary of results from human studies of effects of Cr(VI) exposure on pulmonary
function 3-22
Table 3-5. Summary of results from Lindberg and Hedenstierna (1983) study of effects of Cr(VI)
exposure on pulmonary function 3-23
Table 3-6. Summary of included studies for Cr(VI) respiratory effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome.3 3-26
Table 3-7. Evidence profile table for respiratory effects other than cancer 3-41
Table 3-8. Summary of included studies for Cr(VI) Gl histopathological outcomes and overall
confidence classification 3-47
Table 3-9. Design features of studies that examined Gl tract effects via the oral route of
exposure 3-48
Table 3-10. Evidence profile table for effects in the Gl tract other than cancer 3-59
Table 3-11. Summary of human studies for Cr(VI) cancer of the Gl tract and overall confidence
classification 3-64
This document is a draft for review purposes only and does not constitute Agency policy.
v DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-12. Associations between drinking water exposures to Cr(VI) and cancer in low
confidence epidemiology studies 3-65
Table 3-13. Meta-analyses of Gl tract cancers and Cr(VI) occupational exposure 3-67
Table 3-14. Summary effect estimates from random effects meta-analysis, by cancer site and
type of effect estimate 3-70
Table 3-15. Data on neoplastic lesions in a high confidence study of rats and mice 3-73
Table 3-16. Summary of included human cross-sectional occupational studies for Cr(VI)
mutagenic effects and overall confidence classification [high (H), medium (M),
low (L)] by outcome 3-78
Table 3-17. Associations between Cr(VI) exposure and prioritized genotoxicity outcomes in
epidemiology studies3 3-81
Table 3-18. Summary of prioritized animal studies for investigating Cr(VI)-induced mutagenicity
and overall confidence classification [high (H), medium (M), low (L)] by
endpoint3 3-94
Table 3-19. Prioritized genotoxicity studies in animals exposed to Cr(VI) 3-95
Table 3-20. Evidence for key events and key event relationships involved in Cr(VI)-induced
carcinogenesis 3-114
Table 3-21. Evidence profile table for the carcinogenic mechanisms of inhaled or ingested Cr(VI) 3-144
Table 3-22. Evidence profile table for cancer of the Gl tract3 3-157
Table 3-23. Summary of human studies for Cr(VI) hepatic effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome 3-162
Table 3-24. Associations between Cr(VI) and liver clinical chemistries in epidemiology studies 3-163
Table 3-25. Summary of included animal studies for Cr(VI) liver effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome.3 3-164
Table 3-26. Evidence profile table for hepatic effects 3-180
Table 3-27. Hematologic endpoints commonly evaluated in routine blood testing 3-187
Table 3-28. Summary of human studies for Cr(VI) hematologic effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome 3-189
Table 3-29. Associations between Cr(VI) and hematologic parameters in epidemiology studies 3-190
Table 3-30. Summary of included studies for Cr(VI) hematologic effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome.3 3-191
Table 3-31. Evidence profile table for hematologic effects 3-199
Table 3-32. Summary of human studies for Cr(VI) immune effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome 3-203
Table 3-33. Associations between Cr(VI) exposure and ex vivo WBC function in epidemiology
studies 3-205
Table 3-34. Associations between Cr(VI) exposure and immunoglobulin (Ig) levels in
epidemiology studies 3-206
Table 3-35. Associations between Cr(VI) exposure and WBC counts in epidemiology studies 3-208
Table 3-36. Associations between Cr(VI) exposure and lymphocyte subpopulations in
epidemiology studies 3-210
Table 3-37. Summary of included studies for Cr(VI) immunological effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome.3 3-213
Table 3-38. Evidence profile table for immune effects 3-227
Table 3-39. Summary of human studies for Cr(VI) male reproductive effects and overall
confidence classification [high (H), medium (M), low (L)] by outcome.3 3-237
Table 3-40. Summary of results from human studies of Cr(VI) male reproductive effects 3-239
This document is a draft for review purposes only and does not constitute Agency policy.
vi DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-41. Summary of included animal studies for Cr(VI) male reproductive effects and overall
confidence classification [high (H), medium (M), low (L)] by outcome.3 3-243
Table 3-42. Evidence profile table for male reproductive outcomes 3-256
Table 3-43. Summary of included studies for Cr(VI) female reproductive effects and overall
confidence classification [high (H), medium (M), low (L)] by outcome 3-265
Table 3-44. Evidence profile table for female reproductive outcomes 3-277
Table 3-45. Summary of human studies for Cr(VI) developmental effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome.3 3-284
Table 3-46. Summary of included studies for Cr(VI) developmental effects and overall confidence
classification [high (H), medium (M), low (L)] by outcome.3 3-288
Table 3-47. Evidence profile table for developmental effects of Cr(VI) 3-301
Table 3-48. Individual and social factors that may increase susceptibility to exposure-related
health effects 3-307
Table 3-49. Dose response considerations and rationale for specific routes of exposure and
health effects 3-313
Table 3-50. Available animal studies showing histopathological changes in the duodenum 3-314
Table 3-51. Available animal studies showing histopathological and clinical chemistry changes in
the liver 3-315
Table 3-52. Available animal studies showing histopathological changes and cellular responses in
the lung 3-316
Table 4-1. Design features of high confidence studies that examined Gl tract effects
(histopathology) via the oral route of exposure 4-4
Table 4-2. Design features of studies that examined hepatic effects (clinical chemistry and
histopathology) via the oral route of exposure 4-6
Table 4-3. Summary of derivation of points of departure following oral exposure 4-12
Table 4-4. Effects and corresponding derivation of candidate values 4-16
Table 4-5. Organ/system-specific RfDs and proposed overall RfD for Cr(VI) 4-19
Table 4-6. Evaluation of epidemiology studies on Cr(VI) and nasal effects 4-29
Table 4-7. Dose-response data for effects in the nasal cavity of humans (medium confidence
studies) 4-30
Table 4-8. Design features of inhalation studies that examined effects in animals 4-33
Table 4-9. Summary of derivation of points of departure following inhalation exposure to Cr(VI).
Data for male Wistar rats from Glaser et al. (1990) 4-39
Table 4-10. Summary of derivation of points of departure following human inhalation exposure
to Cr(VI) 4-40
Table 4-11. Effects in the lower respiratory tract and corresponding derivation of candidate
values for Cr(VI) 4-44
Table 4-12. Organ/system-specific reference concentrations (RfCs) and overall RfCfor Cr(VI) 4-47
Table 4-13. Summary of the oral slope factor derivations 4-53
Table 4-14. Application of ADAFs for 70-year exposure to Cr(VI) from ages 0 to 70 4-55
Table 4-15. Summary of uncertainties in the derivation of oral slope factor values for Cr(VI) 4-57
Table 4-16. Summary of included studies considered for the derivation of an inhalation unit risk
for Cr(VI) and overall confidence classification 4-61
Table 4-17. Details of rationale for selecting a principal study on Cr(VI) for IUR derivation 4-64
Table 4-18. Results of Cox proportional hazards modeling of cumulative chromium exposure (mg
Cr03/m3-years) by different lag periods 4-68
Table 4-19. Results for relative exponential exposure-response (R&L) model adjusted for age and
smoking 4-69
This document is a draft for review purposes only and does not constitute Agency policy.
vii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 4-20. Calculation of lifetime cancer unit risk estimate not accounting for assumed
increased early-life susceptibility 4-72
Table 4-21. Calculation of total cancer unit risk estimate from adult-only exposure 4-74
Table 4-22. Total cancer risk from exposure to constant Cr(VI) exposure level of 1 ng/m3 from
ages 0-70 years, adjusted for potential increased early-life susceptibility 4-74
Table 4-23. Overview of air sampling program for the Baltimore cohort 4-76
Table 4-24. Variation in unit risks among the Cox Proportional Hazards model results by lag
length 4-77
FIGURES
Figure ES-1. General assumptions regarding absorption and distribution of Cr(VI) ingested by
rodents during ad libitum drinking water or dietary bioassays xxvi
Figure 1-1. Sources of Cr(VI) in soil and groundwater 1-7
Figure 1-2. Drinking water chromium (VI) concentrations in the United States by EPA region 1-10
Figure 2-1. Literature search flow diagram for Cr(VI) 2-2
Figure 3-1. Overview of the absorption, distribution, metabolism, and excretion of Cr(VI), with
focus on extracellular transport and metabolism at portals of entry 3-2
Figure 3-2. Reduction of Cr(VI) in samples of human gastric juice (fasted subjects) using data
from Proctor et al. (2012). For these experiments, stomach contents were
diluted 10:1 to highlight the effect of pH 3-6
Figure 3-3. Gl tract pH values reported in Mcconnell et al. (2008) (rodents: female BALB/c mice
and female Wistar rats) and Parrott et al. (2009) (humans) 3-6
Figure 3-4. Schematic of the rat oral cavity depicting the gradient of Cr(VI) concentration
following ingestion of Cr(VI) in drinking water, both from anterior to posterior
locations, as well as across the tissue depth 3-9
Figure 3-5. Schematic of the mouse upper Gl tract (stomach and small intestine) depicting the
gradient of Cr(VI) concentration following ingestion of Cr(VI) in drinking water 3-10
Figure 3-6. Schematics of the human respiratory system 3-13
Figure 3-7. Intracellular reduction pathways of Cr(VI) 3-15
Figure 3-8. Relationship between ex vivo reduction models, in vivo gastric models, and whole-
body PBPK models 3-17
Figure 3-9. Lung cellular responses in BALF in male animals 3-29
Figure 3-10. Histopathological results and effects in macrophages in male rat lungs 3-31
Figure 3-11. Lung weight in male animals 3-33
Figure 3-12. Diffuse epithelial hyperplasia in Cr(VI) treated mice in high confidence studies 3-48
Figure 3-13. Diffuse epithelial hyperplasia in Cr(VI) treated rats in high confidence studies 3-49
Figure 3-14. Cr(VI)-induced degenerative changes in the small intestines of mice and rats in high
confidence studies 3-52
Figure 3-15. Fractional incidence of mice with adenomas or carcinomas in the small intestine (SI
tumors), and fractional incidence of rats with squamous cell carcinomas or
papillomas in the oral mucosa or tongue (oral tumors) 3-74
Figure 3-16. Key events and mechanistic pathways induced by Cr(VI) exposure that can lead to
cancer 3-113
Figure 3-17. Reported tumors of the digestive tract tissues for all rodents exposed to Cr(VI) 3-121
Figure 3-18. Cellular processes involved in the mutagenic MOA of Cr(VI) 3-138
This document is a draft for review purposes only and does not constitute Agency policy.
viii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Figure 3-19. Hepatic effects of oral Cr(VI) exposure in animals (histopathology) 3-170
Figure 3-20. Hepatic effects of oral Cr(VI) exposure in animals (clinical chemistry) 3-173
Figure 3-21. Hepatic effects of oral Cr(VI) exposure in animals (relative liver weight) 3-175
Figure 3-22. Hematology findings from NTP (2007) and NTP (2008) in rats and mice exposed by
gavage to Cr(VI) for 90 days or 12 months 3-193
Figure 4-1. Evaluation of studies from the Cr(VI) hazard identification for derivation of toxicity
values 4-3
Figure 4-2. Process for calculating the human equivalent dose for Cr(VI) 4-10
Figure 4-3. Candidate values with corresponding POD and composite UF 4-17
Figure 4-4. Evaluation of animal studies from the Cr(VI) hazard identification for derivation of
toxicity values. Low confidence studies were not considered 4-32
Figure 4-5. Dose-response relationship for lung histopathological in male rats 4-34
Figure 4-6. Dose-response relationship for selected endpoints in male rats using data from
Glaser et al. (1990). Data (± 95% confidence interval) are for 90-day observation
time immediately following exposure, and 120-day observation time (90 days of
exposure followed by a 30-day period of no exposure) 4-35
Figure 4-7. Candidate values with corresponding POD and composite UF 4-45
Figure 4-8. BMDS 3.2 graphical output of selected models for dose-response of cancer data in
male and female rats and mice) 4-52
Figure 4-9. Literature screening results for studies containing exposure-response data of Cr(VI)
and lung cancer 4-60
This document is a draft for review purposes only and does not constitute Agency policy.
ix DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
ABBREVIATIONS
ADAF age-dependent adjustment factors HERO
ADME absorption, distribution, metabolism,
and excretion i.p.
AIC Akaike's information criterion i.v.
ALT alanine aminotransferase IRIS
ALP alkaline phosphatase LCso
Asc ascorbate LDso
AST aspartate aminotransferase LDH
ATSDR Agency for Toxic Substances and LOAEL
Disease Registry MCH
BAL bronchoalveolar lavage MCHC
BALF bronchoalveolar lavage fluid MCV
BMD benchmark dose MEF
BMDL benchmark dose lower confidence limit MMAD
BMDS Benchmark Dose Software MN
BMI body mass index MOA
BMR benchmark response MTD
BMDC bone marrow-derived stem cell CPHEA
BW body weight
CA chromosomal aberration
CASRN Chemical Abstracts Service Registry NOAEL
Number
CHO Chinese hamster ovary (cell line cells)
CPHEA Center for Public Health and NTP
Environmental Assessment NZW
CL confidence limit ORD
CNS central nervous system OSHA
Cr(III) trivalent chromium
Cr(IV] tetravalent chromium PBPK
Cr(V] pentavalent chromium PDC
Cr(VI] hexavalent chromium PND
DAF dosimetric adjustment factor POD
DLCO diffusing capacity of carbon monoxide POD[adj]
DNA deoxyribonucleic acid POD[hed]
ELF epithelial lining fluid POD[hec]
EPA Environmental Protection Agency
ER extra risk RBC
FDA Food and Drug Administration
FEV1.0 forced expiratory volume of 1 second RD
FVC forced vital capacity RfC
GD gestation day RfD
GGT y-glutamyl transferase RDDR
GI gastrointestinal RNA
GLP good laboratory practices SCE
GSD geometric standard deviation SD
GSH glutathione SDH
GST glutathione-S-transferase SE
Hgb hemoglobin SDD
HEC human equivalent concentration PK
HED human equivalent dose TSCATS
Health and Environmental Research
Online
intraperitoneal
intravenous
Integrated Risk Information System
median lethal concentration
median lethal dose
lactate dehydrogenase
lowest-observed-adverse-effect level
mean cell hemoglobin
mean cell hemoglobin concentration
mean cell volume
maximal expiratory flow
mas median aerodynamic diameter
micronuclei
mode of action
maximum tolerated dose
Center for Public Health and
Environmental Assessment NCI
National Cancer Institute
no-observed-adverse-effect level
National Toxicology Program
New Zealand White (rabbit breed)
Office of Research and Development
Occupational Safety and Health
Administration
physiologically based pharmacokinetic
potassium dichromate
postnatal day
point of departure
duration-adjusted POD
human equivalent dose POD
human equivalent concentration POD
red blood cell, also known as
erythrocyte
relative deviation
inhalation reference concentration
oral reference dose
regional deposited dose ratio
ribonucleic acid
sister chromatid exchange
standard deviation
sorbitol dehydrogenase
standard error
sodium dichromate dihydrate
pharmacokinetics
Toxic Substances Control Act Test
Submissions
This document is a draft for review purposes only and does not constitute Agency policy.
x DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
TWA
time-weighted average
UF
uncertainty factor
UFa
animal-to-human uncertainty factor
UFh
human variation uncertainty factor
UFl
LOAEL-to-NOAEL uncertainty factor
UFs
subchronic-to-chronic uncertainty
factor
UFd
database uncertainty factor
WOS
Web of Science
This document is a draft for review purposes only and does not constitute Agency policy.
xi DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
AUTHORS | CONTRIBUTORS | REVIEWERS
Assessment Managers (Lead Authors)
Catherine Gibbons U.S. EPA/Office of Research and Development/Center
Alan Sasso for Public Health and Environmental Assessment
Assessment Team (Authors)
Michelle Angrish U.S. EPA/Office of Research and Development/Center
Xabier Arzuaga for Public Health and Environmental Assessment
Thomas Bateson
Krista Christensen*
Johanna Congleton
Barbara Glenn
Leonid Kopylev
David Lehmann
Roman F. Mezencev
Rebecca M. Nachman
Kathleen Newhouse
Elizabeth Radke
Paul Reinhart
Susan Rieth
Paul Schlosser
Rachel Shaffer
Andre Weaver
Amina Wilkins
Erin Yost
Contributors
Nora Abdel-Gawad U.S. EPA/Office of Research and Development/Center
Ted Berner for Public Health and Environmental Assessment
Todd Blessinger*
Christine Cai
Glinda Cooper
Kelly Garcia
Carolyn Gigot
Andrew Greenhalgh
Shahreen Hussain
Grace Kaupas
Stephanie Kim
Urmila Kodavanti
Alexandra Larsen
Cheng-Kuan (Calvin) Lin
Larissa Pardo
Keith Salazar
Sabah Tariq
Todd Zurlinden*
Stephanie Smith-Roe NIH/National Institute of Environmental Health
Sciences/Division of the National Toxicology Program
This document is a draft for review purposes only and does not constitute Agency policy.
xii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Production Team
Maureen Johnson
Ryan Jones
Dahnish Shams
Vicki Soto
Samuel Thacker
Ashlei Williams
Contractor Support
Michelle Cawley ICF, Inc.
Ali Goldstone
Kim Osborn
Alessandria Schumacher
Nicole Vetter
Executive Direction
Wayne Cascio
Samantha Jones
Kristina Thayer*
Andrew Kraft*
I. Allen Davis. Ir.
Barbara Glenn
Paul White*
Internal Reviewers
Epidemiology Workgroup
General Toxicity/Immunotoxicity/Cancer
Workgroup
Inhalation Workgroup
Neurotoxicity Workgroup
Pharmacokinetics Workgroup
Reproductive and Developmental Toxicity
Workgroup
Scoping and Problem Formulation
Workgroup
Statistics Workgroup
Systematic Review Workgroup
Toxicity Pathways Workgroup
Ila Cote* U.S. EPA (retired), Contractor
Alan Stern* NJDEP (retired), Contractor
*EPA/ORD/CPHEA Executive Review Committee
U.S. EPA/Office of Research and Development/Center
for Public Health and Environmental Assessment
CPHEA Center Director
CPHEA Associate Director for Assessment Science
CPAD Director
CPAD Associate Director
CPAD Senior Science Advisor, Acting
SAM Branch Chief
CPAD Senior Science Advisor
This document is a draft for review purposes only and does not constitute Agency policy.
xiii DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Toxicological Review ofHexavalent Chromium
EXECUTIVE SUMMARY
Summary of Occurrence and Health Effects
Chromium is a ubiquitous element present in soil, water, air, and food that can
originate from both natural and anthropogenic sources. This toxicological review
restricts its focus to hexavalent chromium compounds, which are a group of
substances that contain chromium in the hexavalent (+6) oxidation state, denoted as
Cr(VI). Cr(VI) compounds have many industrial applications, including pigment
manufacturing, corrosion inhibition and metal finishing. Because many Cr(VI)
compounds are water soluble, they are highly mobile in soil and may contaminate
drinking water. Cr(VI) may be emitted into air by industries using Cr(VI) compounds,
and by various other sources such as the burning of fossil fuels.
The systematic review (see Appendix A for methods] conducted to support this
assessment evaluated all cancer outcomes, and noncancer effects for the following
potential target systems: respiratory, gastrointestinal (GI] tract, hepatic, hematologic,
immune, reproductive, and developmental. For cancer and nasal effects via the
inhalation route (which are well established], the systematic review focused on data
that may inform the quantitative dose-response analysis.
Evidence indicates that Cr(VI] is likely to cause GI tract, liver, developmental, and
lower respiratory toxicity in humans. Evidence suggests that Cr(VI] may cause male
reproductive effects, immune effects, and hematologic toxicity in humans. Evidence
is inadequate to assess whether Cr(VI] causes female reproductive toxicity in
humans. Organ/system-specific reference values were derived for GI tract, liver,
developmental, hematological, lower respiratory, and nasal effects. The overall
chronic RfD is 9 x 10"4 mg/kg-d, and the overall chronic RfC is 1 x 10"5 mg/m3.
For cancer via the oral route of exposure, Cr(VI] is likely to be carcinogenic to the
human GI tract. Because a mutagenic mode-of-action (MOA] for Cr(VI]
carcinogenicity is "sufficiently supported in (laboratory] animals" and "relevant to
humans," EPA used a linear low dose extrapolation from the POD in accordance with
Guidelines for Carcinogen Risk Assessment fU.S. EPA. 2005al Furthermore, in the
absence of chemical-specific data to evaluate differences in age-specific susceptibility,
increased early-life susceptibility to Cr(VI] is assumed and EPA applied ADAFs in
accordance with the Supplemental Guidance for Assessing Susceptibility from Early-
Life Exposure to Carcinogens (U.S. EPA. 2005b], The total lifetime oral slope factor
(OSF] for Cr(VI] is 0.5 (per mg/kg-d].
For cancer via the inhalation route of exposure, quantitative exposure-response data
were evaluated, and an inhalation unit risk (IUR] was developed for human lung
cancer. Similar to the oral route of exposure, linear low dose extrapolation and
application of ADAFs were performed for the inhalation route of exposure. The total
lifetime IUR for Cr(VI] is 2 x 10~2 (per [ig Cr(VI]/m3].
This document is a draft for review purposes only and does not constitute Agency policy.
xiv DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
ES.1 EVIDENCE FOR HAZARDS OTHER THAN CANCER: ORAL EXPOSURE
The evidence indicates that Cr(VI) is likely to cause gastrointestinal (GI) tract, hepatic, and
developmental toxicity in humans following oral ingestion (see Sections 3.2.2, 3.2.4, 3.2.9). The
determination that evidence indicates that Cr(VI) is likely to cause GI toxicity in humans was based
on toxicology studies in rodents reporting histological effects in the GI tract For the determination
of hepatic toxicity, toxicology studies in rodents reported histological effects in the liver and serum
indicators of hepatotoxicity. The determination for developmental effects was based on the
observation of decreased offspring growth across most animal studies. For the hazards listed
above, mechanistic evidence supported the human relevance of the effects observed in animals.
The evidence suggests that Cr(VI) may cause immune, hematologic, and male reproductive
toxicity in humans (see Sections 3.2.5, 3.2.6, 3.2.7). Male reproductive effects on sperm parameters
and testosterone were observed in both human and animal studies, however most studies were
considered low confidence, and effects were inconsistent among the high confidence rodent studies.
For hematological effects, high confidence studies in rodents reported changes in hematological
parameters that suggested a pattern consistent with regenerative microcytic hypochromic anemia,
but the confidence in this judgment was diminished due to uncertainty regarding the apparent
transient nature of the effects. The conclusion for immune effects was primarily based on coherent
evidence of effects on 1) ex vivo WBC function across human and animal studies, 2) antibody
responses to T cell-dependent antigen measured in animals, and 3) reduction in host resistance to
bacterial infection reported in animal studies; however, confidence in the evidence was reduced
due to primarily low confidence studies reporting findings that were often inconsistent across
studies.
The evidence is inadequate to assess whether Cr(VI) causes female reproductive toxicity in
humans (see Section 3.2.8). Although an association with female reproductive toxicity was
demonstrated in a single low confidence epidemiology study and a series of low confidence animal
toxicology studies, effects were not observed in medium or high confidence studies aside from a
moderate decrease in maternal body weight
ES.1.1. Oral Reference Dose (RfD)
Hyperplasia in the small intestine of female B6C3F1 mice was selected as the basis for the
overall chronic RfD of 9 x 10~4 mg/kg-d. A LOAEL analysis was used to derive an organ/system-
specific point of departure (POD) for GI tract effects. Human equivalent doses (HEDs) were
calculated using PBPK modeling to account for species differences and human variability in
detoxification of Cr(VI) in the stomach. A composite uncertainty factor of 100 was applied. This
uncertainty factor incorporated: an interspecies uncertainty (UFa) of 3 to account for animal-to-
human extrapolation (pharmacodynamic differences); an intraspecies uncertainty (UFh) of 3 to
account for variation in susceptibility across the human population, and the possibility that the
available data may not be representative of individuals who are most susceptible to the effects; and
This document is a draft for review purposes only and does not constitute Agency policy.
xv DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
a LOAEL-to-NOAEL uncertainty (UFl) of 10 to account for extrapolation from the LOAEL. The
remaining uncertainty factors were equal to 1.
The confidence in the overall chronic RfD is high. The RfD is based on a high confidence
chronic 2-year drinking water study by NTP (20081 that exposed rats and mice of both sexes to
Cr(VI) as sodium dichromate dihydrate (see Section 3.2.2). Multiple high confidence subchronic
studies also support these data, and mechanistic studies support the involvement of oxidative
stress in Cr (VI)-induced cytotoxicity in a variety of tissues, including the GI tract The
organ/system-specific RfD for the liver (hepatic system) is also supportive of the GI tract RfD,
because the GI tract and liver are exposed on first-pass following oral ingestion (so both should get
the highest internal dose). While the human database for Cr(VI) induced GI toxicity was
indeterminate, this did not warrant changing the overall confidence from high. Organ/system-
specific RfDs (osRfDs) are listed in Table ES-1.
Table ES-1. Organ/system-specific RfDs and overall RfD for Cr(VI)
Hazard
Basis
osRfD
mg/kg-d
Study exposure
description
Confidence
Gastrointestinal
system (GI tract)
Hyperplasia in small
intestine of female
mice
9 x 10"4
Chronic drinking
water
High
Hepatic system
Chronic inflammation
in female rats
7 x 10"4
Chronic drinking
water
High
Developmental
toxicity
Decreased F1
offspring postnatal
growth
0.07
Continuous breeding
Low
Hematological
toxicity
Decreased Hgb (male
rats)
0.01
Subchronic drinking
water
High
Overall RfD
GI tract effects
9 x 10"4
Chronic drinking
water
High
The osRfD for hepatic effects was based on chronic inflammation in female F344 rats
reported in NTP (2008). An osRfD of 7x 10-4 mg/kg-d was derived using a LOAEL analysis. Human
equivalent doses (HEDs) were calculated using pharmacokinetic modeling to account for species
differences and human variability in detoxification of Cr(VI) in the stomach. A composite
uncertainty factor of 100 was applied. This uncertainty factor incorporated: an interspecies
uncertainty (UFa) of 3 to account for animal-to-human extrapolation (pharmacodynamic
differences); an intraspecies uncertainty (UFh) of 3 to account for variation in susceptibility across
the human population, and the possibility that the available data may not be representative of
individuals who are most susceptible to the effects; and a LOAEL-to-NOAEL uncertainty (UFl) of 10
to account for extrapolation from the LOAEL. The remaining uncertainty factors were equal to 1.
There is high confidence in this osRfD. It is based on a high confidence chronic study in rats and
This document is a draft for review purposes only and does not constitute Agency policy.
xvi DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Toxicological Review ofHexavalent Chromium
there are other subchronic data and mechanistic evidence to support the liver endpoints (see
Section 3.2.4).
The osRfD for developmental toxicity was based on decreased F1 offspring postnatal
growth from the continuous breeding study in BALBC mice (NTP. 19971. The osRfD was 0.07
mg/kg-d and was based on extrapolation from a NOAEL. A human equivalent dose (HED) was
calculated using PBPK modeling to account for species differences and human variability in
detoxification of Cr(VI) in the stomach. A composite uncertainty factor of 10 was applied. This
uncertainty factor incorporated: an interspecies uncertainty (UFa) of 3 to account for animal-to-
human extrapolation (pharmacodynamic differences); an intraspecies uncertainty (UFh) of 3 to
account for variation in susceptibility across the human population, and the possibility that the
available data may not be representative of individuals who are most susceptible to the effects. The
remaining uncertainty factors were equal to 1. There is low confidence in this osRfD. While it is
based on a high confidence continuous breeding study and similar effects on decreased offspring
growth observed in multiple other studies (see Section 3.2.9), this effect only occurred in high dose
groups where other toxicological effects (as indicated by the lower points of departure in Table
ES-2) may be occurring. Lower confidence in this osRfD was assigned due to the possibility that
other toxicities could be affecting the animals in the high dose groups where developmental effects
were observed.
The osRfD for hematological toxicity was based on decreased Hgb in male F344 rats at 22
days reported in NTP f20081. Hematological effects were observed to have the highest magnitude
at short time periods, and ameliorate over time. As a result, short-term/low-dose data from NTP
f20081 were used, and a subchronic-to-chronic uncertainty factor was not applied. An osRfD of
0.01 mg/kg-d was derived using BMD analysis and PBPK modeling. A composite uncertainty factor
of 10 was applied. This uncertainty factor incorporated: an interspecies uncertainty (UFa) of 3 to
account for animal-to-human extrapolation (pharmacodynamic differences); an intraspecies
uncertainty (UFh) of 3 to account for variation in susceptibility across the human population, and
the possibility that the available data may not be representative of individuals who are most
susceptible to the effects. There is high confidence in this osRfD. It is based on a high confidence
study in rats and there are other subchronic data and mechanistic evidence to support the endpoint
(see Section 3.2.5).
Table ES-2. Summary of reference dose (RfD) derivation
Critical effect
Point of departure
mg/kg-d
UF
Candidat Value
(mg/kg-d)
osRfD
(mg/kg-d)
Gl TRACT TOXICITY
Mice (M) diffuse epithelial hyperplasia of
duodenum3 (NTP, 2008)
BMDLio%er-hed: 0.0443
10
4.43 x 10"3
9 x 10"4
This document is a draft for review purposes only and does not constitute Agency policy.
xvii DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofHexavalent Chromium
Critical effect
Point of departure
mg/kg-d
UF
Candidat Value
(mg/kg-d)
osRfD
(mg/kg-d)
Mice (F) diffuse epithelial hyperplasia of
duodenum3 (NTP, 2008)
LOAELhed: 0.0911
100
9.11 x 10"4
HEPATIC TOXICITY
Rat (M) liver ALT (12 months) (NTP, 2008)
BMDLird-hed: 0.204
10
0.0204
Rat (M) liver ALT (3 months) (NTP, 2008)
NOAELhed: 0.191
30
6.37 x 10"3
Rat (M) liver ALT (90 davs) (NTP, 2007)
LOAELhed: 0.203
300
6.77 x 10"4
Rat (F) liver ALT (90 davs) (NTP, 2007)
LOAELhed: 0.190
300
6.33 x 10"4
7 x 10"4
Rat (F) liver chronic inflammation (2 years)
(NTP, 2008)
LOAELhed: 0.0669
100
6.69 x 10"4
Mouse (F) liver chronic inflammation (2 years)
(NTP, 2008)
BMDLio%er hed: 0.182
10
0.0182
Rat (F) liver fattv change (2 vears) (NTP, 2008)
NOAELhed: 0.0669
10
6.69 x 10"3
DEVELOPMENTAL TOXICITY
Mouse (F) Decreased F1 postnatal growth
(NTP, 1997)
NOAELhed: 0.700
10
0.0700
0.07
HEMATOLOGICAL TOXICITY
Rat (M) decreased Hgb (22 davs) (NTP, 2008)
BMDLisdhed: 0.126
10
0.0126
0.01
aDuodenum: the most proximal subsection of the small intestine, immediately distal to the stomach.
ES.2 EVIDENCE FOR HAZARDS OTHER THAN CANCER: INHALATION EXPOSURE
As stated in the Cr(VI) IRIS Assessment Protocol (Appendix A), EPA did not re-evaluate the
qualitative evidence for an association between inhalation Cr(VI) exposure and nasal effects. Based
on EPA's 1998 evaluation of the literature and the determination that the effects of Cr(VI) on the
nasal cavity have been well established [e.g., OSHA f20061 and U.S. EPA f2014cl], hazard
identification was not performed for nasal effects. Rather, the review of the evidence for nasal
effects focused on identifying studies that might improve the quantitative dose-response analysis
for this outcome.
EPA evaluated qualitative evidence for an association between inhalation Cr(VI) exposure
and lower respiratory toxicity. EPA determined that Cr(VI) is likely to cause lower respiratory
toxicity, based on evidence in six medium confidence animal studies examining lung cellular
responses and/or histopathology. Because histopathological and cellular changes occurred
This document is a draft for review purposes only and does not constitute Agency policy.
xviii DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofHexavalent Chromium
together, and in combination with serum biomarkers indicating an inflammatory response, these
were considered indicators of adverse responses. The human evidence for Cr(VI)-induced lower
respiratory effects is limited in terms of number and confidence of studies. However, three of the
available five studies provide some indication of exposure-related decrements in lung function
assessed using spirometry. Mechanistic evidence supports the respiratory tract effects observed in
animals.
ES-2.2 Inhalation Reference Concentration (RfC)
The overall RfC was based on effects in the upper respiratory tract (ulceration of the nasal
septum) reported by medium confidence studies. Effects of Cr(VI) on the nasal cavity have been
well established to occur in humans, and this was also the most sensitive effect. It is considered
protective of the other noncancer effects. Organ/system-specific RfCs are listed in Table ES-3.
Table ES-3. Organ/system-specific RfCs and overall RfC for Cr(VI)
Hazard
Basis
osRfC
mg/m3
Study exposure
description
Confidence
Respiratory (upper
tract)
Ulcerated nasal
septum in humans
1 X 10-5
Occupational
longitudinal study
Medium
Respiratory3 (lower
tract)
Lung cellular
responses and
histopathological
changes in rats
1 x lO"4
Subchronic study
Medium
Overall RfC
Respiratory effects
1 X 10-5
Occupational
longitudinal study
Medium
aHuman equivalent concentrations were calculated using a dosimetric adjustment factor accounting for
interspecies differences in particle deposition (the regional deposited dose ratio, or RDDR).
Effects in the nasal cavity included irritation/ulceration of the nasal mucosa or septum,
perforation of the septum, and bleeding nasal septum. The osRfC (for upper respiratory tract) was
derived using data of nasal septum ulceration in humans from Gibb etal. (2000a). LOAEL analyses
were used to derive the upper respiratory tract related points of departure (POD). A composite
uncertainty factor of 300 was applied. This uncertainty factor incorporated: an intraspecies
uncertainty factor (UFh) of 3 to account for variation in susceptibility across the human population
and the possibility that the available data may not be representative of individuals who are most
susceptible to the effect; a LOAEL-to-NOAEL uncertainty factor (UFl) of 10 because this endpoint
had a high incidence at the lowest concentration across multiple studies; and a subchronic-to-
chronic uncertainty factor (UFs) of 3 because data were not from chronic lifetime exposures
(however the effects had a short onset time). A database uncertainty factor (UFd) of 3 was applied
because multi-generational inhalation studies were not available in animals, human prenatal
studies were rated low confidence, and effects of Cr(VI) differ by route of exposure due to
This document is a draft for review purposes only and does not constitute Agency policy.
xix DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofHexavalent Chromium
pharmacokinetics1 (thus, the oral database of multi-generational studies does not inform the
quantitative analysis for the inhalation route).
For the lower respiratory tract, the osRfC was derived using data of lung cellular responses
and histopathological changes in rats from Glaser etal. (19901. A LOAEL analysis was used to
derive most organ/system-specific points of departure (PODs). Human equivalent concentrations
were calculated using a dosimetric adjustment factor accounting for interspecies differences in
particle deposition (the regional deposited dose ratio, or RDDR). A composite uncertainty factor of
1000 was applied to the LOAEL-derived PODs (BMD-derived bronchioalveolar hyperplasia had a
composite UF was 300; see Section 4.2.4). The database uncertainty factor, UFd, was 3 for the same
reasons specified above for the nasal osRfC. A subchronic-to-chronic uncertainty factor, UFs, of 3
was incorporated to account for the less-than-lifetime exposure. There was some indication in
Glaser etal. (1990) that the effects were transient, and therefore a 10 was not applied; however,
there is still uncertainty due to the lack of long-term data for continuous chronic exposure. An
interspecies uncertainty factor, UFa, of 3 was applied to account for residual uncertainty in the
extrapolation from laboratory animals to humans (an inhalation dosimetry factor was used to
estimate a human equivalent concentration from animal data, but some pharmacodynamic
uncertainty remained). A LOAEL-to-NOAEL uncertainty factor, UFl, of 3 was applied to LOAELs
because characteristics of the lung histopathological and cellular responses supported a value less
than 10. UFl of 1 was applied when BMD modeling was used (bronchioalveolar hyperplasia). An
intraspecies uncertainty factor, UFh, of 10 was applied to account for variability and uncertainty in
pharmacokinetic and pharmacodynamic susceptibility within the human population (source data
were only available in male inbred rats). Table ES-4 summarizes the derivation of the osRfCs.
Table ES-4. Summary of reference concentration (RfC) derivation
Critical effect
Point of departure
mg/m3
UF
Candidate value
mg/m3
osRfC
mg/m3
UPPER RESPIRATORY TRACT TOXICITY
Ulceration of the nasal septum (Gibb et
a 1., 2000a)
LOAEL: 3.4 x 10"3
300
1.1 X 10-5
1 x 10-5
Nasal mucosal patholoev (Cohen et al.,
1974)
LOAEL: 9.5 x 10"4
300
3.2 x lO"6
Ulceration of the nasal septum
(Lindberg and Hedenstierna, 1983)
LOAEL: 6.6 x 10"4
300
2.2 x 10"6
LOWER RESPIRATORY TRACT TOXICITY
Histopathology: histiocytosis in rats
(Glaser et al., 1990)
LOAELhec: 0.133
1000
1.3 x lO"4
1 X 10-4
1 Because Cr(VI) is detoxified in the gut on first-pass, it is possible that inhalation exposures may induce
systemic effects not observed following ingestion.
This document is a draft for review purposes only and does not constitute Agency policy.
xx DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofHexavalent Chromium
Critical effect
Point of departure
mg/m3
UF
Candidate value
mg/m3
osRfC
mg/m3
UPPER RESPIRATORY TRACT TOXICITY
Histopathology: bronchioalveolar
hyperplasia in rats (Glaser et al., 1990)
BMDLisd-hec: 0.0413
300
1.4 x lO"4
Cell responses: LDH in BALF in rats
(Glaser et al., 1990)
LOAELhec: 0.133
1000
1.3 x lO"4
Cell responses: Albumin in BALF in rats
(Glaser et al., 1990)
LOAELhec: 0.170
1000
1.7 x lO"4
Cell responses: Total protein in BALF in
rats (Glaser et al., 1990)
LOAELhec: 0.133
1000
1.3 x lO"4
ES.3 EVIDENCE FOR HUMAN CARCINOGENICITY
Under EPA's Guidelines for Carcinogen Risk Assessment fU.S. EPA. 2005al Cr(VI) is likely
to be carcinogenic to humans by the oral route of exposure. The evidence of carcinogenicity to the
GI tract from animal studies is robust, and the evidence of carcinogenicity from human studies is
slight. There is strong supporting mechanistic evidence for Cr(VI) involvement in biological
pathways contributing to carcinogenesis.
As noted in the Protocol (see Appendix A), this assessment maintains the previous
determination that Cr(VI) is carcinogenic to humans by the inhalation route of exposure based on
long-standing evidence of a causal relationship between inhalation of Cr(VI) and increased
incidence of lung cancer in humans in occupational settings.
ES.4 QUANTITATIVE ESTIMATE OF CARCINOGENIC RISK: ORAL EXPOSURE
The animal database for cancer by oral exposure consisted of a high confidence chronic
2-year drinking water bioassay which found "clear evidence of carcinogenic activity" of Cr(VI) in
male and female rats and mice (NTP. 2008). These results were based on increased incidences of
squamous cell neoplasms in the oral cavity of rats, and increased incidences of neoplasms in the
small intestine of mice. Using these data, benchmark dose (BMD) modeling was applied to derive
points of departure (PODs) for small intestinal tumors in mice and oral tumors in rats (See
Section 4.3). For mice, human equivalent doses (HEDs) were calculated using PBPK modeling to
account for species differences in detoxification of Cr(VI) in the stomach because tumors occurred
in the small intestine (after stomach reduction to Cr(III)). For rats, HEDs were calculated using
BW3/4scaling in accordance with U.S. EPA (2011c). because tumors occurred in the oral cavity
(prior to stomach reduction to Cr(III)). In the absence of an adequately developed theory or
information to develop and characterize an oral portal-of-entry dosimetric adjustment factor,
application of BW3/4 scaling is recommended (U.S. EPA. 2011c. 2005a).
This document is a draft for review purposes only and does not constitute Agency policy.
xxi DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Toxicological Review ofHexavalent Chromium
The lifetime oral cancer slope factor for humans is defined as the slope of the line from the
lower 95% bound on the exposure at the POD to the control response (slope factor = 0.1/BMDLio).
Using linear extrapolation from the BMDLio, human equivalent oral slope factors were derived for
each sex/species/tumor site combination and are listed in Table ES-5. The adult-based oral slope
factor for Cr(VI) is 0.3 (per mg/kg-d), based on tumors of the small intestine of male and female
mice.
Table ES-5. Summary of oral slope factor (OSF) derivation
Critical effect
Point of departure
mg/kg-d
Human
equivalent
dose mg/kg-d
OSFa
(per mg/kg-d)
Confidence
Adenomas or carcinomas in the
mouse small intestine of male mice
(NTP, 2008)
BMDLio%er: 1.05
0.319b
0.313
High
Adenomas or carcinomas in the
mouse small intestine of female mice
(NTP, 2008)
BMDLio%er: 1.03
0.316b
0.317
High
Squamous cell carcinoma or
squamous cell papilloma in oral
mucosa or tongue of male rats (NTP,
2008)
BMDLio%er: 3.37
0.923°
0.108
High
Squamous cell carcinoma or
squamous cell papilloma in oral
mucosa or tongue of female rats
(NTP, 2008)
BMDLio%er: 2.70
0.645°
0.155
High
Adult-based OSF: 0.3 (mg/kg-d)1 (rounded from either 0.313 or 0.317)
Lifetime OSF for adenomas or carcinomas in the mouse small intestine, after application of the age-dependent
adjustment factors: 0.5 (mg/kg-d)1 (see Section 4.3.4 for derivation)
aOSF prior to application of the age-dependent adjustment factors.
Estimated by PBPK modeling.
CBW3/4 scaling adjustment (administered dose multiplied by (BWa/BWh)1/4, where BWh = 80kg (human body
weight) and BWa (animal body weight) is set to a study-specific value.
Because a mutagenic MOA for Cr(VI) carcinogenicity (see Section 3.2.3) is "sufficiently
supported in (laboratory) animals" and "relevant to humans," and as there are no chemical-specific
data to evaluate the differences between adults and children, increased early-life susceptibility
should be assumed. If there is early-life exposure, age-dependent adjustment factors (ADAFs)
should be applied, as appropriate, in accordance with the EPA's Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA. 2005b).
The total lifetime OSF for Cr(VI) is 0.5 (per mg/kg-d). Partial oral slope factors for
different age groups are provided in Section 4.3.4.
This document is a draft for review purposes only and does not constitute Agency policy.
xxii DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Toxicological Review ofHexavalent Chromium
ES.5 QUANTITATIVE ESTIMATE OF CARCINOGENIC RISK: INHALATION EXPOSURE
In 1998, the EPA IRIS Toxicological Review ofHexavalent Chromium classified Cr(VI) as a
"known human carcinogen by the inhalation route of exposure" based on consistent evidence that
inhaled Cr(VI) causes lung cancer in humans and supporting evidence of carcinogenicity in animals
(U.S. EPA. 1998c). The same conclusion has since been reached by other authoritative federal and
state health agencies and international organizations and the carcinogenicity of Cr(VI) is well
established for inhalation exposures (TCEO. 2014: IPCS. 2013: NIOSH. 2013: IARC. 2012: CalEPA.
2011: NTP. 2011: OSHA. 20061. As stated in the 2014 preliminary packages fU.S. EPA. 2014b. c)
and the Systematic Review Protocol (Appendix A), the review of cancer by the inhalation route
focused on data that may improve the quantitative exposure-response analysis conducted in EPA's
1998 IRIS assessment. An overview of the literature screening for exposure-response data is
contained in Section 4.4.1.
The IUR was based on an occupational cohort by Gibb et al., (2020; 2015: 2000b) of
chromate production workers at a facility in Baltimore, MD. Details of the cohort are contained in
Section 4.4.
Because a mutagenic MOA for Cr(VI) carcinogenicity is "sufficiently supported in
(laboratory) animals" and "relevant to humans," and as there are no chemical-specific data to
evaluate the differences between adults and children, increased early-life susceptibility should be
assumed. If there is early-life exposure, age-dependent adjustment factors (ADAFs) should be
applied, as appropriate, in accordance with the EPA's Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA. 2005b).
The total lifetime IUR for Cr(VI) is 2 x 10"2 (per ng Cr(VI)/m3). Partial unit risks for
different age groups are provided in Section 4.4.4. Table ES-6 summarizes the derivation of the
IUR.
Table ES-6. Summary of inhalation unit risk (IUR) derivation
Critical effect
Basis
IUR
(Hg CrlVIJ/m3)"1
Study exposure
description
Confidence
Cancer
Lung cancer (Gibb et al.,
2020)
2 x 10"2
Occupational
cohort
High
ES.6 SUSCEPTIBLE POPULATIONS AND LIFE STAGES
Susceptible populations and life stages refers to groups of people who may be at increased
risk for negative health consequences following chemical exposures due to factors such as life stage,
genetics, race/ethnicity, sex, health status and disease, lifestyle factors, and other co-exposures.
Populations susceptible to increased risks for negative health consequences of Cr(VI) exposure
include:
This document is a draft for review purposes only and does not constitute Agency policy.
xxiii DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
• Individuals with preexisting health effects that overlap with those caused by Cr(VI)
exposure may be at increased risk. Health conditions that may be exacerbated by Cr(VI)
exposure include gastrointestinal diseases, liver diseases, respiratory diseases, and anemia.
• Individuals with chronically high stomach pH are expected to detoxify Cr(VI) less
effectively, leading to increased uptake of Cr(VI) in the gastrointestinal tract following oral
exposure. High stomach pH can be caused by a number of factors, such as low gastric acid
(hypochlorhydria), usage of medications to treat gastroesophageal reflux disease (GERD),
and population variability.
• Individuals with genetic polymorphisms conveying deficiencies in DNA repair capacity may
have increased susceptibility to Cr(VI)-induced cancer.
• Carriers of a mutated cystic fibrosis transmembrane conductance regulator (CFTR) allele
may be at higher risk of Cr(VI)-induced cancers of the gastrointestinal tract Suppression of
the CFTR gene was shown to enhance intestinal tumorigenesis in animal models. CFTR was
shown to be inactivated in mice exposed to Cr(VI). Thus, individuals with an impaired CFTR
due to genetics may suffer an even further reduction in CFTR expression levels following
oral exposure to Cr(VI).
Life stages susceptible to increased risks for negative health consequences of Cr(VI) exposure
include:
• The developmental life stage (in utero) is considered susceptible because Cr(VI) was
determined to likely cause developmental toxicity in humans.
• Neonates, infants, and young toddlers less than 30 months old, which exhibit elevated
stomach pH and therefore cannot effectively detoxify Cr(VI).
• Elderly populations (aged 65 and older) may be at higher risk because they exhibit some
preexisting health conditions associated with aging that may be exacerbated by oral or
inhalation exposure to Cr(VI). This includes conditions that cause elevated stomach pH.
ES.7 ORAL ABSORPTION UNCERTAINTIES AND ASSUMPTIONS APPLIED IN HAZARD
IDENTIFICATION AND MODE-OF-ACTION ANALYSES
Even under controlled rodent pharmacokinetic studies, assessing the oral absorption and
whole-body distribution of orally administered Cr(VI) at low doses involves uncertainty. Only the
total chromium concentration, which includes the trivalent and hexavalent oxidation states, can be
reliably measured in tissues in vivo, and most total chromium is likely to be Cr (III). Total chromium
measured in tissues of animals orally exposed to Cr(VI) results from:
• Rapid cellular uptake of administered Cr(VI) that was absorbed into the body as Cr(VI), and
subsequently reduced to Cr(III) within that tissue.
• Slow cellular uptake of Cr(III) that was absorbed into the body as Cr(III), formed from
administered Cr(VI) that reduced to Cr(III) extracellularly and outside of systemic
circulation (e.g., gastric juices).
This document is a draft for review purposes only and does not constitute Agency policy.
xxiv DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
• Slow cellular uptake of Cr(III) that was absorbed into the body as administered Cr(VI) and
reduced by other components within systemic circulation (e.g., plasma, liver, red blood
cells). For example, plasma can reduce Cr(VI) extracellularly, and the resulting Cr(III)
absorbed into other tissues. RBCs can reduce Cr(VI) intracellularly, and the resulting Cr(III)
can be released to systemic circulation (to be absorbed by other tissues) after RBCs are
broken down.
• Background uptake and distribution of dietary and drinking water chromium (Cr(III)
and/or Cr(VI)) not administered or controlled in the bioassay.
Additional details are provided in Section 3.1 (Pharmacokinetics) and Appendix C.l.
Elevated chromium concentrations in red blood cells (RBCs) is a strong indicator that Cr(VI) was
absorbed in the GI tract unreduced and was not subsequently reduced by the liver during first-pass
metabolism. Uptake and reduction of Cr(VI) by RBCs is rapid, and the resulting Cr(III) in red blood
cells is bound to hemoglobin and/or diffuses out of the RBC slowly. Therefore, elevated RBC
chromium persists longer relative to plasma chromium levels following systemic Cr(VI) absorption.
Based on analyses of the RBC:plasma ratios of exposed and unexposed rodents from the NTP (2008,
20071 studies (see Appendix C.l.2), general assumptions were made when interpreting animal
studies for hazard identification and MOA:
• At oral ad libitum doses below 1 mg/kg-d, Cr(VI) is absorbed by the GI tract, but most Cr(VI)
absorbed by the GI tract is reduced to Cr(III) by the liver (and to a lesser extent, plasma and
RBCs in the portal vein). At these low doses the GI tract and liver are exposed to Cr(VI), but
exposure to other systems may be low and highly variable. There is high uncertainty as to
whether other systemic tissues receive consistent exposure to Cr(VI) at these doses across
all the studies. Therefore, inconsistent pharmacokinetic and toxicological results among
studies for doses below 1 mg/kg-d are to be expected.
• At oral ad libitum doses greater than or equal to 1 mg/kg-d, Cr(VI) is absorbed by the GI
tract, exceeds the reducing capacity of the liver, and is widely distributed to systemic tissues
(e.g., kidney, lung, brain). Exposure to systemic tissues may still be highly variable, and
there may be some inconsistencies in dose-response between studies.
• For oral gavage doses at any level, Cr(VI) is widely distributed to systemic tissues, and
results in significantly higher internal doses than dietary and drinking water exposure. This
is because the gavage route greatly condenses the timescale of an exposure, surpassing
gastric reduction capacity [ad libitum exposures are distributed over a 24-hour period,
whereas gavage occurs over a very short period).
• Injection studies (intravenous or intraperitoneal) will expose systemic tissues to
significantly greater levels of Cr(VI) than oral gavage studies because there is not a first-
pass effect (reduction of Cr(VI) in the stomach and liver). Following injection, there will
also be (temporarily) more Cr(VI) available in the plasma prior to uptake to RBCs.
This document is a draft for review purposes only and does not constitute Agency policy.
xxv DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofHexavalent Chromium
> 1 mg/kg-d
G.I. + liver + v
exposure
< 1 mg/kg-d
G.I. + liver ex
Kidneys, spleen,
bone, bone marrow,
lung, heart, muscle,
fat, reproductive
organs
Gl tract epithelium
Figure ES-1. General assumptions regarding absorption and distribution of
Cr(VI) ingested by rodents during ad libitum drinking water or dietary
bioassays. At doses <1 mg/kg-d, it is assumed that Cr(VI) is absorbed by the small
intestine, and most of the absorbed Cr(VI) is reduced by the liver. At doses >1
mg/kg-d, it is assumed that systemic absorption and distribution of Cr(VI)
throughout the whole body will occur.
Despite uncertainties below 1 mg/kg-d, these assumptions were adequate for interpreting
the current Cr(VI) database because most studies were conducted using doses greater than 1
mg/kg-d. The 1 mg/kg-d dose level was not used as a cutoff for the inclusion of data or to make
inferences about low-dose extrapolation, but instead was used to generally evaluate the
uncertainties of results. For studies in which the daily oral ad libitum dose was much greater than 1
mg/kg-d, there is higher certainty that Cr(VI) reaches target tissues. For studies in which the daily
oral ad libitum doses were lower than 1 mg/kg-d, there is added uncertainty when analyzing data
outside of the GI or liver, because it cannot be assumed that Cr(VI) reaches other target systemic
tissues at high enough doses that can induce observable effects. In general, it can be assumed that
ingested Cr(VI), even at low doses, will expose at least the surface GI epithelial cells if not the liver.
For chronic exposure collection periods of the NTP (20081 distribution study (collection days 182
and 371, with 2-day washout period), liver chromium concentrations were significantly elevated at
all dose groups (including <1 mg/kg-d) in rats and mice.
This document is a draft for review purposes only and does not constitute Agency policy.
xxvi
DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Toxicological Review ofHexavalent Chromium
1. INTRODUCTION
1.1. OVERVIEW
This Toxicological Review critically evaluates the publicly available studies on Cr(VI) in
order to identify its adverse human health effects and to characterize exposure-response
relationships. This assessment was prepared under the auspices of the U.S. Environmental
Protection Agency's (EPA's) Integrated Risk Information System (IRIS) Program. IRIS assessments
are not regulations but provide critical scientific support for human health risk assessments and
resulting decisions made by EPA, state and local health agencies, other federal agencies, and
international health organizations to protect human health.
This assessment updates a previous IRIS assessment of Cr(VI) (posted in 1998) that
included an oral reference dose (RfD) and inhalation reference concentration (RfC) for effects other
than cancer, a determination of carcinogenic potential, and inhalation unit risk (IUR) for
carcinogenic effects.
As part of the initial steps in assessment development, the IRIS Program undertook scoping
and initial problem-formulation activities. During scoping activities, the IRIS Program consulted
with EPA program and regional offices to identify the nature of the hazard characterization needed,
the most important exposure pathways, and the level of detail required to inform Agency decisions.
A broad, preliminary literature survey was conducted to assist in identifying the extent of the
evidence and health effects that have been studied for Cr(VI). The IRIS Program also undertook
problem-formulation activities to frame the scientific questions that are a focus of this assessment
A summary of the IRIS Program's scoping and problem-formulation conclusions are contained in
the 2014 preliminary packages (U.S. EPA. 2014b. c). The preliminary packages were followed by
development of a Systematic Review Protocol (Appendix A), which presents detailed methods for
conducting the full systematic review and dose-response analysis. As discussed in the preliminary
materials and protocol, the IRIS assessment includes evaluations of the evidence relevant to all
cancer outcomes and noncancer effects for the following potential target systems: respiratory,
gastrointestinal (GI) tract, hepatic, hematologic, immunological, reproductive, and developmental.
For cancer and nasal irritation via the inhalation route, the systematic review focuses on data that
may improve the quantitative dose-response analysis, conducted in EPA's 1998 IRIS assessment
Appendices for additional systematic review methods and results, pharmacokinetics, dose-
response modeling, and public comments are provided as Supplemental Information to this
assessment (see Appendices A to F).
This document is a draft for review purposes only and does not constitute Agency policy.
1-1 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
1.1.1. Background
Elemental chromium is a Group 6 transition metal (atomic number 24 and atomic weight
52) on the periodic table, existing in nature in the form of various oxide minerals (Anger etal..
20051. It is present in the Earth's crust and has oxidation states ranging from -2 to +6, with the +3
(trivalent) and +6 (hexavalent) states being the most common (Losi etal.. 19941. Chromium in the
environment can originate from both natural and anthropogenic sources (discussed in detail in
Section 1.1.3) (Tohnson etal.. 2006: USGS. 1995: Calder. 1988: Pacvna andNriagu. 19881. Cr(VI)
compounds are used for corrosion inhibition, pigment manufacturing (including textile dyeing,
printing inks, and colored glass and plastic), and metal finishing (chrome plating/electroplating)
(NIOSH. 2013: NTP. 20111. Cr(VI) has been used in wood preservatives [as chromated copper
arsenate (CCA) in pressure treated wood; (ATSDR. 2012: Barnhart. 19971]: however, this use began
to decline in 2003 due to a voluntary phaseout of all residential uses of CCA pressure treated wood
fBedinger. 2015: NTP. 20111. Other uses for Cr(VI) that have been discontinued in the United
States include leather tanning and corrosion inhibition within cooling systems (NIOSH. 2013: NTP.
20111. Cr(VI) is also a byproduct of processes in the iron and steel industries (Shaw
Environmental. 20061.
1.1.2. Chemical Properties
A summary of the Cr(VI) compounds assessed in the human, animal, and mechanistic
studies considered pertinent to this assessment are contained in Table 1-1. This table is not an
exhaustive list of all Cr(VI) species that are relevant to human exposure but reflects those with data
to inform a human health assessment. Compounds of chromium complexed to other metals that
could potentially confound the results (such as lead chromate, barium chromate, zinc chromate,
copper dichromate, strontium chromate) were not included. A majority of the Cr(VI) compounds
evaluated by the human, animal, and mechanistic studies relevant to this assessment are known to
be highly water soluble. Calcium chromate, a form with low water solubility, was used in some
animal bioassays and pharmacokinetics studies and was therefore considered. Inhalation
pharmacokinetics differ between soluble and insoluble forms of Cr(VI) fOSHA. 20061 (see Section
3.1). This assessment will not make separate determinations of toxicity or carcinogenicity of
soluble vs. insoluble Cr(VI) compounds because the aim is to evaluate the toxicity and
carcinogenicity of Cr(VI) in all forms. Where applicable, issues related to solubility and particle size
that may impact study or data interpretations are discussed during study evaluation, hazard
identification, and dose-response.
Cr(VI) can exist as chromate (Cr042 ), hydrochromate (HCr042~) and dichromate (Cr2072 )
anions, whose concentrations at equilibrium depend on the metal concentration in the solution and
pH fBrito etal.. 19971. At physiological conditions (pH 7.4) and micromolar Cr(VI) concentrations,
the major form of Cr(VI) is chromate and the minor form is hydrochromate, with the latter
becoming a dominant form at pH<6 (Cieslak-Golonka. 19961. These pH-relationships between
This document is a draft for review purposes only and does not constitute Agency policy.
1-2 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 Cr(VI) species were incorporated into the gastric reduction model used in this assessment
2 fSchlosser and Sasso. 20141. Because multiple Cr(VI) compounds are discussed in this assessment,
3 all exposure levels were converted to Cr(VI) equivalents (see Protocol Section 8.2, Appendix A)2.
4 Even though the physical properties differ between compounds, they are all ionized to Cr(VI) in the
5 body and are considered to exert the same pharmacological and toxicological effects (U.S. EPA.
6 20081.
2In many studies, the administered compound is stated as "sodium dichromate" [NazCrzOy] when the
compound is administered in aqueous solution with mass units based on sodium dichromate dihydrate
(Na2Cr207 2H2O]. Unless otherwise noted, the conversion factor for sodium dichromate dihydrate (0.349]
was used to convert parent compound concentrations and doses to Cr(VI] units for studies labeled as either
sodium dichromate or sodium dichromate dihydrate. Due to variations in reporting, it may be unclear
whether the mass per unit volume of the formulation was based on NazCnOy 2H2O or NazCnOy (which would
yield a conversion factor of 0.397], In situations where the formulation was prepared based on units of
Na2Cr207 mass, doses and concentrations listed in this assessment would underestimate the dose by 12%.
This document is a draft for review purposes only and does not constitute Agency policy.
1-3 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 1-1. Chemical identity and physicochemical properties of Cr(VI)
Name
Calcium chromate
Sodium chromate
Sodium dichromate
Sodium dichromate,
dihydrate
CASRN
13765-19-0
7775-11-3
10588-01-9
7789-12-0
Synonyms
Calcium chromate(VI);
calcium chrome
yellow; calcium
monochromate;
gelbin; yellow
ultramarine; chromic
acid, calcium salt
Sodium chromate(VI);
chromium disodium
oxide; disodium
chromate;
rachromate; chromic
acid, disodium salt;
chromate of soda
Sodium
dichromate(VI);
sodium bichromate;
dichromic acid,
disodium salt;
bichromate of soda
Dichromic acid, disodium
salt, dihydrate
Structure
Ca+2
O O
0" °0
2-
2Na+
V
o o
• •
2-
2Na+
i i
Q>'°
&
°^o
2-
2Na+rn OO nl2"
-df ;bf .2H20
6' "° t>
Molecular
weight
156.07
161.972
261.965
297.995
Molecular
formula
CaCrC>4
Na2CrC>4
Na2Cr2C>7
Na2Cr207*2H20
Conversion
factor3
0.333
0.321
0.397
0.349
Melting point
1020°C (Anger et al.,
2005); decomposition
794°C (Lide, 2008)
357°C (Lide, 2008)
85°C (Lide, 2008);
decomposition
Density
3.12 g/cm3 (Anger et
al., 2005)
2.72 g/cm3 (Lide,
2008)
2.52 g/cm3 (Anger et
al., 2005)
2.35 g/cm3 (Lide, 2008)
Water
solubility
4.5 g/100 g H20
(4.3 wt%) at 0°C
(Anger et al., 2005)
87.6 g/100 g H20 at
25°C (Lide, 2008)
187 g/100 g H20 at
25°C (Lide, 2008)
272.9 g/100 g HzO (73.18
wt%) at 20°C (Anger et al.,
2005)
Stability/
reactivity
Decomposes at
1,000°C (Lide, 2008);
oxidizing agent (Lewis
and Hawley, 2007)
Hygroscopic (Anger et
al., 2005)
Strongly hygroscopic;
decomposes above
400°C (Lide, 2008);
strong oxidizing agent
(Anger et al., 2005)
Very hygroscopic,
deliquesces in air; strong
oxidizing agent in acid
solution (Lide, 2008; Anger
et al., 2005)
Synonyms, structures, and molecular formulas and weights were obtained from ChemID Plus
(https://chem.nlm.nih.eov/chemidplus). unless otherwise noted.
aMass conversion factor from parent compound to Cr(VI) units.
This document is a draft for review purposes only and does not constitute Agency policy.
1-4 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 1-1. Chemical identity and physicochemical properties of Cr(VI)
compounds (continued)
Name
Potassium chromate
Potassium dichromate
Chromium trioxideb
Chromic acidbc
CASRN
7789-00-6
7778-50-9
1333-82-0
7738-94-5 (H2Cr04);
13530-68-2 (H2Cr207)
Synonyms
Potassium
chromate(VI);
bipotassium chromate;
dipotassium chromate;
chromate of potash;
tarapacaite;
chromic acid,
dipotassium salt
Potassium dichromate(VI);
bichromate of potash;
potassium bichromate;
dipotassium bichromate;
dipotassium dichromate;
dipotassium dichromium
heptaoxide; lopezite;
dichromic acid
dipotassium salt
Chromium(VI) oxide;
hexavalent
chromium oxide;
chromic trioxide;
chromic anhydride
Chromic(VI) acid;
chromium hydroxide
oxide; dichromic acid
(H2Cr207)
Structure
CM
[ ° °1
scTf
&
2-
+
CM
O J? °Nv O
vCr Cr
df "b
2-
P
O-Cr
O
O
HO-d'r-OH
ti
Molecular
weight
194.188
294.181
99.993
118.008 (H2Cr04)
218.001 (H2Cr207)
Molecular
formula
K2CrO<
K2Cr207
CrOs
H2CrC>4; H2Cr207
Conversion
factor
0.268
0.353
0.520
0.441 (H2Cr04)
0.477 (H2Cr207)
Melting
point
974°C (Lide, 2008)
398°C (Lide, 2008)
197°C (Lide, 2008)
Not applicable
Density
2.73 g/cm3 (Lide, 2008)
2.68 g/cm3 (Lide, 2008)
2.7 g/cm3 (Lide,
2008)
Not applicable
Water
solubility
65.0 g/100 g H20 at
25°C (Lide, 2008)
15.1 g/100 g H20 at 25°C
(Lide, 2008)
169 g/100 g H20 at
25°C (Lide, 2008)
Not applicable
Stability/
reactivity
Nonhvgroscopic (Anger
etal.,2005). Strong
Nonhygroscopic;
decomposes at 500°C
Deliquescent;
decomposition
Strong oxidizing agent
(Anger et al., 2005)
oxidizing agent, may
explode in contact with
organic materials
(Lewis and Hawlev,
2007)
(Lide, 2008; Anger et al.,
2005)
begins above 198°C
(Anger et al., 2005);
strong oxidizing
agent (O'Neil et al..
2006)
bChromic acid is formed in aqueous solution when chromium(VI) oxide is dissolved in water; it cannot be isolated as a pure
compound out of solution (Anger et al., 2005; Page and Loar, 2004). The term chromic acid is sometimes used to
reference chromium(VI) oxide; however, it should be noted that there is a structural difference between the anhydrous
substance chromium(VI) oxide and the aqueous chromic acid that forms when the oxide is dissolved in water.
cChromic acid exists in solution as both H2Cr04 and H2Cr207(Anger et al., 2005; Page and Loar, 2004; Cotton et al., 1999).
H2Cr04 is the main species in basic solutions (pH > 6) while H2Cr207 is the main species in strongly acidic solutions (pH < 1)
(Anger et al., 2005; Page and Loar, 2004; Cotton et al., 1999). Both species are present in equilibrium in solutions that
have a pH value between 2 and 6 (Anger et al., 2005; Page and Loar, 2004; Cotton et al., 1999).
This document is a draft for review purposes only and does not constitute Agency policy.
1-5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
1.1.3. Sources, Production, and Use
1.1.3.1. Soil
The EPA Toxics Release Inventory (TRI) estimates approximately 53 million pounds of
chromium and chromium compounds were released to the environment via land releases (such as
landfills, land treatment, and surface impoundments, excluding underground injections) fU.S. EPA.
20181. Sources of chromium releases into soil include the disposal of commercial products that
contain chromium, coal fly ash and bottom fly ash from electric utilities and other industries, solid
wastes from metal manufacturing and chrome-plating facilities, chromate production waste,
agricultural and food wastes, leather tannery waste, and cooling tower water containing rust
inhibitors (Oregon DEO. 2014: ATSDR. 2012: U.S. EPA. 2011b: Pellerin and Booker. 2000: Burke et
al.. 1991: Nriagu and Pacvna. 19881. Air deposition to soil from combustion processes also occurs.
Cr(III) in soil may be present predominantly as chromium hydroxide (Cr(OH)3) or
chromium oxide ((>203) (Apte etal.. 2006: Kim and Dixon. 20021. These Cr(III) forms have low
solubility and reactivity. Cr(VI) may exist in soil as chromate (Cr04~2), chromic acid (HCr04~),
dichromate (C^Oy-2), and chromate salts (BaCr04, CaCr04, PbCr04, ZnCr04) (ATSDR. 2012: Apte et
al.. 2006: Kim and Dixon. 20021. Conversion of Cr(VI) to Cr(III) may occur in the environment
under reducing conditions (by ferrous iron, sulfides, and organic matter), while conversion of
Cr(III) to Cr(VI) may occur under oxidizing conditions (by manganese oxide minerals) (Hausladen
etal.. 2018: 2017: McClainetal.. 2017: Tardine etal.. 2011: Cummings etal.. 2007: Oze etal.. 2007:
2004: Kim and Dixon. 2002: Fendorf et al.. 2000: 19951. Fire-induced oxidation of Cr(III)-
substituted iron oxides in soils may also occur during wildfires fBurton etal.. 20191.
Most Cr(III) compounds are insoluble in water and immobile in soils (which helps inhibit
oxidation), while most Cr(VI) compounds are readily soluble in water and highly mobile and
bioavailable (Fendorf etal.. 2000: Fendorf. 19951. In addition to being stabilized by low solubility
and mobility, Cr(III) compounds are more thermodynamically stable than Cr(VI) compounds under
mostpH values encountered in the environment fFendorf. 19951. And therefore, the predominant
direction of chromium transformation in the environment is Cr(VI) -^Cr(III). See Figure 1-1.
This document is a draft for review purposes only and does not constitute Agency policy.
1-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicologiccil Review of Hexavalent Chromium
Direct industrial
inputs of Cr(VI)
Figure 1-1. Sources of Cr(VI) in soil and groundwater. Adapted from Hausladen
etal. f201BI
1.1.3.2. Water
The EPA Toxics Release Inventory (TRI) estimates approximately 66,000 pounds of
chromium and chromium compounds were released to the environment via surface water
discharges, and 315,000 pounds were discharged for wastewater treatment in 2019 (U.S. EPA.
20181. Data from USEPA's Discharge Monitoring Report (DMR) estimates that approximately
90,000 pounds of Cr(VI) was discharged in 2020 (U.S. EPA. 2014a). Most chromium released into
water from anthropogenic sources is ultimately deposited in sediment. Chromium in the aqueous
phase is mostly present as soluble Cr(VI) or as soluble Cr(III) complexes. Reduction of Cr(VI) to
Cr(III) can occur in the presence of reducing agents (e.g., organic matter, hydrogen sulfide, sulfur,
iron sulfide, ammonium, nitrate). The reduction half-life of Cr(VI) in water can be rapid (ranging
from instantaneously to a few days) when reducing agents are present under anaerobic conditions
but can extend from 4-140 days in water with soil and organic sediment fSaleh etal.. 19891.
Oxidation of Cr(III) to Cr(VI) can also occur within aquifers and water treatment systems (Chebeir
and Liu, 2016: U.S. EPA. 1986a). The ratio of Cr(VI) to Cr(III) has been measured to be higher in
groundwater than in surface water fFrev etal.. 2004). Oxidizing conditions within soil, as well as
the natural Cr(VI) content of soil and rocks, also affect Cr(VI) content of water (Vengosh etal..
2016). Above-average groundwater levels of Cr(VI) have been reported in several areas in the
Western US CI I S. EPA. 2014dl.
1.1.3.3. Air
Approximately 222,840 pounds of chromium and chromium compounds were released
from fugitive and point sources into air from reporting facilities in 2020 (U.S. EPA. 2021c). Based on
data from the 2017 EPA National Emissions Inventory (NEI), approximately 64,208 pounds of
Cr(VI), 1,392 pounds of chromic (VI) acid, 86 pounds of Chromium (VI) Trioxide, and 373,891
pounds of chromium (III) were released into the air nationwide fU.S. EPA. 2021bl. The NEI includes
This document is a draft for review purposes only and does not constitute Agency policy,
1-7 ' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
additional emissions sources not reported under TRI (i.e., mobile sources). Atmospheric chromium
particles resulting from industrial emissions have been reported to have a mass mean aerodynamic
diameter (MMAD) of less than 10 |im, were found to remain airborne for 7-10 days, and were
subject to long-range transport (Kimbrough etal.. 19991. Atmospheric particulate matter is
deposited on land and water via wet and dry deposition, and metals may deposit at a higher rate in
urban areas relative to rural and remote locations (Schroeder etal.. 19871. Transport of chromium
from water to the atmosphere is possible via transport in windblown seasalt sprays (Nriagu. 19891.
Major atmospheric chromium emissions from anthropogenic sources in the United States are
outlined in Table 1-2.
Table 1-2. Major anthropogenic sources of atmospheric chromium in the
United States [adapted from ATSDR (2012)]
Industrial processes and production
Cooling towers
Combustion of coal and oil
Ferrochromium production
Chromium chemical manufacturing
Chrome plating
Chrome ore refining
Refractory production
Cement production
Specialty/steel production
Sewage sludge incineration
Municipal refuse incineration
Utility industry cooling towers
Chemical manufacturing cooling towers
Petroleum refining cooling towers
Glass manufacturing cooling towers
Primary metal cooling towers
Comfort cooling towers
Textile manufacturing cooling towers
Tobacco cooling towers
Tire and rubber cooling towers
Data of annual Cr(VI) emissions in the US can be obtained from the EPA National Emissions Inventory (U.S. EPA,
2016a).
Depending on the emission source, different forms of Cr(VI) may be emitted (i.e., Cr(VI) acid
mists/dissolved aerosols, and Cr(VI) dusts). While information is limited regarding
non-occupational inhalation exposures to chromic acid mists for the general U.S. population,
residents of fence-line communities may be exposed to multiple forms of Cr(VI) (OAOPS. 20121.
Chrome-plating facilities and private residencies may exist in close proximity in mixed land use
communities (CARB. 2004: CalEPA. 20031. Chromium trioxide (CrO;j) is the acidic anhydride of
chromic acid (FhCrO^. Chromic acid in mists or vapors dehydrates to CrC>3 upon evaporation, and
some CrC>3 may convert to l-hCrCU in moist environments (including the respiratory tract).
1.1.4. Environmental Occurrence
The mean soil concentration of total chromium in the United States is approximately 36
mg/kg (Smith etal.. 20131. and the ratio of Cr(VI) to Cr(III) depends on several factors (such as soil
pH). Nationwide data for speciated chromium are unavailable, although some site-specific soil
concentrations of Cr(VI) have been reported. For example, soil Cr(VI) concentrations in Montana
were mostly below the limit of detection of 0.29 mg/kg (Hydrometrics. 20131. Cr(VI)
concentrations near industrial facilities in Portland, Oregon were typically below 1 mg/kg but were
This document is a draft for review purposes only and does not constitute Agency policy.
1-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofHexavalent Chromium
measured as high as 3 mg/kg fOregonDEO. 2016a. c). Bioaccumulation of Cr(VI) or Cr(III) from
soil to above ground plants, or biomagnification of chromium in terrestrial or aquatic food chains, is
not expected to occur fATSDR. 20121.
Public water system data from EPA's Third Unregulated Contaminant Monitoring Rule
(UCMR3)3, includes both groundwater and surface water sources (U.S. EPA. 2014dl. Mean Cr(VI)
concentrations in public water systems averaged approximately 0.48 ng/L for large systems (U.S.
EPA. 2014dl. There was wide variability by region (Figure 1-2), and a maximum concentration of
97.4 ng/L.
Ambient air concentrations of Cr(VI) in the United States typically range from 0.01 to 0.05
ng/m3 (U.S. EPA. 2016c)4. but have been measured at values above 1 ng/m3 for urban and
industrial areas (Oregon DEO. 2016b: Huang etal.. 20141. Historically, Cr(VI) concentrations
measured in ambient air downwind of industrial facilities emitting Cr(VI) (such as chrome platers)
have been found to be highly correlated with concentrations measured at the facilities fSCAOMD.
20161. Between May 2001-May 2002, residential air near chrome-plating facilities in San Diego, CA
were measured up to 22.0 ng/m3 Cr(VI) fCalEPA. 2004. 20031.
3Cr(VI) was among 30 contaminants selected for monitoring at public water systems (PWS) for the Third
Unregulated Contaminant Monitoring Rule (UCMR3) between 2013 and 2015. A PWS is a network of pipes
and conveyances constructed to provide water for human consumption (U.S. EPA. 2006a. b). Small systems,
serving 10,000 or fewer people, account for more than 97% of the total number of PWSs, while large systems,
serving more than 10,000 people, account for the remaining 3% (U.S. EPA. 2006a. b). A majority of the U.S.
population is served by large PWSs (nearly 90% fU.S. EPA. 2006a. b)), and all of them (approximately 4,200)
were tested under UCMR3. For small water systems, approximately 800 systems were randomly selected and
used as a representative sample (U.S. EPA. 2012c). Small water systems were omitted from analyses
presented in this section. Cr(VI) was selected for the UCMR3 cycle, and was not selected for monitoring for
the UCMR4 or UCMR5 cycles.
4See also: https://cfpub.epa.gov/roe/indicator.cfm?i=90#7. containing 2008-2014 data from 14 sites across
the United States.
This document is a draft for review purposes only and does not constitute Agency policy.
1-9 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
ID
_Q
Q.
Q.
ŁZ
O
-4—•
CD
-t—»
fZ
CD
O
ŁZ
O
>
O
co
CM
O
123456789 10
EPA Region
Figure 1-2. Drinking water chromium (VI) concentrations in the United States
by EPA region.5 Boxplots are based on the average values of samples of large
public water systems within the region, from EPA's Third Unregulated Contaminant
Monitoring Rule (UCMR3) fU.S. RPA.2014dl. Boxes represent interquartile ranges.
Whiskers are 1.5x the interquartile range away from the 25th/75th percentiles.
1.1.5. Potential for Human Exposure
1 1.1.5.1. General Population
2 General population exposures to Cr(VI) occur via inhalation of ambient air, ingestion of
3 water or food, and non-dietary ingestion of soil or dust. Most human exposure to total chromium
sRegion 1 - CT, ME, MA, NH, RI, and VT
Region 2 - NJ, NY, Puerto Rico, and the U.S. Virgin Islands
Region 3 - DE, DC, MD, PA, VA, WV and 7 federally recognized tribes
Region 4 - AL, FL, GA, KY, MS, NC, SC, and TN
Region 5 - IL, IN, MI, MN, OH, and WI
Region 6 - AR, LA, NM, OK, and TX
Region 7 - IA, KS, MO, and NE
Region 8 - CO, MT, ND, SD, UT, and WY
Region 9 - AZ, CA, HI, NV, American Samoa, Commonwealth of the Northern Mariana Islands, Federated States
of Micronesia, Guam, Marshall Islands, and Republic of Palau
Region 10 - AK, ID, OR, WA and 271 native tribes.
This document is a draft for review purposes only and does not constitute Agency policy.
1-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
(sum of Cr(VI) and Cr(III)) is from dietary intake of Cr(III) that is naturally present in foods
fWisconsin DHS. 20101. Cr(III) is generally understood to be essential to normal glucose, protein,
and fat metabolism and is thus an element with an Adequate Intake (Al)6 values flOM. 20111.
although no Recommended Daily Allowance (RDA) has been established due to insufficient
evidence to establish a level of Cr(III) that is necessary for human health (NIH. 2017: Vincent. 2017:
Vincent. 2013: Stearns. 20001. Dermal exposure may also occur during the use of consumer
products that contain chromium, such as some metals or leather treated with chromium-containing
compounds CATSDR. 2012: NTP. 20111.
Quantifying the non-dietary exposure to Cr(VI) via soil ingestion (hand-to-mouth contact
and pica behavior in children) is uncertain due to limited data on chromium speciation in soil. As
noted earlier, the Cr(VI)/Cr(III) concentration ratio in soil can vary due to factors such as soil pH
and mineral content, and no nationwide data on soil Cr(VI) currently exist EPA's Office of Pesticide
Programs (OPP), in its reregistration eligibility decision (RED) for chromated copper arsenate
(CCA) pesticides (U.S. EPA. 20081. determined that dietary, residential, or other non-occupational
exposures to Cr(VI) was not expected to occur from wood preservative uses of chromated
arsenicals.
Dietary exposure to Cr(VI) via food ingestion is uncertain due to limited data on speciation
in food. Typical total chromium (sum of Cr(VI) and Cr(III)) levels in most foods have been reported
to range from <10 to 1,300 ng/kg, with the highest concentrations being found in meat, fish, fruits,
and vegetables fWHO. 20031. Dietary total chromium intake in the general U.S. population has been
estimated to range from 0.293-0.867 ng/kg-day (ATSDR. 2012: Moschandreas etal.. 20021. It is
possible that a fraction of this intake is in the form of Cr(VI) f Hamilton etal.. 20181. Mathebula et
al. f20171 found that 33-73% of total chromium in bread may exist as Cr(VI) (at concentrations
between 19-64 |ig/kg), and that oxidation of Cr(III) to Cr(VI) can occur from toasting. That study
also detected Cr(VI) in breakfast cereals at concentrations between 41-470 |J.g/kg. Spares et al.
(20101 estimated that 12% of total chromium in bread was hexavalent However, nationwide data
for Cr(VI) content in food is limited. It is assumed that (total) chromium exposure to infants via
breastmilk is low (EFSA CONTAM Panel. 20141: however, no studies investigating levels of
speciated Cr(VI) in human milk were identified.
According to data collected between 2013 and 2015 under EPA's Third Unregulated
Contaminant Monitoring Rule (UCMR3), Cr(VI) has been reported above the minimum reporting
limit (0.03 ng/L) in approximately 90% of public water systems in the United States (U.S. EPA.
2014dl. More detailed concentration data for Cr(VI) in large U.S. water systems are provided in
Section 1.1.4 (above) and in Appendix C.4. Based on this information, drinking water is expected to
be a significant source of exposure for the general population.
The general population may be exposed to Cr(VI) in air but will likely receive a lower
inhaled dose when compared to the oral ingestion pathway. A 70 kg individual drinking 2L/day
6Adequate intakes of chromium for adult males and females are 35 |Ag/day and 25 |Ag/day, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
1-11 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
water containing 0.5 |ig/L Cr(VI) will ingest a dose of 1.4 x 10"5 mg/kg-d Cr(VI). A 70 kg individual
with a respiratory rate of 20 m3/day inhaling air containing 4 x 10"5 |J.g/m3 Cr(VI) will inhale Cr(VI)
at a body weight-normalized rate of 1.1 x 10"8 mg/kg-d. Both air and water concentrations may
vary from the approximate mean values by a factor of 100 in extreme cases (see Appendix C.4).
Only in extreme cases is it possible for the inhaled dose to be comparable to the ingested dose for
people living in an area with low Cr(VI) in water and high Cr(VI) in air.
Inhalation of Cr(VI) in water droplets during showering can also occur. Since Cr(VI) cannot
volatilize, and because Cr(VI) compounds are typically water soluble, the metal will exist only in
water droplets and aerosols. An analysis of this exposure pathway was performed by California
EPA, and determined that a 70-kg adult breathing 20 m3 of air per day, taking a 10-minute shower
would inhale 27 mg of liquid water per shower (3.86 x 10"7 L/kg-d) fCalEPA. 20111. Assuming
water contains 0.5 |ig/L Cr(VI) yields an inhaled dose of 1.9 x 10"10 mg/kg-d, which is five orders of
magnitude lower than the dose resulting from 2 L/day water ingestion at the same Cr(VI)
concentration (1.4 x 10"5 mg/kg-d).
Humans may be exposed via inhalation and incidental ingestion of house dust A study of
house dust in areas with no known soil contamination by Cr(VI) in New Jersey measured a mean
Cr(VI) surface loading of 10 |J.g/m2 (maximum of 169.3 ng/m2), and mean Cr(VI) concentration of
4.6 ng/g (maximum of 56.6 ng/g) (Stern etal.. 20101. Nationwide data of Cr(VI) in house dust are
unavailable.
1.1.5.2. Occupational Exposure
Occupational exposures to Cr(VI) occur primarily via inhalation or dermal contact fNIOSH.
20131 and typically exceed non-occupational exposures (NTP. 20111. Workers can potentially
inhale Cr(VI) during its processing or manufacture and when working with mixtures containing the
chemical or chemical precursors. Dermal exposures may potentially result from the splashing or
spilling of chromium-containing materials that contact the skin or from contact with construction
materials containing Portland cement (due to a Cr(VI) impurity) (NIOSH. 20131. Portal-of-entry
sites may be exposed via hand-to-mouth contact and hand-to-nose contact fOSHA. 20061. and the
extent of these transfers depends on the industry, exposure matrix, and workplace hygiene
practices (Cohen etal.. 19741. Industries that may have workers who are in contact with Cr(VI)-
containing materials include stainless-steel welding, painting, electroplating, steel mill, iron and
steel foundries, wood preserving, and occupations that produce paints, coatings, inks, plastic
colorants, chromium catalyst, and other chemicals (such as chromium dioxide and chromium
sulfate) (NIOSH. 20131. Other industries with limited potential exposures to Cr(VI) compounds
include textile dyeing, glass production, printing, leather tanning, brick production, woodworking,
solid waste incineration, oil and gas well drilling, construction and Portland cement production
fNIOSH. 2013: NTP. 20111. EPA's OPP, in its RED for CCA pesticides fU.S. EPA. 20081. determined
that inhalation exposure to chromium may occur from these pesticide components in occupational
settings. Because exposure to Cr(VI) outside of the workplace is possible via contaminated
This document is a draft for review purposes only and does not constitute Agency policy.
1-12 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofHexavalent Chromium
clothing, OSHA f20061 implemented workplace rules to ensure that clothing contaminated with
Cr(VI) is not carried to employees' cars and homes (which would expose both the workers and
other individuals). Table 1-3 provides a list of industries that are potential sources of chromium
exposure.
Table 1-3. Industries and occupations that may be sources of chromium
exposure
Group 1: Industry sectors where majority of occupational
exposures occur to
hexavalent chromium
Group 2: Industry sectors with limited potential for
occupational exposure to
hexavalent chromium
Electroplating
Welding
Painting
Producers of Chromates and Related
Chemicals from Chromite Ore
Chromate Pigment Production
Chromated Copper Arsenate Producers
Chromium Catalyst Production
Paint and Coatings Production
Printing Ink Producers
Plastic Colorant Producers and Users
Plating Mixture Production
Wood Preserving
Chromium Metal Production
Steel Mills
Iron and Steel Foundries
Chromium Dioxide Producers
Chromium Dye Producers
Chromium Sulfate Producers
Chemical Distributors
Textile Dyeing
Producers of Colored Glass
Printing
Leather Tanning
Chromium Catalyst Users
Producers of Refractory Brick
Woodworking
Solid Waste Incineration
Oil and Gas Well Drilling
Portland Cement Producers
Non-Ferrous Superalloy Producers and Uses
Construction
Producers of Pre-Case Concrete Products
Source: Analysis performed by OSHA (Shaw Environmental, 2006).
1.2. SUMMARY OF ASSESSMENT METHODS
The systematic review and dose-response methods used to conduct this assessment are
summarized in the remainder of this section. A detailed description of these methods is provided in
the preliminary materials released in 2014 fU.S. EPA. 2014b. c) and in the Systematic Review
Protocol for Cr(VI), released in 2019 fU.S. EPA. 20191. which has been updated to reflect
refinements made to the protocol during the assessment process. A link to the updated protocol
can be found in the Supplementary Materials released with this Toxicological Review in Appendix
A.
1.2.1. Literature Search and Screening
Literature search strategies were developed using key terms and words related to the PECO
criteria and potentially relevant supplemental material. Relevant subject headings and text-words
were crafted into a search strategy that was designed to maximize the sensitivity and specificity of
the search results. The search strategy was run, and the results were assessed to ensure that all
This document is a draft for review purposes only and does not constitute Agency policy.
1-13 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 previously identified relevant primary studies were retrieved in the search. Because each database
2 has its own search architecture, the resulting search strategy was tailored to account for the unique
3 search functionality of each database.
4 The following databases were searched:
5 • PubMed (National Library of Medicine)
6 • Web of Science (Thomson Reuters)
7 • Toxline (National Library of Medicine)7
8 Searches were not restricted by publication date, and no language restrictions were applied.
9 Web of Science results were limited using the research areas filter. All Web of Science research
10 areas identified in the search results were prioritized by a technical advisor as high priority
11 (e.g., toxicology), low priority (e.g., chemistry), and not relevant (e.g., forestry). Literature searches
12 were conducted in bibliographic databases as described in Appendix B and uploaded to EPA's
13 Health and Environmental Research Online (HERO) database.8
14 Additional relevant literature not found through database searching was sought by:
15 • Manually searching citations from review articles and studies considered to meet PECO
16 criteria after screening ("included" studies).
17 • Searches of gray literature, including primary studies that are not indexed in databases of
18 peer-reviewed literature (e.g., technical reports from government agencies or scientific
19 research groups; unpublished laboratory studies conducted by industry; working papers
20 from research groups or committees; and white papers), or other nontypical searches. Gray
21 literature is typically identified by searching the EPA Chemical Dashboard
22 fhttps: //comptox.epa.gov/dashboardl during problem formulation, by engaging with
23 technical experts, and during solicitation of Agency, interagency, and public comment at
24 multiple steps in the IRIS process.
25 • "Backward" searches (to identify articles cited by included studies, reviews, or prior
26 assessments by other agencies).
27 The results returned (i.e., the number of "hits" from each electronic database or other
28 literature source), including the results of any literature search updates, are documented in the
29 literature flow diagrams, which also reflect the literature screening decisions (see Section 2.1).
30 The IRIS Program takes extra steps to ensure identification of pertinent studies by
31 (1) encouraging the scientific community and the public to identify additional studies and ongoing
32 research; (2) searching for publicly available data submitted under the Toxic Substances Control
33 Act and the Federal Insecticide, Fungicide, and Rodenticide Act; and (3) considering late-breaking
7T0XLINE was phased out in December 2019 and integrated into other NLM resources.
8Health and Environmental Research Online: https: //hero.epa.gov/hero/.
This document is a draft for review purposes only and does not constitute Agency policy.
1-14 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
studies that would impact the credibility of the conclusions, even during the review process.
Studies identified after peer review begins will only be considered for inclusion if they meet the
PECO criteria and may fundamentally alter the assessment's conclusions.
1.2.2. Evaluation of Individual Studies
The detailed approaches used for the evaluation of epidemiologic and animal toxicology
studies used in the Cr(VI) assessment are provided in the protocol (Appendix A). The general
approach for evaluating health effect studies meeting PECO criteria is the same for epidemiology
and animal toxicology studies although the specifics of applying the approach differ; thus, they are
described in detail in protocol Sections 6.2 and 6.3, respectively, in Appendix A.
• The key concerns for the review of epidemiology and animal toxicology studies are
potential bias (factors that affect the magnitude or direction of an effect in either direction)
and insensitivity (factors that limit the ability of a study to detect a true effect; low
sensitivity is a bias towards the null when an effect exists). In terms of the process for
evaluating individual studies, two or more reviewers independently arrive at judgments
regarding the reliability of the study results (reflected as study confidence determinations;
see below) with regard to each outcome or outcome grouping of interest; thus, different
judgments are possible for different outcomes within the same study. The results of these
reviews are tracked within EPA's version of the Health Assessment Workplace
Collaboration (HAWC).
• To develop these judgments, each reviewer assigns a category of good, adequate, deficient
(or not reported, which generally carries the same functional interpretation as deficient), or
critically deficient (listed from best to worst methodological conduct; see Section 6.1 of the
protocol in Appendix A for definitions) to each evaluation domain representing the different
characteristics of the study methods that were evaluated based on the criteria outlined in
HAWC. Reviewers assigning categories to each domain are guided by core and prompting
questions as well as additional considerations specific to Cr(VI) or the outcome of interest.
Exposure-specific considerations in epidemiology studies are described in Section 6.2.
Briefly, air concentration measurements were preferred to biomarker measurements.
Studies in which human exposure was quantified by measurements of total chromium in
urine, blood, plasma, or erythrocytes were considered for determination of hazard only if
conducted in workers with known occupational exposure to Cr(VI).
Once all evaluation domains were evaluated, the identified strengths and limitations are
considered as a whole by the reviewers in order to reach a final study confidence classification:
• High confidence: No notable deficiencies or concerns were identified; the potential for bias
is unlikely or minimal, and the study used sensitive methodology.
• Medium confidence: Possible deficiencies or concerns were noted, but the limitations are
unlikely to be of a notable degree or to have a notable impact on the results.
• Low confidence: Deficiencies or concerns were noted, and the potential for bias or
inadequate sensitivity could have a significant impact on the study results or their
This document is a draft for review purposes only and does not constitute Agency policy.
1-15 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
interpretation. Low confidence results were given less weight compared to high or medium
confidence results during evidence synthesis and integration (see Sections 1.2.4 and 1.2.5).
• Uninformative: Serious flaw(s) were identified that make the study results unusable.
Uninformative studies were not considered further, except to highlight possible research
gaps.
Using the HAWC platform (and conflict resolution by an additional reviewer, as needed), the
reviewers reached a consensus judgment regarding each evaluation domain and overall
(confidence) determination. The specific limitations identified during study evaluation were
carried forward to help inform the synthesis (Section 1.2.4) within each body of evidence for a
given health effect along with other considerations. Additional details regarding study evaluation
are provided in Sections 6.1-6.5 of the protocol (Appendix A).
1.2.3. Data Extraction
The detailed data extraction approach is provided in Section 8 and Appendix B of the
protocol (Appendix A). Animal data extraction and content management were carried out using
HAWC, while data extracted from epidemiology studies were summarized in tabular format in the
assessment and appendices. Studies evaluated as being uninformative were not considered further
and study details are not summarized. In addition, study details and results for outcomes not
prioritized during PECO refinement (e.g., kidney and neurological) were not extracted or were only
partially extracted (Appendix A). The same was typically true for low confidence studies where a
number of medium and high confidence studies were available, unless the low confidence studies
included study designs lacking in the higher confidence studies (e.g., testing lower exposure levels,
or susceptible populations or life stages). The level of extraction for specific outcomes within a
study may differ (i.e., ranging from a narrative to full extraction of dose-response effect size
information). Data extraction was performed by one member of the evaluation team and checked
by at least one other member.
For animal data already extracted to evidence tables released in 2014 fU.S. EPA. 2014bl.
data extraction procedures depended on data type (e.g., dichotomous, continuous, or qualitative).
For human data already extracted to evidence tables released in 2014 (U.S. EPA. 2014c). data
extraction procedures depended on the study evaluation judgment and the study design. Large-
scale epidemiological datasets, which are typically stored in databases and under the custody of
scientific researchers or institutions, were not extracted or uploaded into HAWC. A detailed
discussion of the methods used for data extraction are provided in Section 8 of the protocol
(Appendix A). Extracted data are available in HAWC and are also summarized in tabular or
graphical form in the hazard identification and dose-response sections.
This document is a draft for review purposes only and does not constitute Agency policy.
1-16 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
1.2.4. Evidence Synthesis and Integration
For the purposes of this assessment, evidence synthesis and integration are considered
distinct but related processes (see Protocol Sections 9 and 10, Appendix A for full details). For each
assessed health effect, the evidence syntheses provide a summary discussion of each body of
evidence considered in the review that directly informs the integration across evidence to draw an
overall judgment for each health effect. The available human and animal evidence pertaining to the
potential health effects are synthesized separately, with each synthesis providing a summary
discussion of the available evidence that addresses considerations regarding causation that are
adapted from Hill (1965). Mechanistic evidence and other supplemental information is also
synthesized to address key science issues and/or to help inform key decisions regarding the human
and animal evidence.
The syntheses focus on describing aspects of the evidence that best inform causal
interpretations, including the exposure context examined in the sets of available studies. The
human and animal health effects evidence syntheses are based primarily on studies of high and
medium confidence. Low confidence studies may be used if few or no studies with higher
confidence are available to help evaluate consistency, or if the study designs of the low confidence
studies address notable uncertainties in the set of high or medium confidence studies on a given
health effect. If low confidence studies are used, then a careful examination of risk of bias and
sensitivity with potential impacts on the evidence synthesis conclusions is included in the narrative.
The synthesis of mechanistic evidence and other supplemental information informs the integration
of health effects evidence for both hazard identification (i.e., biological plausibility of the available
human or animal evidence; inferences regarding human relevance, or the identification of
susceptible populations and life stages across the human and animal evidence) and dose-response
evaluation.
For each assessed health effect, following the evidence syntheses, integrated judgments are
drawn across all lines of evidence. During evidence integration, a structured and documented
process was used, as follows:
• Building from the separate syntheses of the human and animal evidence, the strength of the
evidence from the available human and animal health effect studies was summarized in
parallel, but separately, using a structured evaluation of an adapted set of considerations
first introduced by Bradford Hill (Hill. 1965). These summaries incorporate the relevant
mechanistic evidence (or MOA understanding) that informs the biological plausibility and
coherence within the available human or animal health effect studies.
• The strength of the animal and human evidence was considered together in light of
inferences across evidence streams. Specifically, the inferences considered during this
integration include the human relevance of the animal and mechanistic evidence, coherence
across the separate bodies of evidence, and other important information (e.g., judgments
regarding susceptibility). Note that without evidence to the contrary, the human relevance
of animal findings is assumed.
This document is a draft for review purposes only and does not constitute Agency policy.
1-17 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofHexavalent Chromium
• A summary judgment is drawn as to whether the available evidence base for each potential
human health effect as a whole provides sufficient evidence to indicate that Cr(VI) exposure
has the potential to cause the health effect in humans; insufficient evidence to assess
whether Cr(VI) exposure has the potential to cause the health effect in humans; or, in rare
instances, sufficient evidence that a hazard is unlikely.
The decision points within the structured evidence integration process are summarized in
an evidence profile table for each assessed health effect
The primary focus of this assessment is on the following potential target systems:
respiratory, gastrointestinal (GI) tract, hepatic, hematologic, immunological, reproductive, and
developmental. It is acknowledged that there is evidence for other health effects not assessed here,
including renal and neurological toxicity, which can be induced by toxic metals in general
(Nordberg etal.. 2015). Kidney effects are known to occur following acute exposures to high doses
or concentrations of Cr(VI) fATSDR. 20121. but these effects are not observed following chronic,
low-dose exposure. Neurotoxicity associated with Cr(VI) exposure has recently been reviewed by
Wise etal. (20221: however, the evidence base is still relatively small and more research is needed
in this area. Many studies of chromium and neurotoxicity would not meet PECO criteria due to lack
of exposure information (e.g., studies of unspeciated chromium in organs and tissues of humans
would be excluded) or focus on non-PECO chromium compounds (e.g., lead chromate). In addition,
some endpoints would be difficult to dissociate from one another (e.g., impaired olfactory function
and nasal effects).
For cancer and nasal irritation via the inhalation route, the systematic review will focus on
data that may improve the quantitative dose-response analysis, conducted in EPA's 1998 IRIS
assessment, for these outcomes. Outlines of the major endpoints assessed within each health effect
domain are listed below in Table 1-4.
Table 1-4. Endpoint grouping categories
Relevant human health
effect category3
Endpoints included13
Notes
General toxicity
• Body weight (not maternal or pup weights, or
weights after developmental-only exposure)
• Mortality, survival, or LDsos
• Growth curve
• Clinical observations (non-behavioral)
• Clinical chemistry
endpoints are under
hepatic or hematologic
effects
• Maternal or pup
body-weight endpoints
are under developmental
effects
• Pathology (including
gross lesions) is organ
specific
This document is a draft for review purposes only and does not constitute Agency policy.
1-18 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Relevant human health
effect category3
Endpoints included13
Notes
Hepatic effects
• Liver weight and histopathology (e.g., chronic
inflammation, hyperplasia)
• Serum or tissue liver enzymes (e.g., clinical
chemistry measures such as ALT, ALP, and
AST)*
• Other liver tissue biochemical markers
(e.g., albumin; glycogen; glucose)*
• Liver-specific serum biochemistry
(e.g., albumin; albumin/globulin)*
• Liver tissue lipids: triglycerides, cholesterol
• Serum lipids
• Other liver tissue enzyme
activity (e.g., catalase) or
protein/DNA content are
considered under
mechanistic evidence for
hepatic effects
Hematologic effects
• Red blood cells*
• Blood hematocrit or hemoglobin*
• Cell volume*
• Blood platelets or reticulocytes*
• White blood cell count
and globulin are under
immune effects
• Serum liver markers are
under hepatic effects
Immune effects
• Thymus weight and histopathology
• Host resistance
• General immune assays (e.g., white blood cell
counts, immunological factors or cytokines in
blood, lymphocyte phenotyping or
proliferation)*
• Any measure in lymphoid tissues (weight;
histopathology; cell counts; etc.)
• Immune cell counts or immune-specific
cytokines in non-lymphoid tissues
• Other immune functional assays (e.g., natural
killer cell activity, mixed lymphocyte response,
phagocytosis or bacterial killing by monocytes)
• Immune responses in the respiratory system
• Red blood cells are under
hematologic effects
• Immune responses in the
respiratory tract (such as
phagocytosis, cytokine
signaling, inflammatory
responses) are also under
respiratory effects
• Endpoints related to
Cr(VI)-induced allergic
hypersensitivity were
considered under
mechanistic evidence for
immune effects
Male Reproductive effects
• Reproductive organ weight and histopathology
• Markers of sexual differentiation or
maturation (e.g., preputial separation)
• Mating parameters (e.g., success, mount
latency)
• Reproductive hormones*
• Birth parameters
(e.g., litter size;
resorptions,
implantations, viability)
are under developmental
effects
• If data indicate altered
birth parameters are
likely attributable to
This document is a draft for review purposes only and does not constitute Agency policy.
1-19 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Relevant human health
effect category3
Endpoints included13
Notes
• Sperm and semen parameters*
female fertility, these
data may be discussed
under female
reproductive effects
Female Reproductive
effects
• Reproductive organ weight and histopathology
• Markers of sexual differentiation or
maturation (e.g., vaginal opening or estrous
cycling)
• Birth parameters, if attributable to female
fertility
• Reproductive hormones*
• Birth parameters
(e.g., litter size;
resorptions,
implantations, viability)
are under developmental
effects
Developmental effects
• Dam health (e.g., weight gain, food
consumption)
• Pup viability/survival or other birth
parameters (e.g., number of pups per litter)
• Pup weight or growth (includes measures into
adulthood after developmental-only exposure)
• Developmental landmarks (eye opening, etc.,
but not including markers for other
organ/system-specific toxicities)
• Pregnancy outcomes (e.g., spontaneous
abortion, early pregnancy loss, pregnancy
complications, infant health, congenital
malformations/anomalies) [human only]
• Histopathology and
markers of development
specific to other systems
are organ/system-specific
(e.g., vaginal opening is
under female
reproductive effects;
offspring liver weight is
under hepatic effects)
Lower respiratory effects
Note: Systematic review of
evidence for nasal irritation
via the inhalation route will
focus on data for
quantitative dose-response
analysis.
• Lung weight and histopathology
• Biochemical markers of cell industry (e.g., total
protein, albumin, and lactate dehydrogenase
activity in bronchioalveolar lavage fluid)
• Cellular responses (e.g., number of
macrophages, neutrophils/granulocytes, and
lymphocytes)
• Pulmonary function (e.g., FVC, FEV1.0, DLCO)
[human only]
• Immune responses in the
respiratory tract (such as
phagocytosis, cytokine
signaling, inflammatory
responses) are also under
immune effects
Gastrointestinal tract
effects
• Histopathology (e.g., chronic inflammation,
hyperplasia, ulceration)
• Endpoints related to
precancerous legions are
also considered under
carcinogenicity
This document is a draft for review purposes only and does not constitute Agency policy.
1-20 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofHexavalent Chromium
Relevant human health
effect category3
Endpoints included13
Notes
Carcinogenicity
Note: Systematic review of
evidence for cancer via the
inhalation route will focus
on data for quantitative
dose-response analysis.
• Tumors
• Precancerous lesions (e.g., dysplasia)
ALT = alanine aminotransferase; AST = aspartate transaminase; DNA = deoxyribonucleic acid; LDso = median lethal
dose; FVC: forced vital capacity; FEV1.0: forced expiratory volume in first second; DLCO: the ratio of FEV1.0/FVC,
and diffusing capacity of lung for carbon monoxide.
aHealth effect-relevant endpoints observed after developmental exposure will be discussed primarily in the health
effect category indicated and then referenced in developmental effects.
bEndpoints refer to animal data unless otherwise noted. An asterisk (*) indicates endpoints that are also measured
in humans. Endpoints that are only measured in humans are noted by descriptive text. Some endpoints are
relevant to multiple health effects. These endpoints may be categorized under only a single health effect for
clarity. However, in the assessment, such outcome data may be discussed in each relevant health effect
synthesis, with cross-referencing to the synthesis containing most of the evidence. The evidence (for or against
an effect) will contribute to evidence integration decisions for all relevant health effects.
1.2.5. Dose-Response Analysis
Dose-response analysis to support derivation of toxicity values for Cr(VI) were performed
consistent with EPA guidelines and support documents, especially EPA's Benchmark Dose Technical
Guidance (U.S. EPA. 2012b). EPA's Review of the Reference Dose and Reference Concentration
Processes fU.S. EPA. 20021. Guidelines for Carcinogen Risk Assessment fU.S. EPA. 2005al. and
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens fU.S. EPA.
2005b)- Section 11 of the Protocol (Appendix A) describes the general approach to dose-response
analysis used in this assessment
This assessment includes development of a reference dose (RfD), a reference concentration
(RfC), an inhalation unit risk (IUR), and an oral slope factor (OSF). From among the body of
evidence used for the hazard identification assessment, selection of the studies for dose-response
assessment used information from the study confidence evaluations, with particular emphasis on
conclusions regarding the characteristics of the study population, the accuracy of the exposure
estimates for epidemiology studies or dosing methods for toxicology studies, the severity of the
observed effects, and the exposure levels analyzed (see Table 11-1 in U.S. EPA (2020b)).
When suitable data are available, as described in U.S. EPA (2020b). toxicity values should
always be developed for evidence integration conclusions of evidence demonstrates and
evidence indicates (likely) as well as for carcinogenicity descriptors of carcinogenic to humans
or likely to be carcinogenic to humans. In general, toxicity values would not be developed for
"evidence suggests" for noncancer hazard or "suggestive evidence of carcinogenic potential"
for cancer hazard conclusions, respectively.
Additional special considerations were made when selecting studies for dose-response for
Cr(VI), and these are discussed in greater detail in Section 4:
This document is a draft for review purposes only and does not constitute Agency policy.
1-21 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 • Oral animal studies which did not include an exposed group below 20 mg/kg-d were not
2 considered for quantitative analysis9
3 • Inhalation animal studies which did not report measures of particle size and distribution
4 were not considered for quantitative analysis10.
5 • Human studies for nasal cavity effects which did not report clinical outcomes diagnosed by
6 a trained examiner (e.g., physician, otolaryngologist, or trained researcher) were not
7 considered for quantitative analysis. The preferred clinical outcome measures were
8 atrophy of the nasal mucosa; ulceration of the nasal mucosa or septum; perforation of the
9 septum; and bleeding nasal septum.
9 A similar exposure consideration was not necessary for inhalation studies. Fewer animal inhalation studies
were available, and concentrations were below levels that would cause severe toxicity.
"Availability of particle size distribution information for each study is provided in HAWC.
This document is a draft for review purposes only and does not constitute Agency policy.
1-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
2.LITERATURE SEARCH AND STUDY EVALUATION
RESULTS
2.1. LITERATURE SEARCH AND SCREENING RESULTS
Literature searches for studies relevant to the assessment of Cr(VI) have been conducted on
a yearly basis since 2013, with the most recent update current through August 2022.
The results of the screening process outlined in Section 4.3 of the protocol (Appendix A)
have been posted on the project page for this assessment in the HERO database
(https://hero.epa.gov/hero/index.cfm/project/page/project_id/2233), and studies have been
"tagged" with appropriate category descriptors (e.g., "included", "potentially relevant supplemental
material," "excluded"). Results have also been annotated and reported in a literature flow diagram
(see Figure 2-1). Note that because studies reporting multiple types of evidence may have more
than one tag, the sum of all tags in a category may be greater than the number of individual studies
in that category.
Of the 17,898 unique records undergoing title and abstract screening, 14,320 were excluded
because they either did not meet PECO criteria outlined in protocol Section 3.3 (Appendix A) or
were not determined to be potentially relevant supplemental material according to the criteria
outlined in protocol Section 4.3 (Appendix A). Using the sorting criteria outlined in protocol
Section 4.4 (Appendix A) for studies not meeting PECO criteria but still having information relevant
to the specific aims of the assessment, 3,933 records were identified. A total of 138 studies were
considered eligible for study evaluation (56 human health effects studies and 83 animal health
effects studies).
This document is a draft for review purposes only and does not constitute Agency policy.
2-1 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
Database Searches, n = 21,590
PubMed
WOS
Toxline
TSCATS
n - 10,589
n - 9.670
¦
n - 5,342
\
in
in
II
c
Additional Search Strategies
(n = 102)
(manual search of other regulatory
documents)
Combined Dataset after electronic duplicate removal
[n = 17,898)
i
Title/Abstract/Full Text Screen
Potentially relevant supplemental
materia) (n = 3,933)
Did not meet PECO criteria, buttagged as
supplementalstudies:
¦ PBPK models {n = 12)
• ADME studies |n = 173)
¦ Mechanisticstudiesjn =1,574)
• Dermal studies (/? = 733)
• I n j e ct i o n/i ntra tra ch ea I st u d ie s (n = 115)
• Acute exposures {n = 61)
• Human case reports (n = 131)
• ExposureAssessment(n=218)
• Reviews (/) = 347)
• Other Agency Assessments (n = 39)
• Non-Peer Reviewed [n =458)
¦ Related to included study (e.g., previous
cohort studies) (n=43)
» Unabletodetermine(n = 104)
Excluded (n= 14,320)
Did not meet PECO criteria:
• Other Cr compounds (n = 187)
• Othernotpertinentje.g., co-exposures,
ecology, detection methods,
engineering/remediation) \n =13,699)
Studies Eligible for Study Evaluation
(n = 138)
Human health effects studies (n = 56)
Animal health effect studies (n = 83)
Figure 2-1. Literature search flow diagram for Cr(Vl).
This document is a draft for review purposes only and does not constitute Agency policy,
2-2 ' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
2.2. STUDY EVALUATION RESULTS
Human and animal studies have evaluated potential respiratory, gastrointestinal (GI) tract,
hepatic, hematological, immunological, reproductive, and developmental effects following exposure
to Cr(VI). The evidence informing these potential health effects is presented and assessed in
Section 3.2. Detailed rationales for each domain and overall confidence rating are available in
Health Assessment Workspace Collaborative (HAWC).
Overall confidence classifications are presented by effect in Section 3.2. Over 170 studies
met PECO criteria (with about an even number of human and animal studies). Many human and
animal studies contained information on multiple endpoints. With the exception of male
reproductive effects (which had some medium confidence human studies), all human studies
meeting PECO criteria that were included in the hazard identification analysis were rated low
confidence for all hazard domains. Hazard domains having strong animal databases (containing
medium and high confidence studies) were GI, hepatic, hematological, immune, and male and
female reproductive. Most animal respiratory studies were medium confidence, and most of the
animal developmental studies were rated low confidence.
For human health studies evaluated for dose-response data of nasal effects, three were
considered medium (Gibb etal.. 2000a: Lindberg and Hedenstierna. 1983: Cohen etal.. 1974). and
one was considered low confidence (Hanslian et al.. 1967). For human health studies evaluated for
dose-response data of lung cancer, one was considered high confidence fGibb etal.. 20201. one was
considered medium confidence fProctor etal.. 20161. and two were considered low confidence (Birk
etal.. 2006: Gerin etal.. 1993). No quantitative dose-response data for respiratory tract tumors
outside of the lung were suitable for IUR derivation. For example, studies that were identified of
tumors of the nasal cavity were classified as either case reports or review articles without suitable
dose-response data. Exclusion rationale for individual studies for lung cancer and noncancer
effects of the nasal cavity are provided in Appendix D.4.
Graphical representations focusing on outcome specific ratings are presented in the
organ/system-specific integration sections (Hazard Identification, Section 3.2).
This document is a draft for review purposes only and does not constitute Agency policy.
2-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Toxicological Review ofHexavalent Chromium
3.HAZARD IDENTIFICATION
3.1. OVERVIEW OF PHARMACOKINETICS
A detailed review and literature inventory of the database regarding the absorption,
distribution, metabolism, and excretion (ADME) of Cr(VI) is available in Appendix C. This section
primarily focuses on Cr(VI) reduction to Cr(III) (i.e., metabolism) and localized absorption, which
have the greatest impact on assessment conclusions for cancer MOA, susceptibility, interspecies
differences and dose-response.
3.1.1. Pharmacokinetics
Inhaled or ingested Cr(VI) can be reduced to Cr(III) extracellularly by biological fluids
(e.g., blood, gastric juices and epithelial lining fluid) of humans and rodents. In the hexavalent
oxidation state, cellular uptake of chromium oxyanions occurs rapidly via ubiquitous nonspecific
sulfate and/or phosphate anion transporters due to the structural similarity of the chromate and
dichromate anions to these molecules (see Appendix C for more details). Once absorbed by cells,
intracellular reduction generates reactive intermediates Cr(V) and Cr(IV), and finally Cr(III)
fLuczak etal.. 20161. In the trivalent oxidation state, chromium is poorly absorbed by cells via
passive diffusion and has been shown to induce significantly lower tissue chromium burden in
exposed rodents compared to Cr(VI) (Collins etal.. 2010). Thus, extracellular reduction is believed
to be a pathway for detoxification because it decreases the systemic uptake and distribution of
Cr(VI) and reduces the exposure of epithelial cells, the first cells to interact with external factors, to
Cr(VI). In contrast, intracellular reduction of Cr(VI) is considered to be a pathway for its activation
following the cellular uptake of Cr(VI).
Due to site-specific Cr(VI) reduction differences by route of exposure, ingested Cr(VI) will
primarily distribute to gastrointestinal (GI) tract tissues and the liver, while inhaled Cr(VI) will
primarily distribute to the respiratory tract and more readily enter systemic circulation. This was
demonstrated by O'Flaherty and Radike (1991). which is described in further detail in Appendix
C.1.2. These pharmacokinetic factors have implications for Cr(VI)-induced toxicity and
carcinogenicity because target tissue doses will strongly depend on route of exposure. An overview
of ADME for inhaled and ingested Cr(VI) is provided in Figure 3-1.
This document is a draft for review purposes only and does not constitute Agency policy.
3-1 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
ro
E
Lf)
Ingestion
Mucociliary clearance
t,''' "X
¦ Mixingwithsaliva, limitedCr(Vl) reduction, rapid
emptyingto esophagus/stomach
• D i rect co ntaa b etw een w ater an d epithelium
¦Variable/localeed uptaketo oraltissues
* Morphology differences hard palate, buccal
m u cos a, g ing iva, ve ntral/dorsaI to ngue, lip
* Low absorption surface area
¦ Exte nsive m ixingw rth gaartc j u ice, d eiayed
em ptyi ngto s mail intestine
* Si ng )e p ocket of water/gastr ic co ntents m ixture
¦ Fast reduction ofCr(Vl) atlowpH
* Limited uptake to stomach epithelium
• Low absorptionsurfacearea
* Gastric co ntents emptied to small intestine after
variabledelay
<
• M u Iti pie d iscontin uous p o ckets of water, co ntents,
intestinalsecretions; h ighly va riable motilty
• HigherpH, slower reduction than stomach
•Rapid uptaketo intestinal epithelium i '¦N/
•Facilitated transport ' i]/
•Veryhigh absorptionsurfacearea
•Localized uptake highly variable
• Tra ns port to 1 arge i ntestine an d feces
Inhalation
O
¦ Nasaj/nasopharyngeal impaction; deposition at
bronchioles. bifurcation sites, deep lung
¦ M ixing w ith epith eSal liningf lu id an d pulm onary a iveo iar
macrophages
• Variable,/localised reduction of Cr(VI)
> Rapid uptake into respiratory epithelium
JO
fD
vi
-q
-i
CD
r+
o
-1
<
E~
3
fD
U
Systemic blood
• Rapid uptaketo red blood cells [RBCs;
•Fast reduction ofCr(VI) in RBCs
•Cr(lll) "trapped" inRBCs
•Slower reduction by plasma
Other tissues
• Widespread distribution of
Cr(VI) that escapesportal-of-
entry and first-pas Irver
reduction
• Cr(VI) reduction intissues
• Systemic elimination by
urine, likety asCr(lll)
Portal veir
Liver tissue
•Rapid uptake and reduction
• M u Iti pie enzymes an dp athways
'/¦//
VVy
Denotes
epithelium
andtissuesof
gastrointestinal/
respiratory traas
I High transport
• Lowtransport
Figure 3-1. Overview of the absorption, distribution, metabolism, and
excretion of Cr(VI), with focus on extracellular transport and metabolism at
portals of entry.
This document is a draft for review purposes only and does not constitute Agency policy,
3-2 ' DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 Table 3-1 outlines the general findings regarding Cr(VI) pharmacokinetics in different organ
2 systems, and their implications for the toxicological assessment. It is ordered from external/portal -
3 of-entry tissues to internal/systemic tissues and provides additional support for information
4 provided in Figure 3-1.
Table 3-1. Overall findings by system and implications for the toxicological
assessment
System
General findings
Implication for assessment, with rationale
Respiratory
(extracellular)
Reduction of Cr(VI) possible by epithelial lining
fluid (ELF) and pulmonary alveolar
macrophages (PAM). Components of lung
fluids reducing Cr(VI) include glutathione (GSH)
and ascorbate (Asc).11
Extracellular reduction will not be quantified
for inhalation dose-response modeling.
Computational fluid dynamics studies of
inhaled particulates indicate that respiratory
tract deposition does not occur uniformly.
Thus, Cr(VI) will not evenly mix with all
available reducing agent. Particulates may
deposit locally in high amounts in regions of
the respiratory tract with insufficient
extracellular reducing capacity. Impaction in
nasal/nasopharyngeal regions may also occur.
Site-specific respiratory tract particle
deposition and reduction may be highly
variable between individuals.
Respiratory
(cellular/
epithelial)
Rapid uptake of Cr(VI) into epithelial cells, and
reduction to Cr(lll). Reduction by lung tissue
may involve peripheral lung parenchyma (PLP),
Asc, GSH, cysteine, hydrogen peroxide,
riboflavin, iron, and enzymatic pathways.
Intracellular Cr in lung cells may cluster at the
nucleus.12
Oral cavity
(extracellular)
Reduction in saliva is possible13, although the
extent or rate of localized reduction during the
short timescale typical of human or rodent
water swallowing is unknown.
Extracellular reduction in the oral cavity will
not be quantified.
Mixing of drinking water and saliva will not
occur uniformly. High interindividual variability
exists in oral health/saliva status and water
consumption habits. Ingested water
temporarily washes-away saliva from the oral
cavity.
Oral cavity
(cellular/
epithelial)
Uptake to the sensitive oral sites is uncertain.
Higher concentrations in oral tissues were
detected in mice than in rats, but only rats
were susceptible to oral squamous cell
carcinoma in the NTP (2008) studv.
Morphology within different regions of the
oral cavity is highly variable (hard palate,
buccal mucosa, gingiva, ventral/dorsal tongue,
lip), and may impact localized uptake and
reduction.14
A PBPK model will not be used to estimate
oral cavity absorption for dose-response
modeling.
Modeling dynamics of this compartment are
considered too uncertain (see above), although
it will be assumed that direct contact between
water and oral epithelium occurs.
11De Flora etal. T19871 Petrilli et al. T1986I
12Wongetal. (2012). Harris etal. (20051
13Petrilli and De Flora fl9821.
14Kirman et al. f20121. lones and Klein T20131
This document is a draft for review purposes only and does not constitute Agency policy.
3-3 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
System
General findings
Implication for assessment, with rationale
Stomach and
intestine
(extracellular/
lumen)
Gastric juices reduce Cr(VI) by 2nd-order
reaction in a batch system. Total reducing
capacity in all species generally between 10-30
mg/L. Components of gastric juice reducing
Cr(VI) include ascorbate, glutathione, NADH,
and sulfhydryls. Reduction rate decreases as
pH increases.15
A gastric PBPK model of the stomach will be
used to estimate the Cr(VI) dose escaping
stomach reduction. The adjusted daily dose
may be used as the basis for an internal dose
metric for dose-response modeling.
Gastric juice and Cr(VI)-containing water are
expected to have time to become well-mixed,
and the system is single and continuous
(similar to ex vivo batch systems used to study
reduction kinetics). Higher uncertainty exists
for the small intestine lumen. Multiple
discontinuous pockets of water/gastric
contents and intestinal secretions will not be
well-mixed.
Stomach and
intestine
(cellular/
epithelial)
Transport of Cr(VI) occurs rapidly by
nonspecific phosphate and sulfate
transporters. Transport of Cr(lll) believed to
be slower (diffusion). High variability in Gl
absorption for both Cr(VI) and Cr(lll). Cr
uptake may occur primarily in the villi.
Reduction occurs in the tissue.16
A PBPK model will not be used to estimate
epithelial absorption of Cr(VI) in the stomach
or intestine.
There is high uncertainty in simultaneously
quantifying Cr(VI) uptake/reduction, and Cr(lll)
uptake from lumen, plasma, or background
exposure. However, stomach PBPK modeling
of reduction/transit is sufficient for use in
dose-response modeling without incorporating
uptake kinetics. In this assessment, it will be
assumed that the small intestinal epithelium is
exposed to any unreduced Cr(VI) escaping the
stomach.
Blood
Rapid uptake of Cr(VI) into RBCs. Uptake by
anion transporters (i.e., band-3 protein). Rapid
reduction of Cr(VI) in RBCs by GSH. Binding to
hemoglobin and other components in RBC.
Transport of Cr(lll) into or out of RBCs occurs
slowly (thus, bound or unbound Cr(lll) may be
"trapped" in RBC). Cr(VI) uptake into WBCs
also rapid. Reduction of Cr(VI) in plasma
occurs slowly.17
A systemic PBPK model will not be used to
estimate whole-body pharmacokinetics.
Due to rapid clearance and reduction locally by
liver, RBCs, and most other systemic tissues,
BW3/4 scaling of the available dose estimated to
escape reduction in the stomach would be used
for dose-response modeling for systemic
endpoints outside the Gl tract.
Liver
Uptake and reduction of Cr(VI) occurs rapidly.
Reduction by GSH, ascorbate and other
electron donors and enzymes. Uptake into
cells by anion transporters.18
15De Flora et al. f19871 De Flora et al. f19971. Proctor etal. f20121 Kirman et al. T20131.
16Alexander and Aaseth f19951. Shrivastava et al. f20031. Thompson etal. f2015a1.
17Wiegandetal. (19851. Ottenwaelder etal. (19881. Devov etal. (20161.
18Alexander et al. (19821. Alexander et al. (19861. Wiegand et al. (19861. Alexander and Aaseth (19951.
This document is a draft for review purposes only and does not constitute Agency policy.
3-4 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
System
General findings
Implication for assessment, with rationale
All other
systemic
organs and
tissues
In vivo studies at high doses (regardless of
route) have measured widespread Cr in all or
most tissues examined. Distribution may be
dependent on route of exposure.19 Localized
reduction of Cr(VI) to Cr(lll) occurs in all
tissues. Systemic elimination of Cr(lll) from the
whole body occurs primarily via urinary
excretion. Studies also detect chromium in
tissues of control animals due to background
dietary or drinking water chromium (believed
to be in the trivalent form).
1 3.1.1.1. Oral Exposure
2 Extracellular reduction and absorption
3 The extracellular reduction process is important for the oral route of exposure due to the
4 acidity of gastric juice that influences the reduction of Cr(VI). Cr(VI) reduction occurs more rapidly
5 at low pH (Figure 3-2). The pH of the stomach lumen for humans and rodents in the fasted state are
6 approximately 1.3 and 4, respectively (Figure 3-3). Under such conditions, humans would reduce
7 Cr(VI) more effectively than rodents. Because the pH of the small intestinal lumen is higher than
8 that of the stomach, reduction is believed to be slower once Cr(VI) is emptied from the stomach. As
9 a result, Cr(VI) that is not reduced in the stomach compartment may traverse the remaining
10 sections of the GI tract
190'Flahertv and Radike Q9911
This document is a draft for review purposes only and does not constitute Agency policy.
3-5 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
20 40
Time (min)
Figure 3-2. Reduction of Cr(VI) in samples of human gastric juice (fasted
subjects) using data from Proctor etal. (20121. For these experiments, stomach
contents were diluted 10:1 to highlight the effect of pH. Reduction of Cr(VI) in
natural (undiluted) gastric juice occurs faster (see Appendix C.1.3).
1
2
I Stomach ¦ Duodenum ~ Jejunum ~ Ileum
I
Mouse
(A) Fasted
Rat
Human
¦ Stomach ¦ Duodenum ~ Jejunum ~ Ileum
Mouse
(B)Fed
Rat
y
Human
Figure 3-3. GI tract pH values reported in Mcconnell et al. f20081 (rodents:
female BALB/c mice and female Wistar rats) and Parrott et al. f20091
(humans).
Along the GI tract, the concentration of Cr(VI) will be highest at the portal of entry and in
the lumen close to the portals of entry (oral cavity, tongue, esophagus, stomach, duodenum).
This document is a draft for review purposes only and does not constitute Agency policy,
3-6 ' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
Within the epithelium, a concentration gradient will exist across tissue depth, with the greatest
Cr(VI) concentration at the apical surface of the mucosa, and lower levels at deeper components of
the tissue. Differences in tissue morphologies and absorption across the various segments of the GI
tract result in variable Cr(VI) exposures for different tissue and cell types, which have implications
for site-specific uptake and pharmacodynamics (See Sections 3.2.2.3 and 3.2.3.3). Figures 3-4 and
3-5 illustrate how Cr(VI) will distribute and absorb within the GI tract tissues.
The oral epithelium is composed of multiple cell layers (Figure 3-4) (Squier and Kremer.
20011 and regenerates with stem cells located in the relatively deeper layers (e.g., the lamina
propria or basal layer) Hones and Klein. 2013: Marvnka-Kalmani etal.. 20101. The precise location
of the stem cells depends on the region of the oral mucosa (e.g., lip, hard palate, gingiva, tongue)
(Tones and Klein. 2013: Marvnka-Kalmani et al.. 20101. The concentration of ingested Cr(VI) in the
oral cavity may not exhibit a proximal-to-distal gradient because very limited reduction and
dilution will occur in the lumen. However, the surface cell layers will receive higher exposure. The
small intestine is comprised of three anatomical sections, the duodenum, jejunum, and ileum
(Figure 3-5), each of which have different lengths and absorption surface areas fCastelevn etal..
20101. Within the small intestine, the concentration of ingested Cr(VI) that is not reduced in the
stomach will be the highest in the duodenum. The duodenal villi serve as the functional structures
for absorption. Villous epithelial cells are continuously lost and replaced by stem cells in the
bottom two-thirds of the crypt (Potten etal.. 2009: Potten etal.. 19971. Stem cells differentiate as
they move upward from the crypt and are shed at the tip of the villi. Within the stomach, gastric
stem cells are located within glandular pits, and unlike the small intestine, they are nearer to the
lumen and more likely to be exposed to surface irritants fMills and Shivdasani. 20111.
There are species differences in GI tract structure and drinking water consumption patterns
that may impact susceptibility to the effects of ingested Cr(VI). The rodent stomach is segmented
into a glandular stomach and non-glandular (keratinized) forestomach, whereas humans have a
single glandular stomach type (Kararli. 19 9 5 120. Elevated pH has been measured in the
forestomach of rodents (relative to the glandular stomach) fKohl etal.. 2013: Browning etal.. 1983:
Kunstvr etal.. 19761. and pH variation might not follow the same fed/fasted pattern as the
glandular stomach fWard and Coates. 19871. As a result, it is likely that kinetics within the
stomach, and Cr(VI) exposure to the absorptive regions of the stomach, differ between rodents and
humans. Within the oral cavity, the location and type of tissue keratinization (which decreases site-
specific absorption) differs by species, with a greater percentage of the rodent oral epithelium
being keratinized relative to humans (Tones and Klein. 2013). There are also interspecies
differences in the relative lengths and surface areas of small intestinal segments fCastelevn etal..
20101. With respect to the pattern of drinking water consumption, humans ingest beverages
20A comparative 21-day pharmacokinetic study in guinea pigs (which do not have a forestomach), rats, and
mice by NTP (20071 found no fundamental differences in pharmacokinetics that could be attributable to
different stomach structure.
This document is a draft for review purposes only and does not constitute Agency policy.
3-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
sporadically and within a short period of time, whereas rodents consume water at a more sustained
rate over the nocturnal period fYuan. 1993: Spiteri. 19821.
The characterization of interspecies differences in site-specific pharmacodynamics for
Cr(VI) is highly uncertain due to the nature of the observed tumors (see Section 3.2.3). NTP f20081
observed tumors of the oral cavity in rats, and tumors of the small intestine of mice following
exposure to Cr(VI) in drinking water for two years. The lack of oral tumors in mice cannot be
explained by interspecies differences in pharmacokinetics because higher chromium
concentrations have been measured in the oral tissues of mice vs. rats following a 90-day Cr(VI)
drinking water study fKirman et al.. 20121. In addition, rats are generally more prone to oral
cancer development than mice, and mice are more prone to neoplasia in the small intestine
(Ibrahim etal.. 2021: Chandra et al.. 20101 (Appendix D.2).
In GI tract tissues where tumors were not observed in rodents by NTP (20081 (such as the
stomach or colon), there are also interspecies differences that are difficult to model. For example,
chemically induced epithelial tumors of the forestomach in mice and rats are the most common
neoplasms of the GI tract observed by NTP and Carcinogenic Potency databases, but those of the
glandular stomach are rare fChandra etal.. 20101. However, glandular stomach cancer is one of the
major causes of cancer diagnosis and cancer death in humans worldwide (Crew and Neugut. 20041.
It is the 5 th most commonly diagnosed cancer and the 7th most prevalent in the world (Rawla and
Barsouk. 20191. Morphologies of stomach tumors differ greatly between humans and rodents
fHavakawa etal.. 2013: Tsukamoto etal.. 20071. and therefore lack of Cr(VI)-induced stomach
tumors in rodent bioassays may not be directly applicable to humans. Because these interspecies
differences could not be quantified in a pharmacokinetic or pharmacodynamic model, site-specific
internal dose metrics were not derived for GI tract tissues.
This document is a draft for review purposes only and does not constitute Agency policy.
3-8 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
Squamous epithelium (se)
subtypes of the oral cavity
include: Hard palate, buccal
mucosa, gingiva, ventral
tongue, dorsal tongue, and
inner lip
Ingested water and
Cr(VI) (red)
Mucus
Squamous epithelium
(morphology differs
by subtype)
Basement membrane
Lamina propria
Figure 3-4. Schematic of the rat oral cavity depicting the gradient of Cr(VI)
concentration following ingestion of Cr(VI) in drinking water, both from
anterior to posterior locations, as well as across the tissue depth. Drawn based
in part on images by NRC (2011) and lones and Klein (2013). Transmucosal
uptake may lead to systemic absorption.
This document is a draft for review purposes only and does not constitute Agency policy,
3-9 ' DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
Stomach
Ingested water and Cr(VI)
"O
C
03
C
o
4-*
(U
|
o
1—
Q_
-Stem ceils- ;
i
Q) I
X! i
I
\ .
I It
I \ \
I I I
* 1 1
% I •
1 * I
Ingested water and Cr(VI)
High
Isthmus
Neck
Base
c
.2 i
o
Q_
A 1 1
c
o
c
o
!! 1
¦4—'
03
L_
< 1
l_
Q)
&JJ
F
! i
Ł
~o
r
' I
Q
03
' J I
*
\\
J 1
Villi
m
^ v/
H Crypt
A
\ /
Stem cells
Figure 3-5. Schematic of the mouse upper GI tract (stomach and small
intestine) depicting the gradient of Cr(VI) concentration following ingestion of
Cr(VI) in drinking water. Gradient is both from anterior to posterior locations, as
well as across the tissue depth. Drawn based on images by Radtke and Clevers
Ł20051, Fox and Wangf20071. and Kararli f19951.
1 Data limitations of oral pharmacokinetic data
2 Even under controlled rodent pharmacokinetic studies, assessing the oral absorption and
3 whole-body distribution of orally administered Cr(VI) at low doses contains some uncertainty.
4 Only total chromium can be measured in tissues in vivo. Total chromium measured in tissues
5 following oral Cr(VI] exposure results from:
6 Rapid cellular uptake of administered Cr(VI] that was absorbed into the body as Cr(VI).
7 Because Cr(VI) transport is carrier-mediated via nonspecific sulfate and/or phosphate anion
8 transporters, this uptake is rapid in the lumen and systemic tissues. The absorbed Cr(VI) may be
9 transported throughout the body and reduced intracellularly to Cr(III) in tissues and red blood
10 cells. Absorption of Cr(VI) by the intestine and reduction of Cr(VI] in the lumen are competitive
11 processes.
This document is a draft for review purposes only and does not constitute Agency policy,
3-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Toxicological Review ofHexavalent Chromium
1) Slow cellular uptake of Cr(III) that was absorbed into the body as Cr(III), formed from
administered Cr(VI) that reduced to Cr(III) extracellularly and outside of systemic
circulation (e.g., gastric juices). This process is slow and inefficient because Cr(III)
transport occurs by passive diffusion, resulting in a low percent absorption of Cr(III) in the
GI tract, and a low percent absorption of Cr(III) into systemic tissues from plasma.
However, high concentrations of Cr(III) in the lumen may occur during controlled Cr(VI)
studies (via extracellular reduction), leading to more uptake of Cr(III) than would typically
occur from background dietary ingestion.
2) Slow cellular uptake of Cr(III) that was absorbed into the body as administered Cr(VI) and
reduced by other components within systemic circulation (e.g., plasma, liver, red blood
cells). While uptake of Cr(VI) into the intestinal lumen is rapid, systemic reduction to Cr(III)
is also rapid. Once reduced, Cr(III) will diffuse slowly (into or out of) systemic tissues and
circulate throughout the body in plasma. For example, plasma can reduce Cr(VI)
extracellularly, and the resulting Cr(III) absorbed into tissues. RBCs can reduce Cr(VI)
intracellularly, and the resulting Cr(III) can be released to systemic circulation (to be
absorbed by other tissues) after RBCs are broken down.
3) Background uptake and distribution of dietary and drinking water chromium (Cr(III)
and/or Cr(VI)) not administered or controlled in the bioassay. This is supported by the
detection of chromium in the tissues of control animals.
Because chromium becomes trapped within RBCs following exposure to Cr(VI), elevated
RBC chromium persists longer relative to plasma chromium levels following systemic Cr(VI)
absorption. Based on analyses of the RBC:plasma ratios of exposed and unexposed rodents from
the NTP (2008. 2007) studies (see Appendix C.1.2), it may be assumed that a significantly large
percentage of oral ad libitum doses greater than 1 mg/kg-d likely escapes gastric and hepatic
reduction in rodents and is widely distributed throughout the body. At lower doses, it may be
difficult to interpret pharmacokinetic data due to background chromium exposure, and the fact that
a lower percentage of the dose reaches systemic circulation.
3.1.1.2. Inhalation Exposure
Inhalation pharmacokinetics of Cr(VI) differ substantially from ingestion, and there is less
detoxification via extracellular reduction. Deposition of particles along the respiratory tract is not
uniformly distributed and is strongly dependent on particle size. Inhaled particles with a diameter
greater than 5 |im will typically deposit proximal to the trachea (extrathoracic region). Particles
with a diameter in the range of 2.5-5 |im generally deposit in the tracheobronchial region.
Particles with a diameter less than 2.5 |im generally deposit in the pulmonary region. However,
some proportion of larger particles (>2.5 |im) are still capable of reaching the pulmonary region
(OSHA. 2006). Deposition of both larger particles and ultrafine particles (>0.1 |im) can occur in the
head airways, including the nasal passages fHinds. 1999: ICRP. 19941. Particle size distributions in
the air vary between industries or between different processes within the same industrial plant
(OSHA. 2006). Particles of respirable size capable of depositing in the lower respiratory tract have
been observed in some workplace settings (Kuo etal.. 1997a). As a result, this assessment assumes
This document is a draft for review purposes only and does not constitute Agency policy.
3-11 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
deposition in all regions of the respiratory tract is possible, and that some inconsistencies in
observed effects may be due to particle size. Deposition and transmucosal uptake in the oral cavity
are also considered to occur because humans may breathe through both the mouth and nose
(Figure 3-6), as compared to nose-only breathing in rodents.
Within the lower respiratory tract of the lung, particles may locally accumulate at high
quantities in susceptible areas such as airway bifurcation sites (Balashazv etal.. 2003: Schlesinger
and Lippmann. 19781. This is supported by studies showing high chromium deposition at these
sites in the lungs of chromate workers, and a correlation between lung chromium burden and lung
cancer fKondo etal.. 2003: Ishikawa etal.. 1994a. b).
The respiratory environment is less acidic than the gastric environment fKrawic etal..
2017) and would be less likely to effectively reduce Cr(VI) in vivo. Unlike gastric juice, which exists
in the stomach as a single continuous pocket, respiratory tract epithelial lining fluid is a thin,
heterogeneous film fNg etal.. 20041. Inhaled Cr(VI) will not evenly mix with all the available
extracellular components of the lung that are capable of reducing Cr(VI) to Cr(III). Thus,
extracellular components capable of Cr(VI) reduction may be overwhelmed in local regions of the
respiratory tract where high deposition occurs fKrawic etal.. 20171. regardless of the total reducing
capacity of components in the lung. As a result, PBPK modeling of extracellular Cr(VI) reduction in
the lung was not considered for this assessment.
This document is a draft for review purposes only and does not constitute Agency policy.
3-12 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
Figure 3-6. Schematics of the human respiratory system (adapted from
Kleinstreuer et al. f200Bl21l depicting deposition of particles or mists
containing Cr(VI). The term "generation" refers to the branching pattern of
airways. Each division into a major daughter (larger in diameter) and minor
daughter airway is termed a generation fU.S. EPA. 19941.
1 Inhalation pharmacokinetics and target internal doses to the lung and systemic organs will
2 also vary depending on the solubility of the Cr(VI) compound being inhaled. Both high and low
3 soluble forms of Cr(VI) are believed to be absorbed into lung tissue after deposition in the airways
4 fOSHA. 2006}. However, the accumulation rates in the lung and the extent of systemic absorption
5 will differ. Highly soluble Cr(VI] may be rapidly absorbed by cells, leading to high localized Cr(VI)
6 concentrations in the lung tissue. Because the highly soluble Cr(VI) would be rapidly absorbed and
7 cleared, the high localized Cr(VI) lung concentrations may be temporary fl -'Flaherty and Radike.
8 19911. Cr(VI) absorbed by the lungs is rapidly transported to the bloodstream and may expose
21Modified with permission from the Annual Review of Biomedical Engineering, Volume 10 © 2008 by Annual
Reviews, http: //www.anniialreviews.org.
This document is a draft for review purposes only and does not constitute Agency policy,
3-13 ' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
other systemic tissues fOSHA. 20061. Cr(VI) compounds with low solubility may persist in the lung
for longer periods of time and come into close contact with the bronchoalveolar epithelial cell
surface (OSHA. 2006). So while uptake would be slower, there may be a higher exposure over time.
Cr(VI) that is not readily absorbed into the lung may be transported to the stomach by mucociliary
clearance (O'Flaherty and Radike. 1991). As a result, inhaled Cr(VI) compounds with low solubility
may not reach other systemic tissues as readily as soluble Cr(VI), since most Cr(VI) swallowed by
mucociliary clearance would be reduced in the stomach.
Chromium-containing compounds such as the potassium/sodium/ammonium
chromates/dichromates and chromium trioxide are highly soluble in water, while some mixed salt
chromate pigments (such as lead and zinc chromate) are poorly soluble (O'Flaherty and Radike.
1991). While stainless-steel welding fume contains both high and low soluble components, the
Cr(VI) component of the fume is considered highly soluble and may be distributed throughout the
body fAntonini etal.. 2010a: Antonini et al.. 19991.
3.1.1.3. Intracellular Reduction (All Routes of Exposure)
After Cr(VI) uptake by cells, Cr(III) is the ultimate product of the intracellular reduction of
Cr(VI). Depending on the Cr(VI) concentration and reducing agent involved (e.g., ascorbate, or
thiol-containing compounds such as glutathione and cysteine), various amounts of the unstable and
reactive intermediates Cr(V) and Cr(IV) can be generated prior to reduction to Cr(III). This has
implications for pharmacodynamics and mode-of-action (see Section 3.2.3.4). The reduction
pathway via ascorbate occurs with a two-electron reduction to primarily produce Cr(IV) (Reynolds
and Zhitkovich. 20071. although Cr(V) species have been detected following Cr(VI) reduction by
ascorbate (Polisaketal.. 2005: Stearns etal.. 1995: Stearns and Wetterhahn. 1994). When Cr(VI) is
reduced via thiols such as glutathione, there are two distinct one-electron transfers producing both
intermediates Cr(V) and Cr(IV) (Luczak etal.. 2016: O'Brien etal.. 20031. Both the one- and
two-electron reduction steps are immediately followed by one-electron reductions to produce
Cr(III) (Levina and Lay. 2005). Reduction by ascorbate is kinetically favorable, with an estimated
reduction rate 13x faster than cysteine and 61x faster than glutathione fOuievrvn et al.. 20031. and
the reduction pathway via ascorbate accounts for 90% of metabolism in vivo fStandeven and
Wetterhahn. 1992.1991: Suzuki and Fukuda. 1990). It has been shown that in vitro studies may
produce inaccurate results because standard cultured cells contain <1% of the normal in vivo
ascorbate levels (Luczak etal.. 2016). Without adequate ascorbate, glutathione is the major
reducing agent, and the oxidative Cr(V) is the major intermediate; the additional Cr(V) also depletes
glutathione, thereby increasing the abundance of Cr(V) (Luczak et al.. 2016). In addition, the
presence of ascorbate has been shown to stabilize the reactive intermediates generated by the
glutathione pathway, leading to even more potential interaction between Cr(V) and intracellular
components (Martin et al.. 2006). These intracellular reduction pathways are summarized in
Figure 3-7; for further discussion of the biological consequences of the intracellular reduction of
Cr(VI), see Section 3.2.3.4.
This document is a draft for review purposes only and does not constitute Agency policy.
3-14 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Detoxification
Activation
~
Cr(lll)
Cr(VI)
~
Ascorbate
Stabilizing effect
Cr(IV) —~Cr(lll) Primary in vivo
Fast
t
Cr(VI) ~
*
Cr(V)
Cr(VI)
''A
Cr(IV) —~Cr(lll)
Primary in vitro
Slow
Glutathione
Cell membrane
Figure 3-7. Intracellular reduction pathways of Cr(VI). Adapted from
Zhitkovich (2011). The reduction pathway via ascorbate occurs with a
two-electron reduction to Cr(IV), immediately followed by a one-electron reduction
to Cr(III). When Cr(VI) is reduced via thiols such as glutathione, there are two
distinct one-electron transfers producing the intermediates Cr(V) and Cr(IV), and
lastly another electron transfer producing Cr(III). There may be uncertainty
whether the ascorbate pathway truly lacks a Cr(V) intermediate fPoljsak etal..
2005: Stearns et al.. 1995: Stearns and Wetterhahn. 19941. In vivo and in vitro
differences may arise from the media and ascorbate levels used for experiments in
cultured cells. Ascorbate may have a stabilizing effect on the reactive intermediates
produced via the glutathione pathway.
3.1.2. Description of Pharmacokinetic Models
1 A brief description of the available pharmacokinetic models for Cr(VI) are listed below in
2 chronological order in Table 3-2. For this assessment, models adapted from Sasso and Schlosser
3 (20151: Schlosser and Sasso (20141 were used for oral dose-response and rodent-to-human
4 extrapolation (see Appendix C). Physiology parameters defined in Sasso and Schlosser f20151 were
5 revised to account for the fed and fasted states in humans, and to use alternative gastric
6 physiological parameters obtained from literature and other gastric modeling platforms. A minor
7 structural change was also made to harmonize the volumes of stomach lumen and gastric juice (see
8 Appendix C).
This document is a draft for review purposes only and does not constitute Agency policy.
3-15 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-2. Pharmacokinetic models for Cr(VI)
Reference
Species
Notes
O'Flahertv (1996)
O'Flahertv (1993)
O'Flahertv et al. (2001)
O'Flahertv and Radike
(1991)
Rat
Compartments include kidney, liver, bone, GI tract, two lung pools (for
inhalation only), plasma, red blood cells, and lumped compartments for
remaining tissues (rapidly and slowly perfused). A single lumped
compartment represents the GI tract, and reduction kinetics do not include
pH-reduction relationships. This model is not readily extendable to the
mouse.
Calibrated to data from exposure via intravenous injection, gavage,
inhalation (intratracheal), and drinking water (all data are from studies dated
1985 and earlier). Background Cr(lll) exposure is simulated in the model and
contributes to predicted total chromium concentrations.
O'Flahertv et al. (2001)
Human
Kirman et al. (2012)
Rat,
mouse
Compartments include kidney, liver, bone, GI tract, plasma, red blood cells
and a lumped compartment for remaining tissues. A multicompartment
model represents the GI tract (oral cavity, stomach, duodenum, jejunum,
ileum, large intestine), with reduction kinetics based on the model by
Proctor et al. (2012).
Incorporates pharmacokinetic data from experiments designed by the study
authors, and data from other studies. Only data for drinking water and
dietary routes of exposure incorporated. Total concentrations in control
groups subtracted from exposure groups to account for background Cr(lll)
exposure.
Kirman et al. (2013)
Human
Schlosser and Sasso
(2014); Sasso and
Schlosser (2015)
Rat,
mouse,
human
Simulates Cr(VI) reduction kinetics and transit in the stomach.
Incorporates pharmacokinetic model of the stomach lumen by Kirman et al.
(2013; 2012), but with a revised model for Cr(VI) reduction based on
reanalysis of ex vivo data to improve model/data fit.
Kirman et al. (2017;
2016)
Rat,
mouse
human
Same structure as Kirman et al. (2013; 2012), but incorporates a revised
model for Cr(VI) reduction based on additional human gastric juice data.
This model supersedes earlier models by the same investigators.
ICRP (Hiller and
Leggett, 2020)
Human
Biokinetic model assuming linear lst-order transfer rates among different
systemic tissues. Compartments include respiratory tract, stomach, small
intestine, red blood cells, plasma, liver, kidneys, other/soft tissue, trabecular
bone, cortical bone, right colon, left colon, rectosigmoid colon, urinary
bladder, urine, feces. Reduction of Cr(VI) to Cr(lll) not explicitly modeled
(assumed as a linear transfer between different special plasma
compartments).
1 The O'Flaherty Cr(VI) model was adapted from a PBPK model for lead, and it does not
2 describe Cr(VI) kinetics in the target tissue or species of concern (the mouse GI tract). The models
3 by Kirman et al. (2013: 20121 simulate interspecies differences in gastric reduction kinetics in mice,
4 rats, and humans. These models have a structure similar to the human model by O'Flaherty et al.
5 f20011 but differ in their simulation of background Cr(III) exposure and kinetics of the GI tract and
6 bone. The model presented in Sasso and Schlosser f20151 and Appendix C.1.5 only incorporates
This document is a draft for review purposes only and does not constitute Agency policy.
3-16 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review ofHexavalent Chromium
the GI lumen compartments necessary to simulate the non-systemic dose metrics. It incorporates
in vivo gastric kinetics from the Kirman et al. (2013.; 20121 models, but includes a revised ex vivo
reduction model by Schlosser and Sasso (20141 to improve model fit to the ex vivo data of Proctor
etal. (20121 and Kirman etal. (20131. Models of the GI tract incorporate ex vivo reduction models
and may be run independently of the rest of the body if the internal dose is not impacted by blood
or tissue concentrations (Figure 3-8). Some internal dose metrics for GI tract toxicity do not
require estimates of tissue absorption, blood concentrations or systemic elimination. Validation of
whole-body pharmacokinetics is complicated by background exposure and inability to speciate
chromium oxidation states in vivo (see Section ES.7 and 3.1.1.1).
The ICRP model (Hiller and Leggett. 20201 was focused heavily on the distribution of Cr(III)
in the body and had an over-simplified linear assumption for Cr(VI) reduction that would be
inadequate for assessment of effects in the GI tract
Ex vivo
reduction
model
Whole-body model
—>
<-
<-
GI:
V7
Stomach Duodenum Jejunum Ileum
Cr(VI)-
.
->
Gastrointestinal tract model
Figure 3-8. Relationship between ex vivo reduction models, in vivo gastric
models, and whole-body PBPK models.
The Kirman etal. (2017) model made revisions to the previous Kirman et al. models by
incorporating some ex vivo reduction concepts presented in Schlosser and Sasso (2014) (such as
multiple-pathway reactions) and is calibrated to human gastric juice data for fed and fasted
individuals (Kirman etal.. 2016). Ex vivo data provided in Kirman etal. (2016) and De Flora etal.
f20161 were used to assess model uncertainties and population variability and develop a fed-state
gastric reduction capacity (see Appendix C.l). Minor updates to the Sasso and Schlosser (2015) in
vivo model structure and physiology are documented in Appendix C.l.5.
This document is a draft for review purposes only and does not constitute Agency policy.
3-17 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
3.1.2.1. Rationale for Using a Gastric PBPK Model
This toxicological review applies models describing the reduction kinetics and transit of
Cr(VI) in the stomach lumen (as opposed to whole-body PBPK models) for the oral dose-response
assessment and rodent-to-human extrapolation (Appendix C.1.5).
In the GI tract, the extent of reduction in the stomach compartment determines the
maximum Cr(VI) mass or concentration that enters the small intestine. As a result, the stomach
compartment is a major contributor to inter- and intraspecies pharmacokinetic variation. If
reduction does not occur effectively in the stomach, a greater amount of unreduced Cr(VI) will
persist in the small intestinal compartments (duodenum, jejunum, and ileum). Since values of pH in
the small intestinal compartments are higher than in the stomach for all species (Figure 3-3),
reduction may occur less effectively once chromium has emptied from the stomach. Furthermore,
the data underlying the ex vivo reduction model were generated under batch reaction conditions,
which is more similar to the stomach compartment than the dynamic intestine. Modeling the
stomach requires less extrapolation of the data.
The gastric PBPK models are consistent with both ex vivo and in vivo pharmacokinetics
studies. It is estimated that approximately 10% of an ingested dose of Cr(VI) is absorbed in the GI
tract of rodents (Febel etal.. 2001: Thomann et al.. 1994). and this is consistent with the percentage
of unreduced Cr(VI) emptying from the stomach predicted by the gastric PBPK model (Appendix C).
Under typical physiological conditions in the human (gastric pH of below 3, and gastric emptying
half-time of approximately 15-30 minutes), gastric PBPK models predict that approximately 1-
10% of ingested Cr(VI) may be emptied by the human stomach unreduced. This is in agreement
with pooled human gastric juice data by De Flora etal. f20161. which showed that approximately
93% of the chromium is reduced by undiluted gastric juice after 15 minutes. This is also consistent
with a Cr(VI) bioavailability study performed in an in vitro system, which found that human
bioaccessibility could be as high as 20% at low doses (0.005 mg/kg-d) at a gastric pH of 3.0 (but
drastically lower than 20% at low pHlfWang et al.. 20221. Elevated chromium biomarkers (plasma,
red blood cells and urine) have been measured in human volunteers ingesting Cr(VI) fFinlev etal..
1997: Kerger etal.. 1997: Kerger etal.. 1996: Paustenbachetal.. 1996).
While reduction may still occur in small intestinal compartments, effects observed by NTP
(2008) in mice (see Sections 3.2.2 and 3.2.3.2) indicate that unreduced Cr(VI) may traverse the
small intestine. The jejunum and ileum exhibited lower incidences of effects in mice, which may
indicate that Cr(VI) was reduced and/or diluted by intestinal secretions and lumen contents. Data
by Kirman etal. f 20121 also shows chromium concentrations decreasing in the distal direction in
the small intestine of mice exposed to Cr(VI) in drinking water for 90 days. While it is believed that
more Cr(VI) is absorbed in the proximal small intestine, this assessment will not quantify spatial
differences in absorption within the small intestine. It will be assumed that all Cr(VI) which
escapes the stomach and enters the small intestine is capable of exposing the intestinal epithelium
of any region.
This document is a draft for review purposes only and does not constitute Agency policy.
3-18 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
3.2. SYNTHESIS AND INTEGRATION OF HEALTH HAZARD EVIDENCE BY
ORGAN/SYSTEM
3.2.1. Respiratory Tract Effects Other Than Cancer
The respiratory tract is comprised of multiple tissues that are responsible for air intake and
gas exchange. The upper respiratory tract is composed of the nose, nasal cavity, mouth, pharynx,
and larynx. This region filters, warms and humidifies inhaled air prior to entering the lower
respiratory tract, while also facilitating olfactory function. The lower respiratory tract
(i.e., tracheobronchial, and pulmonary regions), which begins at the larynx below the vocal cords, is
composed of the trachea, bronchi, bronchioles, and the alveoli. The pulmonary region facilitates gas
exchange with the blood. The upper and lower airways and gas-exchange region can be affected by
inhaled toxicants that are deposited along the different regions of the respiratory tract, resulting in
a variety of adverse respiratory outcomes. For an overview of how the particle size and solubility
of Cr(VI) compounds will impact the retention and absorption of Cr(VI) in different regions of the
respiratory tract, see Section 3.1.
Effects in the nasal cavity (irritation/ulceration of the nasal mucosa or septum, perforation
of the septum, and bleeding nasal septum) have been documented for decades in humans
occupationally exposed to Cr(VI) in chromium-related industries fBloomfield and Blum. 19281. As
stated in the Cr(VI) IRIS Assessment Protocol (Appendix A), based on EPA's 1998 evaluation of the
literature and the determination that the effects of Cr(VI) on the nasal cavity have been well
established [e.g., OSHA (2006) and U.S. EPA (2014c)]. EPA will not re-evaluate the qualitative
evidence for an association between inhalation Cr(VI) exposure and nasal effects. Rather, the
review of the evidence for nasal effects focuses on identifying studies that might improve the
quantitative dose-response analysis for this outcome. The review of the evidence and dose-
response for nasal effects can be found in Section 4.2.1.1.
For human studies, this assessment focuses on respiratory effects that may be sensitive and
specific to the effects of inhaled Cr(VI) exposure. This includes decrements in lung function
assessed using spirometry, with comparisons against lesser or unexposed individuals. Mortality or
self-reported symptoms (such as cough) that are nonspecific and may be attributed to multiple
other causes were not considered relevant for this assessment. For animal bioassays, this
assessment considered relevant any reported respiratory effects. Animal studies of respiratory
effects following Cr(VI) exposures typically focused on cellular responses (i.e., cell recruitment, cell
function and cellular products), histopathology, and lung weight
3.2.1.1. Human Evidence
Study evaluation summary
Table 3-3 summarizes the human studies considered in the evaluation of the effects of
exposure to Cr(VI) on the lower respiratory tract These comprise five occupational cohort studies
This document is a draft for review purposes only and does not constitute Agency policy.
3-19 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
of workers in industrial settings in which exposure to Cr(VI) is known to occur (predominantly
through inhalation): a chrome electroplating department in Taiwan fKuo etal.. 1997bl. a chromate
production plant in China (Li etal.. 2015b). a chrome electroplating plant in Sweden fLindberg and
Hedenstierna. 19831. several plants in France at which stainless-steel welding was performed
fSobaszek etal.. 19981. and one of the plants participating in the Occupational Chromate Exposure
Dynamic Cohort of China (although it is unclear whether this specific plant produced or applied
chromate, or both) (Zhang etal.. 20221. Five additional studies were considered but were deemed
uninformative due to critical deficiencies fSitalakshmi etal.. 2016: Sharma etal.. 2012: Huvinen et
al.. 2002b: Nielsen etal.. 1993: Bovetetal.. 19771 and are not further discussed (see HAWC for
additional details).
Concentrations of Cr(VI) in air were measured in two of the five studies. Concentrations of
Cr(VI) from stationary monitors and personal samplers at a chrome-plating facility in Sweden
ranged from <0.2 to 46 |ig/m:i fLindberg and Hedenstierna. 19831. Concentrations of Cr(VI) from
personal samplers ranged from 0.2 to 230.0 |ig/m3 in a study of chromium electroplaters in Taiwan
(mean [SD]: 63.2 [67.2] |ig/m3 fKuo etal.. 1997a. b)). In a third study, concentrations of total
chromium from stationary monitors indicated lower exposures compared to these two studies
(median22 [quartile]: 15.45 [19] [ig/m3) in a study of chromate workers in China (Li etal.. 2015b).
After study evaluation, all five studies were categorized as low confidence (Zhang etal..
2022: Li etal.. 2015b: Sobaszek etal.. 1998: Kuo etal.. 1997b: Lindberg and Hedenstierna. 1983). A
lack of air or biomarker measurements in the study of stainless-steel welders fSobaszek etal..
1998). inability to rule out substantial contribution of Cr(III) exposure to biomarker measurements
fZhang etal.. 20221. and potential for residual confounding in the other studies fLi etal.. 2015b:
Kuo etal.. 1997b: Lindberg and Hedenstierna. 19831. raised concerns about the ability of these
studies to appropriately characterize respiratory effects and resulted in low confidence ratings
despite other notable strengths in terms of study design and methods. In all the considered studies,
while the primary focus was on chromium exposure, coexposure to other occupational hazards may
also contribute to observed health effects. For example, other metallic elements in welding fume or
nickel in electroplating work could also impact respiratory health (Antonini etal.. 2010b: ATSDR.
20051. However, similar effects on respiratory outcomes from studies conducted across different
occupational settings, where the specific coexposures would be expected to differ, would alleviate
concern that any observed effects are due solely to coexposures rather than to Cr(VI).
The main results of the five studies considered are summarized in Table 3-4.
22The article states this value as median and quartile; this appears consistent with an inter-quartile range.
This document is a draft for review purposes only and does not constitute Agency policy.
3-20 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Toxicological Review ofHexavalent Chromium
Table 3-3. Summary of human studies for Cr(VI) lower respiratory effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.3
Click to see interactive data graphic for rating rationales.
Author (year)
Industry
Location
Study Design
Pulmonary
Function
Kuo et al. (1997b) (related: Kuo et
al. (1997a))
Chrome electroplating
Taiwan
Cohort (occupational)
L
Li et al. (2015b)
Chromate production
China
Cohort (occupational)
L
Lindberg and Hedenstierna (1983)
Chrome electroplating
Sweden
Cohort (occupational)
L
Sobaszek et al. (1998)
Stainless-steel welding
France
Cohort (occupational)
L
Zhang et al. (2022) (related: Hu et
al. (2022))
Chromate production
China
Cohort (occupational)
L
aStudies excluded due to critical deficiency in one or more domains: Nielsen et al. (1993), Bovet et al. (1977),
Sharma et al. (2012), Sitalakshmi et al. (2016), and Huvinen et al. (2002b) (related: Huvinen et al. (1996)). One of
these studies (Bovet et al., 1977) met the PECO criteria but was found to be uninformative at the study evaluation
stage due to publication prior to the availability of standardized spirometry guidelines from the American
Thoracic Society.
Synthesis of human evidence
Pulmonary function
Four core endpoints were considered in the evaluation of the effects of exposure to Cr(VI)
on pulmonary function: forced vital capacity (FVC), forced expiratory volume in first second
(FEV1.0), the ratio of FEV1.0/FVC, and diffusing capacity of lung for carbon monoxide (DLCO). The
first three of these are measured by spirometry. Other tests of pulmonary function (such as peak
flow, airway responsiveness, and lung volume) were not utilized in any of the four studies
considered. A key consideration for the evaluation of spirometry data is the adherence to
guidelines published by the American Thoracic Society (ATS) (ATS/ERS. 20 1 9 ) 23 and use of
appropriate reference population data for estimation of predicted values. The results from the four
studies evaluating spirometry endpoints are shown in Tables 3-4 and 3-5.
23These guidelines first developed in 1979 with subsequent updates; standardized guidelines were
harmonized with the European Respiratory Society beginning in 2005 with subsequent updates and include
detailed standardized protocols for the collection of spirometry data. Key features of the ATS guidelines
include: recommendations regarding spirometer equipment specifications; protocols to be followed during
the administration of spirometry tests; and the importance of considering age, sex, and height when
interpreting results (ideally by expressing spirometry measurements as a percent of the measurement
predicted, using reference values appropriately matched to the demographic characteristics of the study
population).
This document is a draft for review purposes only and does not constitute Agency policy.
3-21 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-4. Summary of results from human studies of effects of Cr(VI) exposure on pulmonary function
Study
Exposure
Conf.
Result Format
N
FVC
FEV1.0
FEV1/FVC
Li et al. (2015b)
Chromate production
L
Mean (SD)
Exp: 91
Exp: 72.34
Exp: 76.04
Exp: 116.18
expressed as a
Ref: 38
(14.18)
(16.20)
(11.32)
Median total Cra measured in air: 15.45
percent of
Ref: 81.01
Ref: 86.71
Ref: 114.08
Hg/m3 (exposed) and 0.23 (referent) ng/m3
predicted values.
(20.79)
p = 0.196
(24.53)
p = 0.011
(10.79)
p = 0.044
Kuoetal. (1997b)
Chrome electroplating
L
Adjusted
Exp: 26
"=3-
UD
LO
LO
1
cd
P: -368.0 (163.9)
-
regression
Ref: 34
(151.2) mL
mL
Mean Cr(VI) measured in air near
coefficients (SE)
p<0.01
p < 0.05
electroplating tank: 8.0 ng/m3 (Cr factors),
and p-value
2.8 ng/m3 (Cr-Ni factory) and 2 ng/m3) or mixed
exposure to chromic acid and other acids and
metallic salts
groups. See
table below
Sobaszek et al.
Stainless-steel welding
L
Mean (SD)
Exp: 130
Exp: 103 (12)
Exp: 99 (15)
Exp: 95 (8)
(1998)
expressed as per-
Ref: 234
Ref: 101 (13)
Ref: 98 (14)
Ref: 96 (8)
No quantitative exposure measures
cent of predicted
values.
NS
NS
NS
Zhang et al.
Chromate production or application
L
Fully adjusted
Total: 515
P: -1.03 (-2.42,
P: -1.80 (-3.15,
P: -0.77 (-1.43,
(2022)
(unspecified whether one or both)
regression
(918 visits)
0.30) L,
-0.35) L,
-0.10)%,
Cr(VI) exposure based on measured blood
coefficients (95%
p = 0.115;
p = 0.009; Q2:
p = 0.024; - Q2:
chromium concentration (continuous
CI) and p-value
Q2: 0.78 (-2.42,
0.28 (-3.15,
0.93 (-2.69,
variable or quartiles with Q1 as referent).
4.24), Q3: -0.33
3.85)-0.77, Q3:
0.84), Q3: -0.42 (-
Quartiles of blood Cr (ng/L) were Ql: (<1.06),
(-3.94, 3.03),
(-4.55, 2.80),
2.25, 1.41),
Q2: (1.06-2.23), Q3: (2.24-4.90),Q4 :(>4.91)
Q4: -2.41 (-6.06,
1.21), p trend =
0.174
Q4: -4.24 (-8.06,
-0.35), p trend =
0.033
Q4: -1.81 (-3.75,
0.14), p trend =
0.124
aTotal Cr includes Cr(lll) and Cr(VI). No quantitative Cr(VI) exposure measurements reported.
This document is a draft for review purposes only and does not constitute Agency policy.
3-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Toxicological Review ofHexavalent Chromium
Table 3-5. Summary of results from Lindberg and Hedenstierna T19831 study
of effects of Cr(VI) exposure on pulmonary function
Study information
N
FVC
FEV
Exposure
Chrome electroplating
Study confidence
Low
Result format
Mean (SD) expressed as
actual volume (Liters of
air)
Note: Measurements were
taken Monday morning
before work, Thursday
morning before work, and
Thursday afternoon after
work
Males only, Monday
morning before work:
Exp: 26 nonsmokers
Exp: 48 smokers
Ref: 52 nonsmokers
Ref: 67 smokers
Nonsmokers, Exp: 5.61
(0.99)
Nonsmokers, Ref: 5.20 (1.00)
NS
Smokers, Exp: 5.27 (0.90)
Smokers, Ref: 5.66 (1.02)
NS
Nonsmokers, Exp: 4.54
(0.92)
Nonsmokers, Ref: 4.08 (0.85)
NS
Smokers, Exp: 4.31 (0.85)
Smokers, Ref: 4.38 (0.92)
NS
Males and females,
Non-smoker, High Exp
(n = 6)
Mon. morning: 5.96 (1.64)
Thurs. afternoon: 5.75 (1.58)
p<0.01
Mon. morning: 5.13 (1.37)
Thurs. afternoon: 4.92 (1.29)
p < 0.05
Males and females,
Non-smoker, Low Exp
(n = 10)
Mon. morning: 5.41 (1.27)
Thurs. afternoon: 5.35 (1.24)
NS
Mon. morning: 4.45 (1.05)
Thurs. afternoon: 4.43 (0.97)
NS
Males and females,
Non-smoker, Mixed Exp
(n = 15)
Mon. morning: 4.93 (1.17)
Thurs. afternoon: 4.73 (1.22)
p<0.01
Mon. morning: 4.12 (0.92)
Thurs. afternoon: 4.06 (0.95)
NS
Males and females,
Smoker, All Exp (n = 48)
Mon. morning: 5.04 (1.04)
Thurs. afternoon: 4.97 (0.97)
p < 0.05
Mon. morning: 4.07 (0.95)
Thurs. afternoon: 4.00 (0.91)
NS
One low confidence study fLi etal.. 2015bl reported lower FVC and FEV1.0 in chromate
workers compared to referents (workers in the same plant in administrative offices) with little to
no exposure to Cr(VI) in China fLi etal.. 2015b) (Table 3-4). The percent predicted values for FVC
and FEV1.0 in the exposed group were 72.34 (SD: 14.18) and 76.04 (SD: 16.20), respectively,
compared with 81.01 (SD: 20.79) and 86.71 (SD: 24.53), respectively, in the referent group. The
low percent predicted values in both the exposed and referent groups may in part reflect the high
prevalence of smoking (39.56% of exposed and 28.95% of unexposed workers were current
smokers), which was not accounted for in these analyses. Another possible reason for low percent
predicted values across groups is that the referent group had undescribed exposure to Cr(VI) or
other respiratory toxicants. Finally, it is possible that use of reference values from an ethnically
different population (in this case, Japanese and European referent populations, per correspondence
with study author (Tia. 2021)) could have resulted in low percent predicted values (Korotzer etal..
2000). The use of an inappropriate referent to estimate predicted pulmonary function measures
may not impede comparisons of FVC and FEV1.0 between groups within the same study; however,
This document is a draft for review purposes only and does not constitute Agency policy.
3-23 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
the impact could differ for FVC compared with FEV1.0, thus there is greater uncertainty in
FEV1.0/FVC results (mean [SD]: 116.18 [11.32] in exposed, 114.08 [10.79]).
Another low confidence study comparing chrome electroplaters to zinc electroplaters in
Taiwan (Kuo etal.. 1997b) reported average FVC and FEV values were 556.4 mL (SD: 151.2,
p < 0.01) and 368.0 mL (SD 163.9, p < 0.05) lower, respectively, in the group of chrome
electroplaters after adjusting for age and sex (Table 3-4). However, height (an important predictor
for these measures) was not accounted for in comparison of spirometry values.
A low confidence study of chromium electroplaters in Sweden fLindberg and Hedenstierna.
19831 (Table 3-5) did not find significant differences between FVC or FEV1.0 comparing those with
low and high average exposure to chromic acid, nor when comparing exposed workers and a
referent group of auto mechanics. However, when evaluating spirometry measurements over the
course of the work week (pre-shift on Monday morning vs. post-shift on Thursday afternoon), there
were significant decrements in both measures for those in the high exposure group. This finding
demonstrates the potential for short-term effects of chromic acid exposure to impact lung function
within the same individual and is not affected by the potential for confounding by age and height
that is a primary concern for the comparison of exposed and referent group lung function
measures; however, it does not inform the difference between workers exposed to chromic acid
and referent workers.
The fourth low confidence study (Sobaszek etal.. 1998) also did not report significant
differences in FVC, FEV1.0 (or the ratio of FEV1.0/FVC) between exposed and referent groups
(Table 3-4). There were no major concerns regarding selection bias, outcome measurement, or
statistical analyses in this study, which presented results as a percent of predicted values and
followed ATS protocols. Rather, the low confidence rating arose from concerns about the ability of
the study to detect an association in the presence of exposure misclassification arising from the lack
of quantitative exposure data (Sobaszek etal.. 1998). However, an additional analysis conducted in
this study may provide supporting evidence of an association between chronic exposure to
stainless-steel welding fume and decreased pulmonary function. In this analysis, maximal
expiratory flow (MEF) first increased and then decreased with exposure quantified as years of
duration in welding. The initial increase in MEF may indicate that more susceptible workers
quickly left the workforce (i.e., healthy worker effect). Subsequently, the remaining workers
experienced a decrease in MEF after long-term exposure to stainless-steel welding fume (more than
25 years), a pattern that is consistent with the results of the low confidence study reporting
decreases in pulmonary function in workers exposed to Cr(VI) compared to lesser exposed workers
fLietal.. 2015bl.
A fifth low confidence study (Zhang etal.. 2022) reported a statistically signficant decrease
in FEV1 ((3: -1.80 (-3.15, -0.35) L, p = 0.009), as well as a statistically significant decrease in
FEV/FVC (3: -0.77 (-1.43, -0.10) %, p = 0.024, per 1 |ig/L increase in blood chromium
concentration. No significant change in FVC was observed in relation to blood chromium. A
This document is a draft for review purposes only and does not constitute Agency policy.
3-24 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
limitation of this study was the use of blood chromium concentrations to assess exposure to Cr (VI)
in the absence of job, process, or air data. Blood chromium does not distinguish exposure to
trivalent versus hexavalent chromium. In the exposure setting described, trivalent exposure was
likely, and it is not clear how much exposure to Cr (VI) contributed to chromium blood
concentration.
Overall, there is an indication in three low confidence human studies that higher Cr (VI)
exposure is associated with decrements in lung function assessed using spirometry, and the two
remaining low confidence studies may have had insufficient sensitivity to appropriately
characterize such associations.
3.2.1.2. Animal Evidence
Study evaluation summary
The eight animal toxicology studies that were considered in the evaluation of the effects of
Cr(VI) on the respiratory tract are summarized in Table 3-6. All these studies used the inhalation
route of exposure (nose only or whole body) using respirable aerosols24 and examined respiratory
effects in male rats, mice, and rabbits. Female animals were not assessed. The exposure duration
for the mouse studies was 2 years, while the rabbit studies were limited to 4-6 weeks. The rat
studies ranged from 4 weeks to 18 months.
The outcomes reported can be generally grouped into three categories: cellular responses,
lung histology and lung weight Cellular responses include cell recruitment (the transfer of vascular
cells; monocytes, granulocytes/neutrophils, and lymphocytes into the airways), cell function
(macrophage phagocytosis) and release of cellular products (proteins and enzymes). Cell
recruitment is evaluated using bronchoalveolar lavage (BAL) to obtain total cell counts, and relative
abundance of the various resident and recruited populations of cells recovered in the BAL fluid
(BALF) including monocytes, macrophages, granulocytes/neutrophils, and lymphocytes. Cell
function is evaluated by measuring the ability of macrophages to phagocytose foreign particles and
their ability to release protective oxidant enzymes. Cellular products released by protective cells
within the lumen of the lung that can be measured in the BALF include cytokines, intracellular
enzymes, and proteins, as well as other cell signaling chemicals.
Most of the study outcomes focusing on cellular responses and histopathology were rated as
medium confidence with minor concerns that did not negatively affect the overall outcome
confidence rating. Five study outcomes were rated as low confidence (four of these were for lung
weight, and one was for lung histopathology), and one was rated uninformative (Table 3-6).
24For study evaluation, consideration was given to reporting (or lack of reporting) of particle size and
distribution (such as mass median aerodynamic diameter [MMAD] and geometric standard deviation [GSD]).
Lack of reporting on particle sizes negatively impacted the exposure methods sensitivity rating and overall
confidence rating.
This document is a draft for review purposes only and does not constitute Agency policy.
3-25 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Toxicological Review ofHexavalent Chromium
Table 3-6. Summary of included studies for Cr(VI) respiratory effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.3
Click to see interactive data graphic for rating rationales.
Respiratory outcomes
Author (year)
Species (strain)b
Exposure design
Exposure
route
Cellular
responses
Histopathology
Lung weight
Cohen et al. (2003)
Rat (F344)
4, 8, 12, 24, 48 wk
Inhalation
M
Glaser et al. (1985)
Rat (Wistar)
28 and 90 d
Inhalation
M
M
M
Glaser et al. (1986)
Rat (Wistar)
Chronic
Inhalation
U
L
Glaser et al. (1990)
Rat (Wistar)
30 d, 90 d, and 90 d
with 30 d recovery
Inhalation
M
M
L
Johansson et al.
(1986a)
Rabbit (not specified)
4-6 wk
Inhalation
M
L
Johansson et al.
(1986b)
Rabbit (not specified)
4-6 wk
Inhalation
M
M
Kim et al. (2004)
Rat (Sprague-Dawley)
90 d
Inhalation
M
L
Nettesheim et al.
(1971)
Mouse (C57BL/6)
2 yr
Inhalation
L
aln addition to these studies, four studies meeting PECO criteria were found to be uninformative at the study
evaluation stage for all outcomes assessed: Nettesheim et al. (1970), due to incomplete reporting of
histopathological findings in all the groups, and a group of non-English language studies (Adachi et al., (1987;
1986; 1981)), due to the English-language abstract and results indicating that the exposure vehicle purposefully
contained additional contaminants in order to simulate a chromic acid bath. Glaser et al. (1986) was rated
uninformative only for the outcome of histopathology due to incomplete reporting of histopathological findings in
all the groups.
bAII data are for male animals.
Synthesis of animal evidence
Lung cellular responses in BALF
When particulate matter is inhaled, the lungs typically respond by increasing phagocytic cell
populations to aid in clearance of the particles. Populations of macrophages in the lung increase by
replication of the resident lung macrophages (Bitterman etal.. 19841. as well as by recruitment of
monocytes from the bloodstream that travel to the lung and mature to macrophages (van Pud
Alblas and van Furth. 19791. In addition, granulocytes (i.e., neutrophils) can be recruited to assist
in the phagocytosis of the foreign particles (Kodavanti. 20141. These changes in cell populations,
indicative of inflammation, may be accompanied by biochemical markers of cell injury, such as
changes in the amounts of total protein, albumin, and lactate dehydrogenase (LDH) activity in BALF
(Henderson. 1984). These cellular responses are protective immediately following exposure but
This document is a draft for review purposes only and does not constitute Agency policy.
3-26 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
can become injurious to the organism if they are prolonged, leading to long-term changes such as
increased alveolar-capillary permeability (pulmonary edema).
Four of the included studies reported cellular response outcomes, all of which had medium
confidence ratings. Laboratory animals exposed to aerosols of Cr(VI) exhibited changes in the
protective cells that reside in or recruit to the lung. Findings included changes in the number of
macrophages, granulocytes/neutrophils, and lymphocytes, as well as changes in the total BAL cells.
Chromium concentration-related changes in the number of macrophages recovered in the BALF
were observed in all four studies fCohen etal.. 2003: Glaser etal.. 1990: Tohansson et al.. 1986b:
Glaser etal.. 19851. although the direction of the effects were not consistent across studies or
durations of exposure (Figure 3-9).
Statistically significant increases in numbers of alveolar macrophages in BALF were
reported in male rabbits exposed to 0.9 mg/m3 Cr(VI) as sodium chromate aerosol for 4-6 weeks
flohansson et al.. 1986bl and in male Wistar rats exposed to Cr(VI) as sodium dichromate at
concentrations of 0.20 and 0.40 mg/m3 for 30 or 90 days (Glaser etal.. 1990). In contrast, Glaser et
al. f!9851 reported no significant changes in the number of BALF macrophages in male Wistar rats
after 28 days of Cr(VI) exposure, and a significant concentration-dependent decrease in the number
of BALF macrophages from rats exposed to Cr(VI) concentrations of 0.050 and 0.20 mg/m3 for
90 days. The numbers of BALF macrophages in F344 rats exposed to Cr(VI) in the form of calcium
chromate aerosol (0.36 mg/m3) for durations of 4, 8,12, 24, and 48 weeks were decreased relative
to controls at most intervals fCohen etal.. 20031.
While data for the number of BALF macrophages were variable in the available studies,
macrophages were shown by one research group to undergo replication as a consequence of Cr(VI)
exposure via inhalation. Significant increases in specific macrophage populations including
polynuclear macrophages (Glaser etal. (1985). 90 day, LOAEL 0.05 mg/m3), macrophages in
telophase (Glaser etal. (1985). 90 day, LOAEL 0.025 mg/m3) and dividing macrophages (Glaser et
al. (1990). 90 day, LOAEL 0.05 mg/m3) were observed in Wistar rats. In addition, an increase in the
average macrophage diameter was noted following a 90-day exposure f Glaser etal.. 1990: Glaser et
al.. 1985). In contrast, macrophage diameter in male rabbits exposed to 0.9 mg/m3 Cr(VI) for 4-6
weeks was not different from that in controls, although the number of macrophages was
significantly increased flohansson et al.. 1986bl. The inconsistency in effects on BALF macrophages
could be related to the differences in study design (i.e., form of chromium administered, animal
species and strain, exposure design, endpoint methodology). The ability to synthesize results
across studies is limited due to the small number of studies reporting a particular outcome.
Only two studies examined changes in BALF cell populations other than macrophages after
inhalation exposure to Cr (VI). Significant increases in the percentage of BALF lymphocytes were
observed in Wistar rats after 28 and 90 days of exposure to 0.025 mg/m3 and 0.05 mg/m3 Cr (VI).
However, after 90 days of exposure at a higher dose (0.2 mg/m3) the percentage of BALF
lymphocytes was not significantly different from control. Similarly, the percentage of BALF
This document is a draft for review purposes only and does not constitute Agency policy.
3-27 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
granulocytes / neutrophils was significantly increased over control only after exposure to
0.05 mg/m3 Cr(VI), and decreased compared to control at the higher dose of 0.2 mg/m3 fGlaser et
al.. 19851. However, the percentage of BALF granulocytes / neutrophils was demonstrated to
significantly increase over time following exposure to 0.36 mg/m3 Cr(VI) in a different study using
F-344 rats (Cohen etal.. 20031. The differences in rat strains and exposure levels limit ability to
draw conclusions for these other cell populations, but the two studies do demonstrate changes at
both lower and higher levels of exposure.
Limited investigation of BAL cells provides equivocal evidence of changes in functional
activity of the macrophages. Specifically, no functional changes were observed in macrophages
from rabbits exposed to 0.9 mg/m3 Cr(VI) for 4-6 weeks (Johansson et al.. 1986b) based on
measures of oxidative metabolic activity (via ability to reduce nitro blue tetrazolium) and
phagocytic activity (using fluorescently-labeled yeast cells). However, male Wistar rats exposed to
0.05 mg/m3 Cr (VI) for 28 days, and to 0.025 mg/m3 and 0.05 mg/m3 for 90 days displayed
significant increases in phagocytosis of latex particles. Interestingly, at higher concentrations
(0.2 mg/m3) phagocytosis was significantly reduced fGlaser etal.. 19851. In addition, exposure to
0.2 mg/m3 Cr(VI) for 42 days prior and 49 days post challenge with iron oxide particles
demonstrated significant reductions in early and late phase clearance (Glaser etal.. 1985).
One medium confidence study evaluated several biochemical markers of cell injury fGlaser
etal.. 1990). They reported significant increases in total protein, albumin, and LDH activity in the
BALF at all Cr(VI) concentrations in male Wistar rats exposed for both 30 and 90 days (90-day time
point, LOAEL 0.05 mg/m3); increases were concentration-related and were statistically significant
at most concentrations investigated. Glaser etal. f!9901 also included a group of rats exposed for
90 days with a 30-day recovery period. The author found that many of the BALF endpoints,
including total number of macrophages, number of dividing macrophages, and LDH levels, had
returned to approximately control values at the end of the recovery period. However, BALF total
protein remained statistically significantly elevated at all exposure concentrations, and BALF
albumin remained statistically significantly elevated in the two highest concentration groups (0.20
and 0.40 mg/m3) even after recovery (Figure 3-9). Although only evaluated in one medium
confidence study, there is additional support for these findings. Zhao etal. f 20141 (considered a
supplemental study due to use of intratracheal instillation exposure) reported statistically
significant increases in albumin and total protein levels in BALF isolated from male Sprague-
Dawley rats exposed to 0.022 or 0.22 mg/kg Cr (VI) once per week for four weeks via intratracheal
instillation.
Although increases in BALF total protein are characteristic of acute lung injury, this marker
alone is considered insufficient to indicate lung injury due to its nonspecific nature and unknown
source. BALF protein can increase due to leakage of vascular fluid, and/or lung cells releasing more
protein in the alveolar lining fluid. A more specific indicator is the observation of increased BALF
albumin, which comprises a major portion of BALF protein. Albumin in BALF can only come from
This document is a draft for review purposes only and does not constitute Agency policy.
3-28 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 vascular leakage, since lung cells will not make and release albumin to the lumen fKodavanti.
2 20141: consequently, increased albumin indicates an alteration in the epithelial and vascular
3 permeability of the lung. While the database that evaluated BALF albumin, protein and LDH only
4 includes one to two studies, the positive evidence suggests lung epithelial and vascular injury
5 following Cr (VI) exposure.
End point
Study Name
Animal Description Observation Time
significant increase^ Significant decrease
Total protein in BALF
Glaser etal. (1990)
Rat, VWstar (o)
90.0 days
120.0 days
Albumin in BALF
Glaser etal. (1990)
Rat, VWstar (3)
90.0 days
120.0 days
LDH in BALF
Glaser etal. (1990)
Rat, VWstar (3)
90.0 days
120.0 days
Diyding macrophages in BALF
Glaser etal. (1990)
Rat, VWstar (3)
90.0 days
120.0 days
Granulocytes in BALF
Glaser etal. (1985)
Rat, VWstar (3)
90.0 days
Lymphocytes in BALF
Glaser etal. (1985)
Rat, VWstar (3)
90.0 days
Macrophage diameter in BALF
Glaser etal. (1985)
Rat, VWstar (3)
90.0 days
Johannson et al. (198
6) 63708 Rabbit, unspecified (3) 6.0 weeks
Macrophages in telophase in BALF Glaser etal. (1985)
Rat, VWstar (3)
90.0 days
Pol ynu clear macro phages in BALF
Glaser etal. (1985)
Rat, VWstar (3)
90.0 days
Total Cells in BALF
Cohen etal. (2003)
Rat, Fischer F344 (3)
4.0 weeks
8.0 weeks
12.0 weeks
24.0 weeks
48.0 weeks
Total Macrophages in BALF
Cohen etal. (2003)
Rat, Fischer F344 (3)
4.0 weeks
8.0 weeks
12.0 weeks
24.0 weeks
48.0 weeks
Glaser etal. (1985)
Rat, VWstar (3)
90.0 days
Glaser etal. (1990)
Rat, VWstar (3)
90.0 days
120.0 days
Johannson et al. (198
6) 63708 Rabbit, unspecified (3) 6.0 weeks
Total Neutrophils in BALF
Cohen etal. (2003)
Rat, Fischer F344 (3)
4.0 weeks
8.0 weeks
12.0 weeks
24.0 weeks
48.0 weeks
Viability of BAL cells
Glaser etal. (1990)
Rat, VWstar (3)
90.0 days
AA A
AA A
A-»—A-
AA A
A4-
aa
-A
—~
-A
—~
—~
-V
Figure 3-9. Lung cellular responses in BALF in male animals. The 120-day
observation time in Glaser et al. (1990) incorporates 90 days of exposure
followed by a 30-day period of no exposure (recovery time). Click to see
interactive graphic. A graphic containing 30-day data by Glaser et al. (1990) can
be found in HAWC. An expression of dose-response for selected cellular responses
can be found in Section 4.2.1 and in HAWC.
6 Lung histopathology
7 Histopathology is a classic approach used in evaluating effects on the lung and can detect a
8 large range of effects from minor changes in cell populations to significant structural alterations.
9 Seven of the included studies reported histopathological outcomes, comprising five medium
This document is a draft for review purposes only and does not constitute Agency policy.
3-29 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
confidence, one low confidence, and one uninformative study. Nettesheim etal. fl9711 was rated
low confidence for the outcome of histopathology. Results for this study were only provided
qualitatively and without identifying lesions in any specific treatment group or comparison to
control (see HAWC for details).
One of the medium confidence studies dealt specifically with in vitro ultrastructural electron
microscopy of macrophages with no additional tissue characterization (Johansson et al.. 1986b). In
general, three of the four remaining medium confidence, short-term and subchronic studies of
Cr(VI) in rats and rabbits provide consistent evidence of histiocytosis (macrophage accumulation)
in the lung fKim etal.. 2004: Glaser etal.. 1990: Tohansson et al.. 1986al while one subchronic rat
study (Glaser etal.. 1985) reported normal histopathology findings following Cr(VI) exposure
(Figure 3-10).
In one medium confidence study, the incidence of accumulation of macrophages in the
alveolar and peribronchial region of the lung was increased in male Wistar rats exposed to
0.050-0.40 mg/m3 Cr(VI) as sodium dichromate for exposure durations of 30 days (incidence:
30%-80%; the concentration-response curve was nonmonotonic, with maximal incidence at
0.10 mg/m3), 90 days (incidence: 90%-100%), and 90 days with a 30-day recovery period
(incidence: 50%-100%) (Glaser etal.. 1990). A second medium confidence study of similar design
by the same authors did not appear to have investigated these effects (Glaser etal.. 1985).
Additionally, macrophage aggregation and the accumulation of foamy cells were observed
in male Sprague-Dawley rats exposed to Cr(VI) as chromium trioxide aerosol for 90 days Kim et al.
(2004). All rodents in the high concentration group (1.25 mg/m3) exhibited accumulation of
macrophage aggregations and foamy cells in the alveolar region. This effect was observed to a
lesser extent at 0.5 mg/m3 but was not observed at 0.2 mg/m3. This indicates a dose-response
relationship; quantitative data for these effects were not presented in this study but the pattern can
be inferred based on statements regarding number of animals (i.e., 'all', 'less than all', 'none').
Finally, increased intra-alveolar or intrabronchiolar accumulation of macrophages was
reported in 4 of 8 male rabbits exposed to 0.9 mg/m3 Cr(VI) in the form of sodium chromate for
4-6 weeks (Tohansson etal.. 1986a). Some macrophages were enlarged, multinucleated or
significantly vacuolated and accumulated in a nodular formation. In this study and a companion
study that examined macrophages lavaged from the right lung of these rabbits flohansson etal..
1986b). ultrastructural examination of macrophages revealed large lysosomes with dark or
electron-dense patchy inclusions and short membranous fragments or lamellae. The percentage of
cells that contained inclusions and the percentage of macrophages with a smooth surface were
stated to be significantly increased in the Cr(VI)-exposed group (p < 0.02; however, quantitative
data were not presented (Tohansson etal.. 1986b).
Evidence for Cr(VI)-related histopathologic changes in the lungs other than macrophage
accumulation is limited, and there is some suggestion of a transient effect. A high incidence of
bronchioalveolar hyperplasia (70-100%) was reported in male Wistar rats after 30 days of
This document is a draft for review purposes only and does not constitute Agency policy.
3-30 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
exposure to 0.050-0.40 mg/m3 Cr(VI) relative to the control (10%) fGlaser etal.. 19901. The same
study reported lower incidence of this effect after 90 days of exposure, and after 90 days of
exposure with a 30-day recovery period. There was an increased incidence of fibrosis (10-40%) in
the groups exposed for 30 days to concentrations at or above 0.1 mg/m3 Cr(VI), but no increase for
the 90-day exposure groups. Glaser etal. (1990) also stated that the upper airways of male Wistar
rats exposed 0.1-0.40 mg/m3 Cr(VI) showed focal inflammation; however, incidence data were not
reported, and the exposure period was not stated. Other investigators did not discuss examination
of the upper respiratory tract in experimental animals. Glaser etal. f!9851 noted qualitatively that
all Wistar male rats exposed for 90 days to 0.025-0.20 mg/m3 Cr(VI) exhibited normal histologic
findings in the lung. Nettesheim etal. (1971) exposed mice to calcium chromate dust from
6 months to approximately 120 weeks at a single concentration of 13 mg/m3. This concentration
was significantly higher than those used in the Glaser et al. studies. The study observed marked
changes in the small airways (ranging from epithelial necrosis and atrophy to marked hyperplasia).
In addition, the study observed bronchiolarization of the alveoli, and alveolar proteinosis with
distention of the terminal bronchioli and alveoli.
In general, histiocytosis and other effects observed in macrophages were observed in the
lung following Cr(VI) exposure. Less data was available for bronchiolar hyperplasia, and there is
some indication those effects did not persist The study design by Glaser etal. (1990) allowed for
histopathological effects to be observed as a function of concentration and time (including after a
recovery period). Bronchiolar hyperplasia peaked at the earliest time point examined (30 days)
and diminished over time. Histiocytosis peaked at 90 days and only slightly diminished during the
30-day recovery period. Based on the 30- and 90-day experiments, and the recovery period data,
the structural changes in the lung appear to be transient while the influx of cells persists.
Endpoint Study Name Animal Description
Abnormal macrophage reaction Johannson et al. (1986) 63707 Rabbit, unspecified (r-. )
Nodular macrophages Johannson et al. (1986) 63707 Rabbit, unspecified (r-')
Macrophage effects (qualitative) Johannson et al. (1986) 63708 Rabbit, unspecified (r-')
Bronchioalveolar Hyperplasia Glaser et al. (1990)
Glaser et al. (1990)
Histopathology (General)
Inflammatory reactions
Lung histiocytosis
Relative lung weight
Glaser et al. (1985)
Kim et al. (2004)
Glaser et al. (1990)
Rat, Wistar (r-')
Rat, Wistar (c)
Rat, Wistar (c)
Rat, Wistar (c)
Rat, Wistar (c)
Rat, Wistar (<¦')
Rat, Wistar (c)
Observation Time
6.0 weeks
6.0 weeks
6.0 weeks
30.0 days
90.0 days
120.0 days
30.0 days
90.0 days
120.0 days
90.0 days
Rat, Wistar (>¦')
Rat, Wistar (c)
Rat, Wistar (c)
Johannson et al. (1986) 63707 Rabbit, unspecified (c)
Glaser et al. (1990) Rat, Wistar (c )
Rat, Sprague-Dawley(c) 90.0 days
30.0 days
90.0 days
120.0 days
6.0 weeks
30.0 days
) no change^^ significant increase^ Significant decrease
0.4 0.5 0.6 0.7 0.8 0.9
mg/m 3
1 1.1 1.2 1.3
Figure 3-10. Histopathological results and effects in macrophages in male rat
lungs. Results from Kim etal. (2004) were qualitative, and dose ranges and the
noted statistically significant dose groups are presented here for comparative
purposes. The 120-day observation time from Glaser etal. (1990) incorporates 90
This document is a draft for review purposes only and does not constitute Agency policy.
3-31 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
days of exposure followed by a 30-day period of no exposure (recovery time). Click
to see interactive graphic. A figure containing incidence data of selected
histopathological outcomes can be found in Section 4.2.1 or in HAWC.
Lung weight
Increases in lung weight, a nonspecific indicator of lung injury, can occur from a variety of
pulmonary conditions, including edema, inflammation (including macrophage accumulation),
fibrosis, accumulation of foreign matter, or abnormal tissue growth (e.g., tumors). Changes in lung
weight were examined in five of the included studies, one of which was medium confidence while
the remaining four were considered low confidence for this endpoint
The relative lung weight outcome in Glaser etal. fl9901 was rated as low confidence
because the study lacked sufficient methodological details for measuring lung weight and reduced
body weight gain in exposed rats. The relative lung weight outcome in Glaser etal. (1986) was
rated as low confidence because the study lacked sufficient methodological details for measuring
lung weight, only included data for the high dose group, and did not report absolute lung weight
(despite reporting end-of-study body weight loss). The lung weight outcome in Tohansson etal.
(1986a) was rated low confidence for several reasons: inconsistent exposure times on study,
variable weight/age of animals in the control and exposure groups, lack of documentation of end-
of-study weight, and reporting of absolute lung weight only. The Kim etal. f20041 study was also
rated low confidence for lung weight due to reporting of only relative weights, when both relative
and absolute weights of the lung and other organs are preferred for assessing effects from body
weight changes and differing types of lung toxicity.
Increased lung weight, which was attributed to accumulation of macrophages, was
observed in one medium confidence and one low confidence study following subchronic inhalation
exposure to Cr(VI). Glaser etal. f!9851. reported increased mean relative lung weights (9-35%) in
Wistar rats exposed for 90 days to Cr(VI) at concentrations of 0.05-0.20 mg/m3. Study authors also
noted that relative lung weights were also increased after 28 days of exposure to Cr(VI)
concentrations >0.05 mg/m3; however, quantitative lung weight data were not presented for these
higher doses. In a similarly designed study by the same investigators, Glaser etal. (1990) reported
a concentration-dependent increase in relative lung weight in Wistar rats following both 30 and 90
days of exposure (9-48%), and following a 90-day exposure with a 30-day recovery period (5-
23%); the increase was statistically significant at concentrations of 0.10-0.40 mg/m3 at all time
points, and at the lowest concentration (0.05 mg/m3) after 30 days of exposure. In contrast,
statistically significant changes in lung lower left lobe weight were not observed in male rabbits
exposed to 0.9 mg/m3 for 4-6 weeks (Tohansson etal.. 1986a). and changes in relative lung weight
were not observed in male Sprague-Dawley rats exposed at concentrations ranging from 0.2-
1.25 mg/m3 for 90 days (Kim etal.. 2004).
In the only available chronic study (Glaser etal.. 1986). mean relative lung weight in Wistar
rats exposed to 0.10 mg/m3 (highest concentration tested) for 18 months and kept on study for
This document is a draft for review purposes only and does not constitute Agency policy.
3-32 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
another 12 months (total time on study: 30 months) was 15% greater compared with controls,
although this change cannot be interpreted as clearly due to macrophage accumulation given the
observation of lung tumors at this concentration. Lung weights were not reported for the low- and
mid-concentration exposure groups where tumors did not develop, but no changes were noted by
the study authors.
To summarize, although there were some inconsistencies in the evidence, increases in lung
weights in Wistar rats were observed in the only medium confidence study available and a second
low confidence study by the same authors (Figure 3-11). These changes in lung weight may
represent an indicator of nonspecific lung injury or inflammation associated with Cr(VI) inhalation.
The studies reveal that changes in lung weight may vary by species, strains, and exposure duration
and may attenuate over time.
Endpoint
Study Name
Animal Description Observation Time
0 no change^, significant increase ^ Significant decrease
Relative lung weight
Glaser et al. (1985)
Rat: Wistar (5)
90.0 days
Glaser et al. (1986)
Rat Wistar (5)
30.0 months
Glaser et al. (1990)
Rat Wistar (Ł)
30.0 days
Rat: Wistar (
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
particle response25. Increases in BALF total protein, albumin, and lactate dehydrogenase (LDH)
activity are characteristic of acute lung injury. While total protein is a nonspecific indicator, the
concentration of albumin in the BALF is normally very low, and an increase indicates an alteration
in the epithelial and vascular permeability of the lung. Damage to cells releases the cytosolic
enzyme LDH; increased enzymatic activity of LDH in the BALF is a common finding with acute lung
injury (Henderson etal.. 19851. The increase in BALF albumin and LDH activity provide evidence of
lung injury following Cr(VI) exposure via inhalation; however, it should be noted that this evidence
came from a single study, and no other studies examined these effects. These findings were
accompanied by some evidence of histiocytosis (macrophage accumulation) and increased
leukocytes in plasma (see Section 3.2.6), which are supportive of inflammatory lung responses
(Nikula et al.. 2014). although these findings generally lessened with longer chromium exposure
durations and may reflect adaptation or resolution of the cellular responses during these later time
points of exposure.
The evidence base of histopathological effects in the lung were mostly limited to
macrophage accumulation, which were observed by multiple studies of medium quality. Findings
for other histopathological changes, such as bronchioalveolar hyperplasia, were only reported in
one study.
Increased lung weight was observed in the single medium confidence study in Wistar rats,
but not in lower confidence studies in other species and strains. However, lung weight is a
nonspecific indicator of lung injury and may be a consequence of multiple other more sensitive
outcomes (such as increased macrophages).
3.2.1.3. Mechanistic Evidence
Mechanistic evidence indicating the biological pathways involved in respiratory toxicity
following the inhalation of Cr(VI) is summarized below. Studies of human occupational inhalation
exposures, in vivo studies in mammals that were exposed via inhalation or intratracheal instillation,
and in vitro studies in human primary or immortalized lung cells were prioritized for informing
interpretations of respiratory health effects following inhalation exposure to Cr(VI) in humans,
although systemic markers of toxicity following inhalation exposures were also considered. These
studies focused primarily on oxidative stress and cellular toxicity of the lung and are summarized in
Appendix Table C-31 unless otherwise noted.
Oxidative stress
Cr(VI) compounds are strong oxidizers and can readily enter cells, where they interact with
intracellular reductants to form Cr(VI) intermediate species [Cr(V) and Cr(IV)] and the stable
25For control groups, studies typically exposed rodents to filtered air or inert aerosols (with diluent likely
being sterile water, although none of the articles provided details). Neither of these are expected to have
adverse effects on the airways.
This document is a draft for review purposes only and does not constitute Agency policy.
3-34 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Cr(III). These intermediate species form reactive oxygen species (ROS) that at high levels can
damage intracellular components, including DNA. Increased oxidative stress induced by Cr(VI) has
been consistently reported in many species and cell types (reviewed in Cancer, Section 3.2.3)
Twenty-three observational studies measuring various indicators of oxidative stress in
humans exposed to Cr(VI) were identified that detected systemic biomarkers of oxidative damage
in urine and/or blood (Appendix Table C-56). While a few occupational exposure studies did not
detect statistically significant indicators of oxidative stress in exposed workers (Wultsch etal..
2014: Pournourmohammadi etal.. 2008: Kim etal.. 1999: Faux etal.. 1994: Gao etal.. 19941. most
studies reported statistically significant increased incidences of oxidative stress through increased
levels of relevant markers (e.g., 8-OHdG adducts, lipid peroxidation, decreased levels of antioxidant
enzymes) that correlated with exposure to Cr(VI) in urine and blood fEl Saftv etal.. 2018: Hu etal..
2018: Yazar andYildirim. 2018: Pan etal.. 2017: Mozafari etal.. 2016: Elhosarv etal.. 2014:
Zendehdel etal.. 2014: Wang etal.. 2012b: Zhang etal.. 2011: Kalahasthi et al.. 2006: Goulartetal..
2005: De Mattia etal.. 2004: Maeng etal.. 2004: Kuo etal.. 2003: Huang etal.. 1999: Gromadziriska
etal.. 19961. One group investigated welders exposed to Cr(VI), finding significant upregulation of
a glycoprotein, Apolipoprotein J/Clusterin, that correlated with chromium levels in blood and urine;
ApoJ/CLU has been shown to be involved in cellular senescence and is implicated in diseases
related to oxidative stress, inflammation, and aging (Alexopoulos etal.. 2008).
Less evidence is available for oxidative stress measured in the lung. One study in exposed
workers, Kim etal. f!9991. analyzed respiratory epithelial cells from exposed lead chromate
pigment factory workers and did not detect a difference in 8-OHdG levels compared to office
workers in the same factory. However, the chromium levels measured in the blood were similar
between the exposed and referent groups, indicating that perhaps exposure misclassification could
have contributed to the null findings. In animals, Maeng etal. (2003) exposed rats via inhalation to
0.18 or 0.9 mg/m3 sodium chromate for 1, 2, or 3 weeks and reported increased formation of 8-
OHdG adducts after 1 week exposure that resolved at weeks 2-3, despite consistently diminished
activity of the enzymes that repair these lesions at weeks 1-3. These results are supported by two
studies exposing rats to Cr(VI) via intratracheal instillation that detected significantly increased
oxidative DNA lesions (8-OHdG) in the lung following four weekly intratracheal instillations of
0.063 or 0.630 mg Cr/kg fZhao etal.. 20141 or once daily administrations of 0.09 mg Cr(VI)/kg for
three consecutive days (Izzotti etal.. 1998).
Inhalation exposures provide a direct route for Cr(VI) compounds to be absorbed by the
bronchial epithelium, and increased oxidative stress induced by Cr(VI) has been confirmed in
studies of human lung cells. Cells deficient in the ability to repair oxidative DNA lesions were
reported to have a significant increase in cytotoxicity and cell cycle delay following Cr(VI) exposure
fReynolds etal.. 2012: Reynolds and Zhitkovich. 20071. Cr(VI) exposure has also been observed to
cause oxidative stress with minimal or no cytotoxicity, indicating that oxidative stress may in some
instances be induced at levels that do not affect cell viability. Caglieri etal. f20081 noted increased
This document is a draft for review purposes only and does not constitute Agency policy.
3-35 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
lipid peroxidation in BEAS-2B human bronchial epithelial cells with cytotoxicity but also in A549
human lung adenocarcinoma cells at subtoxic levels. Asatiani et al. f2011: 20101 observed
increased ROS and the antioxidant enzymes glutathione peroxidase, glutathione reductase, and
catalase at transiently toxic Cr(VI) concentrations. Martin etal. (20061 found that adding
glutathione to Cr(VI)-treated cells decreased levels of ROS; conversely, addition of ascorbate
(Vitamin C), a primary intracellular reducer of Cr(VI), increased levels of ROS. The authors theorize
that the ascorbate reduction pathway could interact with reactive Cr(V) intermediates that are
generated via the glutathione pathway, stabilizing Cr(V) and leading to more potential interaction
between Cr(V) and intracellular components. In addition, ascorbate reduction of Cr(VI) occurs at a
much faster rate than glutathione and has been shown to result in higher levels of genotoxicity than
glutathione (Zhitkovich. 20111. Another group reported that cellular thioredoxins and
peroxiredoxins are especially sensitive to oxidation by Cr(VI), disrupting redox signaling and
affecting cell survival f Myers etal.. 2011: Myers etal.. 2010: Myers and Myers. 2009: Myers etal..
20081.
Cytotoxicity
Apoptosis, or programmed cell death, typically plays a protective role in eliminating
damaged cells from the body but can also be triggered by excessive levels of ROS, contributing to
tissue damage and inflammation. The evidence from studies of exposed workers for specific
measures of apoptosis is sparse due to inadequate information to characterize Cr(VI) exposures.
Gambelunghe etal. f20031 did not detect an increase in apoptosis in lymphocytes among chrome-
plating workers, although this study was estimating cell death using the comet assay, which is an
insensitive method of measuring apoptosis (Appendix Table C-59). Wultsch etal. (20171 reported
increased cytotoxicity in the exfoliated buccal and nasal cells of electroplaters indicated by
histopathological evidence of nuclear anomalies consistent with apoptosis; however, this study was
evaluated for another nuclear effect, micronuclei (Section 3.2.3.2), and was found to be
uninformative due to critical deficiencies in the exposure domain. Halasova etal. (20101
determined that expression of the apoptosis inhibitor survivin protein was decreased and pro-
apoptotic p53 was increased in former chromium workers with lung cancer compared to
unexposed lung cancer patients, but the authors did not describe methods for exposure assessment
and characteristics of the exposed and unexposed groups that may also affect the apoptosis
measures were not compared. In animal models, one intratracheal instillation exposure study in
rats observed increased apoptosis in bronchial epithelium and lung parenchyma (D'Agostini etal..
20021.
Cytotoxicity occurring at micromolar Cr(VI) levels that increases with dose and duration of
exposure has been consistently observed in numerous in vitro studies in human lung cells fYang et
al.. 2017: Reynolds etal.. 2012: Asatiani etal.. 2011: Asatiani etal.. 2010: Caglieri etal.. 2008:
Reynolds and Zhitkovich. 2007: Martin etal.. 2006: Pascal and Tessier. 2004: Carlisle etal.. 2000:
Popper etal.. 19931. with some studies specifically detecting increases in apoptotic cell death
This document is a draft for review purposes only and does not constitute Agency policy.
3-36 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
fReynolds etal.. 2012: Azad etal.. 2008: Reynolds and Zhitkovich. 2007: Gambelunghe etal.. 2006:
Carlisle etal.. 20001. Evidence for the involvement of a p53-mediated pathway for the induction of
apoptosis was conflicting; Carlisle etal. f20001 observed a 4-6 fold increase in p53 in LL-24 human
lung fibroblasts, and Gambelunghe etal. f20061 observed increased expression ofp53 inMOLT-4
lymphoblastic leukemia cells, but a similar increase in p53 was not observed in BEAS-2B human
bronchial epithelial cells, and Reynolds and Zhitkovich (20071 determined that p53 status had no
effect on apoptosis (or cytotoxicity) in primary human lung IMR90 fibroblasts or H460 human lung
epithelial cells. Similarly, information on the identification of caspases involved in Cr(VI)-induced
apoptosis was conflicting, with one group reporting that inhibiting caspase-3, -8 and -9 did not
reduce apoptosis in MOLT-4 lymphoblastic leukemia cells (Gambelunghe etal.. 20061. while
another group reported a significant decline in apoptosis after specific suppression of caspase-9 in
H460 human lung epithelial cells fAzad etal.. 20081. Autophagy, another cellular defense
mechanism that can alternately induce or suppress cell death, was reported following Cr(VI)
exposure in A549 human lung adenocarcinoma cells (Yang etal.. 20171. The autophagy was
correlated with a transcription factor, HMGA2, that is highly expressed in lung cancer patients, and
was suppressed by silencing HMGA2.
Cytotoxicity appeared to be dependent on cell type, possibly reflecting underlying
differences in sensitivity, with A549 lung adenocarcinoma cells slightly more resistant to
cytotoxicity than BEAS-2B bronchial epithelial cells derived from non-tumorigenic cells. Asatiani et
al. (20111 observed that at doses <5 [J.M, the cytotoxicity in HLF fetal human lung fibroblasts and L-
41 human epithelial-like cells resolved after 24 h, but these concentrations were sufficient to
induce oxidative stress and an upregulation of antioxidant enzymes. Increasing levels of ascorbate
to better simulate physiological levels, were found to potentially increase oxidative damage (Martin
etal.. 20061 or promote cytotoxicity and apoptosis by forming Cr-DNA adducts (Reynolds etal..
2012: Reynolds and Zhitkovich. 2007: Carlisle etal.. 20001. This evidence implies that the
pathways for Cr(VI)-induced apoptosis and toxicity in human lung cells are complex and likely to
differ substantially among species and cell type.
Lung cellular inflammation
Specific support for the lung cellular responses in animals discussed in the above evidence
synthesis is also provided by two supplemental studies in animals that did not meet PECO criteria
due to the route of exposure used (intratracheal instillation). Zhao etal. f20141 reported
statistically significant increases in relative lung weight and in albumin and total protein levels in
BALF isolated from male Sprague-Dawley rats exposed to 0.063 or 0.630 mg Cr(VI)/kg once per
week for four weeks via intratracheal instillation. These effects were concurrent with increases in
oxidative damage (8-OHdG lesions) and NF-kB, consistent with oxidative stress and inflammation.
In another study in rats exposed to 0.0035, 0.017, or 0.087 mg Cr(VI)/kg, 5x/week, or 0.017, 0.087,
or 0.44 mg/kg, lx/week via intratracheal instillation for 30 weeks, lungs of animals dosed with
This document is a draft for review purposes only and does not constitute Agency policy.
3-37 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
<0.087 mg/kg Cr(VI) contained macrophage foci, while in the high dose group, in addition to benign
and malignant tumors, severe damage and fibrosis to the bronchioloalveolar region of the lung was
observed, alongside inflammatory foci that included alveolar macrophages, epithelial cell
proliferation, and inflammatory thickening of the alveolar septa (Steinhoffetal.. 19861.
Studies investigating immune toxicity (Section 3.2.6) have observed changes in various
cytokine signaling in the blood, serum, and plasma of chromate workers exposed to Cr(VI) (Oian et
al.. 2013: Mignini etal.. 2009: Kuo and Wu. 20021 (Appendix Table C-38), although one study
specific to the lung in rats exposed via inhalation to 0.119 mg Cr(VI)/m3 for 5 h/d for 5 consecutive
days reported no detectable changes in several cytokines in BALF fCohen etal.. 20101. In human
lung cells in vitro, cytotoxicity was shown to correlate with a net loss of urokinase-type
plasminogen activator activity that has been shown to promote pulmonary fibrosis (Shumilla and
Barchowskv. 19991. as well as an inflammatory response via protein phosphorylation and cytokine
signaling fPascal and Tessier. 20041. Although the direction of these changes was not consistent
across studies, fluctuations in systemic cytokine levels and redox imbalance are characteristic of an
inflammatory response and may be indicative of a disruption in the regulatory balance that dictates
normal immune system function.
3.2.1.4. Integration of Evidence
Overall, the available evidence indicates that Cr(VI) likely causes lower respiratory tract
effects in humans. Cr(VI) is a known lung carcinogen, but the evidence for noncancer effects in the
respiratory tract (with the exception of nasal effects) is more sparse. This evidence integration
conclusion is based on observations of decreased lung function among chromium-exposed workers
in three of the five low confidence human studies and of biochemical effects indicative of lung injury
(albumin, LDH, and total protein in BALF) in medium confidence animal studies, supported by
supplemental and mechanistic observations consistent with an inflammatory tissue response
following Cr(VI) exposure. The exposure conditions relevant to these effects are further defined in
Section 4.2.
The development of the ATS guidelines in 1987 greatly increased the reliability of
spirometry measurements. These improvements to outcome measurement technology and
methods coincide with or came after changes to industrial processes aimed at reducing Cr(VI)
exposures in workers. Thus, while researchers were in a better position to reduce outcome
measurement error after the ATS guidelines become available, at the same time, the contrast in
exposures was reduced compared to previous decades, impacting study sensitivity. All five of the
included human studies thus had potential for decreased sensitivity due to lower exposure levels
attributed to industrial hygiene and process changes in more recent years. All five included human
studies were found to be low confidence, and three of these reported decreases in lung function in
chromate workers compared to referents (Zhang etal.. 2022: Li etal.. 2015b: Kuo etal.. 1997b).
Given the consistency of the findings from these three low confidence studies and biological
plausibility provided by supporting evidence for changes in inflammatory, oxidative stress, and
This document is a draft for review purposes only and does not constitute Agency policy.
3-38 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
cytotoxicity biomarkers in workers exposed to Cr(VI) (described under "Mechanistic Evidence"),
the human studies are interpreted to provide slight evidence for lower respiratory tract effects.
The pathogenesis of chronic pulmonary disease induced by chemicals toxic to the lung
involves the accumulation of inflammatory macrophages (Laskin et al.. 20191. In the available
animal studies, which together provide moderate evidence of lung inflammation, histopathological
changes in the lung following Cr(VI) exposure included histiocytosis (macrophage accumulation)
observed in four out of the five medium confidence animal studies. Infiltration of histiocytes was
also observed in multiple other organs following oral exposure in rodents (see a broader discussion
in Section 3.2.6, Immune Effects), which increases confidence that this inflammatory effect is a
result of Cr(VI) exposure. For inhalation exposure, histiocytosis was biologically significant
because it accompanied markers in bronchoalveolar lavage fluid (BALF), and increased leukocytes
in plasma (see Section 3.2.6), which are observations supportive of inflammatory lung responses
fNikula et al.. 20141. Cellular responses consistent with injury in the lung following Cr(VI) exposure
were also observed in animal studies, including increased albumin, total protein, and LDH activity
in BALF, biomarkers known to be evidence of injury and vascular leakage in the lower airway and
deep lung fKodavanti. 20141. Additionally, findings of increased lung weights in a single study of
Wistar rats (but not other strains or species examined in lower confidence studies) and clinical
findings in two rodent studies of obstructive respiratory dyspnea (Glaser etal.. 19901 and "peculiar
sound during respiration" and periodic nose bleeds (Kim etal.. 20041. are coherent with the
inflammatory changes consistently indicated in the available animal studies.
As described in Section 3.1, inhaled chromium can accumulate in high concentrations at
portal-of-entry tissues (such as the respiratory epithelium), resulting in absorption into the
epithelial cells in the lung and lung airways, and particles may accumulate in susceptible areas such
as airway bifurcation sites. Studies investigating the underlying mechanisms involved in Cr(VI)-
induced lung toxicity report significant cytotoxicity at micromolar concentrations in vitro,
concurrent with indications of an inflammatory response (oxidative stress, cytokine and nuclear
transcription factor activation) as well as increased programmed cell death (apoptosis, autophagy)
in response to Cr(VI) exposure. These data support the biological plausibility of the inflammatory
tissue responses observed in Cr(VI)-exposed animals. Although the available mechanistic studies in
humans were measuring systemic markers of oxidative stress and inflammation in the blood and
urine rather than specifically in the lung, consistent evidence of increased reactive oxygen species
generation and cytokine modulation in exposed workers is consistent with an inflammatory
response that contributes to health effects.
For lower respiratory tract effects, there were inconsistencies in the data that may be
explained by differences in study design and particle size. Large inhaled particles (with diameter
>5 |im) will deposit in the extrathoracic region, particles greater than 2.5 |im are generally
deposited in the tracheobronchial regions, and particles less than 2.5 [im are generally deposited in
the pulmonary region (OSHA. 20061. The rodent study of sodium dichromate aerosols by Glaser et
This document is a draft for review purposes only and does not constitute Agency policy.
3-39 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Toxicological Review ofHexavalent Chromium
al. (1990; 19851 likely induced effects in the lower respiratory tract due to the small particle sizes
achieved by the experiment (MMAD < 0.4 [im). For the human occupational studies, particle sizes
may have been larger and more variable (Kuo etal.. 1997a). causing a lower proportion of Cr(VI) to
deposit in the pulmonary region. However, human studies of occupationally exposed workers still
provide some evidence for pulmonary function deficits with increased Cr(VI) exposure. Animal and
human studies also differed with respect to the types of data collected, which precluded the ability
to directly compare effects. Human data were based on functional measures (pulmonary function
evaluated using spirometry), whereas animal data were based on histopathological measures and
cellular responses. The endpoints reported by studies in humans and animals were
complementary; overall the currently available evidence indicates that Cr(VI) is likely to cause
lower respiratory toxicity in humans.
This document is a draft for review purposes only and does not constitute Agency policy.
3-40 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-7. Evidence profile table for respiratory effects other than cancer
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that
increase certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
Evidence from studies of exposed humans
®©o
PULMONARY FUNCTION
Five low confidence studies
in occupationally exposed
adult workers:
Kuoetal. (1997b)
Li et al. (2015b)
Lindberg and Hedenstierna
(1983)
Sobaszek et al. (1998)
Zhang et al. (2022)
Exposure to Cr(VI) was
associated with decreased FVC
and FEV1.0 in three low
confidence studies (association
not statistically significant for
FVC in two of the three studies).
Exposure to Cr(VI) was
associated with decreased
FEV1/FVC in one of two low
confidence studies that
included that endpoint.
No association between Cr(VI)
and FVC, FEV1.0, or FEV1/FVC
was found in the two remaining
low confidence studies.
• Coherence of
observed
effects on
multiple
measures of
pulmonary
function (apical
studies)
• Exposure-
response
gradient
• Imprecision
of effect
estimates
• Low
confidence
studies
• Lack of
consistency,
though
partially
explained by
differences in
study
sensitivity
and exposure
levels
©oo
Slight
Based on
decreased
pulmonary
function with
higher exposure
to Cr(VI) in three
low confidence
studies.
The evidence indicates that
Cr(VI) inhalation is likely to cause
lower respiratory toxicity in
humans given sufficient exposure
conditions, based on moderate
evidence in rats showing
increases in biochemical
indicators of lung injury and
evidence of lung inflammation.
This is supported by slight human
evidence of decreased pulmonary
function from low confidence
studies of exposed workers and
supportive mechanistic evidence
for increases in oxidative stress
and cytotoxicity biomarkers.
The findings in animals are
consistent with known
biomarkers of human pulmonary
dysfunction and thus considered
Evidence from animal studies
relevant to humans.
The evidence is inadequate to
determine whether oral Cr(VI)
exposure might be capable of
causing noncancer respiratory
effects. No respiratory effects
were observed following
ingestion. As described in Section
3.1, Cr(VI) can expose portal-of-
LUNG CELLULAR and
BIOCHEMICAL RESPONSES,
including
HISTOPATHOLOGY
Six medium confidence
studies in rats and rabbits:
Kim et al. (2004)
Inflammatory changes in BALF
Increases in neutrophils/
granulocytes in two medium
confidence studies, and
increased lymphocytes up to 90
days in one medium confidence
study.
• Consistent
evidence of
some
inflammatory
changes in two
medium
• Indirect
biomarker
evidence of
lung injury is
less specific
®©o
Moderate
Coherent and
largely
consistent
increases in
This document is a draft for review purposes only and does not constitute Agency policy.
3-41 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Cohen et al. (2003)
Glaser et al. (1985)
Glaser et al. (1990)
Johansson et al. (1986a)
Johansson et al. (1986b)
One low confidence study
in mice:
Nettesheim et al. (1971)
Summary of key findings
Increased macrophages in two
medium confidence studies, but
no changes or slight decreases
in two others.
Macrophage Functional
changes
Increased phagocytosis in one
medium confidence study (at
concentrations <0.05 mg/m3),
but no change in another.
BALF Biochemistry
Increased protein, albumin and
LDH in one medium confidence
study.
Histiocytosis
Four of five medium confidence
studies reported the
accumulation of macrophages
in the lung by histopathology.
Other Histological Changes
Mixed evidence for bronchiolar
hyperplasia (one medium
confidence study); epithelial
hyperplasia, atrophy, and
necrosis (one low confidence
study); and normal
histopathology (one medium
confidence study).
Factors that
increase certainty
confidence
studies in two
rat strains
Coherence of
observed
effects across
different
biomarkers of
lung injury
Medium
confidence
studies
Concentration-
response
gradient for
most effects
Large effect
magnitude for
histopathologic
a I effects
Biological
plausibility
(mechanistic
evidence of
lung oxidative
stress and
apoptosis in
animal models,
primarily from
Factors that
decrease
certainty
than
pathology
Lack of
duration-
dependence
(some effects
weakened
with longer
exposures)
Some
unexplained
inconsistency
in findings for
macrophages
in BALF and
their
functional
changes
Unclear
adversity of
some
inflammatory
changes and
lack of
expected
coherence
with more
overt
histopatholog
Judgments and
rationale
biomarkers of
pulmonary injury
and
inflammatory
cells in BALF and
lung tissue, as
well as
mechanistic
findings
supportive of
inflammatory
changes in lung.
Inferences and summary
judgment
entry tissues, and reduction of
Cr(VI) in these tissues and red
blood cells decreases uptake by
other organ systems.
This document is a draft for review purposes only and does not constitute Agency policy.
3-42 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that
increase certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
instillation and
in vitro studies)
ical markers
of injury
LUNG WEIGHT
One medium confidence
study in rats:
Glaser et al. (1985)
Four low confidence studies
in rats and rabbits:
Glaser et al. (1986)
Kim et al. (2004)
Glaser et al. (1990)
Johansson et al. (1986a)
Lung Weight
Increased lung weights were
reported in the only medium
confidence study and one low
confidence study, both in
Wistar rats, with exposures for
up to 90 days and for 18
months; however, effects were
not observed in other low
confidence studies of male
rabbits exposed for 4-6 weeks
or male Sprague-Dawley rats
exposed for 90 days.
• Concentration-
response
gradient in two
studies
• Effect
magnitude (up
to 48%
increased
relative lung
weight)
• Coherence
with some
evidence of
increased
macrophages
(leading to
increased lung
weight)
• Some
inconsistency
across
studies,
although
inconsistent
studies were
low
confidence
©oo
Slight
Changes in lung
weight were
reported in one
rat strain but not
in low
confidence
studies of a
different strain
or in rabbits.
Mechanistic evidence
Biological events or
pathways
Summary of key findings and interpretations
Judgments and
rationale
Oxidative stress
Interpretation: Inhalation exposure to Cr(VI) induces a disruption of the
cellular redox balance in the lung that is a key component of Cr(VI)-
induced lung toxicity.
Key findings:
Biologically
plausible,
consistent, and
coherent
This document is a draft for review purposes only and does not constitute Agency policy.
3-43 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that
increase certainty
Factors that
decrease
certainty
Judgments and
rationale
• Consistent evidence of significant increases in oxidative stress in 17
studies of workers exposed to Cr(VI) that correlated with levels of
Cr(VI) in urine and blood
• Increased formation of 8-OHdG DNA adducts in one study of rats
exposed to Cr(VI) via inhalation
• In vitro evidence of oxidative stress with exposure to Cr(VI), including
increased ROS production, oxidation of lipids and proteins, and
increased antioxidant enzyme activity, in human primary and
immortalized lung cells
• Deficiency in DNA repair of 8-OHdG lesions led to increased
cytotoxicity and cell cycle delay following Cr(VI) exposure in vitro
observations of
oxidative stress,
leading to
cytotoxicity and
possibly
involving
inflammation,
which are
interrelated
processes
involved in
cellular stress
signaling that
can underlie the
respiratory
effects reported
in humans and in
animals exposed
to Cr(VI).
Fluctuations in
cytokine levels
and redox
imbalance are
characteristic of
an inflammatory
response and
may be
indicative of a
disruption in the
regulatory
balance that
dictates normal
immune system
function.
Cytotoxicity
Interpretation: Inhaled Cr(VI) is presumed to be cytotoxic to portal-of-
entry tissues; this toxicity, primarily shown by one study in animals and
multiple studies of human cells in vitro, may involve programmed cell
death in the lung.
Key findings:
• Increased apoptosis in the lung of rats exposed to Cr(VI) via
intratracheal instillation in one study
• Consistent in vitro evidence of dose- and time-dependent increases
in apoptosis following Cr(VI) exposure in human lung cells
• Some evidence of increased p53 (which can be pro-apoptotic) with
Cr(VI) exposure in humans or human lung cells in vitro
Inflammation
Interpretation: Inflammation induced by inhalation exposure to Cr(VI)
may involve pro-inflammatory cytokine signaling and enhanced ROS
generation.
Key findings:
This document is a draft for review purposes only and does not constitute Agency policy.
3-44 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that
increase certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
• Supplemental evidence of inflammatory cellular changes,
histopathology, and increased lung weight in Cr(VI) animal
intratracheal instillation studies support animal evidence judgments;
these effects were concurrent with increases in oxidative stress and
inflammatory cell signaling
• Cytokine signaling changes in chromate workers (Appendix C.2.5.2)
This document is a draft for review purposes only and does not constitute Agency policy.
3-45 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
3.2.2. Gastrointestinal Tract Effects Other Than Cancer
Studies of the GI tract following ingestion of Cr(VI) in humans and animals have generally
reported an increased incidence in nonneoplastic lesions in the stomach and portions of the small
intestine. The GI tract is responsible for the digestion, absorption, and excretion of ingested
substances. The main function of the stomach is storage and digestion; it is lined with epithelial
cells with tight junctions that lack the absorptive villi found in the intestines. In the small intestine,
the villi in the semipermeable mucosa consist of epithelial cells characterized by a brush border of
microvilli that further increase absorptive capacity. Between the villi are deep cavities called
crypts. Both crypts and villi contain epithelial enterocytes and goblet cells that secrete mucus. A
schematic of the epithelial morphologies of the stomach and small intestine is provided in
Section 3.1.1 Pharmacokinetics, Figure 3-5. While the small intestine has a large absorptive
capacity it also serves as a barrier (e.g., by mucus secretion) that prevents potentially toxic
substances in the lumen, including bacteria, from entering systemic circulation. The crypts in the
small intestine supply rapidly dividing stem cells for the renewal of the intestinal epithelium, which
turns over within days (Potten etal.. 2009: Potten etal.. 19971. Within the stomach, gastric stem
cells are located within glandular pits, and unlike the small intestine, they are nearer to the lumen
and more likely to be exposed to surface irritants (Mills and Shivdasani. 20111. In animal studies,
the areas of the small intestine that are closer to the stomach (the duodenum and jejunum) appear
to be more susceptible to injury than the ileum.
3.2.2.1. Human Evidence
The literature search for this assessment did not identify epidemiological studies that met
PECO criteria for this health effect The ATSDR Toxicological Profile (ATSDR. 20121 describes
multiple case reports of deaths among adults and children resulting from ingesting Cr(VI)
compounds and subsequent damage to the GI tract and other organs. GI effects reported in acute
oral poisoning studies identified in the literature search for this assessment include stomach and
esophageal pain, diarrhea, lesions of the stomach and duodenum, hemorrhage of the GI tract, and
gut mucosal necrosis (Goulle etal.. 2012: Baresic etal.. 2009: Hantson etal.. 2005: Kolacinski etal..
2004: Sharma etal.. 2003: Stiftetal.. 2000: Kolacinski etal.. 1999: Loubieres etal.. 1999: Stiftetal..
1998: Kurosaki etal.. 1995: van Heerden et al.. 19941. The ATSDR Toxicological Profile (ATSDR.
20121 also describes reports of stomach pain, GI ulcer, and gastritis among workers employed in
electroplating and chromate production from studies published from 1950-1978. The exposures
could have occurred via both inhalation and ingestion of Cr(VI) dusts in the workplace. ATSDR
concluded that these studies included no or inappropriate comparison groups and therefore a
direct association between Cr(VI) exposure and these signs and symptoms could not be drawn.
This document is a draft for review purposes only and does not constitute Agency policy.
3-46 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofHexavalent Chromium
3.2.2.2. Animal Evidence
Study evaluation summary
Table 3-8 summarizes the four animal bioassays that were considered in the evaluation of
noncancer effects in the GI tract from ingested Cr(VI). The studies, conducted by two organizations,
the US National Toxicology Program (NTP) fNTP. 2008. 20071 and ToxStrategies, Inc. fThompson et
al.. 2012b: Thompson etal.. 20111. exposed mice and rats of both sexes to Cr(VI) in drinking water,
and were of subchronic duration except for the NTP f20081 2-year bioassay. Results in all studies
were limited to histopathological observations and mechanistic evidence; the latter is also
described with the evidence for GI tract cancer in Section 3.2.3.2.
Table 3-8. Summary of included studies for Cr(VI) GI histopathological
outcomes and overall confidence classification. Click to see interactive data
graphic for rating rationales.
>
M
o
o
-C
+¦»
(0
Q.
O
"K
Author (year)
Species (strain)
Exposure design
Exposure route
X
NTP(2007)
Rat (F344/N), male and female;
Mouse (B6C3F1, BALB/c, C57BL/6),
male and female
Subchronic
Drinking water
H
NTP(2008)
Rat (F344/N), male and female;
Mouse (B6C3F1), male and female
Chronic
Drinking water
H
Thompson et al. (2011)
Mouse (B6C3F1), female
Subchronic
Drinking water
H
Thompson et al. (2012b)
Rat (F344), female
Subchronic
Drinking water
H
High (H), medium (M), low (L), or uninformative (U).
Synthesis of evidence in animals
All four high confidence studies in rats and mice reported various histological effects in the
GI tract associated with oral exposure to Cr(VI). In the small intestine these included diffuse
epithelial/crypt cell hyperplasia, histiocytic cellular infiltration, and degenerative changes in the
villi (vacuolization, atrophy, and apoptosis); in the glandular stomach these included squamous
metaplasia and gastric ulceration (Thompson etal.. 2012b: Thompson etal.. 2011: NTP. 2008.
20071. Across studies, the most commonly observed nonneoplastic GI lesion was epithelial cell
hyperplasia in the mouse small intestine (Thompson et al.. 2012b: Thompson et al.. 2011: NTP.
2008. 20071. Results from studies in mice and rats are summarized in Figures 3-12 and 3-13, and
study design differences are outlined in Table 3-9 (detailed results are summarized in Appendix
Table C-32). Dose-dependent histiocytic infiltration, described by NTP (20081 as being of unknown
This document is a draft for review purposes only and does not constitute Agency policy.
3-47 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
biological significance, was also observed in the small intestine of exposed animals across studies,
sexes, and species.
Table 3-9. Design features of studies that examined Gl tract effects via the oral
route of exposure
Study reference
Species/st rain
and sex
Exposure
duration
Number of
animals/group
Dose groups (mg Cr(VI)/kg-d)
NTP (2008)a
B6C3F1 mouse,
male and female
2 years
50
0, 0.450, 0.914, 2.40, 5.70 (M)
0, 0.302, 1.18,3.24, 8.89 (F)
NTP (2008)
F344 Rat, male
and female
2 years
50
0, 0.200, 0.796, 2.10, 6.07 (M)
0, 0.248, 0.961, 2.60, 7.13 (F)
NTP(2007)
F344 Rat, male
and female
90 days
10
0, 1.74, 3.14, 5.93, 11.2, 20.9 (M)
0, 1.74, 3.49, 6.28, 11.5, 21.3 (F)a
NTP(2007)
B6C3F1 mouse,
male and female
90 days
10
0, 3.1, 5.3, 9.1, 15.7, 27.9 (M+F)
NTP (2007)
B6C3F1 mouse,
male
90 days
5
0, 2.8, 5.2, 8.7
NTP(2007)
BALB/c mouse,
male
90 days
5
0, 2.8, 5,2, 8.7
NTP(2007)
am-C57BL/6
mouse, male
90 days
5
0, 2.8, 5.2, 8.7
Thompson et al.
(2012b)
F344 Rat, female
7 days
5
0, 0.015, 0.21, 2.9, 7.2, 20.5
90 days
10
Thompson et al. (2011)
B6C3F1 mouse,
female
7 days
5
0, 0.024, 0.32, 1.1, 4.6, 11.6, 31.1
90 days
10
aNote: In the synthesis, male and female doses were rounded to the same values for simplicity.
study identifier
animal description
Endpolnl
Observation
NTP (2007)
Mouse, BALB/c (cf>
Epidreltal hyperplasia of duodenum
days
Epithelial hyperplasia of jejunum
Wdny*
Mouse, BIiC3F;I (9)
Epithelial hyperplasia of duodenum
W0 days
Epithelial hypoplasia of jejunum
90 day*
Mouse. B6CJPI icf >
Epithelial hyperplasia of diuxJenum
'JO days
W days
Epithelial hyperplasia of jejunum
W days
Wdays
Mouse. C57BL/6 «f)
Epithelial hyperplasia of duodenum
M0 days
Epithelial hyperplasia of jejunum
90 days
a al. (2011)
Mouse, RGC3FI (9)
Crypi cell hyperplasia nf the duodenum
40 day*
Crypt cell hyperplasia erf (lie jejunum
40 days
NTP (2U08)
Mouse, B6C3P! (9)
Di ft use epithelial hypcn>Lasla of the duodenum
2 years
Dil l use epithelial hyperplasia of ihe jejunum
2 years
Mouse. B6C3FI frf i
Diffuse epithelial hyperplasia of ihe duodenum
2ycat>
Diffuse epithelial hyperplasia of (he jejunum
2 years
Hyperplasia in the small inlcstinr
AA A
r-A A
AA A-
AA A
Ł no chanjte
A tigililiciUK
I—i rime Range
20
mg/l;g-day
Figure 3-12. Diffuse epithelial hyperplasia in Cr(VI) treated mice in high
confidence studies. Note: NTP (2008, 20071 did not present quantitative no-effect
This document is a draft for review purposes only and does not constitute Agency policy,
3-48' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Toxicologiccil Review of Hexavalent Chromium
data. However, the dose levels and ranges for the exposure groups without effects
are displayed here for comparative purposes. Click to see interactive data graphic.
study Identifier
NTP 120071
animal description
Rju, K344/N f 9 J
Ral, F344/N ( (f)
Thompson ci nl. (2012) Rat. P344/f< 19 >
End point
Epithelial hyperplasia of duodenum
FpidwHal lnyncrpUtM.! of jejunum
EpiUji-Ual hyperplasia uf duodenum
Epithelial hyperplasia of jejunum
Crypt cell hyperplasia of the duodenum
Ci > pi cell hypoplasia of the jejunum
Ohstrvallon time
90 days
W day n
lJO days
90 days
90 days
90day>
Hyperplasia in tllr small intirslfnr
^ mo change
A significant increase
|—| Dwsf Range
Rat. F344/N 19 J Diffuse epithelial hyperplasia of the dm ilenum 2 years
Diffuse epithelial hyperplasia of the Jejunum 2 years
Kai. F344/N (cf > Diffuse epithelial hypeqilasia of (he dmiden urn 2 years
Diffuse epithelial hyperplasia of Ihe jejunum 2 years
mg/tff-day
Figure 3-13. Diffuse epithelial hyperplasia in Cr(VI) treated rats in high
confidence studies. Note: NTP f2008. 20071 did not present quantitative no-effect
data. However, the dose levels and ranges for the exposure groups without effects
are displayed here for comparative purposes. Click to see interactive data graphic.
In sttbchronically exposed B6C3F1 mice, statistically significant elevated incidences of
minimal to mild26 diffuse duodenal epithelial cell hyperplasia were observed in both males and
females at all doses (>3 mg Cr(VI)/kg-d, incidence increasing with dose) (NTP. 20071. In a
companion subchronic strain comparison study, statistically significant increases in the incidence
of diffuse epithelial hyperplasia in the duodenum were also observed across all three strains of
male mice tested (i.e., B6C3F1. BALE/c. and am3C57BL/61 fNTP. 20071. A separate subchronic
study also showed a significant increase in duodenal hyperplasia in B6C3F1 mice at doses >11.6 mg
Cr(VI)/kg-d fThompson et al.. 20111. This study did not show increasing incidence with dose, but
the lowest dose level at which the epithelial hyperplasia was observed in Thompson etal. f20111
(~12 mg Cr(VI)/kg-d) was about 4x higher than for NTP (20071 (~3 mg Cr(VI)/kg-d), and resulted
in dose-dependent apoptosis (which was statistically significant at the highest dose of 31.1 mg
Cr(VI)/kg-d), which likely degenerated the duodenal tissue. The subchronic results of hyperplasia
in the duodenum were consistent with a 2-year study that showed statistically significant elevated
incidences of minimal to mild diffuse epithelial cell hyperplasia in the duodenum of the same
severity but at lower doses (>0.3 mg Cr(VI)/kg-d, incidence increasing with dose with the exception
of the high dose males that had a slightly lower incidence than the second highest dose group)
(NTP. 2008). In the jejunum, there were no significantly elevated increases in epithelial cell
hyperplasia in either sex of B6C3F1 mice in a subchronic study at doses up to 28 mg Cr(VI)/kg-d
(NTP. 2007). but in a second subchronic study, female mice of the same strain showed statistically
significant elevated epithelial cell hyperplasia in the jejunum at doses >11,6 mg Cr(VI)/kg-d
fThompson et al.. 20111. In the 2-year mouse study, this effect was observed in the jejunum of
26According to NTP severity grading: l=minimal, 2=mild, 3=moderate, 4-marked.
This document is a draft for review purposes only and does not constitute Agency policy.
3-49' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
female mice at the highest dose (8.89 mg Cr(VI)/kg-d) fNTP. 20081. Together, these results show a
consistent pattern of minimal to mild diffuse epithelial hyperplasia in mice, which was present in
subchronic studies at higher doses compared to the chronic study.
In subchronic and chronic NTP studies in F344 rats, increased diffuse epithelial hyperplasia
was not observed in the small intestine (NTP. 2008. 20071. In contrast, a statistically significant
increase in these lesions was observed following >7.2 mg Cr(VI)/kg-d exposures for 7 and 90 days
in female F344 rats in a study by a separate group (Thompson etal.. 2012b). The differences in the
presence or absence of these lesions in F344 rats across studies is unknown, but this may have
been affected by differences in water intake between the two study groups, leading to higher
exposures to the rats in the the Thompson etal. (2012b) study. At the administered Cr(VI)
concentrations, which were nearly equivalent between the studies, the mg/kg-d doses in the NTP
subchronic bioassay fNTP. 20071 and the time weighted average doses from weeks 1-13 in the NTP
chronic bioassay (NTP. 20081 were approximately twofold lower than the mg/kg-d doses in
Thompson etal. (2012bll. In addition, Thompson etal. (2012b) noted that the animal vendor
sources for the F344 rats were different between groups (NTP used animals from Taconic Farms,
Inc. (NTP. 2008. 20071 and Thompson etal. (2012b) used animals from Charles River Laboratories
International, Inc.), although the mice used by each group were also procured from these two
different sources, respectively.
In the rat glandular stomach, there were also observations of epithelial hyperplasia along
with several other lesion types in a subchronic but not chronic studies. These lesions were not
observed in a subchronic study of three different strains of mice, nor in a chronic mouse study.
Statistically significant increased incidences of epithelial hyperplasia, squamous metaplasia, and
ulcers in the glandular stomach were reported in male and female F344 rats exposed to 21 mg
Cr(VI)/kg-d (the highest dose) in the 13-week NTP study (NTP. 2007). No statistically significant
increased incidences of glandular stomach or forestomach lesions were reported in the 2-year
studies of F344 rats and B6C3F1 mice fNTP. 20081. or in the NTP f20071 13-week studies of
B6C3F1, BALB/c, or am3-C 57BL/6 mice. Neither of the Thompson et al. (2012b: 2011) 13-week
studies conducted histologic examinations of the forestomach or glandular stomach of mice or rats.
The inconsistency between subchronic and chronic study results in rats is likely attributable to
dose selection; in the 13-week study, stomach lesions occurred at an exposure that was threefold
higher than the highest dose administered in the 2-year chronic assay.
Degenerative changes to the cells lining the GI tract can manifest as necrosis, apoptosis, and
subsequent villous stunting, resulting in crypt abscess and ulceration fBetton. 20131. The NTP
subchronic bioassay reported that the duodenal villi of B6C3F1 mice were short, thick, and blunted,
with cytoplasmic vacuolization in the epithelial cells lining the villi tips at doses up to 27.9 mg
Cr(VI)/kg-d (results were not presented quantitatively) (NTP. 2007). Consistent with these results,
the NTP 2-year bioassay qualitatively reported degenerative effects in mouse duodenal villi
(described as short, broad, and blunt) at doses up 8.89 mg Cr(VI)/kg-d. These effects were not
This document is a draft for review purposes only and does not constitute Agency policy.
3-50 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
reported in F344 rats at doses up to 21 or 7.13 mg Cr(VI)/kg-d after subchronic exposure or
chronic exposure respectively fNTP. 2008. 20071. GI tissue atrophy and apoptosis were not
reported in the NTP bioassays in either species fNTP. 2008. 20071. Although cytoplasmic
vacuolization, when irreversible, can be considered a marker of cell death due to cytoprotective
autophagy in response to stress (Shubin etal.. 2016). the vacuolization observed in epithelial cells
at the tips of villi in mice in the subchronic study was not interpreted by NTP to be indicative of
atrophy or apoptosis and was not observed in the 2-year bioassay fNTP. 2008. 20071. There was an
increased incidence of minimal to mild salivary gland atrophy in female rats after two years at the
two highest doses (the effect at the highest dose lacked statistical significance), although this effect
is of unknown biological significance.
Thompson etal. (2011) reported degenerative changes in the intestines of female B6C3F1
mice after subchronic exposure including statistically significant atrophy in villi of the duodena and
jejuna (31.1 mg Cr(VI)/kg-d, highest dose), apoptosis in the duodenal villi (31.1 mg Cr(VI)/kg-d),
and cytoplasmic vacuolization in the duodena and jejuna (>4.6 mg Cr(VI)/kg-d) (Figure 3-14).
These results are generally consistent with the descriptive observations reported by NTP in mice
after subchronic and chronic exposure. While the subchronic NTP study did not report identical
histopathological findings, it stated that "the epithelial cells lining the tips of the villi of many of the
exposed mice were swollen and had vacuolated cytoplasm. Collectively, these duodenal lesions
suggest regenerative hyperplasia secondary to previous epithelial cell damage or degeneration"
fNTP. 20071. The subchronic study in female F344 rats by Thompson etal. f2012bl also reported
apoptosis of the duodenal villi at the two highest doses (7.2 and 20.5 mg Cr(VI)/kg-d), but no
atrophy or vacuolization (Figure 3-14).
Two follow-up publications using the same experimental subchronic dataset in female
B6C3F1 mice (Thompson et al.. 2011) reported increases in some markers of duodenal villus
cytotoxicity described as karyorrhectic nuclei, desquamation, villous blunting, and disruption of
cellular architecture in the duodenal villi at doses >4.6 mg Cr(VI)/kg-day (Thompson etal.. 2015a:
O'Brien etal.. 20131. It should be noted that O'Brien etal. f20131 only evaluated one animal in the
next-lowest dose group (1.1 mg Cr (VI)/kg-day) for desquamation and disruption of cellular
arrangement In the crypt compartment, although increases in crypt length, area, and number of
crypt enterocytes were reported, there were no statistically significant or dose-responsive changes
in mitotic or apoptotic indices (Thompson et al.. 2015a: O'Brien etal.. 20131. Observations after 7-
day exposures reported by this group (considered supporting evidence due to the short duration)
include duodenal hyperplasia, villous atrophy, and cytoplasmic vacuolization, but again with no
changes in crypt apoptosis indices, mitotic activity, or increases in karyorrhectic nuclei in the crypt
compartment (Thompson et al.. 2015b: Thompson etal.. 2011). The authors attribute this
discrepancy to either the 24-hour period without Cr(VI) exposure prior to sacrifice and/or to the
sudden increase in the number of crypt enterocytes that then migrated toward the villus and
This document is a draft for review purposes only and does not constitute Agency policy.
3-51 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
became post-mitotic in that 24-hour period, apparently as mitotic figures were being measured
fThompson et al.. 2015a: O'Brien etal.. 20131.
While NTP (20081 noted short, broad, and blunt duodenal villi in mice, they did not report
observing duodenal villus atrophy. In a second review of the NTP 2-year bioassay mouse
histopathology slides by Cullen etal. (20151. these authors reported villus atrophy and blunting in
all mice in the highest dose group. Cullen etal. (20151 also only observed cytoplasmic vacuolization
in males; NTP made a general statement that vacuolization was observed in the tips of the villi
without presenting incidence or details. While there were some descriptive reporting differences
across studies for nonneoplastic histopathological lesions, an independent expert pathology review
(Francke and Mo g. 2021) of the diagnostic criteria used by these reports (Cullen etal.. 2015:
Thompson etal.. 2015a: NTP. 2008. 20071 confirmed there was no meaningful difference or
improvement when comparing the five histological diagnoses applied by this second review (ACC.
2015: Cullen etal.. 20151 to those used by NTP. In fact, NTP addressed four of the five diagnostic
terms used by Cullen etal. (20151 (i.e., histiocytic cellular infiltrates, atrophy/blunting, enterocyte
vacuolation, and epithelial hyperplasia), with the exception of single-cell necrosis (i.e., apoptosis).
Thus, the "short, broad, blunt" duodenal villi of exposed mice reported by NTP f20081 are
analogous to the Cullen etal. (2015) report of "atrophy/blunting" of the villus.
Study name animal description
Thompson etal. 2011 Mouse, B6C3F1 (9)
Thompson etal. 2012 Rat, F344/N(?
Endpoint
Villous cytoplasmic vacuolization of the jejunum
Villous cytoplasmic vacuolization of the duodenum
Vllous atrophy of the jejunum
Vllous atrophy of the duodenum
Apoptosis of the duodenum
Vllous cytoplasmic vacuolization of the jejunum
Vllous cytoplasmic vacuolization of the duodenum
Vllous atrophy of the jejunum
Vllous atrophy of the duodenum
Apoptosis of the duodenum
Effects in the small intestine
—~
—*
—+ O no change
—~ A significant increase
-jK |—| Dose Range
20
mg/kg-day
25
30
35
Figure 3-14. Cr(VI)-induced degenerative changes in the small intestines of
mice and rats in high confidence studies. Click here to see interactive graphic.
Increased infiltration of histiocytes (macrophage immune cells) in the duodenum and
jejunum was consistently observed in both sexes of rats and mice orally exposed both chronically
and subchronically to Cr(VI) (Thompson etal.. 2012b: Thompson etal.. 2011: NTP. 2008. 20071.
NTP (2008) indicated that the biological significance of the histiocytic infiltration is not known, but
surmised that the infiltration of macrophages may reflect phagocytosis of an insoluble chromium
precipitate. It should be noted that while macrophage accumulation may be associated with
inflammation, NTP did not report chronic inflammation in the GI tract, or the influx of other
inflammatory cells associated with the histiocytic infiltration in the small intestine fNTP. 2008.
20071.
This document is a draft for review purposes only and does not constitute Agency policy.
3-52 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
In summary, diffuse epithelial hyperplasia of the small intestine was consistently observed
in the three high confidence studies in mice, occurring at higher doses in the subchronic studies
compared to the chronic study, with similar severity across studies. Diffuse epithelial hyperplasia
was also observed in the rat small intestine, but these findings were inconsistent between the two
reporting groups. Similar degenerative changes in the duodenal villi were consistently observed
across studies, and although the description of these effects varied, the results were essentially the
same. Histiocytic infiltration was also consistently observed, though this effect was interpreted by
the report authors to be of unknown biological significance fNTP. 20081 and is likely not adverse on
its own.
3.2.2.3. Mechanistic Evidence
The screening and identification of mechanistic studies for evidence relevant to Cr(VI)-
induced oxidative stress, cell proliferation and cell death in the GI tract prioritized both oral
exposure studies in animals and studies via all routes in animals if results were presented for GI
tissues, as well as in vitro studies in human cells derived from GI tissues (primary and
immortalized); this prioritization strategy and a summary of the studies can be found in Appendix
C.2.2.2. No human oral exposure studies or human studies of cytotoxicity or cell proliferation
specific to the GI tract were identified. Because mechanistic evidence from studies of non-
malignant toxic effects specific to the GI tract (in vivo or in vitro) following the ingestion of Cr(VI) is
also relevant to cancer of the GI tract, a summary of this evidence is presented in Section 3.2.3.3.
The evidence supports a consistent, coherent, and biologically plausible role for oxidative stress,
cytotoxicity, and cell proliferation induced by Cr(VI) exposure in both the nonneoplastic toxicity
and carcinogenic effects of Cr(VI) in the GI tract.
Three in vivo studies were identified that reported biomarkers of oxidative stress in GI
tissues after oral exposure (Thompson etal.. 2012b: Thompson etal.. 2011: De Flora et al.. 2008).
In addition, a gavage study (Sengupta et al.. 1990) reported various oxidative stress parameters in
GI tissue after administration of potassium dichromate at doses of 1500 mg/kg for three days and
300 mg/kg for 30 days. However, the inclusion of doses that are higher than the LD50 (130 mg/kg)
for rats fThermo Fisher. 20091 is considered a limitation for interpreting the results of this study.
In female B6C3F1 mouse GI tract tissues, the reduced-to-oxidized glutathione ratio
(GSH/GSSG), which is considered a biomarker of redox status, showed statistically significant, dose-
dependent decreases in the oral and duodenal epithelium in mice exposed to Cr(VI) in drinking
water (>11.6 mg Cr(VI)/kg-d and >4.6 mg Cr(VI)/kg-d, respectively) after 7 days of exposure,
indicating an increase in oxidative stress, with no correlated change in the GSH/GSSG ratio in
plasma fThompson etal.. 20111. After 90 days, there was still a significant decrease in the
GSH/GSSG ratio in the small intestinal epithelia of the duodenum (up to a 38.5% decrease at the top
dose) and jejunum (up to a 52% decrease at the top dose), but not in the ileum, at concentrations
>1.1 mg Cr(VI)/kg-d and decreases in plasma at higher concentrations (>11.6 mg Cr(VI)/kg-d), but
no decreases were detected in the oral mucosa despite a measurable chromium concentration in
This document is a draft for review purposes only and does not constitute Agency policy.
3-53 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
these tissues. While GSH/GSSG ratio measurement is a generally accepted indicator of oxidative
stress, ascorbate is the preferred in vivo reductant accounting for 90% of Cr(VI) oxidative
metabolism (described in detail in Section 3.1.1.). Although the expected primary oxidative
pathway is not captured in these experiments, the decreased GSH/GSSG ratio with increasing dose
implies some level of Cr(VI)-induced oxidative stress was occurring in the duodenum. However,
protein carbonyls, an indicator of protein oxidation, were only slightly elevated in the duodenum
after 90 days (Thompson etal.. 20111. possibly indicating that the ROS mediated damage is being
preferentially directed at nucleic acids rather than proteins, although the reason for this preference
is not known.
This study also did not observe increases in 8-OHdG DNA adducts in the oral cavity or
duodenal tissue of mice (Thompson etal.. 20111. The absence of oxidative ly induced 8-OHdG
adducts in mouse GI tissues is consistent with a study by De Flora et al. (20081. which found no
increase in these lesions in the forestomach, glandular stomach, or duodenum after female SKH-1
mice were exposed for 9 months via drinking water at concentrations of 1.20 and 4.82 mg
Cr(VI)/kg-d. The reason for the lack of oxidative DNA lesions associated with the oxidative stress in
these studies is not known.
In female F344/N rats, Thompson etal. (2012b) reported no statistically significant changes
in GSH/GSSG ratios in either the oral cavity or the small intestine of female rats after 7 days of
Cr(VI) exposure to concentrations 0.1-180 mg/L Cr(VI), with the exception of decreases in the
jejunum at the high concentration of 180 mg/L Cr(VI) and a decrease at 0.1 mg/L Cr(VI) in the oral
mucosa. After 90 days, statistically significant and dose-dependent reductions in the GSH/GSSG
ratio in the oral mucosa and jejunum were observed at concentrations >20 mg/L Cr(VI) (Thompson
etal.. 2012b). These results are in partial contrast to experiments in mice from the same research
group (described above), which showed decreases in GSH/GSSG ratio in the duodenum but not the
oral mucosa at 90 days despite mice having measurable total chromium concentrations in the oral
cavity fThompson etal.. 20111. The plasma GSH/GSSG ratio was also decreased at concentrations
>60 mg/L Cr(VI). No changes in the GSH/GSSG ratio were observed in the duodenum at 90 days,
and there were no changes in 8-isoprostane, a marker of lipid peroxidation, in the oral mucosa or
duodenum.
Although in vitro exposures may lead to exaggerated cell stress and oxidative responses,
limiting their ability to predict physiological conditions in vivo, these studies can provide
supplemental evidence indicating the potential contribution of oxidative stress and the signaling
pathways involved. Evidence from cells exposed in vitro consistently demonstrates increased
oxidative damage induced by Cr(VI), where ROS levels, lipid and protein oxidation, and decreased
levels of antioxidant enzymes correlate with DNA damage that is increased in test systems with
disabled DNA excision repair processes or abrogated with antioxidant pretreatment (Appendix
Table C-57). This includes studies performed with human colon and gastric cancer cell lines to
study oxidatively induced DNA damage and cytotoxicity. In vitro, it appears that Cr(VI) exposure
This document is a draft for review purposes only and does not constitute Agency policy.
3-54 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
can result in oxidative stress with minimal or no cytotoxicity, as shown in human colorectal
adenocarcinoma Caco-2 cells fThompson etal.. 2012al. Thompson etal. f2012al measured both
8-OHdG adducts and levels of phosphorylated histone H2AX (yH2AX), a marker of DNA double-
strand breaks that could arise from various sources including ROS and/or direct chemical
interactions. After 24 hours, cytotoxic concentrations of Cr(VI) increased 8-OHdG and yH2AX
levels, while non-cytotoxic concentrations only elevated 8-OHdG, suggesting that oxidative stress
could be a mechanism for DNA damage other than double-strand breaks at lower concentrations in
in vitro test systems. Notably, these results conflict with the in vivo study results following
subchronic Cr(VI) exposure in drinking water presented above, which consistently showed no
changes in 8-OHdG.
In the same study, Thompson etal. (2012a) reported that differentiated Caco-2 cells were
more resistant to cytotoxicity than undifferentiated cells. There were no reported changes in
immunofluorescence staining of differentiated Caco-2 cells for p53 or annexin-V (apoptosis
markers) or LCB3 (an autophagy indicator). There was, however, a dose-dependent translocation
of ATF6 to the nucleus in differentiated cells, which is an indicator of endoplasmic reticulum stress
and supports in vivo toxicogenomic data indicating this response in duodenal tissue fKopec etal..
2012b: Thompson et al.. 2012a). A study by a separate group with the human gastric cancer cell
line SGC-7901 showed that Cr(VI) treatment in cells modified by knockdown of URI (a transcription
factor and oncogene) enhanced ROS production and cell death compared to control cells treated
with Cr(VI) fLuo etal.. 20161. This suggests URI may have a role in suppressing Cr (VI)-induced
oxidative stress and apoptosis.
Tissue injury induced by cytotoxicity and oxidative stress in the GI tract may lead to
necrosis and/or regenerative proliferation, evidenced by the histological degenerative changes in
the small intestinal villi of mice exposed to Cr(VI) up to 2 years, as well as in the small intestine and
glandular stomach of rats exposed for 3 months. While ultimately only mice developed intestinal
tumors, the observations of hyperplasia, metaplasia, and ulcer in the stomach and villous wounding
in the intestine of rats are similarly demonstrative that Cr(VI) may cause GI toxicity through tissue
injury. As described in the synthesis of animal evidence, observations indicative of degenerative
changes in the mouse small intestine were reported across studies and suggest a regenerative
response to epithelial cell injury fThompson etal.. 2011: NTP. 2008. 20071. These Cr(VI)-specific
effects in the small intestine are supported by X-ray fluorescence data showing ingested Cr
concentrates in the duodenal villi of mice (Thompson et al.. 2015b: Thompson etal.. 2015a: O'Brien
etal.. 2013: Thompson etal.. 2011). In the duodenum, diffuse hyperplasia was observed at all
doses after both subchronic (>3 mg Cr(VI)/kg-d) and chronic (>0.3 mg Cr(VI)/kg-d) exposure, and
focal hyperplasia was observed after chronic exposure at doses >2.4 mg Cr(VI)/kg-day.
Tissue injury in the mouse duodenal villi may lead to a compensatory proliferative response
in the crypt compartment and hyperplasia observed in the intestinal mucosa as observed by dose-
dependent crypt enterocyte proliferation (Thompson et al.. 2015b: O'Brien etal.. 2013). although
This document is a draft for review purposes only and does not constitute Agency policy.
3-55 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
the relationship between this measure of increased cell proliferation after a 7-day exposure and the
observations of villous hyperplasia after 3 months or 2 years of exposure are unclear. These
investigators observed increased numbers of crypt enterocytes but did not detect a treatment-
related increase in mitotic indices in these crypts, which would appear to be inconsistent with
regenerative crypt hyperplasia (Thompson et al.. 2015b: O'Brien etal.. 20131.
Perturbations in cell signaling pathways that enhance cellular proliferation may contribute
to the hyperplastic effects observed in the small intestine of B6C3F1 mice. Gene expression
profiling studies of the tissues collected in the subchronic drinking water exposure study by
Thompson etal. f20111 found that Ki-67 expression, a protein associated with cell proliferation
used to label proliferative intestinal crypt compartment cells (Li etal.. 2015a: Basak etal.. 20141.
was increased within the duodenal mucosa in mice at the two highest doses (11.6 and 31 mg/kg-d
Cr(VI)) by day 91 (with dose-dependent increases at >4.6 mg/kg-day Cr(VI) at day 8) (Rager etal..
2017: Kopec etal.. 2012al. A separate group reported that after 60 days of exposure to Cr(VI) in
drinking water, the c-Myc oncogene showed a dose-dependent increase in the stomach (gene
expression and protein levels >3.5 mg/kg-day Cr(VI)) and colon (gene expression >1.7 mg/kg-day
and protein levels >5.2 mg/kg-day Cr(VI)) of male Wistar rats fTsao etal.. 20111. consistent with
the promotion of cell cycle progression and cell proliferation. The same study also reported a
decrease in the expression of RKIP (Raf kinase inhibitor protein; >5.2 mg/kg-day Cr(VI)), which is
thought to negatively regulate MAPK (mitogen activated protein kinase) signaling involved in
cellular proliferation fVandamme etal.. 20141. The gene expression and protein levels of tumor
suppressor and cell cycle regulator p53 were also downregulated in the stomach (gene expression
>3.5 mg/kg-day and protein levels >1.7 mg/kg-day Cr(VI)) and colon (gene expression and protein
levels >5.2 mg/kg-day Cr(VI)) fTsao etal.. 20111. Consistent with these studies, toxicogenomic
analyses of GI tissues in Cr(VI)-treated animals have identified differentially expressed genes
(DEGs) associated with activation of c-Myc, MAPK, and a variety of additional pathways associated
with cell cycle, proliferation, and apoptosis. A summary of gene expression changes and
toxicogenomic results most pertinent to both noncancer and cancer GI effects can be found in
Appendix C.3.3 and C.3.4, respectively, and is discussed in the context of cancer MOA in Section
3.2.3.
Although the molecular pathways leading to the cytotoxic effects of Cr(VI) in the GI tract
following oral exposures are not clear, it is likely to involve chronic oxidative stress known to occur
across multiple tissues following Cr(VI) exposures (see Section 3.2.3.3), though there are also
indications of oxidative stress occurring in the absence of cytotoxicity. The data from studies of
Cr(VI) provide consistent support for oxidative stress as a mechanism of Cr(VI) toxicity in the lung
(Section 3.2.1), liver (Section 3.2.4), male and female reproductive organs (Sections 3.2.7 and 3.2.8,
respectively), and fetal development (Section 3.2.9), though in vivo results specific to the GI tract
are mixed (Thompson etal.. 20131. Proliferative cell signaling pathways show upregulation in the
GI tract that is generally consistent with the pathological evidence of tissue regeneration in the
This document is a draft for review purposes only and does not constitute Agency policy.
3-56 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
mouse small intestine, though it cannot be conclusively determined whether these dose-dependent
gene expression and protein level changes are associated with compensatory cell proliferation
following cytotoxicity or are induced by Cr(VI) exposure via another pathway.
3.2.2.4. Integration of Evidence
Overall, the currently available evidence indicates that oral exposure to Cr(VI) likely
causes GI tract toxicity in humans. This evidence is summarized in Table 3-10; the exposure
conditions sufficient to elicit these effects are further defined in Section 4.1. This conclusion is
based on robust studies in rodents that found Cr(VI) causes nonneoplastic effects in the GI tract.
These effects include dose-responsive diffuse epithelial hyperplasia in mice after both subchronic
and chronic exposure at all doses, and degenerative changes in the rat and mouse intestine. Human
evidence for nonneoplastic effects in the GI tract was indeterminate due to a lack of studies of
chronic, nonneoplastic GI effects in humans. The ATSDR Toxicological Profile fATSDR. 20121
described multiple case reports of Cr(VI) induced GI toxicity or deaths among adults and children
but none included an appropriate comparison group.
The animal toxicological database provides robust evidence that Cr(VI) is toxic to the GI
tract. The primary nonneoplastic effects associated with both chronic and subchronic oral
exposure to Cr(VI) in the GI tract are consistent and biologically coherent, and include epithelial cell
hyperplasia, degenerative changes, and histiocytic cellular infiltration in the small intestine. Diffuse
epithelial hyperplasia of the small intestine was predominant in mice across all studies, with
incidence increasing with dose. NTP observed diffuse epithelial hyperplasia, which involved the
entire small intestinal mucosa, in all exposed groups (>0.3 mg/kg-d Cr(VI)) of males and females in
both subchronic and chronic studies (NTP. 2008. 20071. The incidence rate was high (>26%) at the
lowest dose. Other subchronic experiments, including a strain comparison study by NTP, also
observed these lesions in mice (Thompson etal.. 2011: NTP. 20071. The dose-response relationship
for epithelial hyperplasia was stronger in the proximal small intestine (duodenum) than it was in
the jejunum (see Figure 3-12), indicating the effects of Cr(VI) are diminished by a decrease in
concentration as the chemical traverses the small intestine27. In addition to diffuse hyperplasia,
there was a low, nonsignificant incidence of focal epithelial hyperplasia in the duodenum observed
by NTP after 2 years in both male and female mice at the mid and high doses. These lesions are
discussed further in Section 3.2.3.2 as they may be more indicative of a direct treatment-related
preneoplastic response.
In rats, epithelial hyperplasia and villus atrophy/blunting were only reported in one
subchronic study limited to females (>7.2 mg and 31.1 mg/kg-d Cr(VI) respectively) (Thompson et
al.. 2012bl. Histopathological discrepancies in the rat small intestine between these findings and
27As Cr(VI) traverses the small intestine, the concentration of Cr(VI) in the lumen decreases due to 1)
reduction of Cr(VI) to Cr(III), 2) uptake to the small intestine epithelium, 3) dilution by GI contents (including
by ongoing intestinal secretions). See Section 3.1 for more detail.
This document is a draft for review purposes only and does not constitute Agency policy.
3-57 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofHexavalent Chromium
the NTP (2008, 20071 studies are a source of uncertainty, but could involve differences in study
variables such as those described by Thompson etal. f2012bl (e.g., different vendor sources,
differences in water intake), or differences in analyses (i.e., comprehensive pathology reporting by
NTP vs. hypothesis-driven MOA studies by Thompson et al. (Francke and Mog. 202111. In the
glandular stomach, a significantly increased incidence of nonneoplastic lesions was seen in male
and female F344 rats exposed to the highest dose (21 mg/kg-d Cr(VI)) in the subchronic NTP study;
this effect was not observed at any dose after two years (NTP. 2008. 20071. This is likely explained
by differences in dosing, as the rat stomach lesions observed after 13 weeks occurred at an
exposure threefold higher than the highest dose in the 2-year chronic assay.
Observations of histiocytic infiltration in the small intestine were consistent across studies,
sexes, and species; however, this effect is of unknown biological significance. Histiocytic infiltration
(to varying degrees) was also observed in the liver and the pancreatic and mesenteric lymph nodes
fNTP. 2008. 20071. A plausible explanation for this effect is increased phagocytosis due to an
insoluble precipitate of the test material. Cr(III), the reduced form of Cr(VI), is not a substrate for
active transport through the cell membrane and would therefore enter cells through passive
diffusion or phagocytosis fEastmond etal.. 20081. Therefore, the observed histiocytosis is most
compatible with phagocytically active macrophages containing Cr(III). An alternative explanation
could be that histiocytosis occurred as a result of chronic inflammation; however, neither pathology
consistent with inflammation nor the presence of other inflammatory cells types were observed in
rats or mice following drinking water exposures fNTP. 2008. 20071.
Together, these effects provide consistent, biologically coherent evidence of GI toxicity
involving tissue wounding by the test substance leading to degenerative changes, regenerative
proliferation, and hyperplasia. The hyperplasia in the GI tract following oral exposures is
considered to be representative of the constellation of histopathological observations that together
result in a change in tissue function that is considered an adverse noncancer effect, independently
from the significance of this lesion as a preneoplastic effect in the potential progression to cancer.
Mechanistic evidence from in vitro and in vivo models provides additional support for GI tissue
cytotoxicity and apoptosis occurring as a result of Cr(VI) exposure, as well as a proliferative
response that may be directly associated with a Cr(VI)-induced stimulation of proliferative cell
signaling pathways, an indirect consequence of compensatory cell proliferation following tissue
injury, or a combination of both.
This document is a draft for review purposes only and does not constitute Agency policy.
3-58 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-10. Evidence profile table for effects in the GI tract other than cancer
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Evidence from studies of exposed humans (occupational multi-route)
No human studies met
PECO criteria for
nonneoplastic GI effects
For human evidence of cancer of
the GI tract, see Section 3.2.3.2
and Table 3-24 (Evidence profile
table for cancer of the GI tract).
ooo
Indeterminate
Evidence from animal studies (oral)
HISTOPATHOLOGICAL
CHANGES
High confidence:
NTP(2008)
NTP(2007)
Thompson et al. (2012b)
Thompson et al. (2011)
Degenerative changes in
intestinal villi and hyperplasia of
the small intestine observed in
male and female mice by NTP
(2008, 2007). and in female mice
and rats by Thompson et al.
(2012b; 2011).
Histiocytic cellular infiltration
observed in the small intestine of
male and female rats and mice in
all studies and bioassays.
Because these effects can also
represent preneoplastic lesions
that are part of the morphologic
and biologic continuum leading
to cancer (Boorman et al., 2003).
additional discussions are
provided in Section 3.2.3.2
(Gastrointestinal Tract Cancer)
and Table 3-24.
Consistent
• Inconsistent
findings in mice
observations
in four high
of hyperplasia
confidence
between mice
studies reporting
and rats,
multiple
though this is
bioassays (both
explained in
sexes and
part by
multiple strains
pharmacokine
of mice)
tic differences
Coherent,
biologically
related findings
across studies
Large magnitude
of effects
Strong dose-
response
gradient
©0©
Robust
Histopathological
changes reported in
high confidence studies
(proliferative changes)
observed across the
animal evidence base
database are coherent
following chronic and/or
subchronic oral
exposures in rats and
mice and suggest
adverse effects of Cr(VI)
on the GI tract
(specifically, the small
intestine), findings that
are supported by
mechanistic evidence of
oxidative stress and cell
proliferation.
®©o
The evidence indicates
that Cr(VI) is likely to
cause GI toxicity in
humans given sufficient
exposure conditions.
Robust evidence in rats
and mice shows
consistent findings of
histopathological changes
indicative of epithelial
damage and changes in
GI epithelial architecture
following oral exposure.
Although these effects
are presumed to be
relevant to humans, the
lack of human evidence
demonstrating that the
changes observed in
rodents would occur and
progress in humans
precludes a higher
conclusion level (i.e.,
evidence demonstrates).
Mechanistic findings in
animals provide some
evidence supportive of
oxidative stress in the GI
tract as a potential
mechanism for
This document is a draft for review purposes only and does not constitute Agency policy.
3-59 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
• Mechanistic
evidence
(oxidative stress,
cell
proliferation)
provides
plausibility
degenerative Gl effects in
multiple animal species.
This mechanism is
presumed relevant to
humans.
The evidence is
inadequate to determine
whether Cr(VI) inhalation
Mechanistic evidence
exposure might be
capable of causing
Biological events or
pathways
Summary of key findings and interpretations
Judgments and
rationale
noncancer Gl effects. No
noncancer Gl effects
were observed following
Oxidative stress
Interpretation: Cr(VI) is a potent oxidizer that can produce reactive oxygen
species and oxidative stress via intracellular intermediate species, leading to
cytotoxicity in the Gl tract following oral exposures. This supports evidence
of degenerative lesions in the Gl tract (see animal evidence, above).
Key findings:
• Decreased GSH/GSSG ratio in small intestinal epithelium at 8 and 90
days in mice and 90 days in rats, and in oral mucosa in mice at 8 days
and rats at 90 days, although no 8-OHdG adducts or protein oxidation in
anv tissues (Thompson et al., 2011; De Flora et al., 2008)
• In vitro evidence of increased oxidative stress in human colorectal
adenocarcinoma Caco-2 cells, though this also occurred at
concentrations that induced minimal or no cvtotoxicitv (Thompson et
al., 2012a)
Biologically plausible
mechanistic evidence
supports involvement of
oxidative stress in the
histopathological
findings of degenerative
effects, although there
are some
inconsistencies in the
animal findings in the Gl
tract following oral
exposures. Evidence of
increased cell
proliferation in affected
tissues is consistent with
hyperplasia but cannot
inhalation. As described
in Section 3.1, Cr(VI) can
expose portal-of-entry
tissues, and reduction of
Cr(VI) in these tissues and
red blood cells decreases
uptake by other organ
systems.
Cell proliferation
Interpretation: Evidence of increased cell proliferation is consistent with the
histopathological observations of hyperplasia in the mouse small intestine
following oral exposure to Cr(VI) (see animal evidence, above), although
these measures do not indicate the molecular stimuli for the proliferation
and it is unknown whether they are indicative of regenerative proliferation.
Increased cell proliferation has not been detected in the rat oral cavity.
be conclusively
associated with tissue
regeneration following
injury.
This document is a draft for review purposes only and does not constitute Agency policy.
3-60 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Factors that increase Factors that
Summary of key findings certainty decrease certainty
Judgments and
rationale
Key findings:
• The cellular replication marker Ki-67 was increased in isolated duodenal
mucosal cells from the small intestine of mice exposed to Cr(VI) via
drinking water for 7 and 90 davs (Rager et al., 2017; Kopec et al., 2012a)
• Dose-dependent increases in the protein and gene expression of c-Myc,
an oncogenic cell proliferation promoter, and downregulation of cell
cycle regulator p53, in rat stomach and colon exposed to doses as low as
5 mg/kg-d Cr(VI) in drinking water for 60 days of exposure to Cr(VI) in
drinking water (Tsao et al., 2011)
• Toxicogenomic analyses of Gl tissues in Cr(VI)-treated animals have
identified differentially expressed genes (DEGs) associated with
activation of c-Myc, MAPK, and a variety of additional pathways
associated with cell cycle and proliferation (see Appendix C.3.4)
This document is a draft for review purposes only and does not constitute Agency policy.
3-61 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
3.2.3. Cancer
3.2.3.1. Respiratory Tract Cancer
In 1998, the EPA IRIS Toxicological Review ofHexavalent Chromium classified Cr(VI) as a
"known human carcinogen" by the inhalation route of exposure (U.S. EPA. 1998c). This
determination was based on the revised carcinogenicity guidelines, which were proposed at that
time fU.S. EPA. 1996cl and finalized in 2005 fU.S. EPA. 2005al. The "known human carcinogen"
classification replaced the classification as a "Group A - known human carcinogen" by the inhalation
route of exposure under the previous carcinogenicity guidelines (U.S. EPA. 1986b). This
classification was based on consistent evidence that inhaled Cr(VI) causes lung cancer in humans
and supporting evidence of carcinogenicity in animals. The same conclusion has since been
reached by other authoritative federal and state health agencies and international organizations
and the carcinogenicity of Cr(VI) is considered to be well established for inhalation exposures
CTCEO. 2014: IPCS. 2013: NIOSH.2013: TARC. 2012: CalEPA. 2011: NTP. 2011: OSHA. 20061. Thus,
the current review of cancer by the inhalation route adopts the same EPA cancer descriptor for this
route, "carcinogenic to humans," and the analyses focus on data that may improve the quantitative
exposure-response analysis conducted in EPA's 1998 IRIS assessment, as stated in the 2014
preliminary packages (U.S. EPA. 2014b. c) and the Systematic Review Protocol (Appendix A). An
overview of the literature screening and study evaluation for exposure-response data is presented
in Section 4.4.
3.2.3.2. Gastrointestinal Tract Cancer
Human evidence via the oral route of exposure
Study evaluation summary
Three studies analyzed stomach cancer risk in populations exposed to Cr(VI) in drinking
water. Three additional studies were identified but excluded due to critically deficient ratings in at
least one domain, and are not discussed further fFrvzek etal.. 2001: Bick etal.. 1996: Bednar and
Kies. 19911. The three included, low confidence studies are ecological analyses of cancer mortality
in residential populations with potential exposure to Cr(VI)-contaminated drinking water in China
and Greece (Table 3-11).
Two of the studies were ecological analyses of cancer mortality in relation to groundwater
contamination in the same exposed population in Liaoning Province, China fKerger etal.. 2009:
Beaumont etal.. 2008). The Beaumont et al. (2008) and Kerger etal. (2009) studies are reanalyses
of Zhang and Li f!987bl. the original scientific report published in the Chinese Journal of
Preventive Medicine. Another publication, Zhang and Li (1997). has been challenged for conflict-of-
This document is a draft for review purposes only and does not constitute Agency policy.
3-62 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
interest due to undisclosed funding28. Investigators compared cancer mortality rates (total
between 1970-1978) between five contaminated regions identified along a groundwater plume of
Cr(VI) and four presumed uncontaminated regions surrounding a ferrochromium production plant.
The contaminated areas included five communities downgradient of the alloy plant along a dry
riverbed where plant wastewater effluent from chromium smelting had been disposed since 1960.
The communities without contamination included the town adjacent to the alloy plant (TangHeZi)
and three agricultural areas to the north, west and south. Another study with an ecological design,
Linos et al. (20JJJ, analyzed cancer mortality and Cr(VI) exposure via drinking water in Oinofita
municipality, Greece, with data on residents from 1999-2009. Processed liquid industrial waste
containing Cr(VI) was dumped into Asopos River starting around 1969, which was the source for
drinking water in wells within the municipality from 1970-2009 (Linos etal.. 20111.
The definition of Cr(VI) exposure in these studies was based on living in towns or areas
proximate to contaminated rivers, which were the source of drinking water, and assumed
consumption. Individual-level data on the source or amount of drinking water consumed was not
collected. Sampling to measure Cr(VI) concentrations in drinking water was limited in terms of
timespan as well as geographical coverage. In addition, only drinking water in the areas with
suspected contamination was sampled; Cr(VI) concentrations were not measured in drinking water
in areas considered to be unexposed, which could lead to unrecognized exposure and subsequent
misclassification f Linos etal.. 2011: Kerger etal.. 2009: Beaumont etal.. 20081. Based on data for
Liaoning Province reported by the Jinzhou Health and Anti-epidemic Station in 1986,
concentrations of Cr(VI) in drinking water (analytical methods were not available) in 1965, when
the contamination was identified, ranged between 0.002-20.0 mg/L in villages along the plume
that extended from the disposal site located near the chromium alloy plant (Kerger etal.. 2009:
Beaumont etal.. 20081. Well water samples collected in Oinofita municipality between 2007-2010
ranged between 0.010-0.156 mg/L (Linos etal.. 20111.
The studies of both populations were classified as low confidence, primarily due to
limitations in the exposure assessment In each study, exposure was defined at the population
level; no individual-level exposure assignments were possible. Beaumont et al. (2008) and Kerger
et al. (2009) assigned exposure status based on residence information in the death certificate.
Residence at the time of death may not represent residence location - and thus inferred Cr(VI)
exposure - at the critical time window for initiation and progression of cancer, although such
misclassification of the exposure proxy is expected to be nondifferential. In addition, the duration
of follow-up in both studies was not adequate to allow for the long latency of cancer development.
These limitations are expected to result in bias in a direction toward a null association. Finally, age-
adjusted site-specific cancer mortality by region for the study years in China was not available to
the investigators and had to be estimated using other available data.
28Zhang and Li fl9971 was retracted by the journal because "financial and intellectual input to the paper by
outside parties was not disclosed" fSmith. 2008: Brandt-Rauf. 20061.
This document is a draft for review purposes only and does not constitute Agency policy.
3-63 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Toxicological Review ofHexavalent Chromium
Table 3-11. Summary of human studies for Cr(VI) cancer of the GI tract and
overall confidence classification. Click to see interactive data graphic for rating
rationales.
Author (Year)
Location
Exposure Assessment
Study Design
Selection
Exposure
Outcome
Confounding
Analysis
Sensitivity
Self-reporting
Overall confidence
Beaumont et
al. (2008)a
Liaoning
Province, China
Communities
downstream of a
ferrochromium plant
versus unexposed
communities (assumed)
Semi-ecologic
cancer
mortality
D
D
D
D
A
D
A
Low
Kerger et al.
Liaoning
Province, China
Communities
downstream of a
ferrochromium plant
versus unexposed
communities (assumed)
Semi-ecologic
cancer
mortality
D
D
D
D
A
D
A
Low
(2009)a
Linos et al.
(2011)
Oinofita, Greece
Residents of Oinofita, a
contaminated region
versus surrounding
residents
Semi-ecologic
cancer
mortality
A
D
A
A
A
D
A
Low
G = good; A = adequate; D = deficient.
aStudies are reanalyses of Zhang and Li (1997; 1987a).
Each of the three studies selected the referent, or unexposed population, as residents of the
larger area surrounding the exposed area (Linos etal.. 2011: Kerger etal.. 2009: Beaumont etal..
20081. and were not able to account for differing lifestyles, occupational histories, or background
rates of cancer in the referent population that may influence cancer risk. Beaumont et al. (20081
compared cancer mortality in the contaminated villages to mortality in either the surrounding
unexposed villages, or the entire Liaoning Province, with both comparison groups including the
industrial city of TangHeZi. Larger populations, such as a province or state, have the advantage of
providing relatively stable estimates, particularly for low-incident events such as site-specific
cancers, but may obscure differences by demographic and other characteristics important for the
study population. Kerger etal. (2009) compared cancer mortality in the chromium-exposed
agricultural areas to the unexposed agricultural areas and to the unexposed city of TangHeZi
separately to address potential residual confounding by demographic and socioeconomic factors.
Mortality rates for stomach cancer in TangHeZi were lower than those in the unexposed
agricultural areas. Although an analysis of gastric cancer rates in China in 1990-1992 showed
lower mortality rates in urban areas (15.3 per 100,000) compared with rural areas (24.4 per
100,000), possibly in response to economic development and urbanization (e.g., sanitation,
refrigeration) (Yang. 2006). this same study reported little difference between urban and rural
This document is a draft for review purposes only and does not constitute Agency policy.
3-64 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review ofHexavalent Chromium
rates in 1973-1975 (20.1 and 19.4 per 100,000 in urban and rural areas, respectively), the relevant
time period with respect to the Liaoning Province studies given the anticipated latency of cancer
development and diagnosis following the onset of exposure. Therefore, while it is possible that
demographic differences influenced the difference in mortality rates, another factor may have been
statistical instability due to small population sizes.
Synthesis of human evidence
Results of the studies on Cr(VI) oral exposure and cancer are presented in Table 3-12. The
analyses of stomach cancer in two exposed populations in Liaoning Province, China, and Oinofita,
Greece, showed an association with Cr(VI), although effect estimates were imprecise. While the
results of two reanalyses of Zhang and Li (1987a) indicated an increased risk when comparing the
exposed villages to the unexposed referent group, inclusion of the industrial city of TangHeZi in the
referent group increased the magnitude of the relative risk, which became statistically significant
(including TangHeZi, RR 1.82, 95% CI: 1.11, 2.91; excluding TangHeZi, RR 1.22, 95% CI: 0.74, 2.01)
fKerger etal.. 2009: Beaumont etal.. 20081. The mortality rate from stomach cancer was much
lower in TangHeZi, the reason why inclusion of the city was influential. However, Beaumont et al.
(2008) also used the mortality experience of the larger province as a referent and observed an
elevated, statistically significant risk (SMR: 1.69, 95% CI: 1.12-2.44). The number of deaths from
stomach cancer was not reported for one of the villages with higher contamination levels, which
makes it difficult to compare results between the two studies.
Table 3-12. Associations between drinking water exposures to Cr(VI) and
cancer in low confidence epidemiology studies
Reference
Exposure
Cancer Deaths (N)
Relative
Risk
Ratio Measure
(95% CI) N
Linos et al.
(2011)
Oinofita, Greece
Cr(VI) in drinking water
Mortality in exposed areas
compared to surrounding area
(assumed to be unexposed)
All cancers (118)
Stomach (6)a
SMR (95%
CI)
All cancers: 113.6 (94.1,
136.1)
Stomach: 120.9 (44.4,
263.2)
Beaumont et al.
(2008)
Liaoning
Province, China
Cr(VI) in drinking water
Mortality in exposed
communities compared to
nearby regions (assumed to be
unexposed) and to province as a
whole
All cancer (262)b
Stomach cancer
(75)b,c
Rate ratio
(95% CI)
Compared to unexposed
regions:
All cancers: 1.13 (0.86,1.46)
Stomach: 1.82 (1.11, 2.91)
Compared to larger
province:
All cancers: 1.23 (0.97,1.53)
Stomach: 1.69 (1.12, 2.44)
Kerger et al.
(2009)
Liaoning
Province, China
Cr(VI) in drinking water
Mortality in exposed
communities (C) compared to
(A) industrial town, or (B) three
unexposed agricultural villages
All cancer (263)b
Stomach cancer
(89)b,c
Rate ratio
(95% CI)
C vs. B
All cancers: 1.10 (0.80,1.51)
Stomach: 1.22 (0.74, 2.01)
B vs. A
All cancers: 1.03 (0.77,1.39)
This document is a draft for review purposes only and does not constitute Agency policy.
3-65 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
Reference
Exposure
Cancer Deaths (N)
Relative
Risk
Ratio Measure
(95% CI) N
Stomach: 1.70 (1.00, 2.89)
C vs. A
All cancers: 1.14 (0.85,1.52)
Stomach: 2.07 (1.25, 3.44)
aSite-specific cancer risk presented for number of cases >5.
bNumber deaths in the study villages were estimated as described by authors.
cMortality rates were missing for stomach cancer in one contaminated village, Nuer River Village.
The studies of both of the populations exposed to Cr(VI) in drinking water reported
increased SMRs when their mortality experience was compared to unexposed communities in the
surrounding areas. These estimates were imprecise and changed in magnitude depending on the
definition of the unexposed communities. The lack of individual estimates of exposure, the
uncertain nature of the mortality data, and the potential impact of confounding by differences in
SES between comparison groups make it difficult to draw any conclusions.
Human evidence via the inhalation route of exposure
EPA conducted a review and meta-analysis of GI cancer risk from studies of workers with
occupational inhalation exposure to Cr(VI). Exposure via inhalation may pose an increased risk of
cancer in the GI tract in occupationally exposed populations either as a result of systemic
absorption and distribution, or via deposition in airways, mucociliary clearance, and swallowing of
particles (Sedman etal.. 20061. Numerous studies have evaluated the association between Cr(VI)
exposure and cancers of the GI tract, including at least three recent meta-analyses (Deng etal..
2019: Suh etal.. 2019: Welling etal.. 20151 and two older meta-analyses (Gatto etal.. 2010: Cole
and Rodu. 20051 (Table 3-13). These meta-analyses varied in their scope and the specific research
question under study. Among the more recent meta-analyses, the Welling et al. study (Welling et
al.. 20151 concluded that Cr(VI) exposure was associated with increased risks of stomach cancer,
while Suh et al. (2019) had the opposite conclusion; the work by Deng et al. (2019), which
considered additional cancer sites, concluded that there was no evidence for increased risk of death
due to digestive system cancers overall, but that the findings for rectal cancer specifically were
suggestive of increased risk, and the risk of oral cancer incidence (not mortality) was significantly
increased. EPA performed an updated literature search to identify studies for inclusion in a new
meta-analysis of Cr(VI) exposure in relation to GI tract cancers. The goal of the meta-analysis was
to calculate summary effect estimates for persons with likely occupational exposure to Cr(VI) from
an updated set of studies with similar design. Methods for the systematic review and meta-analysis
are in Appendix C.3.1.
This document is a draft for review purposes only and does not constitute Agency policy.
3-66 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-13. Meta-analyses of GI tract cancers and Cr(VI) occupational
exposure
Study
Outcome
Included
Excluded
Summary effect
estimate and 95%
confidence interval
for specified cancer
sites (number of
included studies)
Cole and
Rodu
(2005)
Relative risk
(RR) estimates
for stomach
cancer
Began with set of known
relevant studies, then
performed a literature
search; included those
published after 1950
'no usable data'; 'occupational
settings with little or no chrome
exposure'
Stomach (n = 32):
1.13 (1.03,1.24)
Gatto et al.
Measures of
effect or data
available to
calculate
relative risk
(RR) for GI tract
cancers
Published after 1950;
occupational exposure
(inhalation or ingestion);
exposure potential stated
explicitly or from industry
with recognized exposure
potential: chromate
production, stainless-steel
welding, chrome pigment
production, chrome plating,
ferrochrome production
Esophagus (n = 15):
1.17 (0.90,1.51)
Stomach (n = 29):
1.09 (0.93, 1.28)
Colon (n = 13): 0.89
(0.70, 1.12)
Rectum (n = 20):
1.17 (0.98, 1.39)
(2010)
Welling et
al. (2015)
Relative risk
(RR) estimates
for stomach
cancer
Chromate or chromium
production and plating;
leather work and tanning;
Portland cement work; and
stainless-steel production,
welding, polishing and
grinding
Occupations such as painting,
general foundry work,
construction and shoe (non-
leather) manufacturing; Welding
or metal plating studies that did
not evaluate stainless-steel or
chromium work; Studies involving
work with asbestos cement
Stomach (n = 56):
1.27 (1.18, 1.38)
Deng et al.
(2019)
Standardized
mortality or
incidence ratio
(SMR or SIR)
estimates for
cancer of the
digestive
system
"the exposure factor was
clear and exposure was to
Cr(VI)"
Chromate production,
cement production, cement
industry workers, aircraft
manufacturing workers,
chromium platers, tanners,
welders, masons
Occupational exposure to
materials other than Cr(VI), such
as asbestos or nickel; professions
such as shoemaking (non-leather)
or general building work. Based
on study quality evaluation using
Newcastle-Ottawa scale, excluded
studies with ratings <6
Esophagus (n = 14):
0.88 (0.73, 1.05)
Stomach (n = 33):
0.93 (0.78, 1.09)
Colon (n = 12): 1.06
(0.93, 1.21)
Rectum (n = 23):
1.14 (0.98, 1.33)
Suh et al.
(2019)
Stomach cancer
morbidity
and/or
mortality
Chromate production,
stainless-steel welding,
chrome pigment
production, chrome
plating/ electroplating,
PMR studies, Registry studies
where 'Specifications of Cr(VI)
exposures are not indicated by
the authors"—includes studies
such as Andersen et al. (1999)
Stomach (n = 44):
1.08 (0.96, 1.21)
This document is a draft for review purposes only and does not constitute Agency policy.
3-67 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Toxicological Review ofHexavalent Chromium
Study
Outcome
Included
Excluded
Summary effect
estimate and 95%
confidence interval
for specified cancer
sites (number of
included studies)
ferrochrome production
industries, Leather tanners
(if indicate exposure to
Cr(VI) or process such as
"two bath" process),
Cement workers (if
involved cement
production); Other
occupations if Cr(VI)
exposure indicated by
authors
and Pukkala et al. (2009). Based
on study quality evaluations using
NTP OHAT Risk of Bias
Rating Tool for Human and
Animal Studies, tiered studies and
excluded tier 3.
Occupational studies that analyzed cancer risks related to Cr(VI) exposure were identified
as part of the overall assessment search strategy process described in the Cr(VI) Protocol fU.S. EPA.
20191. This search strategy, which was conditioned on terms for Cr(VI), identified 35 potentially
relevant citations. Since these searches only identified references that mentioned chromium or
related terms in the title or abstract, an additional search strategy was developed to identify studies
of occupational groups with likely exposure to Cr(VI). The search terms and literature
identification results are found in Appendix C.3.1. In total, 35 references from the previous
literature searches for the assessment, 93 references from the subsequent occupationally-focused
search for the meta-analysis, and 20 references identified by looking through the reference lists in
the three most recent meta-analyses were included in this review. Of these, 21 studies were not
included because they were earlier follow-ups with more recent reports available, the cohorts were
not exposed to Cr(VI), or they did not contain results for site-specific GI tract cancers.
A comparison of the studies included in the three most recent meta-analyses and this
analysis with a rationale for decisions to exclude is in the appendix (Appendix Table C-44, Section
C.3.1). The studies included in each meta-analysis comprised a partially overlapping set of studies
reflecting the various time periods used for the literature searches, the inclusion criteria, and the
results of the evaluations of study "quality" used in the studies. The meta-analyses focused on the
studies considered to be medium or high overall confidence, for which EPA had greater certainty in
the exposure assessment for Cr(VI) and minimal concern for other sources of bias. In this analysis,
the primary reason for considering a study to be of low confidence was that exposure to Cr(VI) in
the population was too uncertain.
The studies included in EPA's meta-analysis reported a variety of effect estimates, including
standardized incidence or mortality ratios, standardized risk ratios, odds ratios, and proportionate
mortality ratios. Studies that calculated proportionate mortality ratios were not included. In some
instances, multiple risk estimates were reported—for example, for men or women separately, for
This document is a draft for review purposes only and does not constitute Agency policy.
3-68 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
exposure or occupational subgroups, or by latency period. A priori, risk estimates were preferred if
they (1) were adjusted for potential confounders including age, sex, time period, and geographic
region; (2) were estimated for the longest latency period; (3) were from the most recent follow-up
of a specific study cohort; (4) were estimated for the most highly exposed subgroup of the study
population. When reviewing the studies captured by the literature search and evaluation of the
studies, there were some cancer sites or groupings that were difficult to reconcile across studies
due to differences in ICD codes included, for example, or changes in coding practices and diagnostic
naming conventions over time and across geographical sites. Consequently, it was hard to
determine whether the same cancer sites were contained within some of the groupings. Further, in
some cases the number of studies for a given cancer site was small enough (and heterogenous
enough) that a meta-analysis seemed unlikely to yield useful information. Consequently, a
quantitative meta-analysis was performed to derive summary risk estimates for a subset of GI tract
cancers by site: esophagus, stomach, colon, and rectum. For each of these four sites, there was a
larger number of studies to include in a summary effect estimate, and these studies used relatively
consistent definitions for these specific cancer sites.
Separate meta-analyses were performed to obtain summary estimates from studies
reporting odds ratios (stomach cancer, esophageal cancer), and from studies reporting SMR, SIR, or
SRR estimates (all four sites). All analyses were performed using the 'metafor' package in R
(Viechtbauer. 2010). with a random effects model. This package was also used to generate forest
plots (see Figures 3-15 to 3-21). The potential for publication bias was evaluated using the Egger's
test (Egger etal.. 1997) for funnel plot asymmetry. The I2 statistic value for each study is used to
represent the percentage of variation across studies that is due to heterogeneity rather than
chance.
As shown in Table 3-14, the summary effect estimates showed small increases in risk for
each cancer site associated with Cr(VI) exposure, although only the estimate for rectal cancer was
statistically significant There were few studies reporting odds ratios, but in each case (esophagus
and stomach), summary effect estimates based on these studies were somewhat higher compared
with summary estimates based on other relative risk measures (although neither odds ratio-based
estimate was statistically significant). There was no evidence of funnel plot asymmetry based on
Egger's regression test, indicating that publication bias was not likely to be present.
Summary effect estimates were also derived for each cancer site, stratified by occupational
grouping (see Appendix Table C-45). This separation by occupational grouping did show some
expected patterns for colon cancer risk estimates in that the occupations with a higher certainty of
exposure to Cr(VI) (i.e., ferrochromium, chromate production, stainless-steel workers, chromium
pigment exposed workers) showed higher summary effect estimates. However, there remained
inconsistencies among the studies overall, and the results for cancer of the rectum did not show a
similar pattern of risk. The results of these more detailed analyses are discussed in Appendix
C.3.1.3.
This document is a draft for review purposes only and does not constitute Agency policy.
3-69 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofHexavalent Chromium
Table 3-14. Summary effect estimates from random effects meta-analysis, by
cancer site and type of effect estimate
Cancer
Site
Effect Estimate Type
Number of
individual effect
estimates
Summary effect
estimate (95%
confidence interval)
p-value for funnel
plot asymmetry
Esophagus
Odds Ratio
2
1.43 (0.19, 11.09)
Not computed
Relative Risk (SMR, SIR, or
SRR)
21
1.08 (0.92, 1.37)a
0.33
Stomach
Odds Ratio
4
1.38 (0.77, 2.49)
0.79
Relative Risk (SMR, SIR, or
SRR)
48
1.01 (0.89, 1.15)
0.08
Colon
Relative Risk (SMR, SIR, or
SRR)
19
1.10 (0.97, 1.25)
0.53
Rectum
Relative Risk (SMR, SIR, or
SRR)
32
1.18 (1.01, 1.37)
0.94
aWarning displayed during estimation of the summary estimate indicates that results may not be stable due to the
large range of sampling variance between included estimates.
Due to misclassification and heterogeneity of Cr(VI) exposure among and within the
included studies, there may have been a decreased ability to detect an association if it existed.
Although this analysis included studies that analyzed associations among occupational groups or
subgroups with greater certainty of exposure to Cr(VI), variation in the prevalence, frequency and
magnitude of exposure is likely within the exposure groups. Other factors that could contribute to
the observed heterogeneity of risk estimates include presence of coexposures and bias due to the
use of occupational cohorts. Cancer risk in these industries is likely affected by prevalent exposures
to other carcinogens in addition to Cr(VI), which would vary both within and across occupational
groupings. As noted in Appendix Table C-43, two industry groupings with higher certainty of Cr(VI)
prevalence, ferrochromium, chromate production, and stainless-steel workers, and chromium
pigment exposed workers, had occupational settings characterized by different coexposures, which
argues against a strong common confounder. In some cases, authors did attempt to adjust for
coexposures or restrict the study population to minimize their effect. The majority of the studies
estimated relative risk using SMRs, which also are subject to a bias toward the null due to the
healthy worker effect The summary effect estimates for esophageal and stomach cancers
calculated using odds ratios from the few case-control studies was not subject to this bias and
indicated a higher risk. However, these odds ratio estimates are based on very few studies and are
highly uncertain.
Previous meta-analyses reported summary effect estimates for stomach cancer which
ranged between 0.93 (Deng etal.. 2019) to 1.27 fWelling et al.. 20151. A statistically significant
increase in risk of stomach cancer was reported from two of the previous five estimates (Welling et
al.. 2015: Cole and Rodu. 20051. This assessment's finding of no increased risk (summary relative
This document is a draft for review purposes only and does not constitute Agency policy.
3-70 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
risk of 1.01) is within the range of these previous estimates. Two of the five previous meta-analyses
included estimates for cancers of the esophagus, colon and rectum fDeng etal.. 2019: Gatto etal..
20101. This assessment's summary estimate of 1.08 for esophageal cancer was not significantly
elevated, and was slightly less than that from Gatto etal. (20101. The effect estimate for colon
cancer of 1.10 (95% CI: 0.97,1.25), was close to the estimate reported by Deng et al. (20191.
Finally, this assessment's estimate of rectal cancer risk was significantly elevated, and very similar
to those previously reported (1.18, 95% CI: 1.01,1.37), compared with 1.17 (Gatto etal.. 20101 and
1.14 ("Dengetal.. 201911.
Animal evidence via the oral route of exposure
Synthesis of neoplastic animal evidence
Neoplastic lesions following oral administration of Cr(VI) via drinking water were observed
in the 2-year study conducted by NTP f20081 in both sexes of B6C3F1 mice and F344/N rats. This
was the only animal study examining the potential for tumor development via the oral route of
exposure and was rated as high confidence. An overview of the confidence classification for the GI
histopathology reported in this study can be found in Section 3.2.2, Table 3-8 and in HAWC.
In this study, both sexes of F344/N rats exhibited an increased incidence and trends of
squamous cell carcinomas or papillomas in the oral cavity (mucosa or tongue), uncommon tumor
types. Tumor incidence was statistically significant at the highest doses tested, 6.07 and 7.13 mg
Cr(VI)/kg-d in male and female rats, respectively. The overall tumor incidence at the high dose was
14% in male rats and 22% in female rats fNTP. 20081. as compared to no tumors in control males
and 2% incidence in females. There was also a nonsignificant, low incidence (4%) of oral cavity
tumors in female rats receiving 2.6 mg Cr(VI)/kg-d. In both male and female rats, the increasing
trend of oral tumors was statistically significant. Microscopic examination of the tumors present in
the oral cavity of rats indicated they were highly invasive, originating in the oral mucosa of the
palate adjacent to the upper molar teeth with spread to the tongue, Harderian gland, the soft tissues
surrounding the nose, and the brain fNTP. 20081.
In the same study, male and female B6C3F1 mice exhibited increased incidences and trends
of adenomas and carcinomas in the small intestine, with most tumors occurring in the duodenal
section (proximal small intestine, nearest to the stomach). In male mice, there was a significant
trend for increased incidence of adenoma and carcinomas in the small intestine. Statistically
significant increases in adenomas or carcinomas were observed at doses >2.4 mg Cr(VI)/kg-d with
an overall incidence of 40% at the high dose (NTP. 20081. Female mice also showed a significant
trend for increased incidence of adenomas and carcinomas in the small intestine. At doses >3.24
mg Cr(VI)/kg-d, incidence of adenomas was statistically significantly increased and reached up to
44%. While most tumors in both sexes were located in the duodenum (first section of the small
intestine), female mice also showed a significant increase (10%) in overall incidence in the jejunum
(middle section of the small intestine) at the highest dose. Three adenomas were observed in the
This document is a draft for review purposes only and does not constitute Agency policy.
3-71 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 male mouse jejunum at the highest dose (and while not significant when compared to controls, this
2 constitutes a statistically significant increasing trend and exceeded the historical control range for
3 drinking water studies and for all routes of administration (NTP. 200811. Histopathological
4 evaluation of the adenomas in mice were described as discrete, broad based and focally extensive;
5 composed of irregular, elongated crypts; epithelial cells with oval to elongated nuclei; and
6 increased mitotic activity (NTP. 2008). Carcinomas were characterized as extensive with invasion
7 of the submucosa and/or muscularis mucosa; epithelial cells with round, oval, or elongated nuclei;
8 and with atypical mitosis that was of greater extent than observed in adenomas.
9 The data for both species and sexes are summarized in Table 3-15 and Figure 3-15.
This document is a draft for review purposes only and does not constitute Agency policy.
3-72 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-15. Data on neoplastic lesions in a high confidence study of rats and
mice fNTP. 20081
Tumor type and species/sex
Administered mg/L, mg/kg-d Cr(VI)a and
incidence/total
Male B6C3F1 mice
0 mg/L
5
10
30
90
0 mg/kg-d
0.450
0.914
2.40
5.70
Adenomas (duodenum)
1/50
0/50
1/50
5/50
15/50*
Carcinomas (duodenum)
0/50
0/50
0/50
2/50
3/50
Adenomas or carcinomas
(duodenum, jejunum, or ileum)
Incidence / Total
1/50
3/50
2/50
7/50*
20/50*
Incidence / Total (adj)b
1/50
3/49
2/49
7/50*
20/50*
Animals dead prior to day 365
0
1
1
0
0
Female B6C3F1 mice
0 mg/L
5
20
60
180
0 mg/kg-d
0.302
1.18
3.24
8.89
Adenomas (duodenum)
0/50
0/50
2/50
13/50*
12/50*
Carcinomas (duodenum)
0/50
0/50
0/50
1/50
6/50*
Adenomas or carcinomas
(duodenum, jejunum, or ileum)
Incidence / Total
1/50
1/50
4/50
17/50*
22/50*
Incidence / Total (adj)b
1/49
1/50
4/49
17/50*
22/49*
Animals dead prior to day 365
1
0
1
0
1
Male F344 rats
0 mg/L
5
20
60
180
0 mg/kg-d
0.200
0.796
2.10
6.07
Squamous cell carcinoma (oral mucosa)
0/50
0/50
0/49
0/50
6/49*
Squamous cell papilloma (oral mucosa)
0/50
0/50
0/49
0/50
1/49
Squamous cell carcinoma (tongue)
0/49
1/50
0/47
0/49
0/48
Squamous cell papilloma (tongue)
0/49
0/50
0/47
0/49
1/48
Squamous cell carcinoma or
papilloma (oral mucosa or tongue)
Incidence / Total
0/50
1/50
0/49
0/50
7/49*
Incidence / Total (adj)b
0/50
1/47
0/47
0/50
7/49*
Animals dead prior to day 365
0
3
2
0
0
Female F344 rats
0 mg/L
5
20
60
180
0 mg/kg-d
0.248
0.961
2.60
7.13
Squamous cell carcinoma (oral mucosa)
0/50
0/50
0/50
2/50
11/50*
Squamous cell carcinoma (tongue)
0/45
0/49
0/48
1/48
0/48
Squamous cell papilloma (tongue)
1/45
1/49
0/48
0/48
0/48
This document is a draft for review purposes only and does not constitute Agency policy.
3-73 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Tumor type and species/sex
Administered mg/L, mg/kg-d Cr(VI)a and
incidence/total
Squamous cell carcinoma (oral
mucosa or tongue)
Incidence / Total
1/50
1/50
0/50
2/50
11/50*
Incidence / Total (adj)b
1/50
1/50
0/50
2/50
11/50*
Animals dead prior to day 365
0
0
0
0
0
aTime-weighted average daily doses calculated from NTP water consumption data.
bTumor incidences adjusted based on the number of animals surviving beyond 365 days. First tumor onset: 451
days for intestinal tumors in mice, and 506 days for oral tumors in rats (both occuring at the highest doses).
* Denotes significant difference from the control group reported by NTP (2008) using the Poly-3 test (p < 0.05).
O SI tumors {mice M+F)
X Oral tumors (rats M+F)
0.50
g 0.40 +
c
"O
u 0.30
(T>
I 0.20
o
TO
Ł o.io
QQ
tiX X
0.00 T X| r - f I r
0 25 50 75 100 125 150 175 200
Cr(VI) mg/L
(A) Data by mg/L
0.50
OSI tumors (mice M+F)
X Oral tumors (rats M+F)
0.40
o
c
CD
"O
u 0.30 4
re
c
O
u
re
0.20
0.10
0.00
0 2 4 6 8
Cr(VI) mg/kg-d
(B) Data by mg/kg-d
10
Figure 3-15. Fractional incidence of mice with adenomas or carcinomas in the
small intestine (SI tumors), and fractional incidence of rats with squamous
cell carcinomas or papillomas in the oral mucosa or tongue (oral tumors).
Data presented on a basis of (A) administered mg/L Cr(VI), where incidence data for
male and females were combined, and (B) administered mg/kg-d Cr(VI), where
incidence data for males and females are separated due to differences in water
intake and dose. For mice, both males and females were exposed to 5 mg/L, while
all other nonzero doses differed between males and females. For rats, both males
and females were exposed to the same mg/L Cr(VI) concentration levels. Incidence
data adjusted for rodents surviving at least one year.
1 Notably, at the lower doses, incidences of specific neoplasms in the GI tract observed during
2 the 2-year study exceeded NTP historical controls in both B6C3F1 mice and F344 rats. Therefore,
3 some tumors which were not statistically significant versus concurrent controls at low doses may
4 be biologically significant due to the increasing trend and low historical control incidence
5 (Appendix D.2). Tumors of the oral cavity are rare (Ibrahim etal.. 2021: Leininger and Schutten.
6 2018: Chandra etal.. 20101. In the 2-year NTP (20081 bioassay, one squamous cell carcinoma was
7 identified in the tongue of a male rat in the lowest dose group (0.2 mg Cr(VI)/kg-d), and in the
8 tongue of a female rat at 2.6 mg Cr(VI)/kg-d. The historical controls for squamous cell carcinoma of
This document is a draft for review purposes only and does not constitute Agency policy.
3-74 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
the tongue are 1/1499 for male rats and 0/1449 for female rats (see Appendix D.2). The historical
rates of squamous cell carcinomas and papillomas in the whole oral cavity in rats are less than 1%
in both males and females. In the 2-year bioassay, there was an increasing trend in these tumor
types in both male and female rats (Figure 3-15), with a 22% incidence in female rats at the highest
dose. Tumors of the small intestine of mice are also rare (historical rates of 2.6% and 0.6% in males
and females, respectively). These tumors were observed in all exposed groups of mice (including
3/49 at the lowest dose in males), with an incidence of >40% in the highest dose groups in both
sexes. One tumor each was observed in the control groups of male and female mice (leading to a
2% incidence for controls). In general, historically, rats are more prone to oral cancer development
than mice, and mice are more prone to neoplasia in the small intestine (Ibrahim etal.. 2021:
Chandra etal.. 2010) (Appendix D.2). The reason is unknown, but likely multifactorial in nature,
possibly involving differences in the microbiome (Ibrahim etal.. 2021).
3.2.3.3. Genotoxicity Evidence (AllRoutes)
Cr(VI) is a human lung carcinogen when inhaled. When ingested, Cr(VI) has been shown to
cause tumors in the GI tract in animals exposed in drinking water (NTP. 2008). Evidence relevant
to the potential key events and pathways involved in Cr(VI)-induced cancer via oral or inhalation
exposures was systematically identified (Section 1.2) and is summarized in the next section, 3.2.3.4,
and in Appendix C.3.2 organized by the key characteristics of carcinogens (Smith etal.. 2016). The
majority of studies informing these key events were not evaluated for risk of bias and sensitivity
concerns. However, a set of genotoxicity studies with designs best suited to examining whether and
to what extent Cr(VI)-induced tumorigenesis involves a mutagenic MOA were prioritized and
subject to an additional level of review (discussed in more detail below). This includes studies
measuring gene or chromosomal mutation endpoints in occupationally exposed humans and
studies in experimental animals in inhalation or oral exposure scenarios. An increased focus of
analysis on these studies is warranted because the results of the analyses of whether Cr(VI) acts via
a mutagenic MOA for cancer influences dose-response decisions, including the application of age-
dependent adjustment factors (ADAFs) and low-dose linear extrapolation fU.S. EPA. 2005bl It is
also for this reason that this MOA analysis includes consideration of both GI and lung tumors;
although the hazard for lung cancer is not being revisited (see Section 3.2.3.1), a determination of
whether a mutagenic MOA is applicable to lung tumors is important to consider for dose-response.
The summary and evaluation of the mechanistic evidence most informative to evaluating the role of
mutagenicity is synthesized in the following sections. These studies were initially tagged as
mechanistic supplemental literature and prioritized for analysis as described in the Cr(VI) Protocol
fU.S. EPA. 20191. The inferences drawn from these syntheses form the basis of mutagenic MOA
analysis for carcinogenesis; this analysis, and whether a mutagenic MOA could be secondary to
tissue injury and compensatory proliferation induced by Cr(VI), are presented in Section 3.2.3.4,
"Cancer mode-of-action summary."
This document is a draft for review purposes only and does not constitute Agency policy.
3-75 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
A mutation is a permanent, transmissible change in the genetic material of an organism.
Mutations can be caused by alterations in the DNA sequence of a gene, as well as structural
(clastogenic) and numerical (aneugenic) chromosome alterations (Eastmond et al.. 20091.
Genotoxicity is a more comprehensive term, referring to the ability of an exogenous agent to alter
genetic material. Some genotoxicity assays directly measure mutations, while others measure DNA
damage; proficient DNA repair of these genetic alterations depends on many factors including the
type of genetic damage and the repair capacity of the individual. Although both terms will be used
in the following sections, the more inclusive term "genotoxicity" will be used when discussing
evidence for a mutagenic MOA in a broader context Consideration of both types of genotoxicity
evidence and a broad survey of multiple genotoxicity endpoints, when available, is important for a
comprehensive characterization of an agent's genotoxicity and the underlying genotoxic processes.
A large body of evidence is available to inform the genotoxicity of Cr(VI). Many genotoxicity
studies of Cr(VI) were conducted in test systems primarily used to screen substances for genotoxic
potential, which are useful but also include endpoints measuring genetic damage that may not
represent damage that is transmissible to daughter cells, or that use exposure methods that are
expected to result in higher concentrations of Cr(VI) at the cell membrane, including i.p.
administration and in vitro studies, leading to a greater quantity of Cr(VI) being taken up by the cell
and reduced to Cr(III). These studies have largely shown that intracellular Cr(III) can form DNA
adducts (reviewed in Zhitkovich (201111 and is mutagenic (reviewed in Chen etal. (20191. Wise et
al. f20181 and Nickens etal. f201011. This section is focused on the phenotypic evidence for Cr(VI)-
induced genotoxicity; the evidence for the mechanisms underpinning this genotoxicity, including
cellular uptake and reduction of Cr(VI) and the formation of Cr-DNA adducts and oxidative DNA
lesions, is summarized in the key events for the cancer MOA in Section 3.2.3.4. All studies informing
genotoxic mechanisms are considered, but a more specific and critical analysis below focuses on
evidence that most directly informs the ability of Cr(VI) to cause mutations in exposed humans.
Namely, using the study prioritization and evaluation criteria described in Appendix C.3.2.2, this
analysis focuses on studies that use assays to detect transmissible genetic damage (i.e., gene
mutation, micronuclei, and chromosomal aberrations) observed in exposed humans or in
mammalian test systems in vivo utilizing routes of exposure more applicable to humans (i.e., oral
and inhalation).
Human study evaluation summary
Studies of occupationally or environmentally exposed humans were considered to be most
relevant to a mutagenic MOA analysis for cancer if they included measures of gene mutation (prior
to tumorigenesis), micronuclei induction, or chromosomal aberrations. Human studies were only
considered if they included a comparison or referent population exposed to Cr(VI) at lower levels
(or no exposure/exposure below detection limits) or for shorter periods of time. Twenty-nine
studies of chromosomal aberrations and/or micronuclei in humans were identified according to
these prioritization considerations (see Appendix C.3.2.2 and Table C-47) and evaluated for risk of
This document is a draft for review purposes only and does not constitute Agency policy.
3-76 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
bias and sensitivity. Six studies were considered but deemed uninformative due to critical
deficiencies in either the exposure or outcome domain fWultsch etal.. 2017: Coelho etal.. 2013:
Sellappa etal.. 2010: Hilali etal.. 2008: Cid etal.. 1991: Sarto etal.. 19901 and are not discussed
further. The confidence judgments of the 23 informative studies, all conducted in workers
occupationally exposed to Cr(VI) that are expected to primarily be inhalation exposures, are
summarized in Table 3-16. All of the included studies were cross-sectional in design, comparing
individuals employed in occupations with known potential for chromium exposure to referent
groups, with the specific occupations, geographic locations, and exposure measurement methods
are summarized in Table 3-16. No oral exposure studies in humans were identified.
All studies were categorized as low or medium confidence. Among low confidence studies,
common reasons for decreased confidence ratings included small sample size/low power (Linqing
etal.. 2016: Wultsch etal.. 2014: Medeiros etal.. 2003: Benovaetal.. 2002: Vaglenovetal.. 1999:
Deng etal.. 1988: Husgafvel-Pursiainen etal.. 1982: Sarto etal.. 19821. presence of coexposures to
other occupational hazards that may also contribute to the observed genotoxicity (e.g., nickel) not
accounted for in the design or analysis fWultsch etal.. 2014: Oavyum etal.. 2012: Iarmarcovai etal..
20051. residual confounding due to minimal or no control for covariates fBalachandar etal.. 2010:
Vaglenov et al.. 1999: Koshi etal.. 19841. limitations in outcome assessment techniques or
inadequate reporting f Oayvum etal.. 2012: Balachandar etal.. 2010: Danadevi et al.. 2004: Koshi et
al.. 1984: Littorin etal.. 1983: Sarto etal.. 19821. and insufficient description to allow for evaluation
of potential for bias (including selection bias) f Linqing etal.. 2016: Oayvum etal.. 2012:
Balachandar etal.. 2010: Halasova etal.. 2008: Iarmarcovai etal.. 2005: Danadevi etal.. 2004:
Maeng etal.. 2004: Medeiros etal.. 2003: Benova etal.. 2002: Koshi etal.. 1984: Sarto etal.. 19821.
Among medium confidence studies, the most common reason for decreased confidence rating was
insufficient description to allow for evaluation of potential for bias (including selection bias) (Long
etal.. 2019: El Saftv etal.. 2018: Hu etal.. 2018: Halasova etal.. 20121.
For all studies, exposure to chromium was inferred based on occupational group. Given the
likelihood of chromium exposure in the industries evaluated an exposure assessment that did not
include a precise estimate of exposure levels was not identified as a primary limitation in most of
these studies for consideration with respect to mechanistic interpretations. However, lack of
certainty about differentiation of exposure between comparison groups (including the potential for
exposure among "controls") was a concern in several studies (Halasova etal.. 2012: Vaglenov etal..
1999: Migliore etal.. 1991: Deng etal.. 19881. In all but two studies (Sudha etal.. 2011: Migliore et
al.. 19911. chromium biomarker and/or air concentrations were also measured; these data served
to confirm that exposure occurred and provided context for results, but these measurements were
not a requirement in the evaluation criteria.
This document is a draft for review purposes only and does not constitute Agency policy.
3-77 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-16. Summary of included human cross-sectional occupational studies
for Cr(VI) mutagenic effects and overall confidence classification [high (H),
medium (M), low (L)] by outcome. Click to see interactive data graphic for rating
rationales.
Author (year)
Industry
Location
Exposure Measurement/Cr
Validation Measures
Gene mutation
Chromosomal
aberrations
Micronuclei
Balachandar et al.
(2010)
Tannery
India
Job category/urine and air samples
L
L
Benova et al. (2002)
Chrome electroplating
Bulgaria
Job category/urine and air samples
L
L
Danadevi et al.
(2004)
Welding
India
Job category/ blood samples
L
Deng et al. (1988)
Chrome electroplating
China
Job category/ air, hair, and stool
samples
L
El Saftv et al. (2018)
Chrome electroplating
Egypt
Job category/ serum samples
M
Halasova et al.
(2008)
Welding
Slovak
Republic
Job category/ blood samples
L
Halasova et al.
(2012)
Welding
Slovak
Republic
Job category/blood samples
M
Hu et al. (2018)a
Unspecified factory
work with exposure
to chromate
China
Job category/ blood and air samples
M
Husgafvel-
Pursiainen et al.
(1982)
Welding
Finland
Job category/ urine samples
L
larmarcovai et al.
(2005)
Welding
France
Job category/blood and urine
samples
L
Koshi et al. (1984)
Stainless-steel
welding
Japan
Job category/ urine samples
L
This document is a draft for review purposes only and does not constitute Agency policy.
3-78 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Author (year)
Industry
Location
Exposure Measurement/Cr
Validation Measures
Gene mutation
Chromosomal
aberrations
Micronuclei
Linqing et al. (2016)
Chrome electroplating
China
Job category/blood samples
L
Littorin et al. (1983)
Stainless-steel
welding
Sweden
Job category/ urine and air samples
L
U
Long et al. (2019)
Chromate production
China
Job category/blood samples
M
Maeng et al. (2004)
Chrome electroplating
and buffing
South
Korea
Job category/urine, blood, and air
samples
L
Medeiros et al.
(2003)
Stainless-steel
welders; Tannery
Portugal
Job category/plasma and urine
samples
L
Migliore et al.
(1991)
Tannery
Italy
Job category
L
Qavvum et al.
(2012)
Chrome electroplating
India
Job category/plasma samples
L
Sarto et al. (1982)
Chrome electroplating
Italy
Job category/urine samples
L
Sudha et al. (2011)
Welding
India
Job category
M
Vaglenov et al.
(1999)
Hydraulic machinery;
Chrome electroplating
Bulgaria
Job category/air, red blood cells,
urine samples
L
Wultsch et al.
(2014)
Chrome electroplating
Austria
Job category/whole blood samples
L
Xlaohua et al.
(2012)
Chromate production
China
Job category/urine, blood, air
samples
L
This document is a draft for review purposes only and does not constitute Agency policy.
3-79 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
aTwo other studies by the same group (Li et al., 2014a; Li et al., 2014b) reported the same micronucleus frequency
data and were tagged as "related to included study" supplemental material.
1 Synthesis of human genotoxicitv evidence
2 Among the 23 informative studies prioritized for evaluating mutagenicity, 16 evaluated
3 micronucleus incidence and 10 evaluated chromosomal aberrations (three studies evaluated more
4 than one of these endpoints). The study details are summarized in Table 3-17 and Appendix Table
5 C-47.
This document is a draft for review purposes only and does not constitute Agency policy.
3-80 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-17. Associations between Cr(VI) exposure and prioritized genotoxicity outcomes in epidemiology studies3
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
El Saftv et al. (2018)
Cross-sectional study in Egypt
Exposed: 41 electroplating
workers
Referents: 41 administrative
workers
26.68 (11.21)
Air (mg/m3)
Total Cr
Exposed: median: 15.5 (IQR: 19.0)
Referents: median: 0.2 (IQR: 0.4)
Blood (ug/L)
Exposed: 8.5 (1.3)
Referents: 4.1 (1.4)
In exfoliated buccal cells:
1" MN in exposed compared to controls
(p < 0.001)
1" serum Cr correlates with T* MN
1" serum 8-OHdG in exposed compared
to controls (p < 0.001)
Medium confidence
Halasova et al. (2012)
Cross-sectional study in Slovak
Republic
Exposed: 73 welders
Referents: 73 individuals
without known exposures
10.2 (1.7)
Blood (umol/L)
In cultured lymphocytes:
No significant differences in CAs
between exposed and control groups
1" CAs in individuals with Gln/GIn
genotype compared to Arg/GIn or
Arg/Arg genotypes in XRCC1 Arg399Gln;
more pronounced in Cr-exposed
workers (p = 0.01) (no correlation with
XRCC3 polymorphisms)
Medium confidence
Total Cr
Exposed: 0.07 (0.04)
Referents: 0.03 (0.007)
Hu et al. (2018)
Cross-sectional study in China
Exposed: 87 workers at factory
with chromate exposure
Referents: 30 administrative
workers
Median: 5.0
IQR: 7.0
Air (ng/m3)
Exposed: median: 15.5 (IQR:19.0)
Referents: median: 0.2 (IQR: 0.4)
Blood (ug/L)
Exposed: GM: 8.5 (1.3)
Referents: GM:4.1(1.4)
1" MN in peripheral lymphocytes in
exposed workers compared with
referent
Medium confidence
Long et al. (2019)
Medium confidence
Cross-sectional study in China
Exposed: 120 chromate
production facility workers
Referents: 97 unexposed
workers at the same factory
14.57 (5.85)
Blood (ug/L)
Exposed: median: 2.81 (IQR: 3.86)
Referents: median: 0.99 (IQR: 1.21)
1" MN frequency ratio in lymphocytes
of exposed
Interactions between Cr exposure and
MN frequency in lymphocytes for some
SNPs
This document is a draft for review purposes only and does not constitute Agency policy.
3-81 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
Sudha et al. (2011)
Cross-sectional study in India
Exposed: 66 welders
Referents: 60 general
population controls
Range: 5-20
NR
In exfoliated buccal cells:
1" MN frequency and comet tail length
in welders compared to controls;
increased with duration of work
(p < 0.05)
Medium confidence
Balachandar et al.
(2010)
Cross-sectional study in India
Exposed 1: 36 directly exposed
(DE) through tannery work
Exposed 2: 36 indirectly
exposed (IE) through
residential proximity to
tanneries
Referents: 36 unexposed
individuals
DE (tannery) workers
(% by duration)
0-5: 17%
5-10: 33%
10-15: 36%
15-20: 11%
20-25:3%
Air (mg/m3)
Cr(VI)
DE 0.021 (0.003)
IE: 0.013 (0.005)
Referents: 0.006 (0.001)
Urine
DE: 2.11 (1.01)
IE: 1.81 (0.88)
Referents: 0.54 (0.39)
(Units not provided)
In cultured lymphocytes:
1" CAs in DE group compared to IE
group and controls
1" MN among directly exposed subjects
compared to indirectly exposed &
controls; further elevated in those with
longer duration of exposure
1" mean tail length for comet assay in
DE group compared to IE group and
controls
Low confidence
Benova et al. (2002)
Low confidence
Cross-sectional study in
Bulgaria
Exposed: 15 chrome-plating
workers
Referents: 23 individuals (15
workers and 8 rural residents)
N by duration:
2-5: 3
6-10:1
11-15: 4
16-20: 4
>20: 3
Air (mg/m3)
Cr(VI)
High exposed workers: 0.0249 (SE:
0.004)
Low exposed workers: 0.0075 (SE:
0.001)
Referents: 0.0004 (SE: 0)
Urine (ng/L)
High exposed workers: 104.22 (SE:
27.51)
Low exposed workers: 18.63 (SE:
3.16)
Referents: 1.18 (SE:0.23)
In cultured lymphocytes and exfoliated
buccal cells:
No significant difference in frequencies
of CAs or SCEs in exposed workers
compared to controls
1" MN in workers compared to controls
(lymphocytes: p < 0.01; buccal:
p< 0.001)
This document is a draft for review purposes only and does not constitute Agency policy.
3-82 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
Danadevi et al. (2004)
Cross-sectional study in India
Exposed: 102 welders
Referents: 102 general
population controls
Range: 1-24
Blood (ug/L)
1" MN in exfoliated buccal cells
compared to controls (p < 0.001),
correlated with duration of work, age,
and Cr level in blood
1" mean comet tail length in whole
blood cells compared to controls
(p < 0.001)
Low confidence
Exposed: 151.65 (SD not provided)
Referents: 17.86 (SD not provided)
Deng et al. (1988)
Low confidence
Cross-sectional study in China
Exposed 1: 7 electroplating
workers exposed to chromium
Exposed 2: 7 electroplating
workers exposed to nickel
Referents: 10 officer workers
12.8 (range: 4-18)
Air (mg/m3)
Total Cr
Workers: 8 x 106
(SE: 3.7 x 10s)
Stool (ng/g)
Workers: 8.5 (SE: 3.2)
Hair (ng/g)
Workers: 35.7 (11.5)
In cultured lymphocytes:
1" CAs in chromium workers compared
to nickel workers & controls
1" SCE in chromium & nickel workers
compared to controls
Halasova et al. (2008)
Low confidence
Cross-sectional study in Slovak
Republic
Exposed: 39 welders
Referents: 31 individuals
without known exposures
10.2 (1.7)
Blood (umol/L)
Total Cr
Exposed: 0.07 (0.04)
Referents: 0.03 (0.007)
In cultured lymphocytes:
Nonsignificant T* CAs in exposed
compared to control groups
1" CAs (p < 0.05) in lymphocytes in
individuals with Gln/GIn genotype
compared to Arg/GIn or Arg/Arg
genotypes in XRCC1 Arg399Gln (no
correlation with XRCC3 polymorphisms)
Husgafvel-Pursiainen
et al. (1982)
Cross-sectional study in
Finland
Exposed: 23 welders
Referents: 22 employees at
printing company
21 (10)
Urine (umol/L)
Total Cr
Exposed: range: 0.20-1.55
In cultured lymphocytes:
No significant differences in frequency
of CAs or SCEs between welders and
controls
Low confidence
This document is a draft for review purposes only and does not constitute Agency policy.
3-83 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
larmarcovai et al.
Cross-sectional study in France
Exposed: 60 welders
n = 27 working in areas
"without any collective
protection device"
n = 33 working in places with
"smoke extraction systems
Referents: 30 office workers
Range: 0.5-45
Blood (ug/L)
In cultured lymphocytes:
1" MN in non-protected welders
compared to controls (p = 0.03)
1" mean comet tail length in welders at
the end of the work week (p < 0.001);
not significant at the start of the week
1" mean comet tail length in individuals
with Gln/GIn genotype compared to
Arg/GIn or Arg/Arg genotypes in XRCC1
Arg399Gln (no correlation with XRCC3
polymorphisms)
(2005)
Low confidence
Exposed: 123.8-145.8 (58.8-87.7)°
Referents: 92.0 (15.0)
Urine (ug/g creatinine)
Exposed: 18.6-33.0(11.0-21.4)°
Referents: 12.8 (6.6)
Koshi et al. (1984)
Low confidence
Cross-sectional study in Japan
Exposed: 51 stainless-steel
welders
Referents: 33 office/research
workers
12 (range: 5-20)
Urine (ug/L)
Exposed: 9.8 (9.2)
Referents: 4.2 (1.2) ug/L
In cultured lymphocytes:
1" CAs and SCEs in welders compared to
controls
Linaing et al. (2016)
Cross-sectional study in China
Exposed: 29 chrome-plating
workers
Referents: 29 workers without
chromate exposure history
NR
Blood (ug/L)
In cultured lymphocytes:
1" MN frequencies in workers
compared to controls (p = 0.0048)
No correlation between blood Cr
concentration and MN
4/ methylation of MT-TF and MT-RNR1
genes in mitochondrial DNA correlated
with blood Cr
Low confidence
Exposed: 15.2 (range: 2.1-42)
Referents: 4.6 (range: 0.2-28)
This document is a draft for review purposes only and does not constitute Agency policy.
3-84 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
Littorin et al. (1983)
Cross-sectional study in
Sweden
Exposed: 24 stainless-steel
welders
Referents: 24 matched
individuals without
occupational mutagenic
exposures
19 (range: 7-41)
Air (mg/m3)
CrVI
Exposed: 0.055
(range: 0.005-0.321)
Urine (umol/L)
Exposed: 47
(range: 5-155)
Referents: 1.5
(range: <0.4-7.0)
In cultured lymphocytes:
No significant differences in CAs or SCEs
between exposed and control groups
No significant differences in MN
between exposed and control groups
Low confidence
Maeng et al. (2004)
Low confidence
Cross-sectional study in South
Korea
Exposed: 51 male chrome-
plating/buffing workers
Referents: 31 office workers
9.1 (range: 0-40)
Air (mg/m3)
CrVI
Exposed: GM: 0.0032 (range:
0.0003-0.09)
Referents: GM: 3 x 10"5 (range:
1.4 x 10"5-6.1 x 10"5)
Blood (ue/dL)
Exposed: GM: 0.86 (range: 0.11-
8.99)
Referents: GM: 0.17 (range: 0.00-
0.67)
Urine (ue/g creatinine)
Exposed: GM: 12.82 (range: 0.66-
8.74)
Referents: GM: 3.39 (range: 0.40-
9.04)
In cultured lymphocytes:
Nonsignificant T* CAs detected by solid
Giemsa staining in exposed compared
with unexposed that were statistically
correlated with higher blood Cr
1" CAs with 1" frequency of
chromosome translocations in exposed
compared with unexposed (p < 0.01)
detected by FISH
1" MDA in blood plasma in exposed
compared to controls (p < 0.01)
This document is a draft for review purposes only and does not constitute Agency policy.
3-85 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
Medeiros et al. (2003)
Cross-sectional study in
Portugal
Exposed 1: 5 welders
Exposed 2: 33 tannery workers
Referents: 20-30 unexposed
individuals
NR
Plasma (ug/L)
In cultured lymphocytes:
1" MN in tannery workers compared to
controls (p < 0.01)
Nonsignificant T* MN in welders (n = 5)
1" DNA-protein crosslinks in tanners
(p < 0.001) and welders (p < 0.05)
compared to controls
Low confidence
Tannery workers: 2.43 (2.11)
Welders: 1.55 (0.67)
Referents: 0.41 (0.11)
Urine: (ug/g creatinine)
Tannery workers: 2.63 (1.62)
Welders: 1.90 (0.37)
Referents: 0.70 (0.38)
Migliore et al. (1991)
Low confidence
Cross-sectional study in Italy
Exposed: 17 tannery workers
and 2 reference groups from
different industries
NR
NR
No effects on MN frequency in cultured
lymphocytes
Qawum et al. (2012)
Low confidence
Cross-sectional study in India
Exposed: 100 electroplating
workers (grouped by length of
work)
Referents: 50 individuals with
no known exposure to nickel
or chromium
Group 1: range: 1-9
Group 2: range: 10-
25
Plasma (ug/L)
Group 1: 2.9 (0.8)
Group 2: 1.7 (0.6)
Referents: 0.6 (0.8)
In buccal cells of Group II compared to
Group 1, and in Group III compared to
Group II:
1" MN frequency (p < 0.05)
MN also correlated with Cr levels in
plasma (p < 0.01)
Sarto et al. (1982)
Cross-sectional study in Italy
Exposed: 38 plating factory
workers (bright plating and
hard plating)
Referent 1: 35 sanitary
workers
Referent 2:14 healthy blood
donors
Hard plating: 7 (3)
Bright plating: 9 (11)
Urine (ug/g creatinine)
In cultured lymphocytes:
1" CAs (mostly CSAs) among all exposed
bright platers (p < 0.001) and hard
platers (p < 0.01) compared to controls
1" SCEs in hard platers compared to
blood donors
Low confidence
Exposed—Hard plating: 10.0 (7.5)
Exposed—Bright plating: 6.1 (2.8)
Referents: 1.9 (1.4)
This document is a draft for review purposes only and does not constitute Agency policy.
3-86 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
Vaglenov et al. (1999)
Cross-sectional study in
Bulgaria
Exposed: 30 hydraulic
machinery workers (grouped
by high and low exposure) &
10 hospitalized electroplating
workers
Referents: 18 administrative
workers
Overall range: 4-25
High exposed mean:
11.63
Low exposed mean:
10.44
Air (mg/m3)
High exposed: 0.083 (SE: 0.010)
Low exposed: 0.043 (SE: 0.01)
Referents: 0.0003 (SE: 0.0001)
Ervthrocvtes (ue/L)
High exposed: 8.40 (SE: 1.93)
Low exposed: 4.31 (SE: 1.03)
Referents: 0.57 (SE: 0.05)
Urine (ng/L)
High exposed: 5.0 (SE: 1.52)
Low exposed: 3.97 (SE: 1.98)
Referents: 0.49 (SE: 0.06)
1" MN and binucleated cells carrying
MN in lymphocytes of exposed
compared to control
Correlations of Cr measured in air,
erythrocytes and urine, with higher MN
in lymphocytes
Low confidence
Wultsch et al. (2014)
Cross-sectional study in
Austria
Exposed: 22 chrome-plating
workers
Referents: 22 jail warden
controls
NR
Blood (ug/L)
In exfoliated cells of exposed chrome
platers compared to referent:
1" MN frequency in nasal cells
(p = 0.005)
No significant effect on MN frequency
in buccal cells (23% increase; p = 0.516)
1" nuclear anomalies in buccal and
nasal cells
Low confidence
Exposed: 2.3 (1.5)
Referents: 0.2 (0.2)
This document is a draft for review purposes only and does not constitute Agency policy.
3-87 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
Population
Duration of work in
exposed group
(mean (SD) yrs)
Cr measurements
(mean (SD) unless otherwise
indicated)
Endpointsb
Xiaohua et al. (2012)
Cross-sectional study in China
Exposed: 79 chromate
production workers
Referents: 112 peasant
volunteers without
occupational chromate
exposure
Mean: 14.89
SE: 8.65
Air (ng/m3)
Exposed: 13.01 (range:l.03-56.60)
Referents: 0.073 (range: 0.023-
0.235)
Blood (ug/L)
Exposed: 9.19 (range: 1.17-51.88)
Referents: 3.44 (range: 0.25-22.51)
Urine (ue/g creatinine)
Exposed: 17.03 (range: 2.78-97.23)
Referents: 2.49 (range: 0.39-26.82)
1" MN in binucleated blood cells in
exposed group compared to controls
Moderate correlations (0.353-0.517)
between BNMN and Cr concentrations
in blood, urine, air
Low confidence
GM = geometric mean; IQR = interquartile range; SE = standard error; CA = chromosomal aberration; MN = micronuclei; NR = not reported.
aStudies presented by study confidence (high to low) first, then alphabetically by author.
bSome endpoints reported by the same study but not included in the PECO are also included here for context, p-values are added to provide additional context
but should not be the sole focus for interpretation.
This study reported subgroup means and SD; therefore, this table reports the range of means and the range of SDs for these groups.
This document is a draft for review purposes only and does not constitute Agency policy.
3-88 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Micronuclei
Micronuclei are formed when dividing cells contain whole chromosomes or acentric
chromosome fragments that have lagged behind during anaphase, indicating aneuploidy or the
presence of chromosomal aberrations. Additional procedures to detect the presence of a
centromere in the micronucleus can distinguish between loss of a whole chromosome or
chromosome fragments. All prioritized studies in humans focused on the detection of micronuclei
or chromosomal aberrations in peripheral blood lymphocytes or exfoliated nasal or buccal cells
(epithelial cells inside the mouth/cheek). In humans, it has been shown that an increased
frequency of micronuclei in circulating blood is positively associated with an increased risk of
cancer (Bonassi et al. (2011b; 200711. In addition, micronuclei detected in exfoliated epithelial cells
from the oral buccal or nasal mucosa is an effective measure of genetic damage in directly exposed
tissues (Bonassi etal.. 2011a).
Among the 16 studies evaluating micronuclei, four were rated as medium confidence and 12
were rated as low confidence. All four of the medium confidence studies reported increased
micronuclei, with two studies reporting these increases in lymphocytes (Long etal.. 2019: Hu etal..
20181. and two reporting increases in buccal cells (El Safty etal.. 2018: Sudha etal.. 20111. These
studies included populations from several industries with chromium exposure including
electroplating, chromate production, and welding. While these studies compared groups defined by
job category, three of the four studies augmented the exposure assessment by including data from
supplemental biomarker and/or air measures that showed total Cr levels were higher in exposed
workers and in exposure settings, confirming that exposures occurred and providing context for
the positive results (Long etal.. 2019: El Saftv etal.. 2018: Hu etal.. 20181 (see Table 3-17).
Among the 11 low confidence studies, there were ten that reported increased micronuclei
for at least one cell type. Three evaluated buccal cells (Oayvum etal.. 2012: Danadevi etal.. 2004:
Benova et al.. 2002). six evaluated lymphocytes and/or leukocytes in peripheral blood (Linqing et
al.. 2016: Balachandar etal.. 2010: Iarmarcovai etal.. 2005: Medeiros et al.. 2003: Benova etal..
2002: Vaglenov etal.. 19991. and one evaluated nasal cells fWultsch etal.. 20141 (this study also
reported a slight nonsignificant increase in micronuclei in buccal cells). These studies were
comprised of populations exposed to chromium via welding, electroplating, hydraulic machinery,
and tanneries. These studies also confirmed exposure in biomarker and/or air measures of total Cr
or Cr(VI), though Linqing etal. (2016) did not detect a significant correlation between the increased
blood Cr levels and statistically significantly increased micronucleus frequency in exposed workers
(Table 3-17 and Appendix Table C-47). The potential direction of bias in these low confidence
studies could not be determined.
One low confidence study reported no significant effects on micronucleus endpoints. In this
study, Migliore etal. (1991). there is uncertainty regarding the potential for chromium exposure
among the tannery workers evaluated and no accompanying biomarker measurements to provide
confirmation; misclassification of individuals with regards to exposure group may produce bias
This document is a draft for review purposes only and does not constitute Agency policy.
3-89 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
towards the null. In the absence of quantitative measures of exposure, it cannot be determined
whether a negative result reflects low exposures, a lack of mutagenicity, or both.
Dose-response concordance for the observed increases in micronuclei was observed in
several studies, most reliably based on correlations between Cr levels measured in blood and
micronuclei in buccal cells in the medium confidence study by El Saftv etal. (20181 and also in the
low confidence studies by Oayvum etal. (20121 and Danadevi et al. (20041. Danadevi et al. (20041
also noted a correlation between Cr levels and duration of work and age. A correlation between
work duration and increased micronucleus frequency was also noted in buccal cells in the medium
confidence study by Sudha etal. f20111 and in lymphocytes in the low confidence study by
(Balachandar etal.. 20101.
Several of these studies also reported other significantly increased systemic genotoxicity
markers in exposed workers that may be coherent with the observed micronuclei increases,
including serum 8-OHdG fEl Saftv etal.. 20181 and comet tail length in blood cells f Sudha etal..
2011: Danadevi etal.. 20041.
Overall, all four medium confidence studies across different study populations and
industrial settings (Table 3-16) and covering both lymphocytes and exfoliated epithelial cells
provide evidence for an association between chromium exposure and increased micronuclei. These
results are supported by the large majority of the available low confidence studies. Despite their
limitations, low confidence studies provide supporting evidence for this endpoint in conjunction
with the conclusions from medium confidence studies. In addition, when looking broadly across
studies and evaluating the evidence base as a whole, concerns about any particular study deficiency
is attenuated given that ten of the 11 low confidence studies demonstrated increases in micronuclei
despite differences in population and exposure scenarios.
Chromosomal aberrations
Structural or numerical chromosomal aberrations, observable during metaphase in cells
undergoing mitosis, are typically detected using simple, solid-staining techniques that allow visual
identification of chromosome and chromatid breaks, but do not detect translocations or other more
complex forms of chromosomal damage. Use of G-banding techniques or molecular fluorescent
probes (e.g., FISH) increase the type and complexity of detectable cytogenetic damage. In humans,
it has been shown that an increased frequency of chromosomal aberrations in circulating blood is
positively associated with an increased risk of cancer (Bonassi etal.. 2008: Norppa et al.. 20061.
All included studies evaluating chromosomal aberrations were rated as low confidence
except for one medium confidence study, Halasova et al. (20121. that identified chromosomal
aberrations only within genetically susceptible populations but did not identify differences
between the broader exposed and control groups. It should be noted, however, that a concern for
bias towards the null due to potential insensitivity was identified for this study. The mean levels of
blood chromium among the exposed group in this study were low (0.07 [imol/L = 3.64 |ig/L) and
within the range reported for the referent groups in other studies of chromosomal aberrations
This document is a draft for review purposes only and does not constitute Agency policy.
3-90 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Toxicological Review ofHexavalent Chromium
(e.g., Maengetal. f20041: 2.0 |ig/L) and micronuclei (e.g., Linqing etal. f20161: 4.6 |J.g/L). Lack of
control for potential confounders is also a concern in this study fHalasova etal.. 20121.
Among the nine low confidence studies, six reported increased chromosomal aberrations
among exposed compared to unexposed individuals (Balachandar etal.. 2010: Halasova et al.. 2 0 0 8:
Maeng etal.. 2004: Deng etal.. 1988: Koshi etal.. 1984: Sarto etal.. 19821. These studies examined
individuals exposed to chromium in a range of settings, such as tanneries, mining, electroplating,
and welding. While several studies had deficiencies that pose substantial concern for bias, such as
limited evaluation of confounders or potential for selection bias fKoshi etal.. 1984: Sarto etal..
19821. others had deficiencies that primarily relate to sensitivity, such as small sample size and
unclear differentiation between exposure groups (Balachandar etal.. 2010: Halasova etal.. 2008:
Deng etal.. 19881. Identification of effects on chromosomal aberrations despite sensitivity
concerns in these studies that may bias results towards the null can provide stronger evidence of
effect despite the individual overall study evaluation ratings of low.
Three low confidence studies evaluating populations of welders or chrome-plating workers
reported no changes in chromosomal aberrations in exposed individuals compared to controls
fHalasova et al.. 2008: Benova etal.. 2002: Littorin etal.. 19831. It should be noted that two of these
studies may have limited power to detect the outcome of interest due to small sample size (Benova
etal.. 2002: Husgafvel-Pursiainenetal.. 19821.
Overall, while the evidence base is mostly consistent regarding the association between
chromium exposure and chromosomal aberrations across a variety of exposure scenarios,
biomarkers, and geographic regions, these observations are only available from studies rated as low
confidence and a single medium confidence study with mixed results. Although considering the
entire evidence base mitigates concerns about any particular deficiency in a single low confidence
study and some of these studies detected effects despite limitations in power and sensitivity
(Coelho etal.. 2013: Balachandar etal.. 2010: Halasova et al.. 2008: Deng etal.. 19881. it is difficult
to draw definitive judgments from the predominantly low confidence evidence base on
chromosomal aberrations.
This document is a draft for review purposes only and does not constitute Agency policy.
3-91 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Animal study evaluation summary
As described above in the introduction to the mutagenic MOA evaluation approach and in
Appendix C.3.2.2, the available animal evidence prioritized as the most relevant for informing a
mutagenic MOA analysis for cancer includes measures of gene mutation (prior to tumorigenesis),
micronuclei induction, and chromosomal aberrations. These studies were prioritized for evaluation
and synthesis in this section based on study design, namely if they were conducted in animals
exposed via inhalation or intratracheal instillation, or via the oral route, including drinking water,
diet, or gavage. Gavage and intratracheal instillation exposures were considered with the
acknowledgment that these dosing regimens condense the exposure time while potentially
inhibiting reduction kinetics leading to increased point-of-contact Cr(VI) exposure. Studies
measuring DNA damage or indicators of DNA damage or using more direct methods of chemical
administration (i.e., i.p. injection) were not prioritized but are still considered as supplemental
evidence to mutation and are summarized in Appendix C.3.2.2.
Table 3-18 summarizes the overall classification judgments for 15 animal studies of Cr(VI)-
induced mutagenicity via inhalation or oral exposures (reporting 16 total endpoints) that were
prioritized for evaluation. These consist of six studies measuring mutation frequency following
short-term and subchronic exposures to drinking water (Aoki etal.. 2019: Thompson etal.. 2017:
Thompson etal.. 2015c: O'Brien etal.. 2013: Kirpnick-Sobol etal.. 20061 or via intratracheal
instillation (Cheng et al. (2000: 19981: the preliminary and primary study results were reported in
two separate publications), three studies detecting chromosomal aberrations following a single
gavage dose fMukheriee etal.. 1997: Sarkar etal.. 1996.19931. six studies measuring micronucleus
incidence following acute, short-term, or chronic drinking water and/or gavage exposures
(Thompson et al.. 2015b: O'Brien etal.. 2013: NTP. 2007: De Flora etal.. 2006: Mirsalis etal.. 1996:
Shindo etal.. 19891. and one dominant lethal test in rats exposed via intragastric instillation (Marat
etal.. 20181. Three additional studies reporting the micronucleus test in rats (Elshazlv etal.. 20161
and chromosomal aberrations in mice (Mukheriee etal.. 1999: Go'ldina et al.. 19891 were found to
be uninformative for these endpoints and were not considered further.
The endpoints specific to mutation, identified using the prioritization criteria for
mutagenicity evidence relevant to cancer (Appendix C.3.2.2), were evaluated separately from any
apical endpoints that may have also been reported in these animal bioassays (see Table 3-8). The
majority of the prioritized studies are in vivo assays considered to be complementary, as the
transgenic rodent assay primarily detects point mutations and small deletions (Dobrovolskv and
Heflich. 20181. and the micronucleus assay can detect chromosomal aberrations and aneuploidy
fHavashi. 20161. Following study evaluation, all 15 studies of mutagenic endpoints were
categorized as low confidence.
For many of the considered studies (Aoki etal.. 2019: Thompson etal.. 2017: Thompson et
al.. 2015c: Thompson et al.. 2015b: O'Brien etal.. 2013: NTP. 2007: De Flora etal.. 2006: Mirsalis et
al.. 19961. the concern was not with the "quality" of the study, but rather with study designs that
This document is a draft for review purposes only and does not constitute Agency policy.
3-92 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Toxicological Review ofHexavalent Chromium
were not optimized for genotoxic endpoints and thus lacked sensitivity for detecting an effect if one
were to be present, leading to deficiencies in the exposure sensitivity domain. According to the test
guidelines (TG) adopted by the Organisation for Economic Cooperation and Development (OECD)
for the transgenic rodent assay (TG 488, (OECD. 202011 and the mammalian erythrocyte
micronucleus test (TG 474, (OECD. 2016all. the two endpoints reported in most of the prioritized
studies, these studies should include a range of doses with the top dose representing the maximum
tolerated dose (MTD) that produces non-lethal toxicity in the animals (or, if not achievable, a daily
dose of 1000 mg/kg for a 28 day administration)29. This is to ensure the study is capable of
characterizing the mutagenic potential of the chemical on the target tissue(s) by confirming the
substance has reached the target tissue at levels high enough to induce toxicity, which is often the
bone marrow for standard micronucleus tests in polychromatic erythrocytes. Testing for
mutagenicity up to toxic levels is particularly important for increasing confidence in null findings in
vivo for a substance known to be mutagenic in vitro, such as Cr(VI). The motivation for selecting a
dose range to specifically study the induction of mutagenic effects at the same dose levels (albeit
with shorter exposure durations) that caused preneoplastic lesions and tumors in these animals
(e.g., up to 31.1 mg/kg-d Cr(VI) in female mice) is understandable. However, a bioassay properly
designed to detect potential mutagenic effects from ingested Cr(VI)30, a known carcinogen and a
mutagen via other routes of exposure, was not identified.
Other concerns about the ability of these studies to appropriately characterize mutagenicity
also contributed to their low confidence ratings. Deficiencies in the outcome sensitivity domain
included studies that counted too few plaque-forming units in the transgenic rodent assay (Cheng
et al. (2000; 199811 or polychromatic erythrocytes in the micronucleus assay fO'Brien etal.. 2013:
Shindo etal.. 19891. a mutation frequency background too high to reliably detect an effect fO'Brien
etal.. 20131. or failed positive controls (Thompson etal.. 2015b). A few studies were deficient in
results display sensitivity, including a failure to account for litter effects in a mutation study of
29TG 474 (OECD. 2016a): "The study should aim to identify the maximum tolerated dose (MTD), defined as
the highest dose that will be tolerated without evidence of study-limiting toxicity, relative to the duration of
the study period (for example, by inducing body weight depression or hematopoietic system cytotoxicity, but
not death or evidence of pain, suffering or distress necessitating humane euthanasia. The highest dose may
also be defined as a dose that produces toxicity in the bone marrow (e.g., a reduction in the proportion of
immature erythrocytes among total erythrocytes in the bone marrow or peripheral blood of more than 50%,
but to not less than 20% of the control value)...If the test chemical does not produce toxicity in a range-
finding study or based on existing data, the highest dose for an administration period of 14 days or more
should be 1000 mg/kg body weight/day, or for administration periods of less than 14 days, 2000
mg/kg/body weight/day." TG 488 (OECD. 2020): "The top dose should be the Maximum Tolerated Dose
(MTD). The MTD is defined as the dose producing signs of toxicity such that higher dose levels, based on the
same dosing regimen, would be expected to produce lethality."
30There were issues with Cr(VI) palatability at high drinking water concentrations (above ~90 mg/L Cr(VI) in
the NTP (2007) strain comparison study and at higher doses in the toxicity study), but in these cases it would
also be acceptable to use gavage administration to confirm delivery of a sufficient dose of Cr(VI). Only one
study included a gavage-administered dose that reached sufficient bone marrow toxicity, but this study was
judged low confidence due to deficiencies in the reporting, confounding, and endpoint sensitivity domains
fShindo etal.. 19891.
This document is a draft for review purposes only and does not constitute Agency policy.
3-93 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 exposures in mice in utero fKirpnick-Sobol etal.. 20061. not reporting the total number of cells
2 scored for micronuclei fO'Brien etal.. 20131. or pooling total micronuclei from multiple animals
3 (Thompson et al.. 2015bl. One dominant lethal test identified did not report the strain of animals,
4 test compound, or vehicle used (Marat etal.. 20181. And three low confidence studies were
5 identified that used a single gavage dose of Cr(VI) in mice to induce chromosomal aberrations in
6 order to test the effectiveness of anticlastogenic botanicals and were thus not optimized for an
7 objective assessment of genetic damage (Mukheriee etal.. 1997: Sarkar etal.. 1996.19931. The
8 prioritized studies are summarized in Table 3-19.
Table 3-18. Summary of prioritized animal studies for investigating Cr(VI)-
induced mutagenicity and overall confidence classification [high (H), medium
(M), low (L)] by endpoint3. Click to see interactive data graphic for rating
rationales.
Mutagenic endpoints
Author (year)
Species (strain)
Exposure
duration
Exposure route
Gene mutation
Chromosomal
aberrations
Micronuclei
Dominant Lethal
test
Aoki et al. (2019)
Mouse (transgenic gpt
delta), male
28 and 90 days
Drinking water
L
—
—
—
Cheng et al. (2000;
1998)
Mouse (C57BL/6 Big
Blue® and
nontransgenic C57BL/6),
female
1, 2, or 4 wks
post-instillation
Intratracheal
instillation
L
De Flora et al. (2006)
Mouse (BDFi), male and
female;
Mouse (Swiss albino)
pregnant dams and
fetuses
20 or 210 days
or pregnancy
duration
Drinking water,
gavage, i.p.
L
Kirpnick-Sobol et al.
(2006)
Mouse
(C57BL/6Jpun/pun),
pregnant dams and
offspring
GD 10.5-20.5
Drinking water
L
Marat et al. (2018)
Rat ("mature white
outbred"), male
60 days
Intragastric
administration
-
-
-
L
Mirsalis et al. (1996)
Mouse (Swiss-Webster),
male and female
2 days
Drinking water,
gavage
—
—
L
—
Mukheriee et al.
(1997)
Mouse (Swiss albino),
male
Bolus dose
(acute)
Gavage
—
L
—
—
NTP(2007)
Mouse (B6C3Fi), male
and female;
90 days
Drinking water
-
-
L
-
This document is a draft for review purposes only and does not constitute Agency policy.
3-94 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Mutagenic endpoints
Author (year)
Species (strain)
Exposure
duration
Exposure route
Gene mutation
Chromosomal
aberrations
Micronuclei
Dominant Lethal
test
Mouse (B6C3Fi, BALB/c,
am3-C57BL/6), male
O'Brien et al. (2013)
[related study:
(Thompson et al.,
2011)1
Mouse (B6C3Fi), female
90 days
Drinking water
L
L
Sarkar et al. (1993)
Mouse (Swiss albino),
male
Bolus dose
(acute)
Gavage
—
L
—
—
Sarkar et al. (1996)
Mouse (Swiss albino),
male
Bolus dose
(acute)
Gavage
—
L
—
—
Shindo et al. (1989)
Mouse (MS/Ae and CD-
1), male
Bolus dose
(acute)
Gavage, i.p.
—
—
L
—
Thompson et al.
(2015b)
Mouse (B6C3Fi), female
7 days
Drinking water
—
—
L
—
Thompson et al.
(2015c)
Rat (transgenic Big
Blue® TgF344), male
28 days
Drinking water
L
—
—
—
Thompson et al.
(2017)
Rat (transgenic Big
Blue® TgF344), male
28 days
Drinking water
L
-
-
—
aStudies excluded due to critical deficiency in one or more domains: Elshazly et al. (2016), Mukherjee et al. (1999),
and Go'ldina et al. (1989).
Synthesis of animal ge no toxicity evidence
The studies prioritized for being most informative for a mutagenic MOA analysis are
summarized in Table 3-19.
Table 3-19. Prioritized genotoxicity studies in animals exposed to Cr(VI)
Reference
System/
Exposure
Endpoint/Resultsa
Comments
Tests in lung tissue
This document is a draft for review purposes only and does not constitute Agency policy.
3-95 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
System/
Exposure
Endpoint/Results3
Comments
Cheng et al.
(2000; 1998)
Low confidence
Mouse, transgenic
C57BL/6 Big Blue®, female
Intratracheal instillation
(single administration): 0,
1.7, 3.4, or 6.8 mg/kg
Cr(VI)
Measured mutation
frequency in lung at 1, 2,
or 4 weeks post-exposure
Significantly increased
mutation frequency at all
doses; increased with dose
and duration post-treatment
Mutation spectrum:
increased frequency of G:C to
T:Atransversions, associated
with oxidative damage
Preliminary experiment identified
doses >6.75 mg/kg were lethal
Potentially underpowered with 4
mice per dose group
Positive control not concurrently
tested with Cr(VI)-treated group
Inconsistent/low numbers of
PFUs scored per animal
Spontaneous mutations primarily
G:C to A:T transitions
Tests in Gl tissue
Aoki et al. (2019)
Low confidence
Mouse, transgenic gpt
delta, male
Drinking water, 28 d: 0, 30,
or 90 mg/L Cr(VI) (0,13, or
30 mg/kg-d Cr(VI))
Drinking water, 90 d: 0, 3,
10, or 30 mg/L Cr(VI) (0,
1.6, 6, or 17 mg/kg-d
Cr(VI))
Measured mutation
frequency in duodenum at
28 and 90 days
In mouse duodenum:
No increased mutation
frequency (gpt delta locus)
relative to control at 28 or 90
d
Mutation spectrum: slightly
increased A:T to T:A
transversions at 28 d but not
at 90 d (significance
unknown)
Study selected doses based on
NTP 2-yr bioassay and did not
include a top MTD, potentially
biasing toward the null
Positive control not concurrently
tested with Cr(VI)-treated groups
90-d study potentially
underpowered with 4 mice per
dose group
Spontaneous mutations primarily
G:C to A:T transitions
Positive control potassium
bromate (but not Cr(VI)) had
increased G:C to T:A
transversions, associated with
oxidative damage
Thompson et al.
(2015c)
Low confidence
Rat, transgenic Big
Blue((R)) TgF344, male
Drinking water: 180 mg/L
Cr(VI), 28 d
In oral mucosa (upper inner
gingiva and adjacent palate
tissue and the upper outer
gingiva and adjacent buccal
tissue):
No increase in mutation
frequency (ell gene) relative
to control
Study used single dose group
based on NTP 2-yr bioassay top
dose and did not include a top
MTD, potentially biasing toward
the null
Cr levels in the gingival/ buccal
and gingival/palate regions were
0.66 and 1.0 |Jg/g, respectively,
compared to untreated Tg344
rats, which were 0.17 and 0.33
Hg/g respectively in the
gingival/buccal and
gingival/palate regions
Authors reported in vitro results
showing enriched responses for
p53, cell proliferation and
apoptosis
This document is a draft for review purposes only and does not constitute Agency policy.
3-96 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
System/
Exposure
Endpoint/Results3
Comments
Thompson et al.
(2017)
Low confidence
In duodenum:
No increase in mutation
frequency (ell gene) relative
to control
Study used single dose group
based on NTP 2-yr bioassay top
dose and did not include a top
MTD, potentially biasing toward
the null
Positive control not concurrently
tested with Cr(VI)-treated group
Rat small intestine is not a tumor
target tissue
O'Brien et al.
(2013)
Low confidence
Mouse, B6C3F1, female
Drinking water:
0, 0.1, 1.4, 4.9, 20.9, 59.3,
and 181 mg/L Cr(VI)
(0, 0.024, 0.32, 1.1, 4.6,
11.6, or 31.1 mg/kg-d
Cr(VI))
7 or 90 d
(Continued analysis of
tissues from Thompson et
al. (2011))
Micronucleus assay, in crypt
and villous cells from scraped
duodenal epithelium:
No increase in micronucleus
frequency in crypt cells
Statistically significantly
increased micronuclei in
villous cells from animals
exposed to 11.6 mg/kg-d
Cr(VI) for 90 days or 31.1
mg/kg-d Cr(VI) for 7 or 90
days
ACB-PCR, in scraped
duodenal epithelium:
No induction of GGT to GAT
mutations in KRAS codon 12
detected by ACB-PCR relative
to control
Micronucleus assay:
No baseline incidence of
micronuclei established in these
tissues
Crypt cell data pooled from all
animals per dose group and large
variation in total cells counted
per dose
Total number of villous cells
analyzed not presented
ACB-PCR:
High background mutant
frequency
Both endpoints:
Study selected doses based on
NTP 2-yr bioassay and did not
include a top MTD, potentially
biasing toward the null
Positive control not concurrently
tested with Cr(VI)-treated group
Thompson et al.
(2015b)
Low confidence
Mouse, B6C3F1, female
Drinking water:
0,1.4, 20.9, and 180 mg/L
Cr(VI)
(0, 0.32, 4.6, and 31.1
mg/kg-d Cr(VI))
7 d
In duodenal crypts (villi not
reported):
No increase in micronucleus
frequency relative to control
No effect on levels of yH2AX
Study selected doses based on
NTP 2-yr bioassay and did not
include a top MTD, potentially
biasing toward the null
No baseline MN incidence
established for these tissues,
positive control DMH was null,
number of cells analyzed
inadequate to measure an effect
21 and 180 mg/L Cr(VI)
significantly increased the
number of crypt enterocytes,
although no increase in crypt
mitotic activity was detected
No aberrant crypt or villous foci;
no apoptosis in crypt cells
This document is a draft for review purposes only and does not constitute Agency policy.
3-97 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
System/
Exposure
Endpoint/Results3
Comments
Tests in other tissues
Kirpnick-Sobol et
al. (2006)
Low confidence
Mouse, C57BL/6Jpun/pun,
female
Drinking water: 0, 22, or
44 mg/L Cr(VI) at 10.5 to
20.5 days postcoitum
(average dose of 4.4 or 8.8
mg/kg-day)
In 20-day-old offspring
harvested to visualize
eyespots corresponding to
DNA deletions in their retinal
pigment epithelium (RPE):
Increased deletions with
dose (p < 0.01)
Failed to account for litter
effects, potentially biasing away
from the null
No information on blinding,
concerning for this type of assay
that requires manual counting of
eyespots
Positive control not concurrently
tested with Cr(VI)-treated group
No signs of toxicity observed
Marat et al.
(2018)
Low confidence
Rat, white outbred males
Intragastric
administration, 1 mg Cr/kg
body mass, single dose, 60
days prior to mating with
virgin female rats
Survival of F1 fetuses from F0
males exposed to Cr(VI):
Ratio of live fetuses in the
Cr(VI) treatment group
compared to the control
group = 0.665 indicating
increased dominant lethal
mutation frequency in
exposed male rats
Deficiencies in reporting and
information on lab
proficiency/reproducibility
Study also reported increased
micronucleus frequency in bone
marrow in rats exposed to a
single i.p. dose of K2Cr207
NTP(2007)
Low confidence
Study 1: Mouse, B6C3Fi
(5/sex/group)
Drinking water: 0, 21.8,
43.6, 87.2, 174.5, or 350
mg/L Cr(VI), 90 d
NTP estimated daily doses
at 0, 3.1, 5.2, 9.1, 15.7, or
27.9 mg Cr(VI)/kg
In peripheral blood:
B6C3Fi: No effect on %MN
NCEs (males: p = 0.857;
females: p = 0.158)
The reduction of PCE/NCE ratio in
treatment groups was slight,
indicating mild bone marrow
toxicity, though this did not
increase with dose
Study 2: Mouse, B6C3Fi
(5/group), BALB/c
(5/group), and am3-
C57BL/6 (10/group), males
Drinking water: 0, 21.8,
43.6, or 87.2 mg/LCr(VI),
90 d
NTP estimated average
daily doses at 0, 2.8, 5.2,
or 8.7 mg Cr(VI)/kg
In peripheral blood:
B6C3Fi: NTP determined this
result to be equivocal due to
a trend test p-value very
nearly significant (p = 0.031;
a level = 0.025) and a
significant response
(p = 0.0193) in the highest
dose group of 87.2 mg/L.
BALB/c: No effect on %MN
NCEs (p = 0.680)
om3-C57BL/6: f %MN NCEs
(p < 0.001)
No effect on PCE/NCE ratio and
no clinical signs of toxicity
observed; failure to include an
MTD potentially biases toward
the null
om3-C57BL/6 transgenic mice
intended to measure mutation
frequency, but technical
difficulties prevented completion
of this study
This document is a draft for review purposes only and does not constitute Agency policy.
3-98 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference
System/
Exposure
Endpoint/Results3
Comments
Mirsalis et al.
(1996)
Low confidence
Mouse, Swiss-Webster,
M&F (5/sex/group)
Drinking water: 0,1, 5, or
20 mg/L Cr(VI), 48 h
Gavage: 20 mL/kg of 0,1,
5, or 20 mg/L Cr(VI), 2
doses, 24 and 48 h
In bone marrow:
No effect on %MN PCEs
Study did not include enough
information to accurately
calculate a dose for either
experiment
Study did not include a top MTD
(no effect on PCE/NCE ratio)
potentially biasing toward the
null
De Flora et al.
(2006)
Low confidence
Experiment 1:
Mouse, BDFi males
Drinking water: 0,10, or
20 mg/L Cr(VI), 20 d
Daily intake estimated at 3
and 6 mg/kg for 10 and 20
mg/L, respectively
Gavage or i.p.: 0 or 17.7
mg/kg Cr(VI), single dose,
24 h
Drinking water, in peripheral
blood, day 0, 5,12, and 20:
no effect on %MN NCEs
Drinking water, in bone
marrow, day 20: no effect on
%MN PCEs
Gavage, in bone marrow, 24
h: no effect on %MN PCEs
i.p. injection, in bone
marrow, 24 h: significant
increase in %MN PCEs
(p < 0.001)
Results of %MN NCEs at day 5-20
are uninterpretable; evaluation
of MN in mature erythrocytes
requires 4 weeks of continuous
treatment (Macgregor et al.,
1990)
Per os exposure groups did not
include a top MTD (no effect on
PCE/NCE ratio) potentially biasing
toward the null
Experiment 2:
Mouse, BDFi M&F
Drinking water: 0, 5, 50,
and 500 mg/L Cr(VI), 210 d
Daily intake estimates per
dose group, respectively:
Males: 1.65,16.5, and 165
mg Cr(VI)/kg
Females: 1.4,14, and 140
mg Cr(VI)/kg
In peripheral blood, day 0,
14, 28, 56, and 147: no effect
on %MN NCEs
In bone marrow, day 210: no
effect on %MN PCEs
Results of %MN NCEs at day 14
are uninterpretable; evaluation
of MN in mature erythrocytes
requires 4 weeks of continuous
treatment
Study did not include a top MTD
(no effect on PCE/NCE ratio)
potentially biasing toward the
null
Cr(VI) groups had similar drinking
water consumption at all doses
Slight decrease in body weight in
Cr(VI)-treated animals, especially
females
Experiment 3:
Mouse, pregnant Swiss
albino
Drinking water: 0, 5, or 10
mg/L Cr(VI) (as both
sodium dichromate
dihydrate (SDD) and
potassium dichromate
(PDC)) throughout
pregnancy duration, 18 d
i.p.: 0 or 17.7 mg/kg Cr(VI)
(as both SDD and PDC), PD
17, 24 h
In the bone marrow of dams
or in the liver or peripheral
blood of fetuses:
Drinking water: no effect on
%MN PCEs
i.p. exposures: micronuclei
significantly increased in all
tissues (p < 0.001)
Per os exposure groups did not
include a top MTD (no effect on
PCE/NCE ratio) potentially biasing
toward the null
No effect on fetus body weights
This document is a draft for review purposes only and does not constitute Agency policy.
3-99 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofHexavalent Chromium
Reference
System/
Exposure
Endpoint/Results3
Comments
Shindo et al.
(1989)
Low confidence
Mouse, MS/Ae and CD-I,
male
Gavage and i.p. injection:
2.68, 5.36, 10.7, 21.4,
42.8, and 85.7 mg
Cr(VI)/kg, bolus dose, 24 h
Gavage, in bone marrow: No
effect on %MN PCEs up to
acutely toxic oral gavage
doses that reduced PCE/NCE
ratio >50%
i.p. injection, in bone
marrow: Dose-dependent
increase in %MN PCEs and
decrease in PCE/NCE ratio
Calculated LD50s:
MS/Ae mice LD50: 80.3 mg
Cr(VI)/kg p.o., 13.4 mg Cr(VI)/kg
i.p.
CD-I mice LD50: 48.2 mg
Cr(VI)/kg p.o., 8.57 mg Cr(VI)/kg
i.p.
Study reported mean/SD per
dose group but did not report the
number of animals tested per
group
Baseline MN incidence extremely
low
Tests using Cr(VI) to induce genotoxicity
Mukheriee et al.
(1997)
Low confidence
Mouse, Swiss albino, male
Gavage: 7.1 mg Cr(VI)/kg
In bone marrow: significant
increase in chromosomal
aberrations (excluding gaps)
per cell (p < 0.01)
Study designed to test the
effectiveness of black tea in
preventing Cr(VI)-induced
clastogenicity
Sarkar et al.
(1993)
Low confidence
Mouse, Swiss albino, male
Gavage: 10.4 mg Cr(VI)/kg
In bone marrow: significant
increase in chromosomal
aberrations (excluding gaps)
per cell (p < 0.001)
Study designed to test the
effectiveness of chlorophyllin in
preventing Cr(VI)-induced
clastogenicity
Sarkar et al.
(1996)
Low confidence
Mouse, Swiss albino, male
Gavage: 7.1 mg Cr(VI)/kg
In bone marrow: significant
increase in chromosomal
aberrations (excluding gaps)
per cell (p < 0.05)
Study designed to test the
effectiveness of a spinach-beet
leaf extract in preventing Cr(VI)-
induced clastogenicity
aResults reported in the same study of genotoxicity endpoints or exposure routes that did not meet PECO have
also been included here for study context.
Gene mutations
Three studies in mice and rats were identified that used transgenic models to measure
mutation frequency in tumor target tissues after short-term or subchronic exposures to Cr(VI) in
drinking water fAoki etal.. 2019: Thompson etal.. 2015cl or in the lung following intratracheal
instillation (Cheng et al. (2000; 199811. The rodents contain transgenes (i.e., reporter genes
integrated into their genome) that can detect point mutations in any tissue studied. Cheng et al.
(2000: 19981 exposed female transgenic C57BL/6 Big Blue® mice to Cr(VI) via intratracheal
instillation, then measured the mutation frequency in the lacl transgene in lung tissues after 1, 2, or
4 weeks post-instillation. This study was found to be low confidence, primarily due to concerns
regarding the number of animals per dose group (four; five is the current minimum
recommendation (OECD. 202011 and the low and inconsistent number of plaque-forming units
evaluated, which were pooled per dose group and not reported per mouse. A preliminary study
determined that doses <6.75 mg/kg were not lethal; the second experiment included dose groups
This document is a draft for review purposes only and does not constitute Agency policy.
3-100 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
exposed to 0, 6.8, 3.4, and 1.7 mg/kg Cr(VI). The study reported increasing mutation frequency
with dose and time post-instillation; at the top dose after 4 weeks, the mutation frequency was 4.7-
fold of background levels, although there is some concern that the mutation frequency in the
vehicle control providing comparison was only assessed at 1 week post-treatment. The observed
increase of mutation frequency with time up to 4 weeks post-treatment corresponds to the average
cell turnover time of 28 days in lung tissue.
In a study conducted by members of the same group that created the transgenic gpt delta
mouse used in the study fNohmi etal.. 19961. Aoki etal. f20191 used male mice to examine
mutation frequency in the duodenum after 28 or 90 days of exposure via drinking water, at
concentrations of 0, 30, and 90 mg/L Cr(VI) (28 days) or 0, 3,10, and 30 mg/L Cr(VI) (90 days).
This group selected doses for both exposure periods based on the doses used in the NTP 2-year
bioassay with the exception of the lowest dose selected [3 mg/L Cr(VI)], which was less than the
lowest dose used by NTP [5 mg/L Cr(VI)]. No significant increase in mutation frequency was
detected after either time period. Although this study was otherwise well-conducted, deficiencies
in study design led to sensitivity concerns indicating potential for bias toward the null, leading to
overall low confidence. Use of concurrently run positive controls and inclusion of a dose that
induced clear clinical signs of toxicity would have increased confidence in the negative findings for
this assay.
A transgenic 28-day Big Blue® TgF344 rat study conducted by Thompson et al. (2017;
2015c) reported exposure to 180 mg/L Cr(VI) in drinking water also did not significantly increase
the mutant frequency in the gingival/buccal or gingival/palate regions in the oral cavity of rats or in
the rat duodenum. Similar to Aoki etal. (2019). the selection of a single Cr(VI) exposure group that
was not high enough to induce systemic toxicity in a short-term bioassay led to reduced confidence
in the sensitivity of this study design to detect a positive result and an overall low confidence
judgment In addition, the inclusion of rat duodenal tissues in this mutation assay provides little
value to mechanistic interpretation given the small intestine is not a tumor target tissue in rats.
In another low confidence mutation study by the same group, O'Brien etal. (2013)
conducted an analysis of KRAS codon 12 GGT to GAT mutations in mice, which are associated with
human colorectal cancer and metastasis (Tones etal.. 2017: Margonis etal.. 2015). The study used
tissues obtained from a previous subchronic bioassay in female mice (Thompson et al.. 2011). The
detection method, allele-specific competitive blocker polymerase chain reaction (ACB-PCR), was
developed and validated by one of the study authors fMckinzie and Parsons. 20021 and is a
sensitive method for detecting specific mutations. There were no statistically significant Cr(VI)
treatment-related increases measured for KRAS codon 12 GAT mutations; however, results were
difficult to interpret due to the lack of a concurrent positive control and the high background
mutation incidence (10-2 to 10"3) compared to previous findings of spontaneous mutation
frequency in mouse lung [3.88 x 10"4; (Meng etal.. 2010)]. rat distal colon [12.9 x 10"5; (Mckinzie
and Parsons. 2011)]. or human colonic mucosa [1.44 x 10~4; (Parsons et al.. 20101], Although this
This document is a draft for review purposes only and does not constitute Agency policy.
3-101 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
was a 90-day study, the dose levels tested in drinking water were selected to replicate those used in
the 2-year NTP bioassay [up to 180 mg/L Cr(VI)] and did not include a higher dose to determine
whether mutations would have been induced at toxic levels, reducing the sensitivity of this study to
detect an effect.
In a mouse model for measuring mutant frequency, the C57BL/6J pun/pun mouse strain takes
advantage of a naturally occurring mutation, a tandem duplication at the pink-eyed dilution (p)
locus, which causes the mice to have pink eyes (Brilliant etal.. 19911. Exposure to mutagens that
induce deletions via homologous recombination during fetal development can lead to reversion of
this unstable mutation back to black-pigmented cells, or eyespots, which are visible and
quantifiable. Although this assay developed by Schiestl et al. (1997) has not become part of the
standard testing battery for the detection of mutagens, it represents a highly sensitive assay for
detecting deletion mutations in single cells that are caused by transplacental exposures during
embryonic development. The Schiestl lab fKirpnick-Sobol etal.. 20061 exposed female C57BL/6J
pun/pun mice to 22 or 44 mg/L Cr(VI) in drinking water from 10.5 to 20.5 days post-coitum. Despite
a somewhat elevated background frequency (~10-4), dose-dependent, statistically significant
increases in mutations were observed in offspring (p < 0.01). However, the results of this study
were presented as the mean of individual pups without taking litter effects into account, potentially
overestimating the statistical significance of experimental findings (Haseman et al.. 2001) and
leading to bias away from the null. Therefore, this study was judged to be low confidence for this
outcome.
One rodent dominant lethal test was identified (Marat etal.. 2018). This assay detects gene
and/or chromosomal mutations produced in male germ cells during a pre-mating exposure period,
causing fetal death fOECD. 2016bl. Marat etal. T20181 reported a dominant lethal mutation
frequency of 0.665 by comparing the number of live F1 fetuses to control after exposure of F0 male
rats to 0.353 mg/kg-day Cr(VI) by oral gavage, with increases in pre- and post-implantation loss.
The dominant lethal test appears to have been conducted appropriately and detected a 10-fold
increase in post-implantation loss, but this study was found to be low confidence due primarily to
reporting deficiencies.
Micronuclei
Mutation studies can also measure increased incidences of heritable genetic alterations due
to numerical or structural changes in the chromosomes of animals exposed to Cr(VI) in vivo. Four
studies measuring changes in micronucleus frequency in the peripheral blood or bone marrow of
mice exposed to Cr(VI) via drinking water or oral gavage were identified. In a bioassay conducted
by NTP f20071. two micronucleus assays were conducted in mice exposed to Cr(VI) in drinking
water for 90 days; a minimum of 30 days is recommended for micronuclei in mature erythrocytes
to reach a steady state when a repeat-dose study design is used (Macgregor etal.. 1990). Study 1
exposed B6C3Fi male and female mice up to 350 mg/L Cr(VI), and Study 2 exposed male B6C3Fi,
BALB/c, and am3-C57BL/6 mice up to 87.2 mg/L Cr(VI). B6C3Fimice did not have increased
This document is a draft for review purposes only and does not constitute Agency policy.
3-102 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
frequencies of micronuclei in Study 1, but in Study 2, the result was considered equivocal due to a
nearly statistically significant increased trend (p = 0.031; the one-tailed trend test required a
p < 0.025 for significance). For the two other strains tested in Study 2, BALB/c mice also showed no
increase in micronucleus frequency, but the top two dose groups of am3-C57BL/6 mice had
statistically significant increases in micronuclei (p = 0.0025 and 0.0001 at 43.6 and 87.2 mg/L,
respectively), as well as a statistically significant trend (p < 0.001), with no evidence of bone
marrow toxicity. Although 5 animals per dose group is the minimum required for this test, it is of
note that the micronucleus test with the only clear, statistically significant positive result reported
by NTP (20071. in am3-C57BL/6 mice, tested twice as many animals (10/dose group), increasing
the power of this study to detect an effect This transgenic strain of mice was specifically included
to perform an analysis of mutation frequency that was unsuccessful due to technical difficulties;
however, there is no known reason to suspect that the endogenous genome of transgenic mice
would be unusually sensitive to clastogenic or aneugenic damage, and no data exist to suggest
strain-specific susceptibility.
The interpretation of negative results for the hazard identification of micronucleus
incidence in erythrocytes requires confirmation that the test agent reached the bone marrow at a
sufficient dose to induce erythropoietic toxicity; although it is possible for an agent to reach the
bone marrow without inducing toxicity, the OECD Test Guidelines (OECD. 2016a) recommend that
the highest dose should reduce the percentage of polychromatic erythrocytes (PCEs, also known as
reticulocytes) among total erythrocytes (normochromatic erythrocytes, or NCEs) by at least 50% to
ensure that any null findings can be interpreted as indicating a lack of genotoxic effect and not a
lack of exposure. In Study 1, a slight decrease in %PCEs among total NCEs was noted, indicative of
toxicity in the bone marrow, but this reduction was relatively small (19% and 25% reduction
compared to controls in male and female mice, respectively, at 350 mg/L) and did not increase with
dose. The mice in Study 2, exposed to lower concentrations of Cr(VI), had no decreases in %PCEs.
These results are consistent with those reported in NTP (2008) in female B6C3Fi mice, where no
changes in reticulocyte or nucleated erythrocyte counts were observed at 22 days, 90 days, 3
months, or 12 months following doses up to 180 mg/L in drinking water. It was noted in NTP
f20071 that the top doses from Studies 1 and 2 caused reductions in body weight gain (which the
study authors attributed to decreased palatability causing reduced food intake and not to Cr(VI)-
induced toxicity) indicating that higher doses could not have been administered in drinking water.
The NTP study, a well-conducted bioassay, was high confidence for the histopathological measures,
but for the reasons described above was found to be low confidence for this endpoint. Although
some toxicity was measured in the bone marrow in one (of two) arm of the study, a study design
including more animals and higher doses, perhaps administered via gavage to avoid palatability
issues, would have increased the sensitivity of this study to detect a positive result and/or
increased confidence in the negative/equivocal findings.
This document is a draft for review purposes only and does not constitute Agency policy.
3-103 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Two otherwise well-conducted in vivo micronucleus studies were found to be low
confidence for sensitivity concerns. Mirsalis etal. f!9961 dosed mice via drinking water and gavage
up to 20 mg/L Cr(VI) for 48 hours and did not detect an increase in micronucleus frequency or any
effect on PCE/NCE ratio in the bone marrow. In another large study using far higher doses for a
longer duration, De Flora etal. (2006) exposed mice to up to 500 mg/L in drinking water for
210 days in addition to exposures to pregnant dams of 10 mg/L in drinking water for the duration
of pregnancy. However, no increased incidence of micronuclei or effect on PCE/NCE ratio was
observed in the peripheral blood or bone marrow of exposed adults or in the liver or blood of
fetuses exposed in utero. In another branch of this study, De Flora et al. f20061 also dosed mice
with single i.p. injections of 17.7 mg/kg Cr(VI), which produced positive results for micronucleus
induction; these subtoxic exposures were considered positive controls for the route comparison
study, emphasizing the importance of pharmacokinetic considerations for Cr(VI) exposures. This
study also screened NCEs from peripheral blood for micronuclei after 10 or 20 mg/L drinking
water exposures for 20 days but these data are not considered (i.e., uninformative) as this exposure
duration is insufficient for detecting micronuclei in mature erythrocytes fMacgregor etal.. 19901.
One study, Shindo etal. f!9891. did include a top dose (85.7 mg Cr(VI)/kg) that reached
sufficient bone marrow toxicity. This well-conducted study was part of a larger effort by the
Collaborative Study Group for the Micronucleus Test to establish best practices for this assay. The
group conducted a pilot test to determine LD50s for each strain and route (oral and i.p.). A
micronucleus test was then conducted, finding no increases in micronucleus frequency from acute
oral exposures that reached a maximum tolerated dose in each strain. This study, however, was
determined to be low confidence due to lack of reporting the number of animals tested and not
establishing a sufficient background level of micronucleated PCEs to ensure adequate detection
sensitivity in the study; for the CD-I mice, the background micronucleus frequency was zero.
While the micronucleus assay has been traditionally performed in PCEs from peripheral
blood or bone marrow, it has been developed for use in other tissues provided the test is optimized
for sensitivity (e.g., ensuring the test captures cells during the first cell division post-exposure).
Notably, some GI tract mutagens [e.g., N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), N-methyl-N-
nitrosourethane (NMUT)], do not show increased micronucleus frequency in the peripheral blood
or bone marrow due to pharmacokinetic considerations, and adapting the MN assay for use in the
GI tract, where the cellular turnover rate is 3-5 days, has yielded positive results for GI carcinogens
known to be mutagenic (e.g., Okada et al. (2019)). Only two studies, conducted by the same group,
were identified that specifically measured micronuclei in duodenal epithelial cells of mice exposed
to Cr(VI) in drinking water fThompson etal.. 2015b: O'Brien etal.. 20131.
O'Brien etal. (2013) identified micronuclei as well as mitotic and apoptotic cells in fully
intact crypts from formalin-fixed and paraffin-embedded duodenal tissues obtained from a
previous subchronic bioassay in female mice (Thompson etal.. 2011). Because the bioassay had
been conducted previously, appropriate positive controls were not run concurrently, which would
This document is a draft for review purposes only and does not constitute Agency policy.
3-104 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
be useful for establishing proficiency in this less standardized tissue for this assay (compared to the
bone marrow or peripheral blood) for which no historical control data is available. In crypt cells,
zero micronuclei were reported for every dose group; this, and the lack of cytotoxicity detected in
these tissues even at the top dose (as measured by mitotic indices), indicate that the study was also
likely not sensitive enough to detect an effect in these tissues, leading to a judgment of low
confidence. At a minimum, scoring enough cells to detect a background rate for micronuclei
incidence would have helped increase confidence in these findings. In the villous cells, however,
statistically significantly increased numbers of cells with micronuclei were observed at the top dose
at day 7 and the two highest dose groups (60 and 180 mg/L Cr(VI)) at day 90. The micronuclei
counts were pooled per dose group, and the total number of cells scored was not reported, so
frequency cannot be determined, contributing to the low confidence judgment for this endpoint
In the second micronucleus study in the GI tract by this group, Thompson etal. (2015b) also
reported no increased micronuclei in duodenal crypt cells, but this study did not investigate
whether there were again increased micronuclei in villous cells. Concerns regarding the sensitivity
of the study design primarily involve the lack of establishing proficiency in this nonstandard assay.
Specifically, again, a baseline number of micronucleated cells in crypts and/or duodenal
enterocytes was not established; two exposure groups [180 mg/L Cr(VI), and the positive control,
65 mg/kg DMH] reported zero micronuclei in 5161 and 3153 cells, respectively. These groups had
lower numbers of cells analyzed than the vehicle control, which screened 6694 cells to identify four
micronucleated enterocytes (0.06%). Therefore, sufficient numbers of cells should have been
counted for all dose groups to increase confidence in the sensitivity of this assay to detect reliable
negative result In addition, the top dose did not induce a change in mitotic indices in the crypts
which was interpreted as a lack of cytotoxicity, indicating a lack of sensitivity for this endpoint (see
above discussion on sensitivity concerns for this assay).
Of primary concern regarding the sensitivity of Thompson etal. (2015b) is the lack of
micronuclei detection or other nuclear damage in animals dosed with 65 mg/kg DMH via gavage, or
the low, nonsignificant levels of micronuclei reported for i.p. injection of DMH. DMH
(1,2-dimethylhydrazine) is a colon carcinogen and alkylating agent widely used to induce colon
tumors in animal models fVanhauwaert etal.. 20011 and has been used as a positive control to
validate the micronucleus assay in the GI tract by other groups fCoffing etal.. 2011: Ohvama etal..
2002: Goldberg etal.. 1983). When administered via gavage or i.p., it induces increased
micronucleus frequency in the mouse colon (Ohvama etal.. 2002: Vanhauwaertetal.. 2001:
Goldberg etal.. 1983). Another study validating the micronucleus assay in GI tissues dosed mice
with DMH via gavage at 16.5, 33, 50, and 66 mg/kg and reported statistically significant, dose-
dependent increases in micronuclei in the duodenum and colon at all doses tested (Coffingetal..
20111. with micronuclei detected at a higher frequency in the duodenum than in the colon.
Therefore, this study was judged to be low confidence for this endpoint.
This document is a draft for review purposes only and does not constitute Agency policy.
3-105 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Chromosomal aberrations
Three studies from the same laboratory group (Mukheriee etal.. 1997: Sarkar etal.. 1996.
19931 used Cr(VI) exposure to induce genetic damage (as a positive control) in order to test
botanical extracts for their ability to mitigate chromosomal damage. Justification for the use of
Cr(VI) administered via gavage as a clastogenic agent in the bone marrow was provided by citing
results from the unpublished Ph.D. thesis by the first author in a review article f Singh etal.. 19901.
These are the only published studies of chromosomal aberrations following oral exposures to
Cr(VI) in animals that were not found to be uninformative, with all three studies reporting a
statistically significant increased incidence in the bone marrow compared to vehicle controls.
However, low confidence in these studies limits the ability to consider these results as informative
to the evaluation of mutagenicity from oral Cr(VI) exposure.
Integration of genotoxicitv evidence
Cr(VI) has been shown to be genotoxic and induce mutations in animals exposed via i.p.
injection and in vitro (Appendix Tables C-52 and C-53), providing mechanistic support for the
mutagenicity of Cr(VI) in these specific exposure scenarios. The evidence is less clear from in vivo
exposures, where pharmacokinetics can influence the ability and extent of Cr(VI) reaching the
tissues at concentrations capable of inducing detectable mutations. Therefore, genotoxicity studies
were prioritized to identify gene and chromosomal mutation studies in vivo using inhalation and
oral routes of exposure more relevant to humans.
Occupational exposure studies provide the most human relevant information for mutagenic
risk from Cr(VI) exposures. Consistent evidence of the mutagenic and genotoxic effects associated
with Cr(VI) exposure is provided by human studies across a diversity of study populations and
industrial settings (summarized inTable 3-17 and Appendix Table C-47). In studies detecting
transmissible genetic damage (i.e., micronuclei and chromosomal aberrations), increased
micronucleus frequency and, to a lesser extent, chromosomal aberrations were consistently
detected in the peripheral blood lymphocytes and exfoliated nasal and buccal epithelial cells of
exposed workers. These biomarkers have been shown to be positively associated with an increased
risk of cancer in humans (Bonassi et al. f2011b: 2008: 20071. fNorppa et al.. 200611. The data for
micronuclei and chromosomal aberrations are supported by additional evidence of genotoxic
responses to Cr(VI) exposure in humans, including DNA strand breaks, adducts, and crosslinks
(summarized in Appendix Table C-49).
No studies investigating genotoxicity in nonneoplastic lung tissues were identified in the
occupational exposure studies, but there was consistent evidence of increased micronucleus
frequency in buccal cells from workers occupationally exposed to Cr(VI) via chrome plating and
welding from two medium confidence studies fEl Saftv etal.. 2018: Sudha etal.. 20111 supported by
findings reported in three low confidence studies (Oayvum etal.. 2012: Danadevi et al.. 2004:
Benova et al.. 20021. Although occupational exposure occurs primarily via inhalation, changes in
This document is a draft for review purposes only and does not constitute Agency policy.
3-106 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
buccal cells can serve as a surrogate of direct Cr(VI) exposures to the GI tract in humans if ingested
Cr(VI) is able to reach those tissues in comparable amounts. Micronucleus frequency in these
workers was found to correlate with blood chromium levels (El Saftv etal.. 2018: Oavyum etal..
2012: Danadevi etal.. 20041. with work duration (Danadevi etal.. 20041. and with systemic
measures of DNA damage (e.g., 8-OHdG adducts, DNA strand breaks) (El Saftv etal.. 2018: Sudha et
al.. 2011: Danadevi et al.. 20041.
The experimental evidence base of gene and chromosomal mutation studies in animals is
smaller and composed entirely of low confidence studies (see Appendix Figures C-2 2 to C-2 5 for a
visual comparison of the reported findings from the oral exposure studies). One study was
identified that exposed animal lung tissues directly to Cr(VI) (via intratracheal instillation) and
reported dose-dependent increases in mutation frequency that increased with time from 1 to 4
weeks post-exposure (Cheng et al. (2000: 199811. Although this is only one low confidence study, it
is coherent with the findings in exposed humans and demonstrates the mutagenicity of Cr(VI) when
it comes into direct contact with tissues.
A slightly higher number of studies investigating mutagenicity via the oral route are
available. Four drinking water and/or gavage studies in mice measured micronucleus frequency in
the peripheral blood or bone marrow, the tissues most commonly studied in the micronucleus
assay due to the requirement of exposing actively dividing cells. Acute and subchronic studies by
NTP found mixed results among three strains of mice (NTP. 2007). while three additional studies
reported negative results in the bone marrow and/or peripheral blood fDe Flora et al.. 2006:
Mirsalis etal.. 1996: Shindo etal.. 1989). When interpreting genotoxicity results, particularly
negative results for a substance known to be mutagenic in other exposure scenarios, it is important
to confirm that the test substance reached the tissues tested. In vivo micronucleus assays are
designed to inform decisions regarding the mutagenic potential of a chemical (Eastmond etal..
2009). but if the doses selected for testing are lower than levels inducing some toxicity in the target
tissues, it is not possible to conclude the chemical would not be a mutagen at higher, subtoxic or
even toxic doses. For some of these studies, there is reason to suspect the exposures were not high
enough to achieve adequate tissue concentrations in the bone marrow. For example, although
pharmacokinetic findings by NTP f20071 indicate that Cr(VI) can reach the bone (or femur) at
concentrations above 10 mg/L Cr(VI) (approximately 1-2 mg/kg-d), two of these studies exposing
animals to concentrations up to 20 mg/L Cr(VI) in drinking water (De Flora etal.. 2006: Mirsalis et
al.. 1996) did not detect increases in micronuclei, and also did not detect decreases in the PCE/NCE
ratio, which would indicate toxicity in the bone marrow as specified by standard guidance for this
assay fOECD. 2016al. A third study, exposing animals via gavage to much higher doses (bolus dose,
up to 86 mg/kg Cr(VI)31), also reported negative findings that were observed in animals with
significant bone marrow toxicity, but this study was low confidence due to the lack of establishing a
31As a comparison, drinking water exposure of the top concentration of 350 mg/L Cr(VI) in Study 1 by NTP
f~20071 yields a daily dose of approximately 20 mg/kg-d, which is distributed over a longer period of time.
This document is a draft for review purposes only and does not constitute Agency policy.
3-107 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
background spontaneous rate of micronucleus incidence and not reporting the number of animals
tested fShindo etal.. 19891. Three other studies that used lower doses in a single gavage
administration to study chromosomal aberrations in the bone marrow of mice (7.1 or 10.4 mg
Cr(VI)/kg) did report positive findings (Mukheriee etal.. 1997: Sarkar etal.. 1996.19931. but low
confidence in these studies limits the ability to conclude that Cr(VI) can reach the bone marrow and
induce genotoxicity following a gavage exposure to Cr(VI).
The subchronic bioassay by NTP exposed male and female B6C3Fi mice to concentrations in
drinking water up to 350 mg/L Cr(VI) and did not detect increases in micronucleus frequency;
these animals had a slight induction of bone marrow toxicity, though decreased palatability in these
animals prevented these investigators from achieving a higher tissue concentration and led to the
selection of lower doses for their second study (NTP. 20071. There were some positive findings in
the second study, a mouse strain comparison of toxicity responses that dosed up to 87.2 mg/L
Cr(VI), with am3-C57BL/6 positive, BALB/c negative, and B6C3Fi nearly statistically significant
(see Table 3-19). This was despite a complete lack of toxicity in the bone marrow in these animals.
It is possible that, due to pharmacokinetic variability, Cr(VI) concentrations in drinking water do
not always reach sufficient concentrations in the bone marrow to induce significant mutagenicity in
that tissue, making this test in bone marrow tissues or cells a less sensitive measure for detecting
mutagenic potential in GI tissues following drinking water exposures. To enter bone marrow, orally
ingested Cr(VI) must escape 1) extracellular reduction in the GI tract lumen, 2) extracellular
reduction or cellular uptake in the liver and portal blood, and 3) extracellular reduction or cellular
uptake in systemic blood. Unlike gastrointestinal tract tissues which may be more directly exposed
to higher sustained levels of Cr(VI), the bone marrow may receive lower levels of exposure.
Evidence in tumor target tissues, as with the mutation study in the lung, is considered more
informative due to the point of contact uptake of Cr(VI) and intracellular reduction that initiates
potential carcinogenic pathways associated with Cr(VI) exposure (see Section 3.2.3.4). Three
studies directly investigated mutation frequency in tissues in the mouse duodenum or the rat oral
cavity following drinking water exposures. Two are gene mutation studies that examined target
tissues in the mouse duodenum (Aoki etal.. 20191 or the rat oral cavity (Thompson etal.. 2015c) of
transgenic rodents following subchronic drinking water exposures. Neither of these low confidence
studies observed significant increases in mutation frequencies. These studies designed the dosing
regimen based on the NTP 2-year bioassay and did not cover a range of doses that included a toxic
dose, which would have increased confidence in this study's ability to detect an effect.
A third study, O'Brien etal. (2013). did not detect an increase in KRAS codon 12 GGT to GAT
mutations in the mouse duodenal tissues. While KRAS mutations, primarily occurring in codons 12
and 13, have been identified in 35-45% of human colorectal cancers (Nguyen and Duong. 2018).
and many types of codon 12 mutations have been identified in tumors of the GI tract in humans
(Peng and Zhao. 2014). there are no data to establish the presence of codon 12 GGT to GAT
mutations in tumors from Cr(VI)-exposed workers, or in oral rat or duodenal mouse tumors
This document is a draft for review purposes only and does not constitute Agency policy.
3-108 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
induced by Cr(VI). Furthermore, a comparison study with spontaneous mutations in untreated
animals has not been conducted. Considering these factors, and the high background incidence of
mutation frequency in this study decreasing the sensitivity for detecting an effect, no inferences can
be made regarding the significance of these results.
Although micronucleus detection in bone marrow or peripheral blood is standard practice,
this assay can be used for any tissue with actively dividing cells. Two studies by the same group
tested intact duodenal tissues from mice exposed to Cr(VI) in drinking water for 7 or 90 days,
separately counting micronuclei in crypt and villous cells. Both studies, testing dose ranges based
on the NTP 2-year bioassay that did not include a group with a maximum tolerated dose, reported
no increased incidence of micronuclei in crypt cells from Cr(VI)-exposed animals. The first, O'Brien
etal. (20131. did not observe a single micronucleus in crypt cells at any dose. It is possible that cells
with DNA damage were eliminated by apoptosis, explaining the lack of micronuclei. Indeed, in the
original report for these animals fThompson et al.. 20111. duodenal crypt histopathology consistent
with apoptosis was observed in 3/10 animals at 11.6 mg/kg and in 4/10 animals (statistically
significant) at 31.1 mg/kg. However, O'Brien etal. f20131 reported no treatment-related changes
in karyorrhectic nuclei, indices indicative of apoptosis or necrosis, at any dose. The failure of
establishing a background incidence, paired with no concurrent positive controls, make these
results difficult to interpret. In their second study of crypt cells, although an extremely low
background incidence was observed, two exposure groups again were observed to have zero
micronuclei: the top concentration (180 mg/L Cr(VI)), and one positive control, DMH
(1,2-dimethylhydrazine) (Thompson et al.. 2015b). This is of some concern considering this
chemical has been used as a positive control to validate the micronucleus assay in the GI tract by
other groups fCoffing etal.. 2011: Ohvama et al.. 2 0 0 2: Vanhauwaertetal.. 2001: Goldberg etal..
19831.
Of these two studies, only O'Brien etal. (20131 also scored villous enterocytes for
micronuclei and reported a statistically significant increase at the top dose at day 7 and the two
highest exposure groups (60 and 180 mg/L Cr(VI)) at day 90. Although the incidences were pooled
for all animals and the total number of cells scored was not reported, this is an intriguing finding.
Micronuclei cannot be formed in cells that are not actively dividing. Although intestinal villous cells
have a rapid turnover rate of 3-5 days, it is the crypt cells that are the rapidly dividing progenitor
cells; these cells proliferate and differentiate, migrating up the villi to form the cells lining the
intestinal villi (Gelberg. 2018). The nonproliferative, fully differentiated villous enterocytes are
continually sloughed into the lumen as they are replaced by new cells (Potten etal.. 2009).
Therefore, to discover micronuclei in the villous cells, and not in crypt cells (assuming that the
study design was sufficient to detect mutational changes in this region), either demonstrates that
genetic damage occurring in the crypt cells suddenly ceased or was repaired in the 24 hours
between the end of the exposure and sacrifice, pushing the last micronucleated cells into the villus,
or, that in response to Cr(VI), the villous enterocytes absorbing Cr(VI) began dedifferentiating and
This document is a draft for review purposes only and does not constitute Agency policy.
3-109 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
migrating back toward the crypt cells, leaving them vulnerable to the genotoxic effects of Cr(VI)
(consistent with the "top-down" theory for colorectal cancer, fHanahan. 2022: Schwitalla etal..
2013: Shih etal.. 200111. Either instance indicates a potential for Cr(VI) to induce genetic damage
in intestinal villi; however, better designed experiments would be needed to draw any
interpretations with confidence.
In vitro studies of GI tissues comparing genotoxicity across species have shown that cellular
responses are similar in gastric mucosal cells between humans and rodents fPool-Zobel etal..
19941. However, other genotoxicity endpoints from in vivo oral exposure studies specific to GI
tissues were negative, including yH2 AX, a marker of DNA double-strand breaks fThompson etal..
2015b: Thompson et al.. 2015al. and DNA-protein crosslinks were not increased in the
forestomach, glandular stomach, and duodenum fDe Flora et al.. 20081. In addition, several in vivo
studies found no increase in 8-OHdG adducts in target tissues across species fThompson etal..
2012b: Thompson etal.. 2011: De Flora et al.. 20081. suggesting that oxidative DNA damage may
not be a primary source of permanent DNA alteration.
Two positive but low confidence in vivo mutation studies were not conducted in portal-of-
entry or tumor target tissues but were designed to detect mutations induced in germ cells and the
developing fetus. Although the focus of this analysis is to inform an MOA for cancer, an agent that
causes mutation in germ cells is of added concern due to the potential for generating heritable
mutations that can be passed to offspring if the agent is anticipated to reach the germinal tissues
fU.S. EPA. 1986cl. Marat etal. f 20181 reported increased dominant lethal mutation frequency,
indicative of increased chromosomal aberrations and/or gene mutations arising in the exposed F0
male. The second study found a significant dose-dependent increase in mutations in mice after
gestational drinking water exposures despite elevated background frequency fKirpnick-Sobol et al..
20061. although there are indications this study may have been biased away from the null.
Although it cannot be determined from these two low confidence studies that ingested Cr(VI)
reaches these tissues in sufficient concentrations to conclude there is a potential mutagenic hazard
to germ cells and the developing fetus, further research is needed.
Although the current evidence base has not consistently identified signature mutations
associated with Cr(VI) exposure, there may be some indications from in vitro studies that Cr(VI)
induces mutations in vivo primarily through larger deletions or structural changes, versus smaller
point mutations or frameshifts that would be detected by the transgenic rodent assay. Additional
investigation of preserved tissues from animal bioassays could allow the analysis of higher
numbers of cells to increase the sensitivity of micronucleus detection. Future testing for mutation
induction in the GI tract could increase sensitivity by harvesting dissociated mucosal epithelial cells
to increase the number of cells for analysis fOkada etal.. 2019: Coffing etal.. 20111. and flow
cytometric scoring of micronucleated cells can dramatically increase the sensitivity of this assay
(Dertinger etal.. 2011). Updated technologies in DNA sequencing and the identification of
This document is a draft for review purposes only and does not constitute Agency policy.
3-110 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
mutational signatures are also capable of resolving these evidence gaps (e.g., Riva etal. f20201:
Valentine etal. f202011.
In conclusion, there is consistent and coherent evidence that a mutagenic MOA for Cr(VI)-
induced carcinogenesis is biologically plausible and relevant to humans. The implications of this
evidence in the context of human pharmacokinetics and the full complement of carcinogenic
pathways, including interpretations regarding tissue type-specific induction in the lung and GI tract
that can be initiated by Cr(VI) exposure, is discussed in the next section.
3.2.3.4. Mode-of-Action Integration of Evidence for Carcinogenesis
Cr(VI) is a human lung carcinogen when inhaled. When ingested, Cr(VI) has been shown to
cause tumors in the GI tract in animals exposed in drinking water (NTP. 20081. Evidence relevant
to the potential key events and pathways involved in Cr(VI)-induced cancer via oral or inhalation
exposures was systematically identified (Section 1.2) and is presented in Appendix C.3.2 organized
by the key characteristics of carcinogens fSmith etal.. 20161. The key characteristics of Cr(VI) with
the largest evidence bases and most relevant study designs are DNA reactivity
(electrophilicity/formation of DNA adducts), genotoxicity, altered DNA repair processes and
genomic instability, epigenetic effects, oxidative stress, and altered cell division and death. This
evidence, along with the evidence of tumors and preneoplastic lesions from animal bioassays and
from gene expression (Appendix C.3.3) and toxicogenomic studies (Appendix C.3.4), informed the
identification of the steps and key events involved in Cr(VI)-induced cancer as described in EPA's
cancer guidelines fU.S. EPA. 2005al.
There are multiple mechanistic processes induced by Cr(VI) exposure that appear to
contribute to carcinogenesis. The large majority of the mechanistic evidence relevant to
interpretations of upstream mechanistic processes induced by Cr(VI) that may lead to
tumorigenesis is summarized here. The key events identified to be involved in the carcinogenic
process induced by Cr(VI) are the distribution, cellular uptake, and intracellular reduction of Cr(VI);
the DNA reactivity of chromium and the formation of Cr-DNA adducts; oxidative stress and free
radical-induced cytotoxicity and DNA damage; epigenetic modifications; altered DNA repair; the
silencing of tumor suppressor genes and the activation of oncogenes; genomic instability; gene and
chromosomal mutation; the suppression of apoptosis; cytotoxicity and degenerative cellular
changes; cell proliferation and regenerative hyperplasia; and chronic inflammation. The studies
informing these key events were not evaluated for risk of bias and sensitivity concerns using
predefined metrics. A prioritized set of studies with designs best suited to examining whether and
to what extent Cr(VI)-induced tumorigenesis involves a mutagenic MOA were subject to an
additional level of review (Section 3.2.3.3).
Figure 3-16 summarizes the key events (organized by levels of biological complexity) and
mechanistic pathways that have been identified to be involved in the carcinogenic process induced
by Cr(VI). Evidence supporting each key event (boxes) and key event relationship (arrows) is
presented in more detail in Table 3-20. The corresponding key characteristic of carcinogens
This document is a draft for review purposes only and does not constitute Agency policy.
3-111 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 (Appendix C.3.2, fSmith etal.. 201611 is identified with each key event where applicable, as well as
2 whether the key event is recognized to be a hallmark or enabling characteristic of cancer fHanahan.
3 2022: Hanahan and Weinberg. 20111. The visualization of key events in this figure resembles the
4 layout commonly used in adverse outcome pathway (AOP) networks, but this diagram is chemical-
5 specific. Although some events clearly precede others, due to the complexity of the key event
6 pathways the key events themselves have not been numbered to avoid the suggestion of an
7 overarching temporal order.
This document is a draft for review purposes only and does not constitute Agency policy.
3-112 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review of Hexavalent Chromium
1
Figure 3-16, Key events and mechanistic pathways induced by Cr(VI) exposure that can lead to cancer.
This document is a draft for review purposes only and does not constitute Agency policy,
3-113 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-20. Evidence for key events and key event relationships involved in Cr(VI)-induced carcinogenesis
Key event
Key event relationship and evidence
References for Cr(VI)-specific evidence3
Pharmacokinetic-dependent molecular initiating event
Distribution, cellular
uptake and
intracellular
reduction of Cr(VI)
Once Cr(VI) reaches the target tissue(s) in sufficient amounts, the Cr(VI) oxyanion is
taken up by cells via nonspecific anion transporters where it is reduced via intracellular
reductants to Cr(V), Cr(IV), and the kinetically stable Cr(lll). The predominant
intracellular reduction pathways and intermediates depend on available ascorbate,
glutathione, and cysteine.
Reviewed in Section 3.1.1, Zhitkovich
(2011), Nickens et al. (2010) (see below
summary of key events)
Macromolecular
DNA reactivity,
adduct and crosslink
formation, and DNA
double-strand
breaks
Cr(VI) is not DNA reactive, but Cr(lll), the final reduction product, can form bulky Cr-DNA
and Cr-protein adducts and crosslinks, leading to replication fork stalling and DNA
double-strand breaks.
Reviewed in Zhitkovich (2005) (see below
summary of key events)
Oxidation of
biological
macromolecules and
ROS generation
Redox reactions during the intracellular reduction of Cr(VI) generates reactive
intermediates Cr(V) and Cr(IV) that produce reactive oxygen species, directly damaging
intracellular molecules including DNA, proteins and lipids, and inducing cell signaling
pathways and transcription factors associated with inflammation, cytotoxicity, apoptosis
and necrosis, including TNF-a, NF-kB, and NRF2. Cr(VI) is a strong oxidizing agent and can
abstract electrons from a number of intracellular ligands, forming oxyradical species and
leading to oxidative stress and cytotoxicity.
Reviewed in Levina and Lav (2005),
Zhitkovich (2011) (see below summarv of
key events)
Oxidative DNA
damage
Reactive oxygen species generated by intracellular reduction of Cr(VI) can cause DNA
strand breaks, both directly through free radical damage and base modifications (e.g., 8-
OHdG adducts), and indirectly via ROS generation, lipid and protein peroxidation, and
depletion of intracellular antioxidants and DNA repair capacity. DNA damage correlates
with ROS levels and treatment with antioxidants reduces DNA damage.
Reviewed in Shi et al. (2004)
This document is a draft for review purposes only and does not constitute Agency policy.
3-114 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Key event
Key event relationship and evidence
References for Cr(VI)-specific evidence3
Epigenetic
modifications
Cr(VI) exposure induces extensive promoter-specific methylation, global
hypomethylation, post-translational histone modifications, and microRNA dysregulation,
affecting the expression of an extensive number of genes shown to be altered by Cr(VI)
exposure, including those involved in cytotoxicity/cell proliferation and DNA repair. This
pattern of hypermethylation of CpG islands, downregulating tumor-suppressor genes,
and concomitant hypomethylation of global (non-CpG) regions, upregulating tumor
promoter genes, contributes to genomic instability, and has been observed in many
idiopathic cancers including adenocarcinomas of the Gl tract.
Reviewed in Chen et al. (2019); also Rager
et al. (2019)
Altered DNA repair
Cr(VI) exposure alters DNA repair processes by the suppression of DNA repair genes via
epigenetic silencing of mismatch repair (MMR) genes. Epigenetic silencing of DNA repair
genes leads to suppression of proficient DNA repair pathways, including mismatch repair
(MMR), leading to microsatellite instability, and homologous recombination repair (HR),
leading to an increased frequency of replication fork stalling and DNA double-strand
breaks. Increased global hypomethylation and increased promoter-specific
hypermethylation of CpG islands in DNA repair genes have been observed in the lung
tumors of chromate-exposed workers, contributing to mutagenesis and genomic
instability, a hallmark of cancer.
Reviewed in Chen et al. (2019); see also
Guo et al. (2019); Wang and Yang (2019);
Hu et al. (2018); Li et al. (2016); Wang et al.
(2012b)
Inadequate DNA
repair (connector
event)
If the DNA damage produced by Cr(VI) reduction and the formation of DNA adducts and
ROS damage cannot be adequately repaired (or removed by programmed cell death),
this can lead to gene mutations, aneuploidy, and genomic instability. In humans,
decreased DNA repair synthesis has been observed in lymphocytes among individuals
exposed to chromium occupationally. The suppression of DNA damage response and
repair genes increases the probability that Cr(VI)-induced genetic damage will lead to
mutations.
Rudnvkh and Zasukhina (1985)
Silencing of tumor
suppressor genes
and activation of
oncogenic pathways
A number of tumor suppressor genes have been shown to be downregulated by Cr(VI)
exposure, with some known to be due to epigenetic silencing, including APC, P16mk4a,
CFTR, and possibly p53, though there is conflicting evidence for p53 involvement.
Activation of the c-Myc and Wnt/p-catenin oncogenic pathways has also been
implicated.
AN et al. (2011), Hu et al. (2016), Kondo et
al. (2006), Tsao et al. (2011), Li et al. (2017),
Lu et al. (2018), Park et al. (2017),
Mezencevand Auerbach (2021)
Cellular and Tissue Level
This document is a draft for review purposes only and does not constitute Agency policy.
3-115 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Key event
Key event relationship and evidence
References for Cr(VI)-specific evidence3
Genomic instability
Genomic and chromosomal instability induced by Cr(VI) contributes to tumorigenesis
and manifests primarily as microsatellite instability, caused by the epigenetic
suppression of mismatch repair genes, and aneuploidy.
Reviewed in Wise and Wise (2010); also AN
et al. (2011), Hirose et al. (2002), Peterson-
Roth et al. (2005), Takahashi et al. (2005)
Gene and
chromosomal
mutation
Bulky Cr-DNA lesions lead to replication fork stalling and DNA double-strand breaks,
which can become fixed mutations if not efficiently repaired or targeted for cell death by
apoptosis. Some of these mutation may confer a growth advantage, leading to a clonal
outgrowth of the mutated cells and tumorigenesis, a process that is more likely to occur
in rapidly proliferating cells.
See mutagenic MOA evidence synthesis,
Section 3.2.3.3
Suppression of
apoptosis
Unlike the cytotoxicity-related apoptosis induced by the direct cellular injury caused by
initial Cr(VI) exposures, the downstream suppression of programmed cell death via
apoptosis contributes to the fixation of mutations and unchecked cell proliferation,
leading to tumorigenesis. Cr(VI) was shown to initiate signaling pathways that promote
cell proliferation and inhibit apoptosis in Gl target tissues in rats exposed via drinking
water for 60 days.
Tsao et al. (2011)
Cytotoxicity
The oxidative damage induced by Cr(VI) can lead to frank cytotoxicity, which has been
observed as increased levels of apoptosis in the lung and small intestine in animals
following inhalation and drinking water exposures, respectively. This cytotoxicity
contributes to degenerative changes and regenerative hyperplasia. Cytotoxicity has not
been detected in the rat oral cavity.
Reviewed in Levina and Lav (2005), Shi et al.
(2004)
Cell proliferation
Cr(VI) exposure to the lung and Gl tract has been shown to induce cell proliferation, both
by inducing proliferative signaling pathways and by evading apoptotic signals that
regulate uncontrolled cell growth in normal cells, contributing to hyperplasia and
tumorigenesis. Increased cell proliferation can lead to increased genomic instability and
the potential for the clonal selection of mutations that confer tumorigenic hallmarks.
Cell proliferation has not been detected in the rat oral cavity.
Kopec et al. (2012a), Rager et al. (2017),
Tsao et al. (2011), Katabami et al. (2000)
This document is a draft for review purposes only and does not constitute Agency policy.
3-116 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Key event
Key event relationship and evidence
References for Cr(VI)-specific evidence3
Degenerative cellular
changes
Biochemical and histopathological evidence of cellular injury has been observed in the
rat lung following inhalation exposures and in the mouse and rat small intestine
following drinking water exposures, indicative of degenerative changes that can initiate
compensatory cell proliferation. No observations of degenerative cellular changes have
been observed in the rat oral cavity.
Glaser et al. (1990), NTP (2007), NTP (2008),
Thompson et al. (2011), Thompson et al.
(2012b)
Regenerative
hyperplasia
Hyperplasia consistent with regeneration following cell injury has been reported
following oral exposures in the small intestine of mice and rats and following inhalation
exposures in the lung in rats. Hyperplasia has not been observed in the rat oral cavity
following Cr(VI) exposures.
NTP (2008), NTP (2007), Glaser et al. (1990),
Thompson et al. (2011), Thompson et al.
(2015b), Thompson et al. (2012b)
Inflammation
Chronic inflammation is an enabling characteristic of cancer. Evidence consistent with
inflammatory lung responses has been observed following Cr(VI) inhalation. However, no
histopathological evidence of chronic inflammation has been reported in the Gl tract
following oral exposures in animals or humans. Some suggestive evidence from
oxidative stress, cytokine fluctuations, and proinflammatory signaling pathways (e.g., NF-
kB) may be indirectly indicative but this evidence in inconclusive.
Johansson et al. (1986b), Glaser et al.
(1990), Glaser et al. (1985), Cohen et al.
(2003), Kim et al. (2004)
Organ
Tumor formation
• Lung (inhalation): Cr(VI) is a human lung carcinogen.
• Oral cavity (ingestion): Increased incidence of squamous cell carcinomas or
papillomas (mucosa or tongue) in both sexes of F344/N rats (NTP 2-year bioassay).
Statistically significant at highest dose (>6 mg/kg-d in males, >7.13 mg/kg-d in
females) with dose-response trend in lower dose groups, in drinking water. See
Figure 3-16 and Table 3-15. Tumors are rare (see Appendix D.2).
• Small intestine (ingestion): Increased incidences of adenomas and carcinomas in
both sexes of B6C3F1 mice (NTP 2-year bioassay). Statistically significant at two
highest exposures (>2.4 mg/kg-d in males, >3.2 mg/kg-d in females) with dose-
response trend in lower dose groups, in drinking water. See Figure 3-16 and Table
3-15. Tumors are rare (see Appendix D.2).
U.S. EPA (1998c), NTP (2008)
aComplete references for the evidence provided in the table can be found in the below summaries of each key event.
This document is a draft for review purposes only and does not constitute Agency policy.
3-117 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
Key events for CrfVIl-induced cancer
Pharmacokinetic-dependent molecular initiating event: The distribution, cellular uptake and
reduction ofCr(Vl)
The effects induced by Cr(VI) can only occur if Cr(VI) reaches the target tissue prior to
extracellular reduction, which essentially inactivates its toxic and carcinogenic potential.
Therefore, consideration of the pharmacokinetics and the competing processes of reduction and
uptake of inhaled or ingested Cr(VI) are central to assessing the carcinogenic potency of Cr(VI).
Chromium (VI) compounds have been traditionally considered nonreactive towards purified DNA
under physiological conditions. Their ability to induce oxidative stress and DNA damage in exposed
cells and tissues in vitro and in vivo (discussed in the following sections) is explained by the
uptake-reduction model of Cr(VI)-mediated genotoxicity fStandeven and Wetterhahn. 19891.
Based on this model and irrespective of target cell type, Cr(VI) is taken up by cellular anion
transporters, where it then undergoes intracellular reduction predominantly driven by ascorbate,
glutathione and cysteine to form the DNA-reactive and/or oxidative damage-inducing
intermediates Cr(V) and Cr(IV), and eventually the thermodynamically stable Cr(III), which
accumulates in cells via its binding to DNA and other molecules (Zhitkovich. 2011. 2005). These
nonspecific anion transporters, present in all cell types, rapidly take up soluble Cr(VI) due to the
structural similarity of the tetrahedral configuration of the chromate (CrCU2-) anion to that of
phosphate (HPO42") and sulfate (SO42") anions (Alexander and Aaseth. 1995: Standeven and
Wetterhahn. 19891.
Reduction of Cr(VI) is a kinetically controlled process, and the role of specific reductants
reflects their reaction rates with Cr(VI) compounds and intracellular concentrations. The highest
rate of Cr(VI) reduction was found for ascorbate, followed by cysteine and glutathione with
respective rate ratios of 61:13:1 (Ouievryn et al.. 2003). Since typical intracellular concentrations
of ascorbate (1-2 mM) and glutathione (1-10 mM) are comparable and considerably higher than
that of cysteine (0.03-0.2 mM) (Tian etal.. 2014). the principal intracellular reducer of Cr(VI) is
ascorbate, accounting for 80-90% of its metabolism fZhitkovich. 2011. 20051. Ascorbate and
glutathione also display a synergistic effect on the reduction of Cr(VI), as the rate of this reduction
by a mixture of ascorbate and glutathione under physiologically relevant conditions was found to
be higher than the sum of the reduction rates of each of these reductants (Suzuki. 1990).
It should be noted that studies performed in cell-free or cell-based systems that do not fully
reflect physiological conditions and concentrations of intracellular reducers may not fully represent
cellular and molecular processes that occur in human tissues under environmental exposures to
Cr(VI). This limitation affects mechanistic cell-free studies that use certain non-physiological
buffers and cell-based studies that employed ascorbate-depleted cells grown in standard growth
media (Ouievryn et al.. 2002). Since ascorbate represents a major intracellular reductant of Cr(VI)
This document is a draft for review purposes only and does not constitute Agency policy.
3-118 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
fSuzuki and Fukuda. 19901. restoration of ascorbate in cell-based systems is necessary for a correct
assessment of the fate of Cr(VI) and DNA damage following its intracellular uptake.
Reduction of Cr(VI) by ascorbate generates variable amounts of Cr(V), Cr(IV), and carbon-
based radicals (Stearns and Wetterhahn. 19941. At physiologically relevant molar ratios of
ascorbate to Cr(VI) exceeding 2:1, the only detectable intermediate reduction product is reportedly
Cr(IV). The presence of Cr(V) is detectable only at non-physiological ratios of equimolar or lower
ratio of ascorbate to Cr(VI), or in ascorbate-depleted cells (Zhitkovich. 2011: Stearns and
Wetterhahn. 19941. Reduction of Cr(VI) by ascorbate under physiologically relevant conditions is a
low oxidant-generating process that differs remarkably from reduction of Cr(VI) by glutathione,
which generates substantially more reactive oxygen species (Wong etal.. 2012). However, in spite
of reduced DNA oxidative damage in cells with restored ascorbate, these cells can still experience a
large increase in genotoxicity, as displayed by an increased frequency of DNA double-strand breaks
fWong etal.. 20121 and DNA-protein crosslinks fSugivama et al.. 19911 (see next section, "DNA
reactivity").
The reduced form of glutathione (GSH) is a major intracellular reducer of Cr(VI) in cells
cultured without restoration of ascorbate (Figure 3-7 in Section 3.1.1). This reduction can be a one-
or two-electron process (Zhitkovich. 2011). but more typically it proceeds as a one-electron process
sequentially producing Cr(V), Cr(IV) and Cr(III) (Marin etal.. 2018). Reduction by cysteine in the
presence of variable amounts of glutathione is also a one- or two-electron process, with the one-
electron process dominating in the physiological range of concentrations fOuievrvn et al.. 20011.
As described in Section 3.1.1.2, inhaled Cr(VI) that deposits in the upper and lower
respiratory tract will come in direct contact with epithelial cells. Reduction of Cr(VI) by epithelial
lining fluid is less effective than gastric fluid, and both high and low-soluble compounds can pose a
hazard to respiratory tract epithelial cells. Although highly soluble Cr(VI) compounds may clear the
lungs faster than low-soluble forms, they have the potential to be more readily taken up by cells.
Low-soluble forms are absorbed more slowly and may be cleared in the mucus but may expose the
epithelial cells for a longer period of time. In addition, high localized accumulation of Cr(VI)-
containing particulates may occur in susceptible lung regions such as airway bifurcation sites
fBalashazv etal.. 2003: Schlesinger and Lippmann. 19781. This is supported by studies showing
high chromium deposition at these sites in the lungs of chromate workers, and a correlation
between lung chromium burden and lung cancer (Kondo etal.. 2003: Ishikawa etal.. 1994a. b).
There is an extensive mechanistic database demonstrating the toxicity and mutagenicity of Cr(VI)
in humans via the inhalation route of exposure (see Section 3.2.3.3 and Appendix C.3.2.2).
Therefore, it will be assumed that inhaled Cr(VI) at any concentration is capable of exposing the
epithelial cells in the respiratory tract, and that compared to GI epithelial cells after Cr(VI) ingestion
(discussed below), the respiratory epithelial cells have an increased potential for Cr(VI) uptake and
Cr(VI)-mediated cytotoxicity and the induction of mutations in these cells.
This document is a draft for review purposes only and does not constitute Agency policy.
3-119 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
Following ingestion, evidence shows that approximately 10% of the Cr(VI) dose is absorbed
in the GI tract of rodents fFebel etal.. 2001: Thomann etal.. 19941. In humans, it is estimated that
<10% is absorbed in the GI tract (depending on the dose and stomach pH), and this number may be
10% or higher in susceptible populations (see Section 3.3.1 and Appendix C.1.5). Therefore, it is
likely that a portion of ingested Cr(VI) interacts with the epithelial cells of the GI tract in all species.
Effects observed by NTP (2008) in mice indicate that unreduced Cr(VI) may traverse the entire
small intestine. The highest incidences of tumors and potentially preneoplastic lesions were
observed in the duodenum, the region immediately distal to the stomach. This region has a higher
surface area per unit length of intestine fCastelevn etal.. 20101. increasing the absorptive capacity
in this tissue. The combination of high Cr(VI) concentration at the epithelial surface and high
absorptive surface capacity are the likely main contributors to the lesions observed in mice by NTP
C20081.
In contrast to the duodenum, the absorption surface area of the stomach is low fCastelevn
etal.. 20101. which may account for the lack of stomach tumors in the NTP (20081 bioassay. The
jejunum and ileum have lower absorption surface areas than the duodenum (but still higher than
the stomach), and these segments exhibited lower incidences of tumors in mice than the
duodenum. Lower tumor incidence also may have been a result of Cr(VI) reduction and dilution by
intestinal secretions and lumen contents. Data by Kirmanetal. (2012) shows chromium
concentrations decreasing in the distal direction in the small intestine of mice exposed to Cr(VI) in
drinking water for 90 days. While the absorption surface area of the oral cavity is also low, as the
first tissue of contact, it is being exposed to the highest concentration of Cr(VI). This may make oral
tissues more prone to neoplastic effects in rats. However, pharmacokinetics cannot explain why
rats and mice differ with respect to oral and small intestinal tumors, since these differences may be
due to a variety of other factors (Ibrahim etal.. 2021: Chandra etal.. 20101. Figure 3-17 illustrates
the ordering of tissues within the GI tract and is annotated with the types of tumors observed by
NTP (2008) in both mice and rats.
This document is a draft for review purposes only and does not constitute Agency policy.
3-120 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Toxicologiccil Review of Hexavalent Chromium
Oral cavity
Duodenum
Jejunum
Ileum
180
E
a
a.
c
u
-a
a>
*•>
i/i
0)
c
60
20
Stomach
Reduction of
Cr(VI) to Cr(lll)
(A) F344/N rats n=100/group
Squamous cell carcinoma
Squamous cell papilloma
180
••••••OOOO
oooooooo
•ooooo
Decreasing Cr(Vl) concentration
90
• ••ooooo o o
oooooooo
~ •ooo
60
oooo
• •oo
30
••ooooo
•
20
oo
• •
10
o
•
5
1 «*o
o
°
1. II
To large
intestine
(B) B6C3F1 mice n=100/group for 0 and 5 ppm
• Carcinoma n=50/group for all other ppm
o Adenoma
Figure 3-17. Reported tumors of the digestive tract tissues for all rodents
exposed to Cr(VI). Points indicate primary adenomas and carcinomas for the
mouse small intestine, and primary squamous cell carcinomas and squamous cell
papillomas for the rat oral cavity (oral mucosa and tongue). Multiple tumors per
animal per tissue are included, but tumors which were known to have metastasized
from other sites were not included.
In the small intestine, the localization of total chromium in different intestinal
compartments provides some mechanistic information on the ability of Cr(VI) to reach the crypts
(where stem cells reside), which could give rise to cytotoxicity as well as fixed mutations in these
highly proliferative cells. Thompson et al. (2015b; 2015a) used X-ray fluorescence
microspectroscopy to examine the concentrations of total chromium in the cells residing within
mouse villi and crypts after 1 and 13 weeks of exposure. All analysis was performed in the middle
section of the duodenum, which may be a significant source of bias because (1) ingested Cr(VI)
tissue concentrations are expected to be highest in the section of the duodenum (proximal small
intestine closest to the stomach) because reduction/dilution will occur as Cr(VI) traverses the
intestine, and (2) the human duodenum is much shorter than that of the rodent duodenum
fCastelevn et al.. 20101. and therefore the middle section of the rodent duodenum may not be as
relevant to humans. After 13 weeks of exposure, Thompson etal. f2015a) detected a weak Cr signal
(0.4 (Jg/g) in the 24 small intestine crypts that were examined, with a 35-fold higher (14 [.ig/gj
mean concentration in the villi. A separate 7-day study reported the absence of Cr in the crypt
compartment without quantitative results; however, these observations may be biased toward the
null due to the rapid movement of cells from the ciypt compartment and the 24-hour recovery time
before imaging was performed (Thompson et al.. 2015bl In a subsequent gene expression study
This document is a draft for review purposes only and does not constitute Agency policy,
3-121' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
that analyzed microdissected crypts and villi in preserved mouse small intestinal tissues from
Thompson etal. f2011I a robust response in gene expression changes was detected in crypts at
>4.6 mg Cr(VI)/kg-d and in villi at all doses (>0.024 mg Cr(VI)/kg-d) after 7 and 90 day exposures,
demonstrating that Cr(VI) does reach the crypts at these concentrations in drinking water
(Chappell et al.. 2022).
In light of the pharmacokinetic evidence, this assessment assumes that ingested Cr(VI)
escaping stomach reduction is capable of coming into contact with cells of the epithelium of the
lower GI tract (small and large intestine), although the Cr(VI) concentration exposing the cells will
be lower than the ingested concentration. Furthermore, this assessment assumes that ingested
Cr(VI) at any concentration is capable of coming into direct contact with the epithelial cells of the
upper GI tract (oral cavity, esophagus, and stomach) prior to stomach reduction. The Cr(VI)
concentration exposing the cells of the oral cavity is likely very close to the ingested concentration.
Ingested Cr(VI) may expose cells of the GI tract, prior to systemic uptake and reduction to Cr(III) by
the liver and red blood cells.
DNA reactivity (KC#1)
Cr(VI) itself is not known to be DNA reactive. In contrast, the intermediate Cr(IV) and Cr(V)
and terminal Cr(III) species that are generated during intracellular reduction of Cr(VI) can induce
DNA damage directly through interactions with DNA and indirectly via oxidative damage (Arakawa
etal.. 2012). The reduction of Cr(VI) in cell-free, cell-based and in vivo systems generates variable
amounts of intermediate chromium species depending on the nature and concentration of the
reductants and concentrations of Cr species fBorges et al.. 19911. The relative abundance of
specific intermediate species is likely to be a major factor in determining the DNA damaging activity
of Cr(VI) (Sugden and Stearns. 2000). Although the specific role of Cr-species and Cr-induced DNA
lesions in the toxicity and carcinogenicity of Cr(VI) has not yet been conclusively established,
depending on experimental conditions, the reduction of Cr(VI) has been found to produce binary
Cr-DNA and ternary ligand-Cr-DNA adducts, interstrand crosslinks, DNA-protein crosslinks,
oxidative damage to bases and deoxyribose, DNA strand breaks, and DNA abasic sites, which have
been associated, to various extents, with cell cycle arrest, DNA repair, cell death and mutagenesis
(Sugden etal.. 2001: Arakawa et al.. 2 0 0 0: Casadevall etal.. 1999: Stearns and Wetterhahn. 1997:
Zhitkovich etal.. 1996: Bridge water et al.. 1994). The kinetics of intracellular reduction are
reviewed in Section 3.1.1.3, and the specific experimental support for the in vivo generation of the
intermediate and terminal Cr species, as well as their direct and indirect genotoxicity potential, is
described in Appendix C.3.2.
Oxidative stress and oxidative DNA damage (KC#5)
Oxidative stress induced by Cr(VI) exposure appears to lead to several toxicity pathways
causing cytotoxicity, inflammation (in the lung), cell proliferation, and DNA damage. Redox
reactions during the intracellular reduction of Cr(VI) generate reactive intermediates Cr(V) and
This document is a draft for review purposes only and does not constitute Agency policy.
3-122 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Cr(IV) that produce reactive oxygen species, which can cause cytotoxicity and directly damage
intracellular molecules including DNA, proteins and lipids, and in the process, induce cell signaling
pathways associated with inflammation and cell proliferation (reviewed in Levina and Lay (200511.
Radical species formed when Cr(VI) oxidizes intracellular macromolecules can also induce
oxidative damage (reviewed in Zhitkovich (2011)). Reactive oxygen species generated by
intracellular reduction of Cr(VI) can cause free radical damage to DNA via base modifications
(e.g., 8-OHdG adducts), lipid and protein peroxidation, and depletion of intracellular antioxidants
(reviewed in Shi etal. f200411. Because these effects have been well-documented in review articles,
this section will focus on evidence of oxidative stress in occupationally exposed humans and in
animals exposed to Cr(VI) via oral or inhalation, or in vitro studies using human cells derived from
lung or GI tissues. Oxidative stress induced by Cr(VI) exposure has been characterized in other
health effects sections of this assessment, including oxidative damage contributing to Cr(VI)-
induced toxicity of the lung (Section 3.2.1), GI tract (Section 3.2.2), liver (Section 3.2.4), male and
female reproductive organs (Sections 3.2.7 and 3.2.8, respectively), and fetal development (Section
3.2.9). Therefore, the evidence from the lung and GI tract in animals will be briefly summarized
again here, along with systemic evidence of oxidative stress following inhalation or oral exposures.
As summarized in Section 3.2.1, many observational studies reported statistically
significantly increased incidences of systemic disruption in cellular redox status that correlated
with exposure to Cr(VI) in urine and blood of industrial workers and rodents exposed to Cr(VI);
these are also summarized in Appendix C.3.2.5. In tumor target tissues, one study relevant to lung
tissues did not detect increased 8-OHdG adducts in the sputum of lead chromate pigment factory
workers fKim etal.. 19991. No studies examining oxidative stress in GI tissues were identified in
exposed humans.
A small number of animal studies were identified that evaluated oxidative stress in tumor
target tissues. Oxidative DNA damage in the rat lung, evidenced by increased formation of 8-OHdG
adducts, was reported following inhalation or intratracheal instillation exposures in rats (Zhao et
al.. 2014: Maeng etal.. 2003: Izzotti etal.. 19981. Three in vivo studies were identified that reported
biomarkers of oxidative stress in GI tissues after oral exposure (Thompson et al.. 2012b: Thompson
etal.. 2011: De Flora etal.. 20081. None of these studies observed an increase in 8-OHdG adducts in
the mouse or rat small intestine or oral cavity following Cr(VI) drinking water exposures. However,
an increased proportion of oxidized glutathione (GSSG) relative to reduced glutathione (GSH),
indicative of oxidative stress, was observed in the mouse small intestine after 7 and 90 days of
exposure, with a correlated change in the GSH/GSSG ratio in plasma after 90 days at doses >59
mg/L Cr(VI) fThompson etal.. 20111. A decreased GSH/GSSG ratio was also observed in the mouse
oral mucosa after 7 days, but this resolved after 90 days despite a significantly higher total
chromium concentration in these tissues compared with the control fThompson etal.. 20111.
Changes in GSH/GSSG ratios were generally not observed in the oral cavity of rats after 7 days of
Cr(VI) exposure (the ratio was decreased at 0.1 mg/L Cr(VI) in the oral mucosa) but were
This document is a draft for review purposes only and does not constitute Agency policy.
3-123 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
significant and dose-dependent in the oral mucosa and jejunum (and not the duodenum) at >20 mg
Cr(VI)/L for 90 days fThompson et al.. 2012bl. with a significantly decreased ratio in plasma at
>170 mg/L. While GSH/GSSG ratio measurement is a generally accepted indicator of oxidative
stress, ascorbate is the preferred in vivo reductant, accounting for 90% of Cr(VI) oxidative
metabolism. Therefore, though the primary oxidative pathway is not captured in these
experiments, the level of involvement of GSH implies extensive oxidative stress was occurring in
these tissues. Other indicators of protein or lipid oxidation were not elevated in the duodenum of
mice after 90 days fThompson etal.. 20111 or in the rat in the oral mucosa or duodenum
fThompson et al.. 2012bl. The reason for the lack of oxidative DNA lesions associated with the
oxidative stress in these studies is not known. The significance of the oxidative stress detected in
tissues that do not develop tumors, or the potential physiological reasons for the inconsistencies
between species, is also not clear.
A large body of evidence from cells exposed in vitro exists to support and investigate the
oxidative damage induced by Cr(VI) (Appendix Table C-57). These studies include tests in model
systems where ROS levels, lipid and protein oxidation, and decreased levels of antioxidant enzymes
all correlate with DNA damage. Although in vitro exposures may lead to exaggerated cell stress and
oxidative responses, limiting their ability to predict physiological conditions in vivo, these studies
can provide supporting evidence indicating the potential contribution of oxidative stress and the
signaling pathways involved. This DNA damage is increased in test systems deficient in processes
involved in repairing free radical damage and is decreased in many test systems with antioxidant
pre-treatment. The evidence base includes studies performed with human lung or colon and gastric
cancer cell lines to study oxidatively induced DNA damage and cytotoxicity. These in vitro studies
have been summarized in "Mechanistic Evidence" in Sections 3.2.1 (Respiratory Tract Effects Other
Than Cancer) and 3.2.2 (Gastrointestinal Tract Effects Other Than Cancer).
In addition to oxidative stress initiating cytotoxicity and DNA damage following Cr(VI)
exposure, there is evidence that oxidative stress can result in pro-inflammatory signaling pathways
that contribute to cancer. The nuclear transcription factor NF-kB is activated in response to redox
cell signaling and cytokines and is involved in cell survival, proliferation and inflammation
fTaniguchi and Karin. 20181. NF-kB has been found to be upregulated in response to Cr(VI)
exposure in numerous studies and test systems, including in the Cr(VI)-exposed rat lung fZhao et
al.. 20141. in human lung cells in vitro (Wang etal.. 2019: He etal.. 2013: Zuo etal.. 2012: Kim etal..
20031. and in other human cells in vitro (Tullv etal.. 2000: Kaltreider et al.. 19991. The increases in
NF-kB levels correlated with increasing ROS levels and were abrogated by antioxidant treatments
fKim etal.. 20031. TNF-a, which activates NF-kB, is a pro-inflammatory cytokine produced by
immune cells that are involved in redox signaling (Blaser etal.. 20161. It has been shown to be
induced systemically by Cr(VI) in rats fMitrov etal.. 20141. in LPS-stimulated mice flin etal.. 20161.
and in HaCaT immortalized human keratinocyte cells in vitro (Lee etal.. 2014: Wang etal.. 2010b).
However, these findings were not predictive of the results in three studies of occupationally
This document is a draft for review purposes only and does not constitute Agency policy.
3-124 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
exposed humans, which did not detect increased systemic TNF-a levels in blood or serum fOian et
al.. 2013: Mignini etal.. 2009: Kuo and Wu. 20021.
The transcription factor NRF2 binds to and activates genes regulated by Antioxidant
Response Element (ARE) in response to oxidative stress, transactivating genes for antioxidant
enzymes and promoting cell survival (He etal.. 20201. NRF2 has been observed to be upregulated
in human liver cells (Zhong et al.. 2017a) and constitutively activated in Cr(VI)-transformed human
lung cells in vitro (Clementino etal.. 20191. In vivo, the gene that codes for NRF2, NFE2L2, was
found to be upregulated in the duodenum of mice exposed for 91 days to Cr(VI) in drinking water
fKopec et al.. 2012al.
Gene expression changes in genes involved in ROS homeostasis have also been observed in
human lung, hepatic, and epithelial cells treated with Cr(VI) in vitro (e.g., NOX, S0D1, S0D2, CAT,
GSR) (Zhong etal.. 2017b: Zhong etal.. 2017a: Zeng etal.. 2013: Russo etal.. 2005: Asatiani et al..
20041. In addition, Cr(VI) was found to oxidize and inhibit mitochondrial and cellular thioredoxins
and peroxiredoxins involved in cell survival and redox signaling in immortalized human bronchial
epithelial cells, leading to increased sensitivity to ROS damage (Myers et al. f2011: 2010: 2009:
200811.
Overall, there is a consistent, coherent, and biologically plausible evidence base available to
describe the intracellular reduction and redox imbalance, oxidative stress, and cellular oxidative
damage due to free radical generation caused by Cr(VI) exposure, potentially contributing to
cytotoxicity, genetic damage, and cell proliferative signaling pathways.
Epigenetic modifications (KC#4)
Epigenetic modifications are heritable changes in gene expression that occur without
altering the genetic material (Sharma etal.. 20101. This "nonmutational epigenetic
reprogramming," which can be mediated through modifications to histones, DNA methylation, and
noncoding RNAs (e.g., microRNA), is considered an enabling characteristic of cancer (Hanahan.
20221. Five studies evaluated epigenetic changes in humans in relation to chromium exposure.
Kondo etal. f20061 reported increased methylation of P16ink4a, a tumor-suppressor gene, in
chromate factory workers with lung cancer who had occupational chromate exposure compared to
those without chromate exposure. Similarly, they observed increased methylation of P16ink4a with
increased duration of chromium exposure (>15 years) among lung cancer cases (Kondo etal..
20061. Increased methylation was also observed in DNA MMR genes hMLHl and hMSH2 when
comparing lung cancer cases with and without chromate exposure (Ali etal.. 2011: Takahashi etal..
20051 and in the CpG islands (promoter regions) of MMR and HR genes (i.e., MGMT, HOGG1, XRCC1,
ERCC3, and RAD51) in exposed factory workers compared to controls fHu etal.. 20181. Another
study identified inverse associations between blood chromium and the microRNA miR-3940-5p,
which functions as an epigenetic tumor-suppressor by targeting cyclin D1 and ubiquitin specific
peptidase-28 (Ren etal.. 20171. as well as between miR-3940-5p and the DNA repair genes BRCC3
and XRCC2, involved in DNA damage response and homologous DNA repair (Li etal.. 2014b). Ali et
This document is a draft for review purposes only and does not constitute Agency policy.
3-125 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
al. f20111 also observed increased methylation at MGMT, which encodes an enzyme that repairs
DNA adducts at the 06 position of guanine, in chromate lung tumors compared to lung tumors in
referents, as well as in APC, a tumor-suppressor gene that is suppressed via promoter
hypermethylation or mutation in over 85% of colorectal cancers (Zhu etal.. 2021: Tuanes. 20201.
Two additional studies reported decreased methylation across global DNA (Wang etal.. 2012b) as
well as mitochondrial genes (MT-TF and MT-RNR1) specifically (Linqingetal.. 20161 in chromium-
exposed workers (chromate production workers and chrome-plating workers, respectively)
compared to controls.
The findings in humans are supported by studies in vitro showing that Cr(VI) exposure
induces extensive promoter-specific hypermethylation, global hypomethylation, post-translational
histone modifications, and microRNA dysregulation, demonstrating that Cr(VI)-mediated
epigenetic alterations may play a role in affecting the expression of an extensive number of genes
shown to be altered by Cr(VI) exposure (reviewed in Chen etal. f201911. The results from
toxicogenomic studies (reviewed in Appendix C.3.4) showing multiple pathways affected by Cr(VI)
with relevance to carcinogenesis are consistent with the scope of genes shown to be affected by
Cr(VI)-induced epigenetic alterations. These findings are coherent with a recent analysis of existing
toxicogenomic data that identified transcriptional alterations corresponding to epigenetic
modifications following Cr(VI) exposure that were found to influence gene expression in pathways
corresponding to cytotoxicity/cell proliferation and suppression of DNA repair (Rager etal.. 2019).
A pattern of hypermethylation of CpG islands and concomitant hypomethylation of global (non-
CpG) regions has been observed in many idiopathic cancers including adenocarcinomas of the GI
tract fLocke etal.. 2019: CGARN. 2018al.
Altered DNA repair (KC#3)
Although there are numerous processes contributing to the repair of genetic damage when
it occurs, these processes are not failsafe, and any alterations to these activities can result in an
increased risk of heritable mutation (Chatteriee and Walker. 2017). As reviewed in the next
section, epigenetic modifications induced by Cr(VI) exposure have been shown to silence genes
involved in DNA repair, an effect that is found in a significant number of lung tumors from chromate
workers compared to lung tumors in people not exposed to Cr(VI) and has been found to increase
with dose (Hu etal.. 2018: Li etal.. 2014b: Ali etal.. 2011: Takahashi etal.. 2005). Hirose etal.
(2002) reported finding microsatellite instability (MSI) attwo or more loci in 78.9% of lung cancers
with chromate exposure compared to lung cancers without chromate exposure. MSI is the result of
a state of genetic hypermutability that is caused by defective mismatch repair and is found in
approximately 15% of colorectal cancers fBoland and Goel. 20101. Subsequent studies identified
hypermethylation of the CpG island promoter regions of MMR genes hMLHl and hMSH2 in lung
tumors of workers exposed to chromate compared to lung tumors from unexposed subjects (Ali et
al.. 2011: Takahashi etal.. 2005). In vitro, Cr(VI) exposure of human colon cells lacking MLH1
protein led to increased resistance to apoptosis, providing a selective growth advantage (Peterson-
This document is a draft for review purposes only and does not constitute Agency policy.
3-126 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Roth etal.. 20051. This epigenetic silencing of genes involved in DNA repair observed in workers
exposed to Cr(VI) may contribute to mutagenesis and genomic instability, a hallmark of cancer.
Another study of workers in the chromium industry investigated the effect of prolonged
exposure to Cr(VI) on the ability of the cell to correct errors during DNA replication. Evidence of
decreased DNA repair synthesis was observed in isolated lymphocytes exposed to UV light to
compare DNA repair synthesis between Cr(VI)-exposed workers and unexposed subjects (Rudnykh
and Zasukhina. 19851. A nonmonotonic relationship with duration of exposure was also identified,
though sample size was limited within each category of duration.
This slowing of DNA replication could be explained by the formation of bulky Cr-DNA
adducts, which can stall replication forks, leading to increased formation of DNA double-strand
breaks. There are two main DNA double-strand break repair pathways: homologous recombination
(HR) and non-homologous end joining (NHEJ). NHEJ is the predominant repair process in the G1
phase of the cell cycle, prior to synthesis, when only one chromatid is present; it is more error-
prone than HR, which occurs primarily in S/G2, using the sister chromatid as a template for repair.
Cr(VI) has been shown to induce DNA double-strand breaks and Rad51 foci formation, inducing HR
in vitro fBrvantetal.. 20061. However, several studies have also reported a specific inhibition of
genes involved in HR, including Rad51 fSpeer etal.. 2021: Hu etal.. 2018: Browning etal.. 2016: Li
etal.. 2016: Oin etal.. 20141. Cr(VI)-induced targeting of Rad51 following prolonged in vitro
exposures to Cr(VI) has also been shown to involve alterations in Rad51-mediated nucleofilament
assembly, which the authors speculated was due to a Cr(VI)-mediated inhibition of Rad51 nuclear
import (Browning and Wise. 2017: Browning etal.. 20161 and Rad51 foci formation at DNA double-
strand breaks fSpeer etal.. 2021: Oin etal.. 20141. This evidence suggests that a Cr(VI)-mediated
influence on Rad51 may result in modifications to HR, increasing reliance on NHEJ and potentially
leading to unrepaired DNA double-strand breaks and increased aneuploidy and genomic instability.
Silencing of tumor suppressor genes and activation of oncogenic pathways
The ability to evade growth inhibition by suppressing genes that limit cell proliferation is a
hallmark of cancer fHanahan and Weinberg. 20111. The decreased expression of a number of
tumor suppressor genes has been observed following Cr(VI) exposure. For some of these genes, the
mechanism of decreased expression involves epigenetic silencing, and it has been observed that GI
tumors have significantly higher frequencies of DNA hypermethylation at CpG islands than non-GI
tumors (CGARN. 2018a). Cr(VI) was found to induce methylation at CpG sites in the promoter
region of the P16ink4a tumor-suppressor gene; inactivation of this gene is commonly found in lung
cancers and was observed in lung tumors of workers exposed to chromate, which increased with
duration of exposure fHu etal.. 2016: Ali etal.. 2011: Kondo etal.. 20061. Methylation of the APC
(adenomatous polyposis carcinoma) gene, a tumor-suppressor gene that maintains genome
integrity by preventing instability, has also been shown to occur more frequently in the lung tumors
of chromate-exposed workers compared to lung tumors in referents (Ali etal.. 20111. APC
This document is a draft for review purposes only and does not constitute Agency policy.
3-127 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
suppression by mutation or CpG island hypermethylation is present in over 85% of colorectal
cancers fZhu etal.. 20211.
P53 is a tumor-suppressor that normally regulates cell cycle arrest and apoptosis to protect
against tumor formation; the induction ofp53 target genes can indicate the presence of DNA
damage, and inactivation of p53 is associated with carcinogenesis (Williams and Schumacher.
20161. P53 gene expression and protein levels were suppressed in the stomach (gene expression
>3.5 mg/kg-day and protein levels >1.7 mg/kg-day Cr(VI)) and colon (gene expression and protein
levels >5.2 mg/kg-day Cr(VI)) of male Wistar rats after 60 days of exposure to Cr(VI) in drinking
water fTsao etal.. 20111. No studies of p53 expression in human GI tissues or nonneoplastic lung
tissues are available, but studies in lung tumor tissues from chromate exposed vs. referent workers
detected either no difference (Katabami etal.. 2000) or increases fHalasova etal.. 20101 in p53
protein expression, or reduced levels of p53 mutations (Kondo etal.. 19971. and two studies of the
peripheral blood of exposed workers detected increased p53 protein expression fElhosarv etal..
2014: Hanaoka etal.. 19971. However, although these studies in humans were not evaluated for
risk of bias and sensitivity, little information was given regarding potential coexposures, making it
difficult to draw conclusions from these findings. In vitro, some studies showp53 activation in
human lung cells increased with higher Cr(VI) concentrations (Hu etal.. 20161 or occurring in vitro
and not in vivo (Rager etal.. 20171. so the nature of how p53 expression may be affected by Cr(VI)
is not understood.
The oncogene c-Myc has also been shown to be differentially methylated in response to
Cr(VI). Myc was found to show a dose-dependent increase (protein and mRNA) in the stomach and
colon of male Wistar rats after 60 days of exposure in drinking water to Cr(VI) in the stomach (>3.5
mg/kg-d) and colon (>1.7 mg/kg-d) fTsao etal.. 20111. In context, these findings are consistent
with the other observed effects of Cr(VI) exposure given the activity of this broad ranging
oncogene, whose transcriptional control overlaps pathways of DNA damage response, cell
proliferation and metabolism. Myc can be activated by another oncogenic pathway, the Wnt/(3-
catenin signaling pathway. Although no studies were identified that specifically investigated this
pathway following Cr(VI) exposure, its involvement has been indirectly implicated by studies of
Cr(VI)-induced methylation and subsequent downregulation of APC, a Wnt antagonist, as well as by
the downregulation of serine/threonine kinase 11 and depletion of the Gene 33 protein fLu etal..
2018: Li etal.. 2017: Park etal.. 20171.
An analysis of the toxicogenomic data reported in Kopec et al. (2012b: 2012a) from mice
exposed to Cr(VI) in drinking water has identified a potential role for CFTR (cystic fibrosis
transmembrane conductance regulator) in the carcinogenic effects of Cr(VI) fMezencev and
Auerbach. 2021). A tumor suppressor function has been demonstrated for CFTR in the GI tract of
Cftr knockout mice fThan etal.. 20161. Cftr gene expression was decreased in mice exposed to
Cr(VI) levels as low as 0.1 mg/L Cr(VI) (0.024 mg/kg-d) in drinking water for 8 days. Loss of CFTR
expression in humans was found to correlate with the severity of colorectal cancer, and in animals
This document is a draft for review purposes only and does not constitute Agency policy.
3-128 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
with a mutated /Ipc gene, to potentiate tumor progression Thanetal. f20161. Although this effect
has not been characterized beyond this single analysis, the implications of a specific Cr(VI)-induced
CFTR suppression contributing to cancer risk in humans warrants further investigation.
Genomic instability (KC#3)
Genomic instability, an increased rate in the acquisition of genomic alterations, is an
enabling characteristic of cancer and is present in nearly all human cancers fHanahan and
Weinberg. 2011: Negrini etal.. 20101. As mentioned above, Cr(VI) exposure induces the
suppression of DNA repair genes involved in mismatch repair. Defective mismatch repair leads to a
form of genomic instability, microsatellite instability, which is a state of genetic hypermutability
that is closely associated with colorectal cancer in humans (Boland and Goel. 20101. Microsatellite
instability has been detected in the lung tumors of chromate workers compared to referent
workers (Hirose etal.. 2002). suggesting that Cr(VI) exposure may facilitate increased genomic
instability, and ultimately cancer initiation and progression.
In addition to microsatellite instability, Cr(VI) exposure is also associated with increased
aneuploidy, a numerical chromosomal aberration that involves chromosome malsegregation and
breakage (Eastmond etal.. 20091 that is endemic of chromosomal instability and is a hallmark of
cancer (Ben-David and Amon. 20201. Delayed, persistent, transmissible genomic instability has
been observed in immortalized human cells in vitro, manifest as increased structural chromosomal
aberrations, micronuclei, and aneuploidy, and decreased clonogenic cell survival (Glaviano etal..
20061. The delayed, persistent effects were confirmed in other in vitro studies that observed
aneuploidy increasing with exposure duration fWise etal.. 2016: Holmes et al.. 20061. Several
additional studies have shown the ability of Cr(VI) to induce aneuploidy in human cells in vitro,
summarized in Appendix Table C-54 and by Wise and Wise (20101. While most of these studies
used solid-stained chromosomal analysis to detect aneuploidy, the findings have been confirmed by
detection in kinetochore-positive micronuclei (Giierci etal. (20001. Seoane et al. (2002: 2001.
199911 or by chromosome painting with fluorescent probes (Figgitt etal.. 20101. methods with
greater specificity. Exogenous agents inducing aneuploidy may act by interfering with the mitotic
spindle apparatus via disruption of the microtubule cytoskeleton, a mechanism that is consistent
with several mechanistic investigations of Cr(VI)-induced aneuploidy (Martino etal. (20151. Niis
and Kirsch-Volders (19861. Seoane et al. (2002: 2001.199911. It is also plausible that altered DNA
damage and repair pathways (e.g., loss of functional p53 and activation of driver oncogenes like
Myc, reviewed above) can increase aneuploidy by promoting cell cycle progression before repair
pathways can be initiated, resulting in chromosome malsegregation. APC, a tumor-suppressor gene
associated with colorectal cancer when suppressed via promoter hypermethylation or mutation,
has also been shown to have a key role in mitotic spindle orientation fluanes. 20201. Although the
mechanism for induction of aneuploidy by Cr(VI) is not known, the APC gene was found to be
silenced by hypermethylation in the lung tumors of chromate-exposed workers (Ali etal.. 2011).
This document is a draft for review purposes only and does not constitute Agency policy.
3-129 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
providing a hypothesis for how aneuploidy may be induced by Cr(VI), disrupting cell division and
contributing to carcinogenesis; further research is warranted.
Gene and chromosomal mutation (KC#2)
The evidence for the genotoxic effects of Cr(VI) is presented and synthesized in Section
3.2.3.3. There is consistent and coherent evidence that a mutagenic MOA for Cr(VI)-induced
carcinogenesis is biologically plausible and relevant to humans. Primary evidence is provided by
medium and low confidence studies of occupationally exposed humans; some evidence is available
in animals exposed directly in the lung or GI tract, but this evidence base is small and consists of low
confidence studies, many of which were not optimized for reliably detecting genotoxicity.
Genotoxicity studies employing more direct exposures to Cr(VI) (e.g., in vitro and in animals
exposed via i.p. injection) are largely positive (summarized in Appendix C.3.2.2), consistent with
what is known regarding the intracellular pharmacokinetics and DNA reactivity of Cr(VI), as
discussed above.
Suppression of apoptosis (KC#10)
The ability to resist cell death is a hallmark of cancer, contributing to the fixation of
mutations and unchecked cell proliferation (Hanahan and Weinberg. 20111. Although initial
exposures to Cr(VI) induce cytotoxicity (see below), there is evidence from one study of longer
duration exposures that Cr(VI) can lead to the downstream suppression of programmed cell death
via apoptosis in tumor target tissues. Tsao etal. f20111 measured protein and mRNA levels in the
stomach and colon of male rats following 60-day exposures to Cr(VI) in drinking water and
reported decreased expression of p53 (gene and protein), the mediator of a primary cellular fate
determination pathway, which would lead to suppression of apoptosis (Tsao etal.. 2011). This
suggests a possible mechanism for a Cr(VI)-specific suppression of apoptosis via disruption of p53-
mediated pathways that respond to cellular stress, although this is an area that requires further
investigation.
Cytotoxicity and degenerative cellular changes (KC#10)
Cr(VI), a strong oxidizer, is known to be cytotoxic in vitro and may trigger apoptosis
through increased oxidative stress, leading to DNA and protein damage, mitochondrial dysfunction,
and modulation of pro-apoptotic signaling pathways. The reduction of Cr(VI) generates reactive
intermediates Cr(V) and Cr(IV) that produce reactive oxygen species that can lead to apoptosis and
necrosis, as well as induce cell signaling pathways associated with cell death (reviewed in Levina
and Lay (2005) and Shi etal. (2004)). Because this evidence is relevant to both cancer and
noncancer mechanisms of toxicity, these effects are reviewed in Sections 3.2.3.1 and 3.2.3.2 for the
lung and GI tract, respectively. To summarize, this evidence supports a toxicity pathway of tissue
injury induced by cytotoxicity in the lung and GI tract that may lead to necrosis and/or regenerative
proliferation. In the lung, studies investigating the underlying mechanisms involved in Cr(VI)-
This document is a draft for review purposes only and does not constitute Agency policy.
3-130 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
induced lung toxicity report significant cytotoxicity at micromolar concentrations in vitro,
concurrent with indications of increased programmed cell death (apoptosis, autophagy) in
response to Cr(VI) exposure. In the GI tract, evidence of GI tract toxicity that involves Cr(VI)-
induced cytotoxicity and apoptosis leading to degenerative changes and regenerative hyperplasia,
as well as cell proliferation directly induced by Cr(VI). Other evidence of gene expression changes
indicate cell signaling pathways induced by Cr(VI) exposure that are involved in the evasion of
apoptosis contributing to tumorigenesis, indicating a downstream role independent of the cytotoxic
effects of Cr(VI) that separately contributes to carcinogenesis by suppressing apoptosis. These
cellular and molecular processes underlie the histopathological changes, including hyperplasia of
the small intestine (described in Animal Evidence), that are considered potentially preneoplastic
events.
Cell proliferation (KC#10)
Cancer is the result of sustained and uninhibited cell proliferation fHanahan and Weinberg.
20111. Several studies have identified proliferative markers and signaling pathways that are
upregulated by Cr(VI) exposure. Increases in transcript expression of Ki-67, a nuclear protein
associated with cellular proliferation, and in some cases malignant metastasis and tumor growth (Li.
etal.. 2015a). was detected in the duodenum of mice after exposure to 11.6 and 31 mg/kg Cr(VI)-
day in drinking water; levels were increased approximately 4-fold after 7 days of exposure but
diminished to approximately 2-fold after 90 days (data from Kopec etal. (2012a) was presented
graphically in Thompson etal. f201311. In another drinking water exposure study, a dose-
dependent upregulation of the c-Myc oncogene was found in the stomach (>3.5 mg/kg-d) and colon
(>1.7 mg/kg-d) of male Wistar rats after 60 days of exposure to Cr(VI) in drinking water (Tsao et
al.. 2011). MYC functions as a transcription factor that upregulates genes involved in cell
proliferation and other processes contributing to neoplastic transformation (Gabav etal.. 2014).
Another transcription factor, AP-1, was found to be significantly activated by Cr(VI)
exposure in studies of gene expression changes in human lung cells (Zuo etal.. 2012: O'Hara etal..
20041 and in human breast cancer and rat hepatoma cells fKaltreider et al.. 19991. The AP-1
complex, which is composed of oncogenic proteins (Jun, Fos, ATF, MAF) fEferl and Wagner. 20031.
is induced by JNK and ERK/MAPK signaling cascades in response to stress and inflammatory
cytokines (Gazon etal.. 2017). leading to increased cell proliferation or inhibition of apoptosis, in
part through the activation of cyclin D1 (Guo etal.. 2020). Cyclin Dl, a regulator and promoter of
cell cycle progression, has been detected at significantly increased levels in the lung tumor tissues
of chromate-exposed patients compared to unexposed lung cancer patients (Katabami etal.. 2000).
Increased expression of cyclin Dl has been associated with cell proliferation and tumorigenesis
fGuo etal.. 20201. These findings are consistent with an induction of biological processes by Cr(VI)
that can lead to sustained cell proliferation and contribute to cancer. It is currently unknown to
what extent these proliferation-promoting pathways are initiated by Cr(VI)-induced epigenetic
repression of transcriptional regulators or are the result of a compensatory response to cytotoxicity
This document is a draft for review purposes only and does not constitute Agency policy.
3-131 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
and DNA damage sensing and repair machinery (discussed below), or if other direct or indirect
factors induced by Cr(VI) are involved.
Regenerative hyperplasia
Hyperplasia is the enlargement of a tissue or organ resulting from increased cell
proliferation and can be induced as an adaptive or compensatory response to cellular and tissue
damage. In the evaluation of noncancer effects in the GI tract from ingested Cr(VI), hyperplasia is
considered to be an adverse effect (Section 3.2.3), but it can also represent preneoplastic lesions
that are part of the morphologic and biologic continuum leading to cancer (Hanahan and Weinberg.
2011: Boorman et al.. 2003). Because hyperplasia can also be a reversible effect, it is important to
consider several relevant factors when determining the contribution of hyperplasia to
tumorigenesis, including whether there is a common cellular origin for hyperplasia and tumors, the
presence or absence of a morphological continuum within the study between hyperplasia and
neoplasia, histologic similarities, whether there is treatment-related toxicity, and other information
about the test compound, including mutagenicity and ADME considerations fBoorman et al.. 20031.
The diffuse intestinal epithelial hyperplasia observed in mice across studies is described in
detail in Section 3.2.2.2. In the NTP (2008) 2-year bioassay, minimal to mild diffuse hyperplasia
was significantly increased in the duodenum of all exposed male and female mice. These animals
also exhibited tumors of epithelial origin (adenomas and carcinomas) that were statistically
significant at the two highest exposures (>2.4 mg/kg-d in males, >3.2 mg/kg-d in females) with a
dose-response trend in lower dose groups. There were multiple shared pathological features
between the diffuse hyperplasia and the neoplastic lesions, including elongated crypts with
increased numbers of epithelial cells and mitotic figures (NTP. 2008). These observations are
generally consistent with the intestinal hyperplasia observed in mice in subchronic studies by NTP
(2007) and Thompson et al. (2015a: 2011). lending further evidence of a consistent response in
animals exposed to Cr(VI) via drinking water.
However, even with the presence of these morphologic similarities, in the absence of
experiments with recovery groups to distinguish these lesions from reversible hyperplasia induced
by Cr(VI), it cannot be concluded with certainty that the hyperplasia observed in the subchronic
studies would have progressed to neoplasia. As discussed in Section 3.2.2.3, some discrepancies
have been noted, including the lack of increased mitotic activity in hyperplastic duodenal crypt cells
in mice (Thompson et al.. 2015b: O'Brien etal.. 2013). although follow-up analysis of the mice
exposed via drinking water for 7 and 90 days (Thompson et al.. 2 011) reported a significant
response in gene expression changes related to cell cycle progression phenotypically anchored to
the histopathological results in duodenal crypts at doses >4.6 mg Cr(VI)/kg-d fChappell etal..
20221. In addition, as discussed above, although Thompson etal. f20131 reported levels of the
cellular replication marker Ki-67 were increased compared to untreated controls in mice exposed
for 7 or 90 days in drinking water, these levels declined in the mice exposed for 90 days, and Ki-67
This document is a draft for review purposes only and does not constitute Agency policy.
3-132 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
cannot distinguish between chemically induced cell proliferation and proliferation secondary to
cellular toxicity without concurrent detection of cellular markers for apoptosis and necrosis.
The presence of tissue injury is also important in interpreting the relevance of these lesions
to neoplasia. Tissue-specific hyperplasia and neoplasia with an inciting factor such as cellular
degeneration and compensatory regeneration may suggest a carcinogenic response that is
secondary to chronic tissue injury (Boorman etal.. 2003). As reviewed in Section 3.2.2.2, the
authors of both sets of studies (Thompson et al.. 2012b: Thompson etal.. 2011: NTP. 2008. 20071
considered the hyperplastic lesions to be consistent with regenerative hyperplasia resulting from
Cr(VI)-induced epithelial damage and degenerative changes seen in the mouse villi. This suggests a
mechanism in the carcinogenic process that may be secondary to chronic tissue injury.
In addition to the diffuse hyperplasia, a non-statistically significant incidence of focal
epithelial hyperplasia was observed in male mice at >2.4 mg/kg Cr(VI)-day that increased slightly
in severity grading (3.0-3.5) with dose. Female mice also showed a low incidence of focal
hyperplasia with increasing severity grading (2.0-3.0) at 1.2 and 3.2 mg/kg Cr(VI)-day with no
reported incidences at the high dose fNTP. 20081. NTP considered the focal hyperplasia to be
biologically significant preneoplastic lesions due to the pathological similarities to neoplastic
growths, including crypts and villi that were lined by increased numbers of cuboidal to tall
columnar epithelial cells that were morphologically similar to those of the adenomas ffNTP. 20081:
see Francke and Mog (2021) for further description). In addition, these lesions, located in the
superficial mucosa rather than the crypt mucosa, arose from the same tissue type (duodenal
epithelium) as the neoplastic growths32. The focal hyperplastic lesions were distinguished from
adenomas by their smaller size and less discrete margins that tended to blend with the normal
surrounding mucosal epithelium.
While diffuse hyperplasia may have an origin in a regenerative response that is secondary
to chemically induced tissue degeneration, focal hyperplasia that is morphologically similar to
neoplasia without evidence of concurrent tissue injury may be indicative of a direct neoplastic
response (Boorman et al.. 2003). Although the focal hyperplasia could be a part of the proliferative
continuum of lesions, progressing from diffuse hyperplasia to focal hyperplasia (preneoplastic), to
adenoma (autonomous growth), to carcinoma (malignant neoplasia) originating from a common
precursor cell type, this cannot be confirmed due to the absence of histopathological observations
from interim sacrifices.
Thompson etal. (2012b) also reported duodenal hyperplasia and villous apoptosis in rats
treated with >7.2 mg Cr(VI)/kg-d in drinking water for 7 and 90 days, as well as villous atrophy at
7.2 mg Cr(VI)/kg-d. Rats were not observed to develop intestinal lesions or tumors in the bioassays
by NTP (2008, 20071. Rats developed tumors in the oral cavity, but there were no observations of
lesions or hyperplasia in the rat oral cavity by any of these studies.
32Most (76%) tumor-bearing animals were observed to have exhibited nonneoplastic lesions in the small
intestine (see Appendix D.5).
This document is a draft for review purposes only and does not constitute Agency policy.
3-133 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Hyperplasia has also been observed in the rat lung following inhalation exposures to Cr(VI)
for 30 and 90 days fGlaseretal. fl9901. see Section 3.2.1.2). A high incidence of bronchioalveolar
hyperplasia (70-100%) was reported in male Wistar rats after 30 days of exposure to 0.050-0.40
mg/m3 Cr(VI) relative to the control (10%) (Glaser etal.. 1990). The same study reported lower
incidence of this effect after 90 days of exposure, and after 90 days of exposure with a 30-day
recovery period, suggesting this may have been a transient effect
Overall, there is evidence for regenerative hyperplasia as a key event for tumors of the small
intestine in mice. Theoretically, any increase in the rate of cell proliferation over the background
basal rate of cell division, even if transient, can increase the probability of the formation and
fixation of mutations that may confer a selective advantage to the cell and promote the subsequent
clonal outgrowth of the mutated cells, leading to tumorigenesis. There are some inconsistencies
that create uncertainty in drawing a conclusion that Cr(VI)-induced regenerative hyperplasia is a
primary event driving carcinogenesis, including hyperplastic responses that did not increase in
severity with dose, and the presence of degenerative lesions and hyperplasia in the rat small
intestine with no induction of tumors at this site in this species. Regenerative hyperplasia may be a
contributing factor to carcinogenicity in the lung, as toxicity and hyperplasia have been observed in
the lung following inhalation exposures, though there is not enough evidence to assume a key role
in this tissue. There is no evidence to conclude regenerative hyperplasia is involved in the
tumorigenic process in the rat oral cavity.
Chronic inflammation (KC#6)
Cr(VI) has been shown to induce effects consistent with an inflammatory response by
generating oxidative stress that can stimulate pro-inflammatory cytokines and activate nuclear
transcription factors associated with inflammation (e.g., NF-kB). The evaluation of evidence for
effects of Cr(VI) on the immune system, presented in Section 3.2.6, suggests that Cr(VI) may have a
stimulatory effect on the immune system, largely based on primary immune response assays
indicating increased antibody responses, WBC function and numbers, and total immunoglobulin
levels following Cr(VI) exposure in animals (see Section 3.2.6). Although exposure-related
stimulation of the immune system can lead to exaggerated inflammatory responses associated with
chronic systemic inflammation, the role of inflammation in the carcinogenesis of the GI tract
induced by Cr(VI) exposure (Section 3.2.2) is not clear.
The GI tract contains the majority of immunoglobulin-producing cells that are present in the
human body, and toxicity to the GI tract commonly results in immune system-mediated
inflammation (Gelberg. 2018). Chronic inflammation could have driven the diffuse hyperplasia
observed prior to carcinogenesis in the mouse small intestine in the NTP subchronic and chronic
bioassays, as this is a well-characterized step in inflammatory neoplastic progression, and is an
enabling characteristic of cancer (Hanahan and Weinberg. 2011: Westbrook etal.. 2010). The
development of idiopathic GI cancers has been shown to involve chronic inflammation that can
induce neoplastic genetic and epigenetic changes mediated by proinflammatory cytokines and ROS
This document is a draft for review purposes only and does not constitute Agency policy.
3-134 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
fChiba etal.. 20121. In addition, immunogenomic profiling of data from over 10,000 tumors
collected by the Cancer Genome Atlas used cluster analysis to identify six immune subtypes
commonly associated across multiple tumor types; one identified immune subtype, "wound
healing," was associated with colorectal cancer, lung squamous cell carcinomas, head and neck
squamous cell carcinomas, and the chromosomal instability (CIN) pathway of colorectal cancer
pathogenesis (CGARN. 2018b). tumor tissues also associated with Cr(VI)-induced cancer. However,
NTP reported that the rat oral cavity had neither hyperplasia nor inflammation preceding tumor
formation, and no signs of inflammation were observed in the mouse small intestine after two years
of drinking water exposure to Cr(VI). NTP did report an increased infiltration of histiocytes
(macrophage immune cells) in the duodenum and jejunum that was consistently observed in both
sexes of rats and mice orally exposed both chronically and subchronically to Cr(VI) (Thompson et
al.. 2012b: Thompson et al.. 2 011: NTP. 2008. 2007). but this was not accompanied by an influx of
other inflammatory cells or other histological features consistent with inflammation in the small
intestine and was interpreted by the authors to be of unknown biological significance.
Evidence following inhalation exposures to Cr(VI) is more robust, with consistent evidence
of histiocytosis in the lung from several studies in animals accompanied by inflammatory markers
in BALF and increased leukocytes in plasma, observations supportive of inflammatory lung
responses (Section 3.2.1). The histiocytic/macrophage infiltration leads to cytokine release and cell
to cell signaling conducive to an inflammatory environment (Kodavanti. 2014). Studies
investigating immune toxicity (Section 3.2.6) in chromate workers have also observed changes in
cytokine signaling (Appendix C.2.5.2). Although the direction of these changes was not consistent
across studies or exposure durations, fluctuations in systemic cytokine levels and increased
oxidative stress are characteristic of an inflammatory response and may indicate a disruption in the
regulatory balance that dictates normal immune system function. However, while there is evidence
of oxidative stress and activation of pro-inflammatory cytokines and nuclear transcription factors
including NF-kB, the characterization of chronic inflammation that may occur prior to the
development of neoplasms induced by Cr(VI) exposure remains an evidence gap.
Tumor formation
Neoplastic effects were not observed in subchronic 13-week studies in mice and rats
(Thompson et al.. 2012b: Thompson etal.. 2011: NTP. 2007). though notably some of the
observations in the subchronic studies, including elongated intestinal crypts and increased mitotic
activity, were also reported in the histopathological analysis of adenomas and carcinomas in the 2-
year bioassay. The lack of tumor formation in the subchronic experiments is likely due to
insufficient latency time. The earliest appearance of tumors of the mouse small intestine reported
by NTP in the two-year bioassay fNTP. 20081 was at 451 days in males and at 625 days in females
exposed to the highest tested Cr(VI) doses (5.7 mg/kg-d and 8.9 mg/kg-d in males and females,
respectively). In all other dose groups, tumors in the mouse small intestine were reported at
terminal sacrifice (729 days). The earliest recorded incidences of tumors of the rat oral cavity
This document is a draft for review purposes only and does not constitute Agency policy.
3-135 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
reported by NTP f20081 were at 506 days in females and at 543 days in males exposed to the
highest tested Cr(VI) doses (7.1 mg/kg-d and 6.1 mg/kg-d in females and males, respectively).
Several models have been proposed for the histopathogenesis of GI cancers that are potentially
relevant to Cr(VI). One example is the classical model of transformation and clonal expansion of
rapidly dividing, self-renewing stem cells at the bottom of the intestinal crypts, or the 'bottom up'
model (Schwitalla etal.. 2013: Shih etal.. 2001: Bach etal.. 20001. Alternatively, a 'top down' model
of adenoma morphogenesis in a transgenic c-Myc mouse model system suggests that dysplastic
cells at the luminal surface of the crypts have the ability to dedifferentiate and spread laterally and
downward, forming new crypt-like foci f Schwitalla etal.. 20131. This type of cellular phenotypic
plasticity driven by oncogenic signaling, observed in colon cancers, is considered a hallmark
capability of cancer (Hanahan. 20221. Expression of c-Myc also increases in the stomach and colon
of rats after subchronic oral Cr(VI) exposure (Tsao etal.. 20111. and toxicogenomic data
demonstrate comprehensive activation of the c-Myc pathway and concurrent changes in known
downstream target genes (Rager etal.. 2017: Kopec etal.. 2012b: Kopec etal.. 2012a: Thompson et
al.. 20111. The dysplastic cells at the luminal surface are stem-like, preneoplastic, and represent
mutant clones containing genetic alterations not found in the morphologically normal cells at the
bottom of the crypt (Shih etal.. 2001). This model is based in part on the frequent observation that
early adenomatous polyps are found at the top of colonic crypts without stem cell compartment
contact (Shih etal.. 2001). Mechanistically, Schwitalla etal. (2013) proposed that NF-kB can
enhance Wnt signaling leading to dedifferentiation of epithelial non-stem villus cells into tumor-
initiating cells. In addition, the cell proliferation marker Ki-67, which was increased in the duodena
of mice after exposure to Cr(VI) in drinking water fRager etal.. 2017: Kopec etal.. 2012al. has been
shown to be increased in the dysplastic crypt orifices of idiopathic human intestinal adenomas
(Shih etal.. 2001).
Evidence favoring the 'bottom up' model is provided by a follow-up analysis of the mice
exposed via drinking water for 7 and 90 days (Thompson etal.. 20111. which determined that a
robust response in gene expression changes was present in the crypts at doses >4.6 mg Cr(VI)/kg-
d, and that the enrichment of gene sets related to cell cycle progression and DNA damage were
more robust in the crypts compared to the villi fChappell et al.. 20221. Alternatively, there is
evidence for the 'top-down' model, as X-ray fluorescence microspectroscopy in a separate study by
this group detected a 35-fold higher mean Cr(VI) concentration in the villi compared to the
intestinal crypts (Thompson et al.. 2015a). The precise mechanism for how Cr(VI) would initiate a
'top-down' process is unknown but could plausibly involve mutagenic processes; although
inconclusive due to incomplete reporting and analysis, O'Brien etal. f20131 reported increased
micronucleus frequency in the duodenal villi of Cr(VI)-exposed mice. Neither model can be reliably
ruled out without further investigation.
There is considerable uncertainty regarding the origin of the tumors observed in the rat
oral cavity by NTP (2008). A recent review of chemicals that have been shown to cause oral
This document is a draft for review purposes only and does not constitute Agency policy.
3-136 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
squamous cell neoplasms by the NTP suggests multiple mechanisms can promote rat oral tumors
flbrahim etal.. 20211. Some studies have indicated that a Cr(VI)-induced effect in the bone marrow
or blood cells may have exacerbated an effect occurring at the epithelium. Two studies exposed the
skin of hairless mice to UV light while simultaneously exposing some groups to Cr(VI) in drinking
water (Uddin etal.. 2007: Davidson et al.. 20041. Ingested Cr(VI) significantly increased
susceptibility to UV-induced skin tumors in a dose-dependent manner. One commonality between
the NTP (20081 studies and the Davidson et al. (20041 and Uddin etal. (20071 studies is that the
skin and alimentary tract are sites where bone marrow-derived stem cells can engraft fKrause et
al.. 20011 and possibly transform to malignant epithelial tumors fBessede etal.. 2015: Ouante etal..
2013: Ouante etal.. 2011: Gonda etal.. 2009: Ouante and Wang. 2009. 2008: Fox and Wang. 2007:
Li etal.. 2006: Houghton et al.. 20041. An unknown mechanism involving these cells could have
contributed to the initiation and/or growth of the skin and oral tumors. Another group reporting
an i.p. injection experiment in female Wistar rats showed effects on the submandibular gland that
may support the findings of oral cancer in rats. Submandibular acinar saliva-secreting cells showed
an increase in cystatin staining, which may play a role in tumorigenesis, metastasis, and
immunomodulation fOchieng and Chaudhuri. 2010: Cohen etal.. 19931. Inducible type 2 cystatin
was not detected in the parotid or sublingual glands, trachea, lung, stomach, small intestine, large
intestine, spleen, liver or pancreas, suggesting that Cr(VI)-induced effects on cystatins are likely to
be localized. These inferences, however, are highly speculative. Overall, the underlying
mechanisms induced by Cr(VI) that lead to oral tumors in rats are unknown.
Cancer mode-of-action summary
The mechanistic events identified above and depicted in Figure 3-18 have some level of
Cr(VI)-specific evidence to indicate their involvement in the carcinogenic effects of Cr(VI). These
events are biologically plausible in that they are known to be associated with carcinogenesis and
can occur in humans, with interrelated pathways that emerge involving mutagenicity, cytotoxicity
and regenerative cellular proliferation. The molecular events involved in these effects are assumed
to be relevant to all routes of exposure. The evidence-based assumption is that some amount of
unreduced Cr(VI) can reach target tissues when ingested or inhaled and can be quickly taken up by
the cells in these tissues, where it will be reduced intracellularly to reactive intermediates that
induce toxic and carcinogenic effects. At the tissue level, differences in the evidence for each tumor
type also emerge, therefore it is unclear whether some mechanistic events are key for every tumor,
as the mechanistic effects may be dependent on the specific pattern or duration of activation of
certain events. These may occur based on cell type-specific properties such as their baseline
proliferative rate or ability to mitigate the effects of oxidative stress. Cr(VI) is a known human lung
carcinogen, therefore specifics of lung tumors will not be discussed here in the context of
mechanistic evidence, but the mechanistic evidence from studies of the exposed lung is considered
relevant and discussed along with mechanistic evidence for the tumors of the mouse small
intestine. There is a lack of empirical mechanistic evidence from the rat oral cavity.
This document is a draft for review purposes only and does not constitute Agency policy.
3-137 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicologiccil Review of Hexavalent Chromium
Figure 3-18. Cellular processes involved in the mutagenic MOA of Cr(VI).
There is extensive evidence of the mutagenicity of Cr(VI] when reduced intracellularly to
Cr(III] in studies conducted among in vitro test systems. A mutagenic MOA is also supported in test
animals when considering the evidence in the context of pharmacokinetic considerations. Although
the evidence of mutation from oral exposures is less consistent, the genotoxicity observed in animal
i.p. studies in vivo has been consistently observed. Therefore, evidence of transmissible and
permanent genetic alterations have been prioritized for the analysis of a mutagenic MOA if
observed following oral or inhalation exposures in GI or lung tissues.
Cr(VI) is a known lung carcinogen, and a mutagenic MOA is supported for lung tumors
following inhalation exposures primarily by evidence of increased micronuclei detected in the
blood and exfoliated nasal and oral epithelial cells from occupationally exposed humans. Mutagenic
activity also correlates with blood chromium levels in medium confidence studies, and several low
confidence human studies that demonstrate increased chromosomal aberrations despite many
having limitations that would potentially lead to bias toward the null. Supporting evidence is also
provided by studies showing increased levels of DNA damage in exposed workers, as well as one
low confidence study of mutations in the mouse lung that increased with dose and time following
intratracheal instillation, providing biological plausibility that mutation is involved in the
development of Cr(VI)-induced lung cancers. Therefore, a mutagenic MOA for lung tumors is
considered to be relevant to humans and sufficiently supported in laboratory animals after
inhalation exposure, based on the following: 1) the evidence-based interpretation that some
amount of inhaled Cr(VI) (at physiologically relevant doses] escapes detoxification and is taken up
This document is a draft for review purposes only and does not constitute Agency policy,
3-138' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
by target cells; 2) this uptake is expected to occur more readily in regions of the lung showing a
high chromate deposition that correlate with sites of lung tumors in exposed workers; 3)
demonstrations of increased chromosomal mutations in the exfoliated nasal and buccal cells and in
the peripheral blood of occupationally exposed workers; 4) gene mutations in the mouse lung that
increased with dose and time post-intratracheal instillation; 5) other genotoxic effects in the
peripheral blood of exposed workers and in lung-derived cell cultures in vitro; and 6) mutagenicity
of Cr(VI) when it reaches cells of various tissue types in vivo and in vitro. The implications of a
mutagenic MOA for the dose-response analysis and inhalation unit risk calculation for lung cancer
are presented in Section 4.4.3.
The evidence for a mutagenic MOA following oral exposures is less clear. There are no
human oral exposure studies of mutation in the GI tract, although consistent evidence of increased
micronucleus frequency in the oral epithelial cells of exposed workers may support the evidence
that Cr(VI) can induce mutagenic effects when it comes into contact with cells in the GI tract, and
contributes to an evaluation of whether mutation may be a primary neoplastic event The database
of in vivo oral animal genotoxicity studies that are specific to GI tissues is limited to a small number
of low confidence studies, most of which have deficiencies in sensitivity for detecting an effect or
other concerns that introduce a large amount of uncertainty.
The mutagenicity assays used by these studies were originally designed and optimized for
purposes of identifying hazard, namely, whether a chemical is capable of inducing increased
mutagenic damage, regardless of dose. Although several doses are typically employed, these assays
are not optimized for dose response, and typically use a minimal number of animals (1-5).
Therefore, it is important that these assays use a range of doses that include a maximum tolerated
dose (MTD) or otherwise indicate that the chemical reached the target tissue to ensure sensitivity
(Havashi. 2016) and strengthen a conclusion that, when reported, null findings represent a true
lack of effect (versus a deficiency in study design). As with all genotoxicity assays, these tests are
often considered in an MOA analysis for cancer, with the hypothesis that evidence of mutation in
the tumor target tissue occurs earlier than the induction of tumors, in the same species, and at the
same doses causing tumors supports a mutagenic MOA. Evaluations of this hypothesis often
presume the converse also applies, in that a negative result will indicate a lack of mutagenicity and
therefore support an alternate MOA that does not involve mutagenicity. This assumption often
relies on testing results within an acute to subchronic exposure period in a small number of
animals. It is difficult to make a definitive conclusion that Cr(VI) is not mutagenic in the GI tract
following oral exposures from an evidence base in animals composed of mostly null results from a
small number of low confidence studies, given that Cr(VI) has been shown to be mutagenic
following more direct exposures (i.e., i.p., in vitro), and we can reasonably expect that ingested
Cr(VI) will reach the GI tract.
High levels of cytotoxicity can lead to the detection of increased DNA damage in some test
systems. For this reason, the interpretation of genotoxicity evidence from chemicals inducing
This document is a draft for review purposes only and does not constitute Agency policy.
3-139 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
excessive toxicity includes efforts to determine whether increases in genotoxicity are potentially
secondary to cytotoxicity. For the Cr(VI) in vivo oral exposure database, there is not enough
evidence to determine whether and to what extent Cr(VI)-induced genotoxicity might be the result
of (secondary) cytotoxic DNA damage in the GI tract Most notably, while many of the animal
studies examining the most relevant genotoxicity endpoints did not detect substantial evidence of
genotoxicity at doses that also caused histological effects in the GI tract, including diffuse
epithelial/crypt cell hyperplasia and degenerative changes in the villi (vacuolization, atrophy, and
apoptosis), one study did observe statistically significantly increased micronuclei in villous cells
from animals exposed to doses that similarly induced villous atrophy and apoptosis. Because no
studies were available that specifically examine the presence or absence of genotoxicity in the GI
tract as the MTD was approached and exceeded, this uncertainty cannot currently be addressed.
Although it is presumed that ingested Cr(VI) can reach the target tissues in at least a
fraction of humans and animals, there are pharmacokinetic differences between oral and inhalation
exposure routes that indicate lower concentrations of Cr(VI) will reach target tissues when
ingested than when inhaled. In this context, however, it is still not possible to conclude that there is
no potential risk of increased mutations occurring in humans ingesting Cr(VI) in drinking water,
particularly when taking into consideration human subpopulations with a diminished ability to
reduce Cr(VI) in the stomach due to low gastric pH (see 'Susceptible populations' in the following
section). Therefore, given the uncertainty in the evidence base of ingestion studies in animals due
to a lack of study designs adequately testing for mutagenicity in target tissues, a mutagenic MOA is
supported for GI tumors after oral exposure, based on the following: 1) the evidence-based
interpretation that some amount of ingested Cr(VI) (at physiologically relevant doses) escapes GI
detoxification and reaches target cells; 2) the demonstrated chromosomal mutations in buccal cells
of occupationally exposed workers; and 3) the demonstrated mutagenicity of Cr(VI) when it comes
into direct contact with any cell type in various tissues in vivo and in vitro.
The mutagenic effects of Cr(VI) in the lung and GI tract are expected to be amplified by
promutagenic effects that are also anticipated to be key events for cancer induced by Cr(VI).
Oxidative stress induced by reactive Cr(VI) intermediates can damage DNA and intracellular
proteins and lead to an imbalance between free radicals and antioxidants. Direct and indirect
suppression of DNA repair processes via epigenetic silencing may lead to increased DNA damage,
DNA double-strand breaks, and genomic instability including microsatellite instability and
aneuploidy. The epigenetic modifications induced by Cr(VI) include extensive promoter-specific
hypermethylation, global hypomethylation, post-translational histone modifications, and microRNA
dysregulation. These perturbations can affect the expression of an extensive number of genes
including tumor suppressors and oncogenes associated with lung and colorectal cancers that
involve the promotion of unchecked cellular proliferation along with the suppression of apoptosis.
Although epigenetic changes are not permanent changes to the gene sequence, their overall effect
can be analogous to mutation in that they are heritable changes affecting gene expression. The
This document is a draft for review purposes only and does not constitute Agency policy.
3-140 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
oxidative stress, oxidative DNA damage, direct or epigenetic suppression of DNA repair processes,
and genomic instability induced by Cr(VI) are all likely to be key events for carcinogenesis
applicable to oral and inhalation exposures for all tumor types. These effects combine to produce a
promutagenic microenvironment that promotes the formation and fixation of mutations from DNA
damage, regardless of whether the genetic damage was produced endogenously, by Cr(VI), or from
another source.
Consistent evidence of an inflammatory response in the lung following inhalation Cr(VI)
exposures in animals indicates this effect is likely to be a key event for lung cancer. Although
idiopathic cancer development in the GI tract has also been shown to involve chronic inflammation
(Chiba etal.. 2012). no histopathological evidence of GI inflammation induced by Cr(VI) oral
exposure was observed in animals exposed via drinking water. However, the inflammatory
response associated with GI tract cancers has been shown to be mediated by proinflammatory
cytokines and ROS, effects that are known to result from Cr(VI) oral exposures and can lead to
genetic and epigenetic changes that promote neoplastic transformation. Combined, these data
suggest that inflammation could still be involved in the neoplastic effects of the small intestine in
mice.
An alternative MOA for carcinogenicity induced by ingested Cr(VI) is regenerative
proliferation caused by tissue injury, leading to a higher probability of spontaneous mutations that
may result in tumorigenesis. Cr(VI), a strong oxidizer, is known to be cytotoxic in vitro and may
trigger apoptosis through increased oxidative stress, mitochondrial dysfunction, and modulation of
pro-apoptotic signaling pathways. Following oral exposures, regenerative hyperplasia interpreted
to be the result of regressive changes such as villous blunting, villous atrophy, and apoptosis of
enterocytes was consistently observed in the mouse small intestine fThompson et al.. 2012b:
Thompson etal.. 2011: NTP. 2008. 20071. Inconsistencies in the hyperplastic responses to these
degenerative changes have been noted, however, including hyperplasia that did not increase in
severity with dose, and no statistically significant or dose-responsive changes in mitotic or
apoptotic indices in tissue regions where increased crypt length, area, and number of crypt
enterocytes were reported. The diffuse hyperplasia of the small intestine is likely to be a key event
for tumors in this tissue, although these hyperplastic lesions, which were also observed in the rat
small intestine by Thompson etal. f2012bl. do not always progress to cancer and can represent a
functionally adverse change on their own.
The GI tract has a high capacity for tissue regeneration following cellular injury, which
makes it more sensitive to exposures that may interfere with the process of cell division (Nolte et
al.. 20161. At least some of the molecular events affecting cell cycle regulation that are altered by
Cr(VI) exposure also appear to underlie the regenerative histopathological changes in animals
exposed to Cr(VI). A toxicogenomic analysis comparing gene expression changes in the duodenal
crypts and villi of the mice exposed via drinking water for 7 and 90 days (Thompson et al.. 2 011)
found a robust response in the crypts at doses >4.6 mg Cr(VI)/kg-d, and that the enrichment of
This document is a draft for review purposes only and does not constitute Agency policy.
3-141 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
gene sets related to cell cycle progression and DNA damage were more robust in the crypts
compared to the villi fChappell etal.. 20221. Other toxicogenomic evidence consistent with
increased cellular proliferation in the mouse small intestine, including increased expression of
oncogenic c-Myc and the proliferative marker Ki-67, provides additional support for increased cell
proliferation occurring in the preneoplastic small intestine, although these markers are not specific
to regenerative hyperplasia. It is also not clear whether the degenerative and regenerative effects
are key events for other tumor types. No lesions or hyperplasia have been reported in the rat oral
cavity, and while cellular injury and hyperplasia were observed in the rat lung following inhalation
exposures, the hyperplasia diminished with longer exposures and following a recovery period.
The focal hyperplasia observed only in the mouse small intestine, although not statistically
significant or dose-dependent, represents a biologically important preneoplastic event that could
result from the interaction between Cr(VI)-induced regenerative processes and mutagenic effects
fNTP. 20081. These lesions were observed closer to the hyperplastic villous region of the
superficial intestinal mucosa, where Cr(VI) has been shown to concentrate (Thompson etal..
2015a). Some evidence of micronuclei and oncogenic transformation has also been observed in this
tissue fO'Brienetal.. 2013: Tsao etal.. 20111. This indicates the potential for a combined MOA for
Cr(VI)-induced tumorigenesis in the small intestine after oral exposure, where mutagenic effects
occur concurrently with hyperplasia, providing an environment that can support the clonal
expansion of mutated cells.
Although no histopathological changes were observed in the rat oral cavity preceding tumor
formation in subchronic or chronic bioassays of Cr(VI) in drinking water, and no increases in
mutation frequency were observed in these tissues in a single study investigating this endpoint,
mutagenicity is a biologically plausible mechanism and is coherent with the evidence of increased
micronuclei in the buccal cells of exposed humans. Although site concordance is not a requirement
when considering the evidence for a mutagenic MOA, there is currently not an understanding of
why humans do not show evidence of oral tumors, or why rats do not have tumors of the small
intestine. It is plausible that extensive epigenetic alterations, which have been shown to account for
phenotypic differences among individuals as well as among different tissue and cell types (Zhang et
al.. 20131. may influence the differences in carcinogenic response and the carcinogenic potency of
Cr(VI) at the tissue level or even among individuals and across species.
In conclusion, the available mechanistic evidence supports key events at the molecular and
cellular level that are expected to be applicable to all exposure types and tumors. These key events
are summarized in Table 3-21 and Figures 3-16 and 3-18. Cr(VI) that is not reduced extracellularly
may be taken up by cells near the point of contact, which is generally expected to be the lung for
inhalation exposures and the GI tract for oral exposures. The GI tract, including the oral cavity, is
expected to be exposed by both of these routes in humans (impaction of dusts in the mouth and
tongue resulting from oral breathing and mucociliary clearance may result in GI exposure via the
inhalation route). Oxidative stress occurs within the cell, generated by the reactive chromium
This document is a draft for review purposes only and does not constitute Agency policy.
3-142 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofHexavalent Chromium
intermediate species, inducing DNA damage and cytotoxicity. Chromium-DNA adducts can be
formed by the ultimate Cr(III) species and, in combination with suppressed DNA repair processes
via epigenetic modifications, these adducts and other oxidative DNA damage may be fixed as
mutations in these cells. Cr(VI) may also promote aneuploidy and microsatellite instability by
suppressing DNA mismatch repair. These promutagenic effects, combined with epigenetic
modifications influencing the suppression of apoptosis and increased cell proliferation, combine to
create a tumor microenvironment supporting the clonal outgrowth of mutated cells. In addition,
there is evidence from the small intestine of mice exposed via drinking water that Cr(VI) exposure
can induce degenerative effects at the tissue level, with a proliferative response that should
promote the selection of cells with a growth advantage, leading to tumorigenesis, though it is
unclear whether this occurs in all tumor types. These processes also likely involve chronic
inflammation, though there is inconsistent evidence of this in all tumor tissues.
This document is a draft for review purposes only and does not constitute Agency policy.
3-143 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-21. Evidence profile table for the carcinogenic mechanisms of inhaled or ingested Cr(VI)
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
Distribution of Cr(VI)
(Sections 3.1.1 and
3.2.3.4; Appendix
C. 1.2)
Lung:
• Inhaled Cr(VI) comes into direct contact with lung epithelial cells and is expected to be
directly absorbed with minimal extracellular reduction (i.e., detoxification) due to a less
favorable reduction environment in lung tissues
• Cr(VI) accumulates at lung bifurcation sites in the lungs of chromate workers
• Cr(VI) burden in the lung correlates with lung cancer incidence
Oral cavity: Following inhalation or oral exposures, cellular uptake may occur in the
epithelium of the oral mucosa, tongue, and esophagus (prior to Cr(VI) reduction in the
stomach), although the surface area for mass transfer is low
Stomach: While reduction (i.e., inactivation) of ingested Cr(VI) occurs in the stomach, it will
compete with gastric emptying of Cr(VI) to the small intestine. Uptake in the stomach
epithelium is also possible, although the surface area for mass transfer is low
Small intestine:
• Cr(VI) bioavailability and kinetic considerations suggest that 10-20% of ingested Cr(VI)
escapes human gastric inactivation and could expose the Gl tract epithelium
• Cr(VI) exposure to the proximal small intestine will be greater than exposure to the
distal small intestine, as the Cr(VI) concentration decreases
• The surface area for mass transfer in the small intestine is high
Following exposure to Cr(VI), it has
been demonstrated that inhaled Cr(VI)
can reach cells in the lung and oral
cavity, and after ingestion, Cr(VI) can
reach cells in the oral cavity (either by
movement through the Gl tract after
inhalation and deposition into the oral
cavity, or by direct ingestion), stomach,
and small intestine, both potentially in
appreciable amounts to elicit an effect.
Distribution is strongly dependent on
route of exposure (inhalation ->
respiratory tract, oral ingestion ->
gastrointestinal tract and liver).
Cellular uptake of
Cr(VI)
(Sections 3.1.1 and
3.2.3.4; Appendix
C.l.l)
All cell types: Cr(VI) is rapidly taken up by nonspecific sulfate and phosphate transporters
due to the structural similarity of Cr(VI).
Lung: Particulates may deposit and absorb locally; the amount taken up is dependent on
location, particle size, and solubility.
Ingested or inhaled Cr(VI) can be taken
up by cells in tumor target tissues.
This document is a draft for review purposes only and does not constitute Agency policy.
3-144 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
Oral cavity: Morphology within different regions of the oral cavity is highly variable (hard
palate, buccal mucosa, gingiva, ventral/dorsal tongue, lip), and may impact localized cellular
uptake.
Stomach: Lower absorptive surface area and different morphology than the small intestine.
Some uptake may occur prior to gastric emptying.
Small intestine: Highly absorptive surface area increases uptake of Cr(VI) (primarily by the
villi).
Intracellular reduction
of Cr(VI)
(Sections 3.1.1.3 and
3.2.3.4; Appendix
C.3.2.1)
All cell types: Following cellular uptake, Cr(VI) is reduced primarily by ascorbate, but other
biological reductants (e.g., cysteine, GSH) are also capable of reducing Cr(VI). This leads to
the intracellular formation of the reactive intermedidate species Cr(V) and Cr(IV) and the
stable Cr(lll).
Intracellular reduction is considered an
activation pathway, generating
reactive intermediates capable of
damaging DNA directly or indirectly via
oxidative damage.
DNA reactivity
(Section 3.2.3.4;
Appendix C.3.2.1)
All cell types: Intracellular Cr(lll) has been demonstrated to be DNA reactive and can form
stable complexes with DNA, RNA, amino acids and proteins, including Cr(lll)-DNA adducts,
DNA-DNA crosslinks, and DNA-protein crosslinks.
Intracellular Cr(lll) can bind to DNA,
which can form bulky adducts that
cause replication fork stalling, DNA
double-strand breaks and mutations if
not adequately repaired or eliminated
by apoptosis.
Oxidative stress and
oxidative DNA
damage
(Section 3.2.3.4;
Appendix C.3.2.5)
Inhalation exposure:
• Consistent evidence of significant increases in oxidative stress in workers exposed to
Cr(VI) that correlated with levels of Cr(VI) in urine and blood (see Appendix C.2.1 and
C.3.9)
• Increased formation of 8-OHdG DNA adducts in rats exposed to Cr(VI) via inhalation
(Maeng et al., 2003) or intratracheal instillation (Zhao et al., 2014; Izzotti et al., 1998)
Oral exposure:
• Decreased GSH/GSSG ratio in small intestinal epithelium after 7 and 90 days of oral
dosing in mice and after 90 days in rats, and in oral mucosa in mice after 7 days and rats
A consistent and coherent evidence
base shows redox reactions during
intracellular reduction of Cr(VI)
produce reactive oxygen species that
cause DNA damage in occupationally
exposed humans, experimental animal
studies, and in vitro studies, although
the evidence in animals exposed orally
is less consistent.
This document is a draft for review purposes only and does not constitute Agency policy.
3-145 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
at 90 davs, although no 8-OHdG adducts or protein oxidation in anv tissues (Thompson
et al., 2011; De Flora et al., 2008)
• Activation of genes involved in oxidative stress in the duodenum of mice exposed to
Cr(VI) for 90 days but not after 7 days
In vitro:
• Detection of reactive intermediates in acellular systems
• Oxidative stress in human primary and immortalized lung or Gl cells after exposure to
Cr(VI), including increased ROS production, oxidation of lipids and proteins, and
increased antioxidant enzyme activity
• Increased intracellular reduction via ascorbate correlates with free radical production,
oxidative DNA damage (e.g., 8-OHdG adducts, DNA strand breaks, DNA-protein
crosslinks, alkali labile sites) and lipid peroxidation
• Addition of antioxidants reduces/eliminates oxidative DNA damage; suppression of
antioxidants or use of DNA repair deficient cell line increases oxidative DNA damage
• Dose-dependent activation of NF-kB and AP-1, pro-inflammatory transcription factors
and redox-sensitive signaling molecules
Epigenetic
modifications
(Section 3.2.3.4;
Appendix C.3.2.4)
Inhalation exposure:
• Hvoermethvlation of tumor-suppressor genes pi6ink4a (Kondo et al., 2006) and APC (AN
et al., 2011) in chromate factorv workers with lung cancer who had occupational
chromate exposure compared to those without chromate exposure, and dysregulation
of tumor suppressor microRNAs that correlate with Cr blood levels (Li et al., 2014b).
• Hypermethylation of DNA mismatch repair and homologous recombination repair genes
in lung cancer cases with chromate exposure (Hu et al., 2018; AN et al., 2011; Takahashi
et al., 2005), leading to microsatellite instability
• Global hvpomethylation in chromium-exposed workers (Linqing et al., 2016; Wang et
al., 2012b)
Consistent, coherent evidence of
epigenetic alterations (heritable
changes in gene expression that are
not caused by changes in DNA
sequence) that correlate with Cr(VI)
exposure in humans and are known to
contribute to microsatellite instability,
mutagenicity, and carcinogenesis.
This document is a draft for review purposes only and does not constitute Agency policy.
3-146 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
In vitro: Extensive evidence of the epigenetic mechanisms of Cr(VI) (including methylation,
histone modifications, and miRNA) (reviewed in Chen et al. (2019)) and increased resistance
to apoptosis in human colon cells lacking a key mismatch repair gene when exposed to
Cr(VI). Transcriptomic changes consistent with epigenetic modifications in genes involved in
cvtotoxicitv/cell proliferation and DNA repair (Rager et al., 2019).
Inhibition of DNA
repair
(Section 3.2.3.4;
Appendix C.3.2.3)
Inhalation exposure: epigenetic suppression of genes involved in DNA repair in Cr(VI)-
exposed workers (summarized above)
In vitro: Inhibition of genes involved in mismatch repair (see above) and homologous
recombination repair, including RAD51 (Browning et al., 2016; Hu et al., 2016; Li et al., 2016;
Bryant et al., 2006)
Consistent, coherent evidence of the
epigenetic suppression of DNA
mismatch repair (see above) and
homologous recombination repair,
leading to increased DNA double-
strand breaks that are more likely to
cause mutations.
Genomic instability
(Section 3.2.3.4;
Appendix C.3.2.3)
In vitro: Consistent evidence of aneuploidv induced bv Cr(VI) ((Figgitt et al.. 2010). (Giierci et
al., 2000), Seoane et al. (2002; 2001,1999))
Besides the microsatellite instability
induced by epigenetic suppression of
DNA mismatch repair (see above),
Cr(VI) may also cause aneuploidy, a
hallmark of cancer. This evidence is
primarily from in vitro studies.
Genotoxicity and
mutagenicity
(Section 3.2.3.3;
Appendix C.3.2.2)
Inhalation exposure:
• Consistent evidence of increased micronucleus frequency from medium confidence
studies of the blood, nasal and oral cavity of exposed workers that correlated with
blood chromium levels (Long et al., 2019; El Saftv et al., 2018; Hu et al., 2018; Sudha et
al., 2011)
• Ten of 11 low confidence studies found increased micronuclei in workers despite
differences in population and exposure scenarios (Linaing et al., 2016; Wultsch et al.,
2014; Qavvum et al., 2012; Balachandar et al., 2010; larmarcovai et al., 2005; Danadevi
et al., 2004; Medeiros et al., 2003; Benova et al., 2002; Vaglenov et al., 1999)
• Consistent evidence of increased chromosomal aberrations in low confidence studies of
workers despite sensitivity concerns that biased toward the null (Balachandar et al.,
Consistent observations of heritable
structural and numerical genetic
damage in exposed humans, supported
by a small number of low confidence
studies in animals exposed via
inhalation or ingestion, with other
supporting evidence of genotoxicity
provided by supplemental studies
humans, animals, and in vitro.
This document is a draft for review purposes only and does not constitute Agency policy.
3-147 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
2010; Halasova et al., 2008; Maeng et al., 2004; Deng et al., 1988; Koshi et al., 1984;
Sarto et al., 1982)
• Increased mutation frequency in the lungs of transgenic rodents exposed via
intratracheal instillation, increasing with dose and post-exposure time, provides
biological plausibility for mutations in exposed target tissues (Cheng et al. (2000; 1998))
• Consistent supporting evidence of genotoxicity in studies of exposed humans and
animals dosed via i.p. injection, including DNA strand breaks, adducts, crosslinks, or
other DNA damage and repair-related endpoints (e.g., sister chromatid exchange)
(Appendix Table C-52)
• Correlation of svstemic Cr levels and other genotoxic endpoints (El Saftv et al., 2018;
Qavvum et al., 2012; Sudha et al., 2011; Danadevi et al., 2004)
• Correlation of MN with work duration (Danadevi et al., 2004)
Oral exposure:
• Some mixed evidence of micronucleus frequency in one low confidence study in the
bone marrow of Cr(VI)-exposed mice (NTP, 2007) and positive findings of mutation in
two low confidence studies in the developing mouse fetus (Schiestl et al., 1997) and in
male rat germ cells (Marat et al., 2018)
• Largely null findings of gene mutation or micronuclei in low confidence studies in the
bone marrow (De Flora et al.. 2006; Mirsalis et al.. 1996; Shindo et al.. 1989) or Gl tract
(Aoki et al., 2019; Thompson et al., 2015c; Thompson et al., 2015b; O'Brien et al., 2013)
of mice or rats, though all but one of these studies lacked sensitivity for detection due
to nontoxic dose ranges tested
In vitro:
• DNA reactivity and genotoxicity has been confirmed in a large evidence base of in vitro
studies (Appendix Table C-53)
Cytotoxicity and
degenerative cellular
changes
Biochemical markers of cell injury in the lung:
Consistent evidence of cytotoxicity and
degenerative cellular changes
observed in the lung and small
This document is a draft for review purposes only and does not constitute Agency policy.
3-148 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
(Sections 3.2.1, 3.2.2,
3.2.3.4)
• Concentration-related increases in total protein, albumin, and LDH activity have been
observed in rats exposed via inhalation for 30 and 90 davs to >0.05 mg/m3 Cr(VI) (Glaser
etal., 1990)
Atrophy and blunting of small intestinal villi:
• Observed to increase with dose in mice following drinking water exposures to >11.6
Cr(VI)/kg-d after 7 and 90 davs (Thompson et al., 2011)
• Observed in a significant proportion of mice at all doses after 90 day (>3 mg Cr(VI)/kg-d)
or 2 vear (>0.3 mg/kg-d) drinking water exposures in mice (not observed in rats) (NTP.
2008, 2007)
• Also observed in rats at 7.2 mg Cr(VI)/kg-d in drinking water (Thompson et al., 2012b)
Cytoplasmic vacuolization of small intestinal villi:
• Observed in mice following >11.6 mg Cr(VI)/kg-d in drinking water for 7 days and >4.6
mg Cr(VI)/kg-d in drinking water for 90 davs (not observed in rats) (Thompson et al.,
2011)
• Observed at all doses (>3 mg Cr(VI)/kg-d) in drinking water after 90 days exposure in
drinking water (qualitative data) (not observed in rats) (NTP, 2007)
Apoptosis in the lung and small intestine:
• Lung: One intratracheal instillation exposure study in rats observed increased apoptosis
in bronchial epithelium and lung parenchyma; in vitro studies support dose and time-
dependent increases in apoptosis following Cr(VI) exposure in human lung cells
(Reynolds et al., 2012; Azad et al., 2008; Reynolds and Zhitkovich, 2007; Gambelunghe
et al., 2006; D'Agostini et al., 2002; Carlisle et al., 2000)
• Small intestine, mouse: Apoptotic villi increasing with dose >11.6 Cr(VI)/kg-d in drinking
water for 90 davs; not observed after 7 davs (Thompson et al., 2015b; Thompson et al.,
2015a; O'Brien et al., 2013; Thompson et al., 2011)
• Small intestine, rat: Apoptotic villi at >7.2 mg Cr(VI)/kg-d in drinking water (Thompson
etal., 2012b)
intestine of animals following
inhalation and drinking water
exposures, respectively.
This document is a draft for review purposes only and does not constitute Agency policy.
3-149 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
Suppression of
apoptosis
(Section 3.2.3.4;
Appendix C.3.2.10 and
C.3.3)
Oral exposures:
• Inhibition of the MAPK inhibitor RKIP was observed in the stomach and colon of male
Wistar rats after 60 days of exposure to Cr(VI) in drinking water, leading to the
activation of the ERK/MAPK signaling pathwav (Tsao et al., 2011)
• Activation of the ERK/MAPK signaling pathway promotes cell proliferation (via c-Myc
expression activation) and has been observed in rat stomach and colon after oral
exposure (Tsao et al.. 2011)
Biologically plausible evidence of the
suppression of apoptosis, a hallmark of
cancer, in the stomach and colon of
animals exposed via drinking water.
Cell proliferation
(Section 3.2.3.4;
Appendix C.3.2.10 and
C.3.3)
Inhalation exposures:
• Cyclin Dl, a regulator and promoter of cell cycle progression, has been detected at
significantly increased levels in the lung tumor tissues of chromate-exposed patients
compared to unexposed lung cancer patients. Increased expression of cyclin Dl has
been associated with cell proliferation and tumorigenesis (Katabami et al., 2000)
Oral exposures:
• The cellular replication marker Ki-67, which is upregulated in human intestinal
adenomas, has been found to be increased in isolated duodenal mucosal cells from the
small intestine of mice exposed to Cr(VI) via drinking water for 7 and 90 davs (Rager et
al., 2017; Kopec etal., 2012a)
• The c-Myc oncogene codes for a pro-proliferation transcription factor and can be
activated by Wnt or the MAPK/ERK pathway, though it can also be blocked by NF-kB
signaling. A dose-dependent increase in the c-Myc oncogene was found in the stomach
and colon of male Wistar rats after 60 days of exposure to Cr(VI) in drinking water (Tsao
etal., 2011)
• Galectin-1, associated with gastric cancer cell motility and overexpressed in gastric
tumor cells and digestive cancers, was increased in the stomach and colon of male
Wistar rats after 60 davs of exposure to Cr(VI) in drinking water (Tsao et al., 2011)
Biologically plausible evidence of
increased cell proliferation, a hallmark
of cancer, as interpreted by the
aberrant expression of genes related to
cell cycle regulation in lung tumor
tissues of humans exposed to Cr(VI)
and in the stomach, duodenum and
colon of animals exposed via drinking
water.
Regenerative
hyperplasia
Focal epithelial hyperplasia of the small intestine:
Consistent evidence of hyperplasia
interpreted to be the result of
This document is a draft for review purposes only and does not constitute Agency policy.
3-150 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
(Sections 3.2.1, 3.2.2,
3.2.3.4)
• Observed in mice exposed to >1.18 mg (females) and >2.4 mg (males) Cr(VI)/kg-d in
drinking water for 2 years. The responses were not statistically significant, but this is
considered a biologically significant pre-neoplastic lesion due to morphologic similarity
to adenoma (NTP, 2008)
Diffuse epithelial hyperplasia of the lung and small intestine:
• Lung: Bronchioalveolar hyperplasia (70-100%) observed in rats following 0.050-0.40
mg/m3 Cr(VI) inhalation exposure for 30 days, but incidence was decreased at 90 days
(Glaser et al.. 1990)
• Small intestinal crypt cells, mice: Hyperplasia reported in mice exposed for 7 days at
31.1 mg Cr(VI)/kg-d (NS) in drinking water with no changes in mitotic activity in crypt
cells and following 90 davs at >11.6 mg Cr(VI)/kg-d (non-dose-dependent) (Thompson et
al., 2015b; Thompson et al., 2011)
• Small intestine, mice: Hyperplasia observed at all doses (>3 mg Cr(VI)/kg-d) in drinking
water for 90 days, minimal to mild severity, 100% incidence at mid/high dose levels,
with increased numbers of mitotic figures in the hyperplastic epithelium (in females and
four male datasets in multiple strains). Also observed at all doses (>0.3 mg Cr(VI)/kg-d)
in drinking water for 2 years, increasing with dose, minimal to mild severity, with
increased numbers of mitotic figures in the hyperplastic epithelium (NTP, 2008, 2007)
• Small intestinal villous cells, rats: Hyperplasia observed at >7.2 mg Cr(VI)/kg-d in
drinking water for 7 and 90 davs (Thompson et al., 2012b)
regeneration following cell injury
following oral exposures in the small
intestine of mice and rats and
following inhalation exposures in the
lung in rats.
Inflammation
(Section 3.2.3.4;
Appendix Table C-38)
In the lung:
• Increases in macrophages in BALF at 0.9 mg/m3 Cr(VI) inhalation exposure for 4-6 weeks
in rabbit and at at 0.20 and 0.40 mg/m3 Cr(VI) for 30 and 90 davs in rats (Glaser et al.,
1990; Johansson et al., 1986b)
• In rats exposed for 28 and 90 days, increased lymphocytes in BALF at 0.025 mg/m3 and
0.05 mg/m3 Cr(VI); increased granulocytes/neutrophils at 0.05 mg/m3 Cr(VI); no change
or decreased number of macrophages at 0.050 and 0.20 mg/m3 Cr(VI) inhalation
exposure. In rats exposed for 4-48 weeks, increased granulocytes/neutrophils; no
Consistent evidence of chronic
inflammation, an enabling
characteristic of cancer, has been
observed in the lung of animals
following Cr(VI) inhalation. There is no
histopathological evidence in the Gl
tract consistent with chronic
inflammation reported following oral
exposures in animals, although some
indirect evidence consistent with
inflammation has been reported.
This document is a draft for review purposes only and does not constitute Agency policy.
3-151 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Biological events (and
relevant sections)
Summary of key findings
Interpretations, judgments, and
rationale
change or decreased number of macrophages at 0.36 mg/m3 Cr(VI) inhalation exposure
(Cohen et al., 2003; Glaser et al., 1985)
• Histiocytosis (macrophage accumulation) associated with inflammation observed in rats
and rabbits exposed via inhalation for 30-90 davs (Kim et al., 2004; Glaser et al., 1990;
Johansson et al., 1986a)
In the Gl tract:
• Cytokine fluctuations observed in the duodenum (and not the oral mucosa) of mice (4/
IL-ip and TNF-a) and rats CT* IL-la, IL-6; 4, IL-4) following Cr(VI) exposure in drinking
water
• Induction of proinflammatory signaling pathways (e.g., NF-kB) in animals following oral
exposures
This document is a draft for review purposes only and does not constitute Agency policy.
3-152 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
Susceptible populations and life stages
A number of different factors were identified that could predispose some populations of
humans to be more susceptible to Cr(VI) carcinogenicity when ingested. These factors are listed
below and in Section 3.3.1.
Low stomach acid
Because extracellular reduction of Cr(VI) to Cr(III) serves as a detoxifying mechanism,
conditions that would lower an individual's ability to effectively reduce Cr(VI) could lead to a
higher rate of Cr(VI) absorption into the cells lining the GI tract Following oral ingestion, gastric
emptying to the small intestine competes with the rapid extracellular reduction to Cr(III) by gastric
juices (Proctor etal.. 2012: De Flora et al.. 19971. However, there is significant inter individual
variability of stomach pH in the human population. Individuals taking medication to treat
gastroesophageal reflux disease (GERD), including calcium carbonate-based acid reducers and
proton pump inhibitors, have an elevated stomach pH during treatment Individuals with a
preexisting low stomach acid condition (hypochlorhydria, also known as achlorhydria) consistently
have a high gastric pH of approximately 8 (Kalantzi etal.. 2006: Feldman andBarnett. 1991:
Christiansen. 19681. This condition may be caused or exacerbated by multiple other preexisting
gastric conditions, including H. pylori infection. The prevalence of hypochlorhydria (see above) is
believed to be high in elderly populations (age 65 and up) (Doki etal.. 2017). The general healthy
population also exhibits high variability in stomach pH. Among adults without hypochlorhydria
and who do not regularly take antacids, 5% of men may exhibit basal pH exceeding 5, and 5% of
women may exhibit basal pH exceeding 6.8 fFeldman and Barnett. 19911. Neonates have neutral
stomach pH at birth (Neal-Kluever etal.. 2019) (see Section 3.3.1.3).
Genetic polymorphisms
Individuals with genetic polymorphisms conveying haploinsufficiencies in DNA repair or
tumor suppressor genes may have increased susceptibility to Cr(VI)-induced cancer. Several
studies in humans have identified polymorphisms in genes related to DNA repair and tumor
suppression that were correlated with increased genetic damage and lung cancer (summarized
above and in Appendix C.3.14; see also (Urbano etal.. 2012)1. DNA adducts formed directly by
chromium or indirectly via oxidative damage are substrates for nucleotide excision repair (for
bulky lesions) and mismatch repair (for misincorporated bases during DNA replication and
homologous recombination); heritable deficiencies in the effectiveness of these repair processes
can cause a higher rate of unprocessed genetic damage leading to the formation of heritable
mutations.
This document is a draft for review purposes only and does not constitute Agency policy.
3-153 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Carriers of the cystic fibrosis mutant allele
The analyses by US EPA (see Appendix C. 3.4.2 and Mezencev and Auerbach (202111 of the
toxicogenomic data reported in Kopec et al. (2012b: 2012a) from mice exposed to Cr(VI) (reviewed
earlier in this section) have identified a potential role for CFTR in the carcinogenic effects of Cr(VI).
Tumorigenicity of impaired CFTR activity in animal models supports the relevance of the Cr(VI)-
mediated inactivation of CFTR for the development of small intestinal tumors in mice exposed to
Cr(VI) in drinking water (Than etal.. 20161. and CFTR has been shown to act as a tumor suppressor
in the human colon fThanetal.. 20161. These findings may indicate that carriers of the mutated
CFTR allele could be more sensitive to the Cr(VI)-mediated carcinogenicity. In the US alone, more
than 10 million people are carriers of a mutated CFTR allele that confers an approximate 50%
reduction in CFTR expression levels. Although these individuals do not develop cystic fibrosis, the
deficit in CFTR function has been shown to lead to an increased risk for several conditions
associated with the disease, including colorectal cancer (OR = 1.44, 95% CI: 1.01-2.05) fMiller etal..
20201. CFTR suppression induced by low Cr(VI) exposures in drinking water can be expected to
occur in all exposed populations, but a more significant effect would be expected in humans already
producing low levels of this protein.
3.2.3.5. Integration of Evidence for Cancer of the GI Tract
The integrated evidence for Cr(VI)-induced cancer of the GI tract is summarized in Table 3-
22. Overall, Cr(VI) is likely to be carcinogenic to the human GI tract. This conclusion is based on
robust evidence of cancer from a high confidence 2-year cancer bioassay conducted by NTP, which
showed a statistically significant increase in oral cavity tumors in male and female F344/N rats and
small intestine neoplasms in B6C3F1 male and female mice fNTP. 20081. Notably, at the lower
doses where tumor occurrence was nonsignificant compared to concurrent controls, incidences
exceeded NTP historical controls in both species. Therefore, some tumors that were not
statistically significant may be biologically significant due to the increasing trend and low historical
control incidence (Appendix D.2).
The evidence of carcinogenicity of the GI tract from human studies is slight based on studies
of the oral route of exposure. Results for two populations exposed to Cr(VI) through drinking water
in China and Greece were available in the epidemiological evidence base that analyzed stomach
cancer risk (Linos etal.. 2011: Kerger etal.. 2009: Beaumont etal.. 20081. The studies reported
increased SMRs when their mortality experience was compared to other communities in the
surrounding areas or to the mortality experience in the province where the exposed communities
were located. While uncertainties in the study methods and analyses resulted in low confidence
ratings, the studies in both populations reported increased risk estimates, supporting a judgment of
slight.
The evidence from the meta-analysis of GI tract cancer risk from the occupational studies of
workers with inhalation exposure to Cr(VI) is indeterminate. The summary effect estimates showed
This document is a draft for review purposes only and does not constitute Agency policy.
3-154 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
small increases in risk for each cancer site, and this increase was statistically significant for rectal
cancer. There were few studies reporting odds ratios, but in each case (esophagus and stomach),
summary effect estimates based on these studies were somewhat higher compared with summary
estimates based on other relative risk measures (although neither odds ratio-based estimate was
statistically significant). However, there were not clear patterns of risk by either occupational
group or specific cancer site. Due to potential misclassification and heterogeneity of Cr(VI)
exposure among and within the included studies, there may have been a decreased ability to detect
an association if it existed.
Although interspecies correlation is lacking for the exact tumor site within the intestinal
tract, the available evidence in animals and humans with overall species concordance spanning the
entire alimentary tract, including the oral cavity, is robust (with the acknowledgment that there is
not a requirement to establish site concordance to draw a conclusion for cancer hazard). While it is
difficult to draw conclusions regarding an association between human exposure to Cr(VI) through
drinking water or inhalation and GI tract cancer from the available epidemiological evidence, there
is consistency among species (human, rat, and mouse) regarding the potential for Cr(VI) to cause
cancer at various sites along the GI tract.
Potential MOAs for carcinogenicity induced by ingested Cr(VI) in the mouse small intestine
include mutagenicity and regenerative proliferation caused by tissue injury leading to a higher
probability of the clonal outgrowth of spontaneous mutations. These mechanistic processes are not
mutually exclusive, and there is evidence that Cr(VI)-induced carcinogenesis in the GI tract after
oral exposure involves both MOAs.
Bioavailability results and kinetic considerations (see Section 3.1 and Appendix C.l) lead to
the conclusion that approximately 10-20% of ingested low dose Cr(VI) escapes human gastric
inactivation and could therefore reach the target cells in appreciable amounts and would thus be
reasonably anticipated to act as a mutagen in the GI tract epithelium. Given the cellular capacity for
uptake of Cr(VI) in highly absorptive intestinal tissues, it is biologically plausible that Cr(VI) can
induce genetic damage in the human GI tract. By assuming significant (80-90%) but incomplete
gastric detoxification, the capacity for autonomous growth may remain latent for weeks, months, or
years, during which time an initiated cell may be phenotypically indistinguishable from other
parenchymal cells in that tissue. The average tumor diagnosis was over 700 days (100 weeks) for
both sexes of mice (first onset at 451 days and most observed at terminal sacrifice). Most human
and animal neoplasms studied to date are of monoclonal origin. There are several salient
characteristics of initiation. It can occur following a single exposure to a known carcinogen.
Changes produced by the initiator may be latent for weeks or months and are considered
irreversible. The hyperplasia observed at the 2-year evaluation endpointmay, therefore, be a
manifestation of intestinal responses to late clonal expansion following an early initiation. Also,
with age, spontaneous DNA replication becomes more error prone resulting in small intestinal
This document is a draft for review purposes only and does not constitute Agency policy.
3-155 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
tumors. Therefore, the hyperplastic changes described could support either MOA (cytotoxicity with
regenerative cell proliferation and mutagenicity).
The hypothesis that continuous wounding results in regenerative proliferation that may
give rise to spontaneous mutations progressing to neoplasia is largely supported by
histopathological findings that indicate degenerative changes including villous blunting/atrophy
accompanied by cytoplasmic vacuolization and crypt hyperplasia. Importantly, it is unlikely that
this MOA is solely operational in the intestinal tumors observed by NTP after 2 years. While a
'wounding and regenerative cell proliferation' MOA is supported by short-term (7 and 28 day) and
subchronic (90 day) bioassays, these studies were (a) too short in duration to show that
regenerative hyperplasia progressed to tumor formation (which could support a threshold dose)
and (b) did not demonstrate that a mutagenic MOA could reliably be excluded. Therefore, whether
the clonal selection and outgrowth of spontaneous mutations is responsible for Cr(VI)
tumorigenesis remains a data gap; DNA sequencing data may assist with assessing the validity of
this hypothesis.
No direct mechanistic evidence in the rat oral mucosa is available to support an MOA for
tumorigenesis of the rat oral cavity induced by ingested Cr(VI). It is important to note that the
apical membrane of the human tongue, oral mucosa, and esophagus will come into direct contact
with Cr(VI) in ingested drinking water before gastric detoxification. This is supported by consistent
observations of increased micronuclei in oral epithelial cells from humans occupationally exposed
to Cr(VI). Importantly, the proposed wounding and regenerative proliferation MOA for the
intestinal tumors in mice does not address the Cr(VI) oral cavity tumors of rats, in which neither
degenerative changes nor hyperplasia were observed. Only one low confidence study investigated
the mutation frequency in the rat oral cavity and did not find an increase after a 7-day exposure to
Cr(VI) in drinking water. Additional studies designed to be sensitive for detecting mutations as
well as other potential mechanisms involved in carcinogenicity of the oral mucosa are needed.
Overall, the determination of a mutagenic MOA, the incompleteness of gastric detoxification,
and the development of oral cavity tumors without any apparent tissue injury or regenerative
proliferation argue against a threshold for low dose extrapolation of cancer risk for both oral and GI
tract tumors from ingested Cr(VI). Because a mutagenic mode of action for Cr(VI) carcinogenicity is
"sufficiently supported in (laboratory) animals" and "relevant to humans," EPA uses a linear low
dose extrapolation from the POD in accordance with Guidelines for Carcinogen Risk Assessment
(U.S. EPA. 2005a). The oral slope factor derivation for cancer is described in Section 4.3.
This document is a draft for review purposes only and does not constitute Agency policy.
3-156 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-22. Evidence profile table for cancer of the GI tract3
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes,
and confidence
Summary of key findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and rationale
Evidence from studies of exposed humans
®©o
Cr(VI) is likely to be
carcinogenic to humans
via the oral route of
exposure.
Robust evidence shows
tumors of the GI tract in
mice (small intestine) and
rats (oral cavity) in both
sexes; the oral cavity
tumors were rare
indicating increased
biological significance.
Evidence from humans is
slight but is consistent in
GASTRIC CANCER
(ORAL)
Low confidence:
Beaumont et al.
(2008)
Kerger et al. (2009)
Linos et al. (2011)
Results for two populations
(three publications) in China
and Greece exposed to
Cr(VI) in drinking water
showed increased SMRs.
Ecological study designs
(lack of individual estimates
of exposure), the uncertain
nature of the mortality data
for that period in China, and
the potential impact of
confounding by differences
in SES between comparison
groups are sources of
uncertainty.
• Consistency across
geographical
locations and
multiple referent
groups
• Low confidence
studies
©oo
Slight
Despite findings of
increased SMRs in two
separate studies, these
low confidence ecological
study designs reported
imprecise estimates that
changed in magnitude
depending on the
definition of the
unexposed communities.
This document is a draft for review purposes only and does not constitute Agency policy.
3-157 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
reporting some risk of
cancers of the Gl tract in
humans exposed via
drinking water.
Biological plausibility for
the small intestinal
tumors is provided by
histopathological
evidence of tissue
degeneration and
hyperplasia in the small
intestine of mice and
molecular evidence of
cell proliferation and
oxidative stress in these
animals prior to tumor
formation.
A primary role for
mutagenicity, evident in
oral cavity tissues of
exposed humans and
known to occur when
Cr(VI) comes into direct
contact with cells, in Gl
tract tumorigenesis (and in
particular, in tumors of the
rat oral cavity) is not clear
Studies, outcomes,
and confidence
Summary of key findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and rationale
Gl TRACT CANCER
(INHALATION/ORAL)
Medium confidence:
43 occupational
studies of cancer
mortality or
incidence
A meta-analysis of Gl tract
cancer risk from
occupational studies of
workers with inhalation and
oral (swallowing dust)
exposure to Cr(VI) showed
small increases in risk for
each cancer site.
Occupations with a higher
certainty of exposure to
Cr(VI) showed higher
summary effect estimates.
The summary estimates for
SMR/SIR analyses of rectal
cancer were statistically
significant. The summary
estimates for the few
studies reporting odds
ratios (esophagus and
stomach) were somewhat
higher (although neither
odds ratio-based estimate
was statistically significant).
• Precision across
studies in meta-
analysis increased
by combining
information across
multiple studies for
certain analyses
• Effects observed
despite reduced
sensitivity resulting
from expected
exposure
misclassifi cation
• Exposure-response
gradient (in studies
with better
exposure
assessment
methods)
• Lack of coherence
by cancer site and
occupational
groupings
ooo
Indeterminate
Although the risk estimate
for rectal cancer was
statistically significant,
and coherent results for
colon cancer risk were
found when stratified by
occupational groupings
expected to have higher
exposures to Cr(VI),
inconsistencies in patterns
of risk across occupational
groups raise uncertainties.
Evidence from animal studies
This document is a draft for review purposes only and does not constitute Agency policy.
3-158 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
but also cannot be ruled
out. A mutagenic mode
of action for Cr(VI)
carcinogenicity is
considered sufficiently
supported in (laboratory)
animals and relevant to
humans.
Susceptibility is assumed
for humans with impaired
ability to reduce Cr(VI) in
the stomach.
Studies, outcomes,
and confidence
Summary of key findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and rationale
Gl TRACT TUMORS
(ORAL)
High confidence:
NTP(2008)
Statistically significant
increases in tumors of the
Gl tract were reported in a
high confidence 2-year
animal bioassay: adenomas
and carcinomas of the small
intestine (male and female
mice), and squamous cell
carcinomas and papillomas
of the oral mucosa and
tongue (male and female
rats).
Tumors of the oral cavity
and small intestine have a
very low historical
incidence.
• Consistent findings
in one high
confidence 2-year
study that
contained
bioassays in rats
and mice of both
sexes
• Coherent,
biologically related
findings within the
Gl tract
• Large magnitude of
effects
• Strong dose-
response gradient
• Mechanistic
evidence provides
biological
plausibility
• No factors noted
©0©
Robust
Consistent findings in one
large high confidence
study finding tumors in
the Gl tract in two species
and both sexes.
Animal mechanistic
evidence informing
biological plausibility
(hyperplasia in mouse
small intestine may be a
precursor event for
tumors).
HISTOPATHOLOGICAL
CHANGES (ORAL)
High confidence:
NTP (2008)
NTP (2007)
Thompson et al.
(2012b)
Thompson et al.
(2011)
Degenerative changes in
intestinal villi and
hyperplasia of the small
intestine observed in male
and female mice by NTP
(2008, 2007), and in
female mice and rats by
Thompson et al. (2012b;
2011).
• Consistent findings
in chronic and
subchronic studies
that contained
multiple bioassays
in rats and mice of
both sexes, and
• Inconsistent
observations of
hyperplasia
between mice and
rats, though this is
explained in part
by
©©©
Robust
Histopathological changes
reported in high
confidence studies (tissue
injury and proliferative
changes) observed across
the animal evidence base
database are coherent
This document is a draft for review purposes only and does not constitute Agency policy.
3-159 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes,
and confidence
Summary of key findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and rationale
Histiocytic cellular
infiltration observed in the
small intestine of male and
female rats and mice in all
studies and bioassays.
multiple strains of
mice
• Large magnitude of
effects
• Strong dose-
response gradient
• High confidence
studies observing
an effect
• Mechanistic
evidence provides
plausibility
• Coherence as
potential
preneoplastic
lesions in the
mouse small
intestine only
pharmacokinetic
differences
following chronic and/or
subchronic oral exposures
in rats and mice and
suggest adverse effects of
Cr(VI) on the Gl tract,
findings that are
supported by mechanistic
evidence.
aSee Table 3-21 for the summary of key mechanistic events involved in Cr(VI)-induced cancer.
This document is a draft for review purposes only and does not constitute Agency policy.
3-160 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofHexavalent Chromium
3.2.4. Hepatic effects
The liver is a common site of toxicity as it functions to metabolize exogenous as well as
endogenous chemicals. The liver is considered an accessory digestive organ because it synthesizes
proteins and compounds necessary for digestion as well as filtering and metabolizing nutrients and
toxicants absorbed by the small intestine (first-pass effect). The liver also metabolizes chemicals
absorbed into the bloodstream from other routes (such as intravenous injection or inhalation).
Because of the first-pass effect, the liver may be affected more severely by toxic chemical exposure
via the oral route as compared to the inhalation route.
3.2.4.1. Human Evidence
Study evaluation summary
There are four studies that reported on the association between Cr(VI) exposure and
hepatic-related clinical chemistry measures, including alanine aminotransferase (ALT), aspartate
aminotransferase (AST), and alkaline phosphatase (ALP). Increases in serum ALT and AST are
considered indicative of hepatocellular damage, with ALT considered to be the more sensitive and
specific indicator (EMEA. 2008: Boone etal.. 2005). Increases in ALP can be associated with liver
cholestasis, however, ALP is not as specific to liver injury as extrahepatic sources of ALP exist
(Boone etal.. 2005). Other serum measures evaluated which can help inform liver toxicity,
included: bilirubin, albumin, total protein, creatinine, and albumin/globulin ratio. In general,
increased serum bilirubin and decreased serum albumin/total protein can indicate impaired liver
function fEMEA. 2008: Boone etal.. 20051. A fifth study, Tong and Zhang f20031. reported
hepatomegaly among chromium workers but was found to be uninformative due to multiple critical
deficiencies and is not further discussed.
With respect to confidence in the human studies, one study (Khan etal.. 2013) was
classified as uninformative because exposure was based on tannery work, and there was insufficient
information provided on the specific tanning processes used at the facility33. This study was not
considered further. The three remaining studies were included and classified as low confidence
(see Table 3-23), with two fSaraswathv and Usharani. 2007: Lin etal.. 19941 conducted in
occupational populations with exposure primarily via inhalation and one (Sazakli et al.. 2014) in the
general population with exposure primarily via the oral route. Lin etal. (1994) had adequate
exposure measurement due to use of air sampling with appropriate methods and categorization
into three levels of exposure, but was deficient due to incomplete reporting of results and
confounding. In the remaining two studies, the primary limitation was deficient exposure
33Leather tanning processes that can potentially lead to Cr(VI) exposure include: (1) use of a two-bath
process, (2) on-site production of tanning liquors, and (3) leather finishing steps that involve Cr(VI) (e.g., use
of Cr (VI)-containing pigments) (Shaw Environmental. 20061. If these processes are not specified by the
study, it cannot be determined whether exposure was to Cr(VI) or Cr(III).
This document is a draft for review purposes only and does not constitute Agency policy.
3-161 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
measurement, primarily due to concerns about potential for nondifferential exposure
misclassification that would be likely to bias the results towards the null fSazakli etal.. 2014:
Saraswathv and Usharani. 20071. In Sazakli etal. (20141. exposure was estimated based on water
intake and blood and hair Cr concentrations, but there were poor correlations across measures. In
Saraswathv and Usharani (2007). no air data was available and there was no quantitative
measurement of exposure. These considerations on exposure measurement are the primary basis
for the clinical chemistry outcome judgments presented in Table 3-23.
Table 3-23. Summary of human studies for Cr(VI) hepatic effects and overall
confidence classification [high (H), medium (M), low (L)] by outcome. Click to
see interactive data graphic for rating rationales.
Author (year)
Industry
Location
Exposure
Measurement
Study Design
Clinical
Chemistry
Lin etal. (1994)
Chrome plating
Taiwan
Urine, Air, Work
category
Cross-sectional
L
Sazakli et al. (2014)
General
population
Greece
Urine, Hair, Modeled
lifetime Cr(VI)
exposure dose
Cross-sectional
L
Saraswathv and
Usharani (2007)
Chrome plating
India
Work category
Cross-sectional
L
aStudies excluded due to critical deficiency in one or more domains: Khan et al. (2013) and Tone and Zhang (2003).
Synthesis of evidence in humans
Two studies fSazakli etal.. 2014: Saraswathv and Usharani. 20071 reported statistically
significant changes consistent with liver dysfunction in at least one of the tests (i.e., higher levels of
ALT, AST, ALP, or bilirubin and/or lower levels of total protein or albumin with higher exposure) as
shown in Table 3-24. These associations were observed despite the potential for exposure
misclassification that may have reduced sensitivity. Saraswathv and Usharani f20071 observed an
exposure-response gradient across the three exposure categories for ALT, AST, ALP, and total
protein. However, there is some inconsistency in the direction of results for total protein and
albumin between the two studies. The third study fLin etal.. 19941 evaluated serum ALT, AST,
creatinine, and albumin/globulin ratio. The study authors did not report quantitative results but
reported that there were no significant differences among workers in the four exposure groups.
This document is a draft for review purposes only and does not constitute Agency policy.
3-162 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-24. Associations between Cr(VI) and liver clinical chemistries in
epidemiology studies
Reference,
confidence
Population
Exposure
comparison and
effect estimate
ALT
AST
ALP
Total
protein
Other
Sazakli et al.
(2014), low
confidence
Cross-sectional in
Greece, general
population; two
drinking water
exposure groups
(n = 237) and
controls (n = 67)
Regression
coefficients for
calculated lifetime
exposure dose and
hair biomarkers
Lifetime:
-0.03
(for In-ALT)
Hair: 0.05
(for In-ALT)
Lifetime:
0.04
Hair: 0.04
Lifetime:
0.12*
Hair: 0.22*
Lifetime:
0.14*
Hair: 0.24*
Lifetime:
Albumin
0.21*
Bilirubin
-0.11
Hair:
Albumin
0.23*
Bilirubin
-0.07
Saraswathv
and
Usharani
(2007), low
confidence
Cross-sectional in
India, two
chrome plater
groups (n = 130)
and male area
residents
(n = 130)
Means ± SD for
control/
exposed
8-15 yrs (A)/
exposed 16-25 yrs
(B)
Control:
22.0 ± 1.7
Exposed A:
34.3 ±2.5*
Exposed B:
43.3 ± 1.7*
Control:
19.2 ±2.1
Exposed A:
32.9 ±3.7*
Exposed B:
38.6 ±4.0*
Control:
60.8 ±5.7
Exposed A:
70.2 ±6.2*
Exposed B:
83.7 ±7.6*
Control:
7.8 ±0.4
Exposed A:
7.5 ±0.1*
Exposed B:
6.1 ±0.1*
NR
Lin et al.
(1994), low
confidence
Cross-sectional in
Taiwan, three
chrome plater
groups (n = 79)
and aluminum
plater referent
group (n = 40)
Analysis and
quantitative
results not
reported.
ALT, AST, serum creatinine and albumin/globulin ratio
evaluated, however, authors report no significant difference
among workers across exposure groups (results not shown).
*p < 0.05.
NR: not reported.
1 In addition, four studies (presented in five publications) reported on mortality attributable
2 to cirrhosis of the liver, all based on occupational cohorts (Birk etal.. 2006: Moulin etal.. 2000:
3 Moulin et al.. 1993b: Moulin etal.. 1993a: Moulin etal.. 19901. These studies indicated no increase
4 in cirrhosis mortality with higher exposure levels, but this evidence is considered inadequate to
5 assess the association with Cr(VI) due to several limitations, including lack of control of potential
6 confounding (such as by alcohol consumption), concerns about sensitivity and specificity of the
7 exposure measures, and the sensitivity of mortality as the outcome measure.
8 Overall, there is an indication in the available human studies that higher Cr(VI) exposure
9 may be associated with increased liver dysfunction, but there is some inconsistency in the available
10 results, and limitations especially with respect to exposure measurement.
This document is a draft for review purposes only and does not constitute Agency policy.
3-163 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 3.2.4.2. Animal Evidence
2 Study evaluation summary
3 Information relevant to the evaluation of an association between Cr(VI) exposure and liver
4 effects comes from oral and inhalation studies in mice and rats involving subchronic, chronic, and
5 gestational exposures. Liver effects evaluated in this synthesis include changes in liver histology,
6 clinical chemistry, and relative liver weight. As displayed in Table 3-25, studies reporting liver
7 effects in the Cr(VI) evidence base were of varying study quality (based on factors including
8 strength of study design and transparency of reporting), with the most informative evidence from
9 the NTP chronic and subchronic drinking water bioassays in rats and mice fNTP. 2008. 20071.
Table 3-25. Summary of included animal studies for Cr(VI) liver effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.3
Click to see interactive data graphic for rating rationales.
Liver outcomes
Author (year)
Species (strain)
Exposure design
Exposure route
Organ weight
Clinical
chemistry
Histopathology
Acharva et al. (2001)
Rat (Wistar), male
Chronic
Drinking water
L
L
Chopra et al. (1996)
Rat (Wistar), female
Subchronic
Drinking water
M
L
L
Elshazlv et al. (2016)
Rat (Sprague-Dawley)
Subchronic
Drinking water
-
M
M
Glaser et al. (1985)
Rat (Wistar)
Subchronic
Inhalation
L
L
M
Glaser et al. (1986)
Rat (Wistar)
Chronic
Inhalation
L
L
U
Kim et al. (2004)
Rat (Sprague-Dawley)
Subchronic
Inhalation
M
M
-
Krim et al. (2013)
Rat (Albino)
Subchronic
Gavage
-
M
-
Meenakshi et al.
(1989)
Rat (Wistar)
Subchronic
Gavage
-
L
U
Mo et al. (2018)
Rabbit (New Zealand), male
and female
Subchronic
Gavage
-
-
L
Navva et al. (2017a)
Rat (Wistar), male
Subchronic
Gavage
-
M
L
NTP (1997)
Mouse (BALB/c)
Reproductive study-
continuous breeding
(F0 to F2)
Diet
H
NTP(1996a)
Mouse (BALB/c)
Subchronic
Diet
NTP (1996b)
Rat (Sprague-Dawley)
Subchronic
Diet
NTP(2007)
Rat (F344/N); Mouse
(B6C3F1, BALB/c, C57BL/6)
Subchronic
Drinking water
This document is a draft for review purposes only and does not constitute Agency policy.
3-164 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofHexavalent Chromium
Liver outcomes
Author (year)
Species (strain)
Exposure design
Exposure route
Organ weight
Clinical
chemistry
Histopathology
NTP(2008)
Rats (F344/N); Mouse
(B6C3F1)
Chronic
Drinking water
H
H
Rafael et al. (2007)
Rat (Wistar)
Subchronic
Drinking water
-
M
L
Wang etal. (2015)
Rat (Sprague-Dawley), male
Subchronic
Drinking water
M
M
M
(Younan et al., 2019)
Rat (Wistar), male
Subchronic
Diet
L
M
U
aSeven studies reporting liver endpoints met PECO criteria but were considered to be uninformative at the study
evaluation stage: Kumar and Barthwal (1991), Geetha et al. (2003); Asmatullah and Noreen (1999), Nettesheim et
al. (1971), Soudani et al. (2013), Sanchez-Martin et al. (2015), and MacKenzie et al. (1958).34
Synthesis of evidence in animals
Histopathology
Several subchronic and chronic studies in rats and mice reported histological lesions in the
liver associated with oral exposure to Cr(VI). These lesions include increased inflammation and
infiltration of immune cells (Elshazlv et al.. 2 016: NTP. 2008. 20071. cytoplasmic vacuolation (fatty
changes) fElshazlv etal.. 2016: NTP. 2008: Acharva et al.. 2001: NTP. 1997: Chopra etal.. 1996:
NTP. 1996a). indications of apoptosis and necrosis (Elshazlv etal.. 2016: Acharva etal.. 2001:
Chopra etal.. 19961. and increased hepatocellular foci fElshazlv etal.. 2016: NTP. 20081. These
findings are presented in more detail below (see also Figure 3-19). While some NTP studies
observed histological lesions, several other NTP studies failed to find altered histological findings in
the liver. These studies include an oral study that exposed male and female SD rats to doses of up
to approximately 10 mg Cr(VI)/kg-day for 9 weeks fNTP. 1996bl. as well as a 3-month study in
three different strains of mice fNTP. 20071. A 3-month study in B6C3F1 mice reported a lack of
histological changes in the liver (NTP. 20071 as well as a 9-week oral study in BALB/c mice
(although some non-statistically significant increases in vacuolation were observed) fNTP. 1996al.
In addition, no treatment-related lesions in the liver were found in male and female BALB/c F0 or
F1 mice exposed orally in a continuous breeding study at doses of 30-50 mg Cr(VI)/kg-day for
approximately 20 weeks (NTP. 19971. Across the evidence base, there is some indication that mice
may be more resistant than rats to Cr(VI)-induced changes in the liver, and that histological
changes that were not observed following subchronic exposure durations may be apparent after
chronic exposure. For instance, a study of male and female B6C3F1 mice exposed at doses up to
~28 mg Cr(VI)/kg-day for 12 weeks fNTP. 20071 did notfind evidence of liver histological changes;
34This study was the basis of the previous RfD posted to IRIS in 1998 fU.S. EPA. 1998cl.
This document is a draft for review purposes only and does not constitute Agency policy.
3-165 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
however, after 2 years of exposure, histiocytic infiltration was noted in female mice (but not males)
CNTP. 20081.
The available inhalation studies (medium and low confidence) investigated, but did not
observe, histological alterations in the liver in rats exposed for 12 weeks at concentrations of up to
1.25 mg Cr(VI)/m3 (Kim etal.. 2004) or 0.2 mg Cr(VI)/m3 (Glaser etal.. 1985). or for longer
durations (18 months followed by a 12 month unexposed period) at concentrations of up to 0.1 mg
Cr(VI)/m3 (Glaser etal.. 1986). However, liver chromium concentration following inhalation
exposure to Cr(VI) is expected to be approximately 1-2 orders of magnitude lower than
concentrations following oral exposure due to the first-pass effect fO'Flahertv and Radike. 19911.
As a result, the extent of hepatotoxicity would be expected to differ by route of exposure.
Inflammation-related hepatotoxicity
Inflammation-related histological changes in the liver (increased inflammation and
infiltration of immune cells) were reported in several high confidence studies of Cr(VI) exposure in
F344 rats (NTP. 2008. 2007) and B6C3F1 mice (NTP. 2008). In female F344 rats, statistically
significantly increased incidences of chronic focal inflammation were reported for females in the
highest dose group following 3 months of exposure at 20.9 mg Cr(VI)/kg-day (NTP. 2007) and at
lower doses (0.2-7 mg Cr(VI)/kg-day) after two years of exposure, with incidences increasing
monotonically with dose fNTP. 20081. In male F344 rats exposed for 3 months, no statistically
significant increase in liver lesions was found (NTP. 2007): however, after 2 years of exposure,
chronic inflammation was increased in males in the second highest dose group (56%) relative to
controls, although control incidence was high (38%) and no clear dose-response was apparent for
this endpoint (NTP. 2008). In a 2-year study, a statistically significantly increased incidence of
chronic inflammation was observed in female B6C3F1 mice in the second highest exposure group
(3.2 mg Cr(VI)/kg-day) but not in other exposed groups (high dose: 8.9 mg Cr(VI)/kg-day) or in
male mice at doses up to 5.7 mg Cr(VI)/kg-day fNTP. 20081. Increased Kupffer cell (stellate
macrophage) activation was observed in a high dose, medium confidence study in male SD rats
exposed to approximately35 25 mg Cr(VI)/kg-day for six months (Elshazlv et al.. 2016). In a
continuous breeding study in BALB/c mice, no increased inflammatory changes in the liver were
observed in F0 or F1 male or female mice exposed for approximately 20 weeks at doses up to 30-
50 mg/kg-day fNTP. 19971.
In damaged tissues, infiltrating histiocytes (macrophages) display functions such as
modulation of inflammatory cells, removal of damaged tissues/cellular debris, and antigen
presentation, as well as fibrogenic stimulation (Yam ate etal.. 2016). The incidence of infiltration of
histiocytes in the liver was statistically significantly elevated in female F344 rats exposed for
3 months at doses >3.5 mg Cr(VI)/kg-day (NTP. 2007) and in female F344 rats exposed at lower
35Elshazlv et al. (20161 did not contain enough information to accurately calculate a dose in mg/kg-d. Using
drinking water factors for SD rats from U.S. EPA T19881. the dose may be as high as 25 mg/kg-d (although this
does not take into account decreased palatability of the drinking water at 180 mg/L).
This document is a draft for review purposes only and does not constitute Agency policy.
3-166 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
doses (>0.96 mg Cr(VI)/kg-day) for two years fNTP. 20081. Histiocytic infiltration was not
observed in male F344 rats exposed for 3 months at doses up to 20.9 mg Cr(VI)/kg-day but was
statistically significantly elevated in high dose male rats (5.9 mg Cr(VI)/kg-day) following 2 years of
exposure. Increased incidences of minimal to mild histiocytic infiltration were also observed in all
exposed groups of female mice (0.3 to 8.9 mg Cr(VI)/kg-day), showing an increasing response with
dose, in a 2-year study, but not in male mice (NTP. 20081. Hepatic infiltration of inflammatory cells
was also noted in a medium confidence study which exposed male rats to approximately 25 mg
Cr(VI)/kg-day for six months fElshazlv etal.. 20161. NTP f20081 stated that the significance of
histiocytic infiltration is unknown but hypothesized that infiltration of macrophages may reflect
phagocytosis of an insoluble precipitate. However, specific data investigating chromium removal
from the liver has not been identified. It is important to acknowledge that activated macrophages
can also damage tissue by secreting cytotoxic factors indicative of an innate inflammatory response
and creating an inflammatory environment ffKovama and Brenner. 2017: Yamate etal.. 20161: see
Francke and Mog (20211 for further description) and chronic hepatic inflammation can lead to
fibrosis fKovama and Brenner. 20171. Histiocytic cellular infiltration with exposure to Cr(VI) was
also observed in several other tissues (including the duodenum and mesenteric and pancreatic
lymph nodes) in both rats and mice (NTP. 20081. See the immune effects section (Section 3.2.6) for
further discussion of this effect.
Necrosis and apoptosis
Few chronic or subchronic studies across the evidence base reported liver necrosis or
indications of apoptosis. The incidence of necrosis was not increased in Cr(VI)-exposed animals in
the large (50/sex/group), high confidence, 2-year NTP bioassay in F344 rats or B6C3F1 mice at
doses of up to 6-9 mg Cr(VI) /kg-day (NTP. 20081 or in an NTP continuous breeding study in F0 and
F1 BALB/c mice fNTP. 19971. However, a high dose, medium confidence study observed necrosis in
all SD rats exposed to 25 mg Cr(VI)/kg-day for six months fElshazlv etal.. 20161. Another medium
confidence study observed bile duct necrosis in rabbits exposed by gavage to doses as low as 0.35
mg Cr(VI)/kg-day for three months (Mo etal.. 20181. Several low confidence studies (discussed
below) of shorter duration in Wistar rats reported evidence of necrosis or apoptosis associated
with Cr(VI) exposure. Rafael etal. (20071 described histological changes indicative of apoptosis as
well as necrosis in Wistar rats exposed to approximately 3 mg Cr(VI)/kg-day for 10 weeks. This
study also reported immunohistochemical evidence for increased expression of caspase-3, a marker
for apoptosis, in male rats fRafael etal.. 20071. Mechanistic markers of apoptosis also have been
observed with Cr(VI). A 28-day study in male rats gavaged with 10.6 mg Cr(VI)/kg-d reported
increased expression of genes involved in apoptosis concurrent with increases in liver enzymes
(ALT, AST, and ALP) (Navva etal.. 2017a). Regarding evidence of necrosis, two related publications
qualitatively described periportal necrosis in Wistar rats exposed to 1.4 mg Cr(VI)/kg-day for 22
weeks (Acharva et al.. 2 0 01: Chopra etal.. 1996). While low levels of hepatocellular apoptosis may
be difficult to detect in chronic and subchronic toxicity studies, numerous short-term mechanistic
This document is a draft for review purposes only and does not constitute Agency policy.
3-167 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
studies indicate the upregulation of apoptotic genes as well as the detection of specific markers of
apoptosis (e.g., caspase-3) following Cr(VI) exposure (see Mechanistic Evidence below and Table 3-
26).
Fatty changes and vacuolation
Fatty changes, or steatosis, the accumulation and retention of fat in hepatocytes, is an early
pathological change associated with liver disease. Histologically, fatty change is sometimes noted
as vacuolation, with lipid accumulating in hepatocytes as vacuoles. Fatty changes often coincide
with hepatic inflammation (Kaiser etal.. 2012: Day and Tames. 19981. If the insult responsible for
steatosis persists, more severe pathologies can develop including fibrosis and cirrhosis (Kaiser et
al.. 2012: Day and Tames. 19981. Liver vacuolation associated with oral exposure to Cr(VI) was
reported in several publications fElshazlv etal.. 2016: NTP. 2008: Acharva etal.. 2001: Chopra etal..
1996: NTP. 1996a) but not others (NTP. 2007.1997.1996b). An increased incidence of scattered
hepatocytes with cytoplasmic vacuoles containing lipid, characterized as "fatty changes," was noted
in female (but not male) F344 rats at doses > 0.96 mg Cr(VI)/kg-day in the high confidence 2-year
NTP (2008) study. Furthermore, two similarly designed low confidence studies qualitatively
reported liver vacuolation in Wistar rats exposed to 1.4 mg Cr(VI)/kg-day for 22 weeks (Acharva et
al.. 2001: Chopra et al.. 19961. A high dose, medium confidence study observed vacuolation in all
male SD rats exposed to approximately 25 mg Cr(VI)/kg-day for six months fElshazlv etal.. 20161.
Hepatic vacuolation was also observed in a high confidence study of male and female BALB/c mice
exposed via diet at doses >5.6 mg Cr(VI)/kg-day for 9 weeks fNTP. 1996al. Study authors reported
that the vacuoles were suggestive of lipid accumulation (NTP. 1996a). However, these findings
were not supported by other high confidence studies of this strain of mice treated for 3 months
(NTP. 2007.1997) or a similarly designed 9-week study in rats (NTP. 1996b). No increase in the
3-month study fNTP. 20071 or in F0 male and female BALB/c mice in a continuous breeding study
at doses up to ~30-50 mg Cr(VI)/kg-day for approximately 20 weeks fNTP. 19971.
Other histological effects
Hepatocellular degeneration, altered hepatocellular foci of mixed type, bile duct
hyperplasia, oval cell hyperplasia, and periductal fibroplasia were observed in a medium confidence
study in male SD rats exposed to approximately 25 mg Cr(VI)/kg-day for six months fElshazlv etal..
20161. Necrosis and bile duct toxicity (bile duct hyperplasia and cholangiofibrosis) were also
observed in rabbits exposed by gavage to doses as low as 0.35 mg Cr(VI)/kg-day for three months
(Mo etal.. 2018). Other isolated histological changes were reported in the evidence base, including
the observation of basophilic hepatocellular foci, a preneoplastic lesion. In F344 rats, authors
reported an exposure-related increased incidence of basophilic hepatocellular foci in the 2-year
study in male rats, but not in females (NTP. 2008).
This document is a draft for review purposes only and does not constitute Agency policy.
3-168 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
Summary of histological effects
Overall, there is consistent evidence of Cr(VI)-induced hepatic histological effects, across
species and sexes, in animals exposed via the oral route (see Figure 3-19). Increases in chronic
inflammation and histiocyte infiltration as well as increased fatty change and associated
vacuolation were reported in several high confidence studies following chronic and/or subchronic
oral exposures in rats and mice. Evidence of cell death (necrosis and apoptosis) was reported in
several low confidence studies and is supported by short-term mechanistic studies; however, these
endpoints were unchanged in higher confidence studies testing similar doses, for longer durations.
Histopathological effects were not observed in low and medium confidence studies following
inhalation exposures, potentially due to differences in target tissue dose across routes of exposure.
In general, female rodents appear to be more sensitive to Cr(VI) induced histological
changes (e.g., hepatic inflammation and fatty changes; NTP (2008)). However, few studies are
available in the database that evaluated both males and females; most study designs used either
male or female animals. In the 2 year rat study fNTP. 20081. chronic inflammation and histiocytic
inflammation and were significantly increased in females at lower doses than males (approximately
6-10 fold lower than in male animals).36 Increased fatty changes were also seen in female rats at
doses as low as 0.94 mg/kg-day and were not significantly elevated in males at doses as high as
5.9 mg/kg-d. However, basophilic foci (often considered a preneoplastic effect), was noted in male
rats at doses as low as 0.77 mg/kg-d and was not observed in female rats, although male rats were
observed to have a much higher background rates of this lesion. For mice, which generally
appeared to be less sensitive than rats to hepatic effects with Cr(VI) exposure, statistically
significant increases in chronic inflammation and histiocytic infiltration were seen in female, but
not male mice (NTP. 2008).37
36Inflammation: click to see rat data in females and males in HAWC.
Infiltration: click to see rat data in females and males in HAWC.
37Inflammation: click to see mouse data in females and males in HAWC. Infiltration: click to see mouse data in
females and males in HAWC.
This document is a draft for review purposes only and does not constitute Agency policy.
3-169 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofHexavalent Chromium
End point
Fatty change
Study name
NTP (2008)
Cytoplasmic \&cuolati on NTP (1997)
NTP (1996a)
Elshazlyetal. (2016)
Chronic infl am mali on NTP (2008)
Animal description
Rat, F344/N(Ł)
Rat, F344/N (100% of the control mean were reported in approximately half
of these studies (Younan etal.. 2019: Elshazlv et al.. 2 016: NTP. 2008. 2007: Rafael etal.. 2007:
Acharva etal.. 2001: Chopra etal.. 19961. ALT is found abundantly in the cytosol of the hepatocyte;
in the case of hepatocellular injury, necrosis, or reparative activity, ALT is released into the
This document is a draft for review purposes only and does not constitute Agency policy.
3-170 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
bloodstream fKim etal.. 2008: Boone etal.. 20051. An increase in ALT of >100% (of the control
mean) generally raises concern for hepatic injury fEMEA. 2008: Boone etal.. 20051 and is
considered biologically relevant. Biologically significant increases in ALT (>100%) were observed
across studies in F344 and Wistar rats that were exposed to Cr(VI) for durations ranging from three
months to two years at doses as low as 1-2 mg/kg-day (NTP. 2008. 2007: Acharva et al.. 2001:
Chopra etal.. 19961. ALT was also statistically significantly elevated in some strains of mice
following three months of exposure; however, these increases were smaller in magnitude (<100%
of control) fNTP. 20071. Click here to see the magnitude of ALT changes in HAWC for NTP (2008,
20071.
Statistically significant increases in AST were also observed across rat studies (of various
subchronic durations), with the magnitude of increase ranging from 60-113% above control mean
(Younan et al.. 2019: Navva etal.. 2017a: Krim etal.. 2013: Soudani etal.. 2013: Acharva et al.. 2 0 01:
Chopra etal.. 1996: Meenakshi et al.. 19891. However, many studies in the evidence base did not
measure AST, including the high confidence NTP bioassays. AST is considered a less specific and
sensitive indicator of hepatocellular injury than ALT fEMEA. 2008: Boone etal.. 20051.
Increases in ALP, an indication of hepatobiliary damage fBoone etal.. 20051. were less
consistent across the evidence base, with some studies noting significant increases and other
studies noting decreases in ALP. Several high confidence studies reported small (10-31%) but
statistically significant decreases in ALP in F344 rats (NTP. 2008. 20071 and in one strain of male
mice fNTP. 20071. However, decreases in ALP are not seen as a reflection of hepatobiliary toxicity,
but are thought to be related to decreased food consumption (Travlos etal.. 19961 or conditions
including malnutrition, mineral deficiencies, and anemia fLum. 19951. a finding noted in the NTP
studies (2008, 20071. Four medium or low confidence studies in rats found statistically significant
increases in ALP of 59-165% (Younan etal.. 2019: Navva etal.. 2017a: Elshazlv et al.. 2 016: Krim et
al.. 2013: Chopra et al.. 19961. An increase in ALP was noted in male Wistar rats exposed to 5.3-
10.6 mg Cr(VI)/kg-day for 28-30 days (Navva etal.. 2017a: Krim etal.. 20131 and in female Wistar
rats treated with 1.4 mg Cr(VI)/kg-day for 5.5 months fChopra et al.. 19961. No change relative to
control was seen in male Wistar rats exposed to 1.4 mg Cr(VI)/kg-day for 5.5 months (Acharva et
al.. 20011.
Sorbitol dehydrogenase (SDH), considered to be a supplemental indicator of hepatotoxicity
(Boone etal.. 20051. was evaluated in two NTP studies (NTP. 2008. 20071. NTP reported
statistically significant increases in SDH of 77-458% compared to controls in F344 male and female
rats exposed to >1.7 mg Cr(VI)/kg-day for 3 months (NTP. 20071. Changes in SDH, in male rats
only, were also observed in a 2-year NTP study conducted in the same rat strain that examined
clinical chemistry endpoints at 3, 6, and 12 months (NTP. 20081. This study found more muted
responses than the 3-month study fNTP. 20071. with statistically increased levels of SDH (24-69%)
in the top two dose groups at the 6-month time point, but not at the 3- or 12-month time points
This document is a draft for review purposes only and does not constitute Agency policy.
3-171 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofHexavalent Chromium
fNTP. 20081. In mice, small but statistically significant decreases in SDH were observed in two
strains of mice; however, decreases in SDH are not indicative of liver damage fNTP. 20071.
Hepatic glycogen levels may be affected by exposure to hepatotoxic chemicals. In animals
exposed to Cr(VI), glycogen depletion was noted in two strains of male mice fNTP. 20071 and in two
related studies in male and female Wistar rats f Acharva et al.. 2001: Chopra etal.. 19961. In NTP
(20071. two strains of mice examined histologically showed glycogen depletion at doses >5.2 mg
Cr(VI)/kg-day (B6C3F1) and >2.8 mg Cr(VI)/kg-day (am3-C57BL/6) butno glycogen depletion was
found in exposed BALB/c mice fNTP. 20071. Acharvaetal. f20011 and Chopra etal. T19961 also
noted statistically significant decreased liver glycogen in rats exposed at 1.4 mg Cr(VI)/kg-day (the
only dose tested) for 5.5 months. Hepatic glycogen levels are also dependent on caloric intake. NTP
(20071 noted that the glycogen depletion was likely a result of depressed food consumption, often
observed when water consumption is decreased, as it was at the high dose in this study; however,
food consumption data was not reported.
Overall, significant increases in serum markers of liver damage were reported in several
high and medium confidence oral exposure studies. Generally consistent elevations of ALT and AST
were seen across multiple well-conducted studies in both rats and mice, with the magnitude of
change in ALT considered to be biologically significant and a specific indication of liver damage.
Changes to ALP and SDH were inconsistent across the evidence base and the biological significance
of decreased glycogen observed in several studies is difficult to interpret No effects on serum
markers of liver damage were reported following inhalation exposures.
This document is a draft for review purposes only and does not constitute Agency policy.
3-172 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Endpoint
Alanine Aminotransferase (ALT)
Alkaline Phosphatase (ALP)
Sorbitol dehydrogenase (SDH)
Study Name
Animal Description
Observation Time
Elshazlyetal. (2016)
Rat, Sprague-Dawley((5
6 months
Krim etal. (2013)
Rat, Albino Wistar (c)
30 days
NTP (2007)
Rat, F344/N (o)
90 days
Rat, F344/N (Q)
90 days
Mouse, B6G3F1 ({.)
90 days
Mouse, BALB/c (c)
90 days
Mouse, G57BL/6 (o)
90 days
NTP (2008)
Rat, F344/N (c)
6 months
12 months
90 days
Navya etal. (2017)
Rat, Albino Wistar (o)
28 days
Rafael etal. (2007)
Rat, Wistar (c)
10 weeks
Wang etal. (2015)
Rat, Sprague-Dawley(o
4 weeks
Elshazlyetal. (2016)
Rat, Sprague-Dawley(c
6 months
Krim etal. (2013)
Rat, Albino Wistar (f)
30 days
NTP (2007)
Rat, F344/N (c)
90 days
Rat, F344/N (2)
90 days
Mouse, B6C3F1 (c)
90 days
Mouse, BALB/c (o)
90 days
Mouse, C57BL/6 (c)
90 days
NTP (2008)
Rat, F344/N (c)
6 months
12 months
90 days
Navya etal. (2017)
Rat, Albino Wistar (c)
28 days
Rafael etal. (2007)
Rat, Wistar (o)
10 weeks
Krim etal. (2013)
Rat, Albino Wistar (c)
30 days
Navya etal. (2017)
Rat, Albino Wistar (o)
28 days
Wang etal. (2015)
Rat, Sprague-Dawley(c
4 weeks
NTP (2007)
Rat, F344/N (f)
90 days
Rat, F344/N (Q)
90 days
Mouse, B6G3F1 (c)
90 days
Mouse, BALB/c (c)
90 days
Mouse, C57BL/6 (o)
90 days
NTP (2008)
Rat, F344/N (c)
6 months
12 months
90 days
I no change^^ significant increase^/ Significant decrease;
15 20 25
mg/kg-day
Figure 3-20. Hepatic effects of oral Cr(VI) exposure in animals (clinical
chemistry). Click to see interactive graphic. To view the magnitude of changes in
ALT from NTP f2008. 20071 data, click here. To view data by Elshazlv et al.
(2016) (where dose could not be estimated), click here.
This document is a draft for review purposes only and does not constitute Agency policy.
3-173 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
Liver weight
Several studies reported statistically significant changes (both increases and decreases) in
absolute and relative liver weight (see Figure 3-21) following short-term or subchronic oral
exposures; liver weight was not measured in the 2-year NTP f20081 bioassay. Liver weight relative
to body weight has been shown to be more informative in the evaluation of liver toxicity, as
compared to absolute liver weight, especially when changes in body weight are observed fBailev et
al.. 20041. Therefore, this discussion focuses on changes in relative liver weight where available.
In the only high confidence study in rats, relative liver weights were decreased by about
10% in F344 males exposed to Cr(VI) in drinking water for three months in the two highest dose
groups (11.2 and 20.9 mg Cr(VI)/kg-day) compared with control values; no significant liver weight
changes were found in any female exposed group (NTP. 2007). Relative liver weight was
substantially increased (>twofold) in female Wistar rats exposed to 1.4 mg Cr(VI)/kg-day in
drinking water for 22 weeks in a medium confidence study fChopra etal.. 19961. A low confidence
study found relative liver weight was increased 20-30% in male Wistar rats exposed in the diet to
3-9 mg Cr(VI)/kg-day for 90 days (Younan etal.. 2019). A shorter duration medium confidence
study (4 weeks) in male Sprague-Dawley rats at doses up to 21 mg Cr(VI)/kg-day reported no
change in liver weight (Wang etal.. 2015).
In mice, several high confidence experiments conducted by NTP across three different
strains observed a consistent pattern of absolute liver weight changes in high dose animals (9-30
mg Cr(VI)/kg-day) exposed to Cr(VI) through drinking water for about 3 months. Statistically
significant decreases in absolute liver weights, but not relative liver weight, were observed in
B6C3F1, BALB/c and am3-C57BL/6 mice (NTP. 2008. 2007). However, study authors reported that
decreases in absolute liver weight in these studies were correlated with decreased body weights
seen at higher doses (NTP. 2008. 2007). Several older NTP studies in BALB/c mice did not measure
liver weight (NTP. 1997.1996a).
Regarding inhalation exposure, no changes in relative liver weight were observed in two
90-day rat studies at concentrations of 0.2 mg Cr(VI)/m3 fGlaser etal.. 19851 or 1.25 mg Cr(VI)/m3
fKim etal.. 20041: however, an 18-month study at concentrations of up to 0.1 mg Cr(VI)/m3
observed a statistically significant increase (13.5%) in relative liver weight (Glaser etal.. 1986).
Overall, inconsistent findings were observed for relative liver weight changes in high and
medium confidence oral exposure and low confidence inhalation studies, with decreases in relative
liver weight observed in high confidence studies, and evidence for increased liver weight primarily
limited to the low confidence studies.
This document is a draft for review purposes only and does not constitute Agency policy.
3-174 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Study Name
NTP (1997)
NTP (2007)
Animal Description Observation Time
Relative Liver weight changes
Wang et al. (2015)
Rat, Wistar (j)
22 weeks
F1 Mouse, BALB/c ( )
-117 days
F1 Mouse. BALB/c (j')
-117 days
P0 Mouse, BALB/c (i)
14 weeks
P0 Mouse. BALB/c (/)
14 weeks
Mouse, C57BL/6 ( )
90 days
Mouse. BALB/c ( ")
90 days
Mouse, B6C3F1 ( )
90 days
Mouse, B6C3F1 ( ¦')
90 days
Rat. F344/N (-)
90 days
Rat, F344/N ( )
90 days
Rat. Sprague-Dawley ( ')
4 weeks
25 30
mg/kg-day
Figure 3-21. Hepatic effects of oral Cr(VI) exposure in animals (relative liver
weight). Click to see an interactive graphic.
This document is a draft for review purposes only and does not constitute Agency policy.
3-175 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
3.2.4.3. Mechanistic Evidence
The mechanistic data for liver toxicity indicates that several key events contribute to the
hepatic effects observed in humans and animals. Exposure to Cr(VI) may cause oxidative and
endoplasmic reticulum stress and mitochondrial dysfunction. These events can lead to
inflammation and apoptosis, which can account for histopathological and serum indicators of liver
injury seen in animals. In vivo experiments in rodents report that ingested and (to a lesser extent)
inhaled Cr(VI) can accumulate in the liver (Tin etal.. 2014: NTP. 2008: Cheng etal.. 20001.
demonstrating the metal can reach the target tissue and further supporting the biological
plausibility for Cr(VI)-induced liver toxicity. For chronic oral exposure in the NTP (20081 tissue
distribution study (collection days 182 and 371, with a 2-day washout period), liver chromium
concentrations were significantly elevated at all dose groups compared to controls, indicating
accumulation of chromium in this organ. A pharmacokinetic study by O'Flaherty and Radike (19911
demonstrated that following inhalation or oral exposure to nearly equivalent target absorbed doses
of Cr(VI), oral exposure resulted in liver concentrations that were 1-2 orders of magnitude higher
than those from inhalation exposure (See Appendix C.1.2). As a result, the extent of hepatotoxicity
would be expected to differ by route of exposure.
A large body of mechanistic information (125 studies) exists to inform the potential
hepatotoxicity of Cr(VI) (see Appendix C.2.3). Therefore, studies which are more informative for
chronic human exposure were prioritized for further analysis and interpretation. These included
mammalian studies that focused on exposure routes more relevant to humans (e.g., oral and
inhalation studies), as well as repeat dose studies of longer durations (>28 days). Shorter duration
studies utilizing oral and inhalation routes of administration and in vitro studies in human cell lines
also provided insight into biological plausibility and human relevance of the observed mechanisms.
Oral repeat dose studies provide support for oxidative stress, mitochondrial damage,
inflammation, and apoptosis as mechanisms of Cr(VI)-induced liver effects. A 36-day dietary study
in male mice receiving 1 and 4 mg/kg/K2Cr207-day (0.35 and 1.41 mg/kg-d Cr[VI]) reported
significant increases in hepatic lipid peroxidation and other markers of ROS-related stress flin etal..
20141. similar to a 10-week gavage study in rabbits receiving 5 mg/kg-day (El-Demerdash et al..
20061. Rafael etal. f20071 described immunohistochemical evidence for increased expression of
Caspase-3, a marker for apoptosis in Wistar rats exposed to approximately 3 mg Cr(VI)/kg-day for
10 weeks. A 28-day study in male rats receiving 30 mg/kg/foC^Oy-day (10.6 mg/kg-d Cr[VI]) by
gavage (Navva etal.. 2017a: Navva etal.. 2017b) also reported increases in lipid peroxidation and
decreased SOD, CAT, and GST activity, concurrent with increases in serum indicators of liver
toxicity (ALT, AST, and ALP) and histological changes in the liver (described as feathery
degeneration). These effects were concurrent with the upregulation of some genes involved in
oxidative stress, inflammation, and apoptosis, such as TNF-a, MAPK, Atf-1, GADD-45, Bax, and
Caspase-1, while anti-ap opto tic genes, including Bcl-2 and OGG-1, were downregulated (Navva et
al.. 2017a: Navva etal.. 2017b). Ninety- and 120-day studies in rats exposed to Na2Cr207 (3.97 mg
This document is a draft for review purposes only and does not constitute Agency policy.
3-176 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Cr(VI)/kg-day and 0.99 mg Cr(VI)/kg-day, respectively) reported lipid peroxidation in hepatic
mitochondria and microsomes accompanied by increased urinary excretion of metabolites
indicative of lipid peroxidation such as MDA (Bagchi etal.. 1997: Bagchi etal.. 1995al.
Oral studies in rats and mice of shorter, acute durations provide further support for an MOA
for Cr(VI)-induced liver effects involving oxidative stress and apoptosis. Similar to longer term
repeat dose studies, shorter term and single-dose studies report increased chromium content in the
liver, increased lipid peroxidation and ALT and AST, free radical production, indicators of
inflammation, upregulation of pro-apoptotic genes and proteins, and down-regulation of anti-
apoptotic genes and proteins in liver tissue fZhongetal.. 2017c: Wang etal.. 2010c: Bagchi etal..
2002: Bagchi etal.. 2001: Bagchi etal.. 2000: Bagchi etal.. 1995b: Kumar and Rana. 1982).
In vitro studies in human cell lines provide additional support for the biological plausibility
of these liver toxicity mechanisms in humans. Human liver carcinoma cell lines show increases in
ROS production and MDA at various concentrations as well as effects on antioxidant enzymes and
mitochondrial function fZhong etal.. 2017a: Zeng etal.. 2013: Patlolla etal.. 2009). Similar results
were observed in human fetal hepatocytes including increased mitochondrial stress, ER stress-
related mechanisms, and the activation of apoptotic and senescence signaling cascades f Liang etal..
2019: Xiao etal.. 2019: Zhang etal.. 2019: Liang etal.. 2018a: Liang etal.. 2018b: Yi etal.. 2017:
Zhang etal.. 2017: Zhongetal.. 2017b: Zhong etal.. 2017c: Zhang etal.. 2016: Xiao etal.. 2014: Xie
etal.. 2014: Xiao etal.. 2012a: Xiao etal.. 2012b: Yuan etal.. 2012b: Yuan etal.. 2012a). In vitro
study results also support the upregulation of pro-inflammatory cytokines and signaling molecules
such as NF-ftB, TNF-a, LBT4, and IL1(3 fZhong etal.. 2017c: Yi etal.. 2016).
Collectively, the data indicate oxidative stress, mitochondrial dysfunction, inflammation,
and apoptosis as possible interconnected mechanisms for liver toxicity. The toxicological evidence
in animals taken together with mechanistic evidence, particularly data from oral, in vivo studies
suggest a possible MOA of Cr(VI)-induced liver toxicity involving the production of free radicals and
reactive intermediates through intracellular Cr(VI) reduction. In this possible MOA, the production
of these reactive species alters antioxidant enzyme activity and stresses the endoplasmic reticulum
and mitochondria, triggering an apoptotic signaling cascade. Oxidative stress may lead to liver
inflammation and the upregulation of genes involved in an inflammatory response.
3.2.4.4. Integration of Evidence
Overall, the available evidence indicates that Cr(VI) likely causes hepatic effects in
humans. This conclusion is based on studies in animals that observed hepatic effects following
exposure to Cr(VI) in drinking water. The human evidence for Cr(VI)-induced liver effects is
limited in terms of number and confidence of studies. However, two of the available three studies
(one occupational and one general population study) provide an indication of exposure-related
alterations of liver clinical chemistry (Sazakli etal.. 2014: Saraswathv and Usharani. 2007). Given
the plausible support for these findings from in vitro studies of human hepatic cells, the human
evidence is interpreted to provide slight evidence of hepatic toxicity associated with Cr(VI).
This document is a draft for review purposes only and does not constitute Agency policy.
3-177 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Integrated evidence of the hepatic effects of Cr(VI) exposure from human, animal, and mechanistic
studies is summarized in an evidence profile table, Table 3-26. The exposure conditions relevant to
hepatic effects are further defined in Section 4.1.
The available animal studies provide moderate evidence for liver effects in rats and mice
orally exposed to Cr(VI) compounds, based primarily on elevated serum enzymes suggestive of
liver toxicity, as well as histological evidence of inflammatory effects and fatty changes in the liver
that are supported by a large and coherent database of in vivo mechanistic studies. This conclusion
is specific to oral exposure to Cr(VI) as few, lower confidence inhalation studies evaluated liver
toxicity and were generally null, possibly owing to the known differences in pharmacokinetics
across routes.
Elevations of ALT and AST were seen across the oral evidence base, with biologically
significant elevations in ALT (>100%) seen in multiple studies. ALT in particular is considered a
sensitive and specific indicator of liver injury fKim etal.. 2008: Boone etal.. 20051. Increased ALT
is roughly correlated with the degree of hepatic inflammation, with patients with high ALT levels
tending to have more severe inflammation in the liver than those with normal ALT values fKim et
al.. 20081.
Chronic inflammation in the liver is a concern as it can lead to liver fibrosis (Kovama and
Brenner. 20171. Dose-dependent increases in chronic inflammation were most evident in female
F344 rats exposed for three months to two years (NTP. 20081. Lesser increases in chronic
inflammation were also seen in male F344 rats and female (but not male) B6C3F1 mice exposed for
two years, although background incidence of this lesion was high (NTP. 2008. 20071.
Fatty change (steatosis) is a common pathological change associated with liver disease,
often leading to, or coinciding with, inflammation. If the insult responsible for steatosis persists,
more severe pathologies can develop, including fibrosis and cirrhosis (Kaiser etal.. 2012: Day and
Tames. 19981. Histological findings of vacuolation and fatty changes were also observed in several
studies (NTP. 2008: Acharva etal.. 2001: Chopra etal.. 1996: NTP. 1996a). Fatty changes are
thought to be mediated by impaired mitochondrial function, which was observed in several studies
of Cr(VI) exposure to human hepatic cells in vitro (Yi etal.. 2017: Zhong etal.. 2017c: Zhongetal..
2017a: Zhang etal.. 2016: Xiao etal.. 2014: Xie etal.. 2014: Zeng etal.. 2013: Xiao etal.. 2012a: Yuan
etal.. 2012a: Patlollaetal.. 20091.
Severe histological changes such as necrosis and fibrosis were not observed in the
high-confidence NTP three-month or two-year studies in F344 rats and B6C3F1 mice (NTP. 2008.
2007). However, several lower confidence subchronic studies in rats noted increased evidence of
apoptosis or necrosis fElshazlv etal.. 2016: Rafael etal.. 2007: Acharva et al.. 2 0 01: Chopra etal..
1996). These effects are supported by mechanistic evidence that suggests a possible MOA of
Cr(VI)-induced liver toxicity involving the production of free radicals and reactive intermediates
through intracellular Cr(VI) reduction resulting in oxidative stress, mitochondrial dysfunction,
inflammation, and apoptosis. Taken together, the serum enzyme and histopathology data from
This document is a draft for review purposes only and does not constitute Agency policy.
3-178 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 human, animal, and in vitro studies support biologically significant changes in the livers of rodents
2 orally exposed to Cr(VI).
This document is a draft for review purposes only and does not constitute Agency policy.
3-179 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-26. Evidence profile table for hepatic effects
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Evidence from studies of exposed humans
CLINICAL CHEMISTRY
Low confidence:
Sazakli et al. (2014)
Saraswathv and
Usharani (2007)
Lin et al. (1994)
Statistically significant changes in
at least one marker of liver
dysfunction (ALT, AST, ALP,
bilirubin or total protein) were
reported in 2 out of 3 low
confidence studies, though the
direction of the associations was
not coherent for all endpoints
across studies (i.e., increases in
ALT, AST, ALP, and bilirubin
would be expected to
accompany decreases in total
protein, but this was not
consistently the case).
Exposure-
response
gradient
between
exposure groups
in one study for
ALT, AST, ALP,
and TP
Lack of
expected
coherence
Low
confidence
studies
©oo
Slight
Based on changes in
clinical chemistry
markers of liver
dysfunction in two low
confidence studies.
Evidence from animal studies
HISTOPATHOLOGY (Oral)
High confidence:
NTP (1996a)
NTP (1997)
NTP (2007)
NTP(2008)
Medium confidence:
Wang et al. (2015)
Elshazlv et al. (2016)
Low confidence:
Acharva et al. (2001)
Chopra et al. (1996)
Increased chronic inflammation,
histiocyte infiltration, fatty
change and vacuolation with
subchronic and chronic
exposures in male and female
rats and mice.
No increase in necrosis in high
confidence studies; however,
lower confidence studies and
numerous mechanistic studies
have indicated an increase in
necrosis and markers of
apoptosis.
Mostly high and
• No increase in
medium
necrosis in
confidence
high
studies
confidence
studies
Generally
consistent
findings
regarding
inflammatory
changes and
fatty
changes/vacuola
®©o
Moderate
Findings of
histopathological
changes (particularly
inflammation-related
effects and fatty
changes/vacuolation)
coupled with significant
increases in ALT and AST
are considered to be
0©O
The evidence indicates
that Cr(VI) is likely to
cause liver toxicity in
humans given sufficient
exposure conditions.
Effects on clinical
chemistry were observed
in both human and
animal studies.
Moderate evidence in
rats and mice shows
consistent findings of
elevated liver enzymes
indicative of
hepatocellular damage
and changes in liver
architecture following
oral exposure.
Mechanistic findings in
animals provide evidence
supportive of
histopathological
endpoints in the liver.
Oxidative stress was
identified as a potential
mechanism for liver
effects in multiple animal
species. This mechanism
This document is a draft for review purposes only and does not constitute Agency policy.
3-180 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
is presumed relevant to
humans.
Hepatic effects were
inconsistent following
inhalation. Because of
the first-pass effect, the
liver may be affected
more severely by Cr(VI)
exposure via the oral
route as compared to the
inhalation route.
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Rafael et al. (2007)
tion across most
species and
sexes
• Coherence with
increases in ALT
and AST
• Mechanistic
evidence
provides
biological
plausibility
adverse and a specific
indication of liver injury.
Hepatic effects were
generally not observed
following inhalation
exposures.
HISTOPATHOLOGY
(Inhalation)
Medium confidence:
Kim et al. (2004)
Low confidence:
Glaser et al. (1985)
No histological changes in rats
treated for 12 weeks or 18
months.
CLINICAL CHEMISTRY
(Oral)
High confidence:
Krim et al. (2013)
NTP (2007)
NTP (2008)
Medium confidence:
Navva et al. (2017a)
Rafael et al. (2007)
Wang et al. (2015)
Elshazlv et al. (2016)
Low confidence:
Statistically significant elevations
of ALT and AST seen across
studies.
Biologically significant increases
in ALT (>100%) were observed
across studies and at doses as
low as 1-2 mg/kg-day.
Changes to ALP were less
consistent across the evidence
base.
• Consistent
increases in ALT
and AST
• High and
medium
confidence
studies
• Magnitude of
effect: large
effect size for
ALT and AST
• No factors
noted
This document is a draft for review purposes only and does not constitute Agency policy.
3-181 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Acharva et al. (2001)
Choora et al. (1996)
Meenakshi et al. (1989)
• Dose-response
gradient within
studies
• Coherence with
histopathology
(inflammation
and fatty
changes)
• Mechanistic
evidence of
oxidative stress
provides
biological
plausibility
CLINICAL CHEMISTRY
(Inhalation)
Medium confidence:
Kim et al. (2004)
Low confidence:
Glaser et al. (1985)
Glaser et al. (1986)
No significant changes in
enzymatic markers of liver
damage (ALT, AST, ALP, SDH)
following inhalation.
This document is a draft for review purposes only and does not constitute Agency policy.
3-182 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
ORGAN WEIGHT (Oral)
High confidence:
NTP (2007)
NTP (1997)
Medium confidence:
Chopra et al. (1996)
Wang et al. (2015)
Low confidence:
(Younan et al., 2019)
Inconsistent findings for relative
liver weight changes in high and
medium confidence oral studies,
with no change or decreased
relative liver weight observed in
high and medium confidence
studies and evidence for
increased relative liver weight
primarily limited to low
confidence studies.
Decreases in absolute liver
weight in mice likely correlated
with body weight decreases seen
at high doses.
• No factors noted
• Unexplained
inconsistency
across studies
ORGAN WEIGHT
(Inhalation)
Medium confidence:
Kim et al. (2004)
Low confidence:
Glaser et al. (1985)
Glaser et al. (1986)
Changes in liver weight were
inconsistent following inhalation
exposures. One 18 month study
observed a statistically and
biologically significant (>10%)
increase in relative liver weight
(Glaser et al., 1986).
Mechanistic evidence
Biological events or
pathways
Summary of key findings and interpretations
Judgments and
rationale
Oxidative and
endoplasmic reticulum
stress
Interpretation: Consistent in vivo and in vitro evidence of Cr(VI)-induced
oxidative and ER stress evidenced by increased lipid peroxidation, ROS, and
decreased antioxidant enzyme activity concurrent with biomarkers of liver
injury.
Key findings:
Biologically plausible,
consistent, coherent
observations of
oxidative stress and
endoplasmic reticulum
This document is a draft for review purposes only and does not constitute Agency policy.
3-183 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Factors that increase Factors that
Summary of key findings certainty decrease certainty
Judgments and
rationale
• Consistent evidence of significant increases in lipid peroxidation in liver
tissue in chronic, subchronic and acute dose animal studies (Navva et al.,
2017a; Zhong et al., 2017c; Jin et al., 2014; Wang et al., 2010c; Bagchi et
al., 2002; Bagchi et al., 2001; Bagchi et al., 2000; Bagchi et al., 1997;
Bagchi et al., 1995b; Bagchi et al., 1995a; Kumar and Rana, 1982)
• Increased oxidative stress (decreased antioxidant enzyme activity)
concurrent with serum biomarkers of liver injury (increased ALT, AST,
and ALP) in a 28-dav studv in rats (Navva et al., 2017a)
• Increased oxidative stress (lipid peroxidation, free radical production)
concurrent with serum biomarkers of liver injury (increased ALT and
AST) in liver tissue in short-term and acute oral exposure studies in rats
and mice (Zhong et al.. 2017c; Wang et al.. 2010c; Bagchi et al.. 2002;
Bagchi et al., 2001; Bagchi et al., 2000; Bagchi et al., 1995b; Kumar and
Rana, 1982)
• In vitro evidence of increased ROS production and MDA and effects on
antioxidant enzvmes in human liver carcinoma cell lines (Zhong et al..
2017a; Zeng et al., 2013; Patlolla et al., 2009)
• In vitro evidence of ER stress-related mechanisms in human cells (Zhang
et al., 2017)
stress, mitochondrial
dysfunction,
inflammation, and
apoptosis concurrent
with apical observations
of liver toxicity following
(oral) exposures to
Cr(VI) in animals,
supported by in vitro
evidence in human cells.
Mitochondrial
dysfunction
Interpretation: In vitro evidence in human liver cell lines of Cr(VI)-induced
mitochondrial dysfunction.
Key findings:
• In vitro evidence of effects on mitochondrial function in human liver
carcinoma cell lines (Zhong et al., 2017a; Zeng et al., 2013; Patlolla et al.,
2009)
• In vitro evidence of increased mitochondrial stress in human fetal
hepatocvtes (Yi et al.. 2017; Zhong et al.. 2017c; Zhang et al.. 2016; Xiao
et al., 2014; Xie et al., 2014; Xiao et al., 2012a; Yuan et al., 2012a)
This document is a draft for review purposes only and does not constitute Agency policy.
3-184 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Inflammation
Apo ptosis
Interpretation: Consistent in vivo and in vitro evidence of Cr(VI)-induced liver
inflammation.
Key findings:
• Increased indicators of inflammation concurrent with serum biomarkers
of liver injury (increased ALT and AST) in liver tissue in short-term and
acute oral exposure studies in rats and mice (Zhong et al., 2017c; Wang
et al., 2010c; Bagchi et al., 2002; Bagchi et al., 2001; Bagchi et al., 2000;
Bagchi et al., 1995b; Kumar and Rana, 1982)
• In vitro evidence of the upregulation of pro-inflammatory cytokines and
signaling molecules such as NF-kB, TNF-a, LBT4, and IL1|5 in human cells
(Zhong et al., 2017c; Yi etal.,2016)
Interpretation: Cr(VI) alters protein and gene expression of biomarkers
associated with apoptosis in vivo concurrent with liver injury.
Key findings:
• Increased expression of caspase-3 and histological changes indicative of
apoptosis in a 10-week rat study (Rafael et al., 2007)
• Upregulated transcription of pro-apoptotic genes and downregulated
transcription of anti-apoptotic genes concurrent with serum biomarkers
of liver injury (increased ALT, AST, and ALP) in a 28-d rat study (Navva et
al.. 2017a)
• Upregulation of pro-apoptotic genes and proteins and downregulation
of anti-apoptotic genes and proteins concurrent with serum biomarkers
of liver injury (increased ALT and AST) in liver tissue in short-term and
acute oral exposure studies in rats and mice (Zhong et al.. 2017c; Wang
et al., 2010c; Bagchi et al., 2002; Bagchi et al., 2001; Bagchi et al., 2000;
Bagchi et al., 1995b; Kumar and Rana, 1982)
• In vitro evidence of the activation of apoptotic signaling cascades in
human fetal hepatocytes (Yi et al., 2017; Zhong et al., 2017c; Zhang et
This document is a draft for review purposes only and does not constitute Agency policy.
3-185 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
al., 2016; Xiao et al., 2014; Xie et al., 2014; Xiao et al., 2012a; Yuan et al.,
2012a)
This document is a draft for review purposes only and does not constitute Agency policy.
3-186 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Toxicological Review ofHexavalent Chromium
3.2.5. Hematologic effects
Hematology is a subgroup of clinical pathology concerned with morphology, physiology,
and pathology of blood and blood-forming tissues. Hematology parameters, as part of a routinely
measured complete blood count (CBC), are described in Table 3-27. A CBC is a common blood test
providing quantitative and qualitative information regarding the general health of a patient or
research subject. Examples of quantitative information include total counts of red blood cells
(RBCs), white blood cells (WBCs) and platelets; qualitative information, such as the RBC indices,
give a morphological estimation of the RBC size and color. RBCs carry oxygen throughout the body,
while WBCs are involved in immune function (discussed in Section 3.2.6) and platelets are involved
in blood clotting. RBCs also carry most of the body's iron, which can be indirectly measured in
blood by measuring transferrin, a membrane-bound transporter of ferric (Fe+3) iron, and total iron
binding in blood. Hematology along with clinical pathology measures (e.g., blood proteins,
enzymes, chemicals and waste products) and other general health status indicators are useful for
assessing overall health status, monitoring disease, and determining if follow-up testing is needed.
RBCs act as a sink for chromium in the blood. Cr(VI) is rapidly taken up by RBCs, where it is
reduced to Cr(III) and remains trapped for the lifetime of the cell (see Section 3.1 and Appendix C
for more details). After RBCs are broken down, the Cr(III) is released to systemic circulation and
may be absorbed by other tissues or excreted in urine. Because Cr(III) cannot readily cross cell
membranes, the RBC chromium level is commonly used as a biomarker for Cr(VI) exposure in
industrial settings fMiksche and Lewaiter. 19971. The focus of this section is primarily on RBCs and
related components. Cr(VI) effects on white blood cell parameters are discussed in the context of
the immune system in Section 3.2.6.
Table 3-27. Hematologic endpoints commonly evaluated in routine blood
testing
Endpoint
Description
Hemoglobin (Hgb, g/dL)
Iron-containing oxygen-transport metalloprotein in RBCs
Hematocrit (Hct)
Percentage (by volume) of the blood that consists of RBCs
Hematocrit (%) = MCV x RBC / 10
Red blood cell (RBC;
erythrocyte) count
The most common blood cell responsible for systemic oxygen delivery. Expressed
as number of RBCs per nL of blood
Reticulocytes
Immature non-nucleated RBCs containing residual RNA; indicates rate of new RBC
production. The normal range depends on your level of hemoglobin. Hemoglobin is
a protein in red blood cells that carries oxygen. The range is higher if hemoglobin is
low, from bleeding or if red cells are destroyed.
Mean cell volume (MCV)
Average volume of the RBC
MCV = hematocrit x 10 / RBC
Low MCV: microcytic (smaller RBCs, possibly caused by iron deficiency and
anemia); high MCV: macrocytic (larger RBCs, possibly caused by excess iron).
Mean cell hemoglobin (MCH)
Average weight of hemoglobin (Hgb) in the RBC
This document is a draft for review purposes only and does not constitute Agency policy.
3-187 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Toxicological Review ofHexavalent Chromium
Endpoint
Description
MCH = Hgb x 10 / RBC, (g/dL)
Hemoglobin concentration normalized as amount of hemoglobin per cell.
High MCH: may indicate macrocytic anemia (large red blood cell volume leading to
low Hgb concentration), while low MCH may indicate other types of anemia
(e.g., from iron deficiency).
Mean cell hemoglobin
concentration (MCHC)
Average concentration of Hgb in the RBC volume
MCHC = Hgb x 100 / hematocrit (g/dL)
Hemoglobin concentration normalized to red blood cell volume.
Low MCHC: hypochromic (RBCs paler than normal); high MCHC: hyperchromic
(RBCs more pigmented than normal)
3.2.5.1. Human Evidence
Study evaluation summary
There are five studies that reported on the association between Cr(VI) exposure and
hematologic parameters pertaining to the erythron (circulating RBC mass); specifically, complete
blood counts (CBC), including RBC, hemoglobin (Hgb), and hematocrit (Hct). Four studies were
classified as low confidence (Table 3-28). Sazakli etal. (2014) was limited due to exposure
measurement; exposure was estimated using water intake and historic water concentration
records as well as hair and blood concentrations. Correlations between these measures were low.
It is likely that any exposure misclassification would be nondifferential and therefore lower the
precision of the effect estimates but is less likely to bias the results away from the null. Sharma et
al. (2012) was limited in most domains, and exposure classification was based on residence in a
geographic area with contaminated groundwater, which does not distinguish the heterogeneity of
exposure across exposed participants. Lacerdaetal. (2019) was limited due to potential for
selection bias and confounding and Song etal. f20121 was limited due to potential for confounding.
The remaining study fKhan etal.. 20131 was classified as uninformative because exposure
classification was based on tannery work and there was insufficient information provided on the
specific tanning processes used at the facility to infer Cr(VI) exposure38.
38Leather tanning processes that can potentially lead to Cr(VI) exposure include: 1) use of a two-bath process,
2) on-site production of tanning liquors, and 3) leather finishing steps that involve Cr(VI) (e.g., use of Cr(VI)-
containing pigments) (Shaw Environmental. 20061. If these processes are not specified by the study, it cannot
be determined whether exposure was to Cr(VI) or Cr(III).
This document is a draft for review purposes only and does not constitute Agency policy.
3-188 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofHexavalent Chromium
Table 3-28. Summary of human studies for Cr(VI) hematologic effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.
Click to see interactive data graphic for rating rationales-
Author (year)
Industry
Location
Exposure
Measurement
Study Design
Clinical Pathology
Lacerda et al. (2019)
Chrome-
plating
workers
Brazil
Exposure group
validated by urine,
blood sampling
Cross-sectional
L
Sazakli et al. (2014)
General
population
Greece
Urine, Hair, Modeled
lifetime chromium
exposure dose
Cross-sectional
L
Sharma et al. (2012)
General
population
India
Residence in
geographic area with
contaminated
groundwater vs.
control
Cross-sectional
L
Song et al. (2012)
Chromate
production
workers
China
Work category
validated by air,
blood sampling
Cross-sectional
L
Khan et al. (2013)
Tannery
Pakistan
Blood, Urine, Work
category
Cross-sectional
U
Synthesis of evidence in humans
One of the included low confidence studies (Sazakli etal.. 20141 reported statistically
significant decreases in Hgb and Hct (Table 3-29), inconsistent with another low confidence study
that reported statistically significant increases in the same endpoints fLacerda etal.. 20191. Song et
al. (20121 reported no association with hemoglobin but did not report on hematocrit Another low
confidence study reported higher RBC counts and lower mean cell volume (MCV) in exposed
participants, stratified by sex (all statistically significant except MCV in women) (Sharma etal..
20121. None of the other studies reported an association between Cr(VI) exposure and RBC count,
and none examined associations with diagnosed anemia, other hematological disease, or
hematologic parameters dichotomized based on clinical adversity. Platelet findings were also
inconsistent Sharma etal. f20121 reported lower platelets in exposed participants, while Sazakli et
al. (20141 reported higher platelets with higher exposure, both statistically significant.
This document is a draft for review purposes only and does not constitute Agency policy.
3-189 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-29. Associations between Cr(VI) and hematologic parameters in
epidemiology studies
Reference,
confidence
Population
Exposure
comparison and
effect estimate
RBC
(1012/L)
Hgb
(g/dL)
Hct
(%)
Lacerda et al.
Cross-sectional in
Brazil, chrome-
plating workers
(n = 50) and controls
(n = 50)
Means ± SD for
chromium
unexposed/
exposed
Unexposed:
5.34 ±0.79
Exposed:
5.95 ±0.90
Unexposed:
14.16 ±0.40
Exposed:
15.70 ±0.14*
Unexposed:
39.18 ±0.49
Exposed:
43.30 ±0.36*
(2019), low
Sazakli et al.
Cross-sectional in
Greece, general
population; Two
exposure groups
(n = 237) and
controls (n = 67)
Regression
coefficients for
calculated lifetime
exposure dose
and Cr in hair
Lifetime:
0.007
Hair:
-0.09
Lifetime:
-0.09 *
Hair:
-0.06
Lifetime:
-0.09 *
Hair:
-0.1 *
(2014), low
Sharma et al.
Cross-sectional in
India, general
population with
residence in
contaminated area
(n = 186) or not
(n = 230)
Means ± SD for
chromium
unexposed/
exposed
Males
Unexposed:
4.28 ±0.69
Exposed: 5.55 ±1.39*
Females
Unexposed:
3.89 ±0.71
Exposed: 5.67 ±1.26*
NR
NR
(2012), low
Song et al. (2012),
Cross-sectional in
China, chromate
production workers
(n = 100) and
controls (n = 80)
Means ± SD for
chromium
unexposed/
exposed
Unexposed:
4.7 ±0.4
Exposed:
4.8 ±0.8
Unexposed:
144.8 ± 12.6
Exposed:
148.8 ±27.2
NR
low
*p < 0.05. Shading indicates results supportive of an association between Cr(VI) and hematologic parameters in
the direction of anemia (i.e., decrease in red blood cells, hemoglobin, and hematocrit).
NR: not reported.
1 Due to inconsistent results across low confidence studies, there is no clear evidence of an
2 association between Cr(VI) exposure and hematologic effects in humans. Conflicting results may
3 stem from differences in exposure scenarios, exposure assessment methods, and study sensitivity.
4 Because this is a very limited evidence base in terms of number and confidence of studies, further
5 exploration of patterns by exposure levels or type of analysis is not possible.
6 3.2.5.2. Animal Evidence
7 Study evaluation summary
8 Table 3-30 provides a summary of the animal toxicology studies considered in the
9 evaluation of the hematologic effects of Cr(VI). The available evidence included 14 studies
10 conducted in rats (three strains) and mice (three strains). Exposure durations and routes included
This document is a draft for review purposes only and does not constitute Agency policy.
3-190 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
one chronic oral study fNTP. 20081. one subchronic oral study fNTP. 20071. seven oral 3-9 week
studies fWang etal.. 2015: Krim etal.. 2013: NTP. 2006a. b, 2005.1996a. b), one study conducted
using NTP's Reproductive Assessment by Continuous Breeding (RACES) protocol fNTP. 19971. and
four inhalation studies ranging from short-term to chronic exposure durations (Kim etal.. 2004:
Glaseretal.. 1990: Glaser etal.. 1986: Glaser etal.. 19851.
Of the 15 included studies, 10 were considered medium or high confidence studies, and
included eight National Toxicology Program (NTP) studies with exposure durations ranging from 4
days to 12 months (Table 3-30). Three of the four inhalation studies and one of the 11 oral studies
that examined hematologic endpoints were considered low confidence mostly because of limited
reporting of study methods and/or results. Six additional studies with hematologic data were
judged uninformative based on critical deficiencies identified when the studies were evaluated
(i.e., Anwar etal. (19611 mixed animals of different breeds; Kumar and Barthwal (19911 did not use
concurrent controls; Shrivastava et al. f2005al lacked information on sex, number of mice, and
control group; and Zabulvte etal. (20091 and Zabulvte etal. (20061 had multiple deficiencies
including randomization procedures, lack of vehicle control, and others). MacKenzie etal. f!958139
was rated uninformative due to insufficient reporting of the outcomes, mortality due to a
respiratory infection, and sample sizes of evaluated animals. Full study evaluation details are
available in HAWC.
Table 3-30. Summary of included studies for Cr(VI) hematologic effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.3
Click to see interactive data graphic for rating rationales.
Author (year)
Species (strain)
Exposure design
Exposure route
Hematologic
outcomes'3
Glaser et al. (1985)
Rat (Wistar), male
28 and 90 d
Inhalation
L
Glaser et al. (1986)
Rat (Wistar), male
18 months
Inhalation
L
Glaser et al. (1990)
Rat (Wistar), male
30 and 90 d
Inhalation
L
Kim et al. (2004)
Rat (Sprague-Dawley), male
90 d
Inhalation
M
Krim et al. (2013)
Rat (Wistar), male
30 d
Oral (Gavage)
M
NTP (1996a)
Mouse (BALB/c)
3, 6, and 9 wk
Oral (Diet)
H
NTP (1996b)
Rat (Sprague-Dawley)
3, 6, and 9 wk
Oral (Diet)
H
NTP (1997)
Mouse (BALB/c)
Continuous
breeding design
Oral (Diet)
H
39Normally in situations concerining poor reporting, authors may be contacted for clarifications that may
result in upgraded confidence ratings, but this was not possible due to the age of the publication. This study
was the basis of the previous RfD posted to IRIS in 1998 (U.S. EPA. 1998cl.
This document is a draft for review purposes only and does not constitute Agency policy.
3-191 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Author (year)
Species (strain)
Exposure design
Exposure route
Hematologic
outcomes'3
NTP(2005)
Mouse (B6C3F1), female
28 d
Oral (Drinking
water)
H
NTP(2007)
Rat (F344/N)
Mouse (B6C3F1)
Mouse (B6C3F1, BALB/c, am3-
C57BL/6), male—comparative
toxicity study
90 d
Oral (Drinking
water)
H
NTP(2008)
Rat (F344/N), male
Mouse (B6C3F1), female
2 yr
Oral (Drinking
water)
H
NTP(2006b)
Rat (Sprague-Dawley), female
28 d
Oral (Drinking
water)
M
NTP(2006a)
Rat (F344), female
28 d
Oral (Drinking
water)
M
Samuel et al. (2012a)
Rat (Wistar), female
Pregnant dams, GD
9-21
Oral (Drinking
water)
L
Wang et al. (2015)
Rat (Sprague-Dawley), male
28 d
Oral (Drinking
water)
M
aStudies in this table were ordered first by route of exposure, and then by confidence rating. Within a confidence
rating, studies were ordered chronologically.
bWithin each study, multiple hematologic outcomes (such as those in Table 3-27) were typically measured using
analytical methods for complete blood counts. For this reason, multiple outcome sensitivity ratings are not
presented.
1 Synthesis of evidence in animals
2 Evidence informing Cr(VI) effects on hematologic endpoints was available from several
3 (mostly short-term) medium and high confidence oral exposure studies (Table 3-30). There were
4 two high confidence studies, one subchronic fNTP. 20071 and one chronic fNTP. 20081 bioassay,
5 reporting hematologic outcomes in F344 rats and B6C3F1 mice that were useful for evaluating the
6 potential subchronic and lifetime hematologic effects of Cr(VI) exposure in humans. Both studies
7 are discussed below in detail and are summarized in HAWC and in Figure 3-22 below (note that
8 only observation times at 90 days and greater are presented).40 Methodological considerations for
9 evaluating hematology findings in general included alterations in water intake, fasted/fed status,
10 life stage, and sex.
40Older data from other medium and high confidence studies performed by the National Toxicology Program
fNTP. 2006a. b, 2005.1997.1996a. b) are consistent with results by NTP (2008. 20071. Only the most recent
NTP results are synthesized, because they provide data at a wide dose range for multiple species and strains,
and also provide data from multiple timepoints within its 2-year study.
This document is a draft for review purposes oniy and does not constitute Agency poiicy.
3-192 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
Endpoint Name
Study Name
Animal Description
Observation time
Cr(VJ) Animal Toxicology Hematology Effects NTP
Red blood cell counl
NTP 2007
Rat, F344/N (9)
90 days
~ AA A
A
A
Rat, F344/N (cf)
90 days
~ AA A
A
V
NTP 2008
Mouse, B6C3F1 (?)
90 days
AA A
A
Rat, F344/N (cf)
90 days
• »—A
A
Mouse, B6C3F1 (9)
12 months
Rat, F344/N (cf)
12 months
•» •—•—
A
Hemoglobin (Hgb)
NTP 2007
Mouse. B6C3FI (9)
90 days
m m m 1
—%
V
Mouse, B6C3F1 >se Range 1
0 0 50 100
150 200 250 300
350 4
50
Dose (ppitt)
Figure 3-22. Hematology findings from NTP f2007) and NTP f20081 in rats and
mice exposed by gavage to Cr(Vl) for 90 days or 12 months (full details
available in HAWC).
This document is a draft for review purposes only and does not constitute Agency policy,
3-193' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofHexavalent Chromium
Direct measures of hematopoietic health include RBCs, Hgb, and Hct levels (see Table 3-27).
RBCs were increased across study designs, sexes, and species in both a high confidence subchronic
study (NTP. 20071 and a high confidence chronic bioassay (NTP. 20081 (Figure 3-22). Statistically
significant treatment effects corresponded with an approximately 2-4% change in the 20 mg/L
dose group, 4-8% change at 60 mg/L, and 5-18% change from controls in the 180 mg/L dose group
(NTP. 20081 (click to viewRBC findings)41. Note that RBC counts were greater at 90 days than 12
months within each dose group and sex. Hgb was decreased in both male and female rats and
female mice at 9 and 12 month observation times at doses >174.5 mg/L. The magnitude of change
was <5% from control mean for all findings except in the >174.5 mg/L dose groups. Hct increased
in female mice at 90 days and decreased in male and female rats at doses >174.5 mg/L. No changes
in Hct were observed in either species at 12 months Cr(VI) exposure.
The RBC, Hgb, and Hct findings at 90 days were considered to be potentially adverse based
on data from high confidence studies showing a large magnitude of change, increasing responses
with dose, and consistency across species and sexes, supported by coherent changes in other RBC
indicators (MCV, MCH, and MCHC). The adversity of effects at 12 months, however, were less
certain and potentially adaptive. Decreased mean cell volume (MCV) values (i.e., smaller RBCs)
were consistently observed across study designs, sexes, and species (although male rats were the
most sensitive) in both high confidence NTP bioassays (NTP. 2008. 2007). but while MCV decreases
were dose-responsive across rat 90 day observation times, with a maximal response of a ~30%
change from control in male rats receiving 349 mg/L for 90 days, when comparing the MCV
response to Cr(VI) exposure from 90 days to 12 months, the 12 month response was less robust
(23% decrease compared with 7% at 12 months). Cr(VI) effects on MCH were consistent and
coherent with MCV; decreases were dose-responsive across 90 day and 12 month observation
times, with a maximal response of ~30% at 349 mg/L (90 days), but similar to MCV, the response
was less intense at 12 months (~8% decrease from control) compared with same dose at 90 day
observation time (~27% change) in rats. The MCHC response to Cr(VI) exposure in rats and mice
was muted compared with MCV and MCH, with a maximum response of 5-10% change from control
in male and female rats exposed for 90 days to >174.5 mg/L. The dose-response was less clear at
12 months exposure. The pattern of response, however, was similar to MCH and MCV when
comparing the MCHC response between exposure durations and species, with a greater response at
90 days compared to 12 months, and in rats compared with mice.
41 Exposures for NTP f20081 and NTP f~20071 are expressed in the text as concentration in drinking water (mg
Cr(VI)/L) rather than daily dose (in mg Cr(VI)/kg-day). Differences in rodent drinking water consumption
rates relative to body weight during the growth period lead to different mg/kg-d doses at the different
collection times within the same exposure group of the 2-year study. Discussion in units of drinking water
concentrations simplifies the group-level comparisons. Estimates of time weighted average daily doses at
different observation time are available here (for NTP (200811 and here (for NTP (2007H. At 20 mg/L Cr(VI)
in rats for the 2-year study, the time weighted average dose was 2 mg/kg-d at 22 days, 1.5 mg/kg-d at 90
days, and 0.88 mg/kg-d at 1 year.
This document is a draft for review purposes only and does not constitute Agency policy.
3-194 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Reticulocytes (RET) and nucleated reticulocytes are immature RBCs and their levels may
indicate alterations in RBC production fWhalan. 2015. 20001. Reticulocytes and nucleated
erythrocytes were increased, but the finding was inconsistent across species and sexes, with no
changes observed in mice, and with increases observed in male rats only at the maximum dose
(NTP. 2008. 20071 and in all female rat dose groups at 90 days (NTP. 20071. Microscopic evaluation
of blood smears at exposure durations up to 90 days identified erythrocyte fragments and
keratocytes (evidence of stress or damage to the bone marrow and evidence of increased RBC
injury or turnover) fNTP. 2008. 20071. Similar microscopic findings from blood smears were not
observed after 12 months Cr(VI) exposure.
Although the focus of the assessment is on the development of chronic reference values, it is
noted that hematologic effects were observed in studies with exposure durations <90 days. In
general, the direction of change was similar to the later time points, but the magnitude of response
was greater at observation times <90 days. For example, Hgb decreased by up to 35% at 22 days,
but 15% at 90 days, and a similar amelioration was observed for MCV and other hematological
markers fNTP. 20081. Other medium and high confidence studies were also available at exposure
durations <9 weeks. In general, these studies reported limited or no statistically significant changes
in hematologic parameters at the same dose levels where effects were observed in the subchronic
and chronic studies. Decreased MCV and MCH levels (<6%) were observed in Sprague-Dawley rats
exposed to >10 mg Cr(VI)/kg-day (via diet) for up to 9 weeks (NTP. 1996b). In two other 28-day
studies by NTP, hematologic effects at doses >9 mg Cr(VI)/kg-day exposure (via drinking water)
were not observed for RBCs, hemoglobin, hematocrit, and MCHC in female Sprague-Dawley and
F344 rats fNTP. 2006a. b). MCV and MCH findings were not dose responsive nor considered
biologically meaningful.
The hematologic effects of inhalation exposure were reported in one medium confidence
study (Kim etal.. 2004) where findings included increased RBC count (8%), decreased hematocrit
(<11%), and decreased hemoglobin (<8%) in Sprague-Dawley rats exposed for 90 days to Cr(VI)
concentrations ranging from 0.2-1.25 mg/m3. No effects on MCV or MCHC were observed. No
effects on RBCs were reported in male Wistar rats in three low confidence studies with exposure
durations that ranged from 28 days to 18 months fGlaser etal.. 1990: Glaser etal.. 1986: Glaser et
al.. 19851. whereas the 30- and 90-day experiments did not specify which hematologic parameters
were examined. The highest concentrations tested ranged from 0.1-0.4 mg/m3; the highest
concentration tested in the 18-month study by Glaser etal. (1986) (0.1 mg/m3) was lower than the
lowest concentration tested by Kim etal. (2004) (0.2 mg/m3).
3.2.5.3. Mechanistic Evidence
The subchronic and chronic studies provide evidence for microcytic hypochromic anemia
(characterized by low Hgb concentrations in abnormally small RBCs) after 90 days. After 12
months exposure, most findings returned to near control levels (Hgb, Hct, MCHC). The clinical
pathology and microscopic evaluation indicated small RBCs (microcytic) that were hypochromic
This document is a draft for review purposes only and does not constitute Agency policy.
3-195 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
(pale in color, consistent with decreased Hgb). The mechanistic studies described below provide
evidence for connecting these findings to upstream events, including altered iron metabolism
leading to iron deficiency, and oxidative stress potentially leading to RBC damage, smaller size, and
increased turnover.
Effects on iron homeostasis
Iron is a critical requirement for metabolic processes including oxygen transport,
deoxyribonucleic acid (DNA) synthesis, and electron transport fAbbaspour etal.. 20141. Iron
imbalance, deficiency, and overload have known health effects in humans including iron-deficient
anemia and iron toxicity. Iron is absorbed from the diet by villous enterocytes in the small
intestine. Cellular iron import involves both receptor-mediated endocytosis (by transferrin) of
ferric iron (Fe+3) as well as uptake of reduced iron ferrous iron (Fe2) by membrane-bound
transporters. A majority of the iron is contained by RBCs where iron is stored in complexes with
ferritin (in the ferric state), complexed by heme in the ferrous state (Fe2+), or to a smaller extent
labile in the cytosolic pool in the ferrous state (Fe2+). Several studies provided evidence that Cr(VI)
intereferes with iron homeostasis, thereby decreasing iron bioavailability. Although blood iron
measures were not available from the NTP studies, a subchronic study by Suh etal. (2014) reported
a dose-responsive reduction in iron levels in serum, duodenum, liver, and bone marrow in F344
rats and B6C3F1 mice administered Cr(VI) (as sodium dichromate dihydrate) in drinking water for
90 days (0.1-180 mg Cr(VI)/L) compared to controls. Decreased iron was accompanied by altered
expression of genes involved in iron transport and absorption. Based on these findings and the
knowledge that Cr(VI), Cr(V), and Cr(IV) can oxidize ferrous iron (Fe+2) to ferric iron (Fe+3) (Buerge
and Hug. 1997: Fendorf and Li. 1996). Suh etal. (2014) hypothesized that Cr(VI) may oxidize
ferrous (Fe+2) iron to ferric (Fe+3), thereby interfering not only with (Fe+2) absorption in the
intestinal lumen, but also competing with (Fe+2) for heme binding and ferric iron (Fe+3) storage by
ferritin in RBCs. Cr(VI), but not Cr(III) (NTP. 2010: Stout etal.. 2009). hinders iron aborption in the
small intestine, leading to iron deficiency in rats and to a lesser extent in mice. Consistent with this
hypothesis, Cr(VI) reduced to Cr(III) has been shown to bind transferrin under physiological
conditions fLevina etal.. 2016: Deng etal.. 20151. Consistent with Suh etal. f20141. Wang et al.
(2015) also observed dose-related decreases in iron levels in the liver, kidney, duodenum, and lung
in rats exposed to concentrations up to 106.1 mg/L Cr(VI) in drinking water for four weeks; no
changes were detected in blood iron levels, but significant decreases in Hgb, MCH, and MCHC levels
and increased RBC counts were observed. This evidence that Cr(VI) can inhibit iron absorption
suggests that humans with preexisting blood conditions (e.g., anemia, iron deficiency, intestinal
bleeding disorders) would be expected to be more sensitive to any potential hematologic effects of
Cr(VI) exposure. This includes pregnant women, who are susceptible to developing iron-deficient
anemia (American Pregnancy Association. 2021: O'Brien and Ru. 2017: Rahman etal.. 2016).
Oxidative stress. RBC membrane damage and ervptosis
This document is a draft for review purposes only and does not constitute Agency policy.
3-196 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Both iron deficiency and Cr(VI) exposure have been shown to independently increase
oxidative damage. Potassium dichromate, like iron, is a charged heavy metal, and it has been
proposed that interaction between iron bound by RBCs alters erythrocyte function and/or
formation particularly by targeting the erythron (NTP. 2007.1997.1996a. b). Cr(VI) redox results
in oxidative damage both to hemoglobin and to the RBC membrane (ATSDR. 2012: NTP. 20071. The
increased oxidative damage can initiate pathways leading to erythrocyte injury and eryptosis
(i.e., erythrocyte apoptosis) as well as smaller RBCs (Kempe etal.. 20061. consistent with
observations of decreased MCV in rats and mice fNTP. 2008. 20071.
As discussed in Section 3.2.1, "Respiratory effects other than cancer," evidence of oxidative
stress (i.e., increased oxidative 8-OHdG DNA adducts and lipid peroxidation levels, decreased
antioxidant levels) has been detected at significant levels in the blood (RBCs, plasma, serum) of
workers exposed to Cr(VI) (El Safty etal.. 2018: Hu etal.. 2018: Xu etal.. 2018: Mozafari etal.. 2016:
Elhosarvetal.. 2014: Zendehdel etal.. 2014: Kalahasthi etal.. 2006: De Mattiaetal.. 2004: Maenget
al.. 2004: Wu etal.. 2001: Huang etal.. 1999: Gromadziriska etal.. 19961 (see Appendix Table C-56).
In animals, one 4-week drinking water study in male F344 rats exposed to 10.6-106 mg Cr(VI)/L
and found increased plasma malondialdehyde (MDA), a reactive marker of lipid peroxidation, and
decreased glutathione peroxidase (GSH-Px), an antioxidant enzyme fWangetal.. 20151. Other
findings consistent across in vitro studies with primary human RBCs included observation of
oxidative stress indicators and eryptosis, including increased MDA levels, changes in antioxidant
activity, increased cytosolic Ca2+, increased phosphatidylserine on the outer membrane surface, and
decreased ATP (Sawicka and Dlugosz. 2017: Zhang etal.. 2014: Lupescu etal.. 2012: Ahmad etal..
2011: Fernandes etal.. 1999: Koutras etal.. 19641. These effects indicate a loss of membrane
integrity, coherent with the microscopic evaluations of blood smears from exposed rats and mice,
where evidence of erythrocyte injury, including poikilocytes, erythrocyte fragments/schizocytes,
and keratocytes, were observed after 90 days of Cr(VI) exposure in drinking water (NTP. 2008.
20071. Collectively, the findings of RBC oxidative stress leading to cell membrane damage and
eryptosis are a possible pathway leading to the observed changes in RBC size, and are correlative
with an erythrogenic response supported by increased RBC counts. However, study durations were
limited to <90 days and it is not clear if these mechanistic effects would be persistent long-term.
3.2.5.4. Integration of Evidence
Overall, the currently available evidence suggests that Cr(VI) exposure may cause
hematologic effects in humans. The conclusion of evidence suggests is based primarily on moderate
animal evidence from high and medium confidence subchronic and chronic studies in rats and mice
reporting consistent (across similar exposure durations and doses, sexes, and species), dose-
related, and coherent findings (i.e., in RBC, Hgb, MCHC, MCH, and MCV) at 90 days exposure. The
indeterminate human evidence consists of four low confidence studies that show inconsistent
effects on hematocrit and hemoglobin (positive, negative, and null associations). One low
This document is a draft for review purposes only and does not constitute Agency policy.
3-197 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofHexavalent Chromium
confidence study identified increased RBC and decreased MCV in exposed humans, whereas the
other low confidence studies identifed no association with RBCs. No epidemiological study was
identified that evaluated associations with exposure to Cr(VI) and anemia or other hematological
diseases or parameters.
Confidence in the findings in animal studies, however, is diminished due to the decrease in
magnitude of the collective effect by 12 months, with many findings returning to normal or near
normal levels (generally, with a magnitude of change <10% compared to controls). Given the
absence of correlative findings of apparent RBC injury from blood smears (other than smaller RBCs
that were hypochromic) and the absence of supportive mechanistic findings (such as iron
deficiency and oxidative stress) at 12 months, there exists uncertainty regarding the adverse versus
adaptive nature of the observed effects at exposure durations greater than 90 days. In particular,
the biological significance of the response at 12 months is uncertain, since most markers were
within 10% of controls.
Although the adversity or clinical relevance of the observed changes in any one of the
individual hematologic parameters in isolation is unclear and there is uncertatinty in the adversity
of the effect at 12 months, the interpretation of the collective animal evidence still signals a
potential concern. Supporting evidence of Cr(VI)-induced iron deficiency and oxidative stress
indicates potential pathways leading to the observed findings of hypochromic microcytic anemia,
consistent with the microscopic evaluation of blood smears (with findings of damage to the
erythron), strengthens the evidence for an effect at 90 days. Information including iron levels and
ferritin tests that are useful for evaluating the amount of stored iron were not available at exposure
durations >90 days, making it difficult to confirm whether the diminished effects at 12 months
should be considered adverse. Therefore, although there remains a (weaker) signal for an effect at
12 months, there exists a large amount of uncertainty as to the adversity of the effect. Integrated
evidence for the hematologic effects of Cr(VI) exposure from human, animal, and mechanistic
studies is summarized in an evidence profile table (Table 3-31). However, the mechanistic
evidence suggests that humans with preexisting blood conditions (e.g., anemia, iron deficiency,
intestinal bleeding disorders) would be expected to be more sensitive to any potential hematologic
effects of Cr(VI) exposure. This includes pregnant women, who are susceptible to developing iron-
deficient anemia fAmerican Pregnancy Association. 2021: O'Brien and Ru. 2017: Rahman etal..
20161.
This document is a draft for review purposes only and does not constitute Agency policy.
3-198 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-31. Evidence profile table for hematologic effects
Evidence summary and interpretation
Studies, outcomes,
and confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Evidence from studies of exposed humans
®©o
HEMATOLOGIC
PARAMETERS
Four low confidence
studies, two in
occupational^
exposed adult
workers and two in
general population
adults
Exposure to Cr(VI) was
associated with lower
hemoglobin and/or hematocrit
in one study (p < 0.05), while
two studies reported in the
opposite direction (higher
hemoglobin and hematocrit in
one study, higher red blood
cells in one), and one study
reported no association.
• No factors noted
• Low confidence
studies
ooo
Indeterminate
The available evidence
is inconsistent across
low confidence
studies.
The evidence suggests that
Cr(VI) may cause hematologic
effects in humans given
sufficient exposure conditions.
Consistent findings in high and
medium confidence animal
studies across species and dose
duration with coherent effects
on RBC indices and decreased
Hgb suggesting microcytic
Evidence from animal studies
anemia, with supportive
mechanistic findings of of
Hematologv
Six hiqh confidence
studies in adult rats
and mice
• 28-day oral
• 9-week oral (2
studies)
• Continuous
breeding oral
• 90-day oral
• 2-year oral
Five medium studies
in adult male and
female rats
Hematologic effects included
consistent decreases in Hgb,
MCV, MCH, and MCHC, and
increased RBC counts at 90
days; marginal (near low-
normal) decreases in MCV,
MCH and increase in RBC at 12
months. Most findings
returned to near normal by 12
month exposures.
90 day findings were coherent
with microscopic findings of
RBC damage including smaller
size and hypochromic
appearance that were
consistent with Cr(VI)-induced
iron deficiency.
• Consistent
findings of
decreased Hgb,
MCH, MCHC,
MCV, and
increased RBC
across species
and sexes in
subchronic and
chronic studies
• Coherence of
decreased Hgb,
MCH, MCHC, and
MCV with
increased RBC
• Lack of duration-
dependence
(effects of Cr(VI)
decreased with
longer-term
exposures)
• Uncertainty of
the biological
significance of
effects at 12
months
®©o
Moderate
Based primarily on
high and medium
confidence subchronic
and chronic studies
with consistent
findings across species
and sexes and
coherent effects
across multiple related
endpoints.
Strong dose response
relationship primarily
at 90 days, though
some uncertainty in
biological relevance of
Cr(VI)-induced iron deficiency
and RBC damage. However, the
confidence in these findings is
reduced by the uncertainty
regarding the adverse versus
adaptive nature of the
observed effects, particularly
given the near amelioration of
effects after one year,
precluding a higher confidence
judgment (i.e., evidence
indicates).
Human evidence was primarily
inconsistent and low
confidence. Without evidence
to the contrary, effects in rats
and mice are considered
relevant to humans.
This document is a draft for review purposes only and does not constitute Agency policy.
3-199 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Mechanistic findings of iron
deficiency and altered
pathways involved in iron
metabolism in rats exposed for
<90 days provide evidence
supportive of hematologic
effects. These mechanisms are
presumed to be relevant to
humans and are consistent
with findings of oxidative stress
in the blood of occupationally
exposed humans.
People with preexisting blood
conditions (e.g., anemia, iron
deficiency, chronic intestinal
bleeding disorders, pregnancy)
are expected to be susceptible
to hematological effects from
Cr(VI).
Studies, outcomes,
and confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
• 28-day oral (3
studies)
• 30-day oral
• 90-day
inhalation
Four low confidence
studies in male rats
and mice
• 28- and 90-day
inhalation (2
experiments, 1
study)
• 30- and 90-day
inhalation (2
experiments, 1
study)
• 18-month
inhalation
• Short-term oral
study during
pregnancy
Hct and reticulocyte changes
were inconsistent across
species and sexes.
• High and
medium
confidence
studies
• Large magnitude
of effect <90
days
• Dose-response
gradient for RBC,
MCH, MCV,
MCHC, Hgb (rat,
90-day)
• Mechanistic
evidence
provides
biological
plausibility
the effect as the
magnitude of the
change compared to
controls decreased by
12 mo.
Strong mechanistic
support for anemia
provided by
mechanistic studies
demonstrating Cr(VI)
induced iron
deficiency and
oxidative damage in
the blood of exposed
humans and animals,
and regenerative
responses consistent
with smaller RBC size.
This document is a draft for review purposes only and does not constitute Agency policy.
3-200 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes,
and confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Mechanistic Evidence
Biological events or
pathways
Summary of key findings and interpretation
Judgments and
rationale
Oxidative Stress
Interpretation: Oxidative stress caused by Cr(VI) reactive intermediates may
lead to erythrocyte lipid peroxidation, membrane damage, and eryptosis.
Key Findings:
• Consistent evidence of oxidative stress in the blood of workers exposed
to Cr(VI) (see Section 3.2.1, "Respiratory effects other than cancer")
• Increased oxidative stress levels in plasma in one in vivo study of rats
exposed in drinking water for 4 weeks
• Cr(VI) increased markers of oxidative stress, cellular injury and death in
primary human RBCs in vitro, including MDA, decreased antioxidant
enzymes, increased cytosolic Ca2+, membrane destabilization, and
decreased ATP
Biologically plausible
pathways leading to
the observed clinical
pathology and
microscopic
evaluation of blood
smears that included
Cr(VI) oxidation of
ferrous to ferric iron,
potentially altering
bioavailability,
oxidative damage to
the RBC leading to
increased turnover
and smaller size, and
Cr(VI) interference
with iron metabolism
leading to iron
deficiency. Support for
oxidative stress
occurring in the blood
of humans is provided
by consistent findings
of increased markers
of oxidative stress in
exposed workers.
Iron Deficiency
Interpretation: Interference with iron homeostasis due to interactions with
Hgb, iron and its transporter proteins may also contribute to hematologic
toxicity.
Key Findings:
• Cr(VI) interaction with iron may alter RBC binding and erythrocyte
function or formation
• Cr(VI) reduced to Cr(lll) may bind transferrin, an iron transporter, under
physiological conditions
• Additional in vivo evidence suggests Cr(VI)-induced alterations in iron
homeostasis including dose-dependent decreases in total iron in various
tissues, altered gene regulation, and increased ratios of RBC
Cr(VI):plasma Cr(VI)
This document is a draft for review purposes only and does not constitute Agency policy.
3-201 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
3.2.6. Immune effects
The purpose of the immune system is to provide protection from infections and, in some
cases, the development of neoplasms. A properly functioning immune system involves a delicate
interplay among many cell types working in concert to properly regulate the immune response.
The immune system is integrated into tissues, organs and peripheral sites throughout the body. For
this reason, xenobiotic exposure by virtually any route can adversely impact components of the
immune system. Modulation of the immune system in either direction can result in dysfunction.
Xenobiotic exposure can alter primary immune sites important for immune cell maturation,
including the bone marrow, liver, thymus, and Peyer's patches. Secondary lymphoid sites
(i.e., spleen, lymph nodes, tonsils) can also be impacted by exposure to immunotoxicants.
Immunotoxicity may be expressed as immunosuppression, unintended stimulation of immune
responses, hypersensitivity, or autoimmunity (IPCS. 20121. Data from functional assays provide the
most sensitive and specific evidence of immune hazard.
This synthesis is organized and summarized based on the World Health Organization's
Guidance for Immunotoxicity Risk Assessment for Chemicals (IPCS. 20121 that describes best
approaches for weighing immunotoxicological data. Within this framework, data from endpoints
observed in the absence of an immune stimulus (e.g., levels of serum immunoglobulins, white blood
cell (WBC) counts, WBC differentials, T cell subpopulations, immune organ weights) are not
sufficient on their own to draw a conclusion regarding immune hazard but may provide useful
supporting evidence, especially when evaluated in the broader context of functional data flPCS.
20121. Consequently, the sections that follow are organized into two categories: the more
informative measures of immune system function and supporting immune system data.
3.2.6.1. Human Evidence
Study evaluation summary
Table 3-32 summarizes the human epidemiology studies considered in the evaluation of the
potential effects of Cr(VI) on the immune system. There were nine included human studies, all of
which were classified as low confidence. Four additional studies were identified and classified as
uninformative due to critical deficiencies in exposure methods sensitivity and/or confounding and
were not considered further (Islam etal.. 2019: Khan etal.. 2013: Kativar etal.. 2008: Snyder etal..
19961. All nine included studies were cross-sectional, and all but one were occupational studies
conducted among workers in industries with known risk of exposure to Cr(VI), in a range of
geographical locations. They include two studies of chrome-plating workers fKuo and Wu. 2002:
Verschoor etal.. 19881. two studies of tannery workers (Mignini etal.. 2009: Mignini etal.. 20041.
two studies of chemical plant workers fOian etal.. 2013: Tanigawa etal.. 19981. one study of
chromate production workers (Wang etal.. 2012a) and one study of plastic workers (Boscolo etal..
1997). In addition, one cross-sectional study assessed the effects of Cr(VI) exposure on the general
This document is a draft for review purposes only and does not constitute Agency policy.
3-202 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
population in Greece fSazakli et al.. 20141. Information on study evaluation is provided in the text
below and in Table 3-32. Available evidence in human studies was limited to ex vivo WBC function,
white blood cells (number, type, and T cell subpopulations), immunoglobulin levels, complement
levels, and cytokine levels.
While cytokines are critical for maintaining immune homeostasis, cytokine data, especially
measures of blood cytokines, can be challenging to interpret as primary evidence of immune hazard
(Tarrant. 20101. Changes in cytokine levels can be associated with many different types of tissues
and toxicities, as part of cell differentiation to different immune cell types, or including site-specific
inflammation, which reflects an immune response to tissue injury but not necessarily an impact on
or impairment of immune function. For this reason, cytokine secretion data (in the absence of a
stimulus) were not considered apical outcomes for the purpose of identifying immune hazard, but
rather as supporting evidence for understanding mechanisms of immune disruption and are
summarized in the Mechanistic and Supporting Evidence section below without systematic review.
Allergic sensitization can occur in some individuals exposed to Cr(VI) (OSHA. 20061.
Because the primary exposure route (i.e., dermal) is outside the scope defined by the PECO criteria,
evidence for allergic hypersensitivity responses following Cr(VI) exposure has not been
comprehensively reviewed, but is briefly summarized in the Mechanistic and Supporting Evidence
section below if the exposures or outcomes were relevant to non-dermal Cr(VI) exposures
Table 3-32. Summary of human studies for Cr(VI) immune effects and overall
confidence classification [high (H), medium (M), low (L)] by outcome. Click to
see interactive data graphic for rating rationales.
Author (year)
Industry
Location
Exposure
Measurement
Study
Design
Ex vivo white blood cell
function3
White blood cells
(hematology)
White blood cells
(subpopulations)
Immunoglobulin levels
Boscolo et al. (1997)
Plastic workers
exposed to
lead chromate
Italy
Air
Cross-
sectional
L
L
L
Kuo and Wu (2002)
Chrome-plating
workers
Taiwan
Urine, air
Cross-
sectional
-
-
L
-
Mignini et al. (2004)
Tannery
workers
Italy
Dust, blood, urine
Cross-
sectional
L
-
L
-
Mignini et al. (2009)
Tannery
workers
Italy
Air, blood, urine
Cross-
sectional
L
L
This document is a draft for review purposes only and does not constitute Agency policy.
3-203 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Toxicological Review ofHexavalent Chromium
Author (year)
Industry
Location
Exposure
Measurement
Study
Design
Ex vivo white blood cell
function3
White blood cells
(hematology)
White blood cells
(subpopulations)
Immunoglobulin levels
Qian et al. (2013)
Chemical plant
workers
China
Work categories,
validated by air,
urine, blood
samples
Cross-
sectional
L
Sazakli et al. (2014)
General
population
Greece
Urine, Hair,
Modeled lifetime
chromium
exposure dose
Cross-
sectional
L
Tanigawa et al.
(1998)
Chemical plant
workers
Japan
Work category
Cross-
sectional
-
-
L
-
Verschoor et al.
(1988)
Chrome
platers,
stainless-steel
welders
Netherland
s
Work categories,
validated by urine
samples
Cross-
sectional
L
Wang et al. (2012a)
Chromate
production
workers
China
Urine
Cross-
sectional
L
aEx vivo white blood cell function is more informative of immune system function, while the other measures
provide supporting immune system data.
Synthesis of human evidence
More informative measures of immune system function
Ex vivo WBC functional assays (e.g., NK cell activity, phagocytosis, proliferative responses)
are performed outside the body using isolated cells collected from exposed individuals. These
assays are considered clear evidence of adverse immunosuppression (IPCS. 20121. Two studies
examined the association between occupational Cr(VI) exposure and ex vivo WBC function (Table
3-32). Both studies of tannery workers were low confidence, with deficient ratings in participant
selection, exposure measurement, and sensitivity domains fMignini etal.. 2009: Mignini et al..
20041. Among Cr(VI) exposed workers, there was no effect on phagocytosis by PMNs or NK cell
activity Mignini et al. (20091: however, there was an increase in mitogen-induced proliferative
response that was not seen in workers without Cr(VI) exposure (Mignini etal.. 2009: Mignini etal..
20041 (Table 3-33). Compared to controls, lymphocytes harvested from the exposed workers were
stimulated to proliferate to a greater extent in the presence of the T cell mitogens
phytohemagglutinin (PHA) (Migninietal.. 20091 andconcanavalin A (ConA) (Mignini etal.. 2009:
This document is a draft for review purposes only and does not constitute Agency policy.
3-204 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 Mignini etal.. 20041. and there was evidence that the effect of Cr(VI) exposure on ConA stimulation
2 may be affected by HLA haplotype fMignini et al.. 20041. Cr(VI) exposure had no effect on
3 lymphocyte proliferation in the presence of the B cell mitogen lipopolysaccharide (LPS) fMignini et
4 al.. 20091.
Table 3-33. Associations between Cr(VI) exposure and ex vivo WBC function in
epidemiology studies
Reference,
confidence
Population
Exposure
measu re
Exposure
levels
Comparison
and effect
estimate
Endpoint
Mignini et al.
(2004), low
Cross-
sectional
study in Italy
of 20
exposed and
Cr levels in
blood and urine
NR
ANOVA and
the
Student's t
test
Significant increase in mitogen-
stimulated lymphocyte
proliferation (ConA) in exposed
groups (pooled data from both
exposure groups)
24
unexposed
workers
Significant increase in mitogen-
stimulated lymphocyte
proliferation in exposed HLA-B8-
DR3-negative group to ConA, but
not in the HLA-B8-DR3-positive
group (pooled data from both
exposure groups)
Mignini et al.
(2009), low
Cross-
sectional
study in Italy
of 40
exposed
Cr levels in
urine,
3 categories
~0.6, 0.4, 0.2
ug/L
Means by
exposure
category
(not
reported)
Significant increase in mitogen-
stimulated lymphocyte
proliferation in high exposure
group to PHA and ConA, but not
to LPS
tannery
workers and
44
unexposed
workers
No effect on percent
phagocytosis, phagocytosis index,
or killing percent by PMNs
No effect on NK cell activity, data
not shown
5 Supporting immune system data
6 Immunoglobulin levels
7 Three studies examined the association between Cr(VI) exposure and nonspecific
8 immunoglobulin levels (Table 3-32). All three studies were low confidence, with deficiencies in
9 participant selection, outcome ascertainment, and confounding fOian etal.. 2013: Boscolo etal..
10 1997: Verschoor et al.. 19881. Immunoglobulin levels are difficult to interpret alone without a
11 controlled immune challenge preceding the measurement. Among these studies (Table 3-34),
12 which did not include controlled immune challenges, Cr(VI)-exposed workers had lower levels of
13 IgAandlgG Oian etal. (20131. butlevels were unaffected in Boscolo etal. (19971. Levels of IgG
14 were also unaffected in Verschoor et al. (19881. Serum levels of IgM were unaffected by Cr(VI)
This document is a draft for review purposes only and does not constitute Agency policy.
3-205 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 exposure in the only two studies that investigated this isotype fOian etal.. 20131. IgE levels were
2 unaffected in the only study that investigated this isotype fBoscolo etal.. 19971.
Table 3-34. Associations between Cr(VI) exposure and immunoglobulin (Ig)
levels in epidemiology studies
Comparison
Reference,
Exposure
Exposure
and effect
confidence
Population
measure
levels
estimate
IgG
IgM
IgA
IgE
Boscolo et al.
Cross-sectional
Exposed/
Chromate
Median in
Exposed:
Exposed:
Exposed:
NA
(1997), low
study in Italy of 15
unexposed.
ranged in
mg/dl
1240(991—
118 (75—
193 (182—
plastic workers
Chromium
air from
(25th—75th)
1296)
140)
282)
exposed to lead
levels
0.1 to 5.7
for exposed
Unexposed:
Unexposed:
Unexposed:
chromate and 15
measured in
M-g/m3
and
1151 (942-
79 (58-111)
277 (186—
unexposed workers
blood and
unexposed
1276)
292)
from the same area
urine. Levels
in exposed
were
significantly
higher in
urine, but
similar to
unexposed in
blood
Verschoor et
Cross-sectional
Work
9, 3, 1,
Mean ± SD
Chrome
NA
NA
NA
al. (1988). low
study in the
categories,
0.4 Mg/g
platers:
Netherlands of 21
validated by
creatinine
11.6 ±3.2
chrome platers, 38
urine samples
in urine
SS welders:
SS welders, 16
11.1 ±2.6
boilermakers, and
Boilermakers:
63 unexposed
11.1 ±2.8
workers
Controls:
11.6 ±2.4
Qian et al.
Cross-sectional
Exposed/
14.4 ± 18.
Except for
Exposed:
Exposed:
Exposed:
Exposed
(2013), low
study in China of 56
unexposed
1 M-g/m3
IgE, mean in
10.9 ±2.5
1.2 ±0.5
2.4 ±0.9
(Median
workers exposed to
validated by
g/L ± SD for
Unexposed:
Unexposed:
Unexposed:
g/L
potassium
air sampling
exposed and
12.4 ±2.1
1.0 ±0.4
2.8 ±1.2
(quartile)
dichromate and 50
unexposed
p = 0.03*
p = 0.04*
] 55.2
unexposed
(157.4)
individuals living 20
Unexpos
km from factory
ed 81.9
(237.1)
NA = not applicable.
3 WBC counts (hematology)
4 Three studies reported WBC counts, or related measures, including counts of total WBCs,
5 lymphocytes and granulocytes (Table 3-35). All studies were low confidence. Sazakli et al. f20141
6 was deficient only in exposure measurement, while the remaining studies were deficient in
7 multiple domains, including participant selection (Wang etal.. 2012a: Boscolo et al.. 19971.
8 confounding (Wangetal.. 2012a: Boscolo etal.. 19971. and outcome ascertainment (Boscolo etal..
This document is a draft for review purposes only and does not constitute Agency policy.
3-206 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 19971. Among these studies, one reported a statistically significant increase in total WBCs with
2 higher exposure to Cr(VI) fWang etal.. 2012al. Non- significant increases were also observed for
3 lymphocytes and neutrophils (Wang etal.. 2012al. Two other studies indicated no increase
4 (Sazakli etal.. 2014: Boscolo etal.. 19971. with one indicating non-statistically significant decreases
5 for lymphocytes and WBCs (Boscolo etal.. 19971 (Table 3-35).
This document is a draft for review purposes only and does not constitute Agency policy.
3-207 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-35. Associations between Cr(VI) exposure and WBC counts in epidemiology studies
Reference,
confidence
Population
Exposure measure
Exposure
levels
Comparison and
effect estimate
Total WBC
(count 109/L)
Lymphocytes
(count 109/L)
Granulocytes
(count 109/L)
Neutrophils
(count 109/L)
Sazakli et al. (2014)
Cross-sectional
in Greece,
general
population; Two
exposure groups
(n = 237) and
controls (n = 67)
Chromium levels
measured in blood
and hair. Estimated
lifetime chromium
exposure dose
calculated using
concentration in
drinking water, intake
rate, and body weight
NR
Regression
coefficients for
calculated
lifetime exposure
dose and Cr in
hair
Lifetime dose:
-0.03
Hair:
0.07
p = 0.59
Lifetime dose:
0.02
Hair:
0.1
p = 0.71
Lifetime dose:
-0.01
Hair:
0.03
p = 0.81
NA
low
Boscolo et al. (1997),
Cross-sectional
study in Italy of
15 plastic
workers exposed
to lead
chromate and
15 unexposed
workers from
the same area
Exposed/ unexposed.
Chromium levels
measured in blood
and urine. Levels in
exposed were
significantly higher in
urine, but similar to
unexposed in blood
Chromate
ranged in
air from
0.1 to 5.7
Hg/m3
Median (25th—
75th) for exposed
and unexposed
Exposed:6764
(5940-7180)
Unexposed:
6776 (5680-
8190)
p > 0.05
Exposed: 2340
(1490-2915)
Unexposed:
2730 (2300-
3090)
p > 0.05
NA
NA
low
Wang et al. (2012a),
Cross-sectional
study in China of
86 chromate
production
workers and 45
unexposed
workers
Exposed/ unexposed.
Chromium levels
measured in urine
were significantly
higher in exposed
workers
<50 ng/m3
Mean (SD) for
exposed and
unexposed
Exposed: 7.0
(1.7)
Unexposed: 6.2
(1.3)
p = 0.03
Mixed WBCa
Exposed: 0.6
(0.3)
Unexposed: 0.4
(0.1)
Exposed: 2.2
(0.7)
Unexposed: 2.1
(0.5)
p = 0.19
NA
Neutrophils
Exposed: 4.1
(1.4)
Unexposed: 3.7
(1.0)
p = 0.06
low
NA = not applicable.
aCell mixture containing neutrophils, eosinophils, basophils and mast cells.
This document is a draft for review purposes only and does not constitute Agency policy.
3-208 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicological Review ofHexavalent Chromium
Lymphocyte subpopulations
Five studies examined the association between Cr(VI) exposure and lymphocyte
subpopulations (Table 3-32). All five studies were low confidence cross-sectional studies of Cr(VI)
exposure and white blood cell counts fOian etal.. 2013: Mignini etal.. 2009: Mignini etal.. 2004:
Tanigawa etal.. 1998: Boscolo etal.. 19971. All studies were deficient in multiple domains,
including selection or performance fOian etal.. 2013: Mignini etal.. 2009: Mignini etal.. 2004:
Tanigawa etal.. 1998: Boscolo etal.. 19971. exposure methods sensitivity fMignini et al.. 2009:
Mignini etal.. 2004: Tanigawa etal.. 19981. outcomes measures and results display sensitivity fOian
etal.. 2013: Boscolo etal.. 19971. confounding fOian etal.. 2013: Tanigawa etal.. 1998: Boscolo et
al.. 19971. analysis fOian etal.. 2013: Tanigawa etal.. 1998: Boscolo et al.. 19971. selective reporting
fOian etal.. 20131. and sensitivity fMignini et al.. 2009: Mignini etal.. 2004: Tanigawa etal.. 1998:
Boscolo et al.. 19971. Three studies reported decreased CD4+, CD8+, and CD3+ cells with higher
exposure to Cr(VI) fKuo and Wu. 2002: Tanigawa etal.. 1998: Boscolo etal.. 19971. Two studies did
not report data for changes in levels of CD3+, CD4+, CD8+, DC19 fMignini etal.. 2009: Mignini etal..
20041. CD56 fMignini etal.. 20041. CD16+/CD56+ and CD4/CD8 fMignini etal.. 20091. but stated
there were no significant associations with measures of Cr(VI) exposure (Table 3-36).
This document is a draft for review purposes only and does not constitute Agency policy.
3-209 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-36. Associations between Cr(VI) exposure and lymphocyte subpopulations in epidemiology studies
Reference,
Exposure
Comparison and
confidence
Population
Exposure measure
levels
effect estimate
CD4+
CD8+
CD3+
CD19
CD56
Boscolo et al.
Cross-sectional
Exposed/
Chromate
Median (25th—
Exposed:
Exposed:
Exposed:
Exposed:
NA
(1997), low
study in Italy of
unexposed.
ranged in
75th) for exposed
870 (585-
710 (435-795)
1630 (1035-
180
15 plastic
Chromium levels
air from 0.1
and unexposed
1135)
Unexposed:
1995)
(150-
workers
measured in blood
to 5.7
Unexposed:
810 (570-870)
Unexposed:
280)
exposed to lead
and urine. Levels in
Hg/m3
1140 (970-
1890 (1680-
Unexpos
chromate and
exposed were
1240)
2170)
ed:
15 unexposed
significantly higher
p < 0.05*
330
workers from
in urine, but similar
(260-
the same area
to unexposed in
460)
blood
Tanigawa et al.
Cross-sectional
Exposed/unexpose
NR
Mean ± SD for
Exposed
Exposed
Exposed
NA
NA
(1998), low
study in Japan
d. No validation of
exposed and
smokers:
smokers:
smokers:
of 19 retired
exposure levels.
unexposed, by
790 ± 260
470 ± 250
1140 ± 380
chromate
smoking status
Exposed
Exposed
Exposed
workers and 13
nonsmokers:
nonsmokers:
nonsmokers:
unexposed
870 ±510
330 ± 200
1150 ±640
workers
Unexposed
Unexposed
Unexposed
smokers:
smokers:
smokers:
1660 ± 570
540 ± 280
2110 ± 530
Unexposed
Unexposed
Unexposed
non-
nonsmokers:
nonsmokers:
smokers:
670 ± 480
1840 ± 650
1250 ± 450
p < 0.05*
p < 0.05*
p < 0.05*
Kuo and Wu
Cross-sectional
Chromium levels in
NR
Beta (SE) for
Moderate:
Moderate: -1.8
NA
NA
NA
(2002), low
study in Taiwan
air samples and
moderate and
-0.03 (2.5)
(2.3)
of 27 workers
urine.
high urine Cr vs.
High: -0.2
High: -6.5 (3.6)
from 5 Cr
low group
(4.0)
This document is a draft for review purposes only and does not constitute Agency policy.
3-210 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Reference,
confidence
Population
Exposure measure
Exposure
levels
Comparison and
effect estimate
CD4+
CD8+
CD3+
CD19
CD56
electroplating
plants and 19
unexposed
workers
Correlation
coefficient with
airborne Cr
-0.06
-0.08
NA
NA
NA
Mignini et al.
Occupational
exposure study
in Italy of 20
exposed and 24
unexposed
workers
Cr levels in blood
and urine
NR
ANOVA and the
Student's t test
No changes
reported,
data not
shown
No changes
reported, data
not shown
No changes
reported, data
not shown
No
changes
reported,
data not
shown
No
changes
reported,
data not
shown
(2004), low
Mignini et al.
Cross-sectional
study in Italy of
40 exposed
tannery workers
and 44
unexposed
workers
Cr levels in urine,
3 categories
~0.6, 0.4,
0.2 ug/L
Mean ± SD for
exposed and
unexposed,
Duncan Multiple
Range,' 'Newman-
Keuls, Mann-
Whitney test
No changes
reported,
data not
shown
No changes
reported, data
not shown
No changes
reported, data
not shown
No
changes
reported,
data not
shown
No
changes
reported,
data not
shown
(2009), low
NA = not applicable.
This document is a draft for review purposes only and does not constitute Agency policy.
3-211 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofHexavalent Chromium
3.2.6.2. Animal Evidence
This section focuses on outcomes considered informative for the identification of
chemical-induced adverse effects on the immune system (IPCS. 2012: U.S. EPA. 1998bl. particularly
changes in response to an immune challenge, including effects on antibody responses, host
resistance, and ex vivo white blood cell (WBC) function. Supporting data collected from animals in
the absence of an immune challenge were also considered, including effects on immune organ
pathology, nonspecific immunoglobulin levels, immune organ weights, WBC counts (spleen,
thymus, bone marrow and hematology), and lymphocyte subpopulations. In addition to the
evidence syntheses below, the study findings have been summarized in Appendix C.2.5.1.
Study evaluation summary
Table 3-37 summarizes the animal toxicology studies considered in the evaluation of the
effects of Cr(VI) on the immune system. These studies consist of one oral diet fNTP. 1996al. one
oral gavage (Krim etal.. 20131.11 drinking water (Karaulov et al.. 2019: Tin etal.. 2016: Wang etal..
2015: NTP. 2008. 2007. 2006a. b, 2005: Shrivastava etal.. 2005a: Shrivastava etal.. 2005b: Snyder
and Valle. 19911. and eight inhalation studies (Cohen etal.. 2010: Cohen etal.. 2006: Kim etal..
2004: Cohen etal.. 1998: Glaser etal.. 1990: Glaser etal.. 1986: Tohansson etal.. 1986b: Glaser etal..
19851. These studies used a variety of mouse and rat strains, including BALB/c, B6C3F1,
am3-C57BL/6, and Swiss mice (NTP. 2008. 2007. 2005: Shrivastava et al.. 2005a: Shrivastava et al..
2005b: NTP. 1996al and Sprague-Dawley, F344, F344/N, Wistar, and albino Wistar rats fKaraulov
etal.. 2019: Wang etal.. 2015: Krim etal.. 2013: Cohen etal.. 2010: NTP. 2008. 2007: Cohen etal..
2006: NTP. 2006a. b; Kim etal.. 2004: Cohen etal.. 1998: Snyder and Valle. 1991: Glaser etal.. 1990:
Glaser etal.. 1986: Glaser etal.. 19851.
This document is a draft for review purposes only and does not constitute Agency policy.
3-212 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-37. Summary of included studies for Cr(VI) immunological effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.3
Click to see interactive data graphic for rating rationales.
More
informative
measu resb
Supporting evidence
Author (year)
Species (strain)
Exposure
design
Exposure
route
Host resistance
Antibody responses
Ex vivo WBC function
Immune organ pathology
Immunoglobulin levels
Immune organ weights
WBC counts and differentials
(spleen, thymus, bone marrow)
WBC counts (hematology)
Cohen et al. (1998)
Rat (F-344)
Short-term
Inhalation
-
-
M
-
-
-
-
-
Cohen et al. (2006)
Rat (F-344)
Short-term
Inhalation
M
Cohen et al. (2010)
Rat (F-344)
Short-term
Inhalation
M
-
-
-
-
-
-
-
Glaser et al. (1985)
Rat (Wistar)
Short-term &
subchronic
Inhalation
-
L
L
-
M
M
-
L
Glaser et al. (1986)
Rat (Wistar)
Chronic
Inhalation
-
-
-
M
L
L
-
M
Glaser et al. (1990)
Rat (Wistar)
Short-term &
subchronic
Inhalation
-
-
-
-
L
-
-
M
Jin et al. (2016)
Mouse (ICR)
Short-term
Drinking
water
-
-
-
-
-
M
-
-
Johansson et al.
(1986b)
Rabbit (strain not
specified)
Chronic
Inhalation
-
-
M
-
-
-
-
-
Karaulov et al.
(2019)
Rat (Wistar)
Chronic
Drinking
water
-
-
M
M
-
M
M
-
Kim et al. (2004)
Rat (Sprague-
Dawley)
Subchronic
Inhalation
-
-
-
-
-
M
-
M
Krim et al. (2013)
Rat (albino
Wistar)
Short-term
Gavage
-
-
-
-
-
-
-
M
NTP (1996a)
Mouse (BALBC)
Subchronic
Diet
-
-
-
-
-
H
NTP(2005)
Mouse (B6C3F1)
Short-term
Drinking
water
-
H
H
H
H
H
H
M
NTP(2006b)
Rat (Sprague-
Dawley)
Short-term
Drinking
water
-
H
H
M
H
H
H
M
This document is a draft for review purposes only and does not constitute Agency policy.
3-213 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
More
informative
measu resb
Supporting evidence
Author (year)
Species (strain)
Exposure
design
Exposure
route
Host resistance
Antibody responses
Ex vivo WBC function
Immune organ pathology
Immunoglobulin levels
Immune organ weights
WBC counts and differentials
(spleen, thymus, bone marrow)
WBC counts (hematology)
NTP (2006a)
Rat (F344)
Short-term
Drinking
water
-
H
H
M
H
H
H
M
NTP (2007)
Rat (F344/N);
Mice (B6C3F1,
BALB/c, am3-
C57BL/6)
Subchronic
Drinking
water
H
H
NTP(2008)
Rat (F344/N);
Mice (B6C3F1)
Chronic
Drinking
water
-
-
-
H
-
-
-
H
Shrivastava et al.
(2005a)
Mouse (Swiss)
Short-term &
subchronic
Drinking
water
-
-
-
-
-
-
-
L
Shrivastava et al.
(2005b)
Mouse (Swiss)
Short-term &
subchronic
Drinking
water
-
-
L
-
-
L
-
-
Snyder and Valle
(1991)
Rat (F344)
Short-term
Drinking
water
-
-
L
-
-
-
-
-
Wang et al. (2015)
Rat (Sprague-
Dawley)
Short-term
Drinking
water
-
-
-
-
-
-
-
M
aln addition to these included studies, there were three animal toxicology studies reporting immunotoxicity
outcomes that met PECO criteria but were found to be uninformative at the study evaluation stage for reporting
or attrition Geetha et al. (2003), outcomes measures Nettesheim et al. (1971), and outcomes measures, exposure
methods, reporting or attrition, confounding variable control, and selection or performance Kumar and Barthwal
(1991).
bHost resistance, antibody responses, and ex vivo WBC function are more informative as measures of immune
system function. The remaining measures provide supporting immune system data.
1 Synthesis of Animal Evidence
2 More informative measures of immune system function
3 Host resistance
4 Host resistance assays are considered the gold standard of immunotoxicity testing because
5 clearance of a self-replicating infectious agent or neoplastic disease requires the integration of
6 immune system responses to protect the host, and disruption of this integrated response at any
This document is a draft for review purposes only and does not constitute Agency policy.
3-214 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
point can be detected as a reduction in host resistance. The effect of exposure to Cr(VI)
(0.119 mg/m3 for 5 h/d for 5 consecutive days) on in situ clearance of pneumonia-inducing Listeria
monocytogenes (24, 48 and 72 h timepoints) was investigated in two medium confidence studies of
male F344 rats (Cohen etal.. 2010: Cohen etal.. 20061. Compared to the air-exposed control,
pathogen clearance was reduced in rats exposed to high soluble (Na2CrC>4) and low soluble
(CaCrCU) Cr(VI), but only when measured at the 72 h timepoint (Cohen etal.. 2010: Cohen etal..
20061. The authors noted that the reduction in pathogen clearance did not correlate with lung
chromium burden fCohen etal.. 2010: Cohen etal.. 20061. Overall, available data suggest that
short-term exposure to chromium may reduce in situ bacterial clearance in the lung (i.e., phagocyte
recruitment and bacterial lysis). Since the model used in these studies is a targeted host resistance
model designed to evaluate local pathogen clearance by macrophages, future studies using a
comprehensive host resistance model (e.g., influenza virus) would be useful for developing a better
understanding of the potential for Cr(VI) exposure to impair host resistance.
Antibody responses
Cr(VI) exposure increased IgM antibody-forming cell responses to sheep red blood cells in
three high confidence 28-day NTP studies (NTP. 2006a. b, 20051. but the effect was only significant
in two of the studies fNTP. 2006a. 20051 and the same effect was not observed in a repeat assay
performed by NTP f20051. One 90-day inhalation study, found to be low confidence due to
deficiencies in the presentation of results, also reported increased IgM antibody-forming cell
responses to sheep red blood cells fGlaser etal.. 19851. These investigations were performed in
female B6C3F1 mice and two different strains of female rat exposed to a broad and overlapping
range of Cr(VI) in drinking water (5-180 mg/L) and according to experimental protocols sufficient
for the detection of alterations in antibody cell forming responses.
Antibody response studies only provide information on the number of antibody producing
plasma cells at the time of assay completion, but these studies do not provide any information on
the levels of antigen-specific antibodies in the serum of Cr(VI)-exposed animals. Three high
confidence NTP studies in mice and rats exposed to Cr(VI) in drinking water for 28 days showed no
effect on serum titers of total IgM antibodies specific for two different T cell-dependent antigens
(NTP. 2006a. b, 20051. Recognizing that serum antibody titers are a relatively insensitive measure
of the antibody response, these findings are not inconsistent with the antibody-forming cell
responses discussed above.
Overall, Cr(VI) exposure increased antibody responses to sheep red blood cells but did not
alter the serum antibody titer following exposure to Cr(VI).
Ex vivo WBCfunction
In a low confidence study by Glaser etal. (19851. phagocytic activity was significantly
increased compared to the control group in alveolar lung macrophages isolated from male Wistar
rats exposed to Cr(VI) (up to 0.050 mg/m3) as sodium dichromate by inhalation for 28 and 90 days
but was decreased significantly following a 90-day exposure to 0.20 mg/m3. Findings by the
This document is a draft for review purposes only and does not constitute Agency policy.
3-215 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
companion study fGlaser etal.. 19901 also showed changes characteristic of acute lung injury and
inflammatory lung responses (see Section 3.2.1.2). In a second, medium confidence inhalation
exposure study, phagocytosis by rabbit alveolar macrophages was unaffected following exposure to
0.9 ± 0.4 mg/m3 Cr(VI) as sodium chromate for 4-6 weeks (Johansson etal.. 1986b). The absence
of an effect in Tohansson etal. f!986bl may have been due to a 3-day gap between cessation of
exposure to Cr(VI) and evaluation of phagocytic activity. In Glaser etal. (19851. the clearance of
inhaled iron oxide was lower in the lungs of rats exposed to 0.20 mg/m3 Cr(VI) for 42 days, though
the number of lung macrophages was also reduced relative to the control group. Consequently, the
observed decrease in lung clearance cannot be attributed definitively to a defect in phagocytosis. In
a third low confidence study, however, phagocytic activity of mouse splenic macrophages was
reduced from 92% in control male Swiss mice to 36% in mice exposed to 14.8 mg/kg-day Cr(VI) in
drinking water for 9 weeks fShrivastava etal.. 2005bl.
Cr(VI) exposure had no effect on natural killer (NK) cell activity, mixed lymphocyte
response (MLR), and anti-CD3 stimulation of lymphocytes in three high confidence drinking water
studies (NTP. 2006a. b, 20051 and one low confidence drinking water study (Snvder and Valle.
19911. The studies were performed in female B6C3F1 mice and two different strains of female rats
(Sprague-Dawley and F344) exposed to a broad and overlapping range of Cr(VI) in drinking water
(5-180 mg/L) and according to experimental protocols sufficient for the detection of alterations in
cell-mediated responses.
Mitogen-induced proliferative response was consistent in three low confidence studies
fShrivastava et al.. 2005b: Snyder and Valle. 1991: Glaser etal.. 19851. Spleen cells isolated from
male Swiss mice exposed to Cr(VI) in drinking water (14.8 mg/kg-day) for 9 weeks were stimulated
to proliferate with ConA, but the investigators did not conduct statistical analyses of the findings
fShrivastava et al.. 2005bl. Increased proliferation was observed in splenocytes isolated from F344
rats exposed to Cr(VI) in drinking water (100 or 200 mg/L) for 3 weeks when stimulated with the T
lymphocyte mitogen ConA or B lymphocyte mitogen lipopolysaccharide (LPS) (Snyder and Valle.
19911. Spleen cells isolated from rats exposed to Cr(VI) by inhalation (0.20 mg/m3) for 90 days
were stimulated to proliferate to a greater extent than controls by ConA fGlaser etal.. 19851.
Mitogen-induced cytokine secretion was evaluated in two medium confidence studies
(Karaulov et al.. 2 019: Cohen etal.. 19981. Spleen cells isolated from rats exposed to Cr(VI) in
drinking water for 45, 90, and 135 days and stimulated with ConA secreted less IL-6 (day 135) and
more IL-4 (day 45, 90, and 135) than controls, while secretion of IL-10 and IFNy were unaffected by
treatment (Karaulov etal.. 20191. Compared to control, secretion of IL-1 and TNFa were decreased
in pulmonary alveolar macrophages harvested from rats exposed to Cr(VI) by inhalation for 4
weeks and stimulated with LPS whereas a nonsignificant increase in IL-6 secretion was observed
(Cohen etal.. 19981.
Compared to the control group, exposure to Cr(VI) (0.36 mg/m3) by inhalation for 28 days
had no effect on spontaneous O2- and H2O2 production in the presence or absence of IFN-y at
This document is a draft for review purposes only and does not constitute Agency policy.
3-216 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
4 weeks, but increased opsonized zymosan-stimulated O2-, and decreased H2O2 production
stimulated by opsonized zymosan in the presence of IFN-y fCohen etal.. 19981. Cr(VI) had no effect
on LPS-stimulated nitric oxide (NO) production at 4 weeks but reduced NO production stimulated
by IFN-y at 4 weeks; the authors did not make statistical comparisons between the LPS-stimulated
andIFN-y-stimulatedgroups (Cohenetal.. 19981.
Overall, Cr(VI) exposure had no effect on NK cell activity, MLR, and anti-CD3 stimulation of
lymphocytes in three high confidence drinking water studies (NTP. 2006a. b, 20051. Other studies
provide some evidence for effects on mitogen-stimulated splenocyte proliferation, reactive oxygen
species production, and phagocytic activity. However, data supporting effects on mitogen-
stimulated splenocyte proliferation come from three low confidence studies (Shrivastava et al..
2005b: Snyder andValle. 1991: Glaser etal.. 19851. Data supporting effects on phagocytosis are
limited to two low (Shrivastava et al.. 2005b: Glaser etal.. 19851 and one medium confidence studies
flohansson et al.. 1986bl whereas data on reactive oxygen species are limited to only one low
confidence study (Cohen etal.. 19981. Consequently, additional studies are necessary to better
understand the potential effect of Cr(VI) on these endpoints, particularly studies that more
thoroughly document exposure conditions, exposure dose, group size, data processing, and
attrition.
Supporting immune system data
Immune organ pathology
No gross pathological changes were reported in six medium or high confidence NTP oral
studies where rats or mice were exposed to Cr(VI) for 28 days to 2 years (NTP. 2008. 2007. 2006a.
b, 2005.1996a) and one medium confidence chronic inhalation study that included a 12-month
recovery period f Glaser etal.. 19861. In one medium confidence drinking water study in male
Wistar rats of unknown age exposed to Cr(VI) (20 mg/kg-day) for up to 135 days, evaluation of the
thymus (day 90) revealed structural changes including decreased epithelial reticular cells and
physiologically important associations between these cells and T cells, potentially leading to
functional impairment of the central immune system (Karaulov etal.. 20191. In the same study,
structural effects including an increased B-zone and a decreased T-zone were observed in the
spleen across all timepoints (45, 90, and 135 days). Although the specific type of lymph node was
not reported, lymph node size was increased and was attributed to changes in cellular elements
including reticulocytes and lymphocytes.
Although unlikely to be an indicator of impaired immune function, infiltration of histiocytes
(macrophages) was observed in liver, small intestine, and mesenteric and pancreatic lymph nodes
in rats and mice in two high confidence NTP studies at oral exposure durations up to 2 years (NTP.
2008. 20071. In damaged tissues, infiltrated macrophages display functions such as modulation of
inflammatory cells, removal of damaged tissues/cellular debris, and antigen presentation, as well as
fibrogenic stimulation fYamate etal.. 20161. Histiocytic infiltrates were characterized by study
authors as small, individual clusters and sometimes as syncytia of histiocytes that were large
This document is a draft for review purposes only and does not constitute Agency policy.
3-217 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
(approximately 20-80 microns in diameter) and had pale, lightly eosinophilic, faintly stippled
cytoplasm and single, small, peripheral, dark basophilic nuclei. This finding was distinct from the
histopathological finding of chronic inflammation in the liver that NTP characterized as small,
randomly scattered aggregates of macrophages, lymphocytes, and neutrophils. Dose-related
findings of histiocytic infiltration were also observed in the lung following inhalation exposure (Kim
etal.. 2004: Glaser etal.. 1990: Glaser etal.. 1986: lohansson etal.. 1986b: Tohansson et al.. 1986a)
(see Section 3.2.1.2). The NTP authors (NTP. 2008. 2007) noted that the biological significance of
the histiocytic cellular infiltrates is unknown but suggested this finding may indicate phagocytosis
of an insoluble chemical precipitate. However, it is important to acknowledge that activated
macrophages can also damage tissue by secreting cytotoxic factors indicative of an innate
inflammatory response and create an inflammatory environment (Yamate etal.. 2016).
Overall, one medium confidence oral study (Karaulov etal.. 2019) reported structural
changes in the thymus and spleen and cellular content of lymph nodes after 90 days. Cr(VI)
exposure had no effect on spleen or thymus pathology in six medium or high confidence oral studies
and one medium confidence inhalation study (28-day or 90 days with a recovery period).
Immunoglobulin levels
Short-term, subchronic and chronic inhalation exposures to Cr(VI) (0.025, 0.05, and 0.1
mg/m3) did not alter total serum immunoglobulin levels in one low confidence study performed in
male Wistar rats (Glaser etal.. 1990). However, in a medium confidence study by the same authors,
Glaser etal. f 19851 observed a dose-dependent increase in serum immunoglobulins in male rats
following inhalation exposure for 90 days (0.025-0.10 mg/m3); serum immunoglobulin levels
returned to baseline when rats were exposed to a higher Cr(VI) concentration (i.e., 0.20 mg/m3).
Although quantitative data were not reported, serum immunoglobulins were also reported to
decrease following inhalation exposure to Cr(VI) (as chromium oxide) for 6 months (0.1 mg/m3) in
a low confidence study f Glaser etal.. 19861. Changes in total serum immunoglobulin levels alone
are not considered sensitive enough to detect mild to moderate immunotoxicity or predictive
enough to identify immunotoxicants (IPCS. 2012: Luster etal.. 1993: Luster etal.. 19921. However,
in combination with data on measures of immune function, these results may provide supporting
evidence of immunomodulation.
Immune organ weight
Absolute thymus weight was unchanged in two high confidence NTP studies performed in
female Sprague-Dawley and F344 rats exposed to a range of Cr(VI) concentrations (5-180 mg/L) in
drinking water for 28 days fNTP. 2006a. b). However, absolute thymus weight was decreased in
one high confidence NTP study performed in male B6C3F1 and am3-C57BL/6 mice exposed to
Cr(VI) (90 mg/L, high dose group only) in drinking water for 3 months fNTP. 20071. When
evaluated using a higher concentration, the absolute thymus weight was unchanged in one high
confidence NTP study performed in male and female mice and rats (B6C3F1, BALB/c, and F344/N)
exposed to a range of Cr(VI) concentrations (20-350 mg/L) in drinking water for 3 months fNTP.
This document is a draft for review purposes only and does not constitute Agency policy.
3-218 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
20071. In one medium confidence study, absolute thymus weight decreased in rats exposed to
Cr(VI) (20 mg/kg-d) in drinking water for up to 135 days fKaraulov etal.. 20191.
NTP (20051 reported a decrease in relative spleen weight in female mice exposed to
11 mg/L Cr(VI) in drinking water for 28 days; these findings were not replicated when the study
authors repeated the experiment. Relative spleen weight was not affected by exposure to Cr(VI) in
drinking water for 28 days in other NTP studies (NTP. 2006a. b). However, relative spleen weight
was also decreased in F344/N rats and am3-C57B mice subchronically exposed to Cr(VI) at
concentrations >90 mg/L in drinking water fNTP. 20071. Similarly, in a low confidence study,
relative spleen weight decreased gradually over time in mice exposed to Cr(VI) (14.8 mg/kg) in
drinking water for nine weeks (Shrivastava etal.. 2005b). In one medium confidence study,
absolute spleen weight and body weight decreased in rats exposed to Cr(VI) (20 mg/kg-d) in
drinking water for up to 135 days (Karaulov et al.. 20191. Relative spleen weight was significantly
increased in a medium confidence drinking water study following exposure to 50 mg/L Cr(VI) for 7
days, but not following 21 days exposure to 200 mg/L (Tin etal.. 20161. These results suggest the
effect may recover with time or there may be a nonmonotonic dose-response. In a medium
confidence inhalation study, relative spleen weight increased following Cr(VI) exposure for 28 or
90 days at concentrations >0.050 mg/m3 (Glaser etal.. 19851. However, this effect was not
observed in a low confidence chronic inhalation study using the same model system when the study
design incorporated a 12-month recovery period following an 18-month exposure (Glaser etal..
19861. Spleen weight was also reported to be unaffected in rats exposed by inhalation to higher
Cr(VI) concentrations (i.e., 0.20-1.25 mg/m3) for 13 weeks (Kim etal.. 20041.
Overall, Cr(VI) exposure only reduced absolute thymus weight in a single drinking water
study and the effect was not observed in a second study exposing the same strain of mice to a
broader and higher range of doses. However, absolute thymus weight was decreased in a longer
duration drinking water study. Depending on the concentration of Cr(VI) tested, the exposure
duration, and the route of administration, Cr(VI) exposure was shown to either have no effect, to
increase, or to decrease relative spleen weight Recognizing that immune organ weights are often
confounded by stress responses, results of immune organ weight is of limited utility for immune
organ pathology.
WBC counts and differentials (spleen, thymus; bone marrow)
No effects on the absolute number of splenic WBCs (total), or lymphocyte subtypes were
observed in two high confidence NTP studies performed in female Sprague-Dawley rats and
B6C3F1 mice exposed to Cr(VI) in drinking water for 28 days (5-180 mg/L) (NTP. 2006b. 20051. In
another high confidence 28-day drinking water study in female F344 rats, the total number of
splenic WBCs was also unaffected, but the numbers of NK cells and macrophages were increased at
doses of 4 mg/kg-d and 0.5 mg/kg-d Cr(VI), respectively (NTP. 2006a). In both instances, the
observed increase in cell number was only detected at 1 out of 4 dose levels tested in the study and
always at levels that fell within the range of concentrations tested in the other two drinking water
This document is a draft for review purposes only and does not constitute Agency policy.
3-219 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
studies fNTP. 2006b. 20051. In one medium confidence drinking water study in male Wistar rats
exposed to Cr(VI) (20 mg/kg-day) for up to 135 days, the absolute number of splenic T cells and T
helper cells was decreased on days 90 and 135, but the relative values were unaffected for these
timepoints (Karaulov etal.. 20191. The absolute and relative number of CD8+ T cells were
decreased in the spleens of rats on day 90, but not at any other timepoint The absolute number of
splenic karyocytes, and myeloid cells decreased, and effects on the absolute number of plasma cells
either increased or decreased depending on the timepoint (Karaulov etal.. 20191. In the same
study, the absolute number of thymocytes decreased. The absolute number of bone marrow
myeloid cells, neutrophils, lymphocytes, and karyocytes were increased at the 135-day timepoint
(Karaulovetal.. 20191.
Overall, recognizing that splenic WBC counts and differentials have only been evaluated in a
small number of drinking water studies, the effects of Cr(VI) exposure on splenic WBC and splenic
WBC differentials varied across studies. These differences in outcome may relate to experimental
design parameters including rodent species, test article concentration and study duration. Based
on a single medium quality study, Cr(VI) exposure has the potential to alter the number of
thymocytes and bone marrow cells. Additional studies are needed to better understand the effects
of Cr(VI) on WBC counts and differentials.
WBC counts (hematology)
Dose-related increases in total WBCs and some WBC types were reported in F344/N rats
exposed to Cr(VI) for up to 14 weeks fNTP. 2008. 20071: however, WBC counts were similar to the
control at 6 months and decreased at 12 months of exposure (NTP. 20081. Increased total WBC
number was also reported in one medium confidence inhalation study performed in rats for 30 and
90 days but the effect reversed in animals exposed for 90 days followed by a 30-day observation
period fGlaser etal.. 19901. In a low confidence drinking water study in Swiss mice, total WBC
number and some WBC types decreased after 3 weeks of Cr(VI) exposure fShrivastava et al..
2005a).
No effects on WBCs (total or differentials) were observed in mice in three high confidence
NTP studies (NTP. 2007. 2005.1996a). in mice or rats in seven medium confidence studies (Krim et
al.. 2013: NTP. 2006a. b; Kim etal.. 2004: Glaser etal.. 1986). and in rats in two low confidence
studies fShrivastava et al.. 2005a: Glaser etal.. 1985). These short-term, subchronic, and chronic
exposure studies included oral exposures via the diet (approximately 1-50 mg/kg-d Cr(VI)) fNTP.
1996a), oral gavage (5.3 mg/kg Cr(VI)) fKrim etal.. 20131. and drinking water (approximately
0.5-10 mg/kg-d Cr(VI)) (NTP. 2007. 2006a. b, 2005) as well as inhalation exposures
(0.025-1.25 mg/m3) (Kim etal.. 2004: Glaser etal.. 1986: Glaser etal.. 1985) in rats and mice.
Overall, evidence for Cr(VI)-related changes in WBC count is inconsistent.
This document is a draft for review purposes only and does not constitute Agency policy.
3-220 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
3.2.6.3. Mechanistic and Supporting Evidence
Available evidence from studies of apical immune endpoints in human and animals suggests
that Cr(VI) exposure may have the capacity to modulate the immune system by stimulating some
elements of immune responses (antibody response, mitogen-stimulated lymphocyte proliferation,
total WBC counts (hematology), complement levels) and suppressing others (pathogen clearance).
The sections that follow describe mechanistic data from studies of mechanistic endpoints that
might inform immune effects derived from human ex vivo and in vivo animal investigations.
Summary tables of mechanistic studies are presented in Appendix C.2.5.2.
Immune modulation
Several lines of mechanistic information support the conclusion that Cr(VI) exposure may
have the potential to modulate the immune system. For organizational purposes, available
mechanistic and supporting evidence was organized into effect categories of key characteristics
common to immunotoxicants; these studies are summarized in Appendix Table C-37.
Effects on immune cell differentiation or activation
Alterations in dendric cell maturation and T cell activation could impact antigen
presentation, a process central to the development of adaptive immune responses. In human
monocyte-derived dendritic cells in vitro, exposure to Cr(VI) increased expression of dendritic cell
maturation marker CD86 but had no effect on expression of CD83 fToebak etal.. 20061. Cr(VI)
exposure decreased anti-CD3/anti-CD28-stimulated expression of T cell activation markers CD69
and CD25 in primary mouse T cells (Dai etal.. 2017).
Effects on immune effector cell function
Effector functions of innate (i.e., myeloid cell-mediated phagocytosis, cytokine production,
and respiratory burst; natural killer cell function) and acquired (i.e., plasma cells and antibody
production, helper T cells and cytokine production, cytotoxic T cell function) immunity cells can be
altered by xenobiotic exposure. The 28-day NTP drinking water studies in rats and mice (reviewed
above, under "Antibody responses") showed no effect on serum titers of total IgM antibodies
specific for two different T cell-dependent antigens (NTP. 2006a. b, 2005). However, in an
additional study where Cr(VI) was administered by a route of administration that did not meet
PECO criteria, serum titers specific for T-l bacteriophage, a T cell-dependent antigen, were reduced
(Figoni and Treagan. 1975). In this study, female Sprague Dawley rats immunized with E. coli
bacteriophage T-l were administered Cr(VI) by subcutaneous injection (4.3 mg/kg Cr(VI)) for up to
44 days. The degree of antibody suppression observed in this study correlated with exposure
duration, which extended longer than the NTP drinking water exposure studies. Differences in
pharmacokinetics due to the different exposure scenarios complicate our ability to compare the
results of the two studies. Nonetheless, this study provides additional evidence that exposure to
Cr(VI) has the potential modulate immune responses.
This document is a draft for review purposes only and does not constitute Agency policy.
3-221 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Phagocytosis is important in both innate and adaptive immune responses by removing
pathogens and debris and as a key event in antigen presentation. The available animal studies
(reviewed above, under "Ex vivo WBC function") reported inconsistent effects of Cr(VI) exposure
on phagocytic activity (i.e., increased, decreased, or no effect) in alveolar macrophages (Johansson
etal.. 1986b: Glaser etal.. 1985) and decreased activity in splenic macrophages (Shrivastava et al..
2005b). In vitro studies were more consistent in demonstrating that exposure to Cr(VI) decreased
phagocytic activity of human PMNs isolated from workers exposed to Cr(VI) (Mignini et al.. 2009).
bovine alveolar macrophages fHooftman etal.. 19881. mouse peritoneal macrophages (Christensen
etal.. 19921. and mouse RAW264.7 macrophages fBadding et al.. 20141. However, only two of these
studies measured cell viability to take into account a potential role for cytotoxicity as a causative
factor (Badding etal.. 2014: Hooftman etal.. 1988). Additional in vivo and in vitro studies would
help to better understand the effects of Cr(VI) exposure on phagocytic activity.
Other in vitro studies reported diminished activity in important effector cell functions
including IgG production (Borella and Bargellini. 1993). cell mobility (Christensen et al.. 1992). and
NK cell degranulation fDai etal.. 20171. Pokeweed mitogen-stimulated IgG production by human
primary lymphocytes was reduced by Cr(VI) exposure fBorella and Bargellini. 19931. Random cell
migration was decreased in stimulated mouse primary peritoneal macrophages (Christensen etal..
1992). Activation of T cells stimulated by anti-CD3 and expression of CDal07a, a marker for NK cell
degranulation, was reduced in mouse splenocytes following Cr(VI) exposure (Dai etal.. 2017).
In general, although conflicting evidence was reported in the three in vivo animal studies
identified, Cr(VI) exposure consistently decreased immune effector cell function in vitro. However,
caution should be taken when interpreting these data, since only the studies by Badding et al.
f20141. Dai etal. f20171. and Hooftman et al. f!9881 evaluated cell viability as a potential causative
factor for observed effects following exposure to Cr(VI).
Effects on immune cell proliferation
As discussed in the section "Ex vivo WBC function" above, the effect of Cr(VI) exposure on
spleen cell proliferation ex vivo has been investigated using three approaches: mitogen stimulation,
anti-CD3 ± anti-CD28 stimulation, and the MLR. Exposure to Cr(VI) in vivo increased spleen cell
proliferation in rats and mice in the presence of ConA, a T cell mitogen (Shrivastava etal.. 2005b:
Snyder and Valle. 1991: Glaser etal.. 1985). Consistent with this finding, ConA-induced spleen cell
proliferation was increased when lymphocytes collected from Cr(VI) exposed workers were
cultured in the presence of Cr(VI) in vitro (Mignini et al.. 2009). Furthermore, in vitro exposure to
Cr(VI) increased activation by ConA in human lymphocytes, but decreased activation when
exposure was to a higher dose fMignini etal.. 20091. Snyder and Valle f!9911 reported inhibition of
in vitro ConA-stimulated proliferation, whereas Mignini etal. f20041 reported no effect. Spleen cell
proliferation was also investigated using PHA. Addition of Cr(VI) to lymphocytes cultured from
exposed workers lead to an increase in proliferation stimulated by PHA (Mignini etal.. 2009).
When exposed to Cr(VI) in vitro, the proliferation response was biphasic in PHA-stimulated human
This document is a draft for review purposes only and does not constitute Agency policy.
3-222 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
primary lymphocytes fMignini etal.. 2009: Borella and Bargellini. 19931. The effect of in vivo
exposure to Cr(VI) on spleen cell proliferation stimulated by LPS has only been investigated in a
single report (Snyder and Valle. 19911 (see Ex vivo WBC function). In that study, the low dose of
LPS (100 mg/L), but not the high dose (200 mg/L), decreased rat splenic lymphocyte proliferation.
LPS-induced spleen cell proliferation was also decreased in lymphocytes cultured in vitro in the
presence of Cr(VI) fMignini et al.. 20091.
Cr(VI) exposure had no effect on anti-CD3 spleen cell proliferation in three rodent studies
fNTP. 2006a. b, 20051. In contrast, exposure to Cr(VI) in vitro decreased anti-CD3 and
anti-CD3/anti-CD28 stimulated primary human lymphocyte proliferation fDai etal.. 2017: Akbar et
al.. 20111.
In vivo studies showed no effect of Cr(VI) exposure on MLR (NTP. 2005: Snyder and Valle.
19911. However, MLR was increased when splenocytes collected from Cr(VI)-exposed rats were
exposed to additional Cr(VI) in vitro fSnyder and Valle. 19911. When the only source of Cr(VI)
exposure was in vitro, either no effect or a stimulatory effect on MLR was observed (Snvder and
Valle. 19911. Recognizing that these the in vitro studies performed by were part of an investigation
fSnvder and Valle. 19911 using the same study design parameters (i.e., rat strain, exposure
duration, Cr(VI) concentration, stimulator), the discrepancy may be attributable to low study
replication.
Effects on communication between immune cells
Complement levels
One low confidence cross-sectional study investigated the effects of Cr(VI) exposure on
complement levels (Table 3-32). In that study, exposure to Cr(VI) increased levels of complement
C3 (mean: 0.91 ± 0.13 g/L unexposed, 1.20 ± 0.24 g/L exposed) and C4 (mean: 0.23 ± 0.05 g/L
unexposed, 0.32 ± 0.07 g/L exposed) in serum fOian etal.. 20131. Serum complement levels
increased two- to threefold above baseline are associated clinically with infection or acute
inflammation fRitchie etal.. 20041. But even subtle increases in baseline complement C3 and C4
are associated with other inflammatory markers and have been identified as a risk factor for
disorders associated with systemic inflammation, including cardiometabolic disease (Hertle etal..
2012: Engstrom etal.. 2007b: Engstrom et al.. 2007a: Engstrom etal.. 20051.
Mitogen-stimulated cytokine secretion
Effects of in vivo Cr(VI) exposure on mitogen-induced cytokine secretion by isolated cells in
vitro was evaluated in two medium confidence studies with ConA (Karaulov etal.. 20191 or LPS
(Cohen etal.. 19981. A single in vivo study observed increased secretion of TNF-a and IL-6 in the
serum of LPS challenged mice (Tin etal.. 20161. There are no in vitro studies available assessing the
effects of Cr(VI) exposure on ConA-stimulated cytokine secretion.
This document is a draft for review purposes only and does not constitute Agency policy.
3-223 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Cytokine measurements in biological media
Twenty-one studies investigated the effects of Cr(VI) on immune cell communication (see
Appendix Tables C-37 and C-38). A primary mechanism of communication for cells of the immune
system is through production and release of cytokines, which are low molecular weight
glycoproteins involved in immune responses and are commonly classified as pro-inflammatory
(i.e., immune stimulating) or anti-inflammatory (i.e., immunosuppressive). In practice, however,
the distinction between the classes of cytokines is not clear cut Interpretation of cytokine data
collected from biological medium is challenging because, depending on context, the same cytokine
can have either activating or suppressing effects on a particular cell type (Nature. 2019).
Furthermore, reduction in the level of a pro-inflammatory cytokine can have an anti-inflammatory
effect and vice versa. The effects of Cr(VI) exposure on levels of 30 cytokines (i.e., IL-la, IL-lb, IL-2,
IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-12, IL-13, IL-17, IL-17A, TNF-a, IFN-y, IFN-a, MIP-2,
CXCL10, CXCL11, CCL5, CCL17, CCL18, CCL20, CCL22, eotaxin, G-CSF, GM-CSF, MCP-1, andMIPla)
have been investigated. These studies include cytokine measurements conducted following in vivo
and in vitro exposures to Cr(VI) in human and animal models. Generally, the specific cytokines
measured included in each study varied, making interpretation of consistency for a given cytokine
difficult Interpretation is further hampered by the mix of responses reported for the same
cytokine. Irrespective, the available data suggest that Cr(VI) exposure has the potential to alter
levels of some cytokines, potentially disrupting the regulatory balance that dictates normal immune
system function. While the predictive value of cytokine levels for hazard assessment is unclear, the
observed alterations in cytokine levels do add to the weight of the evidence evaluation of Cr(VI) and
its potential to modulate the immune system.
Vascular cell adhesion molecule 1 (VCAM-1), endothelial-leukocyte adhesion molecule 1
(ELAM-1), and intracellular adhesion molecule 1 (ICAM-1) play an important role in endothelial
transmigration, the process whereby immune cells enter tissues. Expression of these important
proteins is up-regulated by certain cytokines (e.g., IL-1, TNF-a). Mignini et al. (2009) reported no
effect of Cr(VI) exposure had no effect on levels of these proteins.
Cr(VI) exposure had no effect on E-rosetting by human lymphocytes collected from exposed
workers and treated with additional Cr(VI) in vitro. E-rosetting occurs when human T cells
spontaneously bind to sheep red blood cells, a process that involves CD2 (i.e., the E-rosette
receptor), which plays an important role in T cell activation.
Allergic hypersensitivity
Hypersensitivity responses are the result of an over-reaction of the immune system.
Allergic hypersensitivity to Cr(VI) is generally observed following occupational exposure (Hedberg.
2018). Hypersensitivity reactions are organized into four different classes, Type I, II, III, and IV
fMurphv and Weaver. 20161. There are only a few anecdotal case reports and a small number of
animal studies associating Cr(VI) with Type I hypersensitivity (antibody mediated) responses that
cause allergic asthma fATSDR. 2012: Ban etal.. 2010: Fernandez-Nieto etal.. 2006: OSHA. 2006:
This document is a draft for review purposes only and does not constitute Agency policy.
3-224 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
Bright etal.. 1997: Olaguibel and Basomba. 19891: however, there is strong and compelling
evidence that Cr(VI) causes Type IV hypersensitivity responses. Type IV hypersensitivity responses
are mediated by T cells and are responsible for allergic contact dermatitis (ACD) resulting from
dermal exposure. As described in the protocol (Appendix A), a review of the evidence for
Cr(VI)-induced ACD is not included in this toxicological review because the scope of the Cr(VI) IRIS
assessment is comprised of potential health effects by the inhalation and oral routes of exposure.
Consequently, Cr(VI)-induced ACD was not comprehensively reviewed but was considered as
supporting evidence for the effects of Cr(VI) exposure on the immune system. The strongest
evidence for Cr(VI) Type IV hypersensitivity reactions comes from dermal patch testing in humans
(ATSDR. 2012: OSHA. 20061. Human clinical evidence of Type IV hypersensitivity is supported by
data from in vivo and ex vivo investigations performed in Guinea pigs (Wang etal.. 2010a: Ikarashi
etal.. 1996: Helmbold etal.. 1993: Saloga etal.. 1988: Christensen etal.. 1984: Parker etal.. 1984:
Tirova etal.. 1983: Siegenthaler etal.. 1983: Lindberg etal.. 1982: Turk and Parker. 1977: Miyamoto
etal.. 1975: Schneeberger and Forck. 19741 and mice (Lindemann et al.. 2008: Mandervelt etal..
1997: Basketter et al.. 1994: Ikarashi etal.. 1992: Vreeburg etal.. 1991: Kimber etal.. 1990: Mor et
al.. 1988: Lischka. 19711.
3.2.6.4. Integration of Evidence
Overall, the evidence suggests that Cr(VI) may cause immune effects in humans. Cr(VI)
may modulate the immune system through both stimulatory and suppressive actions. This
conclusion is primarily based on coherent evidence of effects on ex vivo WBC function across
human and animal studies, antibody responses to T cell-dependent antigen measured in animals,
and reduction in host resistance to bacterial infection reported in animal studies. However,
confidence in the evidence was reduced because some of the studies are low confidence and
reported findings often differed across studies. Integrated evidence of immune system effects of
Cr(VI) exposure from human, animal, and mechanistic studies is summarized in an evidence profile
table (Table 3-38).
The evidence of an association between Cr(VI) exposure and immunotoxicological effects in
humans is slight. The available studies are low confidence. Data obtained from supporting immune
system studies lack consistency across studies and across endpoints within studies. However, there
is some evidence from the most informative studies (i.e., ex vivo WBC function) that Cr(VI) has the
potential to stimulate at least some aspects of immune function. In addition, the large evidence
base demonstrating that exposure to Cr(VI) can induce allergic hypersensitivity responses in
humans further supports this conclusion (Hedberg. 2018: ATSDR. 20121.
Evidence from animal toxicology studies and supportive mechanistic data from in vivo and
in vitro studies provide slight evidence that Cr(VI) has both stimulatory and suppressive effects on
the immune system. Cr(VI) exposure increased antibody responses to T cell-dependent antigen
(i.e., sheep red blood cells), and effects on this critical function of the immune system were
observed in mice exposed orally and in rats exposed orally or by inhalation (NTP. 2006a. 2005:
This document is a draft for review purposes only and does not constitute Agency policy.
3-225 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
Glaser etal.. 19851. The body of evidence in support of this effect is small, but the findings are
supported by evidence from some studies of increases in ex vivo WBC function fShrivastava et al..
2005b: Cohen etal.. 1998: Snyder and Valle. 1991: Glaser etal.. 19851. WBC numbers (NTP. 2008.
2007: Glaser etal.. 19901. and total immunoglobulin levels following in vivo Cr(VI) exposure (Glaser
etal.. 19851. Some mechanistic evidence has demonstrated an increased response to antigenic
stimuli in one-way mixed lymphocyte cultures when splenocytes collected from Cr(VI)-exposed
rats were exposed to additional Cr(VI) in vitro (Snyder and Valle. 19911 and increased
mitogen-stimulated spleen cell proliferation with in vitro Cr(VI) exposure fMignini etal.. 2009:
Borella and Bargellini. 19931. Data demonstrating that exposure to Cr(VI) can result in allergic
hypersensitivity responses bolster these findings (ATSDR. 20121.
There is also evidence of an effect on host resistance, with short-term inhalation exposure
decreasing in situ clearance of bacteria from the lungs of Cr(VI)-exposed rats (Cohen etal.. 2010:
Cohen etal.. 20061. The host resistance model used for these studies is designed to evaluate local
pathogen clearance by alveolar macrophages. While the effect cannot be directly attributed to a
defect in phagocytosis, lung clearance of inhaled iron oxide was reduced in rats exposed to Cr(VI)
by the inhalation route f Glaser etal.. 19851. Furthermore, phagocytic activity of PMNs collected
from exposed workers (Mignini etal.. 20091 and splenic macrophages collected from mice exposed
to Cr(VI) in drinking water was reduced fShrivastava et al.. 2005bl. and several in vitro mechanistic
studies showed decreased phagocytic activity by human primary PMNs (Mignini et al.. 20091.
bovine alveolar macrophages fHooftman etal.. 19881. mouse peritoneal macrophages (Christensen
etal.. 19921. and mouse RAW264.7 macrophages (Badding et al.. 20141. Cr(VI) exposure also
impaired the mobility of mouse alveolar macrophages f Christensen et al.. 19921. Together, these
findings suggest that Cr(VI) can alter key functions of cells of the innate immune system, but
additional studies would be useful for identifying the most relevant exposure contexts and the
overall impact of these effects on immunity.
It is not without precedent for a single chemical to exert both stimulatory and suppressive
effects on various immune parameters flPCS. 20121. Exposure-related stimulation of the immune
system might increase susceptibility to allergic disease or autoimmunity and can include
exaggerated or inappropriately prolonged inflammatory responses associated with systemic
chronic inflammation, which can increase risk of developing other serious health conditions such as
cardiometabolic disease or cancer (Furman et al.. 2019: IPCS. 20121. In addition, because
continuous, uncontrolled immune stimulation represents a disruption of the homeostatic processes
required to maintain a balanced immune response, stimulation of the immune system may be
accompanied by immunosuppression, potentially altering host resistance as was observed here in a
limited number of studies. Additional studies are necessary to better understand the effects of
Cr (VI) exposure on the immune system, particularly with respect to studies of host resistance.
This document is a draft for review purposes only and does not constitute Agency policy.
3-226 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-38. Evidence profile table for immune effects
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
Evidence from studies of exposed humans
©OO
EX VIVO WBC
FUNCTION
Low confidence:
Mignini et al. (2004)
Mignini et al. (2009)
Increased lymphocyte proliferation
induced by two different T cell
mitogens but not by a B cell mitogen
in two low confidence studies.
No effect on phagocytosis by PMNs or
NK cell activity in one low confidence
study.
• Coherent
response with
two different T
cell mitogens
• Consistent ex
vivo proliferative
responses to T
cell mitogens in
rats and mice
(see Mechanistic
evidence and
supplemental
information
below)
• Low
confidence
studies
©oo
Slight
Although coherent
changes in T cell
mitogen-induced
lymphocyte
proliferation, WBC
counts, and some
WBC populations and
immunoglobulin
levels were reported,
available data were
inconsistent and
derived from low
confidence studies.
The evidence suggests
that Cr(VI) may cause
immune modulation in
humans given sufficient
exposure conditions
based on:
Slight evidence from low
confidence cross-
sectional studies of
workers with known risk
of Cr(VI) exposure
showing increased ex vivo
WBC function
(i.e., stimulated
proliferative responses to
T cell mitogens).
WBC COUNTS
Low confidence:
Boscolo et al. (1997)
Sazakli et al. (2014)
Wang et al. (2012a)
Kuo and Wu (2002)
Mignini et al. (2004)
Mignini et al. (2009)
A positive association with white
blood cell counts was observed in 1/3
studies, while an inverse association
was also observed in 1/3 studies.
• No factors noted
• Unexplained
inconsistency
in WBC
counts across
studies,
although
some degree
of variability
in these
Slight evidence from high,
medium, and low
confidence studies in
animals demonstrating
stimulatory effects on
antibody response, ex
vivo WBC function, WBC
number, and Ig levels and
suppressive effects on
host resistance.
This document is a draft for review purposes only and does not constitute Agency policy.
3-227 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Supportive mechanistic
evidence from animal in
vivo, ex vivo, and in vitro
models demonstrating
the potential for multiple
mechanisms of immune
system toxicity.
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
measures is
expected
• Low
confidence
studies
WBC SUBPOPULATIONS
Low confidence:
Boscolo et al. (1997)
Kuo and Wu (2002)
Mignini et al. (2004)
Mignini et al. (2009)
Tanigawa et al. (1998)
Decreased CD4+ cell number in
workers (2 of 2 studies) and in
exposed and unexposed smokers and
nonsmokers (1 of 1 studies).
Decreased CD8+ cell number in
workers (1 of 2 studies) and in
exposed smokers (1 of 1 study).
• Consistent
findings
regarding CD4+
subpopulations
in three studies
• Low
confidence
studies
IMMUNOGLOBULIN
LEVELS
Low confidence:
Boscolo et al. (1997)
Qian et al. (2013)
Verschoor et al. (1988)
A consistent stimulatory effect on
serum levels of IgA and IgM was
reported in two studies whereas
effects on IgG were inconsistent in
three studies.
• Coherent
findings
regarding serum
IgA and IgM
levels in two
studies
• Unexplained
inconsistency
in IgG levels
• Low
confidence
studies
Evidence from animal studies
ANTIBODY RESPONSES
High confidence:
NTP (2005)
NTP (2006b)
NTP (2006a)
Low confidence:
Glaser et al. (1985)
Increased IgM antibody-forming cell
responses was associated with
exposures in three high confidence
drinking water studies (one lacked
statistical significance) and one low
confidence inhalation study; the
effect was not internally reproducible
in one high confidence study.
• Consistency
across studies
performed in
rats and mice
following
• Antibody
response was
inconsistent
in high
confidence
studies
©oo
Slight
Cr(VI) induced
changes in the most
meaningful
immunological
endpoints
This document is a draft for review purposes only and does not constitute Agency policy.
3-228 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
exposure via two
different routes
• Mostly high
confidence
studies
(i.e., antibody
response, host
resistance and ex vivo
WBC function) and
endpoints that provide
supporting evidence
(i.e., immune organ
weight,
immunoglobulin
levels, and WBC
counts).
HOST RESISTANCE
Medium confidence:
Cohen et al. (2006)
Cohen et al. (2010)
Exposure to Cr(VI) compounds with
high and low solubility was associated
with decreased in situ bacterial
clearance in the lung.
• Consistent
findings
regarding in situ
bacterial
clearance
• Mechanistic
evidence for
immune effector
function
provides
biological
plausibility
• Medium
confidence
studies
• No factors
noted
EX VIVO WBC
FUNCTION
High confidence:
NTP (2005)
NTP (2006b)
NTP (2006a)
Medium confidence:
Effects on phagocytosis by
macrophages were observed in two
low confidence studies.
Increased mitogen-induced
proliferative response (ConA)
observed in three low confidence
studies.
• Coherent
findings of
effects on
phagocytosis and
mitogen-induced
proliferative
responses across
• Low
confidence
studies
This document is a draft for review purposes only and does not constitute Agency policy.
3-229 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
Cohen et al. (1998)
Johansson et al. (1986b)
Low confidence:
Glaser et al. (1985)
Snvder and Valle (1991)
Shrivastava et al.
(2005b)
Effects on superoxide and hydrogen
peroxide production (with zymosan)
and nitric oxide (with IFN-y) in one
low confidence study.
No effects on NK cell activity, the
MLR, or anti-CD3-stimulated spleen
cell proliferation were observed in
three high confidence short-term
drinking water studies performed in
rats and mice.
animal in vivo
studies
• Consistency with
effects observed
in animal cells in
vitro
IMMUNE ORGAN
PATHOLOGY
High confidence:
NTP (1996a)
NTP (2005)
NTP (2007)
NTP(2008)
Medium confidence:
Karaulov et al. (2019)
NTP (2006b)
NTP (2006a)
Glaser et al. (1986)
Microscopic structural effects of the
rat thymus and spleen were reported
in one medium confidence oral
exposure study.
No effects on immune organ gross
pathology were reported in six
medium or high confidence NTP oral
exposure studies and one medium
confidence inhalation study.
• Medium
confidence study
• No factors
noted
IMMUNOGLOBULIN
LEVELS - TOTAL
Medium confidence:
Glaser et al. (1985)
Low confidence:
Glaser et al. (1986)
Glaser et al. (1990)
A dose-dependent increase in serum
immunoglobulins following inhalation
exposure for 90 days (0.025-0.10
mg/m3 Cr(VI)) in a medium confidence
study; effects not observed at higher
Cr(VI) concentrations (i.e., 0.20
mg/m3). Two other low confidence
inhalation studies of short-term,
subchronic, and chronic exposure
• Dose-response
gradient
• Medium
confidence study
• Coherent with
effects on
• No factors
noted
This document is a draft for review purposes only and does not constitute Agency policy.
3-230 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
duration reported no alterations or
decreases in total serum
immunoglobulin levels.
antibody
responses
IMMUNE ORGAN
WEIGHT
High confidence:
NTP (2005)
NTP (2006b)
NTP (2006a)
NTP(2007)
Medium confidence:
Glaser et al. (1985)
Jin et al. (2016)
Karaulov et al. (2019)
Kim et al. (2004)
Low confidence:
Glaser et al. (1986)
Shrivastava et al.
(2005b)
Following drinking water exposures,
treatment-related decreases in
absolute thymus weight was observed
in one high confidence subchronic
exposure study in mice and one
medium confidence long-term study
in rats but not in three other high
confidence subchronic studies in mice
and rats.
Effects of Cr(VI) exposure on absolute
and relative spleen weight were
observed in some studies, but not
others. Results do not consistently
correlate with dose, route of
administration, exposure duration or
species.
• No factors noted
• Unexplained
inconsistency
across study
types
WBC COUNTS
High confidence:
NTP (2005)
NTP (2006b)
NTP (2006a)
Medium confidence:
Karaulov et al. (2019)
The absolute number of macrophages
and percentage NK cells were
increased in one high confidence
study, and the absolute and/or
relative number of several
lymphocyte subtype populations
varied by timepoint in one medium
confidence study. No effects on
lymphocyte subtypes in two high
confidence studies. No effects on the
absolute number of splenic WBCs in
three high confidence and one
• No factors noted
• Unexplained
inconsistency
across study
types,
although
some degree
of variability
in these
measures is
expected
This document is a draft for review purposes only and does not constitute Agency policy.
3-231 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
medium confidence drinking water
studies.
WBC COUNTS
(HEMATOLOGY)
High confidence:
NTP (1996a)
NTP (2007)
NTP (2008)
Medium confidence:
NTP (2005)
NTP (2006b)
NTP (2006a)
Glaser et al. (1986)
Glaser et al. (1990)
Kim et al. (2004)
Krim et al. (2013)
Wang et al. (2015)
Low confidence:
Glaser et al. (1985)
Shrivastava et al.
(2005a)
Effects on WBC counts were reported
in one of five studies performed in
mice (4 drinking water, 1 diet) and
four of nine studies performed in rats
(2 drinking water, 2 inhalation).
These effects were observed more
often in studies of exposure durations
<90 days, but this was not a
consistent finding.
• No factors noted
• Unexplained
inconsistency
across study
types
ANTIGEN-SPECIFIC
ANTIBODY TITER
High confidence:
NTP (2005)
NTP (2006b)
NTP (2006a)
No effect on serum titer of total IgM
antibodies specific for two different T
cell-dependent antigens in three high
confidence NTP studies of 28-day
exposures in drinking water.
• No factors
• No factors
This document is a draft for review purposes only and does not constitute Agency policy.
3-232 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Mechanistic evidence and supplemental information
Biological events or
pathways
Summary of key findings and interpretation
Judgment(s) and
rationale
Effects on immune
effector function of
specific cell types
Interpretation: Consistent in vitro evidence that Cr(VI) decreases phagocytosis by
macrophages. Phagocytosis is important in both innate and adaptive immune
responses by removing pathogens and debris and also as a key event in antigen
presentation.
Key findings:
• Reduced phagocytosis in PMNs collected from exposed workers and treated
with additional Cr(VI) in vitro (Mignini et al., 2009)
• Consistent in vitro evidence of decreased phagocytic activity by
macrophages (splenic, alveolar) harvested from murine and bovine sources
and bv the RAW2643.7 macrophage cell line (Badding et al., 2014;
Christensen et al., 1992; Hooftman et al., 1988)
• Exposure to Cr(VI) in vitro had no effect on random migration in mouse
primary peritoneal macrophages exposed to non-cytotoxic concentrations of
Cr(VI) (Christensen et al., 1992)
• In vitro evidence of decreased IgG production in human primary
Ivmphocvtes (Borella and Bargellini, 1993)
• In vitro evidence of decreased cell surface expression of CD107a, a marker
for NK cell degranulation {Dai, 2017, 4453480
Biologically plausible
observations of effects
on phagocytosis in
vitro that are
consistent with the in
vivo findings in
animals, and coherent
with other immune
effects (e.g., mitogen-
induced proliferative
responses).
T cell proliferative
responses are
consistent among ex
vivo evidence from
exposed humans and
animals, but less
consistent among in
vitro exposures and
lack coherence with
direction of effects on
T cell activation, likely
This document is a draft for review purposes only and does not constitute Agency policy.
3-233 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Effects on immune cell
differentiation or
activation
Interpretation: In vitro exposure studies indicate Cr(VI) has the potential to affect
activation of dendritic cells, which serve an important role in innate and adaptive
immune responses. Cr(VI) exposure decreased T cell activation in vitro.
Key findings:
• Dose-dependently increased expression of cell surface marker CD86
(dendritic cell maturation marker) but no effect on CD83 (activation marker
for antigen presenting cells) expression in human monocyte-derived
dendritic cells in vitro (Toebak et al., 2006)
• Decreased activation of T cells stimulated with anti-CD3 and anti-CD28 in
vitro (Dai et al., 2017)
due to differing
experimental
conditions.
There is not enough
information for the
remaining mechanistic
evidence base (e.g.,
for effects on immune
cell communication)
to make a
determination.
Effects on immune cell
proliferation
Interpretation: Consistent with findings in human occupational exposure studies,
Cr(VI) exposure in vitro has the potential to alter T cell proliferative responses.
Key findings:
• Ex vivo spleen cell proliferation increased in rats and mice in response to T
cell mitogen ConA (Shrivastava et al., 2005b; Snvder and Valle, 1991; Glaser
et al., 1985)
• Some in vitro evidence of potential alterations in proliferative responses to T
cell mitogens PHA and ConA in cells from humans and rats exposed to Cr(VI),
where lower concentrations appear more likelv to induce an effect (Mignini
et al., 2009; Mignini et al., 2004; Borella and Bargellini, 1993; Snvder and
Valle, 1991)
• In vitro evidence that Cr(VI) exposure decreases proliferation of lymphocytes
stimulated bv anti-CD3/anti-CD28 (Dai et al., 2017; Akbar et al., 2011)
Limitations:
• Inconsistent evidence for effects on the MLR between ex vivo and in vitro
exposures (Snvder and Valle, 1991)
This document is a draft for review purposes only and does not constitute Agency policy.
3-234 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
• Difficulty in comparing results due to differing test conditions
• Some inconsistencies in proliferative responses between ex vivo and in vitro
exposures
Effects on immune cell
communication
Interpretation: Cr(VI) increases complement factors, which may indicate recent
infection or development of inflammatory disease.
Key findings:
• Cr(VI) exposure increased complement factors C3 and C4 in one low
confidence study of serum collected from workers occupationally exposed to
Cr(VI) (Qian et al., 2013)
This document is a draft for review purposes only and does not constitute Agency policy.
3-235 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
3.2.7. Male Reproductive Effects
The male reproductive system consists of internal and external organs that are regulated by
a balanced interplay of hormones from the hypothalamus-pituitary-gonadal (HPG) axis. The
development and function of the male reproductive system can be affected by toxicants that
directly reach reproductive tissues or by the disruption of hormone activity at any point along the
HPG axis fCreasy and Chapin. 20181. Common endpoints evaluated to gauge male reproductive
toxicity include semen parameters and male reproductive hormone levels in human studies, as well
as changes in fertility and fecundity, sperm parameters, reproductive system organ weights and
histopathology, structural abnormalities, and changes in sexual behavior in animal studies fU.S.
EPA. 1996b). This section considers reproductive effects in males exposed to Cr(VI) at any life
stage, including exposures occurring preconception and for all stages of development. This is in
accordance with EPA's Framework for Assessing Health Risk of Environmental Exposures to Children
fU.S. EPA. 2006dl. which recommends that evidence for organ system toxicity be considered for all
life stages in order to identify populations or life stages that may be more susceptible to chemical-
induced toxicity. Reproductive effects resulting from developmental exposures are also considered
in the "Developmental effects" section.
3.2.7.1. Human Evidence
Study evaluation summary
Table 3-39 summarizes the human epidemiology studies considered in the evaluation of the
effects of Cr(VI) on the male reproductive system. These consist of six cross-sectional occupational
studies conducted among workers in two industries with known risk of exposure to Cr(VI) in
Denmark and India. They include five studies of stainless-steel welders (Danadevi etal.. 2003:
Hiollund etal.. 1998: Bonde and Ernst. 1992: Bonde. 1990: Telnes and Knudsen. 19881. Two of
these studies were performed on the same cohort of workers using different analyses (Bonde and
Ernst. 1992: Bonde. 19901 and therefore were evaluated as a single study (Table 3-39), although
there are differences in the analyses and results between the two studies as discussed below. In
addition, one study conducted in chromium (III) sulfate production workers was considered
relevant due to evidence of exposure to Cr(VI) among the workers that could be explained by the
location of the chromium sulfate operations within a chromate production plant (Kumar etal..
20051. The study evaluations resulted in one medium confidence study (Bonde and Ernst. 1992:
Bonde. 19901 and four low confidence studies (Kumar etal.. 2005: Danadevi etal.. 2003: Hiollund et
al.. 1998: Telnes and Knudsen. 19881. Results of the male reproductive effects in these studies—
specifically, semen parameters and serum reproductive hormones—are summarized in Table 3-40.
In all studies, the primary exposure route was inhalation of Cr(VI) in air. Air concentrations
of Cr(VI) (mean [SD] = 3.6 [2.8] ug/m3) were reported in one cohort of stainless-steel welders
(Bonde. 19901 (Protocol, Section 6, Appendix A for more on consideration of welding studies in this
This document is a draft for review purposes only and does not constitute Agency policy.
3-236 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicological Review ofHexavalent Chromium
assessment). No other studies of Cr(VI) exposure and male reproductive effects in humans
reported air concentrations of Cr(VI) or total chromium.
Lack of air concentration measurements in all studies except one fBonde. 1990] contributed
to concerns about potential bias from exposure misclassification. These concerns were mitigated
when job-based dichotomous exposure categories were consistent with reported concentrations of
chromium in urine (Bonde and Ernst. 1992) or blood (Danadevi etal.. 2003). In one study of
workers on a site where both trivalent and hexavalent chromate products were produced (Kumar
etal.. 20051. it is unclear whether blood concentrations of chromium reflected Cr(VI) specifically;
however the high rate of nasal perforation among the workers in this study indicate a history of
Cr(VI) exposure. Other study evaluation concerns included potential residual confounding (Kumar
etal.. 2005: Telnes and Knudsen. 1988) and concerns about outcome measurement (Kumar etal..
2005: Hiollundetal.. 19981.
Table 3-39. Summary of human studies for Cr(VI) male reproductive effects
and overall confidence classification [high (H), medium (M), low (L)] by
outcome.3 Click to see interactive data graphic for rating rationales.
Author (year)
Industry
Location
Study Design
Sperm
Parameters
Hormones
Bonde and Ernst (1992), Bonde
(1990)
SS Welding
Denmark
Cohort
(occupational)
M
M
Danadevi et al. (2003)
SS Welding
India
Cohort
(occupational)
L
-
Hiollund et al. (1998)
SS Welding
Denmark
Cohort
(occupational)
L
Ub
Jelnes and Knudsen (1988)
SS Welding
Denmark
Cohort
(occupational)
Lc
-
Kumar et al. (2005)
Chromium
sulfated
India
Cohort
(occupational)
L
-
SS = stainless steel.
aln addition to these included studies, two additional studies reported male reproductive outcomes that met PECO
criteria but were found to be uninformative at the study evaluation stage: Tielemans et al. (1999): Li et al. (2001).
bAnalysis of hormone concentrations in Hiollund et al. (1998) compared all welders to referents (no analysis
comparing SS welders to referents) and therefore was found to be uninformative for this outcome.
cUninformative for motility only. Low confidence for other sperm parameters.
dThough chromium sulfate is trivalent, there is evidence of simultaneous or recent exposure to Cr(VI) in the
exposed group.
Synthesis of evidence in humans
Semen parameters
Four core endpoints were considered in the evaluation of the effects of exposure to Cr(VI)
on semen parameters: volume, concentration, morphology and motility. A key consideration when
This document is a draft for review purposes only and does not constitute Agency policy.
3-237 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
assessing the quality of outcome measurements for these endpoints was the window of time
following collection of samples fRadke etal.. 20191. Other quality control procedures related to
collection and processing of samples were considered, including but not limited to collection of
more than one sample from the same individual and abstinence period duration before sample
collection.
One medium confidence study reported mild decreases in semen volume and sperm motility
in stainless-steel welders (mean [SD] = 2.4 [1.1] mL; 51.0 [15.7] percent motile) compared to
nonwelders (mean [SD] = 3.1 [1.3] mL; 57.7 [14.8] percent motile), but no differences in sperm
concentration or morphology between these two groups fBonde. 19901. In the same cohort,
comparisons of sperm concentration, morphology and motility among three exposure groups
characterized by urine chromium measurements were indicative of an effect but did not reach
statistical significance (Bonde and Ernst. 19921 (Table 3-40). Both air concentrations and urine
chromium concentration were higher among stainless-steel welders compared to mild steel
welders or nonwelders, and these exposure data lent confidence to the exposure characterization of
participants in both analyses. These data also reveal some exposure misclassification in both
analyses that may have decreased study sensitivity. The detection of a statistically significant
decrease in volume and motility despite limits to study sensitivity increased confidence in the
findings of this study.
This document is a draft for review purposes only and does not constitute Agency policy.
3-238 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-40. Summary of results from human studies of Cr(VI) male reproductive effects
Semen Parameters
Hormones
Study
Exposure
Conf.
Result Format
N
Vol (mL)
Concentration
(million/mL)
% Normal forms
% Motile
T
(nmol/L)
LH
(IU/L)
FSH
(IU/L)
Danish
Welders3
Bonde (1990)
SS welding v.
ref
M
Mean (SD) and
p-value
Exp: 35
Ref: 54
Exp: 2.4 (1.1)
Ref: 3.1 (1.3)
p < 0.05
Exp: 58.4 (16.7)
Ref: 58.6 (23.9)
NS
Exp: 65.8 (15.7)
Ref: 66.7 (17.1)
NS
Exp: 51.0 (15.7)
Ref: 57.7 (14.8)
p < 0.05
Exp: 17.3 (5.8)
Ref: 21.2 (8.0)
p < 0.05
Exp: 6.1 (2.4)
Ref: 7.2 (2.7)
NS
Exp: 4.4 (5.1)
Ref: 4.9 (2.8)
NS
Danish
Welders3
Bonde and
Ernst (1992)
3-level3
M
Unadjusted
regression
beta;
Mean (SD) and
p-value
Low: 60
Med: 24
High: 23
P: 0.2
Low: 2.9 (1.3)
Med: 3.0 (1.6)
High: 3.2 (1.4)
NS
|3: -1.5
Low: 54.5 (26.9)
Med: 62.8 (21.7)
High: 50.7 (20.9)
NS
|3: -1.6
Low: 65.8 (17.8)
Med: 61.0 (17.1)
High: 56.8 (20.5)
NS
|3: -0.5
Low: 55.2 (14.6)
Med: 54.8 (11.9)
High: 51.6 (16.4)
NS
|3: -1.2
Low: 21.0 (7.8)
Med: 18.7 (7.3)
High: 16.4 (5.6)
NS
|3: -0.1
Low: 6.8 (3.0)
Med: 6.8 (2.4)
High: 6.7 (2.8)
NS
P: -0.1
Low: 4.7 (2.9)
Med: 5.0 (2.6)
High: 4.5 (2.2)
NS
Danadevi et
al. (2003)
Welders'5 v.
Controls
L
Mean (SD) and
p-value
Exp: 57
Ref: 57
Exp: 2.4 (0.5)
Ref: 2.5 (0.5)
NS
Exp: 14.5 (24.0)
Ref: 62.8 (43.7)
p < 0.001
Exp: 37.0 (14.3)
Ref: 69.0 (8.0)
p < 0.001
% IMMOTILE:
Exp: 31.0 (16.6)
Ref: 12.4 (7.0)
p < 0.001
Hiollund et
al. (1998)c
SS welding v.
ref
L
Median (crude
and adj)
Exp: NR
Ref: NR
(29, 205
respectively
at
enrollment)
Exp: 56.0 (crude)
Exp: 65.5 (adj)
Ref: 50.0 (crude)
Ref: 46.4 (adj)
Uninformative
for this
endpoint
Uninformative
for this
endpoint
Uninformative
for this
endpoint
Jelnes and
Knudsen
(1988)
SS welding v.
ref
L
Median and p-
value
Exp: 75-77
Ref: 67-68
Exp: 3.0
Ref: 3.0
p = 0.50-0.70
Exp: 58.6
Ref: 58.2
p = 0.95
Exp: 36.0
Ref: 36.5
p = 0.70-0.90
Uninformative for
this endpoint
Kumar et al.
(2005)
Chromate
workers v. ref
L
Mean (SD) and
p-value
Exp: 54-61
Ref: 10-15
Exp: 2.67 (0.964)
Ref: 2.54 (0.641)
p = NR
Exp: 49.57 (36.3)
Ref: 43.75 (29.9)
p = NS
Exp: 27.87 (2.5)
Ref: 45.10 (13.4)
p < 0.005
Exp: 73.77 (11.79)
Ref: 76.89 (5.76)
p = NS
NS = not significant, as reported the study; exact p-values are included in the table when available. NR = not reported.
aTwo analyses in the same cohort (Bonde and Ernst, 1992; Bonde, 1990). Exposure variable characterization by job category (supported by air concentration data) in 1990
analysis, exposure characterization by urine chromium (supported by job history) in 1992 analysis.
bWelding type not specified, blood chromium higher in welders compared to referents, coexposure to Ni.
cStainless steel and non-stainless-steel welders were pooled in the analysis of the male hormone concentrations; therefore, the hormone analysis from this study was considered
uninformative.
This document is a draft for review purposes only and does not constitute Agency policy.
3-239 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
Of the four other studies considered, all four measured sperm concentration and were
judged to be low confidence for that outcome fKumar etal.. 2005: Danadevi etal.. 2003: Hiollund et
al.. 1998: lelnes and Knudsen. 19881 (Table 3-39). Three of the studies also measured semen
volume and sperm morphology and motility and were judged to be low confidence for all outcomes
(Kumar etal.. 2005: Danadevi etal.. 2003: lelnes and Knudsen. 19881. with the exception of one
study that was uninformative for motility (lelnes and Knudsen. 19881. One low confidence study
reported a statistically significant decrease in sperm concentration in occupationally exposed
groups compared to referents fDanadevi etal.. 20031. One low confidence study reported an
increase in sperm concentration in stainless-steel workers that may have been explained by a
shorter period of abstinence before sample collection in that group compared to the referent
(Hiollund etal.. 19981: in addition, sperm samples in this study were frozen before analysis raising
concerns about the quality of the outcome measurements (WHO. 20101. In all other studies,
samples were not frozen and were analyzed within a short time of collection. Also consistent with
the findings of the medium confidence study discussed previously, two low confidence studies that
investigated sperm motility reported decreases in the exposed group compared to referents. These
findings were statistically significant in one of the studies fDanadevi etal.. 20031. but did not reach
significance in the other study (Kumar et al.. 20051. Both studies also reported changes in
morphology (i.e., decreased percent normal forms) in the occupationally exposed group compared
to referents (Kumar etal.. 2005: Danadevi etal.. 20031. One low confidence study reported no
effect of Cr(VI) exposure on volume, concentration, or morphology, but limited description of the
methodology impeded the study evaluation (lelnes and Knudsen. 19881.
Consistency in the findings across several of the five studies, including one medium
confidence study, suggests that Cr(VI) exposure by the inhalation route at levels observed in
occupational settings may impact semen quality. Sperm concentration, morphology, and motility
were decreased in exposed groups compared to referents in three of the five studies (Kumar etal..
2005: Danadevi etal.. 2003: Bonde. 19901. and these results were statistically significant for
concentration fDanadevi etal.. 20031. morphology fDanadevi etal.. 20031. and motility (Kumaret
al.. 2005: D anadevi et al.. 2 0 0 3: Bonde. 19901 despite the likely impact of exposure misclassification
on study sensitivity. Evidence of a dose-response pattern to effects of Cr(VI) exposure on
concentration, morphology, and motility provides further supporting evidence of a relationship
between such exposures and semen quality (Bonde and Ernst. 19921. Two studies reported
findings that were inconsistent with the other studies, but these may be explained by study
limitations such as the use of frozen sperm samples or study quality issues (Hiollund etal.. 1998:
lelnes and Knudsen. 19881. Results for semen volume were inconsistent across studies and within
analyses in the same cohort, suggesting that Cr(VI) exposure is not associated with this specific
endpoint
This document is a draft for review purposes only and does not constitute Agency policy.
3-240 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Male hormones
The male reproductive hormones testosterone, luteinizing hormone (LH), and follicle
stimulating hormone (FSH) were considered when assessing the effects of exposure to Cr(VI) on
male hormones in humans (Radke etal.. 20191. The effects of Cr(VI) on other male reproductive
hormones that potentially serve as endpoints for the evaluation of reproductive effects, especially
for onset of puberty, such as sex hormone binding globulin and dehydroepidandrosterone (DHEA),
were not investigated in the studies included in this analysis. A key consideration in the evaluation
of studies of male hormones is the timing of sample collection; morning collection is recommended
to account for diurnal variation in serum testosterone concentrations.
One medium confidence study described in two publications was considered in the
evaluation of the effect of Cr(VI) exposure on male hormones fBonde and Ernst. 1992: Bonde.
19901. A study by Hiollund et al. (19981 reported male hormones in welders and nonwelders, but
the results were considered uninformative and are not discussed further because stainless-steel and
non-stainless-steel welders were pooled in this analysis. The medium confidence study reported
significantly decreased serum testosterone concentration in stainless-steel welders (mean
[SD] = 17.3 [5.8] nmol/L) compared with nonwelders (mean [SD] = 21.2 [8.0] nmol/L) (Table 3-40)
(Bonde. 19901. A dose-response dependent decrease in serum testosterone was also reported in
the same cohort, though results of that analysis did not reach statistical significance (Bonde and
Ernst. 19921. In the same study, decreased serum LH and FSH concentrations were also reported in
stainless-steel welders compared to nonwelders, but these results did not reach statistical
significance. In an alternative analysis, serum LH and FSH decreased with increased exposure to
Cr(VI) characterized by urine concentration, but evidence of a dose-response trend was not as
strong for these endpoints as it was for testosterone. As discussed previously in the section on
semen parameters, data on air concentrations, urine chromium concentration and job history
support the categorization of exposure in the medium confidence study; however, these data also
point to exposure misclassification in both analyses that may have decreased study sensitivity. The
detection of a statistically significant exposure-dependent decrease in testosterone as well as
nonsignificant decreases in all three hormones measured (testosterone, LH, and FSH) despite
limitations in study sensitivity increased confidence in the findings of this study.
Due to the small number of studies that assessed the relationship between Cr(VI) exposure
and male reproductive hormones, consistency could not be assessed. However, evidence from two
separate analyses in a medium confidence study indicates that exposure may impact serum
concentrations of testosterone and these results are coherent with evidence for semen parameters
described separately. Evidence of a relationship between Cr(VI) and serum concentration of LH
and FSH was not as strong for these hormones as it was for testosterone. The medium confidence
study found a small inverse association between Cr(VI) exposure and serum LH and FSH that was
not statistically significant and was not supported by the findings of the low confidence study.
This document is a draft for review purposes only and does not constitute Agency policy.
3-241 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
3.2.7.2. Animal Evidence
Study evaluation summary
Table 3-41 summarizes the animal toxicology studies considered in the evaluation of the
effects of Cr(VI) on the male reproductive system. These consist of a two-generation reproductive
study with dietary exposure using NTP's Reproductive Assessment by Continuous Breeding (RACES)
protocol (NTP. 19971: subchronic oral exposure studies using diet (NTP. 1996a. b), drinking water
fNTP. 2007: Bataineh etal.. 1997: Elbetieha and Al-Hamood. 19971. or gavage/unspecified oral
administration (Bashandv etal.. 2021: Marat etal.. 2018: Rasool etal.. 2014: Yousef etal.. 20061:
short-term exposure studies using drinking water (Wang etal.. 20151 or unspecified oral
administration (Kim etal.. 20121: a chronic inhalation exposure study (Glaser etal.. 19861:
subchronic inhalation exposure studies (Kim etal.. 2004: Glaser etal.. 19851: and studies that
evaluated F1 males that had been exposed during gestation fNavin etal.. 2021: Shobana etal.. 2020:
Zheng etal.. 2018: Al-Hamood etal.. 19981 or during gestation and lactation (Kumar etal.. 20171.
The three available inhalation studies only reported information on male gonad weights fKim etal..
2004: Glaser etal.. 19861 or histopathology (Kim etal.. 2004: Glaser etal.. 19851. whereas the
available oral exposure studies provided more specific measurements of male reproductive
function including fertility, sperm parameters, hormone levels, and sexual behavior. The report by
NTP (2007) included two separate studies: a 3-month study in rats (F344/N) and mice (B6C3F1),
and a second 3-month comparative study using three strains of mice (B6C3F1, BALB/c, C57BL-6).
NTP's RACB study (NTP. 1997) and subchronic exposure studies (NTP. 2007.1996a. b) and
the gestational exposure study by Zheng etal. f 20181 were well-reported and well-designed to
evaluate reproductive outcomes and were therefore rated as high confidence for almost all
reported outcomes (Table 3-41). The subchronic study by Bashandv et al. (2021) was rated as
medium confidence for the evaluation of sperm parameters and hormone levels, but low confidence
for organ weights and histopathology due to reporting limitations for those endpoints. The
remaining studies had reporting limitations and other substantial concerns raised during study
evaluation and were rated as low confidence across all outcomes. Endpoint-specific concerns
identified during study evaluation are discussed in the respective sections below. Three of the low
confidence studies (Al-Hamood etal.. 1998: Bataineh et al.. 1997: Elbetieha and Al-Hamood. 1997)
exposed animals to high concentrations (350-1770 mg/L) of Cr(VI) in drinking water, which was
considered a potential confounding variable as it is not possible to determine whether reproductive
effects may have been exacerbated by reduced water consumption and/or systemic toxicity; for
instance, drinking water concentrations of 350 mg/L Cr(VI) have been associated in rats with
decreased water consumption and site of contact toxicity (80 and 100% incidence of ulcers in the
glandular stomach of males and females, respectively) fNTP. 20071.
This document is a draft for review purposes only and does not constitute Agency policy.
3-242 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-41. Summary of included animal studies for Cr(VI) male reproductive
effects and overall confidence classification [high (H), medium (M), low (L)] by
outcome.3 Click to see interactive data graphic for rating rationales.
Author (year)
Exposure
route
Species (strain)
Exposure life stage and
duration
Fertility, Fecundity
Sperm evaluation
Histopathology
Hormones
Organ weights
Sexual behavior
Anogenital distance
Al-Hamood et al.
(1998)
Drinking
water
Mouse (BALBC)
F1 offspring; GD 12-
PND20
L
"
"
"
L
"
"
Bashandv et al.
(2021)
Gavage
Rat (Wistar)
Adult males; 8 weeks
-
M
L
M
L
"
"
Bataineh et al.
(1997)
Drinking
water
Rat (Sprague-
Dawley)
Adult males; 12 weeks
L
L
"
-
"
L
L
L
"
Elbetieha and Al-
Hamood (1997)
Drinking
water
Mouse (Swiss)
Adult males; 12 weeks
"
-
"
-
"
Glaser et al.
(1986)
Inhalation
Rat (Wistar)
Adult males; 18 months
-
"
-
"
L
-
"
Glaser et al.
(1985)
Inhalation
Rat (Wistar)
Adult males; 28 or 90
days
-
"
L
"
-
-
"
Kim et al. (2004)
Inhalation
Rat (Sprague-
Dawley)
Adult males; 90 days
-
"
L
"
L
-
"
Kim et al. (2012)
Oral
(unspecified)
Rat (Sprague-
Dawley)
Adult males; 6 days
-
L
L
-
"
L
L
-
"
Kumar et al.
(2017)
Drinking
water
Rat (Wistar)
F1 offspring; GD 9-14
-
L
L
-
L
Marat et al. (2018)
Gavage
Rat (white outbred)
Adult males; 60 days
L
Navin et al. (2021)
Drinking
water
Rat (Wistar)
F1 offspring; GD 9-14
-
-
L
L
L
-
-
NTP (1996a)
Diet
Mouse (BALBC)
Adult males; 3, 6, or 9
weeks
-
H
-
-
H
-
-
NTP (1996b)
Diet
Rat (Sprague-
Dawley)
Adult males; 3, 6, or 9
weeks
-
H
H
-
-
H
H
-
-
NTP (1997)
Diet
Mouse (BALBC)
Reproductive
Assessment by
Continuous Breeding
(F0 to F2)
H
NTP (2007)
Drinking
water
Study 1: Rat
(F344/N), Mouse
(B6C3F1)
Study 2: Mouse
(B6C3F1, BALB/c,
C57BL-6)
Study 1: Adult males; 3
months
Study 2: Adult males, 3
months
H
M
H
Rasool et al.
(2014)
Oral
(unspecified)
Mouse (strain not
reported)
Adult males; 30 or 60
days
-
-
L
-
-
-
-
This document is a draft for review purposes only and does not constitute Agency policy.
3-243 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofHexavalent Chromium
Author (year)
Exposure
route
Species (strain)
Exposure life stage and
duration
Fertility, Fecundity
Sperm evaluation
Histopathology
r" Hormones
Organ weights
Sexual behavior
Anogenital distance
(Shobana et al.,
2020)
Drinking
water
Rat (Wistar)
Fl offspring; GD 9-14 or
GD 15-21
"
"
"
"
"
"
Wang et al. (2015)
Drinking
water
Rat (Sprague-
Dawley)
Adult males; 4 weeks
"
"
L
-
L
"
"
Yousef et al.
(2006)
Gavage
Rabbit (NZ white)
Adult males; 10 weeks
"
L
-
L
L
L
"
Zheng et al. (2018)
Gavage
Rat (Sprague-
Dawley)
Fl offspring; GD 12-21
"
-
H
H
-
-
"
GD = gestation day; PND = postnatal day.
aln addition to these included studies, there were seven animal toxicology studies reporting male reproductive
outcomes that met PECO criteria but were found to be uninformative at the study evaluation stage: Aruldhas et al.
(2006; 2005; 2004); Chowdhury and Mitra (1995); Li et al. (2001); Subramanian et al. (2006); Zabulyte et al. (2009);
and Zahid et al. (1990).
Synthesis of evidence in animals^-^
Fertility and fecundity
No effects on the ability to impregnate females (i.e., fertility parameters) were observed
across the five studies in rats or mice that evaluated this outcome. These consisted of the high
confidence RACES study in mice by NTP f!9971 that evaluated FO and F1 parental animals at oral
doses in diet ranging from 6.8-30.3 mg-kg/day Cr(VI) (FO) or 7.9-37.1 mg-kg/day Cr(VI) (Fl); two
low confidence studies that evaluated adult male rats or mice that had been exposed to 350 mg/L or
up to 1770 mg/L Cr(VI), respectively, in drinking water for 12 weeks prior to mating (Bataineh et
al.. 1997: Elbetieha and Al-Hamood. 19971: one low confidence study that evaluated adult male rats
that had been exposed to 0.353 mg/kg-day Cr(VI) via gavage for 60 days prior to mating (Marat et
al.. 20181: and one low confidence study that evaluated adult Fl male mice that had been exposed
to maternal doses of 350 mg/L Cr(VI) in drinking water during gestation and lactation f Al-Hamood
etal.. 19981. However, Elbetieha and Al-Hamood (19971 observed a statistically significant
42Data are available in HAWC for: NTP (19971 (here)
NTP f!996al There!
NTP f!996bl fherel
NTP ("20071 fmale B6C3F1 mice, male BALBC mice, male am3-C57BL/6 micel.
43For many of the oral studies presented here, it was not possible to estimate an average daily mg/kg dose
due to lack of reporting. To estimate an average daily dose, paired records of body weight and daily intake of
test article are required. This is particularly important for Cr(VI) reproductive and developmental studies,
because rapid changes in maternal body weight are expected during pregnancy, and Cr(VI) affects palatability
(which affects both Cr(VI) intake rate and body weight). Doses of Cr(VI) are presented where possible,
however many cross-study comparisons are done on the basis of mg/L Cr(VI) in drinking water.
This document is a draft for review purposes only and does not constitute Agency policy.
3-244 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
decrease in the number of implantations and viable fetuses when Cr(VI)-exposed male Swiss mice
were mated with untreated females; this effect was observed in 710 or 1410 mg/L Cr(VI) dose
groups, but not the highest dose group (1770 mg/L). Similarly, increased pre- and post-
implantation loss in rats dosed with 0.353 mg/kg-day Cr(VI) by oral gavage prior to mating was
observed by Marat etal. (20181. who reported a dominant lethal mutation frequency of 0.665 by
comparing the number of live fetuses in the Cr(VI) treatment group to the control group. No effects
on offspring viability were observed in rats or mice in other studies following paternal exposure
fAl-Hamoodetal.. 1998: Bataineh et al.. 1997: NTP. 19971. Overall, decreased fetal viability
following paternal-only exposure (indicative of dominant lethal mutations in sperm) was observed
across two studies, but interpretation is limited because these studies were considered low
confidence and the only available high confidence study failed to observe similar effects.
Sperm evaluation
No effects on sperm were observed in the high confidence subchronic exposure studies in
rats and a variety of mouse strains by NTP at oral doses ranging from 0.35-32.5 mg/kg-day Cr(VI)
in drinking water or diet (NTP. 2007.1996a. b), or in the high confidence RACES study in mice that
evaluated F0 and F1 males at doses ranging from 6.8-30.3 mg-kg/day Cr(VI) (F0) or 7.9-
37.1 mg/kg-day Cr(VI) (Fl) in diet (NTP. 19971. These studies reported multiple measurements
aimed at evaluating effects on spermatogenesis. The NTP RACES and 3-month drinking water
studies included measurements of testicular sperm head count (NTP. 2007.19971. epididymal
sperm density fNTP. 2007.19971. epididymal sperm morphology fNTP. 19971. and evaluation of
epididymal sperm motility using computer-assisted sperm motion analysis fNTP. 19971 or visual
motility analysis by two observers (NTP. 20071. Sperm from both F0 and Fl males were evaluated
in the RACB study (NTP. 19971. In the 3-month dietary exposure studies by NTP (1996a. b),
animals underwent whole-body perfusion with fixative after 3, 6, or 9 weeks of exposure and
effects on spermatogenesis were evaluated by counting the ratio of preleptotene spermatocytes
and Sertoli cell nuclei in Stage X or XI tubules, with investigators blinded to the dose group.
Perfusion fixation is considered the gold standard for histopathological evaluation of the testis
fHaschek etal.. 2009: Foley. 20011. and blinding is considered appropriate for reducing observation
bias for this relatively subjective measurement. There were no notable concerns about these
evaluations.
In contrast, one medium confidence study (Bashandv etal.. 20211 and three low confidence
studies (Kumar etal.. 2017: Kim etal.. 2012: Yousefetal.. 20061 observed exposure-related
decreases in sperm quality or quantity. These studies did not indicate whether investigators were
blinded during outcome evaluation and had additional reporting and study design concerns
identified during study evaluation. Bashandv et al. f20211 reported decreased sperm motility and
epididymal sperm counts and increased sperm abnormalities in adult rats following eight weeks of
exposure to 3.5 mg/kg-day Cr(VI) via oral gavage. Yousefetal. (20061 reported a statistically
significant decrease in packed sperm volume, sperm concentration, total sperm output, and sperm
This document is a draft for review purposes only and does not constitute Agency policy.
3-245 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
motility, and a statistically significant increase in the percentage of dead sperm in ejaculates
measured weekly from adult rabbits exposed via oral gavage to 3.6 mg/kg-day Cr(VI) for 10 weeks.
Concerns were raised about the interpretation of results because the numerical data presented by
the authors (means ± SE) appeared to be an average of weekly measurements across 10 weeks of
exposure, which is difficult to interpret. Graphical data were shown for weekly measurements, but
only as means without a measure of variance. Kumar etal. (20171 reported a statistically
significant decrease in epididymal sperm forward motility (measured visually under a microscope),
sperm viability, and sperm count in adult F1 rats that had been exposed during gestation at
maternal doses of 17.7-70.7 mg/L Cr(VI) in drinking water. These measurements were presented
as the mean of individual animals without accounting for potential litter effects, which has the
potential to overestimate statistical significance (Haseman etal.. 20011. Kim etal. f20121 reported
a statistically significant decrease in sperm head count and motility but no effect on the percentage
of abnormal sperm in adult rats exposed to 10 mg/kg-day Cr(VI) for 6 days. This short exposure
duration does not cover the duration of spermatogenesis, and therefore lacks sensitivity for
detecting potential effects on spermatogonia. Overall, although these studies report that Cr(VI)
exposure can affect sperm quality and quantity, interpretation of the low confidence studies is
limited due to the study design and reporting concerns. It is possible that differences in route of
exposure could explain why effects on sperm were observed in the study by Bashandv et al. (20211
(gavage), whereas the NTP studies (drinking water or diet) did not observe effects at equal or
higher dose levels.
Histopathology
Almost all studies that evaluated histopathological outcomes in male reproductive tissues
used conventional fixation in formalin or formaldehyde, which is not recommended for the testis
because it gives poor penetration and may cause artifacts (Hascheketal.. 2009: Foley. 20011. This
was considered a sensitivity concern and reduced the confidence in this dataset Zheng et al.
(20181 is the only study that used Bouin's solution, which is considered a preferable fixative for the
testis fCreasy and Chapin. 2018: Foley. 20011. The study by NTP f20071 reported that slides used
for histopathological evaluation were peer reviewed and the final diagnoses represents a consensus
of contractor pathologists and the NTP Pathology Working Groups, which is considered a best
practice for histopathological evaluations (Crissman et al.. 20041. None of the other studies
indicated that any steps were taken to reduce observational bias.
No dose-related lesions were observed in the testis, epididymis, prostate, or preputial gland
in the 3-month drinking water exposure studies by NTP (20071 in rats and in a variety of mouse
strains at oral doses up to 20.9 mg/kg-day Cr(VI) (Study 1 rats), 27.9 mg/kg-day Cr(VI) (Study 1
mice), or 8.7 mg/kg-day Cr(VI) (Study 2 mice). These studies by NTP f20071 were considered
medium confidence for the testicular evaluation due to the use of formalin fixative and high
confidence for other male reproductive organs. There were also no reported histopathological
changes in the gonad in the low confidence 25- or 90-day inhalation studies in rats by Glaser et al.
This document is a draft for review purposes only and does not constitute Agency policy.
3-246 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
f!9851 and Kim etal. f20041 at concentrations up to 0.2 mg/m3 Cr(VI) or 1.25 mg/m3 Cr(VI),
respectively; or in the low confidence 4-week drinking water study by Wang etal. f20151 at
concentrations up to 106.1 mg/L Cr(VI). However, these four studies exposed adult rodents and
therefore did not expose rodents during critical gestational or developmental windows.
In contrast, a high confidence gestational exposure study (Zheng etal.. 20181 and four low
confidence subchronic oral exposure studies (Bashandv etal.. 2021: Navin etal.. 2021: Kumar etal..
2017: Rasool etal.. 20141 observed histopathological changes in the testis. Zheng etal. (20181
reported altered Leydig cell distribution (increased single-cell clusters and decreased larger
clusters) and decreased Leydig cell size and cytoplasmic size in F1 male rat pups following
maternal exposure to 3-12 mg/kg-day Cr(VI) by oral gavage from GD 12-21, but no change in
Leydig cell number or proliferation. The number of Sertoli cells and the incidence of multinuclear
gonocytes in the pups was not affected. In adult F1 male rats exposed from GD 9-14, Kumar et al.
f20171 observed a statistically significant decrease in the diameter of the seminiferous tubules and
lumen, number of Sertoli cells, and testicular spermatocytes and spermatids at maternal doses of
17.7-70.7 mg/L Cr(VI) in drinking water. A study by the same group of authors fNavin etal.. 20211
similarly observed shrunken tubules with increased interstitial space and sloughing of immature
germ cells from the basal compartment into the lumen at maternal doses of 35.4 - 70.7 mg/L Cr(VI)
in drinking water. In animals exposed for subchronic durations as adults, Bashandv et al. (20211
and Rasool etal. (20141 observed damage to Leydig cells, germinal epithelium, and sperm cells in
rats exposed to 3.5 mg/kg-day Cr(VI) (gavage) and mice exposed to 1.77 mg/kg-day Cr(VI)
(unspecified method of oral administration), respectively. The studies by Zheng etal. (20181 and
Kumar etal. f 20171 provided quantitative data on the incidence of effects, whereas the other three
studies reported only qualitative findings. Data in Kumar etal. f20171 was presented as the mean
of individual animals without accounting for potential litter effects, which has the potential to
overestimate statistical significance (Haseman etal.. 20011.
Within the high confidence study by Zheng etal. (20181. the changes in Leydig cell
distribution may be coherent with the reported effects on testosterone in this study (see next
section). Histopathological changes were also coherent with effects on testosterone and sperm
parameters within low confidence studies, although the interpretation of those studies is more
limited.
Hormones
Effects on reproductive hormone levels were observed across all studies that evaluated this
outcome, which included one high confidence, one medium confidence, and four low confidence
studies. The high confidence study by Zheng etal. T20181 reported a nonmonotonic effect in which
serum testosterone was increased in F1 male rat pups following maternal exposure to 3 mg/kg-day
Cr(VI) by oral gavage from GD 12-21 but decreased in the 12 mg/kg-day Cr(VI) dose group. The
medium confidence study by Bashandv et al. (20211 reported decreased testosterone, decreased
luteinizing hormone (LH), and increased follicle stimulating hormone (FSH) in following an 8-week
This document is a draft for review purposes only and does not constitute Agency policy.
3-247 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
exposure of adult male rats to 3.5 mg/kg-day Cr(VI) by oral gavage. Three low confidence studies
by the same group of authors fNavin etal.. 2021: Shobana etal.. 2020: Kumar etal.. 20171 evaluated
F1 male rats that had been exposed during gestation via maternal drinking water and also reported
decreased serum testosterone levels. Effects on serum testosterone reached statistical significance
at a maternal dose of 17.7 mg/L Cr(VI) in males evaluated on PND 30 (Shobana etal.. 20201 and
PND 60 fNavin etal.. 20211 versus 70.7 mg/L Cr(VI) in males evaluated on PND 120 (Kumar etal..
20171. although effects on testosterone in testicular interstitial fluid at PND 120 reached statistical
significance at 17.7 mg/L Cr(VI). Shobana etal. f20201 and Navin etal. f20211 also reported
increased estrogen, decreased prolactin, and increased LH and FSH, whereas Kumar etal. f 20171
reported decreased LH and FSH. Measurements in these three studies were presented as the mean
of individual animals without accounting for potential litter effects, which has the potential to
overestimate statistical significance (Haseman etal.. 20011. Lastly, the low confidence study by
Yousefetal. f20061 reported a statistically significant decrease in plasma testosterone in rabbits
after a 12-week oral exposure to 3.6 mg/kg-day Cr(VI). Concerns about selective reporting and the
presentation of results were raised because authors stated that testosterone measurements were
performed biweekly but reported only a single mean value for serum testosterone.
These results suggest that Cr(VI) exposure has an anti-androgenic effect at higher dose
levels, although interpretation of results in the low confidence studies is limited. The high
confidence studies by NTP (2007.1997.1996a. b) did not evaluate hormone levels, so a direct
comparison with those studies is not possible; however, one mouse strain in NTP's 3-month
drinking water study observed decreased testis weight (NTP. 20071. which is considered indicative
of changes in androgen levels fFoster and Gray. 2013: Evans and Ganiam. 20111. The lack of effect
on male reproductive organ weights in the other studies by NTP suggests that there was minimal
effect on androgens on those studies.
Organ weight
Except for decreased testis weight observed in one mouse strain in the high confidence
study by NTP f20071. effects on male reproductive organ weights were only seen in low confidence
studies. The 3-month drinking water exposure study by NTP f20071 reported a statistically
significant 11% decrease in absolute testis weight in am3-C57BL/6 mice in the highest dose group
(8.7 mg/kg-day Cr(VI); n = 5/group). No effects were observed in the two other mouse strains
(B6C3F1 and BALB/c) that were tested in this study at doses up to 8.7 mg/kg-day Cr(VI), or in
F344/N rats or B6C3F1 mice at doses up to 20.9 and 27.9 mg/kg-day Cr(VI), respectively fNTP.
20071. No effects on testis or accessory reproductive organ weights were observed in the other
high confidence RACB or 3-month dietary exposure studies in mice or rats by NTP at doses ranging
from 0.35-37.1 mg/kg-day Cr(Vl) fNTP. 1997.1996a. b). There were also no effects on testis
weight in the low confidence studies by Glaser etal. (19861. Kim etal. (20041. Al-Hamoodetal.
(19981. Wang etal. (20151. or Kim etal. (20121. Kim etal. (20121 also reported no effect on
This document is a draft for review purposes only and does not constitute Agency policy.
3-248 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
epididymis weight, although the short exposure duration in this study (6 days) likely limited study
sensitivity.
In contrast, six low confidence subchronic oral exposure studies reported Cr(VI)-induced
changes in testis and accessory male reproductive organ weights. The most notable findings
consisted of a statistically significant decrease in absolute testis, seminal vesicle, and preputial
gland weights in rats after 12-week exposure to 350 mg/L Cr(VI) in drinking water (Bataineh etal..
19971: a statistically significant decrease in testis, vas deferens, epididymis, prostate, and seminal
vesicle weight (unclear whether absolute or relative to body weight) in rats after an 8-week
exposure to 3.5 mg/kg-day Cr(VI) by oral gavage fBashandv etal.. 20211: a statistically significant
decrease in relative testis and epididymis weights in rabbits after a 10-week exposure to 3.6
mg/kg-day Cr(VI) via oral gavage (Yousef etal.. 20061: decreased absolute and relative testis
weights in F1 rats that had been exposed from GD 9-14 and were evaluated on PND 60, reaching
statistical significance at 70.7 mg/L Cr(VI); and a statistically significant decrease in relative testis
weight and absolute epididymal and seminal vesicle weights in adult F1 rats that had been exposed
from GD 9-14 to maternal doses of 17.7-70.7 mg/L Cr(VI) in drinking water fKumar etal.. 20171.
The measurements by Navin etal. f20211 and Kumar etal. f 20171 were presented as the mean of
individual animals without accounting for potential litter effects, which has the potential to
overestimate statistical significance (Haseman etal.. 20011. Additionally, the 12-week drinking
water exposure study in mice by Elbetieha and Al-Hamood (19971 reported a statistically
significant decrease in relative seminal vesicle and preputial gland weight in the 1770 mg/L Cr(VI)
group, but a statistically significant increase in relative testis weight in the 710 and 1770 mg/L
Cr(VI) groups; however, the increase in relative testis weight may have been an artifact of
decreased body weight in these animals. It has been shown that testis weights are not modeled
well by an organ-to-body weight ratio because testis and body weights are not proportional (Bailey
etal.. 20041. so relative organ weights may be a less sensitive measure than absolute testis weight
Decreased body weight was reported in all four studies, including those that reported relative
decreases in organ weights.
Overall, these results suggest that male reproductive organ weights can be decreased by
Cr(VI) exposure, which is consistent with decreased androgen levels as described above. However,
interpretation of these results is limited because effects were predominantly observed in low
confidence studies and were not observed in the majority of the high confidence studies by NTP.
Effects on testis weight observed by Yousefetal. (20061. Kumar etal. (20171. and Navin etal.
(20211 are coherent with the decreased testosterone observed in these studies.
Sexual behavior
Effects on sexual behavior were observed in two low confidence subchronic oral exposure
studies, which were the only studies that evaluated this outcome. Neither of these studies reported
that any steps were taken to reduce observational bias during outcome evaluation, which is a
concern since behavior can be a relatively subjective measurement. In rats, Bataineh etal. (1997)
This document is a draft for review purposes only and does not constitute Agency policy.
3-249 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
reported a statistically significant decrease in mounts and percentage of males ejaculating, and
significant increase in ejaculation latency and post-ejaculatory interval following 12 weeks of
exposure to 350 mg/L Cr(VI) in drinking water. This assessment of sexual behavior was performed
on a separate cohort of animals than those used in the fertility assay by these authors (see earlier
section). In rabbits, Yousefetal. (20061 reported a statistically significant increase in the reaction
time to mounting following 10-week exposure to 3.6 mg/kg-day Cr(VI) by oral gavage. These
results are suggestive of effects on sexual behavior, but interpretation of the results is limited
because these studies are considered low confidence.
Anogenital distance (AGD)
The low confidence gestational exposure study by Kumar et al. (20171 reported a
dose-related decrease in AGD in F1 male rats that had been exposed during gestation from GD 9-14
to maternal doses of 17.7-70.7 mg/L Cr(VI) in drinking water. AGD was measured at multiple
timepoints between PNDs 1-30. AGD is a biomarker of androgen-dependent development, so this
effect is coherent with the decreased androgen levels observed in these animals as adults (see
earlier section). This measurement was presented as the mean of individual animals without
accounting for potential litter effects, which has the potential to overestimate statistical significance
(Haseman et al.. 20011. Overall, while this finding suggests that Cr(VI) exposure decreases AGD via
decreased androgen levels, interpretation of the results is limited because this study is considered
low confidence.
3.2.7.3. Mechanistic Evidence
The Cr(VI) literature provides evidence for potential mechanisms of Cr(VI)-induced male
reproductive toxicity; specifically, oxidative stress in male reproductive tissues, altered cell cycle
regulation and apoptosis in somatic and germ cells, alterations in steroid hormone signaling and
the hypothalamic-pituitary-gonadal (HPG) axis, and effects on Sertoli cells and the blood-testis
barrier. These studies support the biological plausibility that Cr(VI) may have the potential to act
as a male reproductive toxicant acting through several possible modes of action. Mechanistic
studies are tabulated in Appendix C.2.6 and summarized here.
The mechanistic studies reviewed here consisted of in vivo mechanistic data from several of
the included oral exposure studies discussed above (Table 3-41), as well as from intraperitoneal
(i.p.) injection studies that did not meet PECO criteria but were reviewed as informative for
mechanistic analysis. Dosing via i.p. injection is likely to result in higher tissue concentrations of
Cr(VI) compared to oral exposure due to the oral first-pass effect caused by the reduction of Cr(VI)
in the low pH environment of the stomach; less than 10-20% of an ingested dose may be absorbed
in the GI tract, and further reduction will occur in the liver prior to distribution to the rest of the
body (see Section 3.1 and Appendix C). Therefore, systemic effects are expected to be more likely
following i.p. injection or inhalation compared to oral exposure. Given their specific relevance to
the pattern of findings observed in a subset of the in vivo animal studies, in vitro studies that
This document is a draft for review purposes only and does not constitute Agency policy.
3-250 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
evaluated Leydig, Sertoli, or male germ cells were also considered within this synthesis of
mechanistic evidence.
Oxidative stress
Decreased antioxidant enzyme activities [e.g., superoxide dismutase (SOD), catalase (CAT),
glutathione peroxidase (GPx), glutathione-S-transferase (GST), glucose-6-phosphate
dehydrogenase (G-6-PDH), y-glutamyl transpeptidase (y-GT)], decreased nonenzymatic
antioxidants (metallothionein, glutathione, vitamins A, C, E), and increased lipid peroxidation
[measured as malondialdehyde (MDA) or lipid peroxidation potential] were observed in serum or
in male reproductive tissues in rodents and monkeys concurrent with apical outcomes following
oral exposure (Bashandv etal.. 2021: Shobana et al.. 2020: Rasool etal.. 2014: Kim etal.. 2012:
Subramanian etal.. 2006: Aruldhas etal.. 2005) or i.p. injection (El-Demerdash etal.. 2019:
Marouani etal.. 2015a: Hfaiedh etal.. 2014: Acharva et al.. 2006: Acharva etal.. 2004). Similar
markers of oxidative stress were observed in vitro in cultured mouse Leydig cells, Sertoli cells, or
spermatagonial stem cells fLv etal.. 2018: Das etal.. 20151. Although antioxidant levels were
generally decreased across studies, increased GST or metallothionein were observed in some cases
(Das etal.. 2015: Marouani et al.. 2 015 a: Aruldhas etal.. 2005). indicating an antioxidant response.
Several in vivo studies demonstrated that effects on sperm, testicular histopathology,
hormones, and male fecundity were attenuated following cotreatment with antioxidants (Bashandv
etal.. 2021: El-Demerdash etal.. 2019: Lv etal.. 2018: Hfaiedh etal.. 2014: Kim etal.. 2012:
Subramanian etal.. 20061. This may imply that oxidative stress is a mechanism underlying these
effects, but interpretation is difficult because antioxidants can also decrease tissue Cr(VI) levels by
stabilizing lower Cr oxidation states. For instance, Subramanian et al. (2006) reported lower
plasma Cr levels with coadministration of Vitamin C. The authors hypothesized that the protective
effect of Vitamin C may be due to enhanced conversion of Cr(VI) to Cr(III).
Cell cycle regulation and apoptosis in somatic and germ cells
There is evidence that Cr(VI) exposure alters cell cycle regulation and promotes apoptosis
in male reproductive tissues following in vivo exposure. Bashandv et al. (2021) reported increased
p53 expression and decreased DNA content in spermatogenic cells of rats exposed to 3.5 mg/kg-
day Cr(VI) for 8 weeks via oral gavage, suggesting that DNA replication was inhibited in these
animals. Increased expression of the pro-apoptotic protein BAX and increased DNA fragmentation
(measured using DNA ladders or by the biomarker y-H2AX) were observed in the testes of male
rats and mice following i.p. injection fLv etal.. 2018: Marouani et al.. 2 015 al. I.p. injection studies
have also reported degenerative histopathological changes in seminiferous tubules and
spermatogenic cells, absence of spermatocytes in the seminiferous tubules, and lower sperm counts
in rats, mice, and rabbits fEl-Demerdash etal.. 2019: Lv etal.. 2018: Acharva etal.. 2004: Behari et
al.. 19781.
This document is a draft for review purposes only and does not constitute Agency policy.
3-251 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
In vitro studies using mouse Leydig, Sertoli, or spermatagonial stem cells provided
additional evidence of the activation of intrinsic (mitochondria-dependent) apoptotic pathways,
including increased staining in the TUNEL assay, decreased mitochondrial membrane potential,
decreased BAX/BCL-2 ratio, and increased cleavage of caspases 3 and 9 in all three of these cell
types (Lv etal.. 2018: Das etal.. 20151. In vitro studies also found that biomarkers of extrinsic
apoptosis (caspase 8, Fas) were not activated, further supporting intrinsic apoptosis as the
mechanism of cell death (Lv etal.. 2018: Das etal.. 2015). It was demonstrated both in vivo and in
vitro that effects on cell cycle regulation and cell death were attenuated following cotreatment with
an antioxidant fBashandv etal.. 2021: Lv etal.. 2018: Das etal.. 20151.
A single study provides evidence of an effect of Cr(VI) on meiosis, another potential
mechanism for effects on spermatogenesis. Using a bicameral culture chamber of rat Sertoli and
germ cells, Geoffrov-Siraudin etal. (2010) observed that Cr(VI) treatment decreased the number of
late spermatocytes and round spermatids and increased the percentage of cells with alterations in
meiotic prophase.
Altered steroidogenesis and effects on the HPG axis
As described above, hormonal effects in studies meeting PECO criteria included a
nonmonotonic effect on fetal testosterone in F1 male rats (increased at the lowest dose and
decreased at the highest dose) in the high confidence study by Zheng etal. (2018) and decreased
testosterone and effects on gonadotropin levels in medium and low confidence studies in rats
fBashandv etal.. 2021: Navin etal.. 2021: Shobana etal.. 2020: Kumar etal.. 20171 and rabbits
fYousef etal.. 20061. Decreased prolactin and increased estrogen were also reported in F1 rats
(Navin etal.. 2021: Shobana et al.. 2020). Similarly, i.p. injection studies reported decreased
testosterone (El-Demerdash etal.. 2019: Hfaiedh etal.. 2014: Marouani etal.. 2012). decreasedLH,
and increased FSH (El-Demerdash etal.. 2019: Marouani etal.. 2012) in adult male rats. Several of
these studies found that hormone changes were attenuated by cotreatment with an antioxidant
(B ashandv et al.. 2 0 21: El-Demerdash etal.. 2019: Hfaiedh etal.. 2014).
Findings at the molecular level provide supporting evidence that Cr(VI) affects
steroidogenesis, with inhibition occurring at higher dose levels. In Leydig cells of F1 male rats
exposed during gestation, Zheng etal. (2018) reported a nonmonotonic effect on genes/proteins in
the steroidogenesis pathway (increased at low dose and decreased at high dose) and Navin etal.
(2021) reported decreased expression of the steroidogenesis pathway. These changes are
consistent with the effects on testosterone in these two studies, although the molecular changes
reported by Zheng etal. (2018) were less consistent and often differed between the mRNA and
protein levels. Both studies reported decreased Leydig cell LH receptor (LHR) expression at high
dose levels, and Navin etal. f20211 also reported decreased expression of Leydig cell androgen
receptor (AR), prolactin receptor (PRLR), and estrogen receptor alpha (ERa). In Sertoli cells, which
are somatic cells that support germ cell development and play a role in HPG regulation, decreased
expression of FSH receptor (FSHR) and AR was reported in F1 rats by Shobana etal. (2020) and
This document is a draft for review purposes only and does not constitute Agency policy.
3-252 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Kumar etal. f20171. whereas Zheng etal. f20181 reported a low-dose increase in FSHR mRNA
expression but no change at the high dose level. Zheng etal. f20181 also reported that secretion of
the growth factors LIF and PDGFA by Sertoli cells was increased at low doses of Cr(VI) and may
have contributed to Leydig cell stimulation at the low dose level, whereas a high dose of Cr(VI)
caused a decrease in the secretion of insulin-like growth factor-1 (IGF-1) by Sertoli cells that may
have contributed to the suppression of Leydig cell androgen production at the high dose level.
These in vivo observations are supported by an in vitro study (Das etal.. 20151. which reported that
Cr(VI) treatment decreased testosterone secretion and expression of genes in the steroidogenesis
pathway in cultured mouse Leydig cells, and decreased transcriptional expression of FSHR and AR
in cultured mouse Sertoli cells.
Another series of studies specifically suggested that the pituitary and hypothalamus were
targeted by Cr(VI). Male rats exposed to 73.05 mg/kg-day Cr(VI) for 30 days by drinking water
were found to have Cr accumulation in the pituitary and decreased serum prolactin, but no effect on
serum LH, with the same trend observed in primary rat anterior pituitary cells treated with Cr(VI)
in vitro fOuinteros etal.. 20071. A follow-up study using the same experimental design but lower
dose [11.6 mg/kg-day Cr(VI)] reported accumulation of Cr and evidence of oxidative stress in the
pituitary and hypothalamus (Nudler etal.. 2009). Oxidative stress and apoptosis were also
reported in primary anterior pituitary cells treated with Cr(VI) in vitro and were mitigated by
cotreatment with an antioxidant (Ouinteros etal.. 2008: Ouinteros etal.. 20071.
Effects on Sertoli cells and the blood-testis barrier
Several studies reported that Cr(VI) exposure impaired the functionality of Sertoli cells,
including the dynamics of the blood-testis barrier. In F1 rats exposed to Cr(VI) during gestation,
Shobana etal. (2020) reported a decrease in Sertoli cell secretory products (lactate, pyruvate,
retinoic acid, inhibin, androgen binding protein, transferrin), and both Shobana et al. (2020) and
Kumar etal. (2017) reported decreased expression of the tight junction proteins claudin-11 and
occludin. These factors can affect germ cell development and organization of the blood-testis
barrier and are coherent with the histological changes reported in the testis by Kumar etal. f20171.
In rats exposed by i.p. injection, Murthvetal. f 19911 observed leakage of Sertoli cell tight junctions
and adverse effects on late-stage spermatids using electron microscopy. In cultured mouse Sertoli
cells in vitro, Cr(VI) treatment decreased transcriptional expression of tight junction signaling
molecules (Das etal.. 2015). Comparatively, in a bicameral chamber culture of rat primary Sertoli
and germ cells that maintains the blood-testis barrier, gap junction coupling was decreased and the
gap junction protein connexin 43 was delocalized from the membrane to the cytoplasm, but
adherins and tight junction proteins were not affected fCarette etal.. 20131.
3.2.7.4. Integration of Evidence
Overall, the evidence suggests that Cr(VI) may cause male reproductive toxicity in humans.
This conclusion is based on coherent evidence of effects across human and animal studies.
This document is a draft for review purposes only and does not constitute Agency policy.
3-253 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
Decreased testosterone and decreased sperm quantity and quality were observed in both human
and animal studies; however, interpretation of this evidence was limited because most studies that
observed these effects were considered low confidence and there was inconsistency with higher
confidence studies. Integrated evidence of the male reproductive effects of Cr(VI) exposure from
human, animal, and mechanistic studies is summarized in an evidence profile table (Table 3-42).
The evidence of an association between Cr(VI) exposure and male reproductive effects in
humans is slight and indicates an inverse association between occupational exposure to Cr(VI) and
several sperm parameters (concentration, morphology, and motility) and serum testosterone
concentrations. This is largely based on a single medium confidence study in welders fBonde and
Ernst. 1992: Bonde. 1990) and supported by some coherent findings from low confidence studies.
Evidence of a dose-response pattern in these associations further supports this conclusion. Though
some results did not reach statistical significance, this may be explained by the likely impact of
exposure misclassification on study sensitivity in all available studies.
Evidence from animal toxicology studies and supportive mechanistic data from in vivo and
in vitro studies provide slight evidence that Cr(VI) is a male reproductive toxicant. Findings from
high confidence drinking water and dietary exposure studies by NTP that exposed rats or mice as
adults (NTP. 2007.1996a. b) or for multiple generations using an RACB design (NTP. 1997) indicate
that the male reproductive system is not responsive to Cr(VI)-induced toxicity following oral
exposure, with no observed effects on sperm parameters, histopathological outcomes, or male
fertility or fecundity. In contrast, a high confidence gestational exposure study in which maternal
rats were dosed by oral gavage44 reported nonmonotonic alterations in testosterone and Leydig cell
size and distribution fZheng etal.. 20181. and a medium confidence study in which adult male rats
were dosed by oral gavage reported decreased testosterone levels, adverse effects on sperm
parameters and testis histopathology, and decreased reproductive organ weights (Bashandv etal..
2021). The available low confidence developmental and subchronic oral exposure studies likewise
reported effects including decreased male fecundity (suggestive of dominant lethal mutations in
sperm), decreased sperm quantity and quality, decreased testosterone and gonadotropins,
decreased male reproductive organ weights, and altered mating behavior. The low confidence
drinking water exposure studies frequently did not provide sufficient information to support an
estimate of dose, which makes it difficult to compare the dose-response relationships with those
from the higher confidence studies. The doses (in mg/kg-d) of Cr(VI) at which effects were
observed could not be calculated for any of the low confidence drinking water studies because
drinking water consumption data was not reported, but the available information indicates that
some were higher and some were lower than doses used by NTP (both the drinking water studies
and the oral dietary studies). This makes it unlikely that the discrepancy in responses between
high and low confidence studies is due solely to a difference in the dose ranges tested. Support for
44As previously noted, oral gavage administration is likely to achieve higher systemic absorption of un-
reduced Cr(VI) than ad libitum drinking water or dietary administration.
This document is a draft for review purposes only and does not constitute Agency policy.
3-254 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofHexavalent Chromium
biological plausibility of Cr(VI)-induced male reproductive toxicity is provided by mechanistic data
demonstrating evidence of oxidative stress in male reproductive tissues, altered cell cycle
regulation and apoptosis in somatic and germ cells, altered steroid hormone signaling, and
disruption of Sertoli cells and the blood-testis barrier, although much of this evidence was derived
from i.p. injection studies and in vitro studies that have unclear relevance for other routes of
exposure.
In the only human study that provided a quantitative measure of Cr(VI) exposure (Bonde.
19901. effects were observed at air mean (SD) concentrations of 3.6 (2.8) |ig/m3; these reported
concentrations may underestimate exposure in this study population due use of a cellulose fiber
filter during sampling, which can contribute to reduction of Cr(VI) to Cr(III). In animal toxicology
studies, the observation of decreased testis weight occurred at 8.7 mg/kg-day Cr(VI) in the 3-month
drinking water study in mice by NTP (2007). and effects were observed at doses of 3-12 mg/kg-day
Cr(VI) fZheng etal.. 20181. 3.5 mg/kg-day Cr(VI) fBashandv etal.. 20211. 0.353 mg/kg-day Cr(VI)
(Marat etal.. 2018). or 3.6 mg/kg-day Cr(VI) (Yousef et al.. 2006) in oral gavage studies. For the
other drinking water studies in animals, the doses of Cr(VI) at which effects were observed could
not be calculated because drinking water consumption data was not reported. Effects were not
observed in any of the three animal studies that evaluated inhalation exposure, but those studies
did not include specific measures of male reproductive structure and function, so were considered
insensitive. There is therefore inadequate information to evaluate the extent of effects in oral
versus inhalation exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
3-255 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-42. Evidence profile table for male reproductive outcomes
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and rationale
Inferences and summary
judgment
Evidence from studies of exposed humans
©OO
SPERM PARAMETERS
Medium confidence:
Bondeetal. (1992; 1990)
Low confidence:
Danadevi et al. (2003)
Hiollund et al. (1998)
Jelnes and Knudsen
(1988)
Kumar et al. (2005)
Note: Sperm concentration was
measured in all five studies
considered; other endpoints were
measured in some but not all of
the studies.
Decreased sperm motility in 1
medium study and 2 low
confidence studies (1 statistically
significant at p < 0.001,1 no p-
value or significance reported); a
fourth study was uninformative for
this measurement.
Decreased % sperm with normal
morphology in 2 low confidence
studies (out of 4 studies), and
decreased sperm concentration in
1 low confidence study (out of 5
studies).
Decreased semen volume was
reported in 1 medium confidence
study, but no effect on volume was
reported in 3 low confidence
studies.
• Coherence in
direction of
related
parameters
across studies
• Exposure-
response
gradient in one
medium
confidence study
• Detection of
effects despite
limitations to
study sensitivity
• Mechanistic
evidence of
oxidative stress,
cell cycle
dysregulation
and impaired
Sertoli cell
function provides
biological
plausibility
• High
proportion of
low confidence
studies
©oo
Slight
Occupational (inhalation)
Cr(VI) exposure is
inversely associated with
sperm concentration,
normal sperm
morphology, sperm
motility, and serum
testosterone.
These findings are
consistent and coherent
across multiple studies
and endpoints, but
interpretation is limited
because most studies
evaluating sperm were
considered low
confidence.
Evidence of the impact of
Cr(VI) exposure on
semen volume and
serum LH and FSH
concentrations in
humans is unclear.
The evidence suggests
that Cr(VI) may cause
male reproductive toxicity
in humans given sufficient
exposure conditions.
Effects on sperm
parameters and
testosterone were
observed in both human
and animal studies.
Most human and animal
studies were considered
low confidence. Effects in
low confidence animal
studies or in high or
medium confidence
animal studies with
gavage exposures were
generally not seen in the
high confidence RACB and
subchronic studies by
NTP.
Mechanistic findings
(animals and in vitro)
provide evidence
supportive of male
reproductive toxicity.
This document is a draft for review purposes only and does not constitute Agency policy.
3-256 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
These mechanisms are
presumed relevant to
humans.
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and rationale
HORMONES
Medium confidence:
Bondeetal. (1992; 1990)
Exposure associated with
decreased serum testosterone
concentration in Danish stainless-
steel welders. Decreases in serum
LH or FSH concentrations that were
not statistically significant were
also reported.
• Exposure-
response
gradient
• Mechanistic
evidence of
alterations in
steroidogenesis
provides
biological
plausibility
• Uncertainty
about
exposure
measurements
due to multiple
factors that
impact
exposure
among
welders;
direction of
bias is likely
toward the null
Evidence from animal studies
FERTILITY AND
FECUNDITY
High confidence:
NTP (1997)
Low confidence:
Al-Hamood et al. (1998)
Bataineh et al. (1997)
Elbetieha and Al-
Hamood (1997)
Marat et al. (2018)
No effects on ability to impregnate
females.
Decreased fetal viability (indicative
of dominant lethal effects) in two
low confidence studies in rats and
mice following paternal-only
exposure; no effects on fetal
viability in other three studies.
• No factors noted
• Effects
observed only
in low
confidence
studies
©oo
Slight
Evidence of male
reproductive effects was
observed primarily in low
confidence studies
(drinking water or
gavage) and in one high
and one medium
confidence gavage study.
High confidence RACB
and subchronic studies
by NTP observed no male
reproductive effects,
aside from decreased
testis weight in one
mouse strain.
SPERM EVALUATION
High confidence:
NTP (1996a)
NTP(1996b)
NTP (1997)
NTP (2007)
No effects on sperm parameters in
four high confidence studies in rats
or mice, including an RACB study
(F0 and F1 males) and three 3-
month exposure studies.
A medium confidence study in
adult rats and low confidence
• No factors noted
• Effects
observed only
in low
confidence
studies
This document is a draft for review purposes only and does not constitute Agency policy.
3-257 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and rationale
Medium confidence:
Bashandv et al. (2021)
Low confidence:
Kim et al. (2012)
Kumar et al. (2017)
Yousef et al. (2006)
studies in rabbits and F1 rats
report decreased sperm quality
and quantity.
Evidence was insufficient
to evaluate the extent of
effects following
inhalation exposure.
HISTOPATHOLOGY
High confidence:
NTP (2007)
Zheng et al. (2018)
Low confidence:
Bashandv et al. (2021)
Kumar et al. (2017)
Navin et al. (2021)
Rasool et al. (2014)
No dose-related lesions in male
reproductive tissues in a high
confidence 3-month drinking water
study in rats and a variety of
mouse strains.
A high confidence gestational
exposure study in F1 rats reported
Leydig cell alterations.
Low confidence studies in rats and
mice observed histopathological
changes in the testis and
seminiferous tubules.
• High confidence
study showing
Leydig cell
alterations
• Dose-response
gradient
• Coherent with
effects on
testosterone
• Inconsistent
findings in high
confidence
studies
• Changes in
testis and
seminiferous
tubules only
observed in
low confidence
studies
HORMONES
High confidence:
Zheng et al. (2018)
Medium confidence:
Bashandv et al. (2021)
Low confidence:
Kumar et al. (2017)
Navin et al. (2021)
(Shobana et al., 2020)
Yousef et al. (2006)
Nonmonotonic effect on serum
testosterone in a high confidence
gestational exposure study in F1
rats.
Decreased testosterone and
effects on gonadotropins in a
medium confidence study in adult
rats and low confidence studies in
adult rabbits and F1 rats.
• High confidence
study showing
effects on serum
testosterone
• Coherent with
effects on Leydig
cells
• Mechanistic
evidence
provides
biological
plausibility
• Decreased
testosterone
and effects on
gonadotropins
only observed
in medium and
low confidence
studies
This document is a draft for review purposes only and does not constitute Agency policy.
3-258 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and rationale
ORGAN WEIGHT
High confidence:
NTP (1996a)
NTP(1996b)
NTP (1997)
NTP (2007)
Low confidence:
Al-Hamood et al. (1998)
Bashandv et al. (2021)
Bataineh et al. (1997)
Elbetieha and Al-
Hamood (1997)
Glaser et al. (1986)
Kim et al. (2004)
Kim et al. (2012)
Kumar et al. (2017)
Navin et al. (2021)
Wang et al. (2015)
Yousef et al. (2006)
Decreased testis weight in one
mouse strain in the high
confidence 3-month drinking water
studv bv NTP (2007).
Changes (typically, decrease) in
testis and accessory male
reproductive organ weights in low
confidence studies in rabbits, rats,
and mice.
No effects observed in other
mouse strains evaluated in NTP
(2007), or in anv of the remaining
studies.
• Coherent with
decreased
testosterone
within low
confidence
studies
• Unexplained
inconsistency
across high
confidence
studies
SEXUAL BEHAVIOR
Low confidence:
Bataineh et al. (1997)
Yousef et al. (2006)
Decreased mounts, increased
ejaculation latency and post-
ejaculation interval, and decreased
percentage of males ejaculating in
rats exposed as adults.
Increased reaction time to
mounting in rabbits.
• No factors noted
• Low
confidence
studies
ANOGENITAL DISTANCE
Low confidence:
Kumar et al. (2017)
Decreased AGD in developing F1
males.
• No factors noted
• Low
confidence
study
Mechanistic evidence
This document is a draft for review purposes only and does not constitute Agency policy.
3-259 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and rationale
Biological events or
pathways
Summary of key findings and interpretations
Judgments and rationale
Oxidative stress
Interpretation: In vivo and in vitro evidence of Cr(VI)-induced oxidative stress in
male reproductive tissues or in serum concurrent with effects on sperm or
testicular pathology.
Key findings:
• Across most studies, decreased antioxidant activity or expression in male
reproductive tissues or serum observed in animals exposed orally
(Bashandv et al., 2021; Shobana et al., 2020; Rasool et al., 2014; Kim et al.,
2012; Subramanian et al., 2006; Aruldhas et al., 2005) or i.p. (El-Demerdash
et al., 2019; Marouani et al., 2015a; Hfaiedh et al., 2014; Acharya et al.,
2006) and in cultured mouse Leydig, Sertoli, and spermatogonia! stem cells
(Lv et al., 2018; Das et al., 2015)
• Consistent observation of increased testicular or epididymal lipid
peroxidation in animals exposed orallv (Bashandv et al., 2021; Shobana et
al., 2020; Rasool et al., 2014; Kim et al., 2012) or i.p. (El-Demerdash et al.,
2019; Marouani et al., 2015a; Hfaiedh et al., 2014; Acharva et al., 2006;
Acharva et al., 2004), and increased reactive oxveen species in vitro (Lv et
al., 2018; Das et al., 2015)
• Cotreatment of with antioxidants mitigated effects on sperm, testicular
histopathology, male hormones, and male fecundity in Cr(VI)-exposed
animals (Bashandv et al.. 2021; Shobana et al.. 2020; El-Demerdash et al.,
2019; Lv et al., 2018; Hfaiedh et al., 2014; Kim et al., 2012; Subramanian et
al., 2006), and decreased Cr(VI)-induced apoptosis in vitro (Lv et al., 2018;
Das et al., 2015)
Observations of oxidative
stress, altered cell cycle
regulation and apoptosis,
altered steroid hormone
signaling/effects on the
HPG axis, and effects on
Sertoli cells and the
blood-testis barrier.
Oxidative stress was
concurrent with apical
outcomes in some
animal studies.
Testicular degeneration,
decreased testosterone,
and apoptosis are
mitigated by
cotreatment with
antioxidants.
Much of this evidence
was derived from i.p.
injection studies and in
vitro studies that have
unclear relevance for
other routes of exposure.
Cell cycle regulation and
apoptosis in somatic and
germ cells
Interpretation: In vivo and in vitro evidence of Cr(VI)-induced apoptosis in male
reproductive tissues.
Key findings:
• Increased p53 and decreased DNA content of spermatogenic cells after
oral gavage exposure (Bashandv et al., 2021)
This document is a draft for review purposes only and does not constitute Agency policy.
3-260 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Factors that increase Factors that
Summary of key findings certainty decrease certainty
Judgments and rationale
• In vivo expression of BAX and DNA fragmentation in testes following i.p.
injection (Lv et al., 2018; Marouani et al., 2015a)
• Degenerative changes in testis and decreased sperm counts in animals
after i.p. injection (El-Demerdash et al., 2019; Lv et al., 2018; Acharva et al.,
2004; Behari etal., 1978)
• In vitro evidence of intrinsic apoptosis (TUNEL staining, decreased
mitochondrial membrane potential, decreased BAX/BCL-2 ratio, and
increased cleavage of caspases 3 and 9) in cultured Leydig, Sertoli, and
spermatogonia! stem cells fLv et al., 2018; Das et al., 20151
• Evidence of impaired meiotic prophase in a bicameral culture chamber
model using rat primarv Sertoli and germ cells (Geoffrov-Siraudin et al.,
2010)
Altered steroid hormone
signaling and effects on
the HPG axis
Interpretation: Cr(VI) alters steroidogenesis in vivo and in vitro.
Key findings:
• Biphasic effects on testosterone in one oral exposure study in rats
(increased at lowest dose and decreased at highest dose) (Zheng et al.,
2018), and decreased testosterone and altered gonadotropin levels in
other animal studies following oral (subchronic and gestational) (Bashandv
et al., 2021; Navin et al., 2021; Shobana et al., 2020; Kumar et al., 2017;
Yousef et al., 2006) and i.p. exposures (El-Demerdash et al., 2019; Hfaiedh
et al., 2014; Marouani et al., 2012)
• Changes in expression of steroidogenic genes and proteins in testis that are
generally consistent with effect on testosterone (Navin et al., 2021; Zheng
etal., 2018)
• Oxidative stress in pituitary and hypothalamus and decreased prolactin
secretion in rats following 30-dav oral exposure (Nudler et al., 2009;
Quinteros et al., 2007) and in cultured rat primarv anterior pituitarv cells
(Quinteros et al., 2008; Quinteros et al., 2007)
This document is a draft for review purposes only and does not constitute Agency policy.
3-261 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and rationale
• Decreased testosterone production and transcriptional expression of
steroidogenic genes in cultured Levdig and Sertoli cells in vitro (Das et al.,
2015)
Effects on Sertoli cells
and the blood-testis
barrier
Interpretation: In vivo and in vitro evidence of impaired Sertoli cell function and
dynamics of the blood-testis barrier.
Key findings:
• Decrease in Sertoli cell secretory products (lactate, pyruvate, retinoic acid,
inhibin, androgen binding protein, transferrin) (Shobana et al., 2020)
• Leakage of Sertoli cell tight junctions and adverse effects on late-stage
spermatids in rats exposed i.p. (Murthv et al., 1991)
• In vivo and in vitro changes in the expression of molecules that form the
blood-testis barrier (Shobana et al., 2020; Kumar et al., 2017; Das et al.,
2015; Carette et al., 2013)
This document is a draft for review purposes only and does not constitute Agency policy.
3-262 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
3.2.8. Female Reproductive Effects
Female reproductive effects include endpoints related to the structure and function of
reproductive organs in pregnant and non-pregnant females, and the balance and cycling of
hormones from the HPG axis that regulate the development and function of these organs. This
section considers reproductive effects in females exposed to Cr(VI) at any life stage, including
exposures occurring preconception and for all stages of development. This is in accordance with
EPA's Framework for Assessing Health Risk of Environmental Exposures to Children fU.S. EPA.
2006d], which recommends that evidence for organ system toxicity be considered for all life stages
in order to identify populations or life stages that may be more susceptible to chemical-induced
toxicity. Exposure during pregnancy can affect both the mother and the fetus, and it is frequently
not possible to determine whether effects on the fetus are in response to or separate from maternal
toxicity in studies that report both. The maternal endpoints in animal toxicology studies described
in this section (maternal body weight gain and gestation length) must therefore be considered in
conjunction with the fetal endpoints (survival, growth, and structural alterations) that are
discussed in the Developmental Effects Section, 3.2.9.
3.2.8.1. Human Evidence
The majority of human studies with well-characterized exposure to Cr(VI) are conducted in
occupational studies where males predominate. Limited data are available on female reproductive
effects in either the occupational or general population settings. Two studies of female chromate
workers were identified that investigated outcomes on fertility, menstruation, pregnancy
complications, and pregnancy outcomes fRen etal.. 2003: Chen etal.. 19971. but were found to be
uninformative due to multiple deficiencies and thus were not further considered. A single ecologic
study (Re my etal.. 2017) considered female reproductive effects of Cr(VI) exposure in a population
living near a factory that used Cr(VI) in their production processes and where there was
documented contaminated groundwater. This study was considered low confidence due to
potential for exposure misclassification from the ecologic design (exposure was based on location
of residence in relation to the factory), outcome misclassification, and confounding. This study
reported higher relative risk of reproductive organ neoplasm (RR 1.27, 95% CI: 1.08,1.5), pelvic
inflammatory disease (1.31 (1.17,1.47)), endometriosis (1.19 (1.05,1.36)), menstrual disorder
(1.15 (1.03,1.29)), and ovarian cyst (1.43 (1.23,1.65)) in the more exposed geographic area. Due
to concerns for potential bias, however, these data are difficult to interpret on their own.
3.2.8.2. Animal Evidence
Study evaluation summary
Table 3-43 summarizes the animal toxicology studies considered in the evaluation of the
effects of Cr(VI) on the female reproductive system. These consist of a two-generation reproductive
This document is a draft for review purposes only and does not constitute Agency policy.
3-263 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Toxicological Review ofHexavalent Chromium
study with dietary exposure using NTP's Reproductive Assessment by Continuous Breeding (RACES)
protocol fNTP. 19971: subchronic oral exposure studies in adult animals fThompson etal.. 2020:
NTP. 2007: Kanoiiaetal.. 1998: Elbetiehaand Al-Hamood. 1997: Murthvetal.. 1996: NTP. 1996a.
b); gestational exposure studies that were designed to evaluate offspring development but also
reported some F0 maternal outcomes, such as gestational weight gain (Zheng etal.. 2018: Samuel et
al.. 2012a: Elsaieed and Nada. 2002: Tunaid etal.. 1996b. 1995: Trivedi etal.. 19891: and studies that
evaluated effects in F1 females from dams that had been exposed during gestation or lactation
fSivakumar etal.. 2022: Banu etal.. 2016: Banu etal.. 2015: Sivakumar etal.. 2014: Stanley etal..
2014: Stanley etal.. 2013: Samuel etal.. 2012a: Banu etal.. 2008: Al-Hamood etal.. 19981.
The RACB study (NTP. 19971 and subchronic exposure studies by NTP (2007.1996a. b)
were well-reported and well-designed to evaluate reproductive outcomes and were therefore rated
as high confidence for all reported outcomes (Table 3-43). The subchronic exposure study in mice
by Thompson etal. f20201 was also rated as high or medium confidence for most outcomes. The
remaining studies had reporting limitations and other substantial concerns raised during study
evaluation and were rated as low confidence across almost all outcomes. Endpoint-specific
concerns are discussed in the respective sections below. Two of the low confidence studies f Al-
Hamood etal.. 1998: Elbetieha and Al-Hamood. 19971 exposed animals to high concentrations
(350-1770 mg/L) of Cr(VI) in drinking water, which was considered a potential confounding
variable as it is not possible to determine whether reproductive effects may have been exacerbated
by reduced water consumption and/or systemic toxicity; for instance, drinking water
concentrations of 350 mg/L Cr(VI) have been associated in rats with decreased water consumption
and site of contact toxicity (80 and 100% incidence of ulcers in the glandular stomach of males and
females, respectively) fNTP. 20071. There were concerns about scientific integrity for two groups
of authors45 (Banu etal.. 2016: Banu etal.. 2015: Sivakumar etal.. 2014: Stanley etal.. 2014: Stanley
etal.. 2013: Samuel etal.. 2012a: Banu etal.. 2008: Kanoiia etal.. 1998: Tunaid etal.. 1996b: Murthv
etal.. 1996: Tunaid etal.. 1995: Trivedi etal.. 19891. which reduces confidence in these studies and
led to exclusion of three datasets but does not necessarily discount the results.
45Four studies demonstrating self-plagiarism—i.e., publication of identical data presented as separate and
unique experiments—were considered uninformative and were excluded from the assessment. Specifically, 1)
identical data were presented for rats by Kanoiia et al. (19961 and for mice by Tunaid et al. (1996a). despite
these being presented as separate studies in different species; and 2) subsets of the data presented by Samuel
et al. (2012b: 20111 were identical to that in an earlier publication by this laboratory group (Banu et al..
20081. Other studies by the same groups of authors, listed in the text above, were included in the assessment
but considered low confidence.
This document is a draft for review purposes only and does not constitute Agency policy.
3-264 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-43. Summary of included studies for Cr(VI) female reproductive
effects and overall confidence classification [high (H), medium (M), low (L)] by
outcome.3 Click to see interactive graphic with ratings rationale.
Author (year)b
Species
(strain)
Exposure life stage
and duration
Exposure
route
Fertility, Fecundity
Maternal BW gain
Gestation length
Hormones
Estrous cyclicity
Timing of puberty
Organ weight
Oocytes/ovarian
histopathology
Other histopathology
Al-Hamood et al.
(1998)
Mice (BALBC)
F1 females; GD 12-
PND 20
Drinking
water
L
"
"
"
"
L
L
-
"
Banu et al. (2008)
Rat (Wistar)
F1 females; PND 1-21
Drinking
water
-
"
"
L
L
L
-
L
"
Banu et al. (2015)
Rat (Sprague-
Dawley)
F1 females; GD 9.5-
14.5
Drinking
water
-
"
"
-
-
-
-
L
"
Banu et al. (2016)
Rat (Sprague-
Dawley)
F1 females; PND 1-21
Drinking
water
-
"
"
L
-
-
-
L
"
Elbetieha and Al-
Hamood (1997)
Mice (Swiss)
FO dams; 12 weeks
prior to mating
Drinking
water
L
"
"
-
-
-
L
-
"
Elsaleed and Nada
(2002)
Rat (Wistar)
FO dams; GD 6-15
Drinking
water
-
L
L
"
-
-
-
-
-
"
Junaid et al.
(1995)
Mice (Swiss
albino)
FO dams; GD 14-19
Drinking
water
-
"
-
-
-
-
-
"
Junaid et al.
(1996b)
Mice (Swiss
albino)
FO dams; GD 6-14
Drinking
water
L
Kanoiia et al.
(1998)
Rat
(Druckrey)
FO dams; 3 months
prior to mating
Drinking
water
L
L
-
L
-
-
L
"
Murthv et al.
(1996)
Mice (Swiss)
Adult females; 20 or
90 days
Drinking
water
-
-
-
L
-
-
L
"
NTP (1996a)
Mice (BALBC)
Adult females; 3, 6, or
9 weeks
Diet
-
-
-
-
-
-
H
H
NTP (1996b)
Rat (Sprague-
Dawley)
Adult females; 3, 6, or
9 weeks
Diet
H
H
NTP (1997)
Mice (BALBC)
Reproductive
Assessment by
Continuous Breeding
(F0 to F2)
Diet
H
H
H
H
H
NTP (2007)
Rats
(F344/N);
Mice
(B6C3F1)
Adult females; 3
months
Drinking
water
H
H
Samuel et al.
(2012a)
Rat (Wistar)
Study 1\ F0 dams and
F1 females; GD 9-21
Study 2: F1 females;
GD 9-PND 65
Drinking
water
L
L
L
L
L
This document is a draft for review purposes only and does not constitute Agency policy.
3-265 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Author (year)b
Species
(strain)
Exposure life stage
and duration
Exposure
route
Fertility, Fecundity
Maternal BW gain
Gestation length
Hormones
Estrous cyclicity
Timing of puberty
Organ weight
Oocytes/ovarian
histopathology
Other histopathology
Sivakumar et al.
(2014)
Rat (strain
not reported)
F1 females; GD 9.5-
14.5
Drinking
water
L
L
"
Sivakumar et al.
(2022)
Rat (Sprague-
Dawley)
F0 dams; GD 9.5-14.5
Drinking
water
L
"
Stanlev et al.
(2013)
Rat (Sprague-
Dawley)
F1 females; PND 1-21
Drinking
water
-
"
"
L
"
"
"
L
"
Stanlev et al.
(2014)
Rat (Sprague-
Dawley)
F1 females; PND 1-21
Drinking
water
-
"
"
L
L
L
"
L
M
Thompson et al.
(2020)
Mice
(B6C3F1)
5-week-old females;
90 days
Drinking
water
-
"
"
-
-
H
H
Trivedi et al.
(1989)
Mice (albino)
F0 dams; GD 0-19
Drinking
water
-
L
"
-
-
-
-
-
"
Zheng et al.
(2018)
Rat (Sprague-
Dawley)
F0 dams; GD 12-21
Gavage
-
L
"
-
-
-
-
-
"
BW = body weight; GD = gestation day; PND = postnatal day.
aln addition to these included studies, there were four animal toxicology studies reporting female reproductive
effects that met PECO criteria but were found to be uninformative at the study evaluation stage: Junaid et al.
(1996a), Kanoiia et al. (1996), Samuel et al. (2011), and Samuel et al. (2012b).
bData are available in HAWC for NTP (1997] (here), NTP (1996a] (here), NTP (1996b] (here).
1 Synthesis of evidence in animals
2 Fertility and fecundity
3 In the high confidence RACES study in mice (NTP. 19971. Cr(VI) exposure did not affect
4 pregnancy index in FO females at doses up to 50 mg/kg-day Cr(VI) via diet, and had no effect on
5 mating index, pregnancy index, or fertility index in F1 females at doses up to 39 mg/kg-day Cr(VI)
6 via diet Additionally, no effects on pregnancy rate were observed in the low confidence study by
7 Elbetieha and Al-Hamood (19971. in which mice were exposed to 707-1,770 mg/L Cr(VI)46 in
8 drinking water for 12 weeks prior to mating with untreated males.
46For many of the oral studies presented here, it was not possible to estimate an average daily mg/kg dose
due to lack of reporting. To estimate an average daily dose, paired records of body weight and daily intake of
test article are required. This is particularly important for Cr(VI] reproductive and developmental studies,
because rapid changes in maternal body weight are expected during pregnancy, and Cr(VI] affects palatability
(which affects both Cr(VI] intake rate and body weight]. Doses of Cr(VI] are presented where possible,
however many cross-study comparisons are done on the basis of mg/L Cr(VI] in drinking water. Reporting
and nomenclature related to exposure concentration units and water intakes for the studies by Kanoiia et al.
(19981. Murthv et al. (1996], and lunaidetal. (1995] were inconsistent with each other. This assessment
assumes that the drinking water concentrations provided by these studies (from the same laboratory] were
in units of mg/L potassium dichromate.
This document is a draft for review purposes only and does not constitute Agency policy.
3-266 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
In contrast, the low confidence study by Kanoiia etal. f 19981 reported a decrease in mating
index and fertility index in female rats exposed to 88.4-265 mg/L Cr(VI) in drinking water for
3 months prior to mating with untreated males. Two low confidence gestational exposure studies
also observed decreased pregnancy rates in F1 females from dams exposed to 8.8 mg/L Cr(VI) in
drinking water from GD 9.5-14.5 (rats) (Sivakumar et al.. 20141 or 353 mg/L Cr(VI) in drinking
water from GD 12-PND 20 (mice) (Al-Hamood etal.. 1998). Both of the gestational exposure
studies evaluated the F1 animals as individuals without considering the effects of litter, which has
the potential to overestimate statistical significance fHaseman etal.. 20011. Additionally, there is
uncertainty about how pregnancy rates were determined in the study by Sivakumar et al. f 20141.
which bred the animals continuously for 8-10 months and presented data as the percentage of F1
females pregnant at various blocks of age (2-4, 4-6, 6-8, and 8-10 months old); the authors did not
indicate how many times the animals became pregnant within each of these 2-month windows or
provide any additional information on how these percentage were calculated. Overall, although
decreased fertility was observed across several studies, interpretation is limited because these
studies were considered low confidence.
Maternal body weight gain
Decreased maternal body weights at the time of delivery were observed for both F0 and F1
dams in the RACES study in mice (NTP. 1997). which was considered high confidence for this
outcome. For F0 dams, which were allowed to produce up to five litters, the trend was statistically
significant for the first four litters; dam body weights were statistically significantly 5% decreased
compared to controls at doses of 24.4 mg/kg-day Cr(VI) for the first litter and 5-7% decreased
compared to controls at 50.6 mg/kg-day Cr(VI) for the first, second, and third litters, but were not
statistically significantly different from the control group in the fourth or fifth litters. For F1 dams,
the trend towards decreased dam body weights was statistically significant but treated animals did
not differ significantly from controls in any dose group. This study also observed a trend towards
decreased F0 dam body weights during lactation for the final litter; this trend was statistically
significant at PNDs 1, 4, and 14, and dam body weights were statistically significantly different from
controls at doses of 24.4-50.6 mg/kg-day Cr(VI) at these timepoints.
Dose-dependent decreases in maternal gestational weight gain were also observed in five
low confidence studies in which F0 rats or mice were exposed to potassium dichromate in drinking
water and euthanized near the end of gestation. None of these studies adjusted for gravid uterine
weight, which is considered preferable in order to distinguish between maternal and fetal toxicity
(U.S. EPA. 1991). so the magnitude of decreased gestational weight gain in these low confidence
studies likely reflects a combination of maternal toxicity as well as the decreased fetal growth and
survival that was observed in these studies (see "Developmental effects" section). Kanoiia et al.
(1998) exposed female rats for 90 days prior to mating and reported that gestational weight gain
was decreased by 10-22% compared to controls in the 88-265 mg/L dose groups, reaching
statistical significance at 177 mg/L Cr(VI). A10-15% mortality rate and clinical signs of hair loss
This document is a draft for review purposes only and does not constitute Agency policy.
3-267 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
and lethargy were also noted in females in the 177 and 265 mg/L dose groups in this study. In
three studies by the same group of authors that exposed mice for various durations during
pregnancy, gestational weight gain was decreased compared to controls by 11-26% (Tunaid etal..
19951. 8-24% (Tunaid etal.. 1996bl. and 17-20% (Trivedi etal.. 19891 following exposure from
GDs 14-19, 6-14, and 0-19, respectively, reaching statistical significance at 177 mg/L Cr(VI) in all
studies with no mortality or clinical signs of toxicity observed. The study by Trivedi etal. (19891
included a high dose group of 354 mg/L Cr(VI) in which the dams lost weight during the treatment
period and did not produce any litters. Elsaieed and Nada f20021 exposed rat dams to 5 0 mg/L
Cr(VI) from GD 6-15 and observed a 40% decrease in maternal body weight gain.
Lastly, in the low confidence study by Zheng etal. (20181. no effect on maternal body weight
was observed in F0 rat dams exposed from GD 12-21 at oral gavage doses up to 12 mg/kg-day
Cr(VI); however, body weight measurements in this study were taken 10 days after the exposure
ended, so are potentially insensitive due to the lag time between the exposure and endpoint
evaluation.
Gestation length
The only study that evaluated effects on gestation length was the high confidence RACES
study in mice by NTP T19971. There was no effect on the cumulative days to litter for F0 dams over
the course of five litters at doses up to 50.6 mg/kg-day Cr(VI) via diet "Cumulative days to litter" is
the number of days from cohabitation to the birth of each litter and is used as a metric for gestation
length in the RACES in lieu of checking for a copulatory plug. For F1 dams in this study, which were
only allowed to produce one litter and were checked for copulatory plugs to confirm mating, there
was likewise no effect on gestation length at doses up to 39 mg/kg-day Cr(VI) via diet
Hormones
Statistically significant decreases in serum estrogen, testosterone, and progesterone were
observed in weanling and peripubertal F1 females in four low confidence studies in which F0 dams
were exposed to 17.7-70.7 mg/L Cr(VI) in drinking water during lactation (PND 1-21) fBanu etal..
2016: Stanley etal.. 2014: Stanley etal.. 2013: Banu etal.. 20081. The same effects as well as
decreases in prolactin and growth hormone were observed in F1 females in the low confidence
study by Samuel etal. (2012al. in which F0 dams were exposed to 70.7 mg/L Cr(VI) in drinking
water from GD 9-PND 21 and F1 females were continued on the same dosing regimen from
weaning through PND 65. Three of these studies also evaluated gonadotropins and observed a
statistically significant increase in follicle stimulating hormone (Stanley etal.. 2013: Samuel etal..
2012a: Banu etal.. 20081. Luteinizing hormone was statistically significantly increased in the study
by Samuel etal. f2012al. whereas it was not affected in the study by Banu etal. f20081. Across all
five studies, effects were observed at all tested doses and generally at all timepoints evaluated,
which ranged from PND 0-65. Although results were consistent across studies, it should be noted
that all five studies were performed by the same group of researchers, so it is unclear whether
This document is a draft for review purposes only and does not constitute Agency policy.
3-268 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
results would be replicated by an outside research group or by higher confidence studies.
Measurements in all studies were presented as the mean of individual animals without accounting
for potential litter effects, which has the potential to overestimate statistical significance (Haseman
et al.. 20011. Samuel etal. (2012a) reported that body weights were decreased in the F1 females,
whereas the other studies did not report whether there was an effect on body weight or other
evidence of overt toxicity coinciding with the hormonal effects. Overall, the results indicate that
Cr(VI) decreases sex steroid hormone levels in females exposed during development, but
interpretation is limited because all studies were considered low confidence.
Estrous cyclicity
There were no notable effects on estrous cycle length, number of cycles, relative time spent
in estrous stages, or number of females with regular cycles in F1 mice in the high confidence dietary
exposure RACES study by NTP (19971. The proportion of F1 females with irregular cycles increased
with dose from 0/20 in the control group to 3/20 in the 39 mg/kg-day Cr(VI) dose group, but this
effect was not statistically significant and the remaining females had regular cycles with lengths
between 4-5 days. There was also no apparent effect on estrous cyclicity in mice exposed to levels
up to 149.3 mg/L Cr(VI) in drinking water for 90-days in a study by Thompson etal. f20201:
however, the authors did not provide quantitative data and based their conclusion on a single
vaginal smear taken at study termination, so the study was considered low confidence for this
outcome.
Four low confidence studies reported statistically significant increases in estrous cycle
length. A direct comparison between results from these low confidence studies and NTP T19971 is
complicated by the difference in oral administration (feed vs. drinking water), and inadequate
reporting of body weights and/or drinking water consumption by the low confidence studies
(precluding estimates of the mg/kg-d doses47). In adult rats exposed for 90 days, estrous cycle
duration was dose-dependently increased from a mean of 5.15 days in control animals to 8.66 days
at 265 mg/L Cr(VI) (Kanoiia et al.. 19981: however, effects above 88.4 mg/L Cr(VI) may be related
to overt toxicity, as there was a 10-15% mortality rate and decreased body weight among females
in the 177 and 265 mg/L dose groups. In another study in adult mice that used these same dose
levels but a 20-day exposure duration, there was a statistically significant increase in estrous cycle
duration from a mean of 4.4 days in control animals to 7.7 days at 265 mg/L Cr(VI) with no effects
at lower dose levels (Murthv etal.. 1996). The authors did not report whether there was an effect
on body weights or clinical signs of toxicity, which are likely to occur at the 2 65-mg/L dose level
and limits the interpretation of this finding. The remaining two studies investigated estrous
cyclicity in F1 females that had been exposed during development Samuel etal. f2012al exposed
47Based on the information available, the ad libitum drinking water doses from Kanoiia et al. (19981 and
Murthv et al. T19961 were higher than the dietary doses from NTP ("19971. while the doses in Banu et al.
f20081 and Samuel et al. f2012al were lower than NTP f~19971.
This document is a draft for review purposes only and does not constitute Agency policy.
3-269 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
F0 dams to 70.7 mg/L Cr(VI) in drinking water during gestation and lactation (GD 9-PND 21) and
continued F1 females on the same dosing regimen through PND 65 and observed a statistically
significant increase in the number of hours spent in metestrous and diestrous by the F1 animals.
Similarly, Banu etal. (20081 reported a statistically significant increase in the number of hours
spent in diestrous for F1 females from dams exposed to 70.7 mg/L Cr(VI) in drinking water from
PND 1-21, but no change in other estrous phases. None of the available studies indicated whether
investigators were blinded to treatment groups during the evaluation of vaginal cytology, which
would be considered appropriate for reducing observational bias. Measurements in the
developmental exposure studies by Samuel etal. (2012a) and Banu etal. (20081 were presented as
the mean of individual F1 animals without accounting for potential litter effects, which has the
potential to overestimate statistical significance fHaseman et al.. 20011. The finding of increased
estrous cycle duration is coherent with the decreased expression of sex steroid hormones within
the developmental studies by Samuel etal. (2012a) and Banu etal. (20081 (see "Hormones" section
above), but interpretation is limited because effects were observed only in low confidence studies.
Timing of puberty
Four low confidence studies that evaluated F1 females following developmental exposure
reported a statistically significant increase in the age at vaginal opening, which is a biomarker of
female puberty. In F1 mice from dams exposed to potassium dichromate in drinking water from GD
12-PND 20, Al-Hamood et al. f 19981 observed a statistically significant increase in the mean age of
vaginal opening from 24.6 days in control animals to 27 days at 353 mg/L Cr(VI); however, the
authors did not report whether there was overt maternal toxicity, which would be expected at this
high dose level (see "Maternal body weight gain" section above) and could limit the interpretation
of this finding. In two studies that exposed rat dams to potassium dichromate in drinking water
from PND 1-21, there were statistically significant increases in the mean age of vaginal opening in
F1 females from 33 days in control animals to 55 days at 70.7 mg/L Cr(VI) (Banu etal.. 20081. and
from 31 days in control animals to 42 days at 17.7 mg/L fStanlev etal.. 20141. Another study in
developing rats by Samuel etal. (2012a) exposed F0 dams to 70.7 mg/L Cr(VI) in drinking water
during gestation and lactation (GD 9-PND 21) and continued F1 females on the same dosing
regimen through PND 65, and observed a statistically significant increase in the mean age of vaginal
opening from 42.3 days in control animals to 65 days at 70.7 mg/L Cr(V)48. In all four of these
studies, results were presented as the mean of individual F1 animals without accounting for
potential litter effects, which has the potential to overestimate statistical significance (Haseman et
al.. 20011.
Delayed puberty is coherent with decreased estrogen levels in three of these studies
(Stanley et al.. 2014: Stanley etal.. 2013: Banu etal.. 2008) (see "Hormones" section above).
48Numerical values in the study by Samuel etal. f2012al were extracted from a figure using WebPlotDigitizer
software: https://automeris.io/WebPlotDigitizer/.
This document is a draft for review purposes only and does not constitute Agency policy.
3-270 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Delayed puberty can also be closely tied to decreased body weight fGreenspan and Lee. 20181. so
examination of body weight may provide a means for separating direct effects on puberty from
those that are related to general delays in development. Samuel etal. f2012al reported decreased
body weights in Cr(VI) treatment groups at multiple postnatal timepoints, whereas Banu et al.
(20081 and Stanley etal. (20141 did not report body weights. Al-Hamood etal. (19981 reported
that body weight of the F1 females was not affected by Cr(VI) exposure, but the study was not clear
about when the body weight measurements were taken. Thus, the delayed puberty could be
related either to decreases in reproductive hormones or body weight. Overall, interpretation of
these low confidence studies is limited.
Organ weight
Effects on female reproductive organ weight were inconsistent across studies. No effects on
absolute or relative ovary weights were observed in adult F0 or F1 females in the high confidence
RACES study in mice at doses up to 50.6 and 39 mg/kg-day Cr(VI) via diet, respectively fNTP. 19971.
The high confidence study by Thompson etal. f20201 reported no change in the absolute weight of
the ovaries or uterus following a 90-day exposure to 149.3 mg/L Cr(VI) in drinking water. No effect
on relative ovary or uterus weights were observed at PND 50 in F1 female mice exposed to
353 mg/L Cr(VI) in drinking water from GD 12-PND 20 in the low confidence developmental
exposure study by (Al-Hamood et al.. 1998). In the low confidence study in adult mice by (Elbetieha
and Al-Hamood. 19971. relative ovary weight was statistically significantly increased following
exposure to 1770 mg/L Cr(VI) in drinking water for 12 weeks, while relative uterus weight was not
changed. Conversely, in the low confidence study in rats by Samuel etal. f2012al. there was a dose-
dependent decrease in absolute uterus and ovary weight in F0 rat dams exposed to potassium
dichromate in drinking water from GD 9-21 that reached statistical significance at 35.3 mg/L and
70.7 mg/L Cr(VI), respectively. The study by Samuel etal. (2012al also evaluated F1 females that
were continued on the 70.7 mg/L Cr(VI) dosing regimen through PND 65, and observed a
statistically significant decrease in absolute ovary and uterus weight at multiple timepoints
measured between PND 3 and PND 65. Samuel etal. f2012al evaluated F1 animals as individuals
without accounting for potential litter effects, which has the potential to overestimate statistical
significance (Haseman etal.. 20011. Body weights were decreased in both studies that observed
effects, which could have contributed to the increase in relative organ weights and decrease in
absolute organ weights. Overall, interpretation is limited because effects were only observed in low
confidence studies and were not seen in high confidence studies, and the direction of effect was
inconsistent
Oocytes and ovarian histopathology
The high confidence subchronic studies by NTP reported no gross or microscopic changes in
the ovary in adult rats or mice following up to 9 weeks of exposure to doses up to 8.5 or
32.5 mg/kg-day Cr(VI) via diet, respectively (NTP. 1996a. b); or in adult rats or mice following
This document is a draft for review purposes only and does not constitute Agency policy.
3-271 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
3-month exposure to doses up to 20.9 or 27.9 mg/kg-day Cr(VI) via drinking water fNTP. 20071.
respectively. No gross changes were observed in the ovary in F0 or F1 females in the high
confidence RACES study in mice at doses up to 50.6 and 39 mg/kg-day Cr(VI) via diet, respectively
(NTP. 19971. The high confidence study by Thompson etal. (20201 likewise reported no change in
the numbers of small, medium, or large follicles and no change in the incidence of follicular atresia
in mice following 90-day exposure to levels up to 149.3 mg/L in drinking water.
In contrast, nine low confidence studies reported pathological effects in the ovary following
exposure to potassium dichromate in drinking water. Kanoiia et al. T19981 reported a statistically
significant decrease in the number of corpora lutea in maternal female rats that had been exposed
to doses of 177 mg/L Cr(VI) and higher in drinking water for 3 months prior to mating; however,
there was a 10-15% mortality rate and clinical signs of toxicity among rats at these dose levels, so
this effect may be indicative of overt toxicity. Similarly, following exposure in adult mice for
20 days, Murthv et al. f 19961 reported a dose-related statistically significant decrease in follicle
numbers at drinking water concentrations of 88.4 mg/L Cr(VI) and higher, and a statistically
significant decrease in the number of ova recovered when the animals were induced to
superovulate at concentrations of 177 mg/L Cr(VI) and higher. The remaining seven low
confidence studies evaluated ovarian histopathology in developing F1 females and were performed
by a single group of authors (Banu, Stanley, Sivakumar, Samuel, and coauthors). Following
gestational exposure (GD 9.5-14.5) of F0 dams to 8.8 mg/L Cr(VI), F1 female rat fetuses and
newborn pups were found to have decreased oocyte counts and accelerated breakdown of germ
cell nests into primordial follicles49 (Sivakumar etal.. 2022: Banu etal.. 2015: Sivakumar et al..
20141. with an increased number of primary and secondary follicles at PND 4 in treated animals
compared to the control group fBanu etal.. 20151. Following lactational exposure (PND 1-21) of F0
dams to 8.8-70.7 mg/L Cr(VI), F1 female rats were found to have a dose-related increase in
incidence of follicular atresia50 (Banu etal.. 2016: Stanley etal.. 2014: Stanley etal.. 20131 and
decreased numbers of primordial, primary, secondary, and antral follicles (Banu etal.. 20081 at
timepoints between PND 21 and PND 65. Samuel etal. f2012al exposed F0 dams to 70.7 mg/L
Cr(VI) in drinking water during gestation and lactation (GD 9-PND 21) and continued F1 females
on the same dosing regimen through PND 65, and observed pyknotic nuclei and vacuolation in
oocytes, stunted or arrested ovarian follicle development, and abnormalities in thecal cells,
granulosa cells, and luteum in F1 females at various timepoints measured between PND 3-65, but
did not provide quantitative data. These ovarian effects are coherent with the effects on hormones
that were observed in some of these studies (Banu etal.. 2016: Stanley etal.. 2014: Stanley etal..
2013: Samuel etal.. 2012a: Banu etal.. 20081 (see above section) since estrogens and
49Germ cell nests are clusters of oogonia that are formed in the developing ovary during late gestation. Germ
cell nests are present at birth, and then are broken down into primordial follicles during the final stage of
early ovarian development (Wear etal.. 20161.
50Follicular atresia is defined as degenerative changes in the granulosa cell layers or oocyte.
This document is a draft for review purposes only and does not constitute Agency policy.
3-272 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
gonadotropins play a critical role in the growth and development of oocytes. Biological plausibility
is also provided by molecular observations of increased oxidative stress, apoptosis, and effects on
the extracellular matrix in the ovary, which are discussed below as mechanistic evidence.
Interpretation of the histopathological changes in the ovary is limited, however, because effects
were observed only in low confidence studies and were not seen in the high confidence studies.
Other histopathology of the female reproductive system
The high confidence studies by NTP reported no effects on the incidence of gross or
microscopic lesions in the vagina, cervix, uterus, or clitoral gland in adult rats or mice following up
to 9 weeks of exposure to doses up to 8.5 or 32.5 mg/kg-day Cr(VI) via diet, respectively fNTP.
1996a. b); or in adult rats or mice following 3-month exposure to doses up to 20.9 or 27.9 mg/kg-
day Cr(VI) via drinking water fNTP. 20071. respectively. No treatment-related gross lesions were
observed in these organs in F0 or F1 females in the RACES study in mice at doses up to 50.6 and
39 mg/kg-day Cr(VI) via diet, respectively fNTP. 19971. The study by Thompson et al. f20201
likewise reported no significant alterations in the gross and microscopic appearance of the corpus
and cervix uteri, vaginas, or mammary glands, but was considered medium confidence for this
outcome because no quantitative data was reported.
3.2.8.3. Mechanistic Evidence
The Cr(VI) literature provides evidence informing potential mechanisms of Cr(VI)-induced
female reproductive toxicity; specifically, oxidative stress and apoptosis in female reproductive
tissues, altered hormone signaling, and effects on the extracellular matrix. Mechanistic studies are
tabulated in Appendix C.2.7 and summarized here.
The mechanistic studies reviewed here consisted of in vivo mechanistic data from several of
the included oral exposure studies discussed above (Table 3-43), as well as from intraperitoneal
(i.p.) injection studies that did not meet PECO criteria but were reviewed as relevant to the
mechanistic synthesis. Dosing via i.p. injection is likely to result in higher tissue concentrations of
Cr(VI) compared to oral exposure, since an oral first-pass effect exists due to the reduction of Cr(VI)
in the low pH environment of the stomach; less than 10-20% of an ingested dose may be absorbed
in the GI tract, and further reduction will occur in the liver prior to distribution to the rest of the
body (see Section 3.1 and Appendix C.l). Therefore, systemic effects are expected to be more likely
following i.p. injection or inhalation compared to oral exposure. In vitro studies conducted in
relevant cell types, such as thecal and granulosa cells, were also considered for mechanistic
evidence.
Altered steroidogenesis
The effects on hormone levels (described in sections above) are supported by changes in
the ovarian expression of genes involved in steriodogenesis, which were observed in rats and rat
granulosa cells following exposure to potassium dichromate. In F1 rats, Stanley etal. (20131
This document is a draft for review purposes only and does not constitute Agency policy.
3-273 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
reported decreased ovarian FSH receptor gene expression and Banu etal. f20161 reported
decreased ovarian gene expression of steroidogenic acute regulatory protein (StAR),
3(3-hydroxysteroid dehydrogenase, and aromatase. Banu etal. (20161 also reported increased gene
expression of enzymes involved in the metabolic clearance of estradiol (Cyplal, Cyplbl,
UDP-glucuronosyltransferases, Sultlal, NAD(P)H quinone oxidoreductase 1). Similar effects were
observed in an immortalized rat granulosa cell line (Stanley etal.. 2011: Banu etal.. 20081 and in
primary rat granulosa cells (Stanley etal.. 2013: Stanley etal.. 20111. including decreased
expression of LH receptor, FSH receptor, estrogen receptors (ERa, ER(3), StAR, steroidogenic factor
(SF)-l, and 17(3-hydroxysteroid dehydrogenases -1 and -2. In all of these studies, these effects
(including steroid hormone measurements in the in vivo studies) were attenuated by cotreatment
with an antioxidant (vitamin C or resveratrol). Stanley etal. f20141 found that cotreatment of
potassium dichromate-exposed F1 female rats with estradiol restored the expression of several
antioxidant enzymes (Gpxl, catalase, Prdx3, and Txn2), also suggesting a relationship between
hormonal effects and oxidative stress.
Oxidative stress
Decreased antioxidant enzyme expression or activity [e.g., superoxide dismutase (SOD),
catalase (CAT), glutathione peroxidase (GPx), glutathione reductase (GR), peroxiredoxin (PRDX) 3,
and thioredoxin (TXN)], decreased nonenzymatic antioxidants (glutathione, metallothionine,
vitamin C), and increased markers of oxidative stress (lipid peroxidation, superoxide anion, H2O2)
were observed in the ovary in several of the studies in F1 rats described above fBanu etal.. 2016:
Stanley etal.. 2014: Stanley etal.. 2013: Samuel et al.. 2 012 al and in adult mice fRao etal.. 20091
following oral exposure, as well as in the uterus of adult rats following intraperitoneal injection
(Marouani et al.. 2015b). Increased ovarian glutathione-S-transferase (GST) (Stanley et al.. 20131
and SOD expression (Banu etal.. 2016) were observed in some cases. A similar spectrum of effects
was observed in vitro in primary granulosa and theca cells isolated from immature rats and in an
immortalized granulosa cell line (Stanley etal.. 2013). Sivakumar et al. (2014) observed that
potassium dichromate exposure increased colocalization of p53/SOD-2 in the ovary of F1 rats and
hypothesized that this could be contributing to oxidative stress, as p53 has been demonstrated to
reduce SOD-2 antioxidant activity.
Several in vivo studies found that cotreatment of animals with antioxidants (vitamin C,
resveratrol, ginseng edaravone) mitigated apical outcomes including decreased maternal body
weight gain, follicular atresia, and effects on pubertal onset, estrous cyclicity, and hormone levels
(Banu etal.. 2016: Stanley etal.. 2014: Stanley etal.. 2013: Banu etal.. 2008: Elsaieed and Nada.
20021. This may imply that oxidative stress is a mechanism underlying these effects, but
interpretation is difficult because antioxidants can also decrease tissue Cr(VI) levels by stabilizing
lower Cr oxidation states. For instance, (Elsaieed and Nada. 2002) reported lower plasma, placenta,
and fetus Cr levels with coadministration of ginseng, and (Banu etal.. 2008) reported lower plasma
and ovarian Cr levels with coadministration of Vitamin C.
This document is a draft for review purposes only and does not constitute Agency policy.
3-274 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Apoptosis of somatic and germ cells
In a series of studies in F1 rat pups that reported accelerated breakdown of germ cell nests,
follicular atresia, and decreased follicle counts (Sivakumar etal.. 2022: Banu etal.. 2016: Banu etal..
2015: Sivakumar etal.. 2014: Stanley etal.. 2014: Stanley etal.. 20131. these histopathological
changes were accompanied by increased apoptosis of follicular and germ cells. Evidence included
increased staining in the TUNEL assay, increased expression of pro-apoptotic markers [Bax,
cytochrome c, caspase-3, p53, p27, p53-upregulated modulator of apoptosis (PUMA)], decreased
expression of anti-apoptotic markers (Bcl-2, Bcl-XL, Bcl211, HIF-la), and decreased expression of
other signaling molecules that regulate cell survival [p-AKT, p-ERK, X-linked inhibitor of apoptosis
protein (XIAP)]. Increased apoptotic cells and protein expression of Bax in the uterus was also
reported in adult female rats following intraperitoneal injection with potassium dichromate,
accompanied by a decrease in the relative weight of the uterus and ovary (Marouani etal.. 2015bl.
In primary granulosa cells from immature rats, fBanu etal.. 20111 similarly reported upregulation
of apoptotic markers and down-regulation of anti-apoptotic markers and further investigated the
role of signal transduction pathways that regulate cell survival, finding that apoptosis and p53
activity were decreased after treatment with an ERK1 /2 inhibitor. Another study in primary and
immortalized rat granulosa cells reported that potassium dichromate induced cell cycle arrest,
decreased expression of proteins that regulate the progression of the cell cycle [cyclins, cyclin-
dependent kinases (CDKs), and proliferating cell nuclear antigen (PCNA)], and increased expression
of inhibitors of CDKs (pl5, pl6, and p27), although authors stated it was unclear whether these
disruptions to the cell cycle were a cause or a consequence of apoptosis (Stanley etal.. 20111.
Effects on the ovarian extracellular matrix
Banu etal. (20151 proposed a mechanism by which Cr(VI) induces premature ovarian
failure by targeting the metalloenzyme X-propyl aminopeptidase (coded by the gene Xpnpep2),
leading to effects on the extracellular matrix. In F1 female rats from dams that were exposed to
25 mg/L potassium dichromate in drinking water from GD 9.5-14.5, the authors reported increased
ovarian expression of Xpnpep2 during late gestation and decreased ovarian expression of Xpnpep2
during early postnatal life. Levels of ovarian collagen expression (Coll, Col3, Col4) were inversely
proportional to Xpnpep2 at each of the sample time points. The authors hypothesized that Cr(VI)
accelerates the breakdown of germ cell nests by upregulating Xpnpep2 and decreasing the
distribution of collagen in the fetal ovary and alters the histoarchitecture of the ovary in postnatal
animals by downregulating Xpnpep2.
3.2.8.4. Integration of Evidence
Overall, the available evidence is inadequate to assess whether Cr(VI) may cause female
reproductive effects. Although an association with female reproductive toxicity was demonstrated
in a single low confidence epidemiology study and a series of low confidence animal toxicology
studies, effects were not observed in medium or high confidence studies aside from a moderate
This document is a draft for review purposes only and does not constitute Agency policy.
3-275 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
decrease in maternal body weight fNTP. 19971. Integrated evidence of the female reproductive
effects of Cr(VI) exposure from human, animal, and mechanistic studies is summarized in an
evidence profile table (Table 3-44).
The evidence of an association between Cr(VI) exposure and female reproductive effects in
humans is indeterminate. A single low confidence study indicated higher risk of several female
reproductive conditions in a population that was estimated to have higher Cr(VI) exposure, but
there is too much uncertainty to draw conclusions regarding these associations.
Evidence of female reproductive effects from animal toxicology studies and supportive
mechanistic data from in vivo and in vitro studies was also found to be indeterminate. Across high
confidence studies in rats and mice (Thompson etal.. 2020: NTP. 1997.1996a. b), the only notable
female reproductive effect was a 5-7% decrease in F0 and F1 maternal body weights at delivery in
the RACES study in mice (NTP. 19971: fertility, fecundity, and estrous cyclicity were not affected, and
effects on organ weights, follicle counts, and histopathology were not observed. In contrast,
profound effects on female fertility, estrous cyclicity, hormone levels, ovarian follicles and germ
cells, and reproductive development were observed across the other available studies, which were
all considered low confidence and many of which were from a single research group. The doses of
Cr(VI) at which effects were observed could not be calculated for any of the low confidence studies
because drinking water consumption data was not reported, but the available information indicates
that some were higher and some were lower than doses used by NTP; so, it is unlikely that the
discrepancy in responses between high and low confidence studies is simply due to a difference in
the dose ranges tested. Some of the low confidence studies used relatively high dose levels
associated with mortality or other overt toxicity, limiting the ability to interpret the female
reproductive findings. A strength of these low confidence studies is that they evaluated several
indicators of female reproductive toxicity that were not included in the NTP studies: specifically,
steroid hormone and gonadotropin levels, age at pubertal development, and ovarian histopathology
during early developmental stages. The interpretation of the low confidence studies is limited,
however, by deficiencies in study design, conduct, and reporting. Support for biological plausibility
of Cr(VI)-induced female reproductive effects comes from mechanistic data that was also largely
published by the same laboratory group, demonstrating altered expression of steroid hormone
signaling pathways in female rats and rat cells, as well as oxidative stress and apoptosis in rodent
ovarian and uterine tissues and cells. There were no animal studies that evaluated female
reproductive effects following inhalation exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
3-276 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-44. Evidence profile table for female reproductive outcomes
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Evidence from studies of exposed humans
OOO
FEMALE REPRODUCTIVE
EFFECTS
Low confidence:
Remv et al. (2017)
One ecologic study reported
higher relative risk for
reproductive organ neoplasm,
pelvic inflammatory disease,
endometriosis, menstrual
disorder, and ovarian cysts in a
higher exposed geographic area.
• No factors noted
• Low
confidence
study
ooo
Indeterminate
There is some indication
of an association
between Cr(VI) exposure
and female reproductive
effects, but the only
evidence comes from a
single, low confidence
ecologic study so there is
considerable uncertainty
in the findings.
The evidence is
inadequate to assess
whether Cr(VI) causes
female reproductive
toxicity in humans.
The single human study
and most animal studies
were considered low
confidence. With the
exception of decreased
maternal body weight,
effects in low confidence
animal studies were not
seen in the high
confidence RACB and
subchronic and studies.
Evidence from animal studies
FERTILITY AND
FECUNDITY
High confidence:
NTP (1997)
Low confidence:
Kanoiia et al. (1998)
Elbetieha and Al-
Hamood (1997)
Al-Hamood et al. (1998)
Sivakumar et al. (2014)
No effects on mating or pregnancy
rates in mice in the high
confidence RACB study (NTP,
1997) or in a low confidence 12-
week exposure study (Elbetieha
and Al-Hamood, 1997).
Decreased fertility or fecundity in
female rats or mice after
developmental or adult exposure
was reported in 3 low confidence
studies.
• No factors noted
• Effects
observed only
in low
confidence
studies
OOO
Indeterminate
Evidence of female
reproductive effects was
observed in multiple low
confidence studies.
Decreased F0 and F1
maternal body weights
in a RACB study in mice
(NTP, 1997) was the
Mechanistic findings
(animals and in vitro)
provide evidence
supportive of female
reproductive toxicity.
These mechanisms are
presumed relevant to
humans.
This document is a draft for review purposes only and does not constitute Agency policy.
3-277 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
MATERNAL BODY
WEIGHT GAIN
High confidence:
NTP (1997)
Low confidence:
Elsaieed and Nada
(2002)
Junaid et al. (1995)
Junaid et al. (1996b)
Kanoiia et al. (1998)
Trivedi et al. (1989)
Zheng et al. (2018)
Decreased maternal body weight
was reported in 6 out of 7 studies,
including F0 and F1 animals in the
high confidence RACB study.
In low confidence studies,
decreased maternal body weights
during pregnancy were concurrent
with decreased fetal survival
and/or fetal body weight, and
authors did not adjust for gravid
uterine weight to distinguish
between maternal and fetal
effects.
• High confidence
study
• Consistency
• Dose-response
gradient
• Low
confidence
studies did not
adjust for
gravid uterine
weight
only notable effect in
high confidence studies.
GESTATION LENGTH
High confidence:
NTP (1997)
No effects on cumulative days to
litter (F0 dams) or gestation length
(F1 dams) in a high confidence
RACB study in mice.
• No factors noted
• No factors
noted
HORMONES
Low confidence:
Banu et al. (2008)
Banu et al. (2016)
Stanlev et al. (2013)
Stanlev et al. (2014)
Samuel et al. (2012a)
Decreased serum estrogen,
testosterone, and progesterone
and increased FSH and LH in F1
rats in five low confidence studies
from a single laboratory group.
Decreased prolactin and growth
hormone also noted in one of
these studies.
• No factors noted
• Low
confidence
studies, all
from one
research group
ESTROUS CYCLICITY
High confidence:
NTP (1997)
Low confidence:
Kanoiia et al. (1998)
Murthv et al. (1996)
No notable effects on F1 estrous
cyclicity in the high confidence
RACB study in mice.
Increased estrous cycle duration in
four low confidence studies in rats
or mice exposure during
development or as adults.
• No factors noted
• Effects
observed only
in low
confidence
studies
This document is a draft for review purposes only and does not constitute Agency policy.
3-278 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Banu et al. (2008)
Samuel et al. (2012a)
Thompson et al. (2020)
TIMING OF PUBERTY
Low confidence:
Al-Hamood et al. (1998)
Banu et al. (2008)
Stanlev et al. (2014)
Samuel et al. (2012a)
Increase in the age at pubertal
onset (vaginal opening) was
reported in F1 female rats or mice
in four low confidence studies.
• No factors noted
• Effects
observed only
in low
confidence
studies
ORGAN WEIGHT
High confidence:
NTP (1997)
Thompson et al. (2020)
Low confidence:
Elbetieha and Al-
Hamood (1997)
Al-Hamood et al. (1998)
Samuel et al. (2012a)
Increased relative ovary weight
and decreased absolute ovary and
uterus weight in 2 low confidence
studies.
Otherwise, no effects were
observed.
• No factors noted
• Effects
observed only
in low
confidence
studies
• May be
secondary to
decreased
body weight
OOCYTES AND OVARIAN
HISTOPATHOLOGY
High confidence:
NTP (1996a)
NTP (1996b)
NTP (1997)
NTP (2007)
Thompson et al. (2020)
Low confidence:
Kanoiia et al. (1998)
No gross or microscopic changes in
the ovary across 5 high confidence
studies.
Decreased corpora lutea and
decreased follicle numbers and
ova following superovulation in
low confidence studies.
Degenerative effects on the ovary
including accelerated breakdown
of germ cell nests, follicular
atresia, stunted or arrested follicle
• No factors noted
• Effects
observed only
in low
confidence
studies, mostly
from one
research group
This document is a draft for review purposes only and does not constitute Agency policy.
3-279 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Inferences and summary
judgment
Murthv et al. (1996)
Banu et al. (2008)
Banu et al. (2015)
Banu et al. (2016)
Sivakumar et al. (2022)
Sivakumar et al. (2014)
Stanlev et al. (2013)
Stanlev et al. (2014)
Samuel et al. (2012a)
development, and decreased
follicle counts across 7 low
confidence studies from a single
laboratory group.
OTHER
HISTOPATHOLOGY OF
THE FEMALE
REPRODUCTIVE SYSTEM
High confidence:
NTP (1996a)
NTP (1996b)
NTP (1997)
NTP (2007)
Medium confidence:
Thompson et al. (2020)
No gross or microscopic changes
were observed in the vagina,
cervix, uterus, and/or clitoral gland
across 5 high or medium
confidence studies.
• No factors noted
• No factors
noted
Mechanistic evidence
Biological events or
pathways
Summary of key findings and interpretations
Judgments and
rationale
Altered steroidogenesis
Interpretation: Cr(VI) alters steroidogenesis in vivo and in vitro.
Key findings:
• Decreased estrogen, testosterone, and progesterone and increased FSH
and LH in animals in F1 rats following gestational exposure (Banu et al.,
2016; Stanlev et al., 2014; Stanlev et al., 2013; Samuel et al., 2012a; Banu
et al., 2008).
Observations of altered
hormone signaling,
oxidative stress,
apoptosis, and effects on
the ovarian extracellular
matrix.
This document is a draft for review purposes only and does not constitute Agency policy.
3-280 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Factors that increase Factors that
Summary of key findings certainty decrease certainty
Judgments and
rationale
• Decreased ovarian expression of gonadotropin receptors and/or
steroidogenic genes in F1 rats (Banu et al., 2016; Stanlev et al., 2013) and
in cultured rat granulosa cells (Stanlev et al., 2013; Stanlev et al., 2011;
Banu et al., 2008).
• Upregulation of genes involved in metabolic clearance of estradiol in F1
rats (Banu et al., 2016).
Oxidative stress was
concurrent with apical
outcomes in some
animal studies.
Effects on maternal body
weight gain, follicular
atresia, pubertal onset,
estrous cyclicity, and
hormones were
mitigated by
cotreatment of
antioxidants.
Oxidative stress
Interpretation: In vivo and in vitro evidence of Cr(VI)-induced oxidative stress in
female reproductive tissues concurrent with apical measurements of female
reproductive toxicity.
Key findings:
• Decreased antioxidant activity or expression in the ovary was observed in
F1 rats (Banu et al., 2016; Stanlev et al., 2014; Stanlev et al., 2013; Samuel
et al., 2012a), in orallv exposed adult mice (Rao et al., 2009), in the rat
uterus following i.p. injection (Marouani et al., 2015b), and in cultured rat
granulosa and theca cells (Stanlev et al., 2013).
• Cotreatment of with antioxidants mitigated effects on maternal body
weight gain, follicular atresia, and effects on pubertal onset, estrous
cvclicitv, and hormone levels (Banu et al., 2016; Stanlev et al., 2014;
Stanlev et al., 2013; Banu et al., 2008; Elsaieed and Nada, 2002).
Apoptosis of somatic
and germ cells
Interpretation: In vivo and in vitro evidence of Cr(VI)-induced apoptosis in
ovarian follicles and germ cells.
Key findings:
• Increased TUNEL assay staining, increased expression of pro-apoptotic
markers, decreased expression of anti-apoptotic markers, and/or
decreased expression of other signaling molecules that regulate cell
survival reported in ovarian tissue of F1 rats (Sivakumar et al., 2022; Banu
et al., 2016; Banu et al., 2015; Sivakumar et al., 2014; Stanlev et al., 2014;
This document is a draft for review purposes only and does not constitute Agency policy.
3-281 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease certainty
Judgments and
rationale
Stanlev et al., 2013). Similar findings reported in adult female rats
following i.o. injection (Marouani et al., 2015b).
• In vitro evidence of cell cvcle arrest in cultured rat granulosa cells (Stanlev
et al., 2011).
Ovarian extracellular
matrix
Interpretation: In vivo evidence that Cr(VI) induces premature ovarian failure
by altering the extracellular matrix.
Key findings:
• Ovarian expression of the metalloenzyme X-propyl aminopeptidase was
increased in F1 rats during late gestation and increased during early
postnatal life and was inversely proportional to ovarian collagen
expression (Banu et al., 2015).
This document is a draft for review purposes only and does not constitute Agency policy.
3-282 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
3.2.9. Developmental Effects
Developmental toxicity encompasses effects that occur following pre- or postnatal exposure
of the developing organism. The major categories of developmental toxicity discussed in this
section are changes in survival, growth, structural alterations, and effects on the placenta.
Functional effects on specific organ systems following developmental exposures are considered in
their respective sections (e.g., "Male reproductive effects" and "Female reproductive effects"
sections) and are also summarized here. These endpoints are considered relevant for
developmental toxicity risk assessment per U.S. EPA guidelines (U.S. EPA. 19911.
This section considers both indirect (maternal or paternal) and direct routes of exposure to
the developing organism. As noted previously, it is frequently difficult to determine whether effects
on the fetus are in response to or separate from maternal toxicity in studies that report both, so the
fetal endpoints described in this section should be considered in conjunction with the maternal
endpoints described in the "Female reproductive effects" section. Developmental effects produced
at doses that cause minimal maternal toxicity are still considered to represent developmental
toxicity and should not be discounted as maternal toxicity (U.S. EPA. 1991). Less is known about
the potential impact of paternal exposures prior to conception, but it is thought that offspring
development can be affected by genetic or epigenetic changes in sperm or by direct exposure to
toxicant residues in the seminal fluid.
3.2.9.1. Human Evidence
Study evaluation summary
Table 3-45 summarizes the nine human epidemiology studies (eight publications)
considered in the evaluation of the developmental effects of Cr(VI). Four studies were found to be
uninformative due to critical deficiencies in one or more domains fXia etal.. 2016: Ouansah and
Taakkola. 2009: Ren etal.. 2003: Chen etal.. 1997) and were not considered further. Of the six
included studies, three studies (four publications) from the same research group examined
paternally mediated effects on offspring, specifically resulting from paternal occupational
exposures to Cr(VI) from stainless-steel welding f Hiollund etal.. 2005: Hiollund etal.. 2000:
Hiollund etal.. 1995: TP etal.. 19921. Exposure was measured in these studies using questionnaires.
Participants were asked about their past and current welding experiences including type of metal
(stainless or mild steel), welding methods, timing of welding exposures (years welding), and safety
precautions used (ventilation). In each study, exposure was analyzed in three categories (stainless-
steel welding, mild steel welding, and no welding). The questionnaires were not validated, and thus
all the studies were evaluated as low confidence due to concerns in the exposure measurement
domain. Spontaneous abortion was examined in all three studies, and one of these flP etal.. 19921
also examined preterm birth, fetal growth, infant death within one year of birth, and congenital
malformations. In addition to the three studies evaluating effects of paternal occupational
exposure, one general population pregnancy cohort (Peng etal.. 2018) examined fetal growth
This document is a draft for review purposes only and does not constitute Agency policy.
3-283 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofHexavalent Chromium
markers but was limited due to exposure measurement of total chromium in urine with no
additional information to inform Cr(VI) exposure specifically. In addition, two ecologic studies
examined associations based on proximity to a Cr(VI) contaminated site (kilometers from center of
polluted area in Eizaguirre-Garcia etal. (20001. primarily affected town vs. rest of county in Remv
etal. (201711. The developmental effects examined in these studies included spontaneous abortion,
early pregnancy loss (not defined), pregnancy complications, and infant health (Remv etal.. 20171
and congenital malformations/anomalies (Remv etal.. 2017: Eizaguirre-Garcia etal.. 20001.
Table 3-45. Summary of human studies for Cr(VI) developmental effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.3
Click to see interactive data graphic for rating rationales.
Author (year)
Industry
Location
Study design
Spontaneous
abortion
Preterm
birth
Fetal
growth
Other (infant
death, congenital
malformations)
Eizaguirre-
General
population
Scotland
Ecologic
L
Garcia et al.
(2000)
Hiollund et al.
SS Welding
Denmark
Cohort
(occupational)
L
L
L
L
(1995), JP etal.
(1992)b
Hiollund et al.
SS Welding
Denmark
Cohort
(occupational)
L
—
—
—
(2000)
Hiollund et al.
SS Welding
Denmark
Retrospective
cohort
L
—
L
—
(2005)
Peng et al.
General
population
China
Pregnancy
cohort
—
—
—
(2018)
Remv et al.
General
population
U.S.
Ecologic
L
L
—
L
(2017)
SS = stainless steel.
aln addition to these included studies, four additional studies reported developmental outcomes that met PECO
criteria but were found to be uninformative at the study evaluation stage: Quansah and Jaakkola (2009). Xia et al.
(2016). Chen etal. (1997). and Ren et al. (2003).
bOne study was described in two publications (Hiollund et al., 1995: JP et al., 1992) that reported different but
overlapping subsamples. Results from both are described in the text but their results are not considered
independent of each other.
Synthesis of evidence in humans
Spontaneous abortion
Four studies examined associations between spontaneous abortion and Cr(VI) exposure.
Spontaneous abortion is pregnancy loss occurring before approximately 28 weeks gestation and
can be subdivided into early loss (loss before pregnancy is recognized) and clinical loss (loss after
5 weeks gestation) fRadke etal.. 20191. Methods of spontaneous abortion ascertainment can vary
in their ability to identify early losses. When early losses are not detected, there is potential for bias
if a true association with the exposure exists. This can even result in an apparent protective effect.
This document is a draft for review purposes only and does not constitute Agency policy.
3-284 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
In the four available studies, two were designed to ascertain early losses. Hiollund etal. f20001
used daily urine samples to identify pregnancy and early losses, which is the ideal approach.
Hiollund etal. (20051 used registry data from the Danish In Vitro Fertilization Register, which
includes information on clinical pregnancy identification. While this approach may not be as
sensitive as daily urine samples, it is likely that women were closely monitored and pregnancies
were identified early in this population. The other two studies (Remv etal.. 2017: Hiollund etal..
19951 identified spontaneous abortions based on hospital discharge data, which would be limited
to clinical losses, and only those in women who sought medical attention.
Hiollund etal. f2 0001 reported a statistically significant increased risk of spontaneous
abortion with paternal stainless-steel welding (RR = 3.5, 95% CI: 1.3-9.1), which was specific to this
exposure group (i.e., no increase was observed with mild steel welding exposure)51. Conversely,
Hjollund etal. (2005) and Hiollund et al. (1995) reported inverse associations (statistically
significant in Hiollund etal. f200511. although a different analysis of the population in the latter
study (TP etal.. 1992) reported a positive association (OR = 1.9, 95% CI: 1.1-3.2). However, in this
latter analysis, spontaneous abortion was based on registry data providing the number of
spontaneous abortions preceding each birth recorded in the national registry, and this measure
was considered to be less sensitive than measures in other studies. In addition, in TP etal. (1992).
there were similarly higher odds for induced abortion (OR = 2.1, 95% CI: 1.2-3.4), which increases
uncertainty about the reliability of the estimate since there is limited plausibility for Cr(VI) to
influence induced abortions (currently limited data exists on the association between Cr(VI) and
birth defects, as described below). A low confidence ecologic study (Remv etal.. 2017) also
reported higher relative risk of spontaneous abortion with higher exposure (RR 1.80, 95% CI: 1.20,
2.68). Overall, there is some indication that Cr(VI) exposure is associated with spontaneous
abortion, most notably in Hiollund etal. (2000). which had outcome ascertainment methods best
able to ascertain early losses. It is possible that the inverse associations observed in Hiollund etal.
(1995) were due to early losses missed by their outcome ascertainment methods, but there is not
adequate data to assess this. However, given the small number of studies and the limited nature of
the evidence there is considerable uncertainty.
Fetal growth, preterm birth, and infant death
Three studies (Peng etal.. 2018: Remv etal.. 2017: TP etal.. 1992) examined associations
with fetal growth outcomes, though in Remv etal. (2017) the association was reported for a
combination of outcomes that also included preterm birth. Peng etal. (2018) examined birth
weight, length, and ponderal index, as well as fetal ultrasound measurements of head and
abdominal circumference and femur length in all three trimesters. There were statistically
51As noted in Section 3.1.1.2, highly soluble Cr(VI) may be more rapidly absorbed by the lungs and
transported to the bloodstream than Cr(VI) compounds that are less soluble. Cr(VI) components of stainless
steel welding fume are significantly more water soluble than for mild steel welding, and may cause more
persistent and greater inflammatory responses (Shoeb et al.. 20171.
This document is a draft for review purposes only and does not constitute Agency policy.
3-285 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
significant decreases in ponderal index with increased exposure, and non-statistically significant
decreases in birth weight and fetal head and abdominal circumference and femur length (in the
third trimester only). TP etal. (19921 reported no association with low birthweight. Remv et al.
(20171 reported higher relative risk for preterm birth, low birthweight, and small for gestational
age combined (RR 1.14, 95% CI: 1.05,1.25). Thus, there is some indication of fetal growth
restriction with Cr(VI) exposure, but there is considerable uncertainty as the exposure in Peng et al.
(20181 was total chromium and Remv etal. (20171 also included preterm birth, both of which
reduce the interpretability of the findings.
In addition, TP etal. T19921 reported on preterm birth and infant death within the first year.
They reported a non-statistically significant association between higher Cr(VI) exposure levels and
increased odds of preterm birth (OR = 1.3, 95% CI: 0.9-1.9). No association was observed for infant
mortality, but the lack of association could be due at least in part to poor sensitivity as above. In
addition to the preterm birth results already discussed, Remv etal. f20171 reported higher relative
risk for perinatal jaundice (RR 1.13, 95% CI: 1.06,1.20) and some infant health conditions
(infectious/parasitic, nervous system). While both studies reported associations with preterm
birth, this was analyzed in a combined outcome in Remv etal. f20171. which again makes it difficult
to interpret. The other outcomes were observed in a single low confidence study.
Congenital malformations
Three studies examined the association between Cr(VI) exposure and congenital
malformations fRemv etal.. 2017: Eizaguirre-Garcia etal.. 2000: TP etal.. 19921. In TP etal. f!9921.
there was no association between paternal occupational exposure and congenital malformations.
In Eizaguirre-Garcia etal. (20001. risk of congenital malformations was lowest in areas closest to
the center of the polluted area. In Remv etal. (20171. there was higher relative risk of eye, ear, face,
neck, and cleft anomalies in the higher exposed geographic area (RR 1.19, 95% CI: 0.91,1.56), but
this was only observed in one of the two time periods studied. No increase in genitourinary
anomalies was observed. Overall, there is limited evidence of an association between congenital
malformations and Cr(VI) exposure. However, all of the available studies had serious limitations
which limits interpretation of their results.
In summary, there are some indications of an association between Cr(VI exposure and
spontaneous abortion, fetal growth, preterm birth, and congenital malformations, but the evidence
is limited in quality and quantity.
3.2.9.2. Animal Evidence
Study evaluation summary
Table 3-46 summarizes the animal toxicology studies considered in the evaluation of the
developmental effects of Cr(VI). These consist of a continuous breeding study using NTP's
Reproductive Assessment by Continuous Breeding (RACB) protocol (NTP. 19971: studies that
This document is a draft for review purposes only and does not constitute Agency policy.
3-286 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofHexavalent Chromium
evaluated effects in F1 offspring following maternal-only exposure fKanoiia etal.. 1998: Elbetieha
and Al-Hamood. 19971 or paternal-only exposure fMaratetal.. 2018: Al-Hamood et al.. 1998:
Bataineh etal.. 1997: Elbetieha and Al-Hamood. 19971 prior to mating; and studies that evaluated
F1 offspring from dams that were exposed during gestation (Sivakumar etal.. 2022: Navin etal..
2021: Shobana et al.. 2 0 2 0: Zheng etal.. 2018: Arshad etal.. 2017: Banu etal.. 2017a: Banu etal..
2017b: Kumar etal.. 2017: Shobana et al.. 2 017: Banu etal.. 2015: Sivakumar etal.. 2014: Samuel et
al.. 2012a: Bataineh etal.. 2007: De Flora etal.. 2006: Elsaieed and Nada. 2002: Tunaid etal.. 1996b.
1995: Trivedi etal.. 19891 or lactation fSanchez and Ubios. 2021. 2020: Banu et al.. 2016: Sanchez
etal.. 2015: Stanley etal.. 2014: Stanley etal.. 2013: Banu etal.. 20081. All studies were oral
exposures (diet, drinking water, or oral gavage), although exposure to offspring was indirect in all
studies except the RACB study.
The RACB study by NTP (19971 and the gestational exposure study by Zheng etal. (20181
were well-reported and well-designed to evaluate effects in developing animals and therefore were
rated as high confidence for all reported outcomes. The studies by De Flora etal. (20061 and
Shobana etal. f20171 had minor concerns raised during study evaluation and were rated medium
confidence. The remaining studies had reporting limitations and other substantial concerns and
were rated as low confidence across all outcomes. Endpoint-specific concerns are discussed in the
respective sections below. Three of the low confidence studies (Al-Hamood etal.. 1998: Bataineh et
al.. 1997: Elbetieha and Al-Hamood. 19971 exposed animals to high concentrations (350-
1770 mg/L) of Cr(VI) in drinking water, which was considered a potential confounding variable as
it is not possible to determine whether developmental effects may have been exacerbated by
reduced water consumption and/or systemic toxicity; for instance, drinking water concentrations
of 350 mg/L Cr(VI) have been associated in rats with decreased water consumption and site of
contact toxicity (80 and 100% incidence of ulcers in the glandular stomach of males and females,
respectively) (NTP. 20071. There were concerns about scientific integrity for two groups of
authors52 (Banu etal.. 2017a: Banu etal.. 2017b: Kumar etal.. 2017: Samuel et al.. 2 012 a: Kanoiia et
al.. 1998: Tunaid etal.. 1996b. 19951. which reduces confidence in these studies but does not
necessarily discount the results.
52Four studies demonstrating self-plagiarism—i.e., publication of identical data presented as separate and
unique experiments—were considered critically deficient and were excluded from the assessment.
Specifically, 1] identical data were presented for rats by Kanoiia et al. (19961 and for mice by lunaid et al.
(1996a). despite these being presented as separate studies in different species; and 2) subsets of the data
presented by Samuel et al. (2012b: 20111 were identical to that in an earlier publication by this laboratory
group fBanu etal.. 20081. Other studies by the same groups of authors, listed in the text above, were included
in the assessment but considered low confidence.
This document is a draft for review purposes only and does not constitute Agency policy.
3-287 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-46. Summary of included studies for Cr(VI) developmental effects and
overall confidence classification [high (H), medium (M), low (L)] by outcome.3
Click to see interactive graphic for ratings rationale.
Author (year)
Species (strain)
Exposure life stage and
duration
Exposure
route
Survival
Growth
Structural alterations
Placenta
Functional effects
Al-Hamood et al.
(1998)
Mice (BALBC)
F1 males or females
exposed GD 12-PND 20
and mated with untreated
animals
Drinking
water
L
L
Arshad et al. (2017)
Mice (Swiss-
Webster)
GD 6
Gavage
L
L
L
-
-
Banu et al. (2008)
Rat (Wistar)
F1 females; PND 1-21
Drinking
water
-
-
-
-
L
L
Banu et al. (2015)
Rat (Sprague-
Dawley)
F1 females; GD 9.5-14.5
Drinking
water
-
-
-
-
Banu et al. (2016)
Rat (Sprague-
Dawley)
F1 females; PND 1-21
Drinking
water
-
-
-
-
L
Banu et al. (2017b)
Rat (Sprague-
Dawley)
GD 9.5-14.5
Drinking
water
-
L
-
-
-
Banu et al. (2017a)
Rat (Sprague-
Dawley)
GD 9.5-14.5
Drinking
water
-
-
-
L
-
Bataineh et al.
(1997)
Rat (Sprague-
Dawley)
FO males exposed 12
weeks prior to mating with
untreated females
Drinking
water
L
L
Bataineh et al.
(2007)
Rat (Sprague-
Dawley)
GD 1-3 or 4-6
Gavage
-
-
-
-
De Flora et al. (2006)
Mice (Swiss albino)
"Duration of pregnancy"-
GD 18
Drinking
water
M
M
-
-
-
Elbetieha and Al-
Hamood (1997)
Mice (Swiss)
F0 males or females
exposed 12 weeks prior to
mating with untreated
animals
Drinking
water
L
Elsaleed and Nada
(2002)
Rat (Wistar)
GD 6-15
Drinking
water
L
L
L
L
-
Junaid et al. (1995)
Mice (Swiss albino)
GD 14-19
Drinking
water
L
L
L
L
-
Junaid et al. (1996b)
Mice (Swiss albino)
GD 6-14
Drinking
water
L
L
L
L
-
This document is a draft for review purposes only and does not constitute Agency policy.
3-288 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Author (year)
Species (strain)
Exposure life stage and
duration
Exposure
route
Survival
Growth
Structural alterations
Placenta
Functional effects
Kanoiia et al. (1998)
Rat(Druckrey)
F0 females exposed 3
months prior to mating
with untreated males
Drinking
water
L
L
L
L
Kumar et al. (2017)
Rat (Wistar)
GD 9-14
Drinking
water
-
L
-
-
L
Marat et al. (2018)
Rat (white outbred)
Adult males; 60 days
Gavage
L
-
-
-
-
Navin et al. (2021)
Rat (Wistar)
F1 offspring; GD 9-14
Drinking
water
-
-
-
-
L
NTP (1997)
Mice (BALBC)
Reproductive Assessment
by Continuous Breeding
(F0 to F2)
Diet
H
H
H
Samuel et al. (2012a)
Rat (Wistar)
Study 1: GD 9-21
Study 2: GD 9-PND 65
Drinking
water
L
L
-
-
L
Sanchez et al. (2015)
Rat (Wistar)
PND 4-19
Gavage
-
-
L
-
-
Sanchez and Ubios
(2020)
Rat (Wistar)
PND 4-9, PND 4-15, or
PND 4-23
Gavage
-
-
L
-
-
Sanchez and Ubios
(2021)
Rat (Wistar)
PND 4-9 or PND 4-15
Gavage
-
-
L
-
-
Shobana et al. (2017)
Rat (Wistar)
GD 9-14
Drinking
water
-
-
-
-
M
(Shobana et al.,
2020)
Rat (Wistar)
F1 offspring; GD 9-14 or
GD 15-21
Drinking
water
-
-
-
-
L
Sivakumar et al.
(2014)
Rat (strain not
reported)
F0 dams; GD 9.5-14.5
Drinking
water
-
-
-
-
L
Sivakumar et al.
(2022)
Rat (Sprague-Dawley)
F0 dams; GD 9.5-14.5
Drinking water
-
-
-
-
L
Stanlev et al. (2013)
Rat (Sprague-
Dawley)
F1 females; PND 1-21
Drinking
water
-
-
-
-
L
Stanley et al. (2014)
Rat (Sprague-
Dawley)
F1 females; PND 1-21
Drinking
water
-
-
-
-
L
Trivedi et al. (1989)
Mice (albino)
GD 0-19
Drinking
water
L
L
L
L
-
Zheng et al. (2018)
Rat (Sprague-
Dawley)
GD 12-21.5
Gavage
H
H
-
-
H
GD = gestation day; PND = postnatal day.
aln addition to these included studies, there were seven animal toxicology studies reporting female reproductive
effects that met PECO criteria but were found to be uninformative at the study evaluation stage: Junaid et al.
This document is a draft for review purposes only and does not constitute Agency policy.
3-289 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Toxicological Review ofHexavalent Chromium
(1996a), Kanoiia et al. (1996), Soudani et al. (2011b), Soudani et al. (2011a), Soudani et al. (2013), Zahid et al.
(1990), and Borneff et al. (1968).
Synthesis of evidence in animals53' ^
Fetal and postnatal survival
Decreased offspring survival was observed only in low confidence studies. Statistically
significant effects occurred at the same dose or lower compared to decreased maternal body weight
gain or clinical signs of maternal toxicity within a subset of studies that reported both maternal and
fetal endpoints fElsaieed and Nada. 2002: Kanoiia etal.. 1998: Tunaid etal.. 1996b. 1995: Trivedi et
al.. 19891. Other low confidence studies provided little or no data on maternal toxicity, so the
relative sensitivity of maternal and offspring effects could not be compared in those cases.
In the high confidence RACES study in mice (NTP. 19971 there was no effect on the number
of live pups per litter or proportion of pups born alive across the F1 and F2 litters at dietary doses
up to 30.3 mg/kg-day Cr(VI) (F0 parental animals) and 37.1 mg/kg-day Cr(VI) (F1 parental
animals), and no effects on survival of F1 from birth until weaning at PND 21. The high confidence
gestational exposure study by Zheng etal. (20181 also reported no effects on rat pup numbers or
sex ratio (% male pups) following maternal exposure at doses up to 12 mg/kg-d Cr(VI) via oral
gavage from GD 12-21. The medium confidence gestational exposure study by De Flora et al.
(20061 reported no effect on the number of fetuses at GD 18 following maternal exposure to 5 or
10 mg/L Cr(VI) in drinking water throughout the duration of pregnancy.
In contrast to the findings in high and medium confidence studies, all low confidence studies
that exposed dams to Cr(VI) during pregnancy reported increased pre- or post-implantation loss.
Rat dams dosed with 25 mg/kg-day potassium dichromate via oral gavage from GD 1-3 had no
implantations Bataineh etal. (20071: and a dose-related increase in pre-implantation loss was
observed in mice exposed from GD 0-19, reaching statistical significance at 177 mg/L Cr(VI)
(Trivedi etal.. 19891. Statistically significant increases in pre-implantation loss were also reported
in rats exposed to 50 mg/L potassium dichromate in drinking water from GD 6-15 fElsaieed and
Nada. 20021. and a dose-related decrease in implantation index (number of implantation sites /
number of corpora lutea) was reported in rats exposed to 50-400 mg/L Cr(VI) from GD 9-21
fSamuel etal.. 2012al: however, these exposures began around or after the time of implantation in
rats (generally GD 6) and therefore effects may not have been related to treatment fU.S. EPA. 19911.
Statistically significant dose-related increases in post-implantation loss (resorptions or dead
fetuses) were observed in mice following exposure from GD 0-19 (Trivedi etal.. 19891. GD 6-14
53Data are available in HAWC for NTP ("19971 here.
54For many of the oral studies presented here, it was not possible to estimate an average daily mg/kg dose
due to lack of reporting. To estimate an average daily dose, paired records of body weight and daily intake of
test article are required. This is particularly important for Cr(VI) reproductive and developmental studies,
because rapid changes in maternal body weight are expected during pregnancy, and Cr(VI) affects palatability
(which affects both Cr(VI) intake rate and body weight). Doses of Cr(VI) are presented where possible,
however many cross-study comparisons are done on the basis of mg/L Cr(VI) in drinking water.
This document is a draft for review purposes only and does not constitute Agency policy.
3-290 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
fTunaid etal.. 1996bl. and GD 14-19 fTunaid etal.. 19951. reaching statistical significance at 88 or
177 mg/L Cr(VI). In studies that tested a single dose level, post-implantation loss was increased in
rats following exposure to a maternal dose of 50 mg/L Cr(VI) in drinking water from GD 6-15
(Elsaieed and Nada. 20021. in rats given a maternal dose of 8.8 mg/kg-day Cr(VI) via oral gavage
from GD 4-6 fBataineh etal.. 20071. and in mice given a maternal dose of 3.9-16 mg/kg Cr(VI) via
oral gavage on GD 6 fArshad etal.. 20171. The studies by Arshadetal. f20171 and Samuel etal.
f2012al presented results in terms of the number of individual fetuses affected without indication
of means or variance across litters, so there is greater uncertainty in the results of these studies.
Three low confidence studies reported decreased fetal survival when maternal animals
were exposed to Cr(VI) prior to mating. Kanoiia etal. (19981 exposed rat dams to Cr(VI) via
drinking water for 3 months prior to mating with unexposed males and reported a dose-related 2-
to 3.1-fold increase in pre-implantation loss and a 2.2- to 4.2-fold increase in post-implantation loss,
reaching statistical significance at 88 mg/L Cr(VI). Elbetieha and Al-Hamood T19971 exposed adult
F0 female mice to Cr(VI) in drinking water for 12 weeks prior to mating with unexposed males, and
reported a 17-18% decrease in implantations, ~5-6 fold increase in the number of mice with
resorptions, and a 25-32% decrease in viable fetuses, all of which were statistically significant at
both of the tested doses [707 and 1,768 mg/L Cr(VI)]. (Al-Hamood et al.. 1998) exposed F1 female
mice to maternal doses of 353 mg/L Cr(VI) in drinking water during development (from GD 12-
PND 20) and then mated these animals with unexposed males as adults, and reported a statistically
significant 12% decrease in implantations and 14% decrease in viable fetuses.
Male-mediated decreases in fetal survival were observed in two low confidence paternal-
only exposure studies. Elbetieha and Al-Hamood (1997) reported a statistically significant 16-23%
decrease in implantations and viable fetuses when adult F0 male mice were exposed to 707 or
1,414 mg/L Cr(VI) in drinking water for 12 weeks prior to mating with untreated females; these
effects were not observed at the 353 or 1,768 mg/L dose levels, although some resorptions or dead
fetuses were noted. Marat etal. (2018) exposed adult F0 male rats to 0.353 mg/kg-day Cr(VI) via
oral gavage for 60 days prior to mating with untreated females and reported a 1.8-fold increase in
pre-implantation loss, an 8.9-fold increase in post-implantation loss, and a dominant lethal
mutation frequency of 0.665. There were no effects on the number of implantation sites and viable
fetuses in two other low confidence paternal exposure studies, both of which exposed parental
males to a dose level of 353 mg/L Cr(VI) in drinking water during development (Al-Hamood etal..
19981 or as adults fBataineh et al.. 19971 and mated with unexposed females.
Fetal and postnatal growth
Decreased fetal or postnatal growth were observed to some extent in almost all studies that
evaluated these outcomes. Statistically significant effects occurred at the same dose or lower
compared to decreased maternal body weight gain or clinical signs of toxicity within a subset of
studies that reported both maternal and fetal endpoints (Elsaieed and Nada. 2002: Kanoiia etal..
This document is a draft for review purposes only and does not constitute Agency policy.
3-291 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
1998: NTP. 1997: Tunaid etal.. 1996b. 1995: Trivedi etal.. 19891. Other low confidence studies
provided little or no data on maternal toxicity, so the relative sensitivity of maternal and offspring
effects could not be compared in those cases.
In the high confidence RACES study in mice, mean F1 male and female pup body weights in
the highest dose group [F0 dietary exposure of 30.3 mg/kg-day Cr(VI)] were similar to controls at
birth but were 9-15% lower than controls at PNDs 14 and 21, although this effect was not
statistically significant55. By PND 74 ± 10, the effect on F1 body weights was statistically significant;
mean F1 male and female body weights in the highest dose group [37.1 mg/kg-day Cr(VI)] were
decreased by 9% compared to controls, and F1 females in the second highest dose group
[16.1 mg/kg-day Cr(VI)] were decreased by 4% compared to controls fNTP. 19971. Food
consumption was increased in the treated animals compared to controls, so the decrease in growth
does not seem to be attributable to palatability or changes in feed consumption. There was a
statistically significant 11% decrease in F2 female pup birth weights at 37.1 mg/kg-day Cr(VI),
although pup body weights in this group were not statistically significantly lower than controls
when adjusted for litter size. Otherwise, there were no effects on F2 pup birth weights, and F2
animals were not monitored further.
The remaining studies that observed decreased F1 growth were considered low confidence.
Kanoiia et al. (19981 exposed rat dams via drinking water for 3 months prior to mating and
reported a dose-related 21-36% decrease in fetal body weight, reaching statistical significance at
88 mg/L Cr(VI). In drinking water studies that exposed pregnant dams, fetal body weights were
decreased in a dose-related manner compared to controls by 18-47% (Tunaid etal.. 19951. 3-19%
(Tunaid etal.. 1996bl. and 32-44% (Trivedi etal.. 19891 following exposure from GDs 14-19, 6-14,
and 0-19, respectively, reaching statistical significance at 88 or 177 mg/L Cr(VI). Two studies that
exposed pregnant dams to 50 mg/L Cr(VI) observed that fetal body weights were statistically
significantly decreased compared to controls by 33% following maternal exposure from GD 6-14
(Elsaieed and Nada. 20021 and by 31% following maternal exposure from GD 9.5-14.5 (Banu etal..
2017b)56. One study that exposed pregnant mice on GD 6 via oral gavage reported that fetal body
weights were decreased by 17-27% compared to controls, reaching statistical significance at
22 ug/g potassium dichromate (Arshad etal.. 20171. Three of the gestational exposure studies also
reported decreased crown-rump length (Arshad etal.. 2017: Tunaid etal.. 1995: Trivedi etal..
19891. and the study by Arshad etal. (20171 reported decreased morphometric parameters
including head and eye circumference, and fore limb, hind limb, and tail length. In two studies that
assessed postnatal growth, Kumar etal. (20171 reported a dose-related statistically significant
11-20% decrease in body weight at PND 120 in F1 male rats from dams that had been exposed to
35.3 or 70.7 mg/L Cr(VI) in drinking water from GD 9-14, and Samuel etal. f2012al reported a
55Data are available for males (PND14 and PND211 and females (PND14 and PND211.
56Fetal body weights in Banu etal. f2017bl were reported graphically, but were estimated using
WebPlotDigitizer to be 2.64 ± 0.01 g in the control group and 1.82 ± 0.14 g in the Cr(VI] exposure group.
This document is a draft for review purposes only and does not constitute Agency policy.
3-292 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
statistically significant 33-41% decrease in body weights on PNDs 3, 7,18, 45, and 65 in F1 female
rats that had been continuously exposed to 200 mg/L Cr(VI) in drinking water from GD 9-PND
6 5 57. The studies by Banu etal. (2017b). Kumar etal. (20171. and Samuel etal. (2012a) reported
body weights as the mean of individual offspring without accounting for litter effects, and it was not
clear whether results in the studies by Elsaieed and Nada (20021 or Arshad etal. (20171 were litter
means or the means of individual animals; this affects interpretation of the results in these studies,
as failure to consider litter effects has the potential to overestimate statistical significance
fHaseman et al.. 20011.
Three studies reported no effect on F1 growth. The high confidence study in rats by Zheng
etal. (20181 reported no change in newborn pup body weight following maternal exposure at doses
up to 12 mg/kg-d Cr(VI) via oral gavage from GD 12-21. The medium confidence study in mice by
De Flora et al. (20061 reported no change in fetal body weight at GD 18 following maternal
exposure to 5 or 10 mg/1 Cr(VI) in drinking water throughout the duration of pregnancy. The low
confidence study by Al-Hamood et al. (19981 reported no effects on male or female body weight at
PND 50 in F1 mice that had been exposed to maternal doses of 353 mg/L Cr(VI) in drinking water
from GD 12-PND 20.
Structural alterations
A dose-related increase in structural alterations was reported in all studies that evaluated
these outcomes in fetuses or early postnatal animals, which consisted of low confidence studies.
Statistically significant effects occurred at the same dose or lower compared to decreased maternal
body weight gain or clinical signs of toxicity within a subset of studies that reported both maternal
and fetal endpoints fElsaieed and Nada. 2002: Kanoiia etal.. 1998: Tunaid etal.. 1996b. 1995:
Trivedi etal.. 19891. whereas the other two studies did not provide data on maternal toxicity.
Within studies, reduced ossification occurred at doses concurrent with decreased fetal growth
(body weight or morphometric parameters) and was mostly observed in bones that undergo rapid
ossification at the end of gestation (e.g., parietals, interparietals, caudal, frontals). This may indicate
that the delay in ossification is indicative of a generalized growth delay (Carney and Kimmel. 20071.
Four low confidence studies by the same research group evaluated fetuses at GD 19.
Reduced skeletal ossification was observed when F0 rat dams were exposed to potassium
dichromate in drinking water for 3 months prior to mating (Kanoiia etal.. 19981 and when F0
mouse dams were exposed to potassium dichromate in drinking water from GD 0-19 (Trivedi et al..
19891. GD 6-14 (Tunaid etal.. 1996b). or GD 14-19 (Tunaid etal.. 19951. Skeletal effects across
these studies reached statistical significance at levels as low as 88 mg/L Cr(VI). Trivedi et al.
f!9891 also reported that fetuses had decreased number of ribs, which reached statistical
57F1 body weights in Kumar et al. (20171 and Samuel etal. (2012a) were reported graphically and were
estimated using WebPlotDigitizer. The difference in body weights between control and Cr(VI)-exposed
animals on PND 3 in the study by Samuel etal. (2012 a) could not be estimated using WebPlotDigitizer due to
the scale of the figure, so the values shown are for PNDs 7,18,45, and 65.
This document is a draft for review purposes only and does not constitute Agency policy.
3-293 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
significance at 177 mg/L Cr(VI). In addition to skeletal effects, these four studies each reported the
same gross abnormalities (drooping wrist, subdermal hemorrhagic patches, kinking tail, short tail)
and reported that the exposed animals did not have any visceral alterations. Trivedi et al. (19891
also reported "enlarged gap between fingers."
The remaining low confidence studies that evaluated fetal structural alterations have
greater uncertainty due to incomplete reporting of results. Elsaieed and Nada (20021 reported a
statistically significant increase in skeletal and visceral abnormalities in fetuses from F0 rat dams
that were exposed to 50 mg/L Cr(VI) in drinking water from GD 6-15 and euthanized on the day
before delivery, and noted that some animals had incomplete ossification of the skull bone and
increased renal dilation; however, data were reported as the average total skeletal and visceral
abnormalities per litter with no quantitative incidence data provided for specific alterations.
Arshad etal. (2017) reported numerous skeletal and visceral abnormalities in mouse fetuses from
dams that were dosed on GD 6 with 3.8-16 mg/kg Cr(VI) via oral gavage and euthanized on GD 18,
including reduced skeletal ossification; however, most of these abnormalities were described
qualitatively with no information provided on relative incidence. Quantitative incidence data was
provided for some abnormalities (anophthalmia, limb hyperextension, limb hyperflexion, limb
malrotation, limb micromelia, and spina bifida) but was reported as the total number of individual
fetuses affected without indication of potential litter effects. A series of studies by Sanchez and
coauthors evaluated periodontal bone development in rats dosed with 4.4 mg/kg-day Cr(VI) via
oral gavage during lactation and reported delayed tooth eruption, delayed mineralization,
decreased periodontal width and bone volume, and decreased bone resorption and formation
surfaces (Sanchez and Ubios. 2021. 2020: Sanchez etal.. 2015). The publications did not report pup
body weights, but the authors clarified via personal correspondence that pup weights were
decreased in the experimental groups compared to controls, which may suggest that the effects on
tooth development are related to a generalized growth delay. The authors provided quantitative
results for some histomorphometric parameters but concerns were raised due to the lack of
blinding in the analysis and the small sample size (one litter used per experimental group).
Effects on the placenta
Effects on the placenta were evaluated in several low confidence studies that exposed dams
to Cr(VI) prior to or during gestation. Placental effects occurred at the same doses as decreased
maternal body weight gain in three of these studies that provided both maternal and fetal data
(Elsaieed and Nada. 2002: Kanoiia etal.. 1998: Tunaid etal.. 1995). whereas the studies by Tunaid et
al. (1996b) and Trivedi etal. (1989) reported decreased maternal body weight gain but no effect on
placenta weights.
Two low confidence studies evaluated placental histopathology. In rat dams exposed to 50
mg/L potassium dichromate in drinking water from GD 6-15, Elsaieed and Nada (2002) reported
histologic lesions in the placenta including necrosis in the chorionic villi and focal extravasation of
This document is a draft for review purposes only and does not constitute Agency policy.
3-294 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
red blood cells in the decidua basalis. Banu etal. f2017al reported histologic effects including
increased hypertrophy and hemorrhagic lesions in the basal zone in rat dams exposed to 17.7 mg/L
Cr(VI) in drinking water from GD 9.5-14.5. Neither of these studies provided quantitative data on
the incidence or severity of these lesions, so interpretation of these findings is limited.
Changes in placenta weight were also observed in low confidence studies, although the
direction of effect was inconsistent across studies. Rat dams exposed to potassium dichromate in
drinking water for 3 months prior to mating had statistically significantly decreased placenta
weights in the 177 and 265 mg/L Cr(VI) dose groups in the study by Kanoiia etal. f!9981. whereas
a statistically significant dose-related increase in placenta weight was observed at exposure levels
>88 mg/L Cr(VI) in mouse dams exposed from GD 14-19 in the study by Tunaid et al. f!9951. In
other low confidence studies, no effects on placenta weight were observed in mouse dams exposed
to levels up to 265 mg/L Cr(VI) in drinking water from GD 6-14 (Tunaid et al.. 1996bl or up to
177 mg/L Cr(VI) in drinking water from GD 0-19 (Trivedi etal.. 19891.
Functional effects (reproductive, endocrine)
Effects on the developing reproductive system are described in the "Male reproductive
effects" and "Female reproductive effects" sections and summarized briefly here. Effects on F1
male and female fertility and histopathology were not observed in the high confidence RACE! study
(NTP. 19971 at doses up to 37.1 mg/kg-day Cr(VI) via diet, but were documented in several other
studies. In F1 male rats, a nonmonotonic effect on testosterone (increased at 3 mg/kg-day,
decreased at 12 mg/kg-day) and altered Leydig cell distribution were observed following maternal
exposure by oral gavage from GD 12-21 in the high confidence study by Zheng etal. (20181. A
series of low confidence studies by one laboratory group reported effects including decreased
sperm quality, histopathological changes in the testis, decreased testosterone and gonadotropins,
and decreased reproductive organ weights in F1 males exposed from GD 9-14 (Navin etal.. 2021:
Shobana etal.. 2020: Kumar etal.. 20171 or GD 15-21 (Shobana et al.. 20201 to maternal doses of
17.7-70.7 mg/L Cr(VI) in drinking water. In F1 female rats, a series of low confidence studies by
one laboratory group reported pathological effects on oocyte development following gestational
and/or postnatal exposure to maternal doses of 8.8-70.7 mg/L Cr(VI) in drinking water, as well as
decreased sex steroid hormone levels, increased gonadotropin levels, delayed puberty, and changes
in estrous cyclicity (Sivakumar etal.. 2022: Banu etal.. 2016: Banu etal.. 2015: Sivakumar et al..
2014: Stanley etal.. 2014: Stanley etal.. 2013: Samuel et al.. 2 012 a: Banu etal.. 20081. A low
confidence study in mice by Al-Hamood et al. (19981 likewise reported decreased pregnancy rates
and delayed puberty in F1 males that had been exposed to maternal doses of 353 mg/L Cr(VI) from
GD 12-PND 20. Interpretation of the low confidence studies is limited, due to the study design and
reporting concerns discussed in "Male reproductive effects" and "Female reproductive effects"
sections.
This document is a draft for review purposes only and does not constitute Agency policy.
3-295 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
Other evidence of functional effects in developing animals comes from a medium confidence
study that evaluated insulin signaling in F1 rats following maternal exposure to potassium
dichromate in drinking water from GD 9-14 (Shobana etal.. 20171. Serum insulin levels in pubertal
F1 rats evaluated on PND 59 were statistically significantly increased compared to controls at
maternal exposure levels >50 mg/L. Glucose uptake was increased in liver but decreased in
skeletal muscle, and glucose oxidation was increased in both liver and skeletal muscle at 50 mg/L
Cr(VI) but decreased at 100 and 200 mg/L Cr(VI). Despite these changes, there was no effect on
fasting blood glucose or oral glucose tolerance in these animals.
3.2.9.3. Mechanistic Evidence
Studies providing mechanistic evidence on the potential developmental effects of Cr(VI) are
tabulated in Appendix C.2.8 and summarized here. Together, these studies provide supporting
evidence that Cr(VI) may have adverse developmental effects if it were to reach the relevant target
tissues. The mechanistic studies reviewed here consisted of in vivo mechanistic data from several
oral exposure studies, most of which are discussed above (Table 3-46), as well as data from
intraperitoneal (i.p.) injection studies, in vitro studies in whole embryos, and in vitro studies in
trophoblast or osteoblast cell lines that did not meet PECO criteria but were reviewed as
informative to the mechanistic analysis. Dosing via i.p. injection is likely to result in higher tissue
concentrations of Cr(VI) compared to oral exposure, since an oral first-pass effect exists due to the
reduction of Cr(VI) in the low pH environment of the stomach; less than 10-20% of an ingested
dose may be absorbed in the GI tract, and further reduction will occur in the liver prior to
distribution to the rest of the body (see Section 3.1 and Appendix C). Therefore, systemic effects
are expected to be more likely following i.p. injection or inhalation compared to oral exposure.
Effects are also expected to be more likely in in vitro embryonic studies compared to in vivo
studies, since the in vitro studies incubated sperm or blastocytes directly with potassium
dichromate.
Fetal genotoxicity
One study assessed genotoxicity [measured as the frequency of micronucleated (MN)
polychromatic erythrocytes (PCE) in maternal bone marrow and fetal liver and peripheral blood] in
mice exposed to Cr(VI) salts during gestation via i.p. injection or oral exposure (De Flora etal..
20061. Fetuses from dams dosed orally via drinking water with sodium dichromate dihydrate (5 or
10 mg/1) or potassium dichromate (10 mg/1) did not have any changes in the frequency of MN PCE
compared to controls. In contrast, fetuses from dams given a single i.p. injection of 50 mg/kg
potassium dichromate or sodium dichromate dihydrate on GD 17 had significantly increased
frequency of MN PCE frequency in the liver and peripheral blood. The same pattern was observed
in maternal bone marrow. This study suggests that Cr(VI) is genotoxic to fetuses when it reaches
target tissues, although bioavailability is poor through the oral route of exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
3-296 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
In vitro evaluations of embryo development
Three studies in whole embryos provided evidence that Cr(VI) impairs embryonic
development One study incubated mouse sperm with potassium dichromate and used it to fertilize
eggs from untreated mice (Yoisungnern etal.. 20151. It was found that the percentage of
unfertilized oocytes and embryos in the 2-cell stage increased while the percentage in the expanded
and hatching blastocyst stages and total number of blastocysts were decreased, suggesting delays in
embryonic development. These effects were observed at the lowest dose level (1.1 [J.M Cr(VI)), and
differences became more pronounced with increasing doses, although higher doses also produced
statistically significant decreases in sperm viability. Blastocysts in the low dose group also had a
decrease in the number of trophectoderm and inner cell mass cells and decreased expression of
pluripotent marker genes (sox2, pou5fl, and klf4), indicating impaired development of the embryo
and placenta. A second study that incubated mouse blastocysts with potassium dichromate fliiima
etal.. 19831 found a dose-dependent decrease in 2-layer inner cell masses after 6 days of exposure
to 0.088-0.71 [iM Cr(VI), but statistically significant differences in hatching, attachment and
trophoblast outgrowths were not observed. Cultured embryos treated for 24 hours with
0.18-0.71|iM Cr(VI) showed statistically significant decreases in allantois fusion, beating hearts,
and blood islands. Decreased crown-rump length was also observed at doses of 0.35-0.71 [iM
Cr(VI). Additionally, a third study that collected mouse embryos at the 2-cell stage and incubated
them in culture with potassium dichromate or calcium chromate reported that Cr(VI) salts
inhibited blastocyst formation and hatching in a dose-dependent manner, with the high dose of
potassium dichromate (7.1 [iM Cr(VI)) arresting embryonic development at the 4-cell stage (Tacquet
and Drave. 19821.
Mechanisms affecting bone development
Several in vitro and in vivo studies identified mechanisms that are potentially relevant to
skeletal alterations and suggested oxidative stress as an underlying mechanism. In vitro studies
with immortalized rat osteoblasts show that Cr(VI) inhibits cell viability and decreases cellular
activity (protein, DNA, and RNA synthesis; production of collagen fibers) and found that effects
were mitigated by Vitamin C (ascorbic acid), which is an antioxidant fNing etal.. 2002: Ning and
Grant. 2000.19991.
Additionally, thyroid effects [decreased triiodothyronine (T3) and thyroxine (T4), and
follicle size and increased TSH concurrent with morphology changes] were observed in adult male
rats following injection with 21 |J.g/kg Cr(VI) and were partially prevented when animals were
pretreated or cotreated with ascorbic acid (Oureshi and Mahmood. 20101. Thyroid function is
important for skeletal developmental and disruption can result in delays in skeletal ossification;
however, the relevance of this finding to developing animals is unclear since this study was
conducted in adults.
This document is a draft for review purposes only and does not constitute Agency policy.
3-297 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
Mechanisms affecting insulin regulation
The gestational exposure study in rats by Shobana etal. (20171. described in the section
above, also provided mechanistic information relevant to insulin signaling. Insulin receptor protein
expression in liver and gastrocnemius muscle was decreased, suggesting negative feedback
resulting from increased insulin levels, and decreasing trends were observed in the expression of
insulin receptor substrate-1 (IRS-1) and its phosphorylated form (p-IRS-l^632) in these tissues. In
liver, the expression of the downstream signaling molecule Akt was unchanged while the
phosphorylated form (p-AktSer473) increased; whereas in gastrocnemius muscle, Akt expression
decreased and the effects on p-AktSer473 were nonmonotonic (increased at 50 mg/L Cr(VI) but
decreased at 100 mg/L Cr(VI)). GLUT 2 was increased in liver at 50 mg/L Cr(VI) and GLUT 4 was
decreased in gastrocnemius muscle at 200 mg/L, reflecting glucose uptake in these tissues. PPARy
expression in these tissues was increased, which the authors speculated may be involved in the
regulation of glucose transporters.
Oxidative stress and apoptosis in the placenta
Studies in humans, rats, and human cell lines provide supporting evidence for oxidative
damage and apoptosis in the placenta, as well as evidence that chromium reaches human placental
tissue. Placentae collected from healthy women in the general population showed average
chromium concentrations between 0.02 to 1.25 mg/L (Banu etal.. 20181. although these were total
chromium concentrations and it was unclear whether the women were exposed to Cr(VI) or
another form of Cr. Two biomarkers of oxidative stress in the samples with the highest average
chromium concentrations were statistically significantly increased over the lowest concentration
group and differences were also noted in the mRNA and protein expression of some antioxidants,
but there are uncertainties in the interpretation of this data; several apoptotic markers
(e.g., cytochrome C, AIF, Bax and cleaved caspase-3) were elevated in addition to anti-apoptotic
markers Bcl-2 and Bcl-XL, and some results showed sexually dimorphic differences fBanu etal..
20181.
Two studies evaluated placentae in rats administered 17.7 mg/L Cr(VI) in drinking water
during gestation. Banu etal. f2017bl performed immunohistochemical analysis demonstrating
decreased trophoblast cell populations and decreased expression of cyclin D1 in the placentas and
found that placentas of Cr(VI)-treated dams had increased biomarkers of oxidative stress (LPO and
H202) and decreased expression of antioxidant enzymes (SOD, Gpx, Prdx3, and Txn2). Banu et al.
f2017al reported increases in apoptosis and caspase-3 in the maternal compartment (metrial
gland) and the caspase-3 independent apoptotic marker AIF in both the fetal and maternal
compartments. Increases in p53 and related signaling cascade molecules were also observed.
Two studies evaluated placental cells in vitro. Banu etal. f20181 evaluated the human
trophoblastic cell line BeWo and observed a dose-related decrease in the mRNA expression of
antioxidant enzymes (SOD, Gpx, Prdx3 andTxn2) following dosing with 1.8-11 |iM Cr(VI) for 12-
This document is a draft for review purposes only and does not constitute Agency policy.
3-298 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
24 hours. Another in vitro study by Sawicka and Dtugosz f20171 observed increased lipid
peroxidation and decreased antioxidant enzyme activity (SOD, GST) in mitochondria isolated from
human placental tissue following treatment with 0.05-1 |J.g/mL Cr(VI). The increase in lipid
peroxidation and decrease in SOD were mitigated by cotreatment with an estradiol metabolite,
40HE2.
Gestational anemia
Pregnant women are at risk for developing gestational anemia due to the increased
production of blood that occurs during pregnancy (American Pregnancy Association. 20211.
Gestational anemia is associated with adverse developmental effects including low birth weight,
preterm birth, and perinatal and neonatal mortality (Figueiredo etal.. 2018: Rahman etal.. 20161.
Because the evidence suggests that Cr(VI) may produce anemia-like effects such as reduced
hematocrit, hemoglobin, MCV, MCH, and MCHC (see Section 3.2.5), exposure to Cr(VI) may
exacerbate the risk of developing anemia, with pregnant women being a potentially susceptible
subpopulation (see Section 3.3.1.1). Gestational anemia is therefore a potential mechanism for the
low birth weight and preterm birth that are associated with Cr(VI) exposure, although the
relationship between Cr(VI) exposure and gestational anemia has not yet been investigated.
3.2.9.4. Integration of Evidence
Overall, the available evidence indicates that Cr(VI) likely causes developmental effects in
humans. This conclusion is primarily based on the observation of decreased offspring growth
across most animal studies, as evidenced by decreased fetal or postnatal body weights and
decreased skeletal ossification. Other outcomes in animal studies are more uncertain because they
were inconsistent among high and medium confidence studies or were evaluated only in low
confidence studies. Likewise, the available human data were of low confidence and difficult to
interpret. Integrated evidence of the developmental effects of Cr(VI) exposure from human, animal,
and mechanistic studies is summarized in an evidence profile table (Table 3-47). The exposure
conditions relevant to these effects are further defined in Section 4.
The evidence of an association between Cr(VI) exposure and developmental effects in
humans is slight, with an indication of higher rates of spontaneous abortion with higher exposure
levels in two of four low confidence paternal occupational exposure studies and an ecologic study
with exposure evaluated at the zip code level (representing both maternal and paternal exposure).
Results for other outcomes, including preterm birth, fetal growth, infant death, and congenital
malformations indicated no clear association. The available evidence was all considered low
confidence and the studies generally had poor sensitivity, so there is considerable uncertainty in
this judgment.
Animal toxicology studies and supportive mechanistic data provide moderate evidence that
Cr(VI) exposure leads to developmental effects. The strength of evidence was greatest for effects
on fetal and postnatal growth, which were observed to some extent in the high confidence RACB
This document is a draft for review purposes only and does not constitute Agency policy.
3-299 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Toxicological Review ofHexavalent Chromium
study in mice by NTP T19971 as well as all low confidence studies that evaluated these outcomes.
The observation of reduced ossification within several low confidence studies appears to be
consistent with a generalized growth delay, although there is mechanistic evidence suggestive of
effects on osteoblasts or thyroid function that could also affect skeletal development Many studies
reported decreased fetal survival and functional effects on the developing reproductive system, but
there is more uncertainty in these findings because effects were observed primarily in low
confidence studies and were not recapitulated in the high confidence RACES study by NTP (19971
that evaluated effects through the F2 generation. Other outcomes had limited data available
(insulin regulation) or were only evaluated in low confidence studies (effects on the placenta) and
therefore also have greater uncertainty. Within studies that used a maternal route of exposure,
statistically significant effects on fetal development were observed at exposure levels the same or
lower than those that caused maternal toxicity. Most studies did not report maternal body weights
or other measures of overt toxicity, however, so maternal and fetal toxicity could not be compared
within those studies. Decreased fetal survival in paternal-only exposure studies in rats and mice
suggests dominant lethal mutations in sperm (as discussed in the "Male reproductive effects"
section) and is coherent with human paternal occupational exposure studies. There is more
uncertainty in these male-mediated findings because the human and animal studies were rated low
confidence and effects were not consistent across studies.
Postnatal growth in the RACES study by NTP (1997) was decreased in F1 animals at dose
levels of 16.1-37.1 mg/kg-day via diet, and birth weights in F2 females were decreased before
adjusting for litter size at 37.1 mg/kg-day Cr(VI). The doses of Cr(VI) at which effects were
observed in the low confidence drinking water studies in animal models could not be calculated
because drinking water consumption data was not reported, and none of the available human
studies provided a quantitative measure of exposure. There were no animal studies that evaluated
developmental effects following inhalation exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
3-300 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 3-47. Evidence profile table for developmental effects of Cr(VI)
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Evidence from human studies
®©o
The evidence indicates that
Cr(VI) likely causes
developmental effects in humans
given sufficient exposure
conditions.
Decreased offspring growth was
observed across most animal
studies; other effects were
inconsistent in higher confidence
studies, had limited data
available, or were only evaluated
in low confidence animal studies.
Coherence of spontaneous
abortions after paternal
occupation exposure in human
studies with decreased fetal
survival after paternal-only
exposure in animal studies;
however, only in low confidence
studies, and effects were not
consistent.
Mechanistic findings (animals
and in vitro) provide supporting
evidence of fetal genotoxicity,
impaired embryo and fetal
functional development, and
oxidative stress and apoptosis in
the placenta. These mechanisms
SPONTANEOUS
ABORTION
Low confidence
Hiollund et al. (1995)
JPetal. (1992)
Hiollund et al. (2000)
Hiollund et al. (2005)
Remv et al. (2017)
Two studies reported higher rates of
spontaneous abortion with higher Cr(VI)
exposure and two studies reported
lower rates (in one study, the effect
varied by analysis).
• Large effect size
(RR = 3.5) in one
study
• Left truncation
of early losses
could explain
inconsistent
results
• Low
confidence
studies
©oo
Slight
Based on
associations with
paternal
occupational
exposure and
spontaneous
abortion in the
study with the
most sensitive and
specific outcome
ascertainment
(Hiollund et al.,
2000).
OTHER
DEVELOPMENTAL
EFFECTS
Low confidence
JPetal. (1992)
Two studies reported positive
associations between Cr(VI) exposure
and preterm birth and decreased birth
size. Inconsistent associations reported
for congenital malformations.
• No factors noted
• Low
confidence
studies
Eizaguirre-Garcia et al.
(2000)
Peng et al. (2018)
Remv et al. (2017)
Evidence from animal studies
FETAL AND POSTNATAL
SURVIVAL
High confidence:
NTP (1997)
Zheng et al. (2018)
Medium confidence:
No effects on fetal survival (live pups) in
2 high and 1 medium confidence studies,
including NTP's RACB study in mice.
Increased pre- and/or post-implantation
loss in 10 low confidence studies in
which maternal animals were exposed
before mating or during gestation.
• No factors noted
• Effects
observed
only in low
confidence
studies
®©o
Moderate
Based primarily on
the observation of
decreased
offspring growth
This document is a draft for review purposes only and does not constitute Agency policy.
3-301 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
are presumed relevant to
humans.
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
De Flora et al. (2006)
Low confidence:
Al-Hamood et al. (1998)
Arshad et al. (2017)
Bataineh et al. (2007)
Effects were at doses same or lower than
those that caused maternal toxicity.
Increased pre- and/or post-implantation
loss in 2 out of 4 low confidence studies
in which only paternal animals were
exposed to prior to mating.
across most
studies, including
within the high
confidence RACB
in mice by NTP
(1997).
Bataineh et al. (1997)
Elbetieha and Al-
Hamood (1997)
Elsaieed and Nada
(2002)
Junaid et al. (1995)
Junaid et al. (1996b)
Kanoiia et al. (1998)
Marat et al. (2018)
Samuel et al. (2012a)
Trivedi et al. (1989)
FETAL AND POSTNATAL
GROWTH
High confidence:
NTP (1997)
Zheng et al. (2018)
Medium confidence:
De Flora et al. (2006)
Low confidence:
Al-Hamood et al. (1998)
Arshad et al. (2017)
Banu et al. (2017b)
Elsaieed and Nada
(2002)
Junaid et al. (1995)
Decreased F1 postnatal body weights in
NTP's high confidence RACB study.
Effects on F1 and F2 birth weights in this
study were minimal.
Decreased fetal or pup body weight and
other morphometric parameters
(e.g., crown-rump length) in 8 out of 9
low confidence studies.
Within all studies, effects were at same
or lower dose levels that those that
caused decreased maternal body weight
gain.
• High confidence
study
• Consistency
• Effect size
• Dose-response
gradient
• Coherence with
decreased
ossification
within low
confidence
studies
• No factors
noted
This document is a draft for review purposes only and does not constitute Agency policy.
3-302 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
Junaid et al. (1996b)
Kanoiia et al. (1998)
Kumar et al. (2017)
Samuel et al. (2012a)
Trivedi et al. (1989)
STRUCTURAL
ALTERATIONS
Low confidence:
Arshad et al. (2017)
Elsaieed and Nada
(2002)
Junaid et al. (1995)
Junaid et al. (1996b)
Kanoiia et al. (1998)
Trivedi et al. (1989)
Sanchez et al. (2015)
Decreased fetal skeletal ossification as
well as some other structural
abnormalities in low confidence studies,
occurring at the same dose levels as
decreased fetal growth.
Decreased periodontal bone formation
in rat pups exposed postnatally in three
low confidence studies by the same
group of authors, occurring at a dose
level that also caused decreased body
weight.
• Coherence of
decreased
ossification with
decreased
growth
• Low
confidence
studies
EFFECTS ON THE
PLACENTA
Low confidence:
Banu et al. (2017a)
Elsaieed and Nada
(2002)
Junaid et al. (1995)
Junaid et al. (1996b)
Kanoiia et al. (1998)
Trivedi et al. (1989)
Histopathological changes in the
placenta in 2 low confidence studies.
Inconsistent effects on placenta weight
across studies (increased, decreased or
no effect).
• No factors noted
• Low
confidence
studies
FUNCTIONAL
ENDPOINTS
High confidence:
NTP (1997)
Effects on developing male reproductive
system observed in 1 high confidence
study, and effects on developing female
reproductive system observed in
• Dose-response
gradient
• Unexplained
inconsistenc
y across high
This document is a draft for review purposes only and does not constitute Agency policy.
3-303 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Zheng et al. (2018)
Medium confidence:
Shobana et al. (2017)
Low confidence:
Al-Hamood et al. (1998)
Banu et al. (2008)
Banu et al. (2015)
Banu et al. (2016)
Kumar et al. (2017)
Navin et al. (2021)
Samuel et al. (2012a)
Shobana et al. (2020)
Sivakumar et al. (2022)
Sivakumar et al. (2014)
Stanley et al. (2013)
Stanley et al. (2014)
multiple low confidence studies. No
effects in NTP's RACB.
Increased serum insulin levels and
alterations in glucose uptake and glucose
oxidation in F1 rats that had been
exposed during gestation.
• Mechanistic
evidence
provides
biological
plausibility
confidence
studies
Mechanistic evidence
Biological events or
pathways
Summary of key findings and interpretations
Judgments and
rationale
Fetal genotoxicity
Interpretation: In vivo evidence of fetal genotoxicity.
Key findings:
• Increased frequency of fetal micronucleated polychromatic erythrocytes
when mouse dams were exposed via a single i.p. injection, but no effects
following repeat dose oral exposure (De Flora et al., 2006).
Observations of
multiple
mechanisms by
which Cr(VI) can
disrupt fetal
structural and
functional
development.
In vitro evaluations of
embryo development
Interpretation: In vitro evidence that Cr(VI) impairs or arrests embryo
development
Key findings:
This document is a draft for review purposes only and does not constitute Agency policy.
3-304 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
Inferences and summary
judgment
• Impaired embryo development when Cr(VI)-treated mouse sperm were
used to fertilize untreated eggs (Yoisungnern et al., 2015), or when mouse
blastocysts were incubated in solutions of Cr(VI) (liiima et al., 1983; Jacauet
and Drave, 1982).
Mechanisms affecting
bone development
Interpretation: In vitro evidence that Cr(VI) affects viability and activity of
osteoblasts, and in vivo evidence that Cr(VI) decreases thyroid hormone levels.
Key findings:
• Increased cvtotoxicitv (Ning et al., 2002; Ning and Grant, 2000,1999) and
decreased protein, DNA, RNA, and collagen fiber production (Ning et al.,
2002) in an immortalized osteoblast cell line.
• Decreased thyroid hormone levels and follicle size in adult male rats
exposed via i.p. injection (Qureshi and Mahmood, 2010); this mechanism
could affect bone development, but the relevance to developing animals is
unclear.
Mechanisms affecting
insulin regulation
Interpretation: In vivo evidence that Cr(VI) affects insulin signaling in developing
animals.
Key findings:
• Decreased expression of insulin receptor protein and substrates in F1
offspring from dams exposed via drinking water from GD 9-14 (Shobana et
al., 2017).
Oxidative stress and
apoptosis in the
placenta
Interpretation: In vivo and in vitro evidence that Cr(VI) increases oxidative stress
and apoptosis in the placenta.
Key findings:
• Biomarkers of oxidative stress and apoptosis observed in human placenta
samples with relatively high Cr levels (Banu et al., 2018), in rat placentas
following in vivo oral exposure (Banu et al., 2017a; Banu et al., 2017b), and
This document is a draft for review purposes only and does not constitute Agency policy.
3-305 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Evidence summary and interpretation
Inferences and summary
judgment
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that
decrease
certainty
Judgments and
rationale
in placental cells (Banu et al., 2018) or placental mitochondria (Sawicka and
Dtugosz, 2017) cultured in vitro.
Gestational anemia
Interpretation: Evidence suggests that Cr(VI) causes anemia-like effects.
Gestational anemia is a potential mechanism for low birth weight and preterm
birth following Cr(VI) exposure, although this has not yet been investigated.
1
This document is a draft for review purposes only and does not constitute Agency policy.
3-306 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
3.3. SUMMARY OF HAZARD IDENTIFICATION AND CONSIDERATIONS FOR
DOSE-RESPONSE ANALYSIS
3.3.1. Susceptible Populations and Life Stages
Susceptible populations and life stages refers to groups of people who may be at increased
risk for negative health consequences following chemical exposures due to factors such as life stage,
genetics, race/ethnicity, health status and disease, sex, lifestyle factors, and other coexposures.
This discussion of susceptibility focuses on factors for which there are available Cr(VI) data and
factors hypothesized to be important to Cr(VI). It should be noted that while evidence gaps exist
regarding Cr(VI)-specific susceptibilities, it is generally understood that increased negative health
consequences from exposure to pollutants in the air, soil, and groundwater can result from multiple
interacting factors (Table 3-48).
Table 3-48. Individual and social factors that may increase susceptibility to
exposure-related health effects (adapted from U.S. EPA (2020b))
Factor
Examples
Demographic
Sex, age, race/ethnicity, education, income, occupation, geography
Genetic variability
Polymorphisms in genes regulating cell cycle, DNA repair, cell division,
cell signaling, cell structure, gene expression, apoptosis, and metabolism
Lifestage
In utero, childhood, puberty, pregnancy, women of child-bearing age,
old age
Health status
Preexisting conditions or disease such as psychosocial stress, elevated
body mass index, frailty, nutritional status, chronic disease
Behaviors or practices
Diet, mouthing, smoking, alcohol consumption, pica, subsistence, or
recreational hunting and fishing
Social determinants
Income, socioeconomic status, neighborhood factors, health care
access, and social, economic, and political inequality
For noncancer dose-response, the intraspecies UF (UFh) is applied to account for variations
in susceptibility within the human population (interhuman variability) and the possibility (given a
lack of relevant data) that the database available is not representative of the dose/exposure-
response relationship in the subgroups of the human population that are most sensitive to the
health hazards of the chemical being assessed. This is described in greater detail in EPA's A review
of the reference dose and reference concentration processes fU.S. EPA. 20021 and Methods for
Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA.
19941. This assessment will apply an UFh to account for the anticipated broader variability in the
general population, which are outlined in Table 3-48. Toxicokinetics considerations, and use of a
PBPK model to account for some of the physiological human variabilities, will be considered when
This document is a draft for review purposes only and does not constitute Agency policy.
3-307 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
selecting an appropriate UFh. For cancer dose-response, slope factors generally represent an upper
bound on the average risk in a population or the risk for a randomly selected individual but not the
risk for a highly susceptible individual or group. Some individuals face a higher risk and some face a
lower risk. This is described in greater detail in EPA's Guidelines for Carcinogen Risk Assessment
fU.S. EPA. 2005al.
A number of different factors were identified that could predispose some populations of
humans to be more susceptible to Cr(VI) toxicity. These factors depend on the toxicity of concern
and route of exposure. For all endpoints following oral exposure (GI tract cancer and noncancer,
hepatic effects, developmental effects), conditions that elevate stomach pH would lower an
individual's ability to reduce Cr(VI) effectively and could lead to a higher rate of Cr(VI) absorption
(see Section 3.1). Stomach pH may vary according to health status and life stage. For respiratory
effects, preexisting respiratory conditions may be exacerbated by inhalation of Cr(VI). Preexisting
GI, liver, and hematologic conditions may be exacerbated by ingestion of Cr(VI).
3.3.1.1. Health Status and Disease
Low stomach acid
Individuals with chronically high stomach pH are expected to detoxify Cr(VI) less
effectively, leading to increased uptake of Cr(VI) in the GI tract (affecting the GI and other systemic
tissues). Individuals with hypochlorhydria (also known as achlorhydria) have consistently low
stomach acid, causing high stomach pH fKalantzi etal.. 2006: Feldman and Barnett. 1991:
Christiansen. 19681. This condition may be caused or exacerbated by multiple preexisting gastric
conditions, including H. pylori infection. Less than 1% of the adult population may exhibit
hypochlorhydria, whereas 10-20% of the elderly population (age 65 and up) may exhibit this
condition (Russell et al.. 1993). In addition, individuals taking medication to treat gastroesophageal
reflux disease (GERD), including calcium carbonate-based acid reducers and proton pump
inhibitors, have an elevated stomach pH during treatment Approximately 20% of the population
may be afflicted by GERD fLin andTriadafilopoulos. 20151. Sensitivity analyses on high-pH
populations using the PBPK model were performed to inform the dose-response assessment (see
Appendix C.1.5).
In addition to those with medical conditions, there is a significant percentage of individuals
with high stomach pH due to population variability. Among adults without hypochlorhydria,
Feldman and Barnett (1991) estimated that 5% of men may exhibit basal pH exceeding 5, and 5%
of women may exhibit basal pH exceeding 6.8. In the healthy elderly population, the percentage of
individuals with pH > 5 may be higher than for younger adults fRussell etal. f 19931 observed that
11% of elderly subjects hadpH >5).
This document is a draft for review purposes only and does not constitute Agency policy.
3-308 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
GI tract diseases
Individuals with preexisting GI conditions may be at higher risk of Cr(VI)-induced health
effects in the GI tract Cr(VI) contributes to oxidative stress and inflammation in the GI tract As a
result, damage to the gastric and intestinal epithelia due to preexisting inflammatory GI conditions
may be exacerbated by oral Cr(VI) exposure. For stomach cancer, preexisting conditions known to
increase risk in humans include H. pylori bacterial infection fBessede etal.. 2015: Fox and Wang.
20141 and Epstein-Barr virus (CGARN. 20141. Therefore, populations with these preexisting
conditions may also represent a population sensitive to Cr(VI)-induced gastrointestinal tract
cancer.
Liver diseases
Populations with preexisting liver disease represent a population susceptible to Cr(VI).
Cr(VI) contributes to oxidative stress in the liver, causes inflammation, increased fat storage
(histologically noted as vacuolation or fatty changes), and substantial increases in serum ALT and
AST, indicative of hepatocellular injury (see Section 3.2.4). The most common chronic liver disease
in western societies is nonalcoholic fatty liver disease (NAFLD), with an increasing prevalence in
line with obesity (Younossi. 20191. It is estimated that 25% of the US population has NAFLD. This
condition is characterized by excessive fat accumulation, especially triglycerides, in hepatocytes. If
untreated, NAFLD can progress to nonalcoholic steatohepatitis (NASH) and continue to fibrosis,
cirrhosis, and in some cases, hepatocellular carcinoma (Monserrat-Mesquida etal.. 20201.
Increased oxidative stress/pro-inflammatory status is implicated in the pathogenesis of NAFLD
fVidela etal.. 20041 and increased inflammation is associated with increased severity of NASH
(Monserrat-Mesquida etal.. 20201. NAFLD is associated with type 2 diabetes, metabolic syndrome,
obesity and cardiovascular disease (Younossi. 20191. therefore, populations with these preexisting
conditions likely also represent a population sensitive to Cr(VI)-induced liver perturbation.
Respiratory diseases
Inhaled Cr(VI) exposure may exacerbate preexisting respiratory conditions such as asthma,
emphysema and chronic obstructive pulmonary disease (COPD). This is because preexisting
conditions which reduce lung capacity, inflame airways, or obstruct breathing could be
compounded by Cr(VI) exposure, which may induce similar effects. Additionally, respiratory
conditions induced by lifestyle factors (i.e., smoking) or coexposures (i.e., asbestos) may interact
with the effects induced by inhaled Cr(VI) exposure.
Anemia and other blood disorders
Because the evidence suggests that Cr(VI) may produce anemia-like effects such as
reduced hematocrit, hemoglobin, MCV, MCH, and MCHC (see Section 3.2.5), exposure to Cr(VI) may
exacerbate the condition in individuals with preexisting conditions such as anemia, iron deficiency
or bleeding disorders. Pregnant women are at increased risk of developing anemia (American
This document is a draft for review purposes only and does not constitute Agency policy.
3-309 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
Pregnancy Association. 2021: O'Brien and Ru. 20171, and should be considered a susceptible group
for this reason. The prevalence of gestational anemia is highest among women with lower
socioecomonmic status (O'Brien and Ru. 2017: Rahman etal.. 20161. Gestational anemia is also a
risk factor for developmental toxicity, as noted below in Section 3.3.1.3.
3.3.1.2. Genetic Factors
Genetic polymorphisms
As summarized in Cancer MOA, Section 3.2.3.3, individuals with genetic polymorphisms
conveying deficiencies in DNA repair capacity may have increased susceptibility to Cr(VI)-induced
lung cancer. See Section 3.2.3.3 and Appendix C.3.5.1 for more details (see also Urbano et al.
Ł2012}).
Carriers of a mutated cystic fibrosis transmembrane conductance regulator fCFTRl allele
Suppression of the CFTR gene was shown to enhance intestinal tumorigenesis in animal
models fThan etal.. 20161. An analysis of the toxicogenomic data reported in Kopec et al. (2012b;
2012a) from mice exposed to Cr(VI) have identified a potential role for CFTR in the carcinogenic
effects of Cr(VI) (see Appendix C.3.4.2). Data from Kopec et al. (2012b: 2012a) indicate that CFTR
was inactivated in mice exposed to Cr(VI) in drinking water concentrations as low as 0.1 mg/L. In
the US, more than 10 million people are carriers of a mutated CFTR allele that confers an
approximately 50% reduction in CFTR expression levels; the deficit in CFTR function has been
shown to lead to an increased risk for several conditions associated with cystic fibrosis, including
colorectal cancer fMiller etal.. 2020: Scott etal.. 20201. Thus, individuals with this preexisting
condition may suffer an even further reduction in CFTR expression levels following oral exposure to
Cr(VI).
Heritable adenomatous polyposis coli (APC) mutations cause most cases of familial
adenomatous polyposis (FAP), an inherited syndrome associated with a high risk of colorectal
cancers flasperson etal.. 2017: Leoz etal.. 20151. Impaired CFTR activity was also shown to
enhance intestinal tumorigenesis in mice carrying the mutated tumor-suppressor gene
adenomatous polyposis coli [Ape). As a result, carriers of APC mutations may be more susceptible
to the tumorigenicity induced by events that inactivate CFTR, including Cr(VI) exposure, and there
could be additional risk for individuals carrying both the CFTR and APC mutations. Although 95%
of patients with classic FAP develop colorectal cancer by age 35 (Leoz etal.. 2015). there are over
1000 different types of APC mutations, many associated with a milder variant of FAP, that would
also be affected by CFTR inactivation.
This document is a draft for review purposes only and does not constitute Agency policy.
3-310 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
3.3.1.3. Life Stage
Developmental stages and pregnancy
Because the evidence indicates that Cr(VI) likely causes developmental effects in humans
given sufficient exposure conditions, pregnant women are considered a sensitive subpopulation. In
human studies of Cr(VI) focusing on this population, there are some indications of an association
between Cr(VI) exposure and spontaneous abortion, fetal growth, preterm birth, and congenital
malformations, but the evidence is limited in quality and quantity (see Section 3.2.9). Furthermore,
pregnant women, who are susceptible to developing iron-deficient anemia that is associated with
low birth weight, preterm birth, and perinatal and neonatal mortality (Figueiredo etal.. 2018:
Rahman etal.. 20161 are expected to be more sensitive to the hematologic effects of Cr(VI)
exposure.
Early life stages
Neonates, infants, and young toddlers generally have neutral stomach pH for the first 20-30
months, which then lowers to the normal adult range of 1-2 fNeal-Kluever etal.. 2019: Bai etal..
20161. Neonates also have delayed gastric emptying of milk, formula, and other caloric-containing
liquids fNeal-Kluever etal.. 20191. Delayed stomach emptying combined with elevated stomach pH
would lead to a higher uptake of ingested Cr(VI) in the stomach. In addition, incomplete stomach
reduction would lead to increased uptake of Cr(VI) in the small intestine. For chronic noncancer
effects and derivation of the RfD, this short-term change in the potential for absorbed Cr(VI) does
not impact the total lifetime average daily absorbed dose (because it occurs during such a short
time period). It is possible that neonates, infants, and young toddlers may be more susceptible than
adults during the short-term. However, there are no data for Cr(VI) reduction in the gastric acid of
infants and toddlers, and there would be significant uncertainties in applying the adult-based PBPK
model to infant or child physiology. For cancer effects, incorporation of age-dependent adjustment
factors in accordance with the Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens fU.S. EPA. 2005bl account for early-life (birth to 16 years) susceptibility by
using an adjustment to the slope factor.
Later life stages
In general, healthy elderly men and women (age 65 and older) have similar pH profiles as
adults (Russell etal.. 1993). although they may have slightly lower stomach pH than adults, and
higher duodenal pH fBai etal.. 20161. The healthy elderly population has the same gastric
emptying rate as healthy adults, but slower transit time in the small and large intestine fBai etal..
20161. There are age-related changes in the liver affecting hepatic clearance of drugs (Bai etal..
20161. but it is not clear how this may affect hepatic reduction of Cr(VI). As a result, it is uncertain
how Cr(VI) may affect the healthy elderly population differently from adults. However, elderly
populations are more likely to have preexisting health conditions that can lead to increased
This document is a draft for review purposes only and does not constitute Agency policy.
3-311 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
susceptibility to the effects of ingested Cr(VI). The elderly have high prevalence of conditions
associated with hyporchlorhydria such as H. pylori infection fBai etal.. 2016: Morihara etal.. 2001:
Russell etal.. 19931. The elderly also have higher usage of proton pump inhibitors to treat acid
reflux diseases, leading to increased stomach pH (Burdsall etal.. 20131. As a result, it is possible
that the elderly are more susceptible to the health effects of ingested Cr(VI), but mostly due to
pre-existing conditions that are associated with aging.
3.3.1.4. Sex
Males and females can differ greatly in body composition, organ function, and many other
physiological parameters that may influence the pharmacokinetics of chemicals and their
metabolites in the body (Gochfeld. 2007: Gandhi etal.. 20041. On average, males and females are
expected to have the same stomach pH (Shih etal.. 2003: Pressman etal.. 19901. The human
epidemiology studies do not support any specific sex susceptibilities for noncancer effects due to
Cr(VI) exposure. In animals, GI tract toxicity and hepatotoxicity may have been more severe in
females (see Sections 3.2.2 and 3.2.4), but it is unclear if the slight differences in results by sex in
rodents are applicable to humans.
3.3.2. Effects Other Than Cancer
The currently available evidence indicates that Cr(VI) is likely to cause GI, liver,
developmental, and lower respiratory toxicity in humans, given sufficient exposure conditions. The
evidence suggests that Cr(VI) may cause male reproductive, immune, and hematologic toxicity in
humans. The evidence is inadequate to assess whether Cr(VI) causes female reproductive toxicity
in humans. Because the totality of available evidence was sufficient to indicate that exposure to
Cr(VI) has the potential to cause GI, liver, developmental, and lower respiratory toxicity in humans,
organ/system-specific reference values were derived for those health effects, and not for most
health effects with evidence integration judgments of evidence suggests (i.e., male reproductive
effects and immune toxicity). Well-conducted studies for immune toxicity do not indicate chronic
hazards, and lack of sufficient dose-response data are available for male reproductive toxicity. It
was determined that a toxicity value derived for short-term/subchronic hematological effects may
be useful to protect susceptible populations (such as individuals with pre-existing anemia,
including pregnant women). More details are provided in Section 3.3.2.5.
The evidence base consisted of a wide array of animal and human studies (outlined in
greater detail by the health effect summary subsections below). A summary of the justifications for
the evidence integration conclusions for each of the main hazard sections is provided below and
organized by health effect The strength of the evidence for each hazard differed by species and
route of exposure. As discussed in Section 3.1, differences in observed effects between routes of
exposure can be attributed to pharmacokinetics. There was a lack of sufficient dose-response data
for health hazards outside of the respiratory tract following inhalation exposure, and as a result,
derivation of the RfC only considered effects in the respiratory tract. Similarly, respiratory tract
This document is a draft for review purposes only and does not constitute Agency policy.
3-312 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Toxicological Review ofHexavalent Chromium
effects were not observed following oral ingestion, and derivation of the RfD only considered effects
observed following ingestion (GI, hepatic, and developmental effects). Additional considerations,
decisions, and rationale are presented below in Table 3-49 and in Sections 4.1 and 4.4.
Table 3-49. Dose response considerations and rationale for specific routes of
exposure and health effects
Dose response consideration
Decision
Rationale
Animal and human data for RfD
derivation
RfD derivation used animal data
only.
Quantitative dose-response data
from medium and high confidence
oral studies were only available for
rodents.
Appropriate exposure data for RfD
derivation
Gavage studies excluded. Studies
not including a dose group below 20
mg/kg-d excluded.
Concern for frank-effect toxicity.
Health effects for RfC derivation
RfC derivation for respiratory tract
effects only. Route-to-route
extrapolation not performed.
Pharmacokinetic differences are
significant between inhalation and
oral exposure, particularly for portal-
of-entry effects.
Animal and human data for RfC
derivation
RfC derivation of nasal effects used
human data only.
RfC derivation of lower respiratory
effects used animal data only.
Quantitative dose-response data
from medium and high confidence
studies were limited by species and
effects.
3.3.2.1. GI Tract Effects
The judgment that the available evidence indicates that Cr(VI) likely causes GI toxicity in
humans given sufficient exposure conditions is based on four high confidence toxicology studies.
Two of these studies (NTP. 2008. 20071 contained multiple study arms, resulting in both chronic
and subchronic data across multiple species, strains, and sexes (see Table 3-49). All four high
confidence studies in rats and mice reported various histological effects in the GI tract associated
with oral exposure to Cr(VI). These include diffuse epithelial hyperplasia or crypt cell hyperplasia,
histiocytic cellular infiltration, squamous metaplasia, degenerative changes in the villi
(vacuolization, atrophy, and apoptosis), and gastric ulceration fThompson etal.. 2012b: Thompson
etal.. 2011: NTP. 2008. 2007). The literature search for this assessment did not identify
epidemiological studies with analyses of GI effects in humans that met PECO criteria.
Mechanistic evidence supports the GI tract effects observed in animals and suggests a
possible MOA of Cr(VI)-induced GI toxicity involving the production of free radicals and reactive
intermediates through intracellular Cr(VI) reduction resulting in oxidative stress, mitochondrial
dysfunction, inflammation, and apoptosis. Degenerative changes to the cells lining the GI tract can
manifest as necrosis, apoptosis, and subsequent villous stunting, resulting in crypt abscess and
ulceration (Betton. 2013). Irreversible cytoplasmic vacuolization can be a marker of cell death and
cytoprotective autophagy in response to stress (Shubin et al.. 2016).
This document is a draft for review purposes only and does not constitute Agency policy.
3-313 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Toxicological Review ofHexavalent Chromium
The histiocytic cellular infiltration endpoint was considered of unclear biological
significance (see Sections 3.2.2.2 and 3.2.2.4) and therefore was not included for dose-response
analysis. Endpoints observed in subchronic studies such as apoptosis, villous atrophy, and villous
cytoplasmic vacuolization were not considered for dose-response assessment. Only the chronic
data from NTP (20081 were considered for effects in the GI tract.
Diffuse epithelial hyperplasia only occurred in portions of the GI tract where other
degenerative effects were observed. Diffuse epithelial hyperplasia, although predictive of more
severe manifestations of toxicity, is considered minimally adverse. Data for this endpoint are
available from both the chronic and subchronic studies (Table 3-50).
Table 3-50. Available animal studies showing histopathological changes in the
duodenum
Reference
Study arms performed
Observations
NTP(2008)
F344 Rat, male and female
(chronic)
Histiocytic cellular infiltration
B6C3F1 mouse, male and female
(chronic)
Diffuse epithelial hyperplasia, histiocytic
cellular infiltration
NTP(2007)
F344 Rat, male and female
(subchronic)
Histiocytic cellular infiltration
B6C3F1 mouse, male and female
(subchronic)
Epithelial hyperplasia, histiocytic cellular
infiltration
B6C3F1, BALB/c, and am-C57BL/6
mouse, male (subchronic strain
comparison)
Epithelial hyperplasia, histiocytic cellular
infiltration
Thompson et al. (2012b)
F344 Rat, female (subchronic)
Crypt cell hyperplasia, histiocytic
infiltration, apoptosis, villus atrophy
Thompson et al. (2011)
B6C3F1 mouse, female
(subchronic)
Crypt cell hyperplasia, histiocytic
infiltration, apoptosis, villus atrophy,
villous cytoplasmic vacuolization
3.3.2.2. Hepatic Effects
The judgment that the available evidence indicates that Cr(VI) likely causes hepatic
toxicity in humans given sufficient exposure conditions is based on studies in animals that observed
hepatic effects following drinking water exposure. Several studies in rats and mice reported
various histological lesions in the liver associated with oral exposure to Cr(VI). These lesions
include increased inflammation and infiltration of immune cells, fatty changes and vacuolation,
indications of apoptosis and necrosis, and increased incidence of altered hepatic foci. NTP (2008)
described chronic inflammation as "minimal to mild severity" in most dose groups, with "mild to
This document is a draft for review purposes only and does not constitute Agency policy.
3-314 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
moderate" in the higher dose groups. The severity ratings were used to inform BMR selection (see
Section 4.1).
Many studies have examined serum indicators that are potentially informative for
predicting hepatotoxicity following exposure to Cr(VI). The most commonly reported indicators
included alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase
(ALP), and sorbitol dehydrogenase (SDH). These changes were observed across multiple studies,
with ALT changes exceeding twofold which is an indicator of concern for hepatic injury (Sawicka
and Dtugosz. 2017: EMEA. 2010: Boone etal.. 20051. The outcomes rated medium confidence
showing a response were available from chronic and subchronic studies across multiple species,
strains, and sexes (see Table 3-51). These are discussed further in Section 4.1.
The human evidence for Cr (VI)-induced liver effects is limited in terms of number and
confidence of studies. However, two of the available three studies (one occupational and one
general population study) provide some indication of exposure-related alterations of liver clinical
chemistry (Sazakli etal.. 2014: SaraswathvandUsharani. 2007).
Mechanistic evidence supports the hepatic effects observed in animals and humans and
suggests a possible MOA of Cr(VI)-induced liver toxicity involving the production of free radicals
and reactive intermediates through intracellular Cr(VI) reduction resulting in oxidative stress,
mitochondrial dysfunction, inflammation, and apoptosis.
Table 3-51. Available animal studies showing histopathological and clinical
chemistry changes in the liver
Reference
Species/strain and sex
Observations3
NTP(2008)
F344 Rat, male and female
(chronic)
Histopathology: histiocytic cellular infiltration, chronic
inflammation, fatty change, basophilic focus
Clinical chemistry (male rats only): ALT, ALP, SDH, bile
acids
B6C3F1 mouse, male and female
(chronic)
Histopathology: histiocytic cellular infiltration, chronic
inflammation
NTP(2007)
F344 Rat, male and female
(subchronic)
Histopathology: histiocytic cellular infiltration, chronic
focal inflammation
Clinical chemistry: ALT, ALP, SDH, bile acids, cholesterol,
triglycerides, 5'nucleotidase
B6C3F1, BALB/c, and am-
C57BL/6 mouse, male
(subchronic)
Clinical chemistry: ALT, ALP, SDH, bile acids, glycogen
(B6C3F1 and am-C57BL/6 only)
Rafael et al. (2007)
Wistar rat, male (chronic)
Clinical chemistry: ALT, ALP, SDH, glucose, cholesterol,
total protein
NTP (1996a)
BALB/c mouse, male and female
(subchronic)
Histopathology: cytoplasmic vacuolation (fatty change)
This document is a draft for review purposes only and does not constitute Agency policy.
3-315 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofHexavalent Chromium
Reference
Species/strain and sex
Observations3
NTP (1997)
BALB/c mouse, male and female
(continuous breeding)
Histopathology: hepatocyte cytoplasmic vacuolation
(fatty change), hepatocyte individual cell necrosis,
necrosis, acute inflammation
Krim et al. (2013)
Wistar rat, male (subchronic)
Clinical chemistry: ALT, ALP, AST, cholesterol, total lipids,
triglycerides, LDH
Wang etal. (2015)
Sprague-Dawley rat, male
(subchronic)
Clinical chemistry: ALT, AST, cholesterol, triglycerides,
glucose
Navva et al.
(2017a)
Wistar rat, male (subchronic)
Clinical chemistry: ALT, ALP, AST
aOnly endpoints rated medium or high confidence within each study are listed.
3.3.2.3. Respiratory Tract Effects
The judgment that the available evidence indicates that Cr(VI) likely causes respiratory
toxicity in humans given sufficient exposure conditions is based on studies in animals that observed
effects following inhalation exposure. Most animal inhalation studies of lower respiratory effects
contained data for lung histopathology, lung weight, and cellular responses. Because
histopathological and cellular changes occurred together, and in combination with serum
biomarkers indicating an inflammatory response fNikula et al.. 20141. these were considered
indicators of adverse responses and considered for dose-response analysis. Because lung weight is
a less specific endpoint for lung injury (e.g., lung weight increase in the only medium confidence
data by Glaser etal. (1985) may be related to accumulation of macrophages), this endpoint was not
considered for dose-response analysis. The available histopathological changes and cellular
response outcomes that were rated medium confidence are outlined in Table 3-52. These are
discussed further in Section 4.2.
The human evidence for Cr(VI)-induced lower respiratory effects is limited in terms of
number and confidence of studies. However, three of the available five studies provide some
indication of exposure-related decrements in lung function assessed using spirometry (Zhang etal..
2022: Li etal.. 2015b: Kuoetal.. 1997bl.
Mechanistic evidence supports the respiratory tract effects observed in animals and
suggests a possible MOA of Cr(VI)-induced toxicity involving the production of free radicals and
reactive intermediates through intracellular Cr(VI) reduction resulting in oxidative stress.
Table 3-52. Available animal studies showing histopathological changes and
cellular responses in the lung
Reference
Species/ strain and sex
Observations3
Glaser et al. (1990)
Wistar rat, male
(subchronic)
Histopathology: Histiocytosis, bronchioalveolar
hyperplasia, fibrosis
This document is a draft for review purposes only and does not constitute Agency policy.
3-316 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Toxicological Review ofHexavalent Chromium
Reference
Species/ strain and sex
Observations3
BALF: LDH, ALB, total protein, macrophage effects
Glaser et al. (1985)
Wistar rat, male
(subchronic)
BALF: Macrophage effects
Johansson et al. (1986a)
Rabbit, male
(subchronic)
Histopathology: Histiocytosis
Cohen et al. (2003)
F344 Rat, male (chronic)
BALF: Total cells, total macrophages
Johansson et al. (1986b)
Rabbit, male
(subchronic)
BALF: Total macrohpages, macrophage effects
Kim et al. (2004)
Sprague-Dawley Rat,
male (subchronic)
Histopathology: Inflammatory markers (qualitative)
aOnly endpoints rated medium or high confidence within each study are listed.
3.3.2.4. Developmental Effects
The judgment that the available evidence indicates that Cr(VI) likely causes developmental
toxicity in humans given sufficient exposure conditions is based on the observation of decreased
offspring growth across most animal studies, as evidenced by decreased fetal or postnatal body
weights and decreased skeletal ossification. The only data suitable for dose-response analysis were
for fetal and postnatal growth, which were observed to some extent in the high confidence RACES
study in mice by NTP (19971 (all other studies were low confidence and not considered for dose-
response assessment). Within the animal studies, statistically significant effects on fetal
development were observed at doses the same or lower than those that caused decreased maternal
body weight According to EPA Guidelines, developmental effects at doses that cause minimal
maternal toxicity are still considered to represent developmental toxicity and should not be
discounted as maternal toxicity fU.S. EPA. 19911. Because of the correlation between maternal dam
weight and offspring body weight, the maternal dose was used as the basis for dose-response
modeling instead of the averaged F0 male and female dose.
3.3.2.5. Hematological Effects
Although toxicity values are not typically developed for hazards with suggestive conclusions
(e.g., evidence suggests for noncancer hazards and "suggestive evidence of carcinogenic
potential"), it may be useful to develop values for some purposes. For example, providing a sense of
the magnitude and uncertainty of potential risks, ranking potential hazards, or setting research
priorities fU.S. EPA. 2005al. A value may be useful for some purposes when the evidence includes a
well-conducted study (particularly when that study may also demonstrate a credible concern for
greater toxicity in a susceptible population or lifestage) U.S. EPA (2020b). Pregnant women are
more susceptible to developing iron-deficient anemia, making them more susceptible to the
hematological effects of Cr(VI), and hematological effects have been correlated to low birthweight
Because these factors demonstrate a credible concern for greater toxicity in a susceptible
This document is a draft for review purposes only and does not constitute Agency policy.
3-317 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
population and life stage, organ/system-specific reference doses were derived for hematological
effects. Hematological markers affected by Cr(VI) exposure include MCV, MCH, MCHC, Hgb, and Hct
(see Section 3.2.5). Of the available studies collecting complete blood count data, the NTP (20081
and NTP (20071 bioassays provided the most comprehensive data set considering multiple
timepoints and related hematological endpoints in both sexes, and were therefore considered for
dose-response. Additional discussion is provided in Section 4.1.1.
3.3.3. Cancer
Under the 2005 Guidelines for Carcinogen Risk Assessment, Cr(VI) is "carcinogenic to
humans" via the inhalation route of exposure and is "likely to be carcinogenic to humans" via the
oral route of exposure.
In 1998, the EPA IRIS Toxicological Review ofHexavalent Chromium classified Cr(VI) as a
"known human carcinogen by the inhalation route of exposure" based on consistent evidence that
inhaled Cr(VI) causes lung cancer in humans and supporting evidence of carcinogenicity in animals
fU.S. EPA. 1998cl. The same conclusion has since been reached by other authoritative federal and
state health agencies and international organizations and the carcinogenicity of Cr(VI) is
considered to be well-established for inhalation exposures (TCEO. 2014: IPCS. 2013: NIOSH. 2013:
IARC. 2012: CalEPA. 2011: NTP. 2011: OSHA. 20061. As stated in the 2014 preliminary packages
(U.S. EPA. 2014b. c) and the Systematic Review Protocol (Appendix A), the review of cancer by the
inhalation route focused on data that may improve the quantitative exposure-response analysis
conducted in EPA's 1998 IRIS assessment; EPA did not reperform a carcinogenicity determination
for inhalation exposure. An overview of the literature screening for exposure-response data is
contained in Section 4.4.
Determination that Cr(VI) is likely to be carcinogenic to humans by the oral route of
exposure was made based on 1) a high confidence study in rodents showing a clear dose-response
relationship between oral Cr(VI) exposure and incidence of GI tract tumors (NTP. 20081: and 2)
robust evidence that a mutagenic MOA has a key role in Cr(VI)-induced cancer via inhalation and
oral exposures (see Section 3.2.3)t
Because a mutagenic MOA for Cr(VI) carcinogenicity (see Section 3.2.3) is "sufficiently
supported in (laboratory) animals" and "relevant to humans," for both routes of exposure, EPA uses
a linear low dose extrapolation from the POD in accordance with Guidelines for Carcinogen Risk
Assessment (U.S. EPA. 2005a). Furthermore, in the absence of chemical-specific data to evaluate
differences in age-specific susceptibility, increased early-life susceptibility to Cr(VI) is assumed and
EPA applies ADAFs in accordance with the Supplemental Guidance for Assessing Susceptibility from
Early-Life Exposure to Carcinogens fU.S. EPA. 2005bl. Linear low dose extrapolation and ADAFs are
This document is a draft for review purposes only and does not constitute Agency policy.
3-318 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 applied for both the inhalation and oral routes of exposure58. The 2-year drinking water bioassay
2 by NTP f20081 provides the datasets for dose-response modeling of tumors in the GI tract (tumors
3 of the oral cavity in male and female F344 rats, and tumors of the small intestine in male and female
4 B6C3F1 mice).
5 Due to reduction (detoxification) of Cr(VI) in the stomach compartment prior to transit to
6 the small intestine, dose-response modeling of tumors in the mouse small intestine incorporates
7 adjustments by a PBPK model when performing animal-to-human extrapolation. For tumors of the
8 rat oral cavity, PBPK modeling is not applied, because Cr(VI) in drinking water exposes the
9 epithelium of the tongue and oral mucosa prior to detoxification in the stomach.
58Because carcinogenicity determination was not reperformed for lung cancer, this section focuses only on
cancer of the GI tract. A discussion of the considerations for dose-response of lung cancer is contained in
Section 4.4.
This document is a draft for review purposes only and does not constitute Agency policy.
3-319 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofHexavalent Chromium
4. DOSE-RESPONSE ANALYSIS
4.1. ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER
The RfD (expressed in units of mg/kg-day) is defined as an estimate (with uncertainty
spanning perhaps an order of magnitude) of a daily exposure to the human population (including
sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a
lifetime. It can be derived from points of departure (PODs) such as a no-observed-adverse-effect
level (NOAEL), lowest-observed-adverse-effect level (LOAEL), or the 95% lower bound on the
benchmark dose (BMDL), with uncertainty factors (UFs) generally applied to reflect limitations of
the data used.
As discussed in Sections 3.2.2, 3.2.4, 3.2.5, and 3.2.9, based on findings in experimental
animals, the evidence indicates that exposure to Cr(VI) is likely to cause GI, liver, and
developmental toxicity in humans. Because the totality of available evidence was sufficient to
indicate that exposure to Cr(VI) has the potential to cause these health effects, organ/system-
specific reference values were derived for GI, liver, and developmental toxicity, and not for most
health effects with evidence integration judgments of evidence suggests (i.e., male reproductive
effects and immune toxicity). However, for hematologic effects, it was determined that a toxicity
value derived for short-term/subchronic exposures may be useful to protect susceptible
populations (such as individuals with pre-existing anemia, including pregnant women). More
details are provided in Section 3.3.2.5. An overview of the process for deriving candidate values,
osRfDs and osRfCs and overall RfDs and RfCs is provided in Appendix Figure D-l.
4.1.1. Identification of Studies for Dose-Response Analysis of Selected Effects
In order to identify the studies for dose-response analysis, key attributes of the studies
reporting the endpoints selected for each hazard were reviewed (i.e., study size and design,
relevance of the exposure paradigm, and measurement of the endpoints of interest). Exposure
paradigms including a relevant route of human environmental exposure are preferred. When
developing a chronic reference value, chronic or subchronic studies are preferred over studies of
acute exposure durations (with the exception of developmental studies, where exposures only need
to occur during susceptible periods). Studies with a broad exposure range and multiple exposure
levels are preferred to the extent that they can provide information about the shape of the
exposure-response relationship.
Human studies are generally preferred over animal studies as the basis for a reference value
when quantitative measures of exposure are reported, and the reported effects are determined to
be associated with exposure. The available epidemiological studies of worker populations exposed
to Cr(VI) examined the relationship between certain health endpoints and inhalation exposure;
This document is a draft for review purposes only and does not constitute Agency policy.
4-1 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofHexavalent Chromium
however, no sufficient epidemiological studies of ingested Cr(VI) are available and route-to-route
extrapolation was not considered for this assessment (see Protocol, Appendix A). In the absence of
human data, the animal studies were considered for dose-response analysis.
Experimental animal studies considered for each health effect were evaluated using general
study evaluation considerations discussed in the Protocol (Appendix A). The oral animal
toxicological evidence base for Cr(VI) consists of chronic and subchronic studies. Because medium
and high confidence studies were available, low confidence studies were not considered for toxicity
value derivation.
Cr(VI) can induce frank effects in rodents at high doses, which raises considerations of
exposures and study designs appropriate for dose-response analysis. Because Cr(VI) gavage
exposure has been shown to induce frank effects and high mortality in rodents (gut detoxification is
much less effective for gavage exposure), these studies were not considered for dose-response
assessment This criterion resulted in the omission of one high confidence study fZheng etal..
20181 from consideration of dose-response analysis for developmental effects. High dose exclusion
criteria for drinking water and oral feed studies were also considered. At approximately
20 mg/kg-d ad libitum, NTP f20071 reported reduced body weight, chemical-induced stomach
ulcers (80-100% incidence), and reduced water consumption in rats exposed for 90 days. The
study also reported 10-20% decreases in final body weight relative to controls in mice exposed for
90 days at the high doses (approximately 15-25 mg/kg-d). In order to focus on chronic effects
observed in the low dose region (defined here as around 1 mg/kg-d ad libitum based on results
observed by the chronic 2-year NTP (2008) drinking water bioassay), studies which did not include
an exposed group below 20 mg/kg-d were not considered for RfD derivation. This criterion
ultimately did not impact any decisions regarding dose-response, because all such studies were
rated low confidence.
This document is a draft for review purposes only and does not constitute Agency policy.
4-2 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Figure 4-1. Evaluation of studies from the Cr(VI) hazard identification for
derivation of toxicity values. For endpoints where medium or high confidence
studies were available, low confidence studies were not considered.
1 4.1.1.1. GI Tract Toxicity
2 Small intestine histopathology was considered for dose-response analysis of the GI tract
3 effects of oral exposure to Cr(VI). Chronic data from the NTP (20081 2-year bioassay were used for
4 the dose-response assessment The chronic 2-year NTP f20081 bioassay analyzed many of the
5 same endpoints as other shorter term studies (which had smaller sample sizes and typically used
6 higher doses). Thompson et al. (2012b: 20111 were subchronic studies which incorporated lower
7 doses than NTP (20081. However, these studies used smaller sample sizes and shorter exposure
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 durations than NTP f20081. and only examined females (Table 4-1). An overview of design features
2 of the high confidence animal studies containing data for the GI tract is provided below in Table 4-1.
Table 4-1. Design features of high confidence studies that examined GI tract
effects (histopathology) via the oral route of exposure
Study reference
Species/strain
and sex
Exposure
duration
Number of
dose groups'3
Number of
animals/group
Dose range
(mg/kg-d)
NTP (2008)a
B6C3F1 mouse,
male and female
2 years
4
50
0.3-8.9
NTP(2008)
F344 Rat, male
and female
2 years
4
50
0.2-7.1
NTP(2007)
F344 Rat, male
and female
90 days
5
10
1.7-21
NTP(2007)
B6C3F1 mouse,
male and female
90 days
5
10
3.1-27.9
NTP(2007)
B6C3F1 mouse,
male
90 days
3
5
2.8-8.7
NTP(2007)
BALB/c mouse,
male
90 days
3
5
2.8-8.7
NTP(2007)
am-C57BL/6
mouse, male
90 days
3
5
2.8-8.7
Thompson et al.
(2012b)
F344 Rat, female
90 days0
5
10
0.015-20
Thompson et al. (2011)
B6C3F1 mouse,
female
90 days0
6
10
0.024-31.1
Preferred data for dose-response.
bNumber does not include control group.
cNote: Thompson et al. (2012b) and Thompson et al. (2011) also performed an 8-day sacrifice on 5 animals/group.
3 The most sensitive GI effect in mice, diffuse epithelial hyperplasia, was consistently
4 observed at statistically significant incidence levels in mice in all exposure groups (>0.3 mg/kg-d
5 Cr(VI)) of males and females of multiple strains in three high confidence subchronic and chronic
6 studies (Thompson etal.. 2011: NTP. 2008. 20071. The hyperplastic duodenal lesions were
7 described as being suggestive of tissue regeneration following degenerative changes to the
8 intestinal villi. In rats, it was observed less consistently and at higher doses compared to mice
9 fCullen etal.. 2015: Thompson etal.. 2012b: Thompson etal.. 20111. Dose-response modeling was
10 performed on the chronic 2-year data for male and female mice exhibiting diffuse epithelial
11 hyperplasia of the proximal small intestine (duodenum).
This document is a draft for review purposes only and does not constitute Agency policy.
4-4 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
4.1.1.2. Hepatic Toxicity
Liver histopathology changes and serum biomarkers of liver injury were considered for
dose-response analysis of the hepatic effects of oral exposure to Cr(VI). These were considered the
most representative indicators of hepatic toxicity in the database. Fatty liver changes (cytoplasmic
vacuolation) and increased ALT are also clinical markers used in diagnosis of human liver diseases
(see Section 3.3). Dose-response modeling was not performed on liver weight because only
moderate changes were observed (see Section 3.2.4), and changes in liver histopathology and
serum biomarkers were more consistently observed and more sensitive than liver weight changes.
Generally consistent elevations of ALT (biomarkers of liver injury) were seen across various
multiple well-conducted studies in both rats and mice, with the magnitude of change considered to
be biologically significant and a specific indication of liver damage. For dose-response modeling of
clinical chemistry changes, NTP (2008) observed increased alanine aminotransferase (ALT) in male
F344 rats at all three data collection time points (3, 6, and 12 months). Dose-response modeling
was performed on the clinical chemistry endpoint ALT in male F344 rats59 at the 12-month and
90-day collection periods of the NTP f20081 bioassay. ALT changes in male and female rats from
the 90-day NTP (2007) study were also modeled60. ALT changes in male rats at the 90-day
timepoint from the 2-year NTP (2008) study were modeled to provide a comparison with the
90-day NTP (2007) data. In mice, changes in ALT only occurred at high doses during the 90-day
NTP (20071. and there were no changes in the other clinical chemistry parameters like there were
in rats. Therefore, this endpoint was not modeled in mice.
For histopathological changes, increased incidence of chronic liver inflammation was
observed in rodents during the 2-year NTP f20081 bioassay, but this endpoint exhibited a
monotonic dose-response relationship for female rats and mice. In male rats, the increased
inflammation was nonmonotonic and only significantly increased for one dose group. In male mice,
no effect was observed. Fatty liver changes were also observed in female rats during the 2-year
NTP f20081 bioassay. Similar to the chronic inflammation endpoint, this effect was not consistently
observed across species or sex. Dose-response modeling was performed on the incidence data for
chronic liver inflammation and fatty liver changes in female rats from NTP (2008). and chronic
inflammation in female mice from NTP f2008I
An overview of design features of the medium and high confidence animal studies
containing data for hepatic effects considered for oral dose-response is provided below in Table 4-
2. Because there were studies that were rated high and medium for endpoints within this domain
(see Section 3.2.4), low confidence studies were not considered for dose-response assessment
59The NTP f20081 2-year study did not obtain clinical chemistry data in mice or female rats, whereas the 90-
day NTP f2007] study contained data for both male and female F344 rats and mice. While chronic data are
still preferred, subchronic data were evaluated to assess differences between sexes.
60Note: the lowest dose (in mg/kg-d Cr(VI) was the same in males and females for the subchronic study.
When taking into consideration differences in body weight in the pharmacokinetic model, the daily absorbed
dose in males was slightly higher than females (see Appendix C.1.5).
This document is a draft for review purposes only and does not constitute Agency policy.
4-5 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 4-2. Design features of studies that examined hepatic effects (clinical
chemistry and histopathology) via the oral route of exposure
Study reference
(quality)
Species/strain and
sex
Exposure
duration
Number of
dose groups'3
Number of
animals/group
Dose range
(mg/kg-d)
NTP (2008) (hiah)3
F344 Rat, male and
female
2 years
4
50
0.2-7.1
NTP (2008) (hiah)3
B6C3F1 mouse, male
and female
2 years
4
50
0.3-8.9
NTP (2007) (hiah)3
F344 Rat, male and
female
90 days
5
10
1.7-21
NTP (2007) (hiah)
B6C3F1 mouse, male
and female
90 days
5
10
3.1-27.9
NTP (2007) (hiah)
B6C3F1 mouse, male
90 days
3
5
2.8-8.7
NTP (2007) (hiah)
BALB/c mouse, male
90 days
3
5
2.8-8.7
NTP (2007) (hiah)
am-C57BL/6 mouse,
male
90 days
3
5
2.8-8.7
Navva et al. (2017a)
(medium)
Wistar rat, male
28 days
1
6
10.6
Rafael et al. (2007)
(medium)
Wistar rat, male
10 weeks
1
9 control, 19
exposed
2.96
NTP (1996a) (hiah)
BALB/c mouse, male
and female
9 weeks
4
24 males, 48
females (5-6
males, 12
females/group per
timepoint)
1.1-48.4
NTP (1997) (hiah)
BALB/c mouse, male
and female
13-week
continuous
breeding
3
20 (F0), 5-10
(offspring)
6.8-50
Krim et al. (2013)
(medium)
Wistar rat, male
30
1
10
5.3
NTP (1996b) (hiah)
Sprague-Dawley rat,
male and female
9 weeks
4
5
0.35-9.90
Wang et al. (2015)
(medium)
Sprague-Dawley rat,
male
4 weeks
3
8
2.5-7.6
Preferred data for dose-response.
bNumber does not include control group.
1 In summary, dose-response modeling was performed on the following hepatic datasets:
2 • Increased ALT in male rats from NTP f20081 at the 90-dav timepoint and 12-month
3 timepoint
This document is a draft for review purposes only and does not constitute Agency policy.
4-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofHexavalent Chromium
• Increased ALT in male and female rats from NTP (2007) (90 days)61
• Increased chronic liver inflammation in female rats from NTP f20081 (2 years)
• Increased chronic liver inflammation in female mice from NTP (2008) (2 years)
• Fatty liver change in female rats from NTP (2008) (2 years)
4.1.1.3. Developmental Toxicity
As noted in Section 3.2.9, decreases in fetal and postnatal growth were the only consistently
observed effects observed in exposed animals. The two medium to high confidence studies that
observed this effect were NTP (1997) and Zheng etal. (2018). De Flora et al. (2006) did not
observe this effect. The high confidence study by Zheng etal. (2018) was not considered for dose-
response assessment because it was a gavage study; Cr(VI) gavage exposure has been shown to
induce frank effects and high mortality in rodents due to less effective gut detoxification compared
to drinking water exposure. Dose-response modeling was performed on fetal and postnatal growth
outcomes in the F1 generation observed by NTP fl9971. Data are available for males fPND14 and
PND21) and females (PND14 and PND21).
4.1.1.4. Hematological Toxicity
The database of hematological endpoints is extensive due to the number of studies
reporting these endpoints and the comprehensive measures available for multiple markers from
complete blood counts (i.e., MCV, MCH, MCHC, Hgb, and Hct) (see Section 3.2.5). There are eleven
datasets from six high confidence National Toxicology Program studies, and five medium confidence
datasets from NTP and other sources. Data from NTP (2008) and NTP (2007) were particularly
useful because they collected data at multiple timepoints. An overview of the design features of
these studies is presented above in Table 4-2. Because hemoglobin (Hgb) is essential for the
transport of oxygen molecules, it is the marker most closely associated with adverse outcomes
caused by iron-deficient anemia. Because the hematological effects ameliorated over time, the
dose-response will focus on subchronic and short-term data (90 days and 22 days), with chronic
data at 12 months used as a comparison. Only the rat data were modeled, since little or no effects
were observed in mice. While the most sensitive low-dose data were at 22 days from NTP (2008)
(which used lower doses than NTP (2007)). dose response data from the 2007 study were still
evaluated to assess possible sex differences (the 2008 study only collected hematological data in
male rats and female mice). The use of 22-day data for the POD (as opposed to the 12-month data
when effects ameliorated) was determined to be appropriate in order to protect susceptible
61While chronic data are preferred for dose-response, only chronic male data were available for this endpoint.
Subchronic data from both the 90-day study and 2-year study were modeled to evaluate possible difference
between sexes.
This document is a draft for review purposes only and does not constitute Agency policy.
4-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Toxicological Review ofHexavalent Chromium
subpopulations (such as individuals with pre-existing anemia, including pregnant women; see
Section 3.3.1) from both short-term and chronic health effects.
In summary, dose-response modeling was performed on the following hematological
datasets:
• Decreased Hgb in male rats from NTP (20081 at 22 days, 90 days, and 12 months
• Decreased Hgb in male and female rats from NTP f20071 at 23 days, and 90 days
4.1.2. Methods of Analysis
Biologically based dose-response models are not available for Cr(VI). In this situation, EPA
evaluates 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 the observed data. Consistent with this approach, EPA evaluated dose-response information
with the models62 available in EPA's Benchmark Dose Software (BMDS, Version 3.2). EPA
estimated the benchmark dose (BMD) and the 95% lower confidence limit on the BMD (BMDL)
using a benchmark response (BMR) that represents a minimal, biologically significant level of
change (U.S. EPA. 2012b). Endpoint-specific BMRs are described below. Where modeling was
feasible, the estimated BMDLs were used as points of departure (PODs); the PODs are summarized
in Table 4-3. Further details including the modeling output and graphical results for the model
selected for each endpoint can be found in Appendix D.l and U.S. EPA (2021a). Where
dose-response modeling was not feasible, no-observed-adverse-effect levels (NOAELs) or
lowest-observed-adverse-effect levels (LOAELs) were identified; NOAELs and LOAELs are also
summarized in Table 4-3.
4.1.2.1. PBPKModeling and Animal-to-Human Extrapolation
Following ingestion, extracellular reduction of Cr(VI) to Cr(III) in the stomach is a major
pathway for detoxification in both rodents and humans, and may have a significant impact on the
amount of Cr(VI) available for absorption and distribution. Uptake of Cr(VI) into tissues and
intracellular reduction occurs rapidly (see Section 3.1.1 and Appendix C.l.l for overview). While GI
tract PBPK models are capable of estimating the extent of extracellular reduction in the stomach,
the in vivo estimates of localized uptake and reduction of Cr(VI) in GI and systemic tissues exhibit
high uncertainties (particularly for the distal GI). Thus, all unreduced Cr(VI) that escapes stomach
reduction and enters the small intestine (estimated by PBPK modeling) is assumed to have the
62Some statistical models (Gamma, Dichotomous Hill, Weibull, and LogLogistic) were run with constrained
slope or power parameters (>1) (U.S. EPA. 2012b). As noted in Benchmark Dose Software (BMDS) version 3.2
user guide (U.S. EPA. 2020a). some models with unrestricted coefficients can give complicated shapes, in
particular high-degree polynomial models (which produce unrealistic 'wavy' results with negative response
rates). While Bayesian model averaging is an available feature of BMDS 3.2, only frequentist models were run
in this assessment.
This document is a draft for review purposes only and does not constitute Agency policy.
4-8 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 potential for absorption into epithelial cells. The unreduced mg/kg-d Cr(VI) dose escaping stomach
2 reduction in the rodent can be adjusted to an internal dose63 by allometric scaling consistent with
3 Recommended Use of Body Weight3/4 as the Default method in derivation of the oral reference dose
4 (U.S. EPA. 2011c). This assumes that absorbed Cr(VI) is rapidly cleared (reduced or excreted), with
5 interspecies differences following allometry. While there is some uncertainty in how much of the
6 unreduced Cr(VI) escaping the stomach is reduced and absorbed by the GI tissue prior to systemic
7 distribution, the interspecies difference in this amount is likely to be low in relation to the
8 interspecies difference in gastric reduction (which is driven by differences in stomach pH and
9 Cr(VI) reduction capacity).
10 PBPK modeling revealed that the Cr(VI) dose escaping stomach reduction (and therefore
11 the internal dose) increased linearly with oral dose for rats and mice (Appendix C.1.5). Therefore,
12 performing BMD modeling on the orally administered doses and performing PK conversions at a
13 later step would ultimately produce the same POD as if BMD modeling was performed on the basis
14 of internal PK-derived rodent doses. For humans, gastric reduction is nonlinear with respect to
15 ingested dose (Appendix C.1.5).
16 The steps for candidate value derivation are outlined below and in Figure 4-2:
17 • Dose-response modeling was performed on the basis of mg/kg-d Cr(VI) ingested to
18 determine a BMDL or LOAEL/NOAEL. Where possible, time-weighted average daily doses
19 calculated from time-course data (through the time of data collection) were used. For
20 example, for endpoints only measured at the 12-month time point in a 2-year study, the
21 time-weighted average daily doses over 12 months were used for dose-response (as
22 opposed to the average daily doses over the full 2-year study).
23 • The BMDL or LOAEL/NOAEL (in units of mg/kg-d Cr(VI)) was converted to an internal dose
24 using the PK model. The internal dose was the average rodent dose escaping reduction (in
25 mg/kg-d) multiplied by (BWa/BWh)1/4 in accordance with Recommended Use of Body
26 Weight3/4 as the Default Method in Derivation of the Oral Reference Dose (U.S. EPA. 2011c).
27 Study-specific time-weighted average body weights relevant to the data collection time
28 were used in the model and for the BW scaling step.
29 • The adult-based human PBPK model was used to estimate the daily mg/kg Cr(VI) dose that
30 must be ingested to achieve the internal dose calculated in step (2). To account for
31 interindividual variability, the human equivalent dose was determined by Monte Carlo
32 analysis. The lower 1% value of 20000 Monte Carlo PK simulations needed to achieve the
33 internal dose POD was used. As a result, the intraspecies uncertainty factor (UFh) was
34 lowered from 10 to 3 (the pharmacokinetic component of the uncertainty factor was
35 removed as it was accounted for with this analysis). See Appendix C.1.5.
^Alternatively for the small intestine, an internal dose to the small intestine may be derived by scaling the
un-reduced daily Cr(VI) intake rate by intestinal tissue volume (defined as pyloric flux, mg/L-d, by Thompson
etal. f201411. Because organ volumes vary between species by allometric relationships, using the pyloric flux
internal dose metric produces similar results as BW3/4 scaling of the un-reduced Cr(VI) dose.
This document is a draft for review purposes only and does not constitute Agency policy.
4-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
• The uncertainty factors are applied to derive the candidate values.
Rodent
Dose Response Data
Control
#/#
Low
#/#
Mid
#/#
High
#/#
BerKtniirk Dose A-idyns
Dose response
modeling
Solve daily oral dose
producing the equivalent
internal dose in humans
Apply uncertainty
factors
Rodent external dose BMDL
or LOAEL/NOAEL
Internal dose POD in
mg/kg-d Cr(VI)
escaping stomach
reduction
BMDL or LOAEL/NOAEL in
mg/kg-d Cr(VI) escaping
stomach reduction
Figure 4-2. Process for calculating the human equivalent dose for Cr(VI).
4.1.2.2. GI Tract Effects
Incidence data of diffuse epithelial hyperplasia of the duodenum in male mice from NTP
(20081 were amenable to BMD modeling with the highest dose omitted. A BMR of 10% extra risk
(ER) was applied under the assumption that it represents a minimally biologically significant level
of change in the absence of a biologically based BMR (U.S. EPA. 2012b). Diffuse epithelial
hyperplasia, although predictive of more severe manifestations of toxicity, is considered minimally
adverse and does not support using a lower BMR. Incidence data for male mice (all doses included)
are contained in HAWC.
Diffuse epithelial hyperplasia was not amenable to BMD modeling for female mice because
there was too much uncertainty in estimating the BMDL (see Appendix D.l.l). There were three
models which adequately fit the data in accordance with EPA's Benchmark Dose Technical Guidance
(U.S. EPA. 2012b). However, they produced significantly different BMDs and BMDLs, and one
model did not produce useful results due to an extremely low BMDL estimate and high BMD:BMDL
ratio. This is an indication that there was some model dependence of the estimates, and
uncertainty in the estimates was too great to be able to rely on the modeling results. The
uncertainty was primarily caused by the fact that the observed percent incidence at the lowest dose
(38%) was much higher than the BMR (10%). Because there are no data near where the true 10%
response occurs, estimating the BMDio and the 95% lower confidence limit on the BMDio is highly
uncertain. Alternative modeling approaches were explored; however, they could not address the
lack of low dose data near the target 10% extra risk response level. As a result, the LOAEL
approach was used (the LOAEL for hyperplasia in female mice was 0.302 mg/kg-d). Incidence data
for female mouse hyperplasia in the duodenum are available in HAWC.
This document is a draft for review purposes only and does not constitute Agency policy.
4-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofHexavalent Chromium
4.1.2.3. Hepatic Effects
For the liver, data for chronic liver inflammation in female mice from NTP (20081 were
amenable to BMD modeling. A BMR of 10% extra risk (ER) was applied under the assumption that
it represents a minimally biologically significant level of change. NTP (20081 described these
lesions as "minimal to mild severity", with "mild to moderate" in the higher dose groups. As a
result, a BMR lower than 10% was not considered.
Changes in the liver enzyme alanine aminotransferase (ALT) at 12 months in male rats from
NTP f20081 were amenable to BMD modeling. Several expert organizations, particularly those
concerned with early signs of drug-induced hepatotoxicity, have identified an increase in liver
enzymes compared with concurrent controls of two to fivefold as an indicator of concern for
hepatic injury (Sawicka and Dlugosz. 2017: EMEA. 2010: Boone etal.. 2005: Group. 20001. For this
assessment, a twofold increase in ALT is considered indicative of liver injury in experimental
animals. Thus, a BMR of 100% change from control (1 relative deviation from control) was applied.
Data for male and female rats in the subchronic study by NTP (20071 were not amenable to BMD
modeling64, and the lowest dose was identified as the LOAEL. The chronic study by NTP f20081
also provides subchronic data for ALT in male rats at 90 days. Because the chronic study used
lower doses, it was possible to identify a NOAEL65 of 1.58 mg/kg-d, and a LOAEL of 4.16 mg/kg-d
for increased ALT in male rats at 90 days (see Appendix C.1.5 for time-weighted average daily doses
of the first 90 days of exposure during the NTP f20081 2-year study).
Fatty liver change in female rats from NTP f20081 was not amenable to BMD modeling.
Similar to hyperplasia in the female mouse duodenum, uncertainty in estimating the BMDL was too
high (see Appendix D.l.l). As a result, the NOAEL (the lowest dose level, 0.248 mg/kg-d, which
exhibited less than 10% extra risk) was used as the POD for this dataset Similarly, chronic liver
inflammation in female rats from NTP (20081 was not amenable to BMD modeling and the LOAEL
(0.248 mg/kg-d, which exhibited greater than 10% extra risk) was used as the POD.
4.1.2.4. Developmental Effects
For NTP T19971. doses reported for the F0 dams66 were 11.6, 24.4, and 50.6 mg/kg-d Cr(VI)
(via feed). Decreased postnatal growth in the F1 generation was observed beginning at
24.4 mg/kg-d. Data are available for males (PND14 and PND211 and females (PND14 and PND211.
For postnatal growth in the F2 generation, effects were observed at the highest dose only (maternal
doses for females in the F1 generation were 7.27,17.19, 39.15 mg Cr(VI)/kg-d). Datasets for
64For female rats, the first nonzero dose had a very high response relative to other dose levels (click here to
see dose-response data). For male rats, the goodness-of-fit p-values were less than 0.1 for all statistical
models (even when removing the highest dose, which had a low response relative to other exposure levels).
Click here to see dose-response data for male rats.
65Data were not amenable to BMD modeling. No change from control was observed at the first nonzero dose.
66Maternal dam weight is highly correlated to offspring body weight. Because maternal body weight in this
study was also decreased, maternal dose is examined here instead of the averaged F0 male and female dose.
This document is a draft for review purposes only and does not constitute Agency policy.
4-11 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofHexavalent Chromium
postnatal growth were not amenable to BMD modeling because study statistics reported by the
authors were inadequate for use in multi-generational modeling67. A NOAEL of 11.6 mg/kg-d was
used based on outcomes observed in the F1 generation (see Section 3.2.9).
4.1.2.5. Hematological Effects
Male rat data of decreased Hgb at 22 days from NTP (20081 was amenable to BMD modeling
using a BMR of 1 standard deviation from the mean68. These data exhibited the most sensitive
response, and also contained the lowest dose range. With the exception of male rat data at 90 days
from NTP (20071 (which was also amenable to BMD modeling), all other datasets required a
LOAEL/NOAEL analysis. All available hematological data considered for dose-response modeling
are available in Appendix D.l.
4.1.3. Derivation of Candidate Values
This section describes the data and rationale for the selection of uncertainty factors and
derivation of candidate values for each identified human health hazard. The dose-response
modeling results and rodent-to-human extrapolations are summarized in Table 4-3. Further
details, including the BMDS modeling output and graphical results for the model selected for each
endpoint, can be found in Appendix D.l.
Table 4-3. Summary of derivation of points of departure following oral
exposure
Species/
sex
Model
BMR
BMD
mg/kg-d
BMDLor
LOAEiy
NOAEL
mg/kg-d
Internal dosea
mg/kg-d
TWA BW
kg
BW3/4
adjustb
PODhed mg/kg-
dayc
Diffuse epithelial hyperplasia of the duodenum at two vears (NTP, 2008)
Mice/M
Quantal
lineard
10%
ER
0.148
0.121
0.0182
0.05
2.88 x 10"3
0.0443
Mice/F
LOAEL
-
-
0.302
0.0463
0.05
7.32 x 10"3
0.0911
Increase in the liver enzvme alanine aminotransferase (ALT) (NTP, 2008)
Rat/M
12 mo
Expon.2d
1RD
1.83
1.56
0.170
0.395
0.0451
0.204
Rat/M 3 mo
NOAEL
-
-
1.58
0.165
0.246
0.0389
0.191
Increase in the liver enzvme alanine aminotransferase (ALT) at 90 davs (NTP, 2007
Rat/M
LOAEL
-
-
1.74
0.188
0.232
0.0436
0.203
67It was unclear whether standard errors reported for dose groups are based on variation among litters or
among pups across litters, and individual-level data are not available.
68When no biological information is readily available that allows for determining a minimally biological
significant response, the BMD Technical Guidance (U.S. EPA. 2012b) recommends a BMR based on one
standard deviation (SD).
This document is a draft for review purposes only and does not constitute Agency policy.
4-12 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Species/
sex
Model
BMR
BMD
mg/kg-d
BMDLor
LOAEiy
NOAEL
mg/kg-d
Internal dosea
mg/kg-d
TWA BW
kg
BW3/4
adjustb
PODhed mg/kg-
dayc
Rat/F
LOAEL
-
-
1.74
0.181
0.160
0.0383
0.190
Chronic liver inflammation at two vears (NTP, 2008)
Rat/F
LOAEL
-
-
0.248
0.0195
0.260
4.66 x 10"3
0.0669
Mice/F
Log-
logistic
10%
ER
3.70
1.33
0.225
0.05
0.0356
0.182
Liver fatty change at two years (NTP, 2008)
Rat/F
NOAEL
-
-
0.248
0.0195
0.260
4.66 x 10"3
0.0669
Decreased offspring growth (NTP, 1997)
Mouse/F
NOAEL
-
-
11.6
3.09
0.0240
0.407
0.700
Decrease in hemoglobin (Hgb) NTP (2008)
Rat/M 22 d
Exp-4
1SD
1.07
0.816
0.0705
0.138
0.0144
0.126
Rat/M 3 mo
NOAEL
-
-
1.58
0.165
0.246
0.0389
0.191
Rat/M 12
mo
NOAEL
-
-
2.49
0.336
0.395
0.0891
0.286
Decrease in hemoglobin (Hgb) NTP (2007)
Rat/M 90d
Exp-3
1SD
2.99
2.09
0.243
0.232
0.0564
0.227
Rat/M 23d
LOAEL
-
-
2.92
0.367
0.120
0.0722
0.259
Rat/F 90d
NOAEL
-
-
3.50
0.500
0.160
0.106
0.312
Rat/F 23d
LOAEL
-
-
2.97
0.370
0.105
0.0704
0.187
aDose escaping stomach reduction in rodent (mg/kg-d) estimated by PK modeling. Animal BW set to study/sex-
specific time-weighted average values for PK modeling. This explains the discrepancy in internal doses between
male and female rats having the same external-dose LOAEL for ALT changes at 90 days, and differences between
male rats at 3 months and 12 months.
bBW3/4 scaling adjustment: mg/kg-d multiplied by (BWa/80)1/4). Animal BW set to study/sex-specific time-weighted
average values for both BW3/4 scaling and bioassay PK simulation.
cPODhed in units of mg/kg-d Cr(VI) oral dose ingested by humans (lower 1% value of 20000 Monte Carlo PK
simulations needed to achieve the internal dose POD). See Appendix C.1.5 for details.
dData were amenable to BMD modeling with the highest dose omitted.
1
2 For BW3/4 scaling adjustment and PBPK modeling applied above, the mean body weight
3 recommended by EPA's Exposure Factors Handbook fU.S. EPA. 2011al (80 kg) was used. There is a
4 negligible difference in the PODs when using 70 kg fU.S. EPA. 19881 or 80 kg, and the final reference
5 value would be the same under either assumption.
6 Consistent with EPA's A Review of the Reference Dose and Reference Concentration Processes
7 fU.S. EPA. 20021. a series of five UFs were applied to the POD developed for each endpoint/study,
8 specifically addressing the following areas of uncertainty: interspecies uncertainty (UFa) to account
9 for animal-to-human extrapolation, and consisting of equal parts representing pharmacokinetic and
10 pharmacodynamic differences; intraspecies uncertainty (UFh) to account for variation in
This document is a draft for review purposes only and does not constitute Agency policy.
4-13 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Toxicological Review ofHexavalent Chromium
susceptibility across the human population, and the possibility that the available data may not be
representative of individuals who are most susceptible to the effect; LOAEL-to-NOAEL uncertainty
(UFl) to infer an exposure level where effects are not expected when a POD is based on a
lowest-observed-adverse-effect level (LOAEL); subchronic-to-chronic uncertainty (UFs) to account
for the uncertainty in using subchronic studies to make inferences about lifetime exposure, and to
consider whether lifetime exposure would have effects at lower levels (e.g., for studies other than
subchronic studies); and database uncertainty (UFd) to account for database deficiencies if an
incomplete database raises concern that further studies might identify a more sensitive effect,
organ system, or life stage. An explanation of the five possible areas of uncertainty and variability
follows:
• An intraspecies uncertainty factor, UFh, of 3 was applied to account for variability and
uncertainty in pharmacodynamic susceptibility in extrapolating to subgroups of the human
population most sensitive to the health hazards of Cr(VI) fU.S. EPA. 20021. In the case of
Cr(VI), the PODs were derived from studies in inbred animal strains and are not considered
sufficiently representative of the exposure and dose-response of the most susceptible
human subpopulations (see Section 3.3.1). In certain cases, the pharmacokinetic
component of this factor may be replaced when a PK model is available that incorporates
the best available information on variability in pharmacokinetic disposition in the human
population (including sensitive populations). In the case of Cr(VI), a Monte Carlo analysis
using PBPK modeling (see Appendix Section C.1.5) was applied to account for
pharmacokinetic variability, and 3 was retained for pharmacodynamic variability.
• An interspecies uncertainty factor, UFa, of 3 (101/2 = 3.16, rounded to 3) was applied to all
PODs to account for uncertainty in characterizing the pharmacokinetic and
pharmacodynamic differences between rodents and humans. For all datasets used in this
assessment, a PBPK model or BW3/4 scaling was used to convert doses in rodents to
equivalent doses in humans (see rationale in Section 4.1.2.1—Human Extrapolation). This
reduces pharmacokinetic uncertainty in extrapolating from the rodents to humans, but does
not account for interspecies differences due to pharmacodynamics. An UFa of 3 was applied
to account for this remaining pharmacodynamic and any residual pharmacokinetic
uncertainty not accounted for by the PBPK model.
• A subchronic-to-chronic uncertainty factor, UFS, of 1 was applied to all endpoints (GI tract
and liver effects) from the chronic 2-year (lifetime) study in rodents fNTP. 20081 where
exposure occurred for one year or more. For example, ALT changes in rats measured at one
year (12 months) were assigned an UFs of 1. An UFs of 1 was applied to the developmental
endpoint from NTP f!9971. because exposure occurred during the critical window. An UFs
of 1 was applied to decreased Hgb measured at subchronic timepoints from NTP (2008,
20071 because this effect is known to ameliorate over time, and therefore subchronic-
derived PODs will be health-protective for chronic exposure. An UFS of 3 was applied to
ALT changes from the 90-day study in rodents (NTP. 20071. and ALT changes reported at 3
months during the chronic NTP (20081 study. An UFs = 3 (rather than 10) was applied to
90-day data for ALT because data collected at multiple time points from NTP (20081
showed that these effects did not increase in severity between 90 days and 1 year. A value
of 3 was retained to account for the possibility that longer exposure may induce these
effects at a lower exposures fU.S. EPA. 20021. even if the effects themselves do not increase
This document is a draft for review purposes only and does not constitute Agency policy.
4-14 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
in severity. Also, there were no chronic ALT data for female rats, and females may be more
susceptible (based on the observed chronic liver inflammation at 2 years).
• A LOAEL-to-NOAEL uncertainty factor, UFl, of 1 was applied to PODs based on either a
NOAEL or a BMDL. An UFl of 10 (rather than 3) was applied to PODs based on the LOAEL of
ALT changes in rats observed from the 90-day study fNTP. 20071. because the magnitude of
change from control at the lowest dose was very high (180% for males and 585% for
females). These measurements were somewhat volatile (for example, the changes were
typically very large, and the magnitude of changes varied greatly between studies, even
among the NTP studies in the same species and sex which were conducted under very
similar conditions). As a result, the higher UFl was applied. Similarly, an UFl of 10 was
applied to the LOAELs of hyperplasia in the female mouse duodenum and chronic liver
inflammation in female rats from NTP f20081 because responses were high (>20% extra
risk) at the lowest dose. Thus, an UFl of 10 was applied to all PODs that were based on a
LOAEL.
• A database uncertainty factor, UFd, value of 1 was applied for all endpoints. The
toxicological database for oral exposure to Cr(VI) includes several occupational health
studies, and subchronic and chronic toxicity studies in multiple laboratory species. The
database also contains prenatal, multi-gene rational, and gestational oral studies in rodents.
Table 4-4 is a continuation of Table 4-3 and summarizes the application of UFs to each POD
to derive a candidate value for each endpoint, preliminary to the derivation of the
organ/system-specific reference values. These candidate values are considered individually in the
selection of a representative oral reference value for a specific hazard and subsequent overall RfD
for Cr(VI).
This document is a draft for review purposes only and does not constitute Agency policy.
4-15 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 4-4. Effects and corresponding derivation of candidate values
Endpoint and
Reference
PODhed
(mg/kg-day)
POD Type
UFa
UFh
UFl
UFs
UFd
Composite
UF
Candidate
value (mg/kg-d)
GI tract
Mouse (M)
hyperplasia (2
years! fNTP, 20081
0.0443
BMDLio%e
r
3
3
1
1
1
10
4.43 x 10"3
Mouse (F)
hyperplasia (2
years! fNTP, 20081
0.0911
LOAEL
3
3
10
1
1
100
9.11 x 10"4
Liver
Rat (M) liver ALT
C12 months! fNTP,
20081
0.204
BMDLird
3
3
1
1
1
10
0.0204
Rat (M) liver ALT
f3 months! fNTP,
20081
0.191
NOAEL
3
3
1
3
1
30
6.37 x 10"3
Rat (M) liver ALT
f90 davsl fNTP,
20071
0.203
LOAEL
3
3
10
3
1
300
6.77 x 10"4
Rat (F) liver ALT
f90 davsl fNTP,
20071
0.190
LOAEL
3
3
10
3
1
300
6.33 x 10"4
Rat (F) liver
chronic
inflammation (2
vearsl fNTP, 20081
0.0669
LOAEL
3
3
10
1
1
100
6.69 x 10"4
Mouse (F) liver
chronic
inflammation (2
vearsl fNTP, 20081
0.182
BMDLio%e
r
3
3
1
1
1
10
0.0182
Rat (F) liver fatty
change (2 yearsl
fNTP, 20081
0.0669
NOAEL
3
3
1
1
1
10
6.69 x 10"3
Developmental
Mouse (F)
Decreased F1
postnatal growth
fNTP, 19971
0.700
NOAEL
3
3
1
1
1
10
0.0700
Hematological
Rat (Ml Hgb (22d)
fNTP, 20081
0.126
BMDLisd
3
3
1
1
1
10
0.0126
Rat (M) Hgb (3 mo)
fNTP, 20081
0.191
NOAEL
3
3
1
1
1
10
0.0191
This document is a draft for review purposes only and does not constitute Agency policy.
4-16 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicologiccil Review of Hexavalent Chromium
Endpoint and
Reference
PODhed
(mg/kg-day)
POD Type
UFa
UFh
UFl
UFs
UFd
Composite
UF
Candidate
value (mg/kg-d)
Rat (M) Hgb
(12mo) [NTP,
20081
0.286
NOAEL
3
3
1
1
1
10
0.0286
Rat (M) Hgb (90d)
fNTP, 20071
0.227
BMDLisd
3
3
1
1
1
10
0.0227
Rat (M) Hgb (23d)
fNTP, 20071
0.259
LOAEL
3
3
10
1
1
100
2.59x10-3
Rat (F) Hgb [90d]
fNTP, 20071
0.312
NOAEL
3
3
1
1
1
10
0.0312
Rat (F] Hgb (23d1
fNTP, 20071
0.187
LOAEL
3
3
10
1
1
100
1.87x10-3
S Decreased offspring growth (female mice) (NTP, 1997)
Decreased Hb 22d (male rats) (NTP, 2008}
Fatty change 2yr (female rats) (NTP, 2008)
Chronic inflammation 2yr (female mice) (NTP, 2008)
Chronic inflammation 2yr (female rats) (NTP, 2008)
Increased ALT 90d (female rats) (NTP, 2007)
Increased ALT 90d {male rats) (NTP, 2007)
Increased ALT 3mo {male rats) (NTP, 2008)
Increased ALT 12mo (male rats) (NTP, 2008)
Hyperplasia 2yr (female mice) (NTP, 2008)
Hyperplasia 2yr (male mice) (NTP, 2008)
L
Composite UF
a Candidate Value
• POD(HED)
0.0001
0.001 0.01 0.1
log-scale Cr(VI) dose (mg/kg-d)
Figure 4-3. Candidate values with corresponding POD and composite UF.
This document is a draft for review purposes only and does not constitute Agency policy.
4-17' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
4.1.4. Derivation of Organ/System-Specific Reference Doses
Table 4-5 distills the candidate values from Table 4-4 into a single value for each organ or
system (organ/system-specific RfDs, or osRfDs). These organ or system-specific reference values
may be useful for subsequent cumulative risk assessments that consider the combined effect of
multiple agents acting at a common site.
Each candidate value was evaluated with respect to multiple considerations, including
strength of evidence, basis of the POD (i.e., BMD vs. NOAEL vs. LOAEL), and dose-response model
uncertainties. The confidence rating of each osRfD is based on three factors: the level of confidence
in the primary study, the health effect database associated with that reference value, and the
quantification of the POD.
4.1.4.1. GI tract Toxicity
The osRfD for GI effects was based on the incidence of diffuse epithelial hyperplasia of the
duodenum in female B6C3F1 mice reported in NTP f2008I Data in both males and females
indicated that females may be more sensitive to this effect While the result in males may have
higher certainty in quantification (since it is a BMDS-derived result with a lower uncertainty
factor), there is still significant uncertainty for females (the more sensitive group). Because the RfD
is intended to protect the population as a whole including potentially susceptible subgroups (U.S.
EPA. 2002). female data were selected for this osRfD. The hyperplasia in the GI tract following oral
exposures is considered to be representative of the constellation of histopathological observations
that together result in a change in tissue function that is considered an adverse noncancer effect.
An osRfD of 9 x 10~4 mg/kg-d (rounded from 9.11 x 10-4) was derived. There is high confidence in
this osRfD because it is based on chronic 2-year data from a high confidence study, and a strong
dose-response was exhibited in both male and female mice. High confidence subchronic studies
(click the HAWC link for study evaluation details) and mechanistic studies were supportive of these
effects.
4.1.4.2. Hepatic Toxicity
The osRfD for hepatic effects was based on the lowest candidate toxicity value from the
chronic data: chronic inflammation in female F344 rats reported in NTP (2008). Histological
changes were primarily observed in female rats and were less severe in male rats and mice.
Therefore, female rats may be the most sensitive group. Chronic hepatic inflammation can lead to
fibrosis (Kovama and Brenner. 2017). and the candidate value is also protective of the other
endpoints evaluated using chronic data (increased fatty changes and ALT). An osRfD of 7 x 10~4
mg/kg-d (rounded from 6.69 x 10-4) was derived. There is high confidence in this osRfD. It is
based on a high confidence chronic study in rats and there are other subchronic data and
mechanistic evidence to support the liver endpoints.
This document is a draft for review purposes only and does not constitute Agency policy.
4-18 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
4.1.4.3. Developmental Toxicity
The osRfD for developmental toxicity was based on the only candidate value: decreased F1
offspring postnatal growth from the continuous breeding study in BALBC mice (NTP. 19971. The
osRfD was 0.07 mg/kg-d. There is low confidence in this osRfD. While it is based on a high
confidence continuous breeding study and similar effects on decreased offspring growth observed
in multiple other studies (see Section 3.2.9, click the HAWC link for study evaluation details), this
effect only occurred in high dose groups where other toxicological effects (as indicated by the lower
points of departure in this section) may be occurring. For example, female mice in the F0
generation (dams) were exposed to 11.6, 24.4, 50.6 mg/kg-d Cr(VI) (NTP. 1997). The decreased F1
offspring growth effect was observed at maternal dose of 24.4 mg/kg-d, which is a relatively high
dose (NTP (2007) observed high incidence of stomach ulcers in rats at approximately 20 mg/kg-d).
Other studies in the database observing similar effects were lower confidence and used higher (or
unknown) doses. A lower osRfD confidence was assigned due to: 1) a weak health effects database
for this endpoint (most studies were rated low confidence), and 2) the possibility that other
unknown toxicities could be affecting the animals at the high dose. Thus, there was lowered
confidence due to the database of studies examining this endpoint, and lowered confidence in
quantification of the POD.
4.1.4.4. Hematological Toxicity
The osRfD for hematological toxicity was based on decreased Hgb in male F344 rats at 22
days reported in NTP f20081. This effect was observed to have the highest magnitude at short time
periods, and other short-term data (such as 23-day data from NTP f200711 were not as applicable
for low-dose extrapolation due to the higher dose ranges used. An osRfD of 0.01 mg/kg-d (rounded
from 0.0126 mg/kg-d) was derived. There is high confidence in this osRfD. It is based on a high
confidence study in rats that measured data at multiple time points (1 week, 22 days, 3 months, 6
months, 12 months). There are other subchronic datasets, mechanistic evidence, as well as
multiple hematological markers (such as MCV, MCH, MCHC, Hct) that also support this endpoint.
Table 4-5. Organ/system-specific RfDs and proposed overall RfD for Cr(VI)
Effect
Basis
osRfD
(mg/kg-day)
Exposure
Description
Confidence
Gl tract toxicity
Diffuse epithelial hyperplasia in small
intestine (female mice)
9 x 10"4
Chronic
High
Hepatic toxicity
Chronic inflammation (female rats)
7 x 10"4
Chronic
High
Developmental
toxicity
Decreased F1 offspring postnatal growth
(mice)
0.07
Continuous
breeding
Low
Hematological
toxicity
Decreased Hgb (male rats)
0.01
Subchronic
High
This document is a draft for review purposes only and does not constitute Agency policy.
4-19 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Toxicological Review ofHexavalent Chromium
Effect
Basis
osRfD
(mg/kg-day)
Exposure
Description
Confidence
Overall RfD
GI tract effects
9 x 10"4
Chronic
High
4.1.5. Selection of the Overall Reference Dose
Choice of the overall RfD involved consideration of both the level of certainty in the
estimated organ/system-specific values, as well as the level of confidence in the observed effect(s).
An overall confidence level was assigned to the RfD to reflect an interpretation regarding
confidence in the collection of studies used to determine the hazard(s) and derive the RfD, the RfD
calculation itself, as well as the overall completeness of the database on the potential health effects
ofhexavalent chromium exposure.
To estimate an exposure level below which noncancer effects from lifetime oral Cr(VI)
exposure are not expected to occur, the osRfD for GI effects, 9 x 10~4 mg/kg-d, is selected as the
overall RfD for Cr(VI). This was a high confidence value derived from chronic exposure data. The
overall RfD is derived to be protective of all types of noncancer effects for lifetime exposure and is
intended to protect the population as a whole including potentially susceptible subgroups (U.S.
EPA. 20021. While the osRfD for liver was slightly lower, the osRfD for GI effects is still lower than
most other candidate values considered for the liver osRfD (see Figure 4-3). With the exception of
chronic liver inflammation in female rats, candidate values for the osRfD for liver effects that were
based on chronic exposure data (12 months or 2 years; see Figure 4-3) were above 9 x 10~4 mg/kg-
d. Candidate liver values derived from subchronic data that were lower than 9 x 10~4 mg/kg-d had
cumulative uncertainty factors of 300, whereas other candidate values had uncertainty factors of
100 or less. Because the GI tract is exposed to higher concentrations of un-reduced Cr(VI) than the
liver, it is likely to be more susceptible to the effects of ingested Cr(VI). Thus, the osRfD for GI
effects was selected as the overall RfD.
This value (9 x 10~4 mg/kg-d) should be applied in general population risk assessments.
However, decisions concerning averaging exposures over time for comparison with the RfD should
consider the types of toxicological effects and specific life stages of concern. For example,
fluctuations in exposure levels that result in elevated exposures during various life stages could
potentially lead to an appreciable risk, even if average levels over the full exposure duration were
less than or equal to the RfD.
4.1.6. Uncertainties in the Derivation of Reference Dose
The RfD was derived based on GI effects (diffuse epithelial hyperplasia in the duodenum) of
female mice exposed to Cr(VI) in drinking water for two years (NTP. 2008). Some of the
uncertainty considerations related to the RfD derivation are outlined below and in Section 3.3.
This document is a draft for review purposes only and does not constitute Agency policy.
4-20 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
4.1.6.1. Site Concordance and Human Relevance
The GI tract reference value was based on an effect observed in the small intestine of mice,
however it is possible that the effect may be exhibited in different sections of the alimentary tract in
the human (specifically, the oral cavity, esophagus, and stomach). Estimated Cr(VI) exposure to the
stomach epithelium may be similar to exposure to the small intestine epithelium, since both would
be strong functions of gastric pH, Cr(VI) concentration and reduction rate. There are differences in
morphologies between the small intestine and stomach that could potentially impact the tissue
susceptibility. Effects in the rodent stomach only occurred at the high doses of the 90-day NTP
(20071 study. Rodents exposed to Cr(VI) during the 2-year NTP (20081 study did not exhibit effects
in the stomach.
Exposure to the oral cavity and esophagus occurs prior to Cr(VI) reduction in the stomach.
However, no noncancer effects were observed in these tissues during the NTP (2008) or NTP
f20071 bioassays (aside from mild salivary gland atrophy in rats during the 2-year study).
4.1.6.2. Susceptible Populations
A significant fraction of the human population may be highly susceptible to Cr(VI)-induced
effects in the GI tract due to high stomach pH. Individuals with hypochlorhydria (low stomach acid)
have consistently high stomach pH that may exceed 8 (Feldman and Barnett. 1991). Less than 1%
of the adult population may exhibit hypochlorhydria, whereas 10-20% of the elderly population
(aged 65 and up) may exhibit this condition (Russell etal.. 1993). For individuals without this
medical condition, there is still high variability fFeldman and Barnett f!9911 estimated that 5% of
men may exhibit basal pH exceeding 5, and 5% of women may exhibit basal pH exceeding 6.8). Gut
microbiota and gastric juice chemistry in individuals with high gastric pH may differ from those in
the general population. It is not known how effective Cr(VI) can reduce to Cr(III) in this type of
gastric environment Data by Kirman etal. (2016). which included some groups with high stomach
pH, were highly variable.
Individuals taking medication to treat gastroesophageal reflux disease (GERD), including
calcium carbonate-based acid reducers and proton pump inhibitors, have an elevated stomach pH
during treatment This is known to be a significant fraction of the population since up to 20% of the
population may be afflicted by GERD, and the gastric pH for these individuals may be above 4
throughout the day during successful treatment (Delshad etal.. 2020: GBP 2017. 2020: Lin and
Triadafilopoulos. 2015: Burdsall etal.. 2013: Atanassoff et al.. 1995). A sensitivity analysis was
performed on the human model (Appendix C.1.5), assuming a baseline stomach pH = 4 (as opposed
to 1.3). It was found that for internal PODs above 0.001 mg/kg-d (which apply to all the PODs), the
current Monte Carlo approach (taking the lower 1% of 20,000 simulations of the standard
population with baseline stomach pH = 1.3) was protective for the population with baseline pH = 4.
For populations with baseline pH higher than 4, candidate values derived using the
pharmacokinetic approach would not be health-protective. Appendix D.3 contains candidate values
This document is a draft for review purposes only and does not constitute Agency policy.
4-21 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofHexavalent Chromium
calculated by default approaches without adjustment for gastric reduction, which may be health-
protective at low doses for the pH > 4 population (since those results implicitly assume gastric pH
and reduction capacity in rodents and humans are equivalent).
Uncertainties related to extremely high gastric pH, as well as other conditions that could
lead to pharmacokinetic susceptibility (H. pylori infection, gastric bypass, gastrectomy) cannot be
accounted for quantitatively. High interindividual variation was observed in ex vivo data by
Kirman etal. (2016). both in health individuals with high stomach pH and individuals taking proton
pump inhibitors. Additionally, no data are available studying Cr(VI) reduction in the gastric
environments of children, toddlers, or infants. As a result, PBPK modeling was not performed for
these groups, and there may be some residual pharmacokinetic uncertainty not accounted for by
the UFh-
4.1.6.3. Rodent Gastric Modeling Uncertainties
Stomach reduction in the mouse may be impacted by a number of factors. Higher reduction
efficiency may occur during the ingestion of a solid meal, since gastric emptying is delayed, and pH
is decreased (for the mouse, glandular stomach pH is decreased by the fasted state, while the
opposite is true for humans). However, this effect may be counteracted by kinetics in the
forestomach, which humans do not have. The forestomach may not follow the same fed/fasted
pattern as the glandular stomach (Ward and Coates. 1987).
The rodent glandular stomach actively secretes digestive enzymes shortly before, during,
and after a solid meal. The precise dynamics of gastric changes are uncertain, and the "well-mixed"
PBPK model assumption may not be accurate due to ongoing food consumption. In addition, the
rodent forestomach contents may have an elevated pH relative to the glandular stomach (Kohl et
al.. 2013: Browning etal.. 1983: Kunstvr etal.. 1976). and ingested drinking water passes through
both of these stomach regions.
There are also uncertainties related to the pH-kinetic relationship. The dose-response
analysis for this assessment applied rodent pH of greater than 4.0, setting pH to values at which the
rodent ex vivo reduction experiments were performed. Prior to dilution with water, Proctor et al.
f20121 estimated the rodent stomach pH to be approximately 4, but it was increased to
approximately 4.5 after dilution with water for the experiments. The precise relationship between
pH and reduction kinetics in the rodent at lower pH is uncertain, and therefore it was desirable to
perform simulations assuming rodent pH of 4.0 or higher. If the true rodent stomach pH is lower,
or if the reduction kinetics are faster than estimated by the current model, this would ultimately
lead to a decreased RfD. On the other hand, the model already estimates a low percentage of Cr(VI)
escaping the rodent stomach (5-10%). If the true percentage was lower than this, it would mean
that a negligible amount of Cr(VI) enters the mouse small intestine following ingestion. It has been
confirmed by multiple pharmacokinetic studies that Cr(VI) is absorbed systemically in rodents
following exposure via drinking water. Data by Kirman etal. (2012) show chromium
concentrations in the duodenum increasing with a linear or supralinear relationship with respect to
This document is a draft for review purposes only and does not constitute Agency policy.
4-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
dose in mice exposed to Cr(VI) in drinking water for 90 days. Data by NTP f20081 show elevated
tissue chromium for all chronically-exposed groups. Therefore, assuming that in vivo rodent
gastric reduction occurs very effectively (i.e., 99% reduction) would not be consistent with the
available pharmacokinetic data.
4.1.6.4. Human Gastric Modeling Uncertainties
As with the rodent gastric system, there are uncertainties in modeling the human stomach.
There exist complex gastric and intestinal kinetic models, and many of the parameters are highly
variable (Paixao etal.. 2018: Talattof and Amidon. 2018: Yu etal.. 2017: Hens etal.. 2014: Mudie et
al.. 2010: ICRP. 20061. While the PBPK model in this assessment adopts some parameters and
concepts from literature, and incorporates Monte Carlo analysis, it may not account for all
uncertainty and variability. Ex vivo data for Cr(VI) reduction in gastric juices show high
interindividual variability De Flora etal. (20161: Kirman etal. (20161. Interindividual variability in
gastric contents and microbiota likely introduces variation in Cr(VI) reduction. Variability in
reduction kinetic parameters (with the exception of the reducing capacity parameter) was not
incorporated into the model. Furthermore, there are no data for Cr(VI) reduction in the gastric acid
of infants and toddlers, and there would be significant uncertainties in applying the adult-based
PBPK model to infant or child physiology.
4.1.6.5. Uncertainty in Systemic Pharmacokinetics
The current approach uses a PBPK model of the stomach lumen to adjust the average daily
oral Cr(VI) dose to account for detoxification in the stomach compartment It does not explicitly
model systemic whole-body pharmacokinetics. While whole-body PBPK models are available for
Cr(VI), the uncertainties related to the systemic pharmacokinetics in rodents and humans are high,
especially at low doses. However, most endpoints observed following oral ingestion were in or
near the GI tract, and therefore may not require an accounting of systemic chromium. Cr(VI) which
enters the intestinal lumen may expose the systems in which effects were observed (the small
intestine, and the liver by first-pass effect) prior to distribution to systemic circulation. Reduction
of Cr(VI) in the blood and other tissues is rapid, and this assessment neglects the impact that re-
circulating Cr(VI) may have on the liver and small intestine. It is health-protective to assume that
any unreduced Cr(VI) emptying into the human small intestine is absorbed.
For systemic effects, there is some residual pharmacokinetic uncertainty. The modeling
does not take into account how much Cr(VI) may remain in the GI epithelium (or be reduced by the
G.I. tissues, liver, and blood). This loss of Cr(VI) available to absorb into systemic tissues is
neglected in both animals and humans.
This document is a draft for review purposes only and does not constitute Agency policy.
4-23 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
4.1.6.6. Uncertainty in Dose-response Modeling
For the two osRfDs (diffuse epithelial hyperplasia in female mice, and chronic liver
inflammation in female rats from NTP (200811. there was uncertainty related to the dose-response
modeling. Thus a NOAEL/LOAEL approach was used.
As noted in Section 4.1.6, diffuse epithelial hyperplasia was not amenable to BMD modeling
for female mice because there was too much uncertainty in estimating the BMDL. Estimates of the
epithelial hyperplasia RfD from female mice using BMD modeling (without dropping doses) span
from 7.95 x 10"5 mg/kg-d to 2.04 x 10~3 mg/kg-d (see Appendix D.l.l). The GI tract osRfD (derived
by a LOAEL, which resulted in a higher uncertainty factor) falls within this span and differs by
approximately 15% from both the mean and median value of the three adequately fit models (1.06
x lO-3 mg/kg-d). If dropping the two highest doses (as was done by ATSDR (201211 and
performing BMD modeling, the resulting RfD would be 2.6 x 10~3 mg/kg-d (and rounded to 3
x lO-3 mg/kg-d). EPA's Benchmark Dose Technical Guidance fU.S. EPA. 2012bl states dropping dose
groups should only be done when an adequate model fit cannot be achieved. This situation did not
apply to female mouse hyperplasia, because multiple adequate fits were achieved when including
all dose groups, but there was too much uncertainty in the BMD estimate to use these model results
for determining the POD.
Similarly, chronic liver inflammation in female rats from NTP (20081 was not amenable to
BMD modeling. Estimates of the candidate values for this endpoint span from 1.00 x 10~4 to 4.02
x lO-3 (see Appendix D.l.l). The liver osRfD (derived by a LOAEL, which resulted in a higher
uncertainty factor) falls within this span and is about 2x lower than the mean and median values of
the three adequately fit models (mean: 1.80 x 10~3 mg/kg-d, median: 1.28 x 10~3 mg/kg-d).
4.1.7. Confidence Statement
An overall confidence level of High, Medium, or Low was assigned to reflect the level of
confidence in the study(ies) and hazard(s) used to derive the RfD, the overall database, and the RfD
itself, as described in EPA's Methods for Derivation of Inhalation Reference Concentrations and
Application of Inhalation Dosimetry §4.3.9.2 fU.S. EPA. 19941.
The confidence in the overall chronic RfD is high. The RfD is based on a high confidence
chronic 2-year drinking water study by NTP (20081 which exposed rats and mice of both sexes to
Cr(VI) as sodium dichromate dihydrate at drinking water concentrations from 5 mg/L to 180 mg/L
(approximately 0.2 mg/kg-d to 10 mg/kg-d). Multiple high confidence subchronic studies also
support these data (click the HAWC link for study evaluation details), and mechanistic studies
support oxidative stress as a mechanism of Cr(VI) toxicity in a variety of tissues, including the GI
tract Although the value is based on a LOAEL, the final result is supported by BMD modeling
results for hyperplasia in the duodenum of male mice from the same study, and is within the range
of adequately fit models that could not be utilized (see above in Section 4.1.6.6). The osRfD for the
liver is also supportive of the GI tract RfD, because the GI tract and liver are exposed on first-pass
This document is a draft for review purposes only and does not constitute Agency policy.
4-24 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Toxicological Review ofHexavalent Chromium
following oral ingestion (so both should get the highest internal dose). While the human database
for Cr(VI)-induced GI toxicity was indeterminate, this did not warrant changing the overall
confidence from high.
4.1.8. Previous IRIS Assessment: Oral Reference Dose
The previous RfD assessment for Cr(VI) was completed in September 1998. The previous
RfD was based on a NOAEL identified from a 1-year drinking water study in rats in which animals
were exposed to Cr(VI) fMacKenzie etal.. 1958169. MacKenzie etal. f!9581 monitored body weight,
gross external conditions, histopathology and blood chemistry and did not observe any effects at
any level of treatment. A NOAEL of 2.5 mg/kg-day was identified. A composite uncertainty factor
of 300 (10 for interspecies extrapolation, 10 for intraspecies extrapolation, and 3 for subchronic-to-
chronic extrapolation) and a modifying factor of 3 (to account for concerns raised by the
epidemiology study of Zhang and Li (1987a)) were applied to this POD to yield an oral RfD of 3
x lO-3 mg/kg-d.
4.2. INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER
THAN CANCER
The reference concentration (RfC, expressed in units of mg/m3) is defined as an estimate
(with uncertainty spanning perhaps an order of magnitude) of a continuous inhalation exposure to
the human population (including sensitive subgroups) that is likely to be without an appreciable
risk of deleterious effects during a lifetime. It can be derived from a NOAEL, LOAEL, or the 95%
lower bound on the benchmark concentration (BMCL), with uncertainty factors generally applied to
reflect limitations of the data used. As noted in Section 3.3.2, derivation of the RfC was limited to
effects in the respiratory tract
Upper respiratory toxicity in the form of nasal effects in humans has been determined
previously (see Protocol Section 3.1.2, Appendix A), and a set of human studies were evaluated for
data that may inform the quantitative dose-response analysis (this will be discussed in Section
4.2.1). Data suitable for RfC derivation of upper respiratory effects were only available from human
studies (and these were limited to effects in the nasal airways). Data from animals of effects in the
upper respiratory tract (such as reported nosebleeds and other qualitative effect descriptions)
were not considered due to the availability of quantitative dose-response data in humans.
Based on findings from inhalation studies in experimental animals and occupational studies
in humans, evidence indicates that Cr(VI) is likely to cause lower respiratory toxicity in humans
(see Section 3.2.1). Data suitable for RfC derivation of lower respiratory effects were only available
69This study was determined to meet PECO criteria in the current assessment; however, the overall
confidence was rated uninformative due to insufficient reporting of the outcomes, survival, and sample sizes
of evaluated animals. Normally in situations concerining poor reporting, authors may be contacted for
clarifications that may result in upgraded confidence ratings, but this was not possible due to the age of the
publication.
This document is a draft for review purposes only and does not constitute Agency policy.
4-25 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Toxicological Review ofHexavalent Chromium
from animal studies. All human studies of these effects were low confidence and only provided
information on associations (and did not provide dose-response data).
4.2.1. Identification of Studies for Dose-Response Analysis of Selected Effects
4.2.1.1. Upper Respiratory Tract Effects
Effects in the nasal cavity of humans are well-established hazards of inhaled Cr(VI)
exposure, and this review focused on data that may improve the quantitative dose-response
analysis conducted in EPA's 1998 IRIS assessment (see Protocol Section 3.2, Appendix A)70.
Quantitative animal data for effects in the upper respiratory tract were not available. Qualitative
findings in rodents such as obstructive respiratory dyspnea (Glaser etal.. 19901. or "peculiar sound
during respiration" and periodic nose bleeds (Kim etal.. 20041 were not considered for dose-
response assessment due to the availability of human data. No other effects in the upper
respiratory tract outside of the nasal cavity were identified during hazard identification (Section
3.2.1).
The epidemiological database for inhalation of Cr(VI) mainly consists of observational
studies of workers exposed in occupational settings. Human studies were considered suitable for
dose-response analysis and toxicity value derivation if they met the criteria listed below.
Furthermore, preference was given to studies with medium or high overall confidence ratings based
on study evaluation and to studies with larger sample sizes and exposures in the lower range of
human exposures, as these are most likely to represent the relationship between inhalation
exposure to Cr(VI) and adverse effects in the general population.
The following considerations were made during evaluation of studies for derivation of
inhalation toxicity values from human data:
• The study population must be exposed to Cr(VI) (as opposed to Cr(III)) based on air
measurements or job history and industry
• Quantitative estimates relating exposure (or dose) to the core outcomes considered
• Concentration of Cr(VI) in air must be measured at the study site
• Quality of measurements will depend on: type of sampling (personal, stationary, or both);
frequency of sampling; sampling duration; number of samplers; sampling methods
• Exposure to Cr(VI) for individuals or groups of individuals must be estimated with
reasonable accuracy and precision in units of air concentration
70A large literature database exists presenting qualitative evidence for an association between inhalation
Cr(VI) exposure and nasal effects (see ATSDR (201211. These qualitative studies, which presumably would
have varying confidence ratings, were not evaluated in this assessment. Study confidence ratings for the
quantitative data in this assessment do not impact EPA's determination that nasal effects are well established
hazards of inhalation Cr(VI) exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
4-26 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
• If exposure is categorical, it must have corresponding air concentration estimates for each
category
• Exposure is not solely quantified in units of concentration in a biological sample such as
urine or blood
The core outcomes for nasal effects in humans considered for evaluation of dose response
included the following clinical outcomes diagnosed by a trained examiner (e.g., physician,
otolaryngologist, or trained researcher): atrophy of the nasal mucosa, ulceration of the nasal
mucosa or septum, perforation of the septum, and bleeding nasal septum. The development of
these outcomes is highly specific to exposure to Cr(VI) and occurrence outside this exposure
scenario is extremely rare. Consistent with this specificity of outcome, perforation of the septum
has been known as "chrome hole" since the early days of chromium-related industries (including
chromate production and electroplating fBloomfield and Blum. 192811. Furthermore, the presence
of nasal pathologies considered here are occasionally used as supplemental information to confirm
exposure to chromium in studies of non-nasal outcomes (Cimineraetal.. 2016: Gibb etal.. 2015:
Machle and Gregorius. 19481. The specificity of this outcome to Cr(VI) exposure makes it ideal for
the estimation of the dose-response relationship for noncancer effects in humans.
There were over 20 peer-reviewed studies of nasal effects that contained information
related to endpoints in the nasal cavity, but these did not meet all criteria for dose-response
analysis outlined above and were therefore not evaluated. There were also five non-peer-reviewed
reports examining effects in the nasal cavity available from the National Institute for Occupational
Safety and Health (NIOSH). These include Ceballos et al. (20171. Zev and Lucas (19851. Lucas
(19761. Lucas and Kramkowski (19751. Cohen and Kramkowski (19731 and Almaguer and
Kramkowski (1983171. Many of these studies did not have multiple exposure groups (either a
referent or low/high concentration groups). Exposure and health effect data from these studies
were only available for short time periods, and data were only collected after health effects were
reported for the purpose of evaluating plant industrial hygiene practices (potentially leading to
bias). As a result, most of these were excluded for dose-response consideration. Only data from
Cohen and Kramkowski (19731 and its related peer-reviewed study (Cohen etal.. 19741 were
considered since this study contained a referent group. All studies excluded based on criteria above
are listed at the bottom of Table 4-6, and detailed rationale for why each of these were not
considered is provided in Appendix D.4 Table D-25.
Four peer-reviewed studies (some of which were associated with additional related studies
containing exposure or study design information) initially met the criteria to be considered for
toxicity value derivation and underwent formal study evaluation using HAWC. These were Gibb et
al. (2000a). Lindberg and Hedenstierna (1983). Cohen etal. (1974). and Hanslian et al. (1967). All
were conducted in occupational settings and the study populations were workers in either the
71The cited reports were published by the National Institute for Occupational Safety and Health (NIOSH).
Author names listed in these citations are the NIOSH investigators.
This document is a draft for review purposes only and does not constitute Agency policy.
4-27 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicological Review ofHexavalent Chromium
chromate production or chrome electroplating industries. One study of 2,307 chromate production
workers Gibb etal. f2000al though retrospective in design, utilized company records of air
concentration data, individual job and task data, and data from regular medical examinations, to
construct a dataset that included individual exposure estimates for each worker as well as the time
from baseline exposure to the incident event of the health outcome (see Table 4-23 in Section
4.4.5). The other three studies fLindberg and Hedenstierna. 1983: Cohen etal.. 1974: Hanslian et
al.. 19671 were cross-sectional in design and were conducted in smaller study populations
composed of chrome electroplating workers. The populations were adults, and the largest cohort
fGibb etal. f2000al which had a population size of 2307) only had male workers.
Three studies were classified as medium confidence fGibb etal.. 2000a: Lindberg and
Hedenstierna. 1983: Cohen etal.. 1974). and one study was low confidence (Hanslian et al.. 1967).
Because of the availability of medium confidence studies, data from Hanslian etal. (1967) were no
longer considered for dose-response. In addition to the usual factors considered during study
evaluation, diagnosis of nasal outcomes after physical examination of the nasal cavity by a trained
examiner was considered when determining confidence ratings for nasal effects studies. Additional
study details, including the reported endpoint data, are provided in Table 4-7.
This document is a draft for review purposes only and does not constitute Agency policy.
4-28 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 4-6. Evaluation of epidemiology studies on Cr(VI) and nasal effects.
to see interactive data graphic for rating rationales.
Click
¦a
a>
¦c
-a
ai
¦c
Reference
Gibb et al. (2000a) related:
Gibbet al. (2015)
Braver et al. (1985); Haves
et al. (1979)
Lindberg and Hedenstierna
(1983)
Cohen et al. (1974)
Related: Cohen and
Kramkowski (1973)
Hanslian et al. (1967)
Study description
Occupational longitudinal study. Male workers in a
chromate production plant in Baltimore, MD
(n = 2307).
Cross-sectional study. Male and female employees
in chrome-plating industry (n = 104). Office
employees (n = 19) as reference group
Cross-sectional study. White male and female
electroplating workers in nickel-chrome
department (n = 37)
Randomly-chosen workers employed in other areas
of the plant not significantly exposed to chromic
acid as reference group (n = 15)
Cross-sectional study. Male and female chrome-
plating workers (n = 77). 53 working directly with
baths, 23 working directly with chromium.
No reference group.
Study evaluation
aQ a
DAD
DO
c
> t
~ o
.Ł Q.
¦M
-------
Toxicological Review ofHexavalent Chromium
Table 4-7. Dose-response data for effects in the nasal cavity of humans (medium confidence studies)
Study
Exposure
Conf
Result Format
Effects
Lindberg and
Chrome
MED
Number of
Ulceration
Atrophy
Perforation only*
Hedenstierna
plating
cases
8-hr mean air ug Cr(VI)/m3
8-hr mean air ug Cr(VI)/m3
8-hr mean air ug Cr(VI)/m3
(1983)
Group n cases (%)
Group n cases (%)
Group n cases
<1.9 19 0
<1.9 19 4(21)
<1.9 19 0
2-20 24 8 (33)
2-20 24 8 (33)
2-20 24 3 (13)
Highest air ug Cr(VI)/m3
Highest air ug Cr(VI)/m3
Highest air ug Cr(VI)/m3
Group n cases
Group n cases
Group n cases
0.2-1.2 10 0
0.2-1.2 10 1 (10)
0.2-1.2 10 0
2.5-11 12 0
2.5-11 12 8 (67)
2.5-11 12 0
20-46 14 7 (50)
20-46 14 0
20-46 14 3 (21)
*2 w/ulceration also had perfor-
ation (total w/ perforation = 5)
Gibb et al.
Chromate
MED
Cumulative
Ulcerated nasal septum
Perforated nasal septum
Ulcerated septum relative risk
(2000a)
production
incidence (%)
Effect: 62.9%
Effect: 17.3%
Adjusted relative risk for a 0.1 mg
(n = 2307),
Mean (median) exposure:
Mean (median) exposure:
CrC>3/m3 increase (in ambient
onset time, and
0.054 (0.020) mg CrOs/m3
0.063 (0.021) mg CrOs/m3
air) = 1.2 (by Cox proportional
relative risk
or 28 (10) ug Cr(VI)/m3
or 33 (11) ug Cr(VI)/m3
hazards model adjusted for
(ulceration only)
calendar year at hire and age at
Mean (median) time on job
Mean (median) time on job (days)
hire, p = 0.0001).
(days) from date first hired to
from date first hired to date of
date of first diagnosis: 86 (22)
first diagnosis: 313 (172)
Cohen et al.
Chrome
MED
Prevalence (%)
Nasal ulceration parameter cases, number (%)
non-exposed (n = 15)
(1974)
plating
(with grading by
exposed (n = 37)
severity)
nasal mucosa (grade 0)
14 (93) 2 (5)
shallow erosion of septal mucosa (grade 1)
0 8 (22)
ulceration and crusting of septal mucosa (grade 2)
0 12 (32)
avascular, scarified areas of septal mucosa w/o erosion or ulceration
(grade 3)
0 11 (30)
perforation of septal mucosa (grade 4)
1(7) 4(11)
Exposed group area breathing zone: mean = 2.9 (ND-9.1) ug Cr(VI)/m3
Referent area breathing zone: 0.3 (0.1-0.4) ug Cr(VI)/m3
1 mg CrC>3 = 0.52 mg Cr(VI).
This document is a draft for review purposes only and does not constitute Agency policy.
4-30 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofHexavalent Chromium
4.2.1.2. Lower Respiratory Tract Effects
The inhalation animal toxicological database for Cr(VI) consists of studies with chronic,
subchronic, and/or acute data. Many of these studies analyzed similar or identical toxicological
endpoints, particularly for the respiratory system. Within the endpoint-specific databases for
hazard identification, a subset of these studies were considered for toxicity value derivation based
on factors outlined in Section 4.1.1. Preference was given to studies with larger sample sizes and
low concentrations, to facilitate extrapolation to levels typical of environmental human exposure
fU.S. EPA. 2012b). For inhalation studies of particulates, studies that provided measures of particle
size and distribution were preferred. Because of the availability of studies that were rated medium
confidence for lower respiratory tract endpoints, low confidence studies were not considered for
candidate value derivation. An outline of the process used to select candidate animal datasets for
dose-response analysis and candidate value derivation is provided in Figure 4-4.
This document is a draft for review purposes only and does not constitute Agency policy.
4-31 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Figure 4-4. Evaluation of animal studies from the Cr(VI) hazard identification
for derivation of toxicity values. Low confidence studies were not considered.
This document is a draft for review purposes only and does not constitute Agency policy.
4-32 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Toxicological Review ofHexavalent Chromium
Table 4-8. Design features of inhalation studies that examined effects in
animals
Study reference
Species/
strain and
sex
Exposure
duration
Dose
groups3
Animals/
group
Chemical and
particle size
Concentration
range (mg/m3
Cr[VI])
Glaser et al. (1990)
Wistar Rat,
Male
30/90 days (22
hr/day, 7
d/wk)
4
10
Sodium dichromate
MMAD 0.28 (±1.63)
Hm bottom two dose
groups
0.39 (±1.72) nm high
groups
0.05-0.4
Glaser et al. (1985)
Wistar Rat,
Male
28/90 days (22
hr/day, 7
d/wk)
3
10
Sodium dichromate
MMD0.2 (±1.5) nm
0.025-0.2
Johansson (1986b;
1986a)
Rabbit,
Male
4-6 weeks
(inexact), 6
hr/day, 5 d/wk
1
8
Sodium dichromate
MMAD 1 nm
(approx.)
0.9
Cohen et al. (2003)
F344 Rat,
Male
48 weeks, 5
hr/day, 5 d/wk
1
30
Calcium chromate
MMAD 0.6 (±1.7) nm
0.36
Kim et al. (2004)
Sprague-
Dawley
Rat, Male
90 days, 6
hr/day, 5 d/wk
3
5
Chromium trioxide
(size not reported)
0.2-1.25
aNumber does not include control group.
Table 4-4 outlines the inhalation studies rated medium or higher confidence for respiratory
tract endpoints (all were rated medium confidence for lung histopathology cellular responses; see
Section 3.2.1). Of the studies listed in Table 4-8 the Glaser etal. (1990: 19851 studies were
preferred for candidat value derivation due to the number of exposure groups, sample sizes, and
reporting of endpoints, methods, and particle sizes. Kim etal. f20041 did not report quantitative
data for effects or chromium particle size, and effects, and Johansson (1986b: 1986a) and Cohen et
al. (2003) only used a single high exposure group.
Lung histiocytosis, bronchioalveolar hyperplasia, and increased total protein and albumin
in BAL fluid were observed by Glaser etal. (1990) after 90 days of exposure, and these measures
remained slightly elevated after a 30-day recovery period (see Section 3.2.1). Although lactate
dehydrogenase (LDH) in BAL fluid returned to normal following the 30-day recovery period, LDH is
considered a sensitive indicator of cellular injury fHenderson etal.. 19851. and there was a clear
dose-response relationship. Dose-response data from Glaser etal. (1990) following 90 days of
exposure (with and without the 30-day recovery period) are presented in Figures 4-5 and 4-6.
Because histopathological and cellular changes occurred together, and in combination with serum
biomarkers indicating an inflammatory response (Nikula et al.. 2014). all exposure levels were
considered to have induced adverse responses.
This document is a draft for review purposes only and does not constitute Agency policy.
4-33 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
endpoint observation time mg/m3
Bronchioalveolar Hyperplasia 90.0 days
0.054
0.109
U.k!U4
0.403
120.0 days
0.054
0.109
0.204
0.403
Lung histiocytosis 90.0 days
0
0.054
0.109
0.204
120.0 days
0.403
0
0.054
- .
0.109
0.204
0.403
1
1 1 1 1
—I 1 1 1
0 10 20 30 40 50 60 70 80 90 100
% incidence
Figure 4-5. Dose-response relationship for lung histopathological in male rats
using data from Glaser et al. (1990). Data are for 90-day observation time
immediately following exposure, and 120-day observation time (90 days of
exposure followed by a 30-day period of no exposure). N = 10/group. Click here for
interactive graphic.
This document is a draft for review purposes only and does not constitute Agency policy.
4-34 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
endpoint observation time mg/m3 Percent change from control
Albumin in BALF 90.0 days 0
0.054
T
•
1
0.109
h*H
0.204
h« i
0.403
1 •
120.0 days 0
•
1
0.054
1 • -1
0.109
1 • -1
0.204
1 • —1
0.403
•
LDH in BALF 90.0 days 0
0.054
>•'
0.109
¦•4
0.204
1—•—1
0.403
1— •
120.0 days 0
0.054
0.109
>•-
0.204
1 • —1
0.403
• 1
Total protein in BALF 90.0 days 0
0.054
T
•
1
0.109
• i
0.204
• —l
0.403
— • 1
120.0 days 0
i • '
0.054
0.109
0.204
1—•—1
0.403
T
•
1
-50 0 50 100 150 200 250 300 350 400 4*
% change
Figure 4-6. Dose-response relationship for selected endpoints in male rats
using data from Glaser etal. (1990). Data (± 95% confidence interval) are for 90-
day observation time immediately following exposure, and 120-day observation
time (90 days of exposure followed by a 30-day period of no exposure).
N = 10/group. Click here for interactive graphic.
1 The endpoints and datasets used for dose-response of lower respiratory tract effects were:
2 • BAL fluid measurements of total protein, albumin, and LDH from Glaser et al. (1990) at
3 90 days
4 • Lung histopathological findings of histiocytosis and bronchioalveolar hyperplasia Glaser et
5 al. f 19901 at 90 days
6 These endpoints were preferred because they are the most direct and sensitive indicators of
7 cellular lung injury fNikula etal.. 2014: Henderson etal.. 19851.
This document is a draft for review purposes only and does not constitute Agency policy.
4-35 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
4.2.1.3. Other Effects
Inhalation data for effects outside the respiratory system are limited. The only animal
inhalation studies reporting effects outside the respiratory tract were rated low confidence for
these outcomes. No effects were observed in studies rating medium confidence for outcomes
outside the respiratory tract that were determined to be a hazard in Section 3.2, including Kim et al.
f20041 (liver weight and clinical chemistry) and Glaser etal. f!9851 (liver histopathology). As a
result, candidate values were not derived for effects outside of the respiratory tract
4.2.2. Methods of Analysis
4.2.2.1. Analysis of Animal Data
Animal data by Glaser etal. (19901 were used to derive candidate values for lower
respiratory tract effects. As noted earlier, the candidate endpoints were 1) BAL fluid measurements
of total protein, albumin, and LDH; and 2) Lung histopathological findings of histiocytosis and
bronchioalveolar hyperplasia.
Biologically based dose-response models are not available for respiratory effects of Cr(VI).
In this situation, EPA evaluates 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 the observed data. Consistent with this approach, EPA evaluated
dose-response information with the models available in EPA's Benchmark Dose Software (BMDS,
Version 3.2). However, data for lung histiocytosis, and for LDH, albumin, and total protein in BAL
fluid at the 90-day observation from the Glaser etal. fl9901 study in rats, were not amenable to
BMD modeling (see Appendix Section D.1.1.4 for details). As a result, no-observed-adverse-effect
level (NOAEL) and lowest-observed-adverse-effect level (LOAEL) approaches were used for these
effects. Because the lung histopathological changes and cellular responses in BAL fluid occurred
together at the lowest exposure level, and in combination with serum biomarkers indicating an
inflammatory response (Nikula etal.. 2014). all exposure levels were considered to have induced
adverse responses. Therefore, a LOAEL of 0.054 mg/m3 was chosen as the POD for these endpoints.
Data for bronchioalveolar hyperplasia, however, were amenable to BMD modeling. Because the
BMD Technical Guidance (U.S. EPA. 2012b) recommends a BMR based on one standard deviation
(SD) when no biological information is readily available that allows for determining a minimally
biological significant response, a BMR of 1 standard deviation change from the control mean was
applied.
Animal-to-human extrapolation
In accordance with EPA's Methods for Derivation of Inhalation Reference Concentrations and
Application of Inhalation Dosimetry (U.S. EPA. 1994). duration adjustments and dosimetric
adjustment factors (DAFs) were used for extrapolating the selected/candidate PODs from animals
to humans in order to calculate human equivalent concentrations (HECs). Because the RfC is
This document is a draft for review purposes only and does not constitute Agency policy.
4-36 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
intended to apply to continuous lifetime exposures for humans fU.S. EPA. 19941. a duration
adjustment was made to convert study-specific rodent bioassay exposure regimens to continuous
exposures. Next, a dosimetric adjustment factor was applied to account for differences in particle
lung dosimetry between species. Unlike for the RfD, extracellular reduction of Cr(VI) to Cr(III) was
assumed negligible for the inhalation route of exposure, and no additional dosimetric factors were
applied for pharmacokinetics.
The PODs identified from Glaser etal. (19901 were adjusted to account for discontinuous
daily exposure regimens as follows:
PODadj = POD x (hours exposed per day/24 hours) x (days exposed per week/7 days)
Where POD is the external exposure concentration rodent POD (mg/m3, determined by
dose-response modeling of rodent data or from the study NOAEL or LOAEL) and PODadj is the
duration-adjusted experimental exposure concentration (mg/m3).
Next, the PODhec (human equivalent concentration POD) was calculated from the PODadj by
multiplying by a DAF, which in this case was the regional deposited dose ratio (RDDRr) for
respiratory tract region r of interest as described in Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA. 19941.
PODhec = PODadj x RDDRr
The RDDRr can be calculated based on the physiology and respiratory parameters of
rodents and humans, and predicted fractional deposition in each respiratory tract region for each
species:
RDDR = x IMi x PUa
r [SAr)A [Ve)h (Fr)H
where:
SAr = surface area of respiratory tract region r (m2 or cm2)
Ve = ventilation rate (L/minute)
Fr = fractional deposition in respiratory tract region r
Since most effects in the BAL fluid may be indicative of effects due to deposition in the
entire lung (with the exception of the upper airways), the total of the pulmonary (PU) and
tracheobronchial (TB) surface areas and fractional depositions in these regions were used to
calculate an RDDRtb+pu :
nnnn _ (SATB+Pu)H .. (VeJa .. (FTB+Pu)A
KUUKjg+pu
[SAtb+pu)a CVe)h (7tb+pu)h
This document is a draft for review purposes only and does not constitute Agency policy.
4-37 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
The factor RDDRtb+pu was employed for all BAL fluid endpoints (except albumin) because
these effects were believed to be induced by exposure in the conducting airways and deep lung. For
albumin in BAL fluid, RDDRpu was applied, because this effect is believed to be induced by exposure
in the deep lung only.
Fractional depositions in the pulmonary region (Fpu) and tracheobronchial region (Ftb) for
both rats and humans were calculated using the Multi-Path Particle Dosimetry (MPPD) model
version 2. II72, a computational model that can be used for estimating airway particle deposition
and clearance (ARA (2009)).
For the model runs, the Yeh-Schum 5-lobe model was used for the human and the
asymmetric multiple path model was used for the rat (see Appendix D.1.2). Both models were run
under nasal breathing scenarios with the inhalability adjustment selected. The aerosol Cr(VI)
concentrations reported by Glaser etal. (1990) were converted to aerosol sodium dichromate
concentrations by molecular weight conversion (see Appendix D.1.2). It was determined that
aerosol concentration did not affect the predicted fractional lung depositions (human Fr values
were identical if aerosol concentration was set to either 1 or 136 mg/m3). Thus, the aerosol
concentration at the lowest Cr(VI) concentration was applied for rodent-human extrapolation
(reported concentration of 54 mg/m3 Cr(VI) is equivalent to 136 mg/m3 sodium dichromate
aerosol). Mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD)
reported by Glaser etal. (1990) varied slightly with concentration, however this had a negligible
effect on the RDDR (see Appendix Section D.1.2). For MPPD simulations, the particle MMAD ± GSD
(0.28 ± 1.63 |im), which reported for the lower Cr(VI) concentrations, was applied. The density of
sodium dichromate was input as 2.52 g/cm3.
The inhalation parameters used for the rat were: breathing frequency, 102 per minute
(default); tidal volume, 2.1 mL (default); Ve, 0.214 L/minute (calculated); functional residual
capacity, 4 mL (default); and upper respiratory tract volume, 0.42 mL (default). Adult human lung
physiology was: functional residual capacity, 3,300 mL (default); and upper respiratory tract
volume, 50 mL (default). Since human breathing frequency and tidal volume have a significant
impact on the estimated Fr, and these parameters are a strong function of human activity, multiple
different scenarios were simulated: resting, light work, heavy work, and maximal work. Values
defined by EPA's Exposure Factors Handbook fU.S. EPA. 2011al are contained in Appendix Table D-
17 and range from 40 breaths/min at a tidal volume of 3050 mL (maximal work) to 12 breaths/min
at a tidal volume of 500 mL (resting). All other parameters (rodent and human) were set to the
default MPPD software values (see Appendix D.1.2).
For the human, regional-specific surface areas for the respiratory tract (used as normalizing
factors) were 200 cm2 for extrathoracic (ET), 3200 cm2 for tracheobronchial (TB), and 54 m2 for
72EPA has since released newer version of the model. The differences in RDDR between MPPD v2.11
(released by Applied Research Associates) and draft MPPD vl.01 (released by EPA) are less than 10% (see
Appendix D.1.2).
This document is a draft for review purposes only and does not constitute Agency policy.
4-38 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 pulmonary (PU) fU.S. EPA. 19941. For the rat, respiratory tract surface areas were 15 cm2 for ET,
2 22.5 cm2 for TB, and 0.34 m2 for PU fU.S. EPA. 19941. The calculated RDDR values for TB/PU
3 regions ranged from 2.12/7.00 (resting scenario) to 0.12/0.47 (maximal work scenario). Since the
4 maximal and heavy work scenarios would not be representative of average daily lifetime inhalation
5 rates and volumes, the RDDR values were taken to be the average of the mean adult resting and
6 mean light work RDDRs. Values of RDDR were calculated as:
RDDRpu: 3.435
RDDRtb+pu : 2.685
7
8 Table 4-9 summarizes the sequence of calculations leading to the derivation of a
9 human-equivalent point of departure for each data set discussed above.
Table 4-9. Summary of derivation of points of departure following inhalation
exposure to Cr(VI). Data for male Wistar rats from Glaser etal. (1990)
Endpoint
% extra risk at
LOAEL
(mg/m3)a
BMC
(mg/m3)
POD
(mg/m3)
(LOAEL or
BMCL)
PODadj
(mg/m3)
RDDR
PODhec
(mg/m3)
Histopathology:
histiocytosis
87.5%
N/A
0.054
0.0495
2.685
(TB+PU)
0.133
Histopathology:
bronchioalveolar
hyperplasia
30%
BMCisd =
0.0294b
BMCLisd =
0.0168
0.0154
2.685
(TB+PU)
0.0413
Cell responses:
LDH in BALF
17%
N/A
0.054
0.0495
2.685
(TB+PU)
0.133
Cell responses:
Albumin in BALF
49%
N/A
0.054
0.0495
3.435 (PU)
0.170
Cell responses:
Total protein in
BALF
75%
N/A
0.054
0.0495
2.685
(TB+PU)
0.133
a%ER = (% incidence at LOAEL - % incidence at control)/(100 - % incidence at control) x 100
PODadj = (BMCL or NOAEL or LOAEL) x (22/24) x (7/7), since rodents in the Glaser et al. studies were unexposed for
2 hours each day.
PODhec is the human equivalent concentration POD based on the regional deposited dose ratio (RDDR) accounting
for interspecies differences in lung particle deposition. PU+TB: PODhec = (PODadj) x RDDRtb+pu = (PODadj) x 2.685.
PU: PODhec = (PODadj) x RDDRpu = (PODadj) x 3.435.
bLog-logistic model selected.
This document is a draft for review purposes only and does not constitute Agency policy.
4-39 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofHexavalent Chromium
4.2.2.2. Analysis of Human Data
Human data by Gibb etal. (2000a). Lindberg and Hedenstierna (19831. and Cohen etal.
(19741 were used to derive candidate values of upper respiratory tract effects. However, these
effects could not be modeled by Benchmark Dose Software (BMDS) models or other specialized
models. As noted in the analysis of nasal effects by OSHA f20061. the available human data were
insufficient to relate exposures and incidence. Studies either did not have the proper study design
for a quantitative analysis, or lacked short-term airborne Cr(VI) exposure data over an entire
employment period fOSHA. 20061. Because none of the available studies provided data for a no-
observed-adverse-effect-level (NOAEL), PODs were derived using lowest-observed-adverse-effect-
levels (LOAELs). How these uncertainties were accounted for in the quantitative derivation of the
candidate values are described later in this section.
The adjustment factors to account for differences between occupational exposures and
non-occupational exposure follow EPA guidelines fU.S. EPA. 20091 that acknowledges there are
differences in breathing rates between workers (10 m3 per 8-hour day) and non-workers
(20 m3 per 24-hour day) and that workers are exposed 240 days per year while non-workers are
exposed 365 days per year (U.S. EPA. 2016b. 2014e. 2012d. 2011dl. If workplace exposure is
assumed to occur 240 workdays/year:
LOAELHec=LOAEL ([ig/m3) x (VEho/VEh) x 240 days / 365 days
where:
LOAELhec = the LOAEL dosimetrically adjusted to an ambient human equivalent
concentration;
LOAEL = occupational exposure level (time-weighted average);
VEho = human occupational default minute volume (10 m3/8 h); and
VEh = human ambient default minute volume (20 m3/24 h).
Table 4-10. Summary of derivation of points of departure following human
inhalation exposure to Cr(VI)
Study
POD rationale
Notes and
conversions
LOAEL
(Hg/m3)
% incidence
at LOAEL
POD HEC
(Hg/m3)
Lindberg and
Hedenstierna
(1983)
Ulceration of the nasal septum.
The lowest concentration for the
2-20 ng Cr(VI)/m3 group. There is
high uncertainty in the exposure
concentrations.
Table 3
2
33%
0.66
Gibb et al.
Ulceration of the nasal septum.
The median exposure at first
diagnosed nasal ulceration.
Table 1
20 ng CrOs/m3 =
10.4 ng Cr(VI)/m3
10.4
63%
3.4
(2000a)
This document is a draft for review purposes only and does not constitute Agency policy.
4-40 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Study
POD rationale
Notes and
conversions
LOAEL
(Hg/m3)
% incidence
at LOAEL
POD HEC
(Hg/m3)
Gibb et al.
(2000a)
Ulceration of the nasal septum.
The mean exposure at first
diagnosed nasal ulceration
Table 1
54 ng CrOs/m3 = 28
Hg Cr(VI)/m3
28
63%
9.2
Cohen et al.
(1974)
(related study:
Cohen and
Ulceration of the nasal septum.
Mean air concentration for
exposed groups.
Table 6
0.0029 mg
Cr(VI)/m3
(2.9 ng Cr(VI)/m3)
2.9
32%
0.95
Kramkowski
(1973))
Exposure adjustment for all study concentrations to obtain POD HEC used the following occupational/non-
occupational factor: (10/20) x (240/365).
1
2 For ulceration of the nasal septum from Gibb etal. f2000al. the mean exposure
3 concentration was over 2x the median concentration, indicating that the data are skewed. Figure 1
4 in Gibb etal. f2000al indicates that certain job titles were exposed to higher Cr(VI) concentrations
5 early in the study period, and that these job titles experienced lower exposure for most of the later
6 years in the timeline. The median result was chosen instead of the mean for this dataset, because
7 the median is a better estimate of the central tendency for these data.
4.2.3. Derivation of Candidate Values
8 The reference concentration (RfC) is the inhalation concentration likely to be without an
9 appreciable risk of deleterious noncancer health effects during a lifetime fU.S. EPA. 19941.
10 Under EPA's A Review of the Reference Dose and Reference Concentration Processes [fU.S.
11 EPA. 20021: Section 4.4.5], five possible areas of uncertainty and variability were considered. An
12 explanation of the five possible areas of uncertainty and variability follows.
13 For animal-derived PODs using data for lower respiratory effects from Glaser etal. (19901:
14 • An intraspecies uncertainty factor, UFh, of 10 was applied to account for variability and
15 uncertainty in pharmacokinetic and pharmacodynamic susceptibility within the human
16 population. The PODs were derived from studies in inbred animal strains, and data were
17 only available for males. This is not considered sufficiently representative of the exposure
18 and dose-response of the most susceptible human subpopulations. In the case of inhaled
19 Cr(VI), insufficient information is available to quantitatively estimate variability in human
20 susceptibility; therefore, the value of 10 for the intraspecies UF was selected.
21 • An interspecies uncertainty factor, UFa, of 3 was applied to account for residual uncertainty
22 in the extrapolation from laboratory animals to humans in the absence of information to
23 characterize pharmacodynamic differences between rats and humans after inhalation
24 exposure to Cr(VI). This value is adopted when an adjustment from animal to a human
25 equivalent concentration has been performed as described in EPA's Methods for Derivation
26 of Inhalation Reference Concentrations and Application of Inhalation Dosimetry fU.S. EPA.
This document is a draft for review purposes only and does not constitute Agency policy.
4-41 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Toxicological Review ofHexavalent Chromium
1994). For these animal endpoints, an RDDR factor was used to estimate a human
equivalent concentration from animal data.
• A subchronic-to-chronic uncertainty factor, UFs, of 3 was incorporated to account for the
less-than-lifetime exposure in Glaser etal. f!9901 (which was a 90-day study). A value of 3
accounts for the potential that longer exposure may induce effects at a lower concentrations
(TJ.S. EPA. 20021.
• A LOAEL-to-NOAEL uncertainty factor, UFl, of 3 was applied to LOAELs that were based on
lung cellular and histopathological responses in BAL fluid. A value less than 10 was applied
because these responses were highly sensitive indicators of lung injury and inflammation
fNikula etal.. 2014: Henderson etal.. 19851. Additionally, effects began to resolve after a
short recovery time (see Figures 4-5 and 4-6). Considering these characteristics, the
changes were interpreted to approximate adverse responses, albeit with some residual
uncertainty, which do not support application of a UFl =10.
For human-derived PODs using occupational data for effects in the nasal cavity:
• An intraspecies uncertainty factor, UFh, of 3 to account for variation in susceptibility across
the human population and the potential that the available data may not be representative of
individuals who are most susceptible to the effect. The populations evaluated were mostly
adult male workers, which is not representative of individuals who may be most susceptible
to the effect. A value of UFh = 3 (as opposed UFh = 10) was applied because this is aportal-
of-entry effect of a direct-acting corrosive, and therefore the response by different
subpopulations from anatomic or pharmacokinetic/pharmacodynamic variability is
unlikely to differ (NRC. 2001).
• An interspecies uncertainty factor, UFa, of 1 was applied because results were derived from
studies in humans.
• A subchronic-to-chronic uncertainty factor, UFs, of 3 was applied. While data were not from
chronic lifetime exposures, the nasal effects were observed to have a short onset time (Gibb
etal. (2000a) estimated a median onset time of 22 days for ulcerated nasal septum, and
172 days for perforated nasal septum). Studies were generally consistent in showing that
these effects occur after 1-6 months of exposure. This may indicate that nasal effects occur
following short-term occupational exposures to high concentrations of Cr(VI), when
significant impaction of large particulates or mists containing Cr(VI) occurs along the nasal
passages. As noted in U.S. EPA (2020b). if a POD is based on subchronic evidence, the
assessment considers whether lifetime exposure could have effects at lower levels of
exposure. A factor of up to 10 is applied when using subchronic studies to make inferences
about lifetime exposure. However, a factor other than 10 may be used depending on the
magnitude and nature of the response and the shape of the dose-response curve fU.S. EPA.
2002.1998a. 1996a. 1994.1991). Based on the available evidence, it is considered less
likely that exposure to Cr(VI) outside of occupational settings (where particulates are
larger) would induce nasal perforations/ulcerations at much lower concentrations and
smaller particle sizes. (Note: the high response levels at the lowest concentration groups
were already accounted for in the LOAEL-to-NOAEL UF selection; the rate of the effect at
short onset time shows that there cannot be 1 Ox higher incidence due to prolonged
exposure). As a result, a factor of UFs < 10 was applied. Because it is possible that
This document is a draft for review purposes only and does not constitute Agency policy.
4-42 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 prolonged exposures to high concentrations may increase the severity of existing nasal
2 lesions after they occur, a value of UFs = 3 (as opposed to UFs = 1) was applied.
3 • A LOAEL-to-NOAEL uncertainty factor, UFl, of 10 was applied because this endpoint had a
4 high incidence at the lowest concentration across multiple studies. As a result, there was
5 higher uncertainty in the exposure-response relationship at lower concentrations.
6 For PODs derived using either animal (lower respiratory) or human (nasal effect) data:
7 • A database uncertainty factor, UFd, value of 3 was applied. A value of less than 10 was
8 applied because respiratory tract effects of inhaled Cr(VI) are considered portal-of-entry
9 effects, and are therefore likely to be amongst the most sensitive based on current
10 understanding of pharmacokinetics and mechanisms following inhalation. A value of
11 UFd = 3 (as opposed to UFd = 1) was applied because many of the inhalation studies were
12 low confidence (particularly for noncancer effects outside the portal of entry) and limited in
13 scope (working-age and mostly male humans, and only male rodents). Due to
14 pharmacokinetic differences from oral exposure (Cr(VI) is detoxified in the gut and liver on
15 first-pass), the stronger oral database could not be used to inform the UFd for inhalation
16 effects beyond the respiratory tract73.
17 Because of the non-uniform distribution of particulates in the lung, extracellular reduction
18 of Cr(VI) by epithelial lining fluid and BAL fluid was not modeled. Inhaled particles may accumulate
19 in susceptible areas such as airway bifurcation sites (Balashazv etal.. 2003: Schlesinger and
20 Lippmann. 1978). Localized dosimetry of inhaled particulates in susceptible regions could be
21 significantly higher than the average regional dosimetry estimated by MPPD (ARA (2009)). This
22 assessment assumes that the capacity to reduce Cr(VI) extracellularly in the lung fluid is exceeded
23 in both rodents and humans at all concentrations. Thus, uncertainty factor selections for potential
24 interspecies (UFa) or intraspecies (UFh) differences were not influenced by consideration of
25 differences in extracellular lung reduction at concentrations lower than those examined in the
26 available studies.
27 Table 4-11 is a continuation of Tables 4-9 and 4-10 and summarizes the application of UFs
28 to each POD to derive a candidate value for each data set. The candidate values presented in the
29 tables below are preliminary to the derivation of the organ/system-specific reference values. These
30 candidate values are considered individually in the selection of a representative inhalation
31 reference value for a specific hazard and subsequent overall RfC for Cr(VI).
73The database UF is intended to account for the potential for deriving an underprotective RfD/RfC as a result
of an incomplete characterization of the chemical's toxicity fU.S. EPA. 2002). While the database for
respiratory tract effects following inhalation is strong, toxicity information is lacking with respect to effects
outside the respiratory tract following inhalation.
This document is a draft for review purposes only and does not constitute Agency policy.
4-43 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 4-11. Effects in the lower respiratory tract and corresponding
derivation of candidate values for Cr(VI)
Endpoint
PODhec
(Hg/m3)
POD type
UFa
UFh
UFl
UFs
UFd
Composite
UF
Candidate
value
(Hg/m3)
Data for lower resoiratorv tract effects in male Wistar rats bv Glaser et al. (1990)
Histopathology:
histiocytosis
133
LOAEL
3
10
3
3
3
1000
0.13
Histopathology:
bronchioalveolar
hyperplasia
41.3
BMCLisd
3
10
1
3
3
300
0.14
Cell responses: LDH in
BALF
133
LOAEL
3
10
3
3
3
1000
0.13
Cell responses: Albumin
in BALF
170
LOAEL
3
10
3
3
3
1000
0.17
Cell responses: Total
protein in BALF
133
LOAEL
3
10
3
3
3
1000
0.13
Data for effects in the nasal cavity in humans
Ulceration of the nasal
septum (median) (Gibb
et al., 2000a)
3.4
LOAEL
1
3
10
3
3
300
1.1 x 10"2
Nasal mucosal
patholoev (Cohen et al.,
1974)
0.95
LOAEL
1
3
10
3
3
300
3.2 x 10"3
Ulceration of the nasal
septum (Lindberg and
Hedenstierna, 1983)
0.66
LOAEL
1
3
10
3
3
300
2.2 x 10"3
This document is a draft for review purposes only and does not constitute Agency policy.
4-44 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review ofHexavalent Chromium
Total protein in BALfluid (rats) (Glaser et al., 1990)
Albumin in BAL fluid (rats) (Glaser etal., 1990)
$ LDH in BAL fluid (rats) (Glaser et al.( 1990)
CD
5
o
Bronchioalveolar hyperplasia (rats) (Glaser et al., 1990)
Histiocytosis (rats) (Glaser et al., 1990)
>. Nasal mucosa pathology (humans) (Cohen et al., 1974)
o
03
'5.
o3 Ulceration (humans) (Lindberg & Hedenstierna, 1983)
aj
a.
Q_
Ulceration (humans) (Gibb et al., 2000)
Composite UF
A Candidate value
• POD (HEC)
0.0001 0.001 0.01 0.1 1 10 100
log-scale Cr(VI) concentration (ng/m3)
1000
Figure 4-7. Candidate values with corresponding POD and composite UF.
4.2.4. Derivation of Organ/System-Specific Reference Concentrations
Selection of organ/system-specific toxicity values can be based on the most sensitive
outcome, a clustering of values, or a combination. Each candidate value was evaluated with respect
to multiple considerations, including strength of evidence, basis of the POD (i.e., BMC vs. NOAEL vs.
LOAEL), and dose-response model uncertainties. A confidence level of high, medium, or low was
assigned to each osRfC based on the study(ies) used to derive the candidate value, and the
reliability of the associated POD and candidate value calculation(s). Confidence in the POD and
candidate value calculation(s) included considerations of the quality and variability of the exposure
assessment in an epidemiology study or the exposure protocols in an animal study. Moreover,
higher confidence was placed in the osRfC when the POD was identified close to the range of the
observed data and the magnitude of exposure was relevant to those experienced in the general U.S.
population.
4.2.4.1. Lower Respiratory Toxicity
Cr(VI)-induced cytotoxicity has been observed in epithelial tissues following both inhalation
and oral exposures (see Sections 3.2.1 and 3.2.2). Inhaled Cr(VI) in particles, dust, or mists of
respirable size may be absorbed into epithelial cells in the lung and lung airways. The osRfC fol-
lower respiratory tract effects was derived from data in Glaser etal. fl990I Endpoints included
lung cellular responses (LDH, albumin, and total protein in BAL fluid), and changes in lung
histopathology (histiocytosis and bronchioalveolar hyperplasia). Because most of these endpoints
had the same LOAEL and uncertainty factors, they produced essentially the same candidate value
This document is a draft for review purposes only and does not constitute Agency policy.
4-45 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
(note: albumin in BAL fluid differed from the others slightly due to selection of a different RDDR
extrapolation region for animal-to-human extrapolation). BMD modeling was performed on data of
bronchioalveolar hyperplasia, and the resulting candidate value (which used a lower uncertainty
factor) supported the NOAEL-derived candidate values. The osRfC for lower respiratory system
effects was taken as the value of the candidates for cellular responses (total protein and LDH in BAL
fluid) and histopathology findings (histiocytosis and bronchioalveolar hyperplasia) resulting in an
osRfC of 0.1 |J.g/m3 (rounded from 0.13 [ig/m3), or 1 x 10~4 mg/m3.
The relatively small number of medium confidence studies evaluating noncancer lower
respiratory effects decreases the confidence of this osRfC. In addition, the endpoint was derived
from subchronic rodent data. Human data for noncancer lower respiratory tract effects of Cr(VI)
are scarce because studies published prior to the availability of standardized spirometry guidelines
from the American Thoracic Society (first developed in 1979) (ATS/ERS. 2019) were considered
uninformative for pulmonary function. A factor that increased confidence was the clear dose-
response observed for multiple lower respiratory endpoints in rodents.
4.2.4.2. Upper Respiratory Toxicity
As noted earlier, Cr(VI) is cytotoxic and there is high confidence that Cr(VI) induces effects
at the portals of entry. Furthermore, effects in the nasal cavity of humans are well documented by
occupational studies (OSHA. 2006). The osRfC for effects in the upper respiratory tract were based
on ulcerated nasal septum observed by the Gibb etal. (2000a) occupational study. While the study
reported multiple other nasal endpoints (irritated, perforated, and bleeding nasal septum),
ulcerated nasal septum was chosen because of its severity and high incidence (63% of the cohort
having the clinical finding). Gibb etal. (2000a) had higher sample sizes and better exposure data
than the alternative studies by Cohen etal. (1974). and Lindberg and Hedenstierna (1983).
The Baltimore plant studied by Gibb etal. (2000a) had a rigorous personal and air
monitoring system that spanned a period of decades (see Table 4-23 in Section 4.4.5). This greatly
increased confidence in the reported air concentrations and worker exposures. While the Lindberg
and Hedenstierna T19831 used both area and personal air samplers, the recorded data only
spanned 13 days. Furthermore, the defined concentration ranges (<2-20 [ig Cr(VI)/m3) by
Lindberg and Hedenstierna (1983) only constituted average workday concentrations (peak values
were noted to be higher, but only limited concentration data are presented). Characterization of
the nasal endpoints by Cohen etal. (1974) were highly detailed, and the study employed only air
measurements consistent with current NIOSH recommendations (Andrews and O'Connor. 2020:
NIOSH. 20 1 3 ) 74. However, the sample size was small, and the breathing-zone air samples
represented only a snapshot in time (and not the long-term exposure of the workers over time).
The osRfC for upper respiratory tract effects is based on the LOAEL for ulcerated nasal septum in
74This manual chapter fAndrews and O'Connor. 20201 was published by the National Institute for
Occupational Safety and Health (NIOSH). Author names listed for this citation are the NIOSH editors.
This document is a draft for review purposes only and does not constitute Agency policy.
4-46 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofHexavalent Chromium
humans reported by Gibb etal. f2000al resulting in an osRfC of 1 x 10~2 |J.g/m3 (rounded from
1.1 x 10"2 [ig/m3), or 1 x 10"5 mg/m3. Because only LOAELs could be obtained from the datasets,
and because the estimated effect incidences were high at the LOAEL (63%), there is uncertainty in
the dose-response relationship at lower concentrations. For the Gibb etal. (2000a) study, effects in
the nasal cavity were observed after a few months of exposure (median time on the job of 86-418
days), and it is unknown how the effect severity may increase over a lifetime of exposure. These
factors decrease confidence in the osRfC for upper respiratory tract effects. Additional
uncertainties relevant to upper respiratory tract effects are described in detail in Section 4.2.6.
Factors that increase confidence in the osRfC for upper respiratory tract effects include the
consistency at which this effect was observed (generally between 2-20 [ig Cr(VI)/m3 with early
onset time), and the thorough air sampling programs implemented for the Baltimore Cohort (see
Table 4-23) (Gibb etal.. 2000a).
Table 4-12. Organ/system-specific reference concentrations (RfCs) and
overall RfC for Cr(VI)
Effect
Basis
osRfC
mg/m3
Exposure
description
Confidence
Lower respiratory
Increase in total protein and LDH in
BAL fluid, and histiocytosis and
bronchioalveolar hyperplasia in male
rats Glaser et al. (1990)
1 X 10"4
90-day rat study
Medium
Upper respiratory
Ulcerated nasal septum of humans
Gibb et al. (2000a)
1 x 10"5
Occupational
exposure
Medium
Overall RfC
Ulcerated nasal septum
1 x 10"5
Occupational
exposure
Medium
As noted in Section 4.2.8, the prior IRIS assessment developed separate RfCs for "chromic
acid mists and dissolved hexavalent chromium aerosols," and for "hexavalent chromium dusts."
The RfC for chromic acid mists was based on human occupational exposure to chromic acid
(FhCrCU) at a chrome-plating facility by Lindberg and Hedenstierna f!9831. while the RfC for dusts
was based on data for rodent exposure to sodium dichromate (Na2Cr20y) aerosols by Glaser et al.
(1990: 1985). The current database now includes noncancer data from the Baltimore chromate
production plant (Gibb etal.. 2000a). which studied effects in humans occupationally exposed to a
variety of chromium species in dust form, including sodium chromates (Na2CrC>4) and dichromates
(Na2Cr2C>7) (Haves etal.. 1979). Human nasal effects were observed by both the Gibb etal. (2000a)
study (chromium dusts) and the Lindberg and Hedenstierna (1983) study (chromic acid mists).
Lindberg and Hedenstierna T19831 observed that ulceration of the nasal septum occurred only in
the highest peak exposure group (20-48 [ig Cr(VI)/m3) and the highest daily exposure group (>2-
20 [ig Cr(VI)/m3). This is supportive of Gibb etal. (2000a). which reported ulceration of the nasal
septum at a median concentration of 10 [ig Cr(VI)/m3, and a mean concentration of 28 [ig
This document is a draft for review purposes only and does not constitute Agency policy.
4-47 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
Cr(VI)/m3. Therefore, the RfC for upper respiratory tract effects is applicable to both forms of
Cr(VI) (mists and dusts). EPA also considers the RfC for lower respiratory tract effects applicable to
both forms of Cr(VI).
The previous distinction in RfCs drawn between mists and dusts is no longer supported.
However, distinctions are presented via the osRfCs (upper vs. lower respiratory tracts), and these
are a function of particle or droplet size. It is generally known that large inhaled particles (with
diameter >5 |im) will deposit in the extrathoracic region, particles greater than 2.5 |im are generally
deposited in the tracheobronchial regions, and particles less than 2.5 |im are generally deposited in
the pulmonary region fOSHA. 20061. The rodent study of Na2Cr2C>7 aerosols by Glaser et al. (1990j
1985) likely induced effects in the lower respiratory tract due to the small particle sizes achieved
by the experiment (MMAD < 0.4 |im). For the human occupational studies, particle and droplet
sizes may have been larger, causing a larger proportion of Cr(VI) to impact in the nasal airways.
4.2.5. Selection of the Overall Reference Concentration
An overall RfC of 1 x 10"5 mg/m3 was selected. The overall RfC was based on effects in the
upper respiratory tract (ulceration of the nasal septum), because of the two endpoints
representative of respiratory tract effects it is the more sensitive effect and will be protective of
noncancer lower respiratory tract effects and systemic effects. Additional considerations of
uncertainty associated with this RfC are noted here and below in section 4.2.6. It was derived using
a LOAEL, where the incidence of the effect was high and the time of onset relatively short. The
occupational cohort fGibb etal.. 2000al consisted of a population of mostly adult males and may
not have included sensitive individuals. It is uncertain if or how the endpoint severity may be
affected by lifetime chronic exposures.
4.2.6. Uncertainties in the Derivation of Reference Concentration
4.2.6.1. Onset Time for Nasal Effects
The time between first exposure and development of nasal effects varies depending on the
severity of the effect, but nasal effects generally occur within 1 year of initial exposure for more
severe effects, and 1-3 months for less severe effects. Gibb etal. (2000a). the only prospective
study of the development of nasal effects reported the time to event in days (mean [median]) for
irritation (89 [20]), ulceration (86 [22]), perforation of the septum (313 [172]), and bleeding nasal
septum (418 [92]) fGibb etal.. 2000al. Cross-sectional studies reported a similar time to event
periods based on self-reported interview data fLindberg and Hedenstierna. 1983: Cohen etal..
1974). Cohen etal. (1974) reported that severity of pathology increased with longer exposure
times and prevalence of ulceration or perforation in the study population was higher at 94% in
workers who had worked at the plant for more than 1 year at the time data were collected
compared to 57% among workers who had worked for less than a year at the same plant More
recently, Singhaletal. (2015) showed that severity of nasal outcomes increased with years of
This document is a draft for review purposes only and does not constitute Agency policy.
4-48 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
exposure in both chromate manufacturing and chrome electroplating workers. The early
onset-time, combined with the fact that incidences were high at the lowest concentration (the
lowest concentration in this occupational setting is still high relative to environmental levels) leads
to uncertainty in the extrapolation from occupational exposure to continuous lifetime exposure.
4.2.6.2. Hand-to-Nose Transfer
Only one of the candidate value studies reported hand-to-nose transfer of Cr(VI) originating
from surface touching fCohen etal.. 19741. Surface contamination of Cr(VI) throughout workplace
environments (including on gloves and other personal protective equipment), and detection of
Cr(VI) on the hands of employees have been documented (Ceballos etal.. 2017: Lucas and
Kramkowski. 1975: Cohen and Kramkowski. 1973). However, no quantitative data were available
to adjust for this potential route of exposure.
4.2.6.3. Susceptible Populations
Quantitative analysis of effects in the lower respiratory tract were based on animal data,
while analysis of effects in the upper respiratory tract were based on occupational studies of adult
humans. Data for these effects were not available in susceptible populations, such as children or
those with preexisting respiratory conditions.
4.2.7. Confidence Statement
An overall confidence level of High, Medium, or Low was assigned to reflect the level of
confidence in the study(ies) and hazard(s) used to derive the RfC, the overall database, and the RfC
itself, as described in Section 4.3.9.2 of EPA's Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA. 1994).
The confidence in the overall chronic RfC is medium. The RfC for upper respiratory tract
effects is based on the LOAEL for ulcerated nasal septum in humans reported by Gibb etal. (2000a).
resulting in an RfC of 1 x 10"5 mg/m3. While there is high confidence that inhaled Cr(VI) can induce
effects in the nasal cavity of humans, quantitative characterization of these endpoints have
uncertainties. The available studies did not have enough exposure groups or individual-level data
adequate for a dose-response analysis, and only LOAELs could be obtained from all the available
datasets. For the Gibb etal. (2000a) study, effects in the nasal cavity were observed after a few
months of exposure (median time on the job of 86-418 days), and it is unknown how the effect
severity may increase over a lifetime of exposure. Because the estimated effect incidences were
high at the LOAEL (63%), there is uncertainty in the dose-response relationship at lower
concentrations. As a result, the confidence in the RfC for upper respiratory effects is medium.
4.2.8. Previous IRIS Assessment: Inhalation Reference Concentration
The previous IRIS assessment contained two RfCs for Cr(VI). An RfC for "chromic acid mists
and dissolved hexavalent chromium aerosols" and an RfC for "hexavalent chromium dusts" were
This document is a draft for review purposes only and does not constitute Agency policy.
4-49 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
posted on the IRIS database in 1998. As noted in Section 4.2.4, health effects induced by inhalation
exposure to Cr(VI) are expected to differ due to particle size distribution. These differences are
now reflected in the derivation of osRfCs, which are strongly dependent on particle sizes rather
than other chemical properties. Larger particles are more likely to affect the nasal airways, while
smaller particles can affect the lower airways.
The 1998 RfC for Cr(VI) acid mists and dissolved aerosols was based on the human study by
Lindberg and Hedenstierna (19831. A LOAEL for nasal septum atrophy of 2 |ig/m:i was identified
based on the lower bound of the 2-20 |J.g/m3 range, and this value was adjusted using a continuous
exposure adjustment factor, and an adjustment factor for occupational and 24-hour average
breathing rates. This resulted in a LOAEL for continuous exposure of 0.714 |J.g/m3. A total
uncertainty factor of 90 was applied: 3-fold for extrapolation from a subchronic to a chronic
exposure, 3-fold for extrapolation from a LOAEL to a NOAEL, and 10-fold for interhuman variation.
This resulted in an RfC of 0.008 |J.g/m3 (8 x 10~6 mg/m3) for hexavalent chromic acid mists and
dissolved hexavalent chromium aerosols. The current assessment derived a different LOAEL for
the Lindberg and Hedenstierna f!9831 study, because most cases (7/8) of nasal ulceration in the 2-
20 |J.g/m3 group had peak exposure levels at or above 20 |J.g/m3.
The previous RfC for Cr(VI) dusts was based on the studies by Glaser et al. (1990; 19851 and
used the modeling and data analysis of this dataset published by Malsch etal. (19941. Malsch et al.
(19941 developed BMCs for lung weight, lactate dehydrogenase (LDH) in BAL fluid, protein in BAL
fluid, albumin in BAL fluid, and spleen weight. The Malsch etal. T19941 analysis defined the
benchmark concentration as the 95% lower confidence limit on the dose corresponding to a 10%
relative change in the endpoint compared to the control. A continuous exposure adjustment factor
was applied, and the maximum likelihood model was used to fit continuous data to a polynomial
mean response regression, yielding maximum likelihood estimates of 36-78 |J.g/m3 and BMCs of
16-67 |J.g/m3. LDH was the most sensitive endpoint (BMC of 16 ng/m3) and was the basis of the
1998 IRIS assessment RfC for Cr(VI) dusts. An RDDR of 2.1576, derived by methods outlined in U.S.
EPA (19941. was applied to this value to extrapolate a human equivalent concentration. A total
uncertainty factor of 300 was applied: 10-fold for the less-than-lifetime exposure, 10-fold for
variation in the human population, and 3-fold to account for pharmacodynamic differences not
accounted for by the RDDR. This resulted in an RfC of 1 x 10~4 mg/m3 for hexavalent chromium
dusts, which is the same as the value derived in this assessment for lower respiratory tract effects
(using the same study and similar methods).
4.3. ORAL SLOPE FACTOR FOR CANCER
The oral slope factor (OSF) is a plausible upper bound on the estimate of risk per
mg/kg-day of oral exposure. The OSF can be multiplied by an estimate of lifetime exposure (in
mg/kg-day) to estimate the lifetime cancer risk. EPA determined under the 2005 Guidelines for
This document is a draft for review purposes only and does not constitute Agency policy.
4-50 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
Carcinogen Risk Assessment, Cr(VI) is "likely to be carcinogenic to humans" via the oral route of
exposure.
4.3.1. Analysis of Carcinogenicity Data
The animal database for cancer consisted of a chronic 2-year drinking water bioassay which
found "clear evidence of carcinogenic activity" of Cr(VI) in male and female rats and mice fNTP.
20081. These results were based on increased incidences of squamous cell neoplasms in the oral
cavity of rats, and increased incidences of neoplasms in the small intestine of mice. The data from
NTP (20081 indicate a dose-response relationship in both species.
Human dose-response data for cancer via the oral route were not suitable for dose-
response analysis. The lack of individual estimates of exposure, the uncertain nature of the
mortality data, and the potential impact of confounding made it difficult to draw conclusions (see
Section 3.2.3). Human cancer data via the inhalation route of exposure were not used for oral slope
factor derivation because route-to-route extrapolations were not considered in this assessment
(see Protocol, Appendix A).
4.3.2. Dose-Response Analysis—Adjustments and Extrapolations Methods
A benchmark dose (BMD) approach was used to model the dose-response data. This
method is described in detail in Section 4.1.2. Because a mutagenic mode-of-action for Cr(VI)
carcinogenicity via the oral route of exposure (see Section 3.2.3) is "sufficiently supported in
(laboratory) animals" and "relevant to humans," EPA used a linear low dose extrapolation from the
POD in accordance with Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005a). The
multistage model was selected for dose-response analysis because it is consistent with low dose
linearity, it is sufficiently flexible for most cancer bioassay data, and its use provides consistency
across cancer dose-response analyses (Gehlhaus etal.. 2011). Graphical results are provided in
Figure 4-8 below. Further details, including the modeling outputs, can be found in U.S. EPA
(2021a).
This document is a draft for review purposes only and does not constitute Agency policy.
4-51 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicologiccil Review of Hexavalent Chromium
4 6
Dose (mg/kg-d)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Male rat oral tumors
Estimated Probability
Response at BMD
• Linear Extrapolation
Data
BMD
BMDL
2 4 6
Dose (mg/kg-d)
4 6
Dose (mg/kg-d)
4 6
Dose (mg/kg-d)
Figure 4-8. BMDS 3.2 graphical output of selected models for dose-response of
cancer data in male and female rats and mice from NTP (2008).
For tumors of the small intestine of mice, a PBPK model was used to extrapolate the rodent
dose-response model results to a human equivalent dose, using the same methodology applied for
noncancer effects (Section 4.1.2). The internal dose used for mouse-to-human extrapolation was
the BW3/4-adjusted Cr(VI) dose that is estimated to escape gastric reduction. The mean result from
Monte Carlo analysis was used as the POD for the 0SF, as opposed to the lower 1% value (which
was used for the POD of the RfD). This is because intraspecies variability in pharmacokinetics and
pharmacodynamics is not incorporated into cancer risk assessment flJ.S. EPA. 2006cl. with the
exception for early-life considerations noted in the Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens fU.S. EPA. 2005bl Uncertainty factors are not
applied during rodent-to-human extrapolation of cancer dose-response data. For comparative
purposes, the BW3/4 scaling approach without Cr(VI) gastric reduction modeling or Monte Carlo
analysis is also presented. This can be interpreted as the result for a susceptible subpopulation
having high gastric pH (>4.0) and Cr(VI) gastric juice reduction capacity equivalent to rodents (see
Appendix C.1.5).
For tumors in the oral cavity of rats, there is uncertainty regarding the appropriate internal
dose metric. Mice did not exhibit tumors of the oral cavity, but in a separate bioassay were
This document is a draft for review purposes only and does not constitute Agency policy,
4-52' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review ofHexavalent Chromium
observed to have higher oral tissue chromium levels than rats following 90-day drinking water
exposure fKirman etal.. 20121. Mice rarely exhibit oral tumors from NTP bioassays, even for
chemicals inducing oral tumors in rats (Ibrahim etal.. 2021: NTP. 2008175. Thus, mice may be less
susceptible to tumors of the oral cavity due to factors that cannot be accounted for using PBPK
modeling. There were no observed nonneoplastic lesions in the oral mucosa of rats or mice
following either the chronic or subchronic high dose NTP Cr(VI) drinking water bioassays (Witt et
al.. 20131. Unlike for the mouse, where tumors were observed in GI organs posterior to the
stomach (where most Cr(VI) reduction occurs), tumors of the rat oral cavity occur in tissues where
Cr(VI) exposure is not mitigated by extracellular reduction in the stomach. As a result, species
differences in Cr(VI) reduction in the stomach are not relevant for the dose-response analysis of rat
oral tumors. Site-specific PBPK models of Cr(VI) kinetics in the oral cavity epithelium are not
available. In the absence of an adequately developed theory or information to develop and
characterize an oral portal-of-entry dosimetric adjustment factor, application of BW3/4 scaling is
recommended (U.S. EPA. 2011c. 2005a).
4.3.3. Derivation of the Oral Slope Factor
The lifetime oral cancer slope factor for humans is defined as the slope of the line from the
lower 95% bound on the exposure at the POD to the control response (slope factor = 0.1/BMDLio).
This slope, a 95% upper confidence limit represents a plausible upper bound on the true risk.
Using linear extrapolation from the BMDLio, human equivalent oral slope factors were derived for
each sex/tumor site combination. Results for all tumor types are listed in Table 4-13.
Table 4-13. Summary of the oral slope factor derivations
Species/
sex
Model
BMR
BMD
mg/kg-da
BMDL
mg/kg-da
Extrapolation
Method
Internal
rodent
dose
mg/kg-db
Internal
dose POD
mg/kg-dc
PODhed
mg/kg-dd
OSF
Per mg/kg-
d
Adenomas or Carcinomas in the mouse small intestine (NTP, 2008)
Mice (M)
1° MS
10
1.44
1.05
PK
0.173
0.0274
0.319
0.313
BW3/4
N/A
N/A
0.166
0.602
Mice (F)
1° MS
10
1.34
1.03
PK
0.169
0.0267
0.316
0.317
BW3/4
N/A
N/A
0.163
0.613
750f the 24 test articles associated with site-specific neoplasia that produced positive, clear or some evidence
of carcinogenicity in the oral cavity (NTP. 20201. only one (1,2,3-trichloropropane) induced tumors of the oral
cavity in mice. All other test articles induced tumors in the oral cavity of male or female rats. With the
exception of Cr(VI), three chemicals were found to induce both oral and small intestinal tumors
(2,2-bis(Bromomethyl)-l,3-propanediol, C.I. Direct blue 15, C.I. Acid red 114), although they only induced
these effects in rats. In general, tumors of the small intestine are more rare in rats (compared to mice), and
tumors of the oral cavity are more rare in mice (compared to rats) (see Appendix D.2).
This document is a draft for review purposes only and does not constitute Agency policy.
4-53 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Toxicological Review ofHexavalent Chromium
Species/
sex
Model
BMR
BMD
mg/kg-da
BMDL
mg/kg-da
Extrapolation
Method
Internal
rodent
dose
mg/kg-db
Internal
dose POD
mg/kg-dc
PODhed
mg/kg-dd
OSF
Per mg/kg-
d
Squamous cell carcinoma or squamous cell papilloma in oral mucosa or tongue (NTP, 2008)
Rats (M)
1° MS
10
6.04
3.37
BW3/4
N/A
N/A
0.923
0.108
Rats (F)
1° MS
10
4.25
2.70
BW3/4
N/A
N/A
0.645
0.155
aUnits of administered mg/kg-d Cr(VI) dose.
bDose escaping stomach reduction in rodent (mg/kg-d) estimated by PK modeling.
CBW3/4 scaling adjustment of the internal rodent dose (dose escaping reduction multiplied by (BWA/BWH)1/4,
where BWH = 80 kg and BWA is set to study-specific time-weighted average (TWA) values (these same study-
specific BW values were also used in the PK modeling). TWA BWA = 0.450 kg for male rats, and TWA BWA = 0.260
kg for female rats at the 2-year time period in NTP (2008). TWA BWA = 0.05 kg for male and female mice at the 2-
year time period in NTP (2008).
dPODHED in units of mg/kg-d Cr(VI) oral dose ingested by humans. For the PK method, this is the mean value of
20000 Monte Carlo PK simulations needed to achieve the internal dose POD (see Appendix C.1.5 for details). For
the standard BW3/4 method, no additional adjustments beyond BW3/4 scaling of the rodent dose are applied.
The OSF for Cr(VI) was derived from small intestine tumors in male and female mice using
PBPK modeling, 0.3 (mg/kg-d)-1.
For BW3/4 scaling adjustment and PBPK modeling applied above, the mean body weight
recommended by EPA's Exposure Factors Handbook fU.S. EPA. 2011al (80 kg) was used. There is a
negligible difference in the PODs when using 70 kg fU.S. EPA. 19881 or 80 kg, and the final OSF
would be the same under either assumption.
4.3.4. Application of Age-Dependent Adjustment Factors
Because a mutagenic mode-of-action for Cr(VI) carcinogenicity is sufficiently supported in
laboratory animals and is relevant to humans (see Section 3.2.3), and in the absence of chemical-
specific data to evaluate differences in age-specific susceptibility, increased early-life susceptibility
to Cr(VI) is assumed and ADAFs should be applied, as appropriate, in accordance with the
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens fU.S. EPA.
2005b). The oral slope factor of 0.3 (mg/kg-day)-1, calculated from data applicable to adult
exposures, does not reflect presumed early-life susceptibility to this chemical. Example calculations
for estimating cancer risks based on age at exposure are provided in Section 6 of the Supplemental
Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA. 2005b).
The Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to
Carcinogens establishes ADAFs for three specific age groups. The current ADAFs and their
corresponding age groups are 10 for exposed individuals <2 years old, 3 for exposed individuals 2
to <16 years old, and 1 for exposed individuals >16 years old (U.S. EPA. 2005bl The 10- and 3-fold
adjustments to the slope factor are to be combined with age-specific exposure estimates when
estimating cancer risks from early-life (<16 years of age) exposures to Cr(VI).
This document is a draft for review purposes only and does not constitute Agency policy.
4-54 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Toxicological Review ofHexavalent Chromium
To illustrate the use of the ADAFs established in the Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens fU.S. EPA. 2005bl. OSF calculations are
presented for three exposure duration scenarios, including full lifetime. For oral exposures
assuming Cr(VI) exposure-response equivalence across age groups (i.e., equivalent risk from
equivalent exposure levels, independent of body size), the ADAF calculation is fairly
straightforward. The partial and lifetime risks (per mg/kg-d) are presented below in Table 4-14.
Table 4-14. Application of ADAFs for 70-year exposure to Cr(VI) from ages 0 to
70
Age group
ADAF
Slope factor
(per mg/kg-d)
Duration adjustment
Partial risk
(per mg/kg-d)
0-<2 yrs
10
0.3
2 yrs/70 yrs
0.0857
2-<16 yrs
3
0.3
14 yrs/70 yrs
0.180
>16 yrs
1
0.3
54 yrs/70 yrs
0.231
Total risk
0.497
Note that the partial risk for each age group is the product of the values in columns 2-4
(e.g., 10 x 0.3 x 2/70 = 0.0857 for exposures from age 0 to <2 years), and the total risk is the sum of
the partial risks. Thus, a lifetime estimate for the OSF for exposure starting at birth is 0.5 (per
mg/kg-d).
If calculating the cancer risk for a 30-year exposure to a constant average daily dose of
0.0001 mg Cr(VI)/kg-day from ages 0 to 30 years, the duration adjustments would be 2/70,14/70,
and 14/70, and the partial risks would be (10 x 0.3 x 0.0001 x 2/70= 8.6 x 10 6), (3 x 0.3 x 0.0001 x
14/70=1.8 x 10"5.), and (1 x 0.3 x 0.0001 x 14/70 = 6 x 10 6), resulting in a total risk estimate of 3.3
x 10-5.
If calculating the cancer risk for a 30-year exposure to a constant average daily dose of
0.0001 mg Cr(VI)/kg-day from ages 20 to 50 years, the duration adjustments would be 0/70, 0/70,
and 30/70, and the partial risks would be 0, 0, and (1 x 0.3 x 0.0001 x 30/70 = 1.3 x 10 5), resulting
in a total risk estimate of 1.3 x 10 5.
4.3.5. Uncertainties in the Derivation of the Oral Slope Factor
Because the studies and pharmacokinetics methods used to derive the OSF are the same as
those used to derive the RfD, the major uncertainties related to OSF derivation are outlined in
Section 3.3 and Section 4.1.6. Additional information on susceptible populations is provided in
Section 3.3.1. Briefly,
• Uncertainties persists in the PBPK models of the human and mouse stomach. Population
variability in kinetic parameters is unknown, and it is likely that gastric contents and
microbiota contribute to inter individual variation.
This document is a draft for review purposes only and does not constitute Agency policy.
4-55 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofHexavalent Chromium
• Uncertainty in the choice of the tumor type and internal dose metric for cross-species
extrapolation.
• Cr(VI) detoxification in the stomach for populations with elevated stomach pH (consumers
of medicine to treat acid reflux, hypochlorhydria individuals) may differ from standard
health individuals.
• There may be higher susceptibility for carriers of mutated cystic fibrosis transmembrane
conductance regulator (CFTR) gene (see Sections 3.2.3.4 and 3.3.1).
Individuals taking medication to treat gastroesophageal reflux disease (GERD), including
calcium carbonate-based acid reducers and proton pump inhibitors, have an elevated stomach pH
during treatment This is known to be a significant fraction of the population since up to 20% of the
population is afflicted by GERD, and the gastric pH for these individuals may be above 4 throughout
the day during successful treatment fDelshad etal.. 2020: GBP 2017. 2020: Lin and
Triadafilopoulos. 2015: Burdsall etal.. 2013: Atanassoff et al.. 19951. A sensitivity analysis was
performed on the human model (Appendix C.1.5), assuming a baseline stomach pH = 4 (as opposed
to 1.3). It was found that for internal doses near those of the cancer PODs for mice, the mean76
human equivalent dose for a population with baseline gastric pH = 4 would be approximately %
that of the standard population with baseline pH = 1.3. As a result, the OSF for this population
would be 2x more stringent Similarly, the OSF estimated by default approaches (BW3/4 scaling and
no adjustment for gastric reduction) would be health-protective for this population, since that
method implicitly assumes that humans and rodents have the same gastric pH (>4) and reduction
capacity. After rounding, the adult-based OSF for BW3/4scaling (0.6 per mg/kg-d) is exactly 2x the
adult-based OSF estimated by PBPK modeling (0.3 per mg/kg-d). Under the BW3/4scaling
assumption, the lifetime ADAF-adjusted value would also be exactly 2x more stringent (1.2 per
mg/kg-d). The infant and neonatal gastric environments and the lack of data on Cr(VI) reduction
during early life stages are also significant uncertainties, and are not fully addressed by the ADAF or
the adult-based sensitivity analyses.
Table 4-15 provides an overview summarizing the uncertainties and their impact on the
OSF.
76Additional characteristics of the probability distributions can be found in Appendix C.1.5 and Appendix
D.6.2. According to U.S. EPA Guidelines for Carcinogen Risk Assessment fU.S. EPA. 2005a): "Slope factors
generally represent an upper bound on the average risk in a population or the risk for a randomly selected
individual but not the risk for a highly susceptible individual or group. Some individuals face a higher risk and
some face a lower risk." As a result, mean PBPK results are presented in the quantitative cancer assessment.
This document is a draft for review purposes only and does not constitute Agency policy.
4-56 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Table 4-15. Summary of uncertainties in the derivation of oral slope factor
values for Cr(VI)
Consideration
Impact on unit
risk
Decision
Justification
Target organ
4, OSF, 3-fold, if
oral tumors
selected
Small intestine
tumors
(adenomas or
carcinomas of
the duodenum or
jejunum or ileum
in mice)
Tumor site is concordant across rats and mice in the Gl
tract as a whole (small intestine and mouth), increasing
support for its relevance to humans.
As there are no data to support any one result as most
relevant for extrapolating to humans, the most sensitive
result for Gl tract tumors was used to derive the oral slope
factor.
Data set
None
NTP (2008)
NTP (2008) is a hiph confidence study and the only to
evaluate potential carcinogenicity in multiple organs and
multiple species following chronic drinking water
exposure.
Cross-species
scaling dose
metric
Alternatives could
4/ or T* slope
factor
mg/kg-d Cr(VI)
emptied from
stomach,
adjusted by
BW3/4 scaling
The amount of Cr(VI) available for absorption into the
small intestine is a function of how much Cr(VI) will escape
the stomach unreduced.
Applving the pyloric flux dose metric defined in Thompson
et al. (2014) (dailv mg Cr(VI) emptied from stomach, per L
small intestine) would slightly decrease the OSF (BW3/4
scaling is similar as scaling by small intestine volume).
Applying BW3/4scaling without taking into account
interspecies differences in gastric reduction would
increase the OSF by 2x.
Low dose
extrapolation
4/ cancer risk
estimate would be
expected with the
application of
nonlinear low
dose extrapolation
Linear
extrapolation
from POD (based
on mutagenic
MOA)
Available MOA data support linearity (mutagenicity is a
primary MOA of Cr(VI)). See Appendix D.3 for an
uncertainty analysis of the low dose extrapolation method
Statistical
uncertainty at
POD
4, OSF 1.4-fold if
BMD used as the
POD rather than
BMDL
BMDL (preferred
approach for
calculating
plausible upper-
bound slope
factor)
Limited size of bioassay results in sampling variability;
lower bound is 95% confidence interval on administered
exposure at 10% extra risk of alimentary tract tumors.
Dose-response
modeling
Alternatives could
4/ or T* slope
factor
Multistage-
model
No biologically based models for Cr(VI) were available.
Multistage models are sufficiently flexible for most cancer
bioassay data, and their use provides consistency across
cancer dose-response analyses. See Appendix Section
C.1.5 for additional details on the impact of alternative
dose metrics.
Sensitive
subpopulations
-t OSF to
unknown extent
ADAFs are
recommended
for early-life
exposures
No chemical-specific data are available to determine the
range of human pharmacodynamic variability or
sensitivity. Deriving an OSF from populations with high
baseline gastric pH would lead to a significantly higher OSF
(over 2x higher).
This document is a draft for review purposes only and does not constitute Agency policy.
4-57 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
4.3.6. Previous IRIS Assessment: Oral Slope Factor
The previous IRIS assessment for Cr(VI) was posted to the IRIS database in 1998. In that
assessment, EPA concluded that the oral carcinogenicity of Cr(VI) could not be determined (and
was thus classified as Group D under the 1986 classification guidelines). At the time, only one study
in humans suggested an association with stomach cancer, but other human and animal studies did
not report similar effects. Therefore, no oral slope factor was derived.
4.4. INHALATION UNIT RISK FOR CANCER
The inhalation unit risk (IUR) is a plausible upper bound on the estimate of risk per |J.g/m3
air breathed. The IUR can be multiplied by an estimate of lifetime exposure (in |ig/m:i] to estimate
cancer risks over a lifetime or partial lifetime.
In 1998, the EPA IRIS Toxicological Review ofHexavalent Chromium classified Cr(VI) as a
"known human carcinogen by the inhalation route of exposure" based on consistent evidence that
inhaled Cr(VI) causes lung cancer in humans and supporting evidence of carcinogenicity in animals
fU.S. EPA. 1998cl. The same conclusion has since been reached by other authoritative federal and
state health agencies and international organizations and the carcinogenicity of Cr(VI) is
considered to be well-established for inhalation exposures (TCEO. 2014: IPCS. 2013: NIOSH. 2013:
IARC. 2012: CalEPA. 2011: NTP. 2011: OSHA. 20061.
4.4.1. Analysis of Carcinogenicity Data
This section focuses on identifying additional appropriate studies to update the quantitative
exposure-response analysis and the derivation of the IUR. More recent epidemiologic studies have
been identified in the peer-reviewed literature which include higher quality exposure data, longer
follow-up times, larger sample sizes, and more sophisticated analyses than were available in 1998.
While the focus of the updated cancer analysis was evaluation of new information and other studies
that were not evaluated in the 1998 IRIS assessment, EPA did not exclude studies published prior to
1998. Having judged the evidence of hazard for carcinogenicity of inhaled Cr(VI) to be
well-established, EPA focused on studies that could inform estimation of the exposure-response
function which could be used to derive an IUR.
4.4.1.1. Identification of Studies for the Derivation of a Cr(VI) Inhalation Unit Risk
Study selection
A title and abstract screening of human health studies obtained from the literature searches
described in Sections 1.2 and 2.1, and backwards searching using reference lists of screened
studies, identified 64 human lung and respiratory cancer references. These studies then underwent
full-text screening for exposure-response data that may be informative for derivation of a revised
inhalation unit risk. Studies needed to be epidemiological analyses examining quantitative
measures of chromium exposure in relation to lung cancer incidence or mortality risk. Studies
This document is a draft for review purposes only and does not constitute Agency policy.
4-58 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Toxicological Review ofHexavalent Chromium
were excluded if Cr(VI) measurements in air, or convertible equivalents such as CrC>3, were not
presented, or if group-level exposure assignments were based on job title (and not chromium
measurements) (see Table D-26 in Appendix D.4). Applying these criteria, there were 22 lung
cancer references identified as potentially informative for exposure-response analysis.
All 22 studies were based on occupational cohorts, and many followed the same worksites
or worker populations over time. For cohorts with multiple follow-up studies, EPA included only
the most recent follow-up, and used the prior studies to obtain information relevant to analysis of
data and study evaluation (see Table D-27 in Appendix D.4). Of the 22 studies, five independent
cohort studies evaluating Cr(VI) exposure and the risk of lung cancer were obtained after
restricting to the most recent cohort follow-up data (Figure 4-9 andTable 4-16). These were: (1) a
chromate facility in Baltimore, MD (Gibb etal.. 2020: Gibb etal.. 2015: Gibb etal.. 2000b): (2) a
chromate facility in Painesville, OH (Proctor etal.. 2016): (3) two chromate facilities in Germany
(Leverkusen and Uerdingen) fBirk etal.. 20061: (4) the IARC multicenter cohort of welders in the
European Union (Gerin etal.. 1993): and (5) two chromate facilities in the United States (Corpus
Christi TX and Castle Hayne NC) fLuippold etal.. 20051. A sixth study fAEI. 20021 did not include
new data, but was a pooled analysis of the four plants evaluated in Birk etal. f20061 and Luippold
etal. ("20051.
This document is a draft for review purposes only and does not constitute Agency policy.
4-59 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Human lung cancer studies based on title/abstract screen
and backwards bibliography searching
[64 references]
I
Full-text screening for studies with exposure-response data
I
Lung cancer studies containingexposure-response data
[22 references]
Screen for more recent analyses of cohort populations
5 cohorts
[8 references]
3ibb et al (2020, 2015, 2000): Baltimore cohort
Proctor et al. (2016): Painesville cohort
Birk et al. (2006): 2 German cohorts
Luippold et al. (2005): 2 US cohorts
AEI (2002): Pooled cohort of 4 populations
discussed in Birk/Luippold
Gerin et al. (1993): European welder cohort
Screen for preferred
quantitative methods
4 cohorts
Gibb et al (2020, 2015, 2000)
Proctor et al. (2016)
Birk et al. (2006)
Gerin et al. (1993)
Figure 4-9. Literature screening results for studies containing exposure-
response data of Cr(VI) and lung cancer.
1 The next step was to evaluate the quantitative methods used in each of the analyses. It was
2 preferred that exposure-response analyses were conducted using estimated airborne
3 concentrations of speciated Cr(VI) compounds from which a slope77 and its standard error could be
4 obtained. Studies were available that presented results from models using a continuous measure of
77The beta coefficient describing the function of exposure-response relationship between exposure to Cr(VI]
in air, on a continuous scale, and the risk of lung cancer.
This document is a draft for review purposes only and does not constitute Agency policy.
4-60 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofHexavalent Chromium
exposure, so the two that did not (e.g., studies that only presented an overall SMR) were excluded:
Luippold etal. f20051 and AEI f2002I An overview of all studies excluded for exposure-response
analysis of lung cancer in humans is provided in Appendix D.4 Tables D-26 through D-28. The
remaining four studies were then evaluated for risk of bias and sensitivity. Study evaluation
included consideration of exposure assessment, outcome ascertainment, population selection,
confounding, selective reporting sensitivity, and data analysis [see Protocol Section 6.2 (Appendix
A) for more details]. Considerable focus was placed on factors that could notably affect the
magnitude and direction of the effect estimates, including potential for exposure measurement
error, confounding, missing data, and the specific statistical analyses conducted. Summaries of the
study evaluations are presented in Table 4-16 along with the overall confidence rating. Details of
those evaluations are presented in HAWC ( lick here).
Table 4-16. Summary of included studies considered for the derivation of an
inhalation unit risk for Cr(VI) and overall confidence classification. Click to see
interactive data graphic for rating rationales.
Study evaluation
Reference
Study description
Exposure
Outcome
Selection
Confounding
Analysis
Sensitivity
Sel. reporting
Overall confidence
T3
a>
Gibb et al.,
(2020: 2015:
2000b)a
Occupational cohort (n = 2,354 male
workers) in the U.S. exposed 1950-
1985 and followed until 2011.
G
G
A
A
G
A
A
High
_3
U
Proctor et al.
Occupational cohort (n = 714 male
A
A
G
D
G
A
A
Medium
Ł
(2016)
workers) in the U.S. exposed 1940-
1972 and followed until 2011.
Birketal. (2006)
Occupational cohort (n = 901 male
D
A
A
A
A
D
A
Low
workers) in Germany exposed 1958-
1998 and followed until 1998.
Gerin et al.
Pooled IARC multicenter occupational
D
A
A
D
A
A
A
Low
(1993)
cohorts (n = 11,092 male welders)
across 135 companies in 9 EU countries
exposed during various periods 1946-
1986.
aThree studies were used to represent the Baltimore, MD cohort, as they had essentially the same worker
population.
Three studies describing one cohort were classified as high confidence: Gibb etal., (2020:
2015: 2000b) (the Baltimore MD cohort); and one was classified as medium confidence: Proctor et
al. f20161 (the Painesville OH cohort). The remaining studies were low confidence. The high and
medium confidence studies were advanced for further consideration in the derivation of the IUR for
This document is a draft for review purposes only and does not constitute Agency policy.
4-61 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
Cr(VI). Overviews of the two cohorts and their analyses are provided below followed by an analysis
of the preferred characteristics for candidate principal studies for IUR development from
occupational cohorts are described in Table 4-17.
Overview of the Baltimore. MP cohort
Chromate production at the Baltimore, MD site began in 1845 and ultimately ceased in 1985
fGibb etal.. 2000b: Hayes etal.. 19791. The original Baltimore cohort included workers who were
newly employed between 1945 and 1974 fHaves etal.. 19791. The current cohort was defined by
Gibb etal. (2000b) and excluded most workers who began work before August 1,1950. This cutoff
date coincided with when a new chromite ore mill and roasting plant were constructed, exposure
mitigation measures were implemented, and extensive exposure information collection began
(Gibb etal.. 2015: Hayes etal.. 19791. The vital status of 2357 workers were initially followed up
through death or the end of 1992 fGibb etal.. 2000b) and then extended through 2011 for 2354
workers fGibb etal.. 20151 for a total of 91,186 person-years at risk. The mean duration of
employment for the 2011 update of the cohort was 3.1 years and the mean number of years of
follow-up was 38.9 years. The median duration of employment for the cohort was 0.4 years and the
median number of years of follow-up was 39.9 years.
Gibb etal. (2000b) estimated Cr(VI) exposures for each person in each year based on job
titles, the time spent in each sampling zone and exposure estimates based on ~70,000
contemporary measurements of Cr(VI) concentration in air during the study period. Samples
included short-term air sampling in the workers' breathing zones from 1950-1961 followed by
24-hour routine measurements taken by 20 air samplers rotated through 154 fixed sites
throughout the facility, and personal air sampling beginning in 1977. Exposure estimates were
merged with work history data to estimate each workers' cumulative exposures during
employment All air measurements of Cr(VI) were converted to units of mg CrOs/m3 as a common
basis in Gibb etal. (2000b) because the prevailing regulatory standard was from the metric used by
the U.S. Occupational Safety and Health Administration in its past Permissible Exposure Limits for
chromic acid and chromates. The mean cumulative exposure78 to Cr03 reported in Gibb et al.
f20151 Table 2 was 0.14 mg/m3-years which converts to 72.8 [ig/m3-years of Cr(VI).79 The 25th,
50th, and 75th percentiles were 0.52, 5.2, and 41.6 [ig/m3-years of Cr(VI). Company medical
records provided smoking status at the beginning of employment for 91% of the cohort
(Yes/No/Unknown); 74% smoked cigarettes, 16% did not smoke, and smoking status was
unknown for 9%. No information on pack-years of smoking or how smoking status may have
changed over time was available.
78Here the cumulative exposure is unlagged and untransformed.
79Conversion of mass of Cr03 to mass of Cr(VI) is based on the contribution of the molecular weight (MW) of
Cr to MW of Cr03. Since the MW of Cr is 51.996 g/mol and the MW of Cr03 is 99.99 g/mol, the conversion
factor is 51.996/99.99 = 0.52. Units are further converted to |ig/m3 from mg/m3 by multiplying by
1000 ng/mg.
This document is a draft for review purposes only and does not constitute Agency policy.
4-62 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofHexavalent Chromium
Gibb etal. f20151 reported 217 deaths from lung cancer in this cohort compared to 133
expected deaths based on Maryland vital statistics for a SMR of 1.63 (95% CI: 1.42-1.86). The risk
of lung cancer mortality was analyzed using a Cox proportional hazards model with age as the time
variable and cumulative exposure as a time-varying covariate. In a model adjusted for smoking and
age80, each unit increase in logio cumulative Cr(VI) exposure, lagged by 5 years, was associated with
a 1.255-fold (p < 0.001) increase in the hazard ratio.
Gibb etal. (2020) re-analyzed this cohort with the same exposure and outcome data using a
Cox proportional hazards model adjusted for smoking and age, but without log-transforming
cumulative Cr(VI) exposure. In this analysis, untransformed cumulative Cr(VI) exposure, lagged by
5 years, was associated with a 1.64-fold (95% CI: 1.30, 2.04) increase in the hazard ratio. Gibb et al.
(2020) also reported analyses of the untransformed cumulative Cr(VI) exposure using a conditional
Poisson regression approach (Richardson and Langholz. 2012) to estimate the relative risk per unit
of cumulative exposure (controlling for age and smoking) showing that cumulative Cr(VI) exposure,
lagged by 15 years, was associated with a 1.82-fold (95% CI: 1.35, 2.45) increase in the hazard ratio.
Overview of the Painesville. OH cohort
The Painesville, OH chromate production plant was in operation from 1931-1972, with
major renovations occurring in 1949-1950 and 1962-1964 to mitigate exposure and modernize
plant operations (Proctor et al.. 2004). Previous analyses of the Painesville plant relied on indirect
measures of Cr(VI) in air, using measures of air total chromium and soluble/insoluble chromium
dust measurements, and only studied workers employed prior to 1940 fMancuso. 1997.19751. The
current cohort was defined by Proctor etal. f 20161 to include workers employed after December
31,1939. The vital statistics of 714 workers were followed up through death or the end of 2011 for
a total of 24,535 person-years at risk. The mean duration of employment for the cohort was not
explicitly reported but falls within the interval of five to nine years (see Table 1 in Proctor et al.
(2016) and the mean number of years of follow-up was 34.4 years.
The Proctor etal. (2016: 2004) studies obtained 800 measurements of airborne Cr(VI) from
23 historical industrial hygiene surveys for workers employed from 1940-1972. Using historical
records of worker job histories over time and industrial hygiene data (which included Cr(VI)
measurements), a job-exposure matrix (JEM) was constructed (Proctor etal.. 2004). Usable data
were available for 1943,1945,1948,1957, and 1959-1971 (excluding 1962). Exposure estimates
were merged with work history data to estimate each workers' cumulative exposures during
employment All Cr(VI) cumulative exposure estimates were reported in mg/m3-years. The mean
cumulative exposure to Cr(VI) was 1.1 mg/m3-years (Proctor etal.. 2016) which converts to
1.1 x 103 [ig/m3-years with a range of 0.2 x 103 [ig/m3-years to 22.1 x 103 [ig/m3-years. Employee
records provided smoking status for 29% of the cohort (Yes/No/Unknown); of those, 22% smoked
80In this Cox proportional hazards regression, the time scale used was age and this controls for age in the
model.
This document is a draft for review purposes only and does not constitute Agency policy.
4-63 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 cigarettes, 7% did not smoke, and smoking status was unknown for 72%. No information on pack-
2 years of smoking or how smoking status may have changed over time was available.
3 Proctor etal. (20161 reported 77 deaths from lung cancer in this cohort which yielded a
4 SMR of 1.86 (95% CI: 1.45-2.28) compared to lung cancer mortality in Ohio and a SMR of 2.05
5 (95% CI: 1.59-2.50) compared to the U.S. population. Proctor etal. (2016) fit several models
6 within the cohort and concluded that the linear Cox model with age as the time variable and
7 controlling for smoking and age at hire had the best fit and reported a hazard ratio of 1.19 per
8 mg/m3-years increase in Cr(VI) exposure based on a regression coefficient of 0.17 per mg/m3-years
9 (95% CI: 1.11-1.27; p = 0.0006).
Table 4-17. Details of rationale for selecting a principal study on Cr(VI) for IUR
derivation
Attribute
Preferred characteristics for candidate principal
studies for the Cr(VI) IUR
Baltimore, MD Cohort
Painesville, OH Cohort
Study design
Sufficient follow-up time for outcomes to develop
Total person-time at
Total person-time at
characteristics
(this can depend on the health outcome being
risk:
risk:
addressed).
91,186 person-years
24,535 person-years
Study size and participation rates that are
Size of cohort: 2354
Size of cohort: 714
adequate to detect and quantify health outcomes
workers
workers
being studied (without influential biases in study
population selection) are preferred.
Mean follow-up time:
Mean follow-up time:
38.7 years
34.4 years
Use of a study design or analytic approach that
adequately addresses the relevant sources of
Confounding potential:
Confounding potential:
potential confounding, including age, sex, and
Controlled for age and
Controlled for age and
exposures to other risk factors for the outcome of
smoking; no
smoking; six
interest.
mesothelioma deaths
mesothelioma deaths
Effect modification
Effect modification
potential: No known
potential: asbestos
asbestos exposure and
exposure is strongly
no mesothelioma
indicated with six
deaths.
mesothelioma deaths.
This document is a draft for review purposes only and does not constitute Agency policy.
4-64 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
Attribute
Preferred characteristics for candidate principal
studies for the Cr(VI) IUR
Baltimore, MD Cohort
Painesville, OH Cohort
Relevance of
exposure
paradigm
Studies of chronic duration are preferred over
studies of shorter exposure duration because
they are most relevant to environmental
exposure scenarios (potentially including both
continuous exposure from ambient conditions
and episodic activity-related exposures).
Chronic duration
Chronic duration
When available studies observe effects across
different ranges of exposures, studies that include
relatively low exposure intensities that may
represent conditions more similar to
environmental exposures are preferred as there
may be less uncertainty in extrapolation of those
results to lower exposure levels.
Mean exposure Cr(VI):
72.8 ng/m3-years.
The 25th, 50th, and
75th% were 0.52, 5.2,
and 41.6 ng/m3-years
Mean exposure Cr(VI):
1.1 x 103 ng/m3-years.
Range from 0.2 ng/m3-
years to 22.1 x 103
Hg/m3-years
Measurement
of exposure
Emphasis is placed on the specificity of exposure
assessment in time and place with a preference
for greater detail where possible. Exposure
measurements that are site and task specific
provide generally preferred exposure
information. Where available, individual-level
measurements are generally preferred.
Measurement techniques that are more specific
to the agent of concern are preferred over less
specific analytical methods. Better
characterization of airborne concentrations is
preferred.
Stronger studies will often be based upon
knowledge of individual work histories (job
titles/tasks with consideration of changes over
time); however, appropriate group-based
exposure estimates may also be relevant.
Exposure reconstruction and estimating
exposures based on air sampling from other time
periods and/or operations are less preferred
methods of exposure estimation.
~70,000 measurements
during 1950-1974.
Early samples were
short-term air samples
in the workers'
breathing zones, later
24-hours samples from
154 fixed sites, and
full-shift personal air
sampling began in
1977.
Sampling records for 9
years could not be
located (1950-56,
1960-61) and those
values were imputed
based on existing data
to model those job-
specific exposure
values.
Individual work
histories matched to
job-specific exposure
estimates based on
sampling
measurements.
800 measurements
during 1940-1972.
No personal samples.
Uncertainty in short-
term workers'
exposures: Proctor et
al. (2004) "companv
records lacked
sufficient information
on these individuals to
reconstruct their work
histories."
Individual work
histories matched to
job-specific exposure
estimates based on
sampling
measurements.
Measurement
of covariates
Studies that considered the potential effects of
confounding by relevant covariates are preferred
over those without such consideration—unless
confounding is not a major concern.
Age is well measured.
Smoking status was
identified for 93% of
the cohort.
Age is well measured.
Smoking status was
identified for 28% of
the cohort.
This document is a draft for review purposes only and does not constitute Agency policy.
4-65 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofHexavalent Chromium
Attribute
Preferred characteristics for candidate principal
studies for the Cr(VI) IUR
Baltimore, MD Cohort
Painesville, OH Cohort
Measurement
of effect(s)
Cancer incidence data are generally preferred
over cancer mortalitv data (U.S. EPA, 2005a). In
the absence of cancer incidence data, cancer
mortality data are appropriate with preference
for cause of death classified using international
classification disease (ICD) codes at time of death.
Lung cancer data were
obtained from death
certificates.
217 lung cancer cases.
No deaths from
mesothelioma and no
evidence of outcome
misclassifi cation.
Lung cancer data were
obtained from death
certificates.
77 lung cancer cases.
3 deaths from
mesothelioma (of 6
total) were initially
classified as lung
cancer deaths
Analysis
methodology
Studies conducting and reporting regression
results of within cohort comparisons and those
with p and SE(P) are preferred over standardized
mortality ratio (SMR) results. Occasionally
studies reporting standardized rate ratio (SRR) or
SMR results with sufficient specificity by exposure
category may allow for post hoc estimation of p
and SE(P)—although if the lowest exposure
category is defined by the lowest
quantile/category of exposure, such estimates
may be biased towards the null.
Analyses included
multiple model forms
(types of regression)
with multiple
parameterizations of
covariates and lags for
exposure.
Analyses included
multiple model forms
(types of regression)
with multiple
parameterizations of
covariates and lags for
exposure.
Table 4-17 summarizes key considerations related to study attributes that were considered
in the rationale for identifying the principal cohort. The Baltimore, MD cohort was (1) larger than
the Painesville cohort, (2) had longer follow-up time, (3) had more deaths from lung cancer, (4) had
no deaths from mesothelioma, despite having 66,651 additional years of person-time at risk than in
the Painesville cohort, suggesting lower potential for confounding by asbestos exposure, (5) had
more than an order of magnitude lower average exposures which can be more relevant to
estimating effects at lower exposures and requires less extrapolation, (6) had more air samples to
estimate exposures, and (7) had more complete data on smoking. EPA selected the Baltimore, MD
cohort as the basis for deriving the IUR.
4.4.2. Dose-Response Analysis—Adjustments and Extrapolations Methods
The first step towards deriving an inhalation unit risk for lung cancer was to identify
candidate effect estimates (i.e., beta coefficients from the regression analyses) from studies of the
principal cohort. Once the lung cancer effect estimates have been obtained, they are adjusted for
differences in air volumes between workers and other populations due to exposure frequency and
breathing rates. Conversions between occupational Cr(VI) exposures and continuous
environmental exposures were made to account for differences in the number of days exposed per
year, and in the amount of air inhaled per day. Those adjusted values can be applied to the U.S.
population as a whole in EPA life-table analyses. These life-table analyses allow for the estimation
of an exposure concentration associated with a specific extra risk of cancer incidence caused by
This document is a draft for review purposes only and does not constitute Agency policy.
4-66 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
inhalation of Cr(VI); the specific extra risk is called the benchmark response (BMR) and a value of
1% is standard for cancer outcomes in people. Those exposure concentrations serve as points of
departure (POD) from which IURs can be extrapolated. Non-occupational exposure adjustment and
methods applied for the life-table analysis are described in detail in Section 4.4.3.
Because a mutagenic mode of action for Cr(VI) carcinogenicity (see Section 3.2.3) is
"sufficiently supported in (laboratory) animals" and "relevant to humans," EPA used a linear low
dose extrapolation from the POD in accordance with Guidelines for Carcinogen Risk Assessment
flJ.S. EPA.2005al
4.4.2.1. Cancer Risk Models for Cr(VI) Inhalation Exposures
A cancer risk model predicts the probability of cancer in an individual with a specified
history of exposure to a cancer-causing agent. In the case of inhalation exposure to Cr(VI), the lung
cancer effects are of chief concern, and workers' individual cumulative exposure to Cr(VI) are used
to predict cancer risk. Different types of regression analyses were used to model the lung cancer
effect of Cr(VI) in the Baltimore, MD cohort The model forms are described below.
The Cox proportional hazards model (Cox. 1972) is one of the most commonly used
statistical models for the epidemiologic analysis of survival and mortality in cohort studies with
extensive follow-up, including studies of the Baltimore, MD cohort (Gibb etal.. 2020: Gibb etal..
2015: Gibb etal.. 2000b). The Cox proportional hazards model assumes that a function of
covariates (e.g., exposures) result in hazard functions that are a constant proportion of the baseline
hazard function in unexposed individuals over some timescale, typically calendar time or age
(e.g., the background age-specific rates of lung cancer in the population). One of the strengths of
this model is that knowledge of the baseline hazard function is not necessary, and no particular
shape is assumed for the baseline hazard; rather, it is estimated nonparametrically.
Another methodology used to analyze the Baltimore, MD cohort (Gibb etal.. 2020) was the
conditional Poisson regression approach proposed by Richardson and Langholz (R&L) to estimate
the relative risk per unit of cumulative exposure (Richardson and Langholz. 2012). The R&L
approach maximizes a conditional likelihood expression that allows for covariates like age and
smoking to be included in the model, but avoids estimation of all the stratum-specific parameters
by treating them as nuisance terms. This property is made possible by separating and then
cancelling the nuisance terms in the likelihood function. Thus, the R&L approach models the effects
of age and smoking when estimating the effect of Cr(VI), but does not yield the specific effect
estimates for age and smoking.
4.4.2.2. Cancer Risk Parameters
The Cox regression results from the Baltimore, MD cohort are shown in Table 4-18.
This document is a draft for review purposes only and does not constitute Agency policy.
4-67 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofHexavalent Chromium
Table 4-18. Results of Cox proportional hazards modeling of cumulative
chromium exposure (mg Cr03/m3-years) by different lag periods (age and
smoking are included in model). Table adapted from Table 1 of Gibb et al.
f20201.
Lag period (y)
P per mg
Cr03/m3-year
SE(P)
Hazard ratio
Exp(p)
95% CI (P)
-2 log(L)
0
0.4712
0.1133
1.60
1.28-2.00
2830.23
5
0.4868
0.1145
1.63
1.30-2.04
2829.80
10
0.4939
0.1197
1.64
1.30-2.07
2830.52
15
0.4812
0.1333
1.62
1.25-2.10
2833.03
Note: 1 mg CrC>3 = 0.520 mg Cr(VI); CrOs/m3-year = (CrOs/m3)(year).
The measure of fit (-2 Log(L)) of the Cox proportional hazards models of the lung cancer
risk adjusted for age and smoking were very similar for all lag periods, although the fit for the
5-year lag was slightly better than for the other lags—although not statistically better. The
rationale for the lag period is that there is often a latency period for cancer beginning with the
initial incidence of cancer and extending to the time of cancer mortality. In this conceptual model,
the exposures that are experienced by the individual after cancer has begun are no longer expected
to cause lung cancer, and thus those exposures may not be etiologically relevant. Here the results
show little difference in effect size across the different lag times. This is likely due to the fact that
exposures ceased in 1982 and follow-up continued until 2011 so there was little difference in
lagged and unlagged exposures. Section 4.4.5 provides a sensitivity analysis across the different lag
lengths.
The lung cancer effect estimate for the 5-year lag in Table 4-18 above (Gibb etal.. 20201 is
in units of per mg Cr03/m3-year and was converted to unit of per [ig Cr(VI)/m3-year as follows:
1 mg Cr03/m3-year • [0.52 mg Cr(VI)/mg CrOs] • [1000 [ig/mg] = 520 [ig Cr(VI)/m3-year
5-year lag (3 cr(vq = 0.4868 per mg Cr03/m3-year = 0.4868/(1 mg Cr03/m3-year)
= 0.4868/(520 [ig Cr(VI)/m3-year)
= 9.362 x 10_4per [ig Cr(VI)/m3-year
The inhalation unit risk is derived from the one-sided 95th% upper bound of p. Gibb et al.
(20201 reported a two-sided 95% confidence interval as is the standard practice in the
epidemiologic literature (i.e., from the 2.5th% to the 97.5th% bounds). EPA estimated the
one-sided 95th% upper bound (UB) of (3 by assuming the distribution of (3 was normally distributed
(which is appropriate for the Cox Proportional Hazards model) as follows:
One-sided 95th% UB of (3 = (3 + 1.645(se((3))
= 0.4868 per mg Cr03/m3-year +1.645 • (0.1145 per mg Cr03/m3-year)
This document is a draft for review purposes only and does not constitute Agency policy.
4-68 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 = 0.6752 per mg Cr03/m3-year
2 = (0.6752 per mg Cr03/m3-year) / (520 [ig Cr(VI)/m3-year)
3 = 1.298 x 10"3 per [ig Cr(VI)/m3
4
5 This one-sided 95th% upper bound of (3 from the Cox Proportional Hazards analysis in Gibb
6 etal. (20201 will be used to derive an estimate of the IUR using a life-table analysis.
7 R&L regression results from the Baltimore, MD cohort are shown in Table 4-19.
Table 4-19. Results for relative exponential exposure-response (R&L) model
adjusted for age and smoking. Table adapted from Table 2 of Gibb et al.
f20201.
# Age groups3
Lag period (y)
P
SE(P)
RR = exp(P)
95% CI(P)
-2 log(L)
1
0
0.454
0.098
1.57
1.30-1.91
9283.51
5
0.454
0.098
1.57
1.30-1.91
9283.62
10
0.451
0.101
1.55
1.29-1.91
9286.50
15
0.414
0.108
1.51
1.22-1.87
9291.89
2
0
0.454
0.098
1.57
1.30-1.91
9283.50
5
0.461
0.098
1.59
1.31-1.92
9282.79
10
0.463
0.100
1.59
1.31-1.93
9284.08
15
0.474
0.107
1.60
1.30-1.98
9286.46
3
0
0.915
0.047
2.50
2.28-2.74
8854.75
5
0.933
0.048
2.59
2.31-2.79
8846.57
10
0.982
0.050
2.67
2.42-2.94
8845.78
15
1.088
0.056
2.97
2.66-3.31
8848.71
4
0
0.506
0.133
1.66
1.28-2.15
4327.08
5
0.522
0.133
1.69
1.30-2.19
4326.07
10
0.548
0.139
1.73
1.32-2.27
4325.97
15
0.599
0.152
1.82
1.35-2.45
4325.95
5
0
1.179
0.036
3.25
3.03-3.49
8153.85
5
1.246
0.036
3.48
3.24-3.73
8091.17
10
1.387
0.040
4.00
3.70-4.33
8035.39
15
1.559
0.044
4.75
4.36-5.18
8030.41
6
0
1.142
0.036
3.13
2.92-3.36
8253.33
5
1.164
0.036
3.20
2.98-3.44
8235.51
10
1.200
0.038
3.39
3.08-3.58
8238.56
15
1.375
0.043
3.95
3.64-4.30
8223.38
Note: 1 mg CrC>3 = 0.520 mg Cr(VI).
aOne age group (all ages, 15-96); two age groups (>15 to 65 and >65); three age groups (ages >15 to 60, >60 to
>70); four age groups (>15 to 60, >60 to 65, >65 to 75, and >75); five age groups (ages >15 to 60, >60 to 65, >65 to
70, >70 to 75, and >75);six age groups (ages >15 to 55, >55 to 60, >60 to 65,>65 to 70, >70 to 75, and >75).
8 The R&L analysis based on four age groups fit the Baltimore, MD cohort better than the
9 analyses based on other numbers of age groups as evidenced by the lower fit statistics, and within
10 the 4-age group analysis, the fits were very similar for all lag periods, although the fit for the
This document is a draft for review purposes only and does not constitute Agency policy.
4-69 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofHexavalent Chromium
15-year lag was slightly better than for the other lags—although not statistically better. Section
4.4.5 provides a sensitivity analysis across the different lag lengths.
The lung cancer effect estimate for the 15-year lag in Table 2 from Gibb etal. (20201 is
0.599 per mg Cr03/m3-year and was converted to unit of per [ig Cr(VI)/m3-year as follows:
1 mg Cr03/m3-year • [0.52 mg Cr(VI)/mg CrOs] • [1000 [ig/mg] = 520 [ig Cr(VI)/m3-year
5-year lag (3 crtvi) = 0.599 per mg Cr03/m3-year = 0.599/(1 mg Cr03/m3-year)
= 0.599/(520 [ig Cr(VI)/m3-year)
= 1.152 x 10"3 per [ig Cr(VI)/m3-year
One-sided 95th% UB of (3 = (3 + 1.645(se((3))
= 0.599 per mg Cr03/m3-year +1.645 • (0.152 per mg Cr03/m3-year)
= 0.849 per mg Cr03/m3-year
= (0.849 per mg Cr03/m3-year) / (520 [ig Cr(VI)/m3)
= 1.633 x 10"3 per [ig Cr(VI)/m3
This one-sided 95th% upper bound of (3 from the R&L analysis in Gibb etal. (20201 will be
used to derive an estimate of the IUR using a life-table analysis.
4.4.3. Inhalation Unit Risk Derivation
4.4.3.1. Life-Table Analysis to Derive an IUR
The (3 coefficients (slopes) for lung cancer risks attributable to cumulative exposures to
Cr(VI) from the Gibb etal. f20201 are used in life-table analyses to predict the risk of cancer as a
result of the exposure over a lifetime. The life-table analysis divides a lifetime into small
age-specific intervals and sums the risks of lung cancer incidence in each age group in the presence
and absence of Cr(VI) exposure. This is done to assess the age-specific risk of lung cancer incidence
while accounting for competing causes of death. The lung cancer risk in a particular year of life is
conditional on the assumption that the individual is alive, and at risk of incident lung cancer, at the
start of the year for each age-specific interval. Consequently, the risk of a Cr(VI)-related lung cancer
within a specified year of life is calculated as a function of (1) the probability of being alive at the
start of the year, (2) the background probability of getting lung cancer, and (3) the increased risk of
getting lung cancer from Cr(VI) exposure within the specified year. The lifetime risk is then the
sum of all the yearly risks. This procedure is performed to calculate the lifetime risk both for an
unexposed individual (R0) and for an individual with exposure to Cr(VI) (Rx).
"Extra risk" for lung cancer is a calculation of risk which adjusts for background incidence
rates of lung cancer, by estimating risk at a specified exposure level and is calculated as follows
fU.S. EPA. 2012al:
This document is a draft for review purposes only and does not constitute Agency policy.
4-70 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofHexavalent Chromium
Extra Risk = (Rx - R0) / (1 - RO)
The inhalation unit risk (IUR) is the risk of incident lung cancer per unit concentration
(|ig/m3) in inhaled air. The unit risk is calculated by using life-table analysis to find the exposure
concentration (EC) that yields a 1% (0.01) extra risk of lung cancer. The 1% value is referred to as
the Benchmark Response (BMR). This 1% value is used because lung cancer is a severe adverse
effect and 1% also represents a lung cancer response level that is near the low end of the
observable range fU.S. EPA. 2012al. This is also consistent with EPA's Benchmark Dose Technical
Guidance fU.S. EPA. 2012bl. which notes that a BMR of 1% is typically used for epidemiological data
since higher values may involve upward extrapolation.
Because a mutagenic MOA for Cr(VI) carcinogenicity (see Section 3.2.3) is "sufficiently
supported in (laboratory) animals" and "relevant to humans," EPA used a linear low dose
extrapolation from the POD in accordance with Guidelines for Carcinogen Risk Assessment (U.S.
EPA. 2005a). Given the EC associated with a 1% extra risk (ECoi), the unit risk is the slope of a
linear exposure-response line from the origin through the ECoi:
Unit Risk =0.01 / ECoi
A unit risk value may be calculated based on both the best estimate ((3) and the one-sided
95% upper confidence bound (UB) on the best estimate. The value based on the one-sided upper
95% confidence bound is normally used for decision-making, since it corresponds to a one-sided
lower 5% confidence bound (LB) on the exposure level yielding 1% extra risk (LECoi).
IUR = 0.01 / LECoi
Life-table calculations require as input the all-cause mortality rates and lung cancer
incidence rate for the general U.S. population in each year of life. The all-cause mortality data were
obtained from the National Vital Statistics Report Vol 68 No 7 Table 1 (Arias etal. (2017), which
provides data from the U.S. population in 2017. Lung cancer incidence rates were obtained by
downloading 2017 data for malignant neoplasms of bronchus and lung (ICD-10 C33-C34) from CDC
WONDER81. Because cause-specific rates were given for 5-year intervals, the cause-specific rate for
each 5-year interval was applied to each age within the interval.
The detailed equations for calculating lifetime excess cancer risk for a specified exposure
concentration in the presence of competing risks are based on the approach used by NRC (1988) for
evaluating lung cancer risks from radon. The equations are detailed in Appendix E. The SAS code
for lung cancer life-table analysis was provided to EPA by NIOSH82 and was adapted for use by
(1) entering the data noted above; (2) adding adjustment factors to account for differences between
occupational exposures and non-occupational exposure; (3) adding an equation to compute extra
81http://wonder.cdc.gov/ucd-icdl0.html.
82Beta Version. SAS 30NOV18, provided by Randall Smith, National Institute for Occupational Safety & Health.
This document is a draft for review purposes only and does not constitute Agency policy.
4-71 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
risk; and (4) adding a macro to solve for the ECoi or the LECoi. The SAS codes for performing the
lung cancer life-table calculations are provided in Appendix E.
The adjustment factors to account for differences between occupational exposures and non-
occupational exposure follow EPA guidance (U.S. EPA. 20091 that acknowledges there are
differences in breathing rates between workers (10 m3 per 8-hour day) and non-workers (20 m3
per 24-hour day) and that workers are exposed 240 days per year while non-workers are exposed
365 days per year (U.S. EPA. 2016b. 2014e. 2012d. 2011d). Thus, a worker is assumed to inhale
2,400 m3 of workplace air per year while a non-worker is assumed to inhale 7,300 m3 of air per
year. Since the effect estimates for Cr(VI) effects on lung cancer risks are in terms of'per
occupational year', the life-table procedure adjusts for the differences in air volume breathed per
year to represent non-occupational exposures.
4.4.3.2. Summary of Lifetime Unit Risk Estimates—Not Accounting for Assumed
Increased Early-Life Susceptibility
The derivation of the unit risk—not accounting for assumed increased early-life
susceptibility—is based upon the two main regression modeling results in Gibb etal. (2020):
(1) the Cox Proportional Hazard model with exposure lagged by 5 years, and (2) the R&L model
with four age groups and exposure lagged by 15 years. Note that this estimate of the unit risk is
based on the assumption that the relative risks or hazard ratios are independent of age.
Table 4-20. Calculation of lifetime cancer unit risk estimate not accounting for
assumed increased early-life susceptibility
Source
Table in
original
publication
P (Slope)
per mg CrOs/m3
P (Slope)
per |ig Cr(VI)/m3
Exposure
Concentration
associated with BMR
(1% Extra Risk)
[Hg Cr(VI)/m3]
Lifetime Unit Risk
[per ng Cr(VI)/m3]
MLE
95% UB
MLE
95% UB
ECoi
MLE
LECoi
5% LB
MLE
95% UB
Gibb et al.
(2020) Cox
PH Model
Table 1
5-year lag
0.487
0.675
9.36 x 10"4
1.30 x 10"3
1.25
0.899
8.02 x
10"3
1.11 X
10"2
Gibb et al.
(2020) R&L
Model
Table 2
4 age groups
15-year lag
0.599
0.849
1.15 xlO"3
1.63 x 10"3
1.35
0.952
7.41 x
10"3
1.05 x
10"2
The results from the Cox model yielded an estimate of the lifetime unit risk of 1.11 x 10-2
per [ig Cr(VI)/m3 while the results from the R&L model yielded an estimate of the lifetime unit risk
of 1.05 x lO-2 per [ig Cr(VI)/m3. These two estimates are very close to each other and thus mutually
support one another. EPA advanced the estimate of the lifetime unit risk derived from the Cox
proportional hazards models with an exposure lag of 5 years for the following reasons: (1) the Cox
This document is a draft for review purposes only and does not constitute Agency policy.
4-72 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 proportional hazards model is a well-established method for epidemiological analyses that is
2 commonly used in cohort studies, and (2) the results from this type of model have been used as the
3 basis for EPA IRIS IUR derivations for lung cancer (U.S. EPA. 2014el. breast cancer fU.S. EPA.
4 2016b) and lymphohematopoietic cancer (U.S. EPA. 2016b). In the absence of evidence of early-life
5 susceptibility, the lifetime unit risk for lung cancer caused by inhalation exposure to Cr(VI) is
6 considered to be best estimated as 1.11 x 10"2 per |ig Cr(VI)/m3.
7 Because a mutagenic mode of action for Cr(VI) carcinogenicity (see Section 3.2.3) is
8 "sufficiently supported in (laboratory) animals" and "relevant to humans," and as there are no
9 chemical-specific data to evaluate the differences between adults and children, increased early-life
10 susceptibility should be assumed. If there is early-life exposure, age-dependent adjustment factors
11 (ADAFs) are applied, as appropriate, in accordance with the EPA's Supplemental Guidance for
12 Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA. 2005b). See Section 4.4.4
13 below for more details on the application of ADAFs.
4.4.4. Application of Age-Dependent Adjustment Factors
14 The derivation of the IUR when increased early-life susceptibility should be assumed is
15 based on the same main Cox proportional hazards regression modeling results with exposure
16 lagged by 5 years (Gibb etal.. 2020). The process for deriving an IUR when increased early-life
17 susceptibility should be assumed involves an initial estimation of a unit risk based only on
18 adult-only exposures (U.S. EPA. 2016b). followed by the application of age-dependent adjustment
19 factors to age-specific risks for children under age 16 years, and a summary of risks across all ages
20 weighted by the age-dependent adjustment factors. This is accomplished with several steps.
21 • The first step is to apply the effect estimate (i.e., the MLE (3) from the Baltimore, MD cohort
22 and the 95% UB in a life-table initiating exposures at 16 years of age—instead of at birth.
23 This process estimates the unit risks for the 54-year period between age 16 years and age
24 70 years (IRIS' assumption of a lifetime).
25 • The values of the EC01 and LEC01 are derived in the same way using the life-table
26 procedure.
27 • These EC01 and LEC01 values are then divided into the benchmark response of 1% to
28 compute the 'adult-exposure-only' unit risk estimates.
29 • The 'adult-exposure-only' unit risk estimates are multiplied by 70/54 to rescale the 54-year
30 adult period to 70 years. This yields the 'adult-based' lifetime unit risk.
31 • The last step is to apply the ADAFs which adjust the 'adult-based' lifetime age-specific unit
32 risk for children ages less than two years upwards by 10-fold during those years of life, and
33 the unit risk for children ages 2-15 upwards by 3-fold during those years of life, and then
34 applies the unadjusted 'adult-based' lifetime unit risk for people aged 16-70 during those
35 years of life. The weighted sum of these three partial unit risks is the ADAF-adjusted
36 lifetime IUR.
This document is a draft for review purposes only and does not constitute Agency policy.
4-73 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofHexavalent Chromium
Table 4-21. Calculation of total cancer unit risk estimate from adult-only
exposure
Exposure
concentration
associated with BMR
(1% Extra Risk)
Adu It-exposu re-on ly
Adult-based
Starting exposure at
unit risk
unit risk
P (Slope)
age 16 years
[per ng Cr(VI)/m3]
[per ng Cr(VI)/m3]
per |ig Cr(VI)/m3
[HgCr
Vl)/m3]
(54 years)
(70 years)
ECoi(16+)
LECoi(16+)
Source
MLE
95% UB
MLE
5% LB
MLE
95% UB
MLE
95% UB
Gibbetal. (2020)
9.36 x
1.30 x
1.64
1.18
6.12 x 10"3
8.48 x 10"3
7.93 x
1.10 x
Cox PH Model
10"4
10"3
10"3
10"2
5-year lag
The results from the Cox model yielded an estimate of the 'adult-based' unit risk of 1.10 x
10"2 per [ig Cr(VI)/m3. Application of the ADAFs to the 'adult-based' (rescaled as discussed above)
unit risk estimate for Cr(VI) for a lifetime inhalation exposure scenario is presented below. The
unit risk for each age group is the product of the values for the ADAF, the adult-based unit risk, and
the duration adjustment in columns 2-4 [e.g., 10 x (1.10 x 10-2) x 2/70 = 3.14 x 10~3], and the total
risk is the sum of the partial risks. This lifetime inhalation unit risk estimate for a constant
exposure of 1 |ig Cr(VI)/m3 is adjusted for potential increased early-life susceptibility, assuming a
70-year lifetime.
Table 4-22. Total cancer risk from exposure to constant Cr(VI) exposure level
of 1 |ig/m3 from ages 0-70 years, adjusted for potential increased early-life
susceptibility
Age group
ADAF
Adult-based unit risk
(per ng Cr(VI)/m3)
Duration adjustment
Unit risk
[per ng Cr(VI)/m3]
0-<2 years
10
1.10 x 10"2
2 years/70 years
3.14 x 10"3
2-<16 years
3
1.10 x 10"2
14 years/70 years
6.60 x 10"3
>16 years
1
1.10 x 10"2
54 years/70 years
8.49 x 10"3
Total Lifetime Risk
1.82 x 10"2
The lifetime inhalation unit risk for Cr(VI) is 1.82 x 10 2 per ng Cr(VI)/m3. This value is rounded to 2 x 10 2 per ng
Cr(VI)/m3.
If calculating the cancer risk for a 30-year exposure to a constant average concentration of
0.01 ng Cr(VI)/m3from ages 0 to 30 years, the duration adjustments would be 2/70,14/70, and
14/70, and the partial risks would be (10 x 0.011 x 0.01 x 2/70= 3.1 x 10 5), (3 x 0.011 x 0.01 x
This document is a draft for review purposes only and does not constitute Agency policy.
4-74 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Toxicological Review ofHexavalent Chromium
14/70=6.6 x 105), and (1 x 0.011 x 0.01 x 14/70 = 2.2 x 105), resulting in a total risk estimate of
1.2 x io-4.
If calculating the cancer risk for a 30-year exposure to a constant average average
concentration of 0.01 |ig Cr(VI)/m3 from ages 20 to 50 years, the duration adjustments would be
0/70, 0/70, and 30/70, and the partial risks would be 0, 0, and (1 x 0.011 x 0.01 x 30/70 = 4.7 x 10-
5), resulting in a total risk estimate of 4.7 x 10-5.
4.4.5. Uncertainties in the Derivation of the Inhalation Unit Risk
Several potential sources of uncertainty were identified in the derivation of the Cr(VI)
inhalation unit risks. As discussed below, these were not found to be major influences in this
evaluation—including two potential sources of uncertainties generally associated with larger
uncertainty (model uncertainty and low dose extrapolations). Uncertainties related to genetics,
physiological differences, and particle deposition have been discussed previously in this
assessment (see Sections 3.1.1.2 and 3.3.1), and the inhalation unit risk represents an upper bound
on the average risk in a population fU.S. EPA. 2005al.
Sources of uncertainty in this assessment are outlined below.
4.4.5.1. Uncertainty in Exposure Assessment
Routine air sampling was initiated after construction of the new Baltimore, MD facility in
1950 and followed written documentation specifying strategies for air sampling. Sampling was
intended to representthe "typical/usual exposures" to workers fGibb etal.. 2000bl. Table 4-23
below details the sampling regimen over time. In constructing the job-exposure-matrix to assign
individual exposure for each worker, Gibb etal. f2000bl relied on approximately 70,000
measurements across the study period. While the sampling regimes changed over time and can
reasonably be expected to have improved in quantity and specificity, the samples were collected
methodically and used the same analytical method for assessing Cr(VI) concentration in dust over
the study period (Gibb etal.. 2000b).
These exposure estimates were used to construct a job-exposure-matrix (JEM) for each of
the 114 job titles in each of the 36 years of the study period. According to Gibb etal. (2000b). the
JEM was "virtually complete" for the later years (1971-1985) and "fairly complete" for the early
years from 1950-1956 and 1960-1961. While the sampling records for nine years could not be
located, those values were imputed based on existing data to model those job-specific exposure
values. EPA considered uncertainty to be low for the 24 out of 36 years when sampling records was
available and low-to-medium for the missing years that were bookended by actual sampling values.
As exposures may reasonably be assumed to have decreased over the study period as industrial
hygiene practices improved, the interpolation between higher and lower exposure periods was
likely to have captured those interim exposure concentrations.
This document is a draft for review purposes only and does not constitute Agency policy.
4-75 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofHexavalent Chromium
Table 4-23. Overview of air sampling program for the Baltimore cohort
evaluated by Gibb et al., (2015: 2000b)
Exposure measurement system
Years
implemented
Frequency and duration
Airborne dust via high-volume air sampling
pumps and impingers, with sampling wand
held in worker breathing zone.
1950-1961
Short-term samples (tens of minutes).
24-hour routine measurements (fixed-site
monitors) using 20 tape air samplers
(Research Appliance Co., Allison Park, PA).
Observation of how much workers spent in
the vicinity of each of these monitors.
Mid-1960s-1979
24 1-hour samples. Samplers rotated through
154 fixed sites representing exposure zones.
1979-1985
After 1979, frequency reduced to 8 3-hour
samples, and number of fixed sites reduced to
27.
Routine personal sample collection using
NIOSH standard method P and CAM 169
(NIOSH, 1972).
1977-1985
Full-shift sampling.
4.4.5.2. Uncertainty in the Exposure Metric
Gibb etal. f2000bl fit multiple models of lung cancer risks using untransformed and
transformed cumulative exposure to Cr(VI) with log base-10 transformed Cr(VI) providing the
better overall model fit Gibb etal. f20151 also reported updated lung cancer results based on log
base-10 of cumulative exposure to Cr(VI). While log transformation of concentration-based
cumulative exposure is commonplace in epidemiological analyses because those concentrations are
often log-normally distributed, risk calculation based on log-transformed exposure suffer from
exposure-response irregularities such as zero risk whenever the exposure has a numerical value of
one (in any units) [i.e., Iogl0(l) = 0 or ln(l) = 0], and when risks are extrapolated below one unit of
exposure, the sign of the risk estimate flips from positive to negative such that lower exposure
appears to be health protective as an artifact of the transformation. For the purpose of estimating
an IUR, exposure-response results in terms of untransformed cumulative exposures to Cr(VI) can
be more useful than log-transformed exposures. Gibb etal. (20201 reported risks of lung cancer
associated with untransformed cumulative Cr(VI). While a transformed exposure may provide a
better overall model across the entire range of exposures in a study, as in the case of Gibb et al.
f20201. those model results did not meet the needs for estimating an IUR based on a POD in the low
exposure range, and thus EPA selected the results from the models based on untransformed
cumulative Cr(VI)—even if there is some uncertainty concerning the relative fits of different
exposure metrics.
The two candidate IUR's are based on the same cohort that was most highly rated and
preferred on the majority of additional considerations for exposure-response, there are some
aspects of the specific modeling details that were further considered in order to judge their
potential impact on the IUR. Specifically, the exposure lags and the number of age groups that
This document is a draft for review purposes only and does not constitute Agency policy.
4-76 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
1 yielded the better overall fits, often the fit differences were small enough so as to be essentially
2 equal in fit Three additional sets of candidate unit risks were derived to show the differences in
3 those values had those combinations been selected instead, and to allow for comparison between
4 the two candidate IURs on a common basis of exposure lag length.
Table 4-24. Variation in unit risks among the Cox Proportional Hazards model
results by lag length
Cox proportional hazards
Lag period (v) in Gibb et al. (2020) matched in life-table
Lag period (v) in Gibb et al. (2020)
Lifetime unit risk (95%UB) without
ADAFs
[per |ig Cr(VI)/m3]
Lifetime unit risk (95%UB) with
ADAFs
[per |ig Cr(VI)/m3]
0
1.16 x 10"2
2.00 x 10"2
5
1.11 x 10"2
1.82 x 10"2
10
1.05 x 10"2
1.64 x 10"2
15
9.82 x 10"3
1.47 x 10"2
5 4.4.5.3. Uncertainty in the Outcome Metric
6 Lung cancer mortality was ascertained from death certificates according to specific codes
7 from the International Classification of Diseases—eighth edition, and this coding system and those
8 of previous editions have been stable over time. Uncertainty is considered to be very low for lung
9 cancer mortality.
10 4.4.5.4. Uncertainty Due to Length of Follow-up
11 There is little potential uncertainty regarding the length of follow-up for cancer mortality.
12 The hire dates among this cohort ranged from August 1,1950 to December 31,1974 (the mean date
13 of hire was mid-1957) fGibb etal.. 2000b). Follow-up continued until the date of death, age
14 96 years, or December 31, 2011, whichever occurred first Therefore, the range of follow-up was
15 from 37 to 61 years, with a mean of more than 38 years.
16 4.4.5.5. Uncertainty in Model Form
17 For lung cancer mortality, the Cox proportional hazards model is a well-established method
18 for epidemiological analyses that is commonly used in cohort studies because this type of survival
19 analysis takes into account differences in follow-up time among the cohort and is approximately
20 linear at low exposures. This model form allows for the evaluation and control of important
21 potential confounding factors such as age and smoking, and for the modeling of exposure as a
22 continuous variable. There is little uncertainty in the choice of model form. Additionally, the R&L
23 model is an alternative approach to the Poisson model and results from this modeling yielded
24 similar results which further reduces the uncertainty in the choice of model form.
This document is a draft for review purposes only and does not constitute Agency policy.
4-77 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofHexavalent Chromium
4.4.5.6. Uncertainty in Control of Potential Confounding in Modeling Lung Cancer
Mortality
It is well known that smoking is a strong independent risk factor for lung cancer.
Additionally, tobacco smoke contains chromium (Fresquez etal.. 20131. and therefore smokers are
expected to be exposed to higher levels of total chromium than nonsmokers. Company medical
records provided smoking status at the beginning of employment for 91% of the cohort
(Yes/No/Unknown); 74% smoked cigarettes, 16% did not smoke, and smoking status was
unknown for 9% (Gibb etal.. 2000bl. No information on pack-years of smoking or how smoking
status may have changed over time was available. As an important potential confounder of the lung
cancer mortality analysis, smoking was controlled for in the analyses of lung cancer mortality
associated with exposure Cr(VI) (Gibb etal.. 2020: 2015: 2000b). Each of the Cox proportional
hazards analyses showed that smoking at the beginning of employment was a strong predictor of
lung cancer risk. While additional information on the cumulative exposure to smoking may have
been helpful to more completely control for smoking, it is clear that as measured, smoking was a
strong independent predictor of lung cancer risks and was independent of cumulative Cr(VI)
exposure as it was measured at the beginning of employment There remains some uncertainty as
to any potential residual confounding that might be attributed to lack of smoking data on 9% of the
cohort and the lack of information on any changes in smoking over time. However, the Baltimore
cohort had much better data on smoking compared to the Painesville cohort, and thus the selection
of the Baltimore cohort minimizes the potential for confounding by smoking among the available
cohorts.
4.4.5.7. Uncertainty Due to Potential Effect Modification
Among the 217 deaths from lung cancer in workers, only four were among nonsmokers
(Gibb etal.. 20151 and the investigators were unable to evaluate any potential statistical interaction
between smoking and Cr(VI) exposure. It is theoretically possible that the risk of lung cancer
mortality estimated in this current assessment is a reflection of a positive synergy between
smoking and Cr(VI), and that the adverse effect of Cr(VI) among nonsmokers has been
overestimated. However, this possibility cannot be assessed and remains an uncertainty. The unit
risk of the lung cancer risk herein would be health protective for any population that had a lower
prevalence of smoking than that of the Baltimore cohort.
4.4.5.8. Uncertainty in Low Dose Extrapolation
A common source of uncertainty in quantitative cancer risk assessments generally derives
from extrapolating from high doses in animals to low doses in humans. Compared to assessments
based on animal data, the uncertainty from low-dose extrapolation in this assessment, which uses
occupational epidemiology data, is considered to be low because the POD was well within the range
of observed exposure data. The POD for lung cancer was based on 1% extra risk and yielded an
LECoi of 0.899 [ig Cr(VI)/m3 from the Cox analysis and 0.951 [ig Cr(VI)/m3 from the R&L analysis.
This document is a draft for review purposes only and does not constitute Agency policy.
4-78 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofHexavalent Chromium
Table 2 ofGibb etal. f20151 shows that the median cumulative exposure to Cr03 was 0.01 mg
Cr03/m3-years and the 25%-tile of Cr03 was 0.001 mg Cr03/m3-years, and the minimum was zero.
Converting to [ig Cr(VI)/m3, the median was 52 [ig Cr(VI)/m3 and the 25%-tile was 5.2 [ig
Cr(VI)/m3. Here the PODs appear to be between the minimum and the 25%-tile and thus not
outside the range of observed exposures. Thus, there is little uncertainty in extrapolation of the
risk function below the POD associated with a 1% BMR.
4.4.5.9. Uncertainty in Extrapolation of Findings in Adults to Children.
The analysis of lung cancer mortality using the Cox proportional hazards model assumed
that the effect was independent of age, while the analysis using the R&L approach allowed for
effects to be different by age group—although this analysis did not provide any estimates of what
the age effect was beyond showing that the relatively younger cohort members appeared to be at
higher risk of lung cancer mortality than the older cohort members. Given that both of these
analyses yielded approximately the same estimate of the IUR, it appears that while there may be an
age-related effect of Cr(VI) exposure on the risk of lung cancer, two different analyses that treated
age differently yielded essentially the same unit risk when the life-table analysis assumed that the
effect was independent of age.
However, Cr(VI) was found to cause cancer by a mutagenic mode of action, and chemical-
specific data are not available to address early-life exposure. According to EPA's Supplemental
Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA. 2005b).
ADAF are applied for children and risks were based on application of age-dependent risk modifiers
of an "adult-only" unit risks such that effect were independent of age among people age 16 years
and older. There is some uncertainty that these default ADAF would be health-protective of
children although this uncertainty is considered to be low.
The inhalation unit risk (IUR) is a plausible upper bound on the estimate of risk per |J.g/m3
air breathed. The IUR can be multiplied by an estimate of lifetime exposure (in |ig/m3} to estimate
the lifetime cancer risk.
4.4.6. Previous IRIS Assessment: Inhalation Unit Risk
The previous IRIS assessment for Cr(VI) was posted to the IRIS database in 1998. EPA's
1998 IRIS assessment classified Cr(VI) as "Group A—known human carcinogen by the inhalation
route of exposure" under the 1986 guidelines (U.S. EPA. 1986b). This was based on evidence of a
causal relationship between inhalation of Cr(VI) and increased incidence of lung cancer in humans
in occupational settings. An inhalation unit risk (IUR) for Cr(VI) of 1.2 x 10~2 per |J.g/m3 was
calculated based on increased incidence of lung cancer in chromate workers from the Painesville
OH cohort (Mancuso. 1997.1975). Because Mancuso et al. (1997.1975) only provided total
chromium data and contained fewer employee records for smoking status, there was higher
uncertainty in the 1998 IUR. The lack of Cr(VI) data would have led to an underestimation of risk
(because the true Cr(VI) exposure rates were lower relative to total chromium exposure rates),
This document is a draft for review purposes only and does not constitute Agency policy.
4-79 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofHexavalent Chromium
while the lack of smoking data would have led to an overestimation of risk (due to the high
prevalence of smoking during this time period).
This document is a draft for review purposes only and does not constitute Agency policy.
4-80 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Toxicological Review ofHexavalent Chromium
REFERENCES
Abbaspour. N: Hurrell. R: Kelishadi. R. (2014). Review on iron and its importance for human health
[Review], J Res Med Sci 19: 164-174.
ACC (American Chemistry Council). (2015). Pathology peer slide review of non-neoplastic
findings in the duodenum of mice and rats treated with sdd in acc-ts and ntp
studies.
Acharva. S: Mehta. K: Krishnan. S: Rao. CV. (2001). A subtoxic interactive toxicity study of ethanol
and chromium in male Wistar rats. Alcohol 23: 99-108. http://dx.doi.org/10.1016/SQ741-
8329C00100139-7
Acharva. UR: Mishra. M: Mishra. I: Tripathv. RR. (2004). Potential role of vitamins in chromium
induced spermatogenesis in Swiss mice. Environ Toxicol Pharmacol 15: 53-59.
http://dx.doi.Org/10.1016/i.etap.2003.08.010
Acharva. UR: Mishra. M: Tripathv. RR: Mishra. I. (2006). Testicular dysfunction and antioxidative
defense system of Swiss mice after chromic acid exposure. Reprod Toxicol 22: 87-91.
http://dx.doi.Org/10.1016/j.reprotox.2005.ll.004
Adachi. S. (1987). [Effects of chromium compounds on the respiratory system: Part 5. Long term
inhalation of chromic acid mist in electroplating by C57BL female mice and recapitulation of
our experimental studies], Sangyo Igaku 29: 17-33.
http://dx.doi.org/10.1539/johl959.29.17
Adachi. S: Yoshimura. H: Katavama. H: Takemoto. K. (1986). [Effects of chromium compounds on
the respiratory system: Part 4. Long term inhalation of chromic acid mist in electroplating
to ICR female mice], Sangyo Igaku 28: 283-287. http://dx.doi.org/10.1539/iohl959.28.283
Adachi. S: Yoshimura. H: Miyavama. R: Katavama. H: Takemoto. K: Kawai. H. (1981). [Effects of
chromium compounds to the respiratory system: Part 1. An experimental study of
inhalation of chromic acid mist in electroplating], Sangyo Igaku 23: 294-299.
http://dx.doi.org/10.1539/johl959.23.294
AEI (Applied Epidemiology Inc). (2002). Collaborative cohort mortality study of four chromate
production facilities, 1958-1998: Final report (8EHQ-0810-18050; 0SHA Docket H054A Ex.
48-1-2). Pittsburgh, PA: Industrial Health Foundation.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/OTS0602411.xhtml
Ahmad. MK: Svma. S: Mahmood. R. (2011). Cr(VI) Induces Lipid Peroxidation, Protein Oxidation and
Alters the Activities of Antioxidant Enzymes in Human Erythrocytes. Biol Trace Elem Res
144: 426-435. http://dx.doi.org/10.10C)7/sl2011-011-9119-5
Akbar. M: Brewer. TM: Grant. MH. (2011). Effect of chromium and cobalt ions on primary human
lymphocytes in vitro. J Immunotoxicol 8: 140-149.
http://dx.doi.org/10.3109/1547691X.2011.553845
Al-Hamood. MH: Elbetieha. A: Bataineh. H. (1998). Sexual maturation and fertility of male and
female mice exposed prenatally and postnatally to trivalent and hexavalent chromium
compounds. Reprod Fertil Dev 10: 179-183.
Alexander. 1: Aaseth. I. (1995). Uptake of chromate in human red blood cells and isolated rat liver
cells: the role of the anion carrier. Analyst 120: 931-933.
http://dx.doi.Org/l 0.1039/AN9952000931
This document is a draft for review purposes only and does not constitute Agency policy.
R-l DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Alexander. 1: Aaseth. I: Norseth. T. (1982). Uptake of chromium by rat liver mitochondria.
Toxicology 24: 115-122. http://dx.doi.org/10.1016/0300-483Xr82190050-6
Alexander. 1: Mikalsen. A: Ryberg. D. (1986). Microsomal reduction of Cr VI. Acta Pharmacol Toxicol
59: 267-269. http://dx.doi.Org/10.llll/i.1600-0773.1986.tb02759.x
Alexopoulos. EC: Cominos. X: Trougakos. IP: Lourda. M: Gonos. ES: Makropoulos. V. (2008).
Biological monitoring of hexavalent chromium and serum levels of the senescence
biomarker Apolipoprotein J/Clusterin in welders. Bioinorganic Chemistry and Applications
2008: 420578. http://dx.doi.org/10.1155/2008/420578
Ali. AH: Kondo. K: Namura. T: Senba. Y: Takizawa. H: Nakagawa. Y: Toba. H: Kenzaki. K: Sakivama. S:
Tangoku. A. (2011). Aberrant DNA methylation of some tumor suppressor genes in lung
cancers from workers with chromate exposure. Mol Carcinog 50: 89-99.
http://dx.doi.org/10.1002/mc.20697
Almaguer. D: Kramkowski. R. (1983). Health hazard evaluation report no. HETA-82-116-1319,
Modine Manufacturing, McHenry, Illinois. Cincinnati, OH: National Institute for Occupational
Safety and Health.
American Pregnancy Association. (2021). Anemia during pregnancy. Available online at
https://americanpregnancy.org/healthy-pregnancy/pregnancy-concerns/anemia-during-
pregnancv/ (accessed August 10, 2022).
Andersen. A: Barlow. L: Engeland. A: Kiaerheim. K: Lvnge. E: Pukkala. E. (1999). Work-related
cancer in the Nordic countries [Supplemental Data], Scand J Work Environ Health 25: 1-116.
Andrews. R: O'Connor. PF. (2020). NIOSH manual of analytical methods (NMAM), 5th edition.
Washington, DC: National Institute for Occupational Safety and Health.
https://www.cdc.gov/niosh/nmam/default.html
Anger. G: Halstenberg. I: Hochgeschwender. K: Scherhag. C: Korallus. U: Knopf. H: Schmidt. P:
Ohlinger. M. (2005). Chromium compounds. In B Elvers (Ed.), Ullmann's Encyclopedia of
Industrial Chemistry (7th ed., pp. 1-36). Hoboken, NJ: Wiley-VCH Verlag GmbH & Co.
http://dx.doi.Org/10.1002/14356007.a07 067
Antonini. TM: Lawrvk. NT: Murthv. GG: Brain. ID. (1999). Effect of welding fume solubility on lung
macrophage viability and function in vitro. J Toxicol Environ Health A 58: 343-363.
http://dx.doi. org/10.1080/009841099157205
Antonini. TM: Roberts. TR: Chapman. RS: Soukup. TM: Ghio. AT: Sriram. K. (2010a). Pulmonary toxicity
and extrapulmonary tissue distribution of metals after repeated exposure to different
welding fumes. Inhal Toxicol 22: 805-816.
http://dx.doi.Org/10.3109/08958371003621641
Antonini. TM: Taylor. MP: Zimmer. AT: Roberts. TR. (2010b). Pulmonary responses to welding
fumes: Role of metal constituents [Review], J Toxicol Environ Health A 67: 233-249.
http://dx.doi.Org/10.1080/15287390490266909
Anwar. RA: Langham. RF: Hoppert. CA: Alfredson. BY: Bverrum. RU. (1961). Chronic toxicity
studies: III. Chronic toxicity of cadmium and chromium in dogs. Arch Environ Health 3: 456-
460.
Aoki. Y: Matsumoto. M: Matsumoto. M: Masumura. K: Nohmi. T. (2019). Mutant frequency is not
increased in mice orally exposed to sodium dichromate. FoodSaf7: 2-10.
http://dx.doi.org/10.14252/foodsafetyfscj.2018014
Apte. AD: Tare. V: Bose. P. (2006). Extent of oxidation of Cr(III) to Cr(VI) under various conditions
pertaining to natural environment J Hazard Mater 128: 164-174.
http://dx.doi.Org/10.1016/i.ihazmat.2005.07.057
Arakawa. H: Ahmad. R: Naoui. M: Taimir-Riahi. HA. (2000). A comparative study of calf thymus DNA
binding to Cr(III) and Cr(VI) ions. Evidence for the guanine N-7-chromium-phosphate
chelate formation. J Biol Chem 275: 10150-10153.
This document is a draft for review purposes only and does not constitute Agency policy.
R-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Arakawa. H: Weng. MW: Chen. WC: Tang. MS. (2012). Chromium (VI) induces both bulky DNA
adducts and oxidative DNA damage at adenines and guanines in the p53 gene of human lung
cells. Carcinogenesis 33: 1993-2000. http://dx.doi.org/10.lQ93/carcin/bgs237
Arias. E: Heron. M: Xu. 1. (2017). United States life tables, 2013. National Vital Statistics Reports; vol
66 no 3. Hyattsville, MD: National Center for Health Statistics.
Armienta-Hernandez. MA: Rodriguez-Castillo. R. (1995). Environmental exposure to chromium
compounds in the valley of Leon, Mexico. Environ Health Perspectl03 Suppl 1: 47-51.
Arshad. M: Ahmad. N: Khalid. M: Asmatullah: Tahir. M: Naveed. K: Iqbal. A. (2017). Exposure of
pregnant mice to hexavalent chromium causes fetal defects. Pakistan J Zool 49: 1383-1389.
http://dx.doi.Org/10.17582/iournal.piz/2017.49.4.1383.1389
Aruldhas. MM: Subramanian. S: Sekar. P: Vengatesh. G: Chandrahasan. G: Govindaraiulu. P:
Akbarsha. MA. (2005). Chronic chromium exposure-induced changes in testicular
histoarchitecture are associated with oxidative stress: study in a non-human primate
(Macaca radiata Geoffroy). Hum Reprod 20: 2801-2813.
http://dx.doi.org/10.1093/humrep/deil48
Aruldhas. MM: Subramanian. S: Sekhar. P: Hasan. GC: Govindaraiulu. P: Akbarsha. MA. (2004).
Microcanalization in the epididymis to overcome ductal obstruction caused by chronic
exposure to chromium - a study in the mature bonnet monkey (Macaca radiata Geoffroy).
Reproduction 128: 127-137. http://dx.doi.Org/10.1530/rep.l.00067
Aruldhas. MM: Subramanian. S: Sekhar. P: Vengatesh. G: Govindaraiulu. P: Akbarsha. MA. (2006). In
vivo spermatotoxic effect of chromium as reflected in the epididymal epithelial principal
cells, basal cells, and intraepithelial macrophages of a nonhuman primate (Macaca radiata
Geoffroy). Fertil Steril 86: 1097-1105. http://dx.doi.Org/10.1016/i.fertnstert.2006.03.025
Asatiani. N: Abuladze. M: Kartvelishvili. T: Kulikova. N: Asanishvili. L: Holman. HY: Sapoinikova. N.
(2010). Response of antioxidant defense system to chromium (Vl)-induced cytotoxicity in
human diploid cells. Biometals 23: 161-172. http://dx.doi.org/10.1007/sl0534-009-9276-
6
Asatiani. N: Kartvelishvili. T: Abuladze. M: Asanishvili. L: Sapoinikova. N. (2011). Chromium (VI) can
activate and impair antioxidant defense system. Biol Trace Elem Res 142: 388-397.
http://dx.doi. org/10.1007 /si 2011-010-8806-v
Asatiani. N: Sapoinikova. N: Abuladze. M: Kartvelishvili. T: Kulikova. N: Kiziria. E: Namchevadze. E:
Holman. HY. (2004). Effects of Cr(VI) long-term and low-dose action on mammalian
antioxidant enzymes (an in vitro study). J Inorg Biochem 98: 490-496.
http://dx.doi.Org/10.1016/i.iinorgbio.2003.12.014
Asgharian. B: Price. 0. (2009). Multiple-path particle dosimetry model (MPPD v 2.11) (Version 2.11)
[Computer Program], Albuquerque, NM: Applied Research Associates. Retrieved from
https://www.ara.com/mppd/
Asmatullah: Noreen. MA. (1999). Effect of oral administration of hexavalent chromium on total
body weight, chromium uptake and histological structure of mouse liver. Punjab Univ J Zool
14: 53-63.
Atanassoff. PG: Brull. ST: Weiss. BM: Landefeld. K: Alon. E: Rohling. R. (1995). The time course of
gastric pH changes induced by omeprazole and ranitidine: A 24-hour dose-response study.
Anesth Analg 80: 975-979. http://dx.doi.Org/10.1097/00000539-199505000-00021
ATS/ERS (American Thoracic Society and European Respiratory Society). (2019). Standardization
of spirometry 2019 update. Am J Respir Crit Care Med 200: e70-e88.
http://dx.doi.org/10.1164/rccm.201908-1590ST
ATSDR (Agency for Toxic Substances and Disease Registry). (2005). Toxicological profile for nickel
[ATSDRTox Profile], Atlanta, GA: U.S. Department of Health and Human Services, Public
Health Service, http://www.atsdr.cdc.gov/toxprofiles/tpl5.pdf
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
ATSDR (Agency for Toxic Substances and Disease Registry). (2012). Toxicological profile for
chromium. Atlanta, GA: US Department of Health and Human Services, Public Health
Service. http://www.atsdr.cdc.gov/ToxProfiles/tp.asp?id=62&tid=17
Azad. N: Iyer. AK: Manosroi. A: Wang. L: Roianasakul. Y. (2008). Superoxide-mediated proteasomal
degradation of Bel-2 determines cell susceptibility to Cr (VI)-induced apoptosis.
Carcinogenesis 29: 1538-1545. http://dx.doi.org/10.1093/carcin/bgnl37
Bach. SP: Renehan. AG: Potten. CS. (2000). Stem cells: the intestinal stem cell as a paradigm
[Review], Carcinogenesis 21: 469-476. http://dx.doi.Org/10.1093/carcin/21.3.469
Badding. MA: Fix. NR: Antonini. TM: Leonard. SS. (2014). A Comparison of Cytotoxicity and Oxidative
Stress from Welding Fumes Generated with a New Nickel-, Copper-Based Consumable
versus Mild and Stainless Steel-Based Welding in RAW 264.7 Mouse Macrophages. PLoS
ONE 9:el01310. http://dx.doi.org/10.1371/iournal.pone.0101310
Bagchi. D: Bagchi. M: Stohs. ST. (2001). Chromium (Vl)-induced oxidative stress, apoptotic cell death
and modulation of p53 tumor suppressor gene. Mol Cell Biochem 222: 149-158.
http://dx.doi.Org/10.1023/A:1017958028256
Bagchi. D: Balmoori. I: Bagchi. M: Ye. X: Williams. CB: Stohs. ST. (2000). Role ofp53 tumor
suppressor gene in the toxicity of TCDD, endrin, naphthalene, and chromium (VI) in liver
and brain tissues of mice. Free Radic Biol Med 28: 895-903.
http://dx.doi.org/10.1016/S0891-5849r00100173-8
Bagchi. D: Balmoori. 1: Bagchi. M: Ye. X: Williams. CB: Stohs. ST. (2002). Comparative effects of TCDD,
endrin, naphthalene and chromium (VI) on oxidative stress and tissue damage in the liver
and brain tissues of mice. Toxicology 175: 73-82. http://dx.doi.org/10.1016/S030Q-
483XC02100062-8
Bagchi. D: Hassoun. EA: Bagchi. M: Muldoon. DF: Stohs. ST. (1995a). Oxidative stress induced by
chronic administration of sodium dichromate [Cr(VI)] to rats. Comp Biochem Physiol C
Pharmacol Toxicol Endocrinol 110: 281-287. http://dx.doi.org/10.1016/Q742-
8413C94100103-H
Bagchi. D: Hassoun. EA: Bagchi. M: Stohs. ST. (1995b). Chromium-induced excretion of urinary lipid
metabolites, DNA damage, nitric oxide production, and generation of reactive oxygen
species in Sprague-Dawley rats. Comp Biochem Physiol C Pharmacol Toxicol Endocrinol
110: 177-187. http://dx.doi.org/10.1016/0742-8413r94100093-P
Bagchi. D: Vuchetich. PI: Bagchi. M: Hassoun. EA: Tran. MX: Tang. L: Stohs. ST. (1997). Induction of
oxidative stress by chronic administration of sodium dichromate [chromium VI] and
cadmium chloride [cadmium II] to rats. Free Radic Biol Med 22: 471-478.
http://dx.doi.org/10.1016/s0891-5849r96100352-8
Bai. TPF: Burckart. GT: Mulberg. AE. (2016). Literature review of gastrointestinal physiology in the
elderly, in pediatric patients, and in patients with gastrointestinal diseases [Review], J
Pharm Sci 105: 476-483. http://dx.doi.org/10.1002/ips.24696
Bailey. SA: Zidell. RH: Perry. RW. (2004). Relationships between organ weight and body/brain
weight in the rat: What is the best analytical endpoint? Toxicol Pathol 32: 448-466.
http://dx.doi.org/10.1080/01926230490465874
Balachandar. V: Arun. M: Mohana Devi. S: Velmurugan. P: Manikantan. P: Karthick Kumar. A:
Sasikala. K: Venkatesan. C. (2010). Evaluation of the genetic alterations in direct and
indirect exposures of hexavalent chromium [Cr(VI)] in leather tanning industry workers
North Arcot District, South India. Int Arch Occup Environ Health 83: 791-801.
http://dx.doi.Org/10.1007/s00420-010-0562-v
Balashazv. I: Hofmann. W: Heistracher. T. (2003). Local particle deposition patterns may play a key
role in the development of lung cancer. J Appl Physiol (1985) 94: 1719-1725.
http://dx.doi.org/10.1152/iapplphysiol.00527.2002
This document is a draft for review purposes only and does not constitute Agency policy.
R-4 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Ban. M: Langonne. I: Goutet. M: Huguet. N: Pepin. E. (2010). Simultaneous analysis of the local and
systemic immune responses in mice to study the occupational asthma mechanisms induced
by chromium and platinum. Toxicology 277: 29-37.
http://dx.doi.0rg/lO.lOl6/i.tox.2OlO.O8.OlO
Banu. SK: Samuel. IB: Arosh. TA: Burghardt. RC: Aruldhas. MM. (2008). Lactational exposure to
hexavalent chromium delays puberty by impairing ovarian development, steroidogenesis
and pituitary hormone synthesis in developing Wistar rats. Toxicol Appl Pharmacol 232:
180-189. http://dx.doi.0rg/lO.lOl6/i.taap.2OO8.O6.OO2
Banu. SK: Stanley. TA: Lee. 1: Stephen. SD: Arosh. TA: Hover. PB: Burghardt. RC. (2011). Hexavalent
chromium-induced apoptosis of granulosa cells involves selective sub-cellular translocation
ofBcl-2 members, ERK1/2 andp53. Toxicol Appl Pharmacol 251: 253-266.
http://dx.doi.0rg/lO.lOl6/i.taap.2Oll.Ol.Oll
Banu. SK: Stanley. TA: Sivakumar. KK: Arosh. TA: Barhoumi. R: Burghardt. RC. (2015). Identifying a
novel role for X-prolyl aminopeptidase (Xpnpep) 2 in CrVI-induced adverse effects on germ
cell nest breakdown and follicle development in rats. Biol Reprod 92: 67.
http://dx.doi.org/10.1095/biolreprod.114.125708
Banu. SK: Stanley. TA: Sivakumar. KK: Arosh. TA: Burghardt. RC. (2016). Resveratrol protects the
ovary against chromium-toxicity by enhancing endogenous antioxidant enzymes and
inhibiting metabolic clearance of estradiol. Toxicol Appl Pharmacol 303: 65-78.
http://dx.doi.0rg/lO.lOl6/j.taap.2Ol6.O4.Ol6
Banu. SK: Stanley. TA: Sivakumar. KK: Arosh. TA: Taylor. RT: Burghardt. RC. (2017a). Chromium VI -
Induced developmental toxicity of placenta is mediated through spatiotemporal
dysregulation of cell survival and apoptotic proteins. Reprod Toxicol 68: 171-190.
http://dx.doi.0rg/lO.lOl6/j.reprotox.2Ol6.O7.OO6
Banu. SK: Stanley. TA: Sivakumar. KK: Taylor. RT: Arosh. TA: Burghardt. RC. (2017b). Editor's
highlight: Exposure to CrVI during early pregnancy increases oxidative stress and disrupts
the expression of antioxidant proteins in placental compartments. Toxicol Sci 155: 497-511.
http://dx.doi.org/10.1093/toxsci/kfw231
Banu. SK: Stanley. TA: Taylor. RT: Sivakumar. KK: Arosh. TA: Zeng. L: Pennathur. S: Padmanabhan. V.
(2018). Sexually Dimorphic Impact of Chromium Accumulation on Human Placental
Oxidative Stress and Apoptosis. Toxicol Sci 161: 375-387.
http://dx.doi.org/10.1093/toxsci/kfx224
Baresic. M: Gornik. I: Radonic. R: Zlopasa. 0: Gubarev. N: Gasparovic. V. (2009). Survival after severe
acute chromic acid poisoning complicated with renal and liver failure. Intern Med 48: 711-
715.
Barnhart. I. (1997). Chromium chemistry and implications for environmental fate and toxicity.
Journal of Soil Contamination 6: 561-568. http: //dx.doi.org/10.1080 /15320389709383589
Basak. 0: van de Born. M: Korving. 1: Beumer. I: van der Elst. S: van Es. TH: Clevers. H. (2014).
Mapping early fate determination in Lgr5+ crypt stem cells using a novel Ki67-RFP allele.
EMBO J 33: 2057-2068. http://dx.doi.org/10.15252/embi.201488017
Bashandv. SAE: Ebaid. H: Al-Tamimi. I: Ahmed-Farid. OA: Omara. EA: Alhazza. IM. (2021). Melatonin
Alleviated Potassium Dichromate-Induced Oxidative Stress and Reprotoxicity in Male Rats.
BioMedReslnt2021: 3565360. http://dx.doi.org/10.1155/2021/3565360
Basketter. DA: Scholes. EW: Kimber. I. (1994). The performance of the local lymph node assay with
chemicals identified as contact allergens in the human maximization test Food Chem
Toxicol 32: 543-547.
Bataineh. H: Al-Hamood. MH: Elbetieha. A: Bani Hani. I. (1997). Effect of long-term ingestion of
chromium compounds on aggression, sex behavior and fertility in adult male rat. Drug
Chem Toxicol 20: 133-149. http://dx.doi.org/10.3109/01480549709003875
This document is a draft for review purposes only and does not constitute Agency policy.
R-5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Bataineh. HN: Bataineh. ZM: Daradka. H. (2007). Short-term exposure of female rats to industrial
metal salts: Effect on implantation and pregnancy. Reproductive Medicine and Biology 6:
179-183. http://dx.doi.Org/10.llll/i.1447-0578.2007.00183.x
Beaumont. 11: Sedman. RM: Reynolds. SD: Sherman. CD: Li. LH: Howd. RA: Sandy. MS: Zeise. L:
Alexeeff. GV. (2008). Cancer mortality in a Chinese population exposed to hexavalent
chromium in drinking water. Epidemiology 19: 12-23.
http://dx.doi.org/10.1097/EDE.0b013e31815cea4c
Bedinger. GM. (2015). 2014 Minerals yearbook. Arsenic [advance release], Reston, VA: U.S.
Geological Survey.
Bednar. CM: Kies. C. (1991). Inorganic contaminants in drinking water correlated with disease
occurrence in Nebraska. J Am Water Resour Assoc 27: 631-635.
http://dx.doi.Org/10.llll/i.1752-1688.1991.tb01464.x
Behari. 1: Chandra. SV: Tandon. SK. (1978). Comparative toxicity of trivalent and hexavalent
chromium to rabbits III Biochemical and histological changes in testicular tissue. Biomed
Biochim Acta 37: 463-468.
Ben-David. U: Amon. A. (2020). Context is everything: aneuploidy in cancer [Review], Nat Rev Genet
21: 44-62. http://dx.doi.org/l 0.1038/s41576-019-0171 -x
Benova. D: Hadiidekova. V: Hristova. R: Nikolova. T: Boulanova. M: Georgieva. I: Grigorova. M:
Popov. T: Panev. T: Georgieva. R: Nataraian. AT: Darroudi. F: Nilsson. R. (2002). Cytogenetic
effects of hexavalent chromium in Bulgarian chromium platers. Mutat Res 514: 29-38.
Bessede. E: Dubus. P: Megraud. F: Varon. C. (2015). Helicobacter pylori infection and stem cells at
the origin of gastric cancer [Review], Oncogene 34: 2547-2555.
http://dx.doi.org/10.1038/onc.2014.187
Betton. GR. (2013). A review of the toxicology and pathology of the gastrointestinal tract [Review],
Cell Biol Toxicol 29: 321-338. http://dx.doi.org/10.1007/sl0565-013-9257-v
Bick. RL: Girardi. TV: Lack. WT: Costa. M: Titelbaum. D. (1996). Hodgkin's disease in association with
hexavalent chromium exposure. IntJ Hematol 64: 257-262.
Birk. T: Mundt. KA: Dell. LP: Luippold. RS: Miksche. L: Steinmann-Steiner-Haldenstaett. W: Mundt.
DT. (2006). Lung cancer mortality in the German chromate industry, 1958 to 1998. J Occup
Environ Med 48: 426-433. http://dx.doi.org/10.1097/01 .iom.OOOOl94159.88688.f8
Bitterman. PB: Saltzman. LE: Adelberg. S: Ferrans. VI: Crystal. RG. (1984). ALVEOLAR
MACROPHAGE REPLICATION - ONE MECHANISM FOR THE EXPANSION OF THE
MONONUCLEAR PHAGOCYTE POPULATION IN THE CHRONICALLY INFLAMED LUNG. J Clin
Invest 74: 460-469. http://dx.doi.org/10.1172/ICI111443
Blaser. H: Dostert. C: Mak. TW: Brenner. D. (2016). TNF and ROS Crosstalk in Inflammation
[Review], Trends Cell Biol 26: 249-261. http://dx.doi.Org/10.1016/i.tcb.2015.12.002
Bloomfield. II: Blum. W. (1928). Health hazards in chromium plating. Public Health Rep 43: 527-
528. http://dx.doi.org/10.2307/4578958
Boland. CR: Goel. A. (2010). Microsatellite instability in colorectal cancer [Review],
Gastroenterology 138: 2073-2087.e2073. http://dx.doi.Org/10.1053/i.gastro.2009.12.064
Bonassi. S: Coskun. E: Ceppi. M: Lando. C: Bolognesi. C: Burgaz. S: Holland. N: Kirsh-Volders. M:
Knasmueller. S: Zeiger. E: Carnesoltas. D: Cavallo. D: da Silva. I: de Andrade. VM: Demircigil.
GC: Dominguez Odio. A: Donmez-Altuntas. H: Gattas. G: Giri. A: Giri. S: Gomez-Meda. B:
Gomez-Arrovo. S: Hadiidekova. V: Haveric. A: Kamboi. M: Kurteshi. K: Martino-Roth. MG:
Montero Montova. R: Nersesvan. A: Pastor-Benito. S: Favero Salvadori. DM: Shaposhnikova.
A: Stopper. H: Thomas. P: Torres-Bugarin. 0: Yadav. AS: Zuniga Gonzalez. G: Fenech. M.
(2011a). The HUman MicroNucleus project on eXfoLiated buccal cells (HUMN(XL)): the role
of life-style, host factors, occupational exposures, health status, and assay protocol
[Review], Mutat Res 728: 88-97. http: //dx.doi.Org/10.1016/i.mrrev.2011.06.005
This document is a draft for review purposes only and does not constitute Agency policy.
R-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Bonassi. S: El-Zein. R: Bolognesi. C: Fenech. M. (2011b). Micronuclei frequency in peripheral blood
lymphocytes and cancer risk: evidence from human studies [Review], Mutagenesis 26: 93-
100. http://dx.doi.org/10.1093/mutage/geq075
Bonassi. S: Norppa. H: Ceppi. M: Stromberg. U: Vermeulen. R: Znaor. A: Cebulska-Wasilewska. A:
Fabianova. E: Fucic. A: Gundy. S: Hansteen. IL: Knudsen. LE: Lazutka. 1: Rossner. P: Sram. RT:
Boffetta. P. (2008). Chromosomal aberration frequency in lymphocytes predicts the risk of
cancer: results from a pooled cohort study of 22 358 subjects inll countries.
Carcinogenesis 29: 1178-1183. http://dx.doi.org/10.1093/carcin/bgn075
Bonassi. S: Znaor. A: Ceppi. M: Lando. C: Chang. WP: Holland. N: Kirsch-Volders. M: Zeiger. E: Ban. S:
Barale. R: Bigatti. MP: Bolognesi. C: Cebulska-Wasilewska. A: Fabianova. E: Fucic. A: Hagmar.
L: Toksic. G: Martelli. A: Migliore. L: Mirkova. E: Scarfi. MR: Ziino. A: Norppa. H: Fenech. M.
(2007). An increased micronucleus frequency in peripheral blood lymphocytes predicts the
risk of cancer in humans. Carcinogenesis 28: 625-631.
http://dx.doi.org/10.1093/carcin/bgll77
Bonde. TP. (1990). Semen quality and sex hormones among mild steel and stainless steel welders: A
cross sectional study. Br J Ind Med 47: 508-514.
Bonde. TP: Ernst. E. (1992). Sex hormones and semen quality in welders exposed to hexavalent
chromium. Hum Exp Toxicol 11: 259-263.
Boone. L: Meyer. D: Cusick. P: Ennulat. D: Bolliger. AP: Everds. N: Meador. V: Elliott. G: Honor. D:
Bounous. D: Tordan. H. (2005). Selection and interpretation of clinical pathology indicators
of hepatic injury in preclinical studies [Review], Vet Clin Pathol 34: 182-188.
http://dx.doi. org/10.1111 /i.l 939-165X.2005.tb00041 ,x
Boorman. G: Dixon. D: Elwell. M: Kerlin. R: Morton. D: Peters. T: Regan. K: Sullivan. 1: Group. SHW.
(2003). Assessment of hyperplastic lesions in rodent carcinogenicity studies. Toxicol Pathol
31: 709-710.
Borella. P: Bargellini. A. (1993). Effects of trace elements on immune system: results in cultured
human lymphocytes. J Trace Elem Med Biol 7: 231-233.
Borges. KM: Boswell. IS: Liebross. RH: Wetterhahn. KE. (1991). Activation of chromium(VI) by thiols
results in chromium(V) formation, chromium binding to DNA and altered DNA
conformation. Carcinogenesis 12: 551-561. http://dx.doi.Org/10.1093/carcin/12.4.551
Borneff. 1: Engelhardt. K: Griem. W: Kunte. H: Reichert. 1. (1968). Kanzerogene substanzen in wasser
und boden. XXII. Mausentrankversuch mit 3,4-benzpyren und kaliumchromat.
[Kanzerogene substanzen in wasser und boden. XXII. Mausentrankversuch mit 3,4-
benzpyren und kaliumchromat], Archiv fuer Hygiene und Bakteriologie 152: 45-53.
Boscolo. P: Pi Gioacchino. M: Bavazzano. P: White. M: Sabbioni. E. (1997). Effects of chromium on
lymphocyte subsets and immunoglobulins from normal population and exposed workers.
Life Sci 60: 1319-1325. http://dx.doi.org/10.1016/S0024-3205r96100669-8
Bovet. P: Lob. M: Grandiean. M. (1977). Spirometric alterations in workers in the chromium
electroplating industry. Int Arch Occup Environ Health 40: 25-32.
http://dx.doi. org/10.1007/BF00435514
Brandt-Rauf. P. (2006). Editorial retraction. Cancer mortality in a Chinese population exposed to
hexavalent chromium in water. J Occup Environ Med 48: 749.
http://dx.d0i.0rg/l 0.1097/00043764-200607000-00019
Braver. ER: Infante. P: Chu. K. (1985). An analysis of lung cancer risk from exposure to hexavalent
chromium. Birth Defects Res B Dev Reprod Toxicol 5: 365-378.
Bridgewater. LC: Manning. FC: Patierno. SR. (1994). Base-specific arrest of in vitro DNA replication
by carcinogenic chromium: relationship to DNA interstrand crosslinking. Carcinogenesis 15:
2421-2427. http://dx.doi.org/10.1093/carcin/15.ll.2421
Bright. P: Burge. PS: O'Hickev. SP: Gannon. PF: Robertson. AS: Boran. A. (1997). Occupational
asthma due to chrome and nickel electroplating. Thorax 52: 28-32.
This document is a draft for review purposes only and does not constitute Agency policy.
R-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Brilliant. MH: Gondo. Y: Eicher. EM. (1991). Direct molecular identification of the mouse pink-eyed
unstable mutation by genome scanning. Science 252: 566-569.
http://dx.doi.org/10.1126/science.1673574
Brito. F: Ascanio. 1: Mateo. S: Hernandez. C: Arauio. L: Gili. P: Martin-Zarza. P: Dominguez. S:
Mederos. A. (1997). Equilibria of chromate(VI) species in acid medium and ab initio studies
of these species. Polyhedron 16: 3835-3846. http://dx.doi.org/10.1016/S0277-
5387(~97")00128-9
Browning. CL: Oin. 0: Kelly. DF: Prakash. R: Vanoli. F: Tasin. M: Wise. TP. (2016). Prolonged
Particulate Hexavalent Chromium Exposure Suppresses Homologous Recombination Repair
in Human Lung Cells. Toxicol Sci 153: 70-78. http://dx.doi.org/10.1093/toxsci/kfwl03
Browning. CL: Wise. TP. (2017). Prolonged exposure to particulate chromate inhibits RAD51 nuclear
import mediator proteins. Toxicol Appl Pharmacol 331: 101-107.
http://dx.doi.Org/10.1016/j.taap.2017.05.030
Browning. 1: Gannon. BT: O'Brien. P. (1983). The microvasculature and gastric luminal pH of the
forestomach of the rat: A comparison with the glandular stomach. Int J Microcirc Clin Exp 2:
109-118.
Bryant. HE: Ying. S: Helledav. T. (2006). Homologous recombination is involved in repair of
chromium-induced DNA damage in mammalian cells. MutatRes 599: 116-123.
http://dx.doi.0rg/lO.lOl6/i.mrfmmm.2OO6.O2.OOl
Buerge. IT: Hug. ST. (1997). Kinetics and pH dependence of chromium(VI) reduction by iron(II).
Environ SciTechnol 31: 1426-1432.
Burdsall. DP: Flores. HC: Krueger. 1: Garretson. S: Gorbien. Ml: Iacch. A: Dobbs. V: Homa. T. (2013).
Use of proton pump inhibitors with lack of diagnostic indications in 22 Midwestern US
skilled nursing facilities. J Am Med Dir Assoc 14: 429-432.
http://dx.doi.0rg/lO.lOl6/i.iamda.2Oi3.Ol.O2i
Burke. T: Fagliano. 1: Goldoft. M: Hazen. RE: Iglewicz. R: Mckee. T. (1991). Chromite ore processing
residue in Hudson County, New Jersey. Environ Health Perspect 92: 131-137.
Burton. ED: Choppala. G: Karimian. N: Tohnston. SG. (2019). A new pathway for hexavalent
chromium formation in soil: Fire-induced alteration of iron oxides. Environ Pollut 247: 618-
625. http://dx.doi.Org/10.1016/i.envpol.2019.01.094
Caglieri. A: Goldoni. M: De Palma. G: Mozzoni. P: Gemma. S: Vichi. S: Testai. E: Panico. F: Corradi. M:
Tagliaferri. S: Costa. LG. (2008). Exposure to low levels of hexavalent chromium: target
doses and comparative effects on two human pulmonary cell lines. Acta Biomed 79 Suppl 1:
104-115.
Calder. L. (1988). Chromium contamination of groundwater. In JO Nriagu; E Nieboer (Eds.),
Chromium in the natural and human environments (pp. 215-229). New York, NY: Wiley.
https://www.wiley.com/en-us/Chromium+in+the+Natural+and+Human+Environments+-
p-9780471856436
CalEPA (California Environmental Protection Agency). (2003). Ambient air monitoring for
hexavalent chromium and metals in Barrio Logan: May 2001 through May 2002. (Report:
2003-10-14). Sacramento, CA.
CalEPA (California Environmental Protection Agency). (2004). Barrio Logan report: A compilation
of air quality studies in Barrio Logan. November 2004. (Report: 2005-11-00). Sacramento,
CA.
CalEPA (California Environmental Protection Agency). (2011). Public health goal for Hexavalent
Chromium (Cr VI) in drinking water. Sacramento, CA: Pesticide and Environmental
Toxicology Branch, Office of Environmental Health Hazard Assessment
http://oehha.ca.gov/media/downloads/water/public-health-goal/cr6phg072911.pdf
CARB (California Air Resources Board). (2004). Neighborhood scale monitoring in Barrio Logan.
(Paper 70255). Sacramento, CA.
This document is a draft for review purposes only and does not constitute Agency policy.
R-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
Carette. D: Perrard. MH: Prisant. N: Gilleron. 1: Pointis. G: Segretain. D: Durand. P. (2013).
Hexavalent chromium at low concentration alters Sertoli cell barrier and connexin 43 gap
junction but not claudin-11 and N-cadherin in the rat seminiferous tubule culture model.
Toxicol Appl Pharmacol 268: 27-36. http://dx.doi.Org/10.1016/i.taap.2013.01.016
Carlisle. PL: Pritchard. DE: Singh. 1: Owens. BM: Blankenship. LI: Orenstein. TM: Patierno. SR. (2000).
Apoptosis and P53 induction in human lung fibroblasts exposed to chromium (VI): Effect of
ascorbate and tocopherol. Toxicol Sci 55: 60-68. http ://dx.doi.org/10.1093/toxsci/55.1.60
Carney. EW: Kimmel. CA. (2007). Interpretation of skeletal variations for human risk assessment:
Delayed ossification and wavy ribs. Birth Defects Res B Dev Reprod Toxicol 80: 473-496.
http://dx.doi.org/10.1002/bdrb.20133
Casadevall. M: da Cruz Fresco. P: Kortenkamp. A. (1999). Chromium(VI)-mediated DNA damage:
oxidative pathways resulting in the formation of DNA breaks and abasic sites. Chem Biol
Interact 123: 117-132.
Castelevn. C: Rekecki. A: Van der Aa. A: Simoens. P: Van den Broeck. W. (2010). Surface area
assessment of the murine intestinal tract as a prerequisite for oral dose translation from
mouse to man. Lab Anim 44: 176-183. http://dx.doi.org/10.1258/la.2009.009112
Ceballos. D: West. C: Methner. M: Gong. W. (2017). Health hazard evaluation report: HHE-2013-
0011-3278, May 2017. Evaluation of chromium, hexavalent chromium, cadmium, and
isocyanate exposures in an African re finishing plant. (HHE-2 013-0011-3 278). National
Institute for Occupational Safety and Health.
https://ntrl.ntis.gOv/NTRL/dashboard/searchResults/titleDetail/PB2017102698.xhtml
Ceballos. D: West. C: Methner. M: Gong. W. (2019). Hexavalent Chromium Exposure and Nasal
Tissue Effects at a Commercial Aircraft Refinishing Facility. J Occup Environ Med 61: e69-
e73. http://dx.doi.Org/10.1097/TOM.0000000000001510
CGARN (Cancer Genome Atlas Research Network). (2014). Comprehensive molecular
characterization of gastric adenocarcinoma. Nature 513: 202-209.
http://dx.doi.org/10.1038/naturel348Q
CGARN (Cancer Genome Atlas Research Network). (2018a). Comparative molecular analysis of
gastrointestinal adenocarcinomas. Cancer Cell 33: 721-735.
http://dx.doi.Org/10.1016/i.ccell.2018.03.010
CGARN (Cancer Genome Atlas Research Network). (2018b). The immune landscape of cancer.
Immunity 48: 812-830.e814. http://dx.doi.Org/10.1016/i.immuni.2018.03.023
Chandra. SA: Nolan. MW: Malarkev. DE. (2010). Chemical carcinogenesis of the gastrointestinal
tract in rodents: an overview with emphasis on NTP carcinogenesis bioassays [Review],
Toxicol Pathol 38: 188-197. http://dx.doi.org/10.1177/01926233Q9356452
Chappell. GA: Wolf. TC: Thompson. CM. (2022). Crypt and villus transcriptomic responses in mouse
small intestine following oral exposure to hexavalent chromium. Toxicol Sci 186: 43-57.
http: / /dx. do i. or g /10.109 3 /toxsci/kfab 152
Chatteriee. N: Walker. GC. (2017). Mechanisms of DNA damage, repair, and mutagenesis [Review],
Environ Mol Mutagen 58: 235-263. http://dx.doi.org/10.1002/em.22Q87
Chebeir. M: Liu. H. (2016). Kinetics and mechanisms of Cr(VI) formation via the oxidation of Cr(III)
solid phases by chlorine in drinking water. Environ Sci Technol 50: 701-710.
http://dx.doi.org/10.1021/acs.est.5b05739
Chen. 0: Li. H: Chen. W: Yao. W: Li. S: Xu. Y: Meng. A. (1997). [Impact of chromium compounds on
female reproductive function], Gongye Weishengyu Zhiyebin 23: 180-182.
Chen. OY: Murphy. A: Sun. H: Costa. M. (2019). Molecular and epigenetic mechanisms of Cr(VI)-
induced carcinogenesis [Review], Toxicol Appl Pharmacol 377: 114636.
http://dx.doi.Org/10.1016/i.taap.2019.114636
This document is a draft for review purposes only and does not constitute Agency policy.
R-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Cheng. L: Liu. S: Dixon. K. (1998). Analysis of repair and mutagenesis of chromium-induced DNA
damage in yeast, mammalian cells, and transgenic mice. Environ Health Perspect 106: 1027-
1032.
Cheng. L: Sonntag. DM: de Boer. 1: Dixon. K. (2000). Chromium(VI)-induced mutagenesis in the
lungs of big blue transgenic mice. J Environ Pathol Toxicol Oncol 19: 239-249.
Chiba. T: Marusawa. H: Ushiiima. T. (2012). Inflammation-associated cancer development in
digestive organs: mechanisms and roles for genetic and epigenetic modulation [Review],
Gastroenterology 143: 550-563. http://dx.doi.Org/10.1053/i.gastro.2012.07.009
Chopra. A: Pereira. G: Gomes. T: Pereira. 1: Prabhu. P: Krishnan. S: Rao. CV. (1996). A study of
chromium and ethanol toxicity in female Wistar rats. Toxicol Environ Chem 53: 91-106.
http://dx.doi. org/10.1080/02772249609358274
Chowdhurv. AR: Mitra. C. (1995). Spermatogenic and steroidogenic impairment after chromium
treatment in rats. Indian J Exp Biol 33: 480-484.
Christensen. MM: Ernst. E: Ellermann-Eriksen. S. (1992). Cytotoxic effects of hexavalent chromium
in cultured murine macrophages. Arch Toxicol 66: 347-353.
Christensen. OB: Christensen. MB: Maibach. HI. (1984). Flare-up reactions and desensitization from
oral dosing in chromate-sensitive guinea pigs. Contact Derm 10: 277-279.
Christiansen. PM. (1968). The incidence of achlorhydria and hypochlorhydria in healthy subjects
and patients with gastrointestinal diseases. Scand J Gastroenterol 3: 497-508.
http://dx.doi.Org/10.3109/00365526809179909
Cid. MG: Loria. D: Vilensky. M: Miotti. TL: Matos. E. (1991). Leather tanning workers: chromosomal
aberrations in peripheral lymphocytes and micronuclei in exfoliated cells in urine. Mutat
Res 259: 197-201. http://dx.doi.org/10.1016/0165-1218r91190053-0
Cieslak-Golonka. M. (1996). Toxic and mutagenic effects of chromium(VI). A review. Polyhedron 15:
3667-3689. http://dx.doi.org/10.1016/0277-5387r96100141-6
Ciminera. P: Superior. MI: Bullman. T. (2016). Findings from the Department of Veterans Affairs
Qarmat Ali Medical Surveillance Program. Mil Med 181: 307-310.
http://dx.doi.org/10.7205/MTLMED-D-15-00278
Clementino. M: Kim. D: Zhang. Z. (2019). Constitutive Activation of NAD-Dependent Sirtuin 3 Plays
an Important Role in Tumorigenesis of Chromium(VI)-Transformed Cells. Toxicol Sci 169:
224-234. http://dx.doi.org/10.1093/toxsci/kfz032
Coelho. P: Garcia-Leston. I: Costa. S: Costa. C: Silva. S: Dall'Armi. V: Zoffoli. R: Bonassi. S: de Lima. TP:
Gaspar. IF: Pasaro. E: Laffon. B: Teixeira. TP. (2013). Genotoxic effect of exposure to
metal(loid)s. A molecular epidemiology survey of populations living and working in
Panasqueira mine area, Portugal. Environ Int 60: 163-170.
http://dx.doi.Org/10.1016/i.envint.2013.08.014
Coffing. S: Engel. M: Dickinson. D: Thiffeault. C: Spellman. R: Shutskv. T: Schuler. M. (2011). The rat
gut micronucleus assay: a good choice for alternative in vivo genetic toxicology testing
strategies. Environ Mol Mutagen 52: 269-279. http://dx.doi.org/10.1002/em.20616
Cohen. MP: Prophete. C: Sisco. M: Chen. LC: Zelikoff. IT: Smee. IT: Holder. AA: Crans. DC. (2006).
Pulmonary immunotoxic potentials of metals are governed by select physicochemical
properties: chromium agents. J Immunotoxicol 3: 69-81.
http://dx.doi.org/10.1080/15476910600718434
Cohen. MP: Sisco. M: Baker. K: Bowser. D: Chen. LC: Schlesinger. RB. r20031. Impact of coexposure
to ozone on the carcinogenic potential of inhaled chromium. 1. Effects on retention and on
extra- and intracellular distribution. J Toxicol Environ Health A 66: 39-55.
http://dx.doi.org/10.1080/15287390306459
Cohen. MP: Sisco. M: Prophete. C: Yoshida. K: Chen. LC: Zelikoff. IT: Smee. 1: Holder. AA:
Stonehuerner. 1: Crans. PC: Ghio. AT. (2010). Effects of metal compounds with distinct
physicochemical properties on iron homeostasis and antibacterial activity in the lungs:
This document is a draft for review purposes only and does not constitute Agency policy.
R-10 PRAFT-PO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Chromium and vanadium. Inhal Toxicol 22: 169-178.
http://dx.doi.Org/10.3109/08958370903161232
Cohen. MP: Zelikoff. TT: Chen. LC: Schlesinger. RB. (1998). Immunotoxicologic effects of inhaled
chromium: role of particle solubility and co-exposure to ozone. Toxicol Appl Pharmacol
152: 30-40. http://dx.doi.org/10.1006/taap.1998.8502
Cohen. RE: Neiders. ME: Bedi. GS: Comeau. R. (1993). Induction of type 2 cystatin in rat
submandibular glands by systemically administered agents. Arch Oral Biol 38: 319-325.
Cohen. SR: Davis. DM: Kramkowski. RS. (1974). Clinical manifestations of chromic acid toxicity:
Nasal lesions in electroplate workers. Cutis 13: 558-568.
Cohen. SR: Kramkowski. RS. (1973). Health hazard evaluation/toxicity determination. Marsh
Plating Corporation, Ypsilanti, Michigan. (HHE-72-118-104). Cincinnati, OH: National
Institute for Occupational Safety and Health.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?search0uery=PB232740
Cole. P: Rodu. B. (2005). Epidemiologic studies of chrome and cancer mortality: a series of meta-
analyses. Regul Toxicol Pharmacol 43: 225-231.
http://dx.doi.Org/10.1016/i.yrtph.2005.06.009
Collins. BT: Stout. MP: Levine. KE: Kissling. GE: Melnick. RL: Fennell. TR: Walden. R: Abdo. K:
Pritchard. IB: Fernando. RA: Burka. LT: Hooth. MI. (2010). Exposure to hexavalent
chromium resulted in significantly higher tissue chromium burden compared with trivalent
chromium following similar oral doses to male F344/N rats and female B6C3F1 mice.
Toxicol Sci 118: 368-379. http://dx.doi.org/10.1093/toxsci/kfq263
Cotton. FA: Wilkinson. G: Murillo. CA: Bochmann. M. (1999). The elements of the first transition
series. In Advanced Inorganic Chemistry. New York: Wiley.
Cox. DR. (1972). Regression models and life-tables. J R Stat Soc Ser B 34: 187-220.
Creasy. DM: Chapin. RE. (2018). Chapter 17 - Male reproductive system. In Fundamentals of
Toxicologic Pathology (Third Edition). London, UK: Academic Press of Elsevier.
http://dx.doi.Org/10.1016/B978-0-12-809841-7.00017-4
Crew. KD: Neugut. AI. (2004). Epidemiology of upper gastrointestinal malignancies. Semin Oncol
31: 450-464. http://dx.doi.Org/10.1053/i.seminoncol.2004.04.021
Crissman. TW: Goodman. DG: Hildebrandt. PK: Maronpot. RR: Prater. DA: Riley. TH: Seaman. WT:
Thake. DC. (2004). Best practices guideline: Toxicologic histopathology. Toxicol Pathol 32:
126-131. http://dx.doi.org/10.1080/01926230490268756
Cullen. TM: Ward. TM: Thompson. CM. (2015). Reevaluation and Classification of Duodenal Lesions
in B6C3F1 Mice and F344 Rats from 4 Studies ofHexavalent Chromium in Drinking Water.
Toxicol Pathol 44: 279-289. http://dx.doi.org/10.1177/Q192623315611501
Cummings. DE: Fendorf. S: Singh. N: Sani. RK: Peyton. BM: Magnuson. TS. (2007). Reduction of
Cr(VI) under acidic conditions by the facultative Fe(lll)-reducing bacterium Acidiphilium
cryptum. Environ Sci Technol 41: 146-152. http://dx.doi.org/10.1021/esQ61333k
D'Agostini. F: Izzotti. A: Bennicelli. C: Camoirano. A: Tampa. E: De Flora. S. (2002). Induction of
apoptosis in the lung but not in the liver of rats receiving intra-tracheal instillations of
chromium(VI). Carcinogenesis 23: 587-593. http://dx.doi.Org/10.1093/carcin/23.4.587
Dai. L: Xu. W: Li. H: Frank. TA: He. C: Zhang. Z: Chen. G. (2017). Effects of hexavalent chromium on
mouse splenic T lymphocytes. Toxicol In Vitro 45: 166-171.
http: / /dx.doi. org/10.1016/i .tiv. 2017.09.006
Danadevi. K: Rozati. R: Banu. BS: Grover. P. (2004). Genotoxic evaluation of welders occupationally
exposed to chromium and nickel using the Comet and micronucleus assays. Mutagenesis 19:
35-41. http://dx.doi.org/10.1093/mutage/geh001
Danadevi. K: Rozati. R: Reddv. PP: Grover. P. (2003). Semen quality of Indian welders
occupationally exposed to nickel and chromium. Reprod Toxicol 17: 451-456.
This document is a draft for review purposes only and does not constitute Agency policy.
R-ll DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
Das. I: Kang. MH: Kim. E: Kwon. DN: Choi. YT: Kim. TH. (2015). Hexavalent chromium induces
apoptosis in male somatic and spermatogonial stem cells via redox imbalance. Sci Rep 5:
13921. http://dx.doi.org/10.1038/srepl3921
Davidson. T: Kluz. T: Burns. F: Rossman. T: Zhang. 0: Uddin. A: Nadas. A: Costa. M. (2004). Exposure
to chromium (VI) in the drinking water increases susceptibility to UV-induced skin tumors
in hairless mice. Toxicol Appl Pharmacol 196: 431-437.
http://dx.doi.Org/10.1016/i.taap.2004.01.006
Day. CP: Tames. OF. (1998). Hepatic steatosis: Innocent bystander or guilty party? [Review],
Hepatology 27: 1463-1466. http://dx.doi.org/10.1002/hep.510270601
De Flora. S: Badolati. GS: Serra. D: Picciotto. A: Magnolia. MR: Savarino. V. (1987). Circadian
reduction of chromium in the gastric environment. MutatRes Lettl92: 169-174.
http://dx.doi.org/10.1016/0165-7992r87190051-0
De Flora. S: Camoirano. A: Bagnasco. M: Bennicelli. C: Corbett. GE: Kerger. BP. (1997). Estimates of
the chromium(VI) reducing capacity in human body compartments as a mechanism for
attenuating its potential toxicity and carcinogenicity. Carcinogenesis 18: 531-537.
De Flora. S: Camoirano. A: Micale. RT: La Maestra. S: Savarino. V: Zentilin. P: Marabotto. E: Suh. M:
Proctor. DM. (2016). Reduction of hexavalent chromium by fasted and fed human gastric
fluid. I. Chemical reduction and mitigation of mutagenicity. Toxicol Appl Pharmacol 306:
113-119. http://dx.doi.Org/10.1016/i.taap.2016.07.004
De Flora. S: D'Agostini. F: Balanskv. R: Micale. R: Baluce. B: Izzotti. A. (2008). Lack of genotoxic
effects in hematopoietic and gastrointestinal cells of mice receiving chromium(VI) with the
drinking water [Review], MutatRes 659: 60-67.
http://dx.doi.Org/10.1016/i.mrrev.2007.ll.005
De Flora. S: Iltcheva. M: Balanskv. RM. (2006). Oral chromium(VI) does not affect the frequency of
micronuclei in hematopoietic cells of adult mice and of transplacental^ exposed fetuses.
MutatRes 610: 38-47. http://dx.doi.Org/10.1016/i.mrgentox.2006.06.011
De Mattia. G: Bravi. MC: Laurenti. 0: De Luca. 0: Palmeri. A: Sabatucci. A: Mendico. G: Ghiselli. A.
(2004). Impairment of cell and plasma redox state in subjects professionally exposed to
chromium. Am J Ind Med 46: 120-125. http://dx.doi.org/10.1002/aiim.20044
Delshad. SD: Almario. CV: Chev. WD: Spiegel. BMR. (2020). Pevalence of gastroesophageal reflux
disease and proton pump inhibitor-refractory symptoms. Gastroenterology 158: 1250.
http://dx.doi.Org/10.1053/i.gastro.2019.12.014
Deng. C: Lee. HH: Xian. H: Yao. M: Huang. I: Ou. B. (1988). Chromosomal aberrations and sister
chromatid exchanges of peripheral blood lymphocytes in Chinese electroplating workers:
Effectof nickel and chromium. J Trace Elem Exp Med 1: 57-62.
Deng. G: Wu. K: Cruce. AA: Bowman. MK: Vincent. IB. (2015). Binding of trivalent chromium to
serum transferrin is sufficiently rapid to be physiologically relevant. J Inorg Biochem 143:
48-55. http://dx.doi.Org/10.1016/i.iinorgbio.2014.12.004
Deng. Y: Wang. M: Tian. T: Lin. S: Xu. P: Zhou. L: Dai. C: Hao. 0: Wu. Y: Zhai. Z: Zhu. Y: Zhuang. G: Dai.
Z. (2019). The Effectof Hexavalent Chromium on the Incidence and Mortality of Human
Cancers: A Meta-Analysis Based on Published Epidemiological Cohort Studies. 9: 24.
http://dx.doi.Org/10.3389/fonc.2019.00024
Dertinger. SD: Torous. DK: Havashi. M: Macgregor. IT. (2011). Flow cytometric scoring of
micronucleated erythrocytes: an efficient platform for assessing in vivo cytogenetic damage
[Review], Mutagenesis 26: 139-145. http://dx.doi.org/10.1093/mutage/geqO55
Devov. 1: Gehin. A: Miiller. S: Melczer. M: Remv. A: Antoine. G: Sponne. I. (2016). Evaluation of
chromium in red blood cells as an indicator of exposure to hexavalent chromium: An in
vitro study. Toxicol Lett 255: 63-70. http://dx.doi.Org/10.1016/i.toxlet2016.05.008
This document is a draft for review purposes only and does not constitute Agency policy.
R-12 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Dobrovolskv. VN: Heflich. RH. (2018). Chapter eleven: Detecting mutations in vivo. In Mutagenicity:
Assays and applications. Cambridge, MA: Academic Press. http://dx.doi.org/10.1016/B978-
0-12-809252-1.00011-0
Doki. K: Darwich. AS: Patel. N: Rostami-Hodiegan. A. (2017). Virtual bioequivalence for achlorhydric
subjects: The use of PBPK modelling to assess the formulation-dependent effect of
achlorhydria. Eur J Pharm Sci 109. http://dx.doi.Org/10.1016/i.eips.2017.07.035
Dornan. ID. (1981). Occupational dermatoses amongst chrome platers in the Sheffield area 1977—
80. Contact Derm 7: 354-355. http://dx.doi.Org/10.llll/i.1600-0536.1981.tb04116.x
Pressman. IB: Berardi. RR: Dermentzoglou. LC: Russell. TL: Schmaltz. SP: Barnett. TL: Tarvenpaa. KM.
(1990). Upper gastrointestinal (GI) pH in young, healthy men and women. Pharm Res 7:
756-761.
Eastmond. DA: Hartwig. A: Anderson. D: Anwar. WA: Cimino. MC: Dobrev. I: Douglas. GR: Nohmi. T:
Phillips. DH: Vickers. C. (2009). Mutagenicity testing for chemical risk assessment: Update of
the WHO/IPCS harmonized scheme. Mutagenesis 24: 341-349.
http://dx.doi.org/10.1093/mutage/gep014
Eastmond. DA: Macgregor. IT: Slesinski. RS. (2008). Trivalent chromium: Assessing the genotoxic
risk of an essential trace element and widely used human and animal nutritional
supplement Crit Rev Toxicol 38: 173-190.
http://dx.doi.Org/10.1080/10408440701845401
Eferl. R: Wagner. EF. (2003). AP-1: a double-edged sword in tumorigenesis [Review], Nat Rev
Cancer 3: 859-868. http://dx.doi.Org/l0.1038/nrcl209
EFSA CONTAM Panel (EFSA Panel on Contaminants in the Food Chain). (2014). Scientific Opinion
on the risks to public health related to the presence of chromium in food and drinking
water. EFSA J 12: 3595. http: //dx.doi.Org/10.2903 /i.efsa.2014.3595
Egger. M: Smith. GD: Schneider. M: Minder. C. (1997). Bias in meta-analysis detected by a simple,
graphical test. Br Med J 315: 629-634. http://dx.doi.org/10.1136/bmi.315.71Q9.629
Eizaguirre-Garcia. D: Rodriguez-Andres. C: Watt. G. (2000). Congenital anomalies in Glasgow
between 1982 and 1989 and chromium waste. J Public Health (Oxf) 22: 54-58.
http://dx.doi.0rg/lO.lO93/pubmed/22.l.54
El-Demerdash. FM: Tebur. AB: Nasr. HM: Hamid. HM. (2019). Modulatory effect of Turnera diffusa
against testicular toxicity induced by fenitrothion and/or hexavalent chromium in rats.
Environ Toxicol 34: 330-339. http://dx.doi.org/10.1002/tox.22688
El-Demerdash. FM: Yousef. MI: Elaswad. FA. (2006). Biochemical study on the protective role of
folic acid in rabbits treated with chromium (VI). J Environ Sci Health B 41: 731-746.
http://dx.d0i.0rg/l 0.1080/03601230600704282
El Safty. AMK: Samir. AM: Mekkawv. MK: Fouad. MM. (2018). Genotoxic effects due to exposure to
chromium and nickel among electroplating workers. Int J Toxicol 37: 234-240.
http://dx.doi.org/10.1177/1091581818764084
Elbetieha. A: Al-Hamood. MH. (1997). Long-term exposure of male and female mice to trivalent and
hexavalent chromium compounds: effect on fertility. Toxicology 116: 39-47.
http://dx.d0i.0rg/l 0.1016/S0300-483XC96103516-0
Elhosarv. N: Maklad. A: Soliman. E: El-Ashmawv. N: Orebv. M. (2014). Evaluation of oxidative stress
andDNA damage in cement and tannery workers in Egypt Inhal Toxicol 26: 289-298.
http://dx.doi.Org/10.3109/08958378.2014.885100
Elsaieed. EM: Nada. SA. (2002). Teratogenicity of hexavalent chromium in rats and the beneficial
role of ginseng. Bull Environ Contam Toxicol 68: 361-368.
http://dx.doi.Org/10.1007/s00128-001-0262-0
Elshazlv. MO: Morgan. AM: Ali. ME: Abdel-Mawla. E: Abd El-Rahman. SS. (2016). The mitigative
effect of Raphanus sativus oil on chromium-induced geno- and hepatotoxicity in male rats.
JAR 7: 413-421. http://dx.doi.Org/10.1016/i.iare.2016.02.008
This document is a draft for review purposes only and does not constitute Agency policy.
R-13 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
EMEA (European Medicines Agency). (2008). Non-clinical guideline on drug induced hepatotoxicity.
(Doc. Ref. EMEA/CHMP/SWP/15 0115/2006). London, UK.
http://www.ema.europa.eu/docs/en GB/document library/Scientific guideline/2 009/09/
WC500003355.pdf
EMEA (European Medicines Agency). (2010). Reflection paper on non-clinical evaluation of drug-
induced liver injury (DILI). (EMEA/CHMP/SWP/150115/2006). London, UK.
http://www.ema.europa.eu/docs/en GB/document library/Scientific guideline/2010/07/
WC500094591.pdf
Engstrom. G: Hedblad. B: Berglund. G: Tanzon. L: Lindgarde. F. (2007a). Plasma levels of
complement C3 is associated with development of hypertension: a longitudinal cohort
study. J Hum Hypertens 21: 276-282. http://dx.doi.org/10.1038/si.ihh.1002129
Engstrom. G: Hedblad. B: Eriksson. KF: Tanzon. L: Lindgarde. F. (2005). Complement C3 is a risk
factor for the development of diabetes: a population-based cohort study. Diabetes 54: 570-
575. http://dx.doi.Org/10.2337/diabetes.54.2.570
Engstrom. G: Hedblad. B: Tanzon. L: Lindgarde. F. (2007b). Complement C3 and C4 in plasma and
incidence of myocardial infarction and stroke: a population-based cohort study. Eur J
Cardiovasc Prev Rehabil 14: 392-397.
http://dx.doi.Org/10.1097/01.hir.0000244582.30421.b2
Evans. TT: Ganiam. VK. (2011). Chapter 2 - Reproductive anatomy and physiology. In RC Gupta (Ed.),
Reproductive and Developmental Toxicology (lsted., pp. 7-32). San Diego, CA: Academic
Press, http://dx.doi.org/10.1016/B978-0-12-382032-7.10002-5
Fagliano. TA: Savrin. I: Udasin. I: Gochfeld. M. (1997). Community exposure and medical screening
near chromium waste sites in New Jersey. Regul Toxicol Pharmacol 26: S13-S22.
http://dx.doi.org/10.1006/rtph.1997.1134
Faux. SP: Gao. M: Aw. TC: Braithwaite. RA: Levy. LS. (1994). Molecular epidemiological studies in
workers exposed to chromium-containing compounds. Clin Chem 40: 1454-1456.
Febel. H: Szegedi. B: Huszar. S. (2001). Absorption of inorganic, trivalent and hexavalent chromium
following oral and intrajejunal doses in rats. Acta Vet Hung 49: 203-209.
Feldman. M: Barnett. C. (1991). Fasting gastric pH and its relationship to true hypochlorhydria in
humans. Dig Dis Sci 36: 866-869. http://dx.doi.org/10.1007/BF01297133
Fendorf. S: Wielinga. BW: Hansel. CM. (2000). Chromium transformations in natural environments:
The role of biological and abiological. processes in chromium(VI) reduction. International
Geology Review 42: 691-701. http: //dx.doi.org/10.1080 /00206810009465107
Fendorf. SE. (1995). Surface reactions of chromium in soils and waters. Geoderma 67: 55-71.
http://dx.doi.org/10.1016/0016-706ir94100062-F
Fendorf. SE: Li. G. (1996). Kinetics of Chromate Reduction by Ferrous Iron. Environ Sci Technol 30:
1614-1617. http://dx.doi.org/10.1021/es950618m
Fernandes. MA: Mota. IM: Silva. MT: Oliveira. CR: Geraldes. CF: Alpoim. MC. (1999). Human
erythrocytes are protected against chromate-induced peroxidation. Ecotoxicol Environ Saf
43: 38-46. http://dx.doi.org/10.1006/eesa.l998.1755
Fernandez-Nieto. M: Ouirce. S: Carnes. I: Sastre. I. (2006). Occupational asthma due to chromium
and nickel salts. Int Arch Occup Environ Health 79: 483-486.
http://dx.doi.Org/10.1007/s00420-005-0078-z
Figgitt. M: Newson. R: Leslie. II: Fisher. I: Ingham. E: Case. CP. (2010). The genotoxicity of
physiological concentrations of chromium (Cr(III) and Cr(VI)) and cobalt (Co(II)): an in
vitro study. Mutat Res 688: 53-61. http://dx.doi.Org/10.1016/i.mrfmmm.2010.03.008
Figoni. RA: Treagan. L. (1975). Inhibitory effect of nickel and chromium upon antibody response of
rats to immunization with T-l phage. Res Commun Mol Pathol Pharmacol 11: 335-338.
This document is a draft for review purposes only and does not constitute Agency policy.
R-14 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Figueiredo. ACM. G: Gomes-Filho. IS: Silva. RB: Pereira. PPS: Mata. FAF. D: Lvrio. AO: Souza. ES: Cruz.
SS: Pereira. MG. (2018). Maternal anemia and low birth weight: A systematic review and
meta-analysis [Review], Nutrients 10: 601. http://dx.doi.org/10.3390/nul0050601
Finlev. BL: Kerger. BP: Katona. MW: Gargas. ML: Corbett. GC: Paustenbach. DT. (1997). Human
ingestion of chromium (VI) in drinking water: pharmacokinetics following repeated
exposure. Toxicol Appl Pharmacol 142: 151-159.
http://dx.doi.org/10.1006/taap.1996.7993
Foley. GL. (2001). Overview of male reproductive pathology [Review], Toxicol Pathol 29: 49-63.
http://dx.doi.Org/10.1080/019262301301418856
Foster. P: Gray. LE. Tr. (2013). Toxic responses of the reproductive system. In CD Klaassen (Ed.),
Casarett&Doull's toxicology: The basic science of poisons (8th ed., pp. 861-906). New York,
NY: McGraw-Hill Education, https: //www.mheducation.com/highered/product/casarett-
doull-s-toxicology-basic-science-poisons-9th-edition-klaassen/9781259863 745.html
Fox. IG: Wang. TC. (2007). Inflammation, atrophy, and gastric cancer [Review], J Clin Invest 117: 60-
69. http://dx.doi.org/10.1172/TCI30111
Fox. TG: Wang. TC. (2014). Dietary factors modulate Helicobacter-associated gastric cancer in
rodentmodels. Toxicol Pathol 42: 162-181. http://dx.doi.org/10.1177/Q192623313512564
Francke. S: Mog. S. (2021). Pathology review of hexavalent chromium mode of action evidence. U.S.
Department of Health and Human Services, Food and Drug Administration.
Fresquez. MR: Pappas. RS: Watson. CH. (2013). Establishment of toxic metal reference range in
tobacco from US cigarettes. J Anal Toxicol 37: 298-304.
http: / /dx. do i. or g /10.10 9 3 /i at/bktO 21
Frev. M: Seidel. C: Edwards. M. (2004). Occurrence survey of boron and hexavalent chromium. (No.
91044F). Denver, CO: American Water Works Association Research Foundation.
http://www.waterrf.org/PublicReportLibrarv/91Q44F.pdf
Frvzek. TP: Mumma. MT: Mclaughlin. IK: Henderson. BE: Blot. WT. (2001). Cancer mortality in
relation to environmental chromium exposure. J Occup Environ Med 43: 635-640.
Furman. D: Campisi. 1: Verdin. E: Carrera-Bastos. P: Targ. S: Franceschi. C: Ferrucci. L: Gilrov. DW:
Fasano. A: Miller. GW: Miller. AH: Mantovani. A: Wevand. CM: Barzilai. N: Goronzv. IT: Rando.
TA: Effros. RB: Lucia. A: Kleinstreuer. N: Slavich. GM. (2019). Chronic inflammation in the
etiology of disease across the life span. Nat Med 25: 1822-1832.
http: / /dx.doi.org/10.1038 /s41591 -019-0675-0.
Gabav. M: Li. Y: Felsher. DW. (2014). MYC activation is a hallmark of cancer initiation and
maintenance [Review], Cold Spring Harb PerspectMed 4.
http://dx.doi.org/10.1101/cshperspect.a014241
Gambelunghe. A: Piccinini. R: Abbritti. G: Ambrogi. M: Ugolini. B: Marchetti. C: Migliorati. G:
Balducci. C: Muzi. G. (2006). Chromium Vl-induced apoptosis in a human bronchial
epithelial cell line (BEAS-2B) and a lymphoblastic leukemia cell line (MOLT-4). J Occup
Environ Med 48: 319-325. http://dx.doi.org/10.1097/01.iom.000Q197859.46894.7d
Gambelunghe. A: Piccinini. R: Ambrogi. M: Villarini. M: Moretti. M: Marchetti. C: Abbritti. G: Muzi. G.
(2003). Primary DNA damage in chrome-plating workers. Toxicology 188: 187-195.
Gandhi. M: Aweeka. F: Greenblatt. RM: Blaschke. TF. (2004). Sex differences in pharmacokinetics
and pharmacodynamics [Review], Annu Rev Pharmacol Toxicol 44: 499-523.
http://dx.doi.org/10.1146/annurev.pharmtox.44.101802.121453
Gao. M: Levy. LS: Faux. SP: Aw. TC: Braithwaite. RA: Brown. SS. (1994). Use of molecular
epidemiological techniques in a pilot study on workers exposed to chromium. Occup
Environ Med 51: 663-668. http://dx.doi.org/10.1136/oem.51.10.663
Gatto. NM: Kelsh. MA: Mai. DH: Suh. M: Proctor. DM. (2010). Occupational exposure to hexavalent
chromium and cancers of the gastrointestinal tract: a meta-analysis. Cancer Detect Prev 34:
388-399. http://dx.doi.Org/10.1016/i.canep.2010.03.013
This document is a draft for review purposes only and does not constitute Agency policy.
R-15 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Gazon. H: Barbeau. B: Mesnard. IM: Peloponese. TM. (2017). Hijacking of the AP-1 Signaling Pathway
during Development of ATL [Review], FMICB 8: 2686.
http://dx.doi.org/10.3389/fmicb.2017.Q2686
GBP 2017 (GBD 2017 Gastro-oesophageal Reflux Disease Collaborators). (2020). The global,
regional, and national burden of gastro-oesophageal reflux disease in 195 countries and
territories, 1990-2017:Aa systematic analysis for the Global Burden of Disease Study 2017.
Lancet Gastroenterology and Hepatology 5: 561-581. http://dx.doi.org/10.1016/S2468-
1253Q9130408-X
Geetha. S: Sai Ram. M: Mongia. SS: Singh. V: Ilavazhagan. G: Sawhnev. RC. (2003). Evaluation of
antioxidant activity of leaf extract of Seabuckthorn (Hippophae rhamnoides L.) on
chromium(VI) induced oxidative stress in albino rats. J Ethnopharmacol 87: 247-251.
http://dx.doi.Org/10.1016/S0378-8741f03100154-5
Gehlhaus. MW. Ill: Gift. IS: Hogan. KA: Kopvlev. L: Schlosser. PM: Kadrv. AR. (2011). Approaches to
cancer assessment in EPA's Integrated Risk Information System [Review], Toxicol Appl
Pharmacol 254: 170-180. http://dx.doi.Org/10.1016/i.taap.2010.10.019
Gelberg. H. (2018). Pathophysiological mechanisms of gastrointestinal toxicity. In CA McQueen
(Ed.), Comprehensive toxicology, vol 3 (3rd ed., pp. 139-178). Amsterdam, Netherlands:
Elsevier, http://dx.doi.org/10.1016/B978-0-12-801238-3.10923-7
Geoffrov-Siraudin. C: Perrard. MH: Chaspoul. F: Lanteaume. A: Gallice. P: Durand. P: Guichaoua. MR.
(2010). Validation of a rat seminiferous tubule culture model as a suitable system for
studying toxicant impact on meiosis effect of hexavalent chromium. Toxicol Sci 116: 286-
296. http://dx.doi.org/10.1093/toxsci/kfq099
Gerin. M: Fletcher. AC: Gray. C: Winkelmann. R: Boffetta. P: Simonato. L. (1993). Development and
use of a welding process exposure matrix in a historical prospective study of lung cancer
risk in European welders. Int J Epidemiol 22: S22-S28.
Gibb. H: Wang. I: O'Learv. K: Chen. C: Bateson. TF: Kopvlev. L. (2020). The effect of age on the
relative risk of lung cancer mortality in a cohort of chromium production workers. Am J Ind
Med 63: 774-778. http://dx.doi.org/10.1002/aiim.23152
Gibb. HI: Lees. PS: Pinskv. PF: Roonev. BC. (2000a). Clinical findings of irritation among chromium
chemical production workers. Am J Ind Med 38: 127-131.
Gibb. HI: Lees. PS: Pinskv. PF: Roonev. BC. (2000b). Lung cancer among workers in chromium
chemical production. Am J Ind Med 38: 115-126.
Gibb. HI: Lees. PS: Wang. 1: Grace O'Learv. K. (2015). Extended followup of a cohort of chromium
production workers. Am J Ind Med 58: 905-913. http://dx.doi.org/10.1002/ aiim.22479
Glaser. U: Hochrainer. D: Kloeppel. H: Kuhnen. H. (1985). Low level chromium (VI) inhalation
effects on alveolar macrophages and immune functions in Wistar rats. Arch Toxicol 57: 250-
256. http://dx.doi.Org/10.1007/BF00324787
Glaser. U: Hochrainer. D: Kloeppel. H: Oldiges. H. (1986). Carcinogenicity of sodium dichromate and
chromium (VI/III)oxide aerosols inhaled by male Wistar rats. Toxicology 42: 219-232.
http: //dx.doi.org/10.1016/0300-483Xr86190011 -9
Glaser. U: Hochrainer. D: Steinhoff. D. (1990). Investigation of irritating properties of inhaled Cr(VI)
with possible influence on its carcinogenic action. In NH Seemayer; W Hadnagy (Eds.),
Environmental hygiene II (pp. 239-245). Berlin, Germany: Springer-Verlag.
Glaviano. A: Navak. V: Cabuv. E: Baird. DM: Yin. Z: Newson. R: Ladon. D: Rubio. MA: Sliiepcevic. P:
Lvng. F: Mothersill. C: Case. CP. (2006). Effects of hTERT on metal ion-induced genomic
instability. Oncogene 25: 3424-3435. http://dx.doi.org/10.1038/si.onc.1209399
Go'ldina. IR: Saichenko. SP: Nadeenko. VG: Diachenko. OZ. (1989). [Changes in the occurrence of
sister chromatid exchange and chromosome aberrations in bone marrow cells of mice after
administration of chromium with drinking water]. Gig Sanit7-9.
This document is a draft for review purposes only and does not constitute Agency policy.
R-16 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Gochfeld. M. (2007). Framework for gender differences in human and animal toxicology [Review],
Environ Res 104: 4-21. http://dx.doi.Org/10.1016/i.envres.2005.12.005
Goldberg. MT: Blakev. DH: Bruce. WR. (1983). Comparison of the effects of 1,2-dimethylhydrazine
and cyclophosphamide on micronucleus incidence in bone marrow and colon. Mutat Res
Fundam Mol Mech Mutagen 109: 91-98. http://dx.doi.org/10.1016/0027-5107r83190098-2
Gomes. ER. (1972). Incidence of chromium-induced lesions among electroplating workers in Brazil.
Ind Med Surg 41: 21-25.
Gonda. TA: Tu. SP: Wang. TC. (2009). Chronic inflammation, the tumor microenvironment and
carcinogenesis [Review], Cell Cycle 8: 2005-2013. http://dx.doi.Org/10.4161/cc.8.13.8985
Goulart. M: Batoreu. MC: Rodrigues. AS: Laires. A: Rueff. 1. (2005). Lipoperoxidation products and
thiol antioxidants in chromium exposed workers. Mutagenesis 20: 311-315.
http://dx.doi.org/10.1093/mutage/gei043
Goulle. TP: Saussereau. E: Grosiean. 1: Doche. C: Mahieu. L: Thouret. TM: Guerbet. M: Lacroix. C.
(2012). Accidental potassium dichromate poisoning. Toxicokinetics of chromium by ICP-
MS-CRC in biological fluids and in hair. Forensic Sci Int 217: e8-12.
http://dx.doi.Org/10.1016/i.forsciint2011.10.020
Greenspan. LC: Lee. MM. (2018). Endocrine disrupters and pubertal timing [Review], Curr Opin
Endocrinol Diabetes Obes 25: 49-54. http://dx.doi.Org/10.1097/MED.0000000000000377
Gromadziriska. 1: Wasowicz. W: Sklodowska. M: Bulikowski. W: Rydzvriski. K. (1996). The influence
of atmospheric chromium on selenium content and glutathione peroxidase activity in blood
of tannery workers. Environ Health Perspect 104: 1312-1316.
http://dx.doi.org/10.1289/ehp.104-1469531
Group. FW. (2000). Nonclincial assessment of potential hepatotoxicity in man (a concept paper
meant to provide a framework for discussion ata February 12&13 Workshop. FDA Working
Group.
Giierci. A: Seoane. A: Dulout. FN. (2000). Aneugenic effects of some metal compounds assessed by
chromosome counting in MRC-5 human cells. Mutat Res 469: 35-40.
Guo. X: Feng. L: Lemos. B: Lou. 1. (2019). DNA methylation modifications induced by hexavalent
chromium. J Environ Sci Health C Environ Carcinog Ecotoxicol Rev 37: 133-145.
http://dx.doi. org/10.1080/10590501.2019.1592640
Guo. YT: Pan. WW: Liu. SB: Shen. ZF: Xu. Y: Hu. LL. (2020). ERK/MAPK signalling pathway and
tumorigenesis [Review], Exp Ther Med 19: 1997-2007.
http://dx.doi.org/10.3892/etm.2020.8454
Halasova. E: Adamkov. M: Matakova. T: Kavcova. E: Poliacek. I: Singliar. A. (2010). Lung cancer
incidence and survival in chromium exposed individuals with respect to expression of anti-
apoptotic protein survivin and tumor suppressor P53 protein. Eur J Med Res 15: 55-59.
Halasova. E: Matakova. T: Musak. L: Polakova. V: Letkova. L: Dobrota. D: Vodicka. P. (2012).
Evaluating chromosomal damage in workers exposed to hexavalent chromium and the
modulating role of polymorphisms of DNA repair genes. Int Arch Occup Environ Health 85:
473. http://dx.doi.Org/10.1007/s00420-011-0684-x
Halasova. E: Matakova. T: Musak. L: Polakova. V: Vodicka. P. (2008). Chromosomal damage and
polymorphisms of DNA repair genes XRCC1 andXRCC3 in workers exposed to chromium.
Neuro Endocrinol Lett 29: 658-662.
Hamilton. EM: Young. SD: Bailey. EH: Watts. MT. (2018). Chromium speciation in foodstuffs: A
review [Review], Food Chem 250. http://dx.doi.Org/10.1016/i.foodchem.2018.01.016
Hanahan. D. (2022). Hallmarks of cancer: New dimensions [Review], Cancer Discovery 12: 31-46.
http://dx.doi.org/10.1158/2159-8290.CD-21-1059
Hanahan. D: Weinberg. RA. (2011). Hallmarks of cancer: The next generation [Review], Cell 144:
646-674. http://dx.doi.Org/10.1016/i.cell.2011.02.013
This document is a draft for review purposes only and does not constitute Agency policy.
R-17 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Hanaoka. T: Yamano. Y: Katsuno. N: Kagawa. I: Ishizu. S. (1997). Elevated serum levels of pantropic
p53 proteins in chromium workers. Scand J Work Environ Health 23: 37-40.
http: / /dx. do i. or g /10.5 2 71 /si weh. 17 6
Hanslian. L: Navratil. 1: Turak. 1. (1967). [Damage to the upper respiratory tract by a chromic acid
aerosol], Czech J Occup Med 19: 294-298.
Hantson. P: Van Caenegem. 0: Decordier. I: Haufroid. V: Lison. D. (2005). Hexavalent chromium
ingestion: Biological markers of nephrotoxicity and genotoxicity. Clin Toxicol 43: 111-112.
http://dx.doi.org/10.1081/CLT-50414
Harris. HH: Levina. A: Dillon. CT: Mulvani. I: Lai. B: Cai. Z: Lay. PA. (2005). Time-dependent uptake,
distribution and biotransformation of chromium(VI) in individual and bulk human lung
cells: application of synchrotron radiation techniques. J Biol Inorg Chem 10: 105-118.
http://dx.doi.Org/10.1007/s00775-004-0617-l
Haschek. WM: Rousseaux. CG: Wallig. MA. (2009). Male reproductive system. In Fundamentals of
toxicologic pathology (2nd ed.). Cambridge, MA: Academic Press.
http://dx.doi.Org/10.1016/B978-0-12-370469-6.00018-0
Haseman. IK: Bailer. AT: Kodell. RL: Morris. R: Portier. K. (2001). Statistical issues in the analysis of
low-dose endocrine disruptor data. Toxicol Sci 61: 201-210.
Hausladen. DM: Alexander-Ozinskas. A: Mcclain. CN: Fendorf. S. (2018). Hexavalent chromium
sources and distribution in California groundwater. Environ Sci Technol 52: 8242-8251.
http://dx.doi.org/10.1021/acs.est.7b06627
Hausladen. DM: Fendorf. S. (2017). Hexavalent chromium generation within naturally structured
soils and sediments. Environ Sci Technol 51: 2058-2067.
http://dx.doi.org/10.1021/acs.est.6b04039
Havakawa. Y: Fox. TG: Gonda. T: Worthlev. PL: Muthupalani. S: Wang. TC. (2013). Mouse models of
gastric cancer. Cancers 5: 92-130. http://dx.doi.org/10.3390/cancers5010092
Havashi. M. (2016). The micronucleus test-most widely used in vivo genotoxicity test [Review],
Genes Environ 38: 18. http://dx.doi.org/10.1186/s41021-016-0Q44-x
Hayes. RB: Lilienfeld. AM: Snell. LM. (1979). Mortality in chromium chemical production workers: A
prospective study. Int J Epidemiol 8: 365-374. http://dx.doi.Org/10.1093/iie/8.4.365
He. F: Ru. X: Wen. T. (2020). NRF2, a transcription factor for stress response and beyond [Review],
International Journal of Molecular Sciences 21: 4777.
http://dx.doi.org/10.3390/iims21134777
He. I: Oian. X: Carpenter. R: Xu. 0: Wang. L: Oi. Y: Wang. ZX: Liu. LZ: Tiang. BH. (2013). Repression of
miR-143 Mediates Cr (Vl)-Induced Tumor Angiogenesis via IGF-IR/IRS1/ERK/IL-8
Pathway. Toxicol Sci 134: 26-38. http://dx.doi.org/10.1093/toxsci/kftl01
Hedberg. YS. (2018). Metal allergy: Chromium. In J Chen; J Thyssen (Eds.), Metal allergy (pp. 349-
364). Cham, Switzerland: Springer. http://dx.doi.org/10.1007/978-3-319-585Q3-l 27
Helmbold. P: Rvtter. M: Ziegler. V: Haustein. UF. (1993). Induction of type IV hypersensitivity to
contact allergens in guinea pigs by in vitro haptenized allogenic peritoneal exudate cells.
Acta Derm Venereol 73: 365-369. http://dx.doi.org/10.2340/0001555573365369
Henderson. RF. (1984). Use of bronchoalveolar lavage to detect lung damage [Review], Environ
Health Perspect 56: 115-129.
Henderson. RF: Benson. TM: Hahn. FF: Hobbs. CH: Tones. RK: Mauderlv. TL: McClellan. RO: Pickrell.
TA. (1985). New approaches for the evaluation of pulmonary toxicity: Bronchoalveolar
lavage fluid analysis [Review], Toxicol Sci 5: 451-458.
Hens. B: Brouwers. I: Anneveld. B: Corsetti. M: Svmillides. M: Vertzoni. M: Reppas. C: Turner. DB:
Augustiins. P. (2014). Gastrointestinal transfer: In vivo evaluation and implementation in in
vitro and in silico predictive tools. Eur J Pharm Sci 63: 233-242.
http://dx.doi.Org/10.1016/i.eips.2014.07.008
This document is a draft for review purposes only and does not constitute Agency policy.
R-18 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Hertle. E: van Greevenbroek. MM: Stehouwer. CD. (2012). Complement C3: an emerging risk factor
in cardiometabolic disease. Diabetologia 55: 881-884. http://dx.doi.org/10.1007/s00125-
012-2462-z
Hfaiedh. M: Brahmi. D: Zourgui. L. (2014). Protective role of cactus cladodes extract on sodium
dichromate-induced testicular injury and oxidative stress in rats. Biol Trace Elem Res 159:
304-311. http: //dx.doi.Org/l 0.1007/sl 2011 -014-9969-8
Hilali. A: Anane. R: Taaouani. N: Creppv. EE: Verschaeve. L. (2008). Cytogenetic analysis of tannery
workers in Morocco. J Appl Toxicol 28: 439-442. http://dx.doi.org/10.1002/iat 1291
Hill. AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58: 295-
300.
Hiller. MM: Leggett. RW. (2020). A biokinetic model for trivalent or hexavalent chromium in adult
humans. J Radiol Prot 40: 19-39. http://dx.doi.org/10.lQ88/1361-6498/ab4286
Hinds. WC. (1999). 11: Respiratory Deposition. In Aerosol Technology: Properties, Behavior, and
Measurement ofAirborne Particles (2 ed.). New York, NY: Wiley- Interscience.
Hirose. T: Kondo. K: Takahashi. Y: Ishikura. H: Fuiino. H: Tsuvuguchi. M: Hashimoto. M: Yokose. T:
Mukai. K: Kodama. T: Monden. Y. (2002). Frequent microsatellite instability in lung cancer
from chromate-exposed workers. Mol Carcinog 33: 172-180.
Hiollund. NH: Bonde. TP: Ernst. E: Lindenberg. S: Andersen. AN: Olsen. 1. (2005). Spontaneous
abortion in IVF couples--a role of male welding exposure. Hum Reprod 20: 1793-1797.
http://dx.doi.org/10.1093/humrep/deh680
Hiollund. NH: Bonde. TP: Hansen. KS. (1995). Male-mediated risk of spontaneous abortion with
reference to stainless steel welding. Scand J Work Environ Health 21: 272-276.
http: / /dx. do i. or g /10.5 2 71 /si weh. 3 7
Hiollund. NH: Bonde. TP: Tensen. TK: Ernst. E: Henriksen. TB: Kolstad. HA: Giwercman. A:
Skakkebaek. NE: Olsen. T. (1998). Semen quality and sex hormones with reference to metal
welding. Reprod Toxicol 12: 91-95.
Hiollund. NH: Bonde. TP: Tensen. TK: Henriksen. TB: Andersson. AM: Kolstad. HA: Ernst. E:
Giwercman. A: Skakkebaek. NE: Olsen. T. (2000). Male-mediated spontaneous abortion
among spouses of stainless steel welders. Scand J Work Environ Health 26: 187-192.
Holmes. AL: Wise. SS: Sandwick. ST: Lingle. WL: Negron. VC: Thompson. WD: Wise. TP. Sr. (2006).
Chronic exposure to lead chromate causes centrosome abnormalities and aneuploidy in
human lung cells. Cancer Res 66: 4041-4048. http://dx.doi.org/10.1158/00Q8-5472.can-
05-3312
Hooftman. RN: Arkestevn. CWM: Roza. P. (1988). Cytotoxicity of some types of welding fume
particles to bovine alveolar macrophages. Ann Occup Hyg 32: 95-102.
http://dx.doi.Org/10.1093/annhyg/32.l.95
Horiguchi. S: Morinaga. K: Endo. G. (1990). Epidemiological study of mortality from cancer among
chromium platers. Asia Pac J Public Health 4: 169-174.
http://dx.d0i.0rg/l 0.1177/101053959000400316
Houghton. 1: Stoicov. C: Nomura. S: Rogers. AB: Carlson. 1: Li. H: Cai. X: Fox. TG: Goldenring. TR: Wang.
TC. (2004). Gastric cancer originating from bone marrow-derived cells. Science 306: 1568-
1571. http://dx.doi.org/10.1126/science.1099513
Hu. G: Li. P: Cui. X: Li. Y: Zhang. 1: Zhai. X: Yu. S: Tang. S: Zhao. Z: Wang. 1: Tia. G. (2018). Cr(VI)-
induced methylation and down-regulation of DNA repair genes and its association with
markers of genetic damage in workers and 16HBE cells. Environ Pollut 238: 833-843.
http://dx.doi.Org/10.1016/i.envpol.2018.03.046
Hu. G: Li. P: Li. Y: Wang. T: Gao. X. in: Zhang. W: Tia. G. (2016). Methylation levels of P16 andTP53
that are involved in DNA strand breakage of 16HBE cells treated by hexavalent chromium.
Toxicol Lett 249: 15-21. http://dx.doi.Org/10.1016/i.toxlet.2016.03.003
This document is a draft for review purposes only and does not constitute Agency policy.
R-19 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Hu. G: Long. C: Hu. L: Zhang. Y: Hong. S: Zhang. 0: Zheng. P: Su. Z: Xu. I: Wang. L: Gao. X: Zhu. X: Yuan.
F: Wang. T: Yu. S: Tia. G. (2022). Blood chromium exposure, immune inflammation and
genetic damage: Exploring associations and mediation effects in chromate exposed
population. J Hazard Mater 425: 127769. http://dx.doi.Org/10.1016/i.ihazmat2021.127769
Huang. L: Yu. CH: Hopke. PK: Liov. PI: Buckley. BT: Shin. 1Y: Fan. ZT. (2014). Measurement of soluble
and total hexavalent chromium in the ambient airborne particles in New Jersey. Aerosol Air
Qual Res 14: 1939-1949. http://dx.doi.org/10.4209/aaqr.2013.10.0312
Huang. YL: Chen. CY: Sheu. TY: Chuang. IC: Pan. TH: Lin. TH. (1999). Lipid peroxidation in workers
exposed to hexavalent chromium. J Toxicol Environ Health A 56: 235-247.
http: //dx.doi.org/10.1080/009841099158088
Husgafvel-Pursiainen. K: Kalliomaki. PL: Sorsa. M. (1982). A chromosome study among stainless
steel welders. J Occup Environ Med 24: 762-766.
Huvinen. M: Makitie. A: Tarventaus. H: Wolff. H: Stjernvall. T: Hovi. A: Hirvonen. A: Ranta. R:
Nurminen. M: Norppa. H. (2002a). Nasal cell micronuclei, cytology and clinical symptoms in
stainless steel production workers exposed to chromium. Mutagenesis 17: 425-429.
http://dx.doi.Org/10.1093/mutage/17.5.425
Huvinen. M: Uitti. 1: Oksa. P: Palmroos. P: Laippala. P. (2002b). Respiratory health effects of long-
term exposure to different chromium species in stainless steel production. Occup Med
(Lond) 52: 203-212.
Huvinen. M: Uitti. 1: Zitting. A: Roto. P: Virkola. K: Kuikka. P: Laippala. P: Aitio. A. (1996).
Respiratory health of workers exposed to low levels of chromium in stainless steel
production. Occup Environ Med 53: 741-747.
Hydrometrics. (2013). Project report background concentrations of inorganic constituents in
Montana surface soils. Helena, MT: Hydrometrics, Inc.
https://deq.mt.gov/Portals/112/Land/StateSuperfund/Documents/InorganicBackground/
BkgdlnorganicsReportpdf
IARC (International Agency for Research on Cancer). (2012). A review of human carcinogens. Part
C: Arsenic, metals, fibres, and dusts [IARC Monograph], In Arsenic, Metals, Fibres and Dusts.
Lyon, France: World Health Organization.
http://monographs.iarc.fr/ENG/Monographs/vollOOC/monolOOC.pdf
Iarmarcovai. G: Sari-Minodier. I: Chaspoul. F: Botta. C: De Meo. M: Orsiere. T: Berge-Lefranc. TL:
Gallice. P: Botta. A. (2005). Risk assessment of welders using analysis of eight metals by ICP-
MS in blood and urine and DNA damage evaluation by the comet and micronucleus assays;
influence ofXRCCl andXRCC3 polymorphisms. Mutagenesis 20: 425-432.
http://dx.doi.org/10.1093/mutage/gei058
Ibrahim. R: Brix. A: Malarkev. DE: Nvska. A: Asif. M: Ramot. Y. (2021). Chemical-induced oral
squamous cell neoplasms in rodents: An overview of NTP 2-year cancer studies. Toxicol
Pathol, http://dx.doi.org/10.1177/0192623321989956
ICRP (International Commission on Radiological Protection). (1994). Human respiratory tract
model for radiological protection. Ann ICRP 24.
ICRP (International Commission on Radiological Protection). (2006). Human alimentary tract
model for radiological protection. ICRP Publication 100. A report of The International
Commission on Radiological Protection. Ann ICRP 36: 25-327, iii.
Iiiima. S: Spindle. A: Pedersen. RA. (1983). Developmental and cytogenetic effects of potassium
dichromate on mouse embryos in vitro. Teratology 27: 109-115.
http://dx.doi.org/10.1002/tera.142027Q116
Ikarashi. Y: Momma. 1: Tsuchiva. T: Nakamura. A. (1996). Evaluation of skin sensitization potential
of nickel, chromium, titanium and zirconium salts using guinea-pigs and mice. Biomaterials
17: 2103-2108. http://dx.doi.org/10.1016/0142-9612r96100038-5
This document is a draft for review purposes only and does not constitute Agency policy.
R-20 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Ikarashi. Y: Tsuchiva. T: Nakamura. A. (1992). Detection of contact sensitivity of metal salts using
the murine local lymph node assay. Toxicol Lett 62: 53-61.
http://dx.doi.org/10.1016/0378-4274r92190078-x
IOM (Institute of Medicine). (2011). Dietary reference intakes for calcium and vitamin D. In AC
Ross; CL Taylor; AL Yaktine; HB Del Valle (Eds.). Washington, DC: National Academies Press.
http://dx.doi.org/10.17226/13050
IPCS (International Programme on Chemical Safety). (2012). Harmonization project document no.
10: Guidance for immunotoxicity risk assessment for chemicals. (Harmonization Project
Document No. 10). Geneva, Switzerland: World Health Organization.
http://www.inchem.org/documents/harmproi/harmproi/harmproilO.pdf
IPCS (International Programme on Chemical Safety). (2013). Inorganic chromium(VI) compounds.
(78). Geneva, Switzerland: World Health Organization.
http://www.who.int/ipcs/publications/cicad/cicad 78.pdf
Ishikawa. Y: Nakagawa. K: Satoh. Y: Kitagawa. T: Sugano. H: Hirano. T: Tsuchiva. E. (1994a).
Characteristics of chromate workers' cancers, chromium lung deposition and precancerous
bronchial lesions: an autopsy study. Br J Cancer 70: 160-166.
Ishikawa. Y: Nakagawa. K: Satoh. Y: Kitagawa. T: Sugano. H: Hirano. T: Tsuchiva. E. (1994b). Hot
spots of chromium accumulation at bifurcations of chromate workers' bronchi. Cancer Res
54: 2342-2346.
Islam. LN: Rahman. F: Hossain. A. (2019). Serum Immunoglobulin Levels and Complement Function
ofTannery Workers in Bangladesh. 9: 190308. http://dx.doi.org/lQ.5696/2156-9614-
9.21.190308
Izzotti. A: Bagnasco. M: Camoirano. A: Orlando. M: De Flora. S. (1998). DNA fragmentation, DNA-
protein crosslinks, 32P postlabeled nucleotidic modifications, and 8-hydroxy-2'-
deoxyguanosine in the lung but not in the liver of rats receiving intratracheal instillations of
chromium(VI). Chemoprevention by oral N-acetylcysteine. Mutat Res Fundam Mol Mech
Mutagen 400: 233-244. http://dx.doi.org/10.1016/S0027-5107r98100028-l
Tacquet. P: Drave. TP. (1982). Toxicity of chromium salts to cultured mouse embryos. Toxicol Lett
12: 53-57. http:/ /dx.doi.org/10.1016/0378-4274r82190198-9
Tardine. PM: Mehlhorn. TL: Bailey. WB: Brooks. SC: Fendorf. S: Gentry. RW: Phelps. TT: Saiers. IE.
(2011). Geochemical processes governing the fate and transport of chromium(III) and
chromium(VI) in soils. Vadose Zone J 10: 1058-1070.
http://dx.doi.org/10.2136/vzi2010.0102
Tasperson. KW: Patel. SG: Ahnen. PI. (2017). APC-Associated Polyposis Conditions. In MP Adam; HH
Ardinger; RAPagon; SE Wallace; LJH Bean; G Mirzaa; AAmemiya (Eds.), GeneReviews
[Internet], Seattle, WA: University of Washington, Seattle.
Telnes. IE: Knudsen. LE. (1988). Stainless steel welding and semen quality. Reprod Toxicol 2:213-
215. http://dx.doi.Org/10.1016/0890-6238r88190025-l
Tia. G. (2021). [Request sent to Professor Guang Jia for information from publication with reply,
response to request, and follow-up attached]. Available online
Tin. Y: Liu. L: Zhang. S: Tao. B: Tao. R: He. X: Ou. L: Huang. 1: Wang. X: Fu. Z. (2016). Chromium alters
lipopolysaccharide-induced inflammatory responses both in vivo and in vitro. Chemosphere
148: 436-443. http://dx.doi.Org/10.1016/i.chemosphere.2016.01.057
Tin. Y: Zhang. S: Tao. R: Huang. I: He. X: Ou. L: Fu. Z. (2014). Oral exposure of mice to cadmium (II),
chromium (VI) and their mixture induce oxidative- and endoplasmic reticulum-stress
mediated apoptosis in the livers. Environ Toxicol 31: 693-705.
http://dx.doi.org/10.1002/tox.22082
Tirova. D: Taneckovo. V: Znoiemska. S: Pekarek. I. (1983). Adjuvant Activity Of Muramyldipeptide In
A Guinea Pig Model Of Contact Allergy To Chromium. J Hyg 27: 395-402.
This document is a draft for review purposes only and does not constitute Agency policy.
R-21 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Johansson. A: Robertson. B: Curstedt. T: Camner. P. (1986a). Rabbit lung after inhalation of hexa-
andtrivalentchromium. EnvironRes 41: 110-119.
lohansson. A: Wiernik. A: larstrand. C: Camner. P. (1986b). Rabbit alveolar macrophages after
inhalation of hexa- and trivalent chromium. Environ Res 39: 372-385.
http://dx.doi.org/10.1016/S0013-9351C86180063-9
Tohnson. 1: Schewel. L: Graedel. TE. (2006). The contemporary anthropogenic chromium cycle.
Environ SciTechnol 40: 7060-7069. http://dx.doi.org/10.1021/esQ60061i
Tones. KB: Klein. OP. (2013). Oral epithelial stem cells in tissue maintenance and disease: the first
steps in a long journey [Review], International Journal of Oral Science 5: 121-129.
http://dx.doi.org/10.1038/iios.2013.46
Tones. RP: Sutton. PA: Evans. TP: Clifford. R: Mcavov. A: Lewis. 1: Rousseau. A: Mountford. R:
Mcwhirter. D: Malik. HZ. (2017). Specific mutations in KRAS codon 12 are associated with
worse overall survival in patients with advanced and recurrent colorectal cancer. Br J
Cancer 116: 923-929. http://dx.doi.org/10.1038/bic.2017.37
TP. B: Olsen. TH: Hansen. KS. (1992). Adverse pregnancy outcome and childhood malignancy with
reference to paternal welding exposure. Scand J Work Environ Health 18: 169-177.
Tuanes. MA. (2020). Cytoskeletal control and Wnt signaling-APC's dual contributions in stem cell
division and colorectal cancer [Review], Cancers 12: 3811.
http://dx.doi.org/10.3390/cancersl2123811
Tunaid. M: Murthv. RC: Saxena. DK. (1995). Chromium fetotoxicity in mice during late pregnancy.
Vet Hum Toxicol 37: 320-323.
Tunaid. M: Murthv. RC: Saxena. DK. (1996a). Embryo and feto toxicity of chromium in
pregestationally exposed mice. Bull Environ Contam Toxicol 57: 327-334.
Tunaid. M: Murthv. RC: Saxena. DK. (1996b). Embryotoxicity of orally administered chromium in
mice: exposure during the period of organogenesis. Toxicol Lett 84: 143-148.
Kaiser. TP: Lipscomb. TC: Wesselkamper. SC. (2012). Putative mechanisms of environmental
chemical-induced steatosis. Int J Toxicol 31: 551-563.
http://dx.doi.org/10.1177/1091581812466418
Kalahasthi. RB: Rao. RH: Murthv. RB: Kumar. MK. (2006). Effect of chromium(VI) on the status of
plasma lipid peroxidation and erythrocyte antioxidant enzymes in chromium plating
workers. Chem Biol Interact 164: 192-199. http://dx.doi.Org/10.1016/i.cbi.2006.09.012
Kalantzi. L: Goumas. K: Kalioras. V: Abrahamsson. B: Pressman. IB: Reppas. C. (2006).
Characterization of the human upper gastrointestinal contents under conditions simulating
bioavailability/bioequivalence studies. Pharm Res 23: 165-176.
http://dx.doi.Org/10.1007/sll095-005-8476-l
Kaltreider. RC: Pesce. CA: Ihnat. MA: Lariviere. TP: Hamilton. TW. (1999). Differential effects of
arsenic(III) and chromium(VI) on nuclear transcription factor binding. Mol Carcinog 25:
219-229. http://dx.doi.org/10.1002/fSICI") 1098-2 744f 199907125:3<219: :AID-
MC8>3.0.CO;2-X
Kanoiia. RK: Tunaid. M: Murthv. RC. (1996). Chromium induced teratogenicity in female rat. Toxicol
Lett 89: 207-213. http:/ /dx.doi.org/10.1016/S0378-4274C96103812-X
Kanoiia. RK: Tunaid. M: Murthv. RC. (1998). Embryo and feto toxicity of hexavalent chromium: a
long-term study. Toxicol Lett 95: 165-172. http://dx.doi.org/10.1016/SQ378-
4274C98100034-4
Kararli. TT. (1995). Comparison of the gastrointestinal anatomy, physiology, and biochemistry of
humans and commonly used laboratory animals [Review], Biopharm Drug Dispos 16:351-
380. http://dx.doi.Org/10.1002/bdd.2510160502
Karaulov. AY: Renieri. EA: Smolvagin. AI: Mikhavlova. IV: Stadnikov. AA: Begun. DN: Tsarouhas. K:
Buha Diordievic. A: Hartung. T: Tsatsakis. A. (2019). Long-term effects of chromium on
This document is a draft for review purposes only and does not constitute Agency policy.
R-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
morphological and immunological parameters of Wistar rats. Food Chem Toxicol 133:
110748. http://dx.doi.Org/10.1016/i.fct2019.110748
Katabami. M: Dosaka-Akita. H: Mishina. T: Honma. K: Kimura. K: Uchida. Y: Morikawa. K: Mikami. H:
Fukuda. S: Inuvama. Y: Ohsaki. Y: Kawakami. Y. (2000). Frequent cyclin D1 expression in
chromate-induced lung cancers. Hum Pathol 31: 973-979.
http://dx.doi.Org/10.1053/hupa.2000.9081
Kativar. S: Awasthi. SK: Sahu. RK. (2008). Suppression of IL-6 level in human peripheral blood
mononuclear cells stimulated with PHA/LPS after occupational exposure to chromium. Sci
Total Environ 390: 355-361. http://dx.doi.org/10.1016/i.scitotenv.2007.10.031
Kempe. PS: Lang. PA: Duranton. C: Akel. A: Lang. KS: Huber. SM: Wieder. T: Lang. F. (2006).
Enhanced programmed cell death of iron-deficient erythrocytes. FASEB J 20: 368-370.
http://dx.doi.org/10.1096/fi.Q5-4872fie
Kerger. BP: Butler. WT: Paustenbach. DT: Zhang. 1: Li. S. (2009). Cancer mortality in Chinese
populations surrounding an alloy plant with chromium smelting operations. J Toxicol
Environ Health A 72: 329-344. http://dx.doi.org/10.1080/15287390802529898
Kerger. BP: Finlev. BL: Corbett. GE: Dodge. DG: Paustenbach. DT. (1997). Ingestion of chromium (VI)
in drinking water by human volunteers: absorption, distribution, and excretion of single and
repeated doses. J Toxicol Environ Health 50: 67-95.
Kerger. BP: Paustenbach. DT: Corbett. GE: Finlev. BL. (1996). Absorption and elimination of trivalent
and hexavalent chromium in humans following ingestion of a bolus dose in drinking water.
Toxicol Appl Pharmacol 141: 145-158. http://dx.doi.org/10.1006/taap.1996.0271
Khan. DA: Mushtaq. S: Khan. FA: Khan. MOA. (2013). Toxic effects of chromium on tannery workers
at Sialkot (Pakistan). Toxicol Ind Health 29: 209-215.
http://dx.doi.org/10.n77/07482337n430974
Kim. H: Cho. SH: Chung. MH. (1999). Exposure to hexavalent chromium does not increase 8-
hydroxydeoxyguanosine levels in Korean chromate pigment workers. Ind Health 37: 335-
341.
Kim. HY: Lee. SB: Tang. BS. (2004). Subchronic inhalation toxicity of soluble hexavalent chromium
trioxide in rats. Arch Toxicol 78: 363-368. http://dx.doi.Org/10.1007/s00204-004-0553-4
Kim. TG: Dixon. IB. f2002I Oxidation and fate of chromium in soils. Soil Sci Plant Nutr 48: 483-490.
http://dx.doi.Org/10.1080/00380768.2002.10409230
Kim. SH: Lee. I. nC: Baek. HS: Moon. C: Kang. SS. oo: Bae. CS. ik: Kim. SH. o: Shin. PH. o: Kim. TC.
(2012). Pycnogenol (R) prevents hexavalent chromium-induced spermatotoxicity in rats.
Mol Cell Toxicol 8: 249-256. http://dx.doi.Org/10.1007/sl3273-012-0030-8
Kim. WR: Flamm. SL: Pi Bisceglie. AM: Bodenheimer. HC. (2008). Serum activity of alanine
aminotransferase (ALT) as an indicator of health and disease. Hepatology 47: 1363-1370.
http://dx.doi.org/10.1002/hep.221Q9
Kim. YP: An. SC: Ovama. T: Kawamoto. T: Kim. H. (2003). Oxidative stress, hoggl expression and
NF-kappaB activity in cells exposed to low level chromium. J Occup Health 45: 271-277.
http://dx.doi.org/10.1539/ioh.45.271
Kimber. I: Bentlev. AN: Hilton. I. (1990). Contact sensitization of mice to nickel sulphate and
potassium dichromate. Contact Perm 23: 325-330. http: //dx.doi.org/10.ll 11 /i.1600-
0536.1990.tb05166.x
Kimbrough. PE: Cohen. Y: Winer. A: Creelman. L: Mabuni. C. (1999). A critical assessment of
chromium in the environment Crit Rev Environ Sci Tech 29: 1-46.
http://dx.doi.org/10.1080/10643389991259164
Kirman. CR: Avlward. LL: Suh. M: Harris. MA: Thompson. CM: Haws. LC: Proctor. PM: Lin. SS: Parker.
W: Hays. SM. (2013). Physiologically based pharmacokinetic model for humans orally
exposed to chromium. Chem Biol Interact 204: 13-27.
http://dx.doi.Org/10.1016/j.cbi.2013.04.003
This document is a draft for review purposes only and does not constitute Agency policy.
R-23 PRAFT-PO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Kirman. CR: Hays. SM: Avlward. LL: Suh. M: Harris. MA: Thompson. CM: Haws. LC: Proctor. DM.
(2012). Physiologically based pharmacokinetic model for rats and mice orally exposed to
chromium. Chem Biol Interact 200: 45-64. http://dx.doi.Org/10.1016/i.cbi.2012.08.016
Kirman. CR: Suh. M: Hays. SM: Giirleviik. H: Gerads. R: De Flora. S: Parker. W: Lin. S: Haws. LC:
Harris. MA: Proctor. DM. (2016). Reduction of hexavalent chromium by fasted and fed
human gastric fluid. II. Ex vivo gastric reduction modeling. Toxicol Appl Pharmacol 306:
120-133. http://dx.doi.org/10.1016/i.taap.2016.07.002
Kirman. CR: Suh. M: Proctor. DM: Hays. SM. (2017). Improved physiologically based
pharmacokinetic model for oral exposures to chromium in mice, rats, and humans to
address temporal variation and sensitive populations. Toxicol Appl Pharmacol 325: 9-17.
http://dx.doi.Org/10.1016/i.taap.2017.03.023
Kirpnick-Sobol. Z: Reliene. R: Schiestl. RH. (2006). Carcinogenic Cr(VI) and the nutritional
supplement Cr(III) induce DNA deletions in yeast and mice. Cancer Res 66: 3480-3484.
http://dx.doi.Org/10.1158/0008-5472.CAN-05-3944
Kitamura. F: Yokovama. K: Araki. S: Nishikitani. M: Choi. TW: Yum. YT: Park. HC: Park. SH: Sato. H.
(2003). Increase of olfactory threshold in plating factory workers exposed to chromium in
Korea. Ind Health 41: 279-285. http://dx.doi.org/10.2486/indhealth.41.279
Kleinfeld. M: Rosso. A. (1965). Ulcerations of the nasal septum due to inhalation of chromic acid
mist Indian J Med Surg 34: 242-243.
Kleinstreuer. C: Zhang. Z: Donohue. IF. (2008). Targeted drug-aerosol delivery in the human
respiratory system. Annu Rev Biomed Eng 10: 195-220.
http://dx.doi.Org/10.1146/annurev.bioeng.10.061807.160544
Kodavanti. U. (2014). Respiratory toxicity biomarkers. In R Gupta (Ed.), Biomarkers in toxicology
(pp. 217-239). San Diego, CA: Academic Press. http://dx.doi.org/10.1016/B978-Q-12-
404630-6.00012-9
Kohl. KD: Stengel. A: Samuni-Blank. M: Dearing. MP. (2013). Effects of anatomy and diet on
gastrointestinal pH in rodents. J Exp Zool A Ecol Genet Physiol 319: 225-229.
http://dx.doi.org/10.1002/jez.1786
Kotaciriski. Z: Kostrzewski. P: Kruszewska. S: Razniewska. G: Mielczarska. I. (1999). Acute
potassium dichromate poisoning: a toxicokinetic case study. J Toxicol Clin Toxicol 37: 785-
791.
Kolacinski. Z: Kruszewska. S: Winnicka. R: Razniewska. G: Kolacinska. M. (2004). Toxicokinetics of
Chromium Elimination in Hemodialysis and Forced Diuresis Based on a Case of Potassium
Dichromate Intoxication. J Toxicol Clin Toxicol 42.
Kondo. K: Hino. N: Sasa. M: Kamamura. Y: Sakivama. S: Tsuvuguchi. M: Hashimoto. M: Uvama. T:
Monden. Y. (1997). Mutations of the p53 gene in human lung cancer from chromate-
exposed workers. Biochem Biophys Res Commun 239: 95-100.
http://dx.doi.org/10.1006/bbrc.1997.7425
Kondo. K: Takahashi. Y: Hirose. Y: Nagao. T: Tsuvuguchi. M: Hashimoto. M: Ochiai. A: Monden. Y:
Tangoku. A. (2006). The reduced expression and aberrant methylation of pl6(INK4a) in
chromate workers with lung cancer. Lung Cancer 53: 295-302.
http://dx.doi.Org/10.1016/i.lungcan.2006.05.022
Kondo. K: Takahashi. Y: Ishikawa. S: Uchihara. H: Hirose. Y: Yoshizawa. K: Tsuvuguchi. M: Takizawa.
H: Mivoshi. T: Sakivama. S: Monden. Y. (2003). Microscopic analysis of chromium
accumulation in the bronchi and lung of chromate workers. Cancer 98: 2420-2429.
http://dx.doi.org/10.1002/cncr.11818
Kopec. AK: Kim. S: Forgacs. AL: Zacharewski. TR: Proctor. DM: Harris. MA: Haws. LC: Thompson.
CM. (2012a). Genome-wide gene expression effects in B6C3F1 mouse intestinal epithelia
following 7 and 90days of exposure to hexavalent chromium in drinking water. Toxicol Appl
Pharmacol 259: 13-26. http://dx.doi.Org/10.1016/i.taap.2011.ll.012
This document is a draft for review purposes only and does not constitute Agency policy.
R-24 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Kopec. AK: Thompson. CM: Kim. S: Forgacs. AL: Zacharewski. TR. (2012b). Comparative
toxicogenomic analysis of oral Cr(VI) exposure effects in rat and mouse small intestinal
epithelia. Toxicol Appl Pharmacol 262: 124-138.
http://dx.doi.Org/10.1016/i.taap.2012.04.026
Korallus. U: H-T. L: Neiss. A: Zwingers. T. (1982). [Relationships between hygienic measures and the
bronchial carcinoma mortality in the chromate producing industry], Arbeitsmed Sozialmed
Praventivmed 17: 159-167.
Korotzer. B: Ong. S: Hansen. IE. (2000). Ethnic differences in pulmonary function in healthy
nonsmoking Asian-Americans and European-Americans. Am J Respir Crit Care Med 161:
1101-1108. http://dx.doi.Org/10.1164/airccm.161.4.9902063
Koshi. K: Yagami. T: Nakanishi. Y. (1984). Cytogenetic analysis of peripheral blood lymphocytes
from stainless steel welders. Ind Health 22: 305-318.
http://dx.doi.org/10.2486/indhealth.22.305
Koutras. GA: Hattori. M: Schneider. AS: Ebaugh. FG: Valentine. WN. (1964). STUDIES ON
CHROMATED ERYTHROCYTES. EFFECT OF SODIUM CHROMATE ON ERYTHROCYTE
GLUTATHIONE REDUCTASE. J Clin Invest 43: 323-331.
http://dx.doi. org/10.1172 /TCT104917
Kovama. Y: Brenner. DA. (2017). Liver inflammation and fibrosis [Review], J Clin Invest 127: 55-64.
http://dx.doi.org/10.1172/TCI88881
Krause. PS: Theise. ND: Collector. MI: Henegariu. 0: Hwang. S: Gardner. R: Neutzel. S: Sharkis. ST.
(2001). Multi-organ, multi-lineage engraftmentby a single bone marrow-derived stem cell.
Cell 105: 369-377. http://dx.doi.org/10.1016/S0092-8674C01100328-2
Krawic. A: Luczak. M: Zhitkovich. A. (2017). Variation in extracellular detoxification is a link to
different carcinogenicity among chromates in rodent and human lungs. Chem Res Toxicol
30: 1720-1729. http://dx.doi.org/10.1021/acs.chemrestox.7b00172
Krim. M: Messaadia. A: Maidi. I: Aouacheri. 0: Saka. S. (2013). Protective effect of ginger against
toxicity induced by chromate in rats. Ann Biol Clin (Paris) 71: 165-173.
http://dx.doi.org/10.1684/abc.2013.0806
Kumar. A: Barthwal. R. (1991). Hexavalent chromium effects on hematological indices in rats. Bull
Environ Contam Toxicol 46: 761-768. http://dx.doi.org/10.1007/BFQ1689965
Kumar. A: Rana. SVS. (1982). Lipid accumulation in chromium poisoned rats. Int J Tissue React 4:
291-295.
Kumar. KM: Aruldhas. MM: Banu. SL: Sadasivam. B: Vengatesh. G: Ganesh. KM:
Navaneethabalakrishnan. S: Navin. AK: Michael. FM: Venkatachalam. S: Stanley. TA:
Ramachandran. I: Banu. SK: Akbarsha. MA. (2017). Male reproductive toxicity of CrVI: In-
utero exposure to CrVI at the critical window of testis differentiation represses the
expression of Sertoli cell tight junction proteins and hormone receptors in adult F1 progeny
rats. Reprod Toxicol 69: 84-98. http://dx.doi.Org/10.1016/i.reprotox.2017.02.007
Kumar. S: Sathwara. NG: Gautam. AK: Agarwal. K: Shah. B: Kulkarni. PK: Patel. K: Patel. A: Dave. LM:
Parikh. DT: Saived. HN. (2005). Semen quality of industrial workers occupationally exposed
to chromium. J Occup Health 47: 424-430.
Kunstvr. I: Peters. K: Gartner. K. (1976). Investigations on the function of the rat forestomach. Lab
Anim Sci 26: 166-170.
Kuo. HW: Chang. SF: Wu. KY: Wu. FY. (2003). Chromium (VI) induced oxidative damage to DNA:
increase of urinary 8-hydroxydeoxyguanosine concentrations (8-OHdG) among
electroplating workers. Occup Environ Med 60: 590-594.
Kuo. HW: Lai. IS: Lin. TI. (1997a). Concentration and size distribution of airborne hexavalent
chromium in electroplating factories. Am Ind Hyg Assoc J 58: 29-32.
http://dx.doi.org/10.1080/15428119791013044
This document is a draft for review purposes only and does not constitute Agency policy.
R-25 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Kuo. HW: Lai. IS: Lin. TI. (1997b). Nasal septum lesions and lung function in workers exposed to
chromic acid in electroplating factories. Int Arch Occup Environ Health 70: 272-276.
Kuo. HW: Wu. ML. (2002). Effects of chromic acid exposure on immunological parameters among
electroplating workers. Int Arch Occup Environ Health 75: 186-190.
Kurosaki. K: Nakamura. T: Mukai. T: Endo. T. (1995). Unusual findings in a fatal case of poisoning
with chromate compounds. Forensic Sci Int 75: 57-65.
Lacerda. LM: Garcia. SC: da Silva. LB: de Avila Dornelles. M: Presotto. AT: Lourenco. ED: de
Franceschi. ID: Fernandes. E: Wannmacher. CMP: Brucker. N: Sauer. E: Gioda. A: Machado.
AB: Oliveira. E: Trombini. TL: Feksa. LR. (2019). Evaluation of hematological, biochemical
parameters and thiol enzyme activity in chrome plating workers. Environ Sci Pollut Res Int
26: 1892-1901. http://dx.doi.org/10.1007/sl 1356-018-3755-7
Laskin. PL: Malaviva. R: Laskin. ID. (2019). Role of Macrophages in Acute Lung Injury and Chronic
Fibrosis Induced by Pulmonary Toxicants [Review], Toxicol Sci 168: 287-301.
http://dx.doi.org/10.1093/toxsci/kfy309
Lee. HS: Goh. CL. (1988). Occupational dermatosis among chrome platers. Contact Derm 18: 89-93.
http://dx.doi.Org/10.llll/i.1600-0536.1988.tb02746.x
Lee. YH: Su. SB: Huang. CC: Sheu. HM: Tsai. TC: Lin. CH: Wang. YT: Wang. BT. (2014). N-acetylcysteine
attenuates hexavalent chromium-induced hypersensitivity through inhibition of cell death,
ROS-related signaling and cytokine expression. PLoS ONE 9: el08317.
http://dx.doi.org/10.1371/journal.pone.0108317
Leininger. TR: Schutten. M. (2018). Chapter 4 - Oral cavity, teeth and gingiva. In AW Suttie (Ed.),
Boorman's pathology of the rat (2nd ed., pp. 15-21). Cambridge, MA: Academic Press.
http://dx.doi.Org/10.1016/B978-0-12-391448-4.00004-6
Leoz. ML: Carballal. S: Moreira. L: Ocana. T: Balaguer. F. (2015). The genetic basis of familial
adenomatous polyposis and its implications for clinical practice and risk management
[Review], The Application of Clinical Genetics 8: 95-107.
http://dx.doi.Org/l 0.2147 /TACG.S51484
Levina. A: Lay. PA. (2005). Mechanistic studies of relevance to the biological activities of chromium.
Coord Chem Rev 249: 281-298. http://dx.doi.Org/10.1016/i.ccr.2004.02.017
Levina. A: Pham. TH: Lay. PA. (2016). Binding of chromium(III) to transferrin could be involved in
detoxification of dietary chromium(III) rather than transport of an essential trace element.
Angew Chem Int Ed Engl 55: 8104-8107. http://dx.doi.org/10.1002/anie.2016Q2996
Lewis. RT. Sr: Hawlev. GG. (2007). Chromium. In RJ Lewis, Sr.; GG Hawley (Eds.), Hawley's
condensed chemical dictionary (15th ed., pp. 298-299). Hoboken, NJ: John Wiley & Sons.
http://dx.doi.Org/10.1002/9780470114735.hawlev03760
Li. C: Park. S: Zhang. X: Eisenberg. LM: Zhao. H: Darzvnkiewicz. Z: Xu. D. (2017). Nuclear Gene
33/Mig6 regulates the DNA damage response through an ATM serine/threonine kinase-
dependent mechanism. J Biol Chem 292: 16746-16759.
http://dx.doi.org/10.1074/ibc.Ml 17.803338
Li. H: Chen. 0: Li. S: Yao. W: Li. L: Shi. X: Wang. L: Castranova. V: Vallvathan. V: Ernst. E: Chen. C.
(2001). Effect of Cr(VI) exposure on sperm quality: human and animal studies. Ann Occup
Hyg45: 505-511.
Li. HC: Stoicov. C: Rogers. AB: Houghton. I. (2006). Stem cells and cancer: Evidence for bone marrow
stem cells in epithelial cancers [Review], World J Gastroenterol 12: 363-371.
http://dx.doi.org/10.3748/wig.vl2.i3.363
Li. LT: Tiang. G: Chen. 0: Zheng. IN. (2015a). Ki67 is a promising molecular target in the diagnosis of
cancer [Review], Mol Med Rep 11: 1566-1572. http://dx.doi.org/10.3892 /mmr.2014.2914
Li. P: Gu. Y: Yu. S: Li. Y: Yang. I: Tia. G. (2014a). Assessing the suitability of 8-OHdG and micronuclei
as genotoxic biomarkers in chromate-exposed workers: a cross-sectional study. BMJ Open
4: e005979. http://dx.doi.org/10.1136/bmiopen-2014-0Q5979
This document is a draft for review purposes only and does not constitute Agency policy.
R-26 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Li. P: Li. Y: Zhang. I. i: Yu. SF. a: Tong. W. ei: Hu. X: Tia. G. (2015b). Biomarkers for Lung Epithelium
Injury in Occupational Hexavalent Chromium-Exposed Workers. J Occup Environ Med 57:
E45-E50. http://dx.doi.Org/10.1097/TOM.0000000000000436
Li. Y: Hu. G: Li. P: Tang. S: Zhang. 1: Tia. G. (2016). miR-3940-5p enhances homologous
recombination after DSB in Cr(VI) exposed 16HBE cell. Toxicology 344-346: 1-6.
http://dx.doi.Org/10.1016/i.tox.2016.02.003
Li. Y: Li. P: Yu. S: Zhang. 1: Wang. T: Tia. G. (2014b). miR-3940-5p associated with genetic damage in
workers exposed to hexavalent chromium. Toxicol Lett 229: 319-326.
http://dx.doi.Org/10.1016/j.toxlet2014.06.033
Liang. 0: Xiao. Y: Liu. K: Zhong. C: Zeng. M: Xiao. F. (2018a). Cr(VI)-induced autophagy protects L-02
hepatocytes from apoptosis through the ROS-AKT-mTOR pathway. Cell Physiol Biochem 51:
1863-1878. http://dx.doi.org/10.1159/000495713
Liang. 0: Zhang. Y: Huang. M: Xiao. Y: Xiao. F. (2019). Role of mitochondrial damage in
Cr(VI)-induced endoplasmic reticulum stress in L-02 hepatocytes. Mol Med Rep 19: 1256-
1265. http://dx.doi.org/10.3892/mmr.2018.97Q4
Liang. 0: Zhang. Y: Zeng. M: Guan. L: Xiao. Y: Xiao. F. (2018b). The role of IP3R-S0CCs in Cr(VI)-
induced cytosolic Ca2+ overload and apoptosis in L-02 hepatocytes. Toxicology Research 7:
521-528. http://dx.doi.org/10.1039/c8tx00029h
Lide. D. (2008). Chromium and compounds. In CRC handbook of chemistry and physics. Boca Raton,
FL: CRC Press.
Lin. D: Triadafilopoulos. G. (2015). Dual ambulatory pH monitoring in patients with
gastroesophageal reflux rendered asymptomatic with proton pump inhibitor therapy. Dig
Dis Sci 60: 1343-1349. http://dx.doi.org/10.1007/sl0620-014-3324-v
Lin. SC: Tai. CC: Chan. CC: Wang. ID. (1994). Nasal septum lesions caused by chromium exposure
among chromium electroplating workers. Am J Ind Med 26: 221-228.
http://dx.doi.Org/10.1002/aiim.4700260207
Lindberg. E: Hedenstierna. G. (1983). Chrome plating: Symptoms, findings in the upper airways, and
effects on lung function. Arch Environ Occup Health 38: 367-374.
Lindberg. M: Forslind. B: Wahlberg. IE. (1982). Reactions of epidermal keratinocytes in sensitized
and non-sensitized guinea pigs after dichromate exposure: an electron microscopic study.
Acta Derm Venereol 62: 389-396.
Lindemann. M: Rietschel. F: Zabel. M: Grosse-Wilde. H. (2008). Detection of chromium allergy by
cellular in vitro methods. Clin Exp Allergy 38: 1468-1475.
http://dx.doi.Org/10.llll/i.1365-2222.2008.02970.x
Linos. A: Petralias. A: Christophi. CA: Christoforidou. E: Kouroutou. P: Stoltidis. M: Veloudaki. A:
Tzala. E: Makris. KC: Karagas. MR. (2011). Oral ingestion of hexavalent chromium through
drinking water and cancer mortality in an industrial area of Greece-an ecological study.
Environ Health 10:50. http://dx.doi.org/10.1186/1476-069X-10-5Q
Linqing. Y: Bo. X: Xueqin. Y: Hong. D: Desheng. W: Huimin. Z: Gaofeng. I: Tianiun. L: Zhixiong. Z.
(2016). Mitochondrial DNAhypomethylation in chrome plating workers. Toxicol Lett 243:
1-6. http://dx.doi.Org/10.1016/i.toxlet.2015.ll.031
Lischka. G. (1971). [Lymphocyte transformation test in chromium-hypersensitivity], 240: 212-218.
Littorin. M: Hogstedt. B: Stromback. B: Karlsson. A: Welinder. H: Mitelman. F: Skerfving. S. (1983).
No cytogenetic effects in lymphocytes of stainless steel welders. Scand J Work Environ
Health 9: 259-264.
Locke. WT: Guanzon. D: Ma. C: Liew. YT: Duesing. KR: Fung. K: Ross. TP. (2019). DNA methylation
cancer biomarkers: Translation to the clinic [Review], Front Genet 10: 1150.
http://dx.doi.org/10.3389/fgene.2019.01150
Long. C: Liu. 1: Hu. G: Feng. H: Zhou. D: Wang. I: Zhai. X: Zhao. Z: Yu. S: Wang. T: Tia. G. (2019).
Modulation of homologous recombination repair gene polymorphisms on genetic damage in
This document is a draft for review purposes only and does not constitute Agency policy.
R-27 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
chromate exposed workers. Environ Toxicol Pharmacol 66: 126-132.
http://dx.doi.Org/10.1016/i.etap.2019.01.004
Losi. ME: Amrhein. C: Frankenberger. WT. (1994). Environmental biochemistry of chromium
[Review], Rev Environ Contam Toxicol 136: 91-121. http://dx.doi.org/10.1007/978-l-
4612-2656-7 3
Loubieres. Y: de Lassence. A: Bernier. M: Vieillard-Baron. A: Schmitt. TM: Page. B: Tardin. F. (1999).
Acute, fatal, oral chromic acid poisoning. J Toxicol Clin Toxicol 37:333-336.
Lu. 1: Tang. M: Liu. Y: Wang. 1: Wu. Z. (2018). Comparative Proteomics of Chromium-Transformed
Beas-2B Cells by 2D-DIGE and MALDI-TOF/TOF MS. Biol Trace Elem Res.
http://dx.doi. org/10.1007 /si 2011-017-1222-9
Lucas. IB. (1976). Health Hazard Evaluation Determination, ReportNo. HHE-74-111-283, Cook
Paint and Varnish Company, North Kansas City, Missouri, April 1976 (pp. 74-111).
(NIOSH/00099446). Lucas, JB.
Lucas. IB: Kramkowski. RS. (1975). Health Hazard Evaluation ReportNo. HHE-74-87-221, Industrial
Platers, Inc., Columbus, Ohio (pp. 74-87). (NIOSH/00049983). Cincinnati, OH: National Inst,
for Occupational Safety and Health.
Luczak. MW: Green. SE: Zhitkovich. A. (2016). Different ATM Signaling in Response to
Chromium(VI) Metabolism via Ascorbate and Nonascorbate Reduction: Implications for in
Vitro Models andToxicogenomics. Environ Health Perspect 124: 61-66.
http://dx.doi.org/10.1289/ehp.1409434
Luippold. RS: Mundt. KA: Dell. LP: Birk. T. (2005). Low-level hexavalent chromium exposure and
rate of mortality among US chromate production employees. J Occup Environ Med 47: 381-
385. http://dx.doi.Org/10.1097/01.iom.0000158703.32263.0d
Lum. G. (1995). Significance of low serum alkaline phosphatase activity in a predominantly adult
male population. Clin Chem 41: 515-518.
Luo. D: Xu. Z: Hu. X: Zhang. F: Bian. H: Li. N: Wang. 0: Lu. Y: Zheng. 0: Gu. I. (2016). URI prevents
potassium dichromate-induced oxidative stress and cell death in gastric cancer cells. 8:
5399-5409.
Lupescu. A: Tilani. K: Zelenak. C: Zbidah. M: Oadri. SM: Lang. F. (2012). Hexavalent chromium-
induced erythrocyte membrane phospholipid asymmetry. Biometals 25: 309-318.
http://dx.doi.Org/10.1007/sl0534-011-9507-5
Luster. MI: Portier. C: Pait. DG: Rosenthal. GT: Germolec. PR: Corsini. E: Blavlock. BL: Pollock. P:
Kouchi. Y: Craig. W. (1993). Risk assessment in immunotoxicology. II. Relationships
between immune and host resistance tests. Fundam Appl Toxicol 21: 71-82.
Luster. MI: Portier. C: Pait. PG: White. KL. Tr: Gennings. C: Munson. AE: Rosenthal. GT. (1992). Risk
assessment in immunotoxicology. I. Sensitivity and predictability of immune tests. Fundam
Appl Toxicol 18: 200-210. http://dx.doi.org/10.1016/0272-0590r92190047-L
Lv. Y: Zhang. P: Guo. I: Zhu. Z: Li. X: Xu. P: Zeng. W. (2018). Melatonin protects mouse
spermatogonial stem cells against hexavalent chromium-induced apoptosis and epigenetic
histone modification. Toxicol Appl Pharmacol 340: 30-38.
http://dx.doi.Org/10.1016/i.taap.2017.12.017
Macgregor. IT: Wehr. CM: Henika. PR: Shelby. MP. (1990). The in vivo erythrocyte micronucleus
test: Measurement at steady state increases assay efficiency and permits integration with
toxicity studies. Fundam Appl Toxicol 14: 513-522. http://dx.doi.org/10.1016/Q272-
0590C90190255-T
Machle. W: Gregorius. F. (1948). Cancer of the respiratory system in the United States chromate-
producing industry. Public Health Rep 63: 1114-1127. http://dx.doi.org/10.23Q7/4586677
MacKenzie. RP: Bverrum. RU: Pecker. CF: Hoppert. CA: Langham. RF. (1958). Chronic toxicity
studies. II. Hexavalent and trivalent chromium administered in drinking water to rats. AMA
Arch Ind Health 18: 232-234.
This document is a draft for review purposes only and does not constitute Agency policy.
R-28 PRAFT-PO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Maeng. SH: Chung. HW: Kim. KT: Lee. BM: Shin YC Kim. ST: Yu. IT. (2004). Chromosome aberration
and lipid peroxidation in chromium-exposed workers. Biomarkers 9: 418-434.
http://dx.doi.Org/10.1080/13547500400022200
Maeng. SH: Chung. HW: Yu. IT: Kim. HY: Lim. CH: Kim. KT: Kim. ST: Ootsuvama. Y: Kasai. H. (2003).
Changes of 8-OH-dG levels in DNA and its base excision repair activity in rat lungs after
inhalation exposure to hexavalent chromium. MutatRes 539: 109-116.
http://dx.doi. org/10.1016/SI383-5718C03100154-2
Malsch. PA: Proctor. DM: Finlev. BL. (1994). Estimation of a chromium inhalation reference
concentration using the benchmark dose method: A case study [Review], Regul Toxicol
Pharmacol 20: 58-82. http://dx.doi.org/10.1006/rtph.1994.1036
Mancuso. TF. (1951). Occupational cancer and other health hazards in a chromate plant: A medical
appraisal. II. Clinical and toxicologic aspects. Ind Med Surg 20: 393-407.
Mancuso. TF. (1975). Consideration of chromium as an industrial carcinogen. In International
conference on heavy metals in the environment: v III, health; October. Toronto, ON, Canada:
Institute for Environmental Studies.
Mancuso. TF. (1997). Chromium as an industrial carcinogen: Parti. Am J Ind Med 31: 129-139.
http://dx.doi. org/10.1002 /CSTCI11097-0274C199702131:2<129::ATD-ATTM1 >3.0.CQ:2-V
Mandervelt. C: Clottens. FL: Demedts. M: Nemerv. B. (1997). Assessment of the sensitization
potential of five metal salts in the murine local lymph node assay. Toxicology 120: 65-73.
Marat. I: Arstan. M: Galvmzhan. Y: Timur. I: Yerbolat. I: Almasbek. Y. (2018). Impact of chromium
and boron compounds on the reproductive function in rats. Toxicol Ind Health 34: 365-374.
http://dx.doi. org/10.1177/0748233718759162
Margonis. GA: Kim. Y: Spolverato. G: Eiaz. A: Gupta. R: Cosgrove. D: Anders. R: Karagkounis. G: Choti.
MA: Pawlik. TM. (2015). Association between specific mutations in KRAS codon 12 and
colorectal liver metastasis. JAMA Surgery 150: 722-729.
http://dx.doi.org/10.1001/iamasurg.2015.0313
Marin. RA: Bose. R: Dabrowski. B: Kolesnik. S. (2018). Magnetic Characterization of Chromium
Intermediates in the Reduction of Chromium (VI) by Glutathione in Acidic Solutions. 4.
http://dx.doi.Org/10.3390/magnetochemistry4020023
Marouani. N: Tebourbi. 0: Hallegue. D: Mokni. M: Yacoubi. MT: Saklv. M: Benkhalifa. M: Rhouma. KB.
(2015a). Mechanisms of chromium hexavalent-induced apoptosis in rat testes. Toxicol Ind
Health 33: 97-106. http://dx.doi.org/10.1177/0748233715600333
Marouani. N: Tebourbi. 0: Mahioub. S: Yacoubi. MT: Saklv. M: Benkhalifa. M: Rhouma. KB. (2012).
Effects of hexavalent chromium on reproductive functions of male adult rats. Reprod Biol
12: 119-133.
Marouani. N: Tebourbi. 0: Mokni. M: Yacoubi. MT: Saklv. M: Benkhalifa. M: Ben Rhouma. K. (2015b).
Hexavalent chromium-induced apoptosis in rat uterus: Involvement of oxidative stress.
Arch Environ Occup Health 70: 189-195.
http: //dx.doi.org/10.1080/19338244.2013.828673
Martin. BP: Schoenhard. TA: Hwang. TM: Sugden. KD. (2006). Ascorbate is a pro-oxidant in
chromium-treated human lung cells. Mutat Res 610: 74-84.
http://dx.doi.Org/10.1016/i.mrgentox.2006.06.014
Martino. 1: Holmes. AL: Xie. H: Wise. SS: Wise. TP. (2015). Chronic Exposure to Particulate Chromate
Induces Premature Centrosome Separation and Centriole Disengagement in Human Lung
Cells. Toxicol Sci 147: 490-499. http://dx.doi.org/10.1093/toxsci/kfvl46
Marvnka-Kalmani. K: Treves. S: Yafee. M: Rachima. H: Gafni. Y: Cohen. MA: Pitaru. S. (2010). The
lamina propria of adult human oral mucosa harbors a novel stem cell population. Stem Cells
28: 984-995. http://dx.doi.org/10.1002 /stem.425
Mathebula. MW: Mandiwana. K: Panichev. N. (2017). Speciation of chromium in bread and breakfast
cereals. Food Chem 217: 655-659. http://dx.doi.Org/10.1016/i.foodchem.2016.09.020
This document is a draft for review purposes only and does not constitute Agency policy.
R-29 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
McClain. CN: Fendorf. S: Webb. SM: Maher. K. (2017). Quantifying Cr(VI) production and export
from serpentine soil of the California coast range. Environ Sci Technol 51: 141-149.
http://dx.doi.org/10.1021/acs.est.6b03484
Mcconnell. EL: Basit. AW: Murdan. S. (2008). Measurements of rat and mouse gastrointestinal pH,
fluid and lymphoid tissue, and implications for in-vivo experiments. J Pharm Pharmacol 60:
63-70. http://dx.doi.0rg/lO.i2il/ipp.6O.l.OOO8
Mckinzie. PB: Parsons. BL. (2002). Detection of rare K-ras codon 12 mutations using allele-specific
competitive blocker PCR. MutatRes 517: 209-220. http://dx.doi.org/10.1016/sl383-
5718f02100077-3
Mckinzie. PB: Parsons. BL. (2011). Accumulation of K-Ras codon 12 mutations in the F344 rat distal
colon following azoxymethane exposure. Environ Mol Mutagen 52: 409-418.
http://dx.doi.org/10.1002/em.2Q644
Medeiros. MG: Rodrigues. AS: Batoreu. MC: Laires. A: Rueff. 1: Zhitkovich. A. (2003). Elevated levels
of DNA-protein crosslinks and micronuclei in peripheral lymphocytes of tannery workers
exposed to trivalent chromium. Mutagenesis 18: 19-24.
Meenakshi. CE: Padmini. E: Motlag. DB. (1989). Comparative toxicity of trivalent and hexavalent
chromium in rats. Indian J Environ Health 31: 250-256.
Meng. F: Knapp. GW: Green. T: Ross. TA: Parsons. BL. f2010I K-Ras mutant fraction in A/J mouse
lung increases as a function of benzo[a]pyrene dose. Environ Mol Mutagen 51: 146-155.
http://dx.doi.org/10.1002/em.20513
Mezencev. R: Auerbach. SS. (2021). Inferred inactivation of the Cftr gene in the duodena of mice
exposed to hexavalent chromium (Cr(VI)) in drinking water supports its tumor-suppressor
status and implies its potential role in Cr(VI)-induced carcinogenesis of the small intestines.
Toxicol Appl Pharmacol 433: 115773. http://dx.doi.Org/10.1016/i.taap.2021.115773
Migliore. L: Parrini. M: Sbrana. I: Biagini. C: A: Loprieno. N. (1991). Micronucleated lymphocytes in
people occupationally exposed to potential environmental contaminants: The age effect
MutatRes 256: 13-20. http://dx.doi.Org/doi:10.1016/0921-8734r91190028-A
Mignini. F: Streccioni. V: Baldo. M: Vitali. M: Indraccolo. U: Bernacchia. G: Cocchioni. M. (2004).
Individual susceptibility to hexavalent chromium of workers of shoe, hide, and leather
industries. Immunological pattern of HLA-B8, DR3-positive subjects. Prev Med 39: 767-775.
http://dx.doi.Org/10.1016/i.ypmed.2004.02.048
Mignini. F: Tomassoni. D: Traini. E: Vitali. M: Scuri. S: Baldoni. E: Grappasonni. I: Cocchioni. M.
(2009). Immunological pattern alteration in shoe, hide, and leather industry workers
exposed to hexavalent chromium. Environ Toxicol 24: 594-602.
http://dx.doi.org/10.1002/tox.20464
Miksche. LW: Lewalter. I. (1997). Health surveillance and biological effect monitoring for
chromium-exposed workers [Review], Regul Toxicol Pharmacol 26: S94-S99.
http://dx.doi.org/10.1006/rtph.1997.1146
Miller. AC: Cornelias. AP: Hornick. DB: Stoltz. DA: Cavanaugh. IE: Gerke. AK: Welsh. Ml: Zabner. I:
Polgreen. PM. (2020). Cystic fibrosis carriers are at increased risk for a wide range of cystic
fibrosis-related conditions. Proc Natl Acad Sci USA 117: 1621-1627.
http://dx.doi.org/10.1073/pnas.1914912117
Mills. TC: Shivdasani. RA. (2011). Gastric epithelial stem cells [Review], Gastroenterology 140: 412-
424. http://dx.doi.0rg/lO.lO53/i.gastro.2OlO.i2.OOl
Mirsalis. TC: Hamilton. CM: O'Loughlin. KG: Paustenbach. DT: Kerger. BP: Patierno. S. (1996).
Chromium (VI) at plausible drinking water concentrations is not genotoxic in the in vivo
bone marrow micronucleus or liver unscheduled DNA synthesis assays. Environ Mol
Mutagen 28: 60-63. http://dx.doi.org/10.1002/rSICni098-2280ri996128:l<60::AID-
EM9>3.0.CO;2-I
This document is a draft for review purposes only and does not constitute Agency policy.
R-30 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Mitrov. D: Hadzi-Petrushev. N: Stojkovski. V: Gjorgievska. E: Gagov. H: Mladenov. M. (2014).
Influence of chronic chromium exposition on the processes of lipid peroxidation
inflammation and platelet activation in rats [Letter], J Biol Regul Homeost Agents 28: 531-
535.
Miyamoto. T: Ichikawa. K: Kabe. 1: Horiuchi. Y. (1975). Delayed-type hypersensitivity reaction in the
lungs of guinea pigs due to potassium dichromate. J Allergy Clin Immunol 56: 464-472.
Mo. 1: Tang. C: Wang. X: Lu. W: Tong. H. ui: Zhang. C: Zhang. G. (2018). Histopathological analysis of
ductular reaction in rabbit liver after chronic exposure of chromium (VI). Acta med
Mediterr 34: 1213-1219. http://dx.doi.org/10.19193 /0393-6384 2018 5 186
Monserrat-Mesquida. M: Ouetglas-Llabres. M: Abbate. M: Montemavor. S: Mascaro. CM: Casares. M:
Teiada. S: Abete. I: Zulet. MA: Tur. TA: Martinez. TA: Sureda. A. (2020). Oxidative stress and
pro-inflammatory status in patients with non-alcoholic fatty liver disease. Antioxidants 9:
759. http://dx.doi.org/10.3390/antiox9080759
Mor. S: Ben-Efraim. S: Leibovici. I: Ben-David. A. (1988). Successful contact sensitization to
chromate in mice. International Arch Allergy Appl Immunol 85: 452-457.
http://dx.doi.Org/10.1159/000234550
Morihara. M: Aovagi. N: Kaniwa. N: Koiima. S: Ogata. H. (2001). Assessment of gastric acidity of
Japanese subjects over the last 15 years. Biol Pharm Bull 24: 313-315.
http://dx.doi.org/10.1248/bpb.24.313
Moschandreas. DT: Karuchit. S: Berry. MR: O'Rourke. MK: Lo. D: Lebowitz. MP: Robertson. G. (2002).
Exposure apportionment: ranking food items by their contribution to dietary exposure. J
Expo Anal Environ Epidemiol 12: 233-243. http://dx.doi.org/10.1038/si.iea.7500230
Moulin. IT: Clavel. T: Roy. D: Dananche. B: Marquis. N: Fevotte. 1: Fontana. TM. (2000). Risk of lung
cancer in workers producing stainless steel and metallic alloys. Int Arch Occup Environ
Health 73: 171-180. http://dx.doi.Org/10.1007/s004200050024
Moulin. II: Portefaix. P: Wild. P: Mur. TM: Smagghe. G: Mantout. B. (1990). Mortality study among
workers producing ferroalloys and stainless steel in France. Br J Ind Med 47: 537-543.
http://dx.doi.Org/10.1136/oem.47.8.537
Moulin. IT: Wild. P: Haguenoer. TM: Faucon. D: De Gaudemaris. R: Mur. TM: Mereau. M: Gary. Y:
Toamain. TP: Birembaut. Y. (1993a). A mortality study among mild steel and stainless steel
welders. Br J Ind Med 50: 234-243. http://dx.doi.Org/10.1136/oem.50.3.234
Moulin. IT: Wild. P: Mantout. B: Fournier-Betz. M: Mur. TM: Smagghe. G. (1993b). Mortality from lung
cancer and cardiovascular diseases among stainless-steel producing workers. Cancer
Causes Control 4: 75-81. http://dx.doi.org/10.1007/BF00053147
Mozafari. P: Rezazadeh Azari. M: Shokoohi. Y: Savadi. M. (2016). Feasibility of biological effective
monitoring of chrome electroplaters to chromium through analysis of serum
malondialdehyde. Int J Occup Environ Med 7: 199-206.
http://dx.doi.org/10.15171/iioem.2016.782
Mudie. DM: Amidon. GL: Amidon. GE. (2010). Physiological parameters for oral delivery and in vitro
testing [Review], Mol Pharm 7: 1388-1405. http:/ /dx.doi.org/10.1021 /mp 100149i
Mukheriee. P: Poddar. S: Talukder. G: Sharma. A. (1999). Protection by black tea extract against
chromosome damage induced by two heavy metals in mice. Pharmaceutical Biology 37:
243-247. http://dx.doi.Org/10.1076/phbi.37.3.243.6309
Mukheriee. P: Sarkar. D: Sharma. A. (1997). Effects of dietary consumption of black tea infusion
alone and in combination with known clastogens on mouse bone marrow chromosomes in
vivo. Food Chem Toxicol 35: 657-661. http://dx.doi.org/10.1016/s0278-6915(97)00044-6
Murphy. K: Weaver. C. (2016). Janeway's immunobiology (9th ed.). New York, NY: Garland Science.
Murthv. RC: Tunaid. M: Saxena. DK. (1996). Ovarian dysfunction in mice following chromium (VI)
exposure. Toxicol Lett 89:147-154. http://dx.doi.org/10.1016/S0378-4274(96)03803-9
This document is a draft for review purposes only and does not constitute Agency policy.
R-31 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Murthv. RC: Saxena. DK: Gupta. SK: Chandra. SV. (1991). Ultrastructural observations in testicular
tissue of chromium-treated rats. Reprod Toxicol 5: 443-447.
http://dx.doi.org/10.1016/0890-6238r91190008-4
Myers. CR: Antholine. WE: Myers. 1M. (2010). The pro-oxidant chromium(VI) inhibits mitochondrial
complex I, complex II, and aconitase in the bronchial epithelium: EPR markers for Fe-S
proteins. Free Radic Biol Med 49: 1903-1915.
http://dx.doi.Org/10.1016/i.freeradbiomed.2010.09.020
Myers. TM: Antholine. WE: Myers. CR. (2008). Hexavalent Chromium Causes the Oxidation of
Thioredoxin in Human Bronchial Epithelial Cells. Toxicology 246(2-3): 222-233.
http://dx.doi.Org/10.1016/i.tox.2008.01.017
Myers. TM: Antholine. WE: Myers. CR. (2011). The intracellular redox stress caused by hexavalent
chromium is selective for proteins that have key roles in cell survival and thiol redox
control. Toxicology 281: 37-47. http://dx.doi.Org/10.1016/i.tox.2011.01.001
Myers. TM: Myers. CR. (2009). The effects of hexavalent chromium on thioredoxin reductase and
peroxiredoxins in human bronchial epithelial cells. Free Radic Biol Med 47: 1477-1485.
http://dx.doi.Org/10.1016/i.freeradbiomed.2009.08.015
Nature. (2019). Cytokines in the balance [Editorial], Nat Immunol 20: 1557.
http://dx.doi. org/10.1038/s41590-019-0557-0
Navin. AK: Aruldhas. MM: Navaneethabalakrishnan. S: Mani. K: Michael. FM: Srinivasan. N: Banu. SK.
(2021). Prenatal exposure to hexavalent chromium disrupts testicular steroidogenic
pathway in peripubertalFl rats. Reprod Toxicol 101: 63-73.
http://dx.doi.Org/10.1016/i.reprotox.2021.01.014
Navva. K: Kumar. GP: Anilakumar. KR. (2017a). Ameliorating effect of Curculigo orchoides on
chromium(VI) induced oxidative stress via, modulation of cytokines, transcription factors
and apoptotic genes. J Appl Biomed 15: 299-306.
http://dx.doi.Org/10.1016/i.iab.2017.03.003
Navva. K: Phani Kumar. G: Chandrasekhar. Y: Anilakumar. KR. (2017b). Evaluation of potassium
dichromate (K2Cr207)-induced liver oxidative stress and ameliorative effect of Picrorhiza
kurroa extract in Wistar albino rats. Biol Trace Elem Res 184: 154-164.
http://dx.doi. org/10.1007 /si 2011 -017-1172-2
Neal-Kluever. A: Fisher. 1: Grvlack. L: Kakiuchi-Kivota. S: Halpern. W. (2019). Physiology of the
Neonatal Gastrointestinal System Relevant to the Disposition of Orally Administered
Medications. Drug Metab Dispos 47: 296-313. http://dx.d0i.0rg/l0.1124/dmd. 118.084418
Negrini. S: Gorgoulis. VG: Halazonetis. TP. (2010). Genomic instability — an evolving hallmark of
cancer [Review], Nat Rev Mol Cell Biol 11: 220-228. http: //dx.doi.org/10.1038/nrm2858
Nettesheim. P: Hanna. MG. Tr: Doherty. DG: Newell. RF: Hellman. A. (1971). Effect of calcium
chromate dust, influenza virus, and 100 R whole-body X radiation on lung tumor incidence
in mice. J Natl Cancer Inst 47: 1129-1144.
Nettesheim. P: Hanna. MG. Tr: Doherty. DG: Newell. RF: Hellman. A. (1970). Effect of chronic
exposure to air pollutants on the respiratory tracts of mice: Histopathological findings. In P
Nettesheim; MG Hanna, Jr.; JW Deatherage, Jr. (Eds.), Morphology of experimental
respiratory carcinogenesis: proceedings of a biology division, Oak Ridge National
Laboratory conference; May; Gatlinburg, TN. Oak Ridge, TN: U.S. Atomic Energy
Commission.
Ng. AW: Bidani. A: Heming. TA. (2004). Innate host defense of the lung: effects of lung-lining fluid
pH [Review], Lung 182: 297-317. http://dx.doi.Org/10.1007/s00408-004-2511-6
Nguyen. HT: Duong. HO. (2018). The molecular characteristics of colorectal cancer: Implications for
diagnosis and therapy [Review], Oncol Lett 16: 9-18.
http://dx.doi.org/10.3892/ol.2018.8679
This document is a draft for review purposes only and does not constitute Agency policy.
R-32 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
Nickens. KP: Patierno. SR: Cervak. S. (2010). Chromium genotoxicity: A double-edged sword
[Review], Chem Biol Interact xx: xxx. http://dx.doi.Org/10.1016/i.cbi.2010.04.018
Nielsen. 1: Dahlqvist. M: Welinder. H: Thomassen. Y: Alexandersson. R: Skerfving. S. (1993). Small
airways function in aluminium and stainless steel welders. Int Arch Occup Environ Health
65: 101-105. http://dx.doi.Org/10.1007/BF00405727
NTH (National Institutes ofHealth). (2017). Dietary supplement fact sheet: Chromium — health
professional fact sheet Available online athttps://ods.od.nih.gov/factsheets/Chromium-
HealthProfessional/ (accessed June 12, 2017).
Niis. M: Kirsch-Volders. M. (1986). Induction of spindle inhibition and abnormal mitotic figures by
Cr(II), Cr(III) and Cr(VI) ions. Mutagenesis 1: 247-252.
http://dx.doi.Org/10.1093/mutage/l.4.247
Nikula. KT: Mccartnev. IE: Mcgovern. T: Miller. GK: Odin. M: Pino. MY: Reed. MP. (2014). STP
Position Paper: Interpreting the Significance of Increased Alveolar Macrophages in Rodents
Following Inhalation of Pharmaceutical Materials. Toxicol Pathol 42: 472-486.
http://dx.doi.Org/10.1177/0192623313507003
Ning. 1: Grant. MH. (1999). Chromium (Vl)-induced cytotoxicity to osteoblast-derived cells. Toxicol
In Vitro 13: 879-887.
Ning. 1: Grant. MH. (2000). The role of reduced glutathione and glutathione reductase in the
cytotoxicity of chromium (VI) in osteoblasts. Toxicol In Vitro 14: 329-335.
Ning. 1: Henderson. C: Grant. MH. (2002). The cytotoxicity of chromium in osteoblasts: Effects on
macromolecular synthesis. J Mater Sci Mater Med 13: 47-52.
http://dx.doi. org/10.1023 /A:l 013630401959
NIOSH (National Institute for Occupational Safety and Health). (1972). Chromic acid mist in air. In
Physical and Chemical Analysis Branch. (NIOSH/00051221). Rockville, MD.
NIOSH (National Institute for Occupational Safety and Health). (2013). Criteria for a recommended
standard: Occupational exposure to hexavalent chromium. (NIOSH Publication Number
2013-128). Centers for Disease Control and Prevention.
https://www.cdc.gov/niosh/docs/2Q13-
128/pdfs/2013 128.pdf?id=l 0.26616/NIOSHPUB2Q13128
Nohmi. T: Katoh. M: Suzuki. H: Matsui. M: Yamada. M: Watanabe. M: Suzuki. M: Horiva. N: Ueda. 0:
Shibuva. T: Ikeda. H: Sofuni. T. (1996). A new transgenic mouse mutagenesis test system
using Spi- and 6-thioguanine selections. Environ Mol Mutagen 28: 465-470.
http://dx.doi.org/10.1002 /fSICPl 098-2280fl 996"I28:4<465::AID-EM24>3.0.CO:2-C
Nolte. T: Brander-Weber. P: Dangler. C: Deschl. U: Elwell. MR: Greaves. P: Hailev. R: Leach. MW:
Pandiri. AR: Rogers. A: Shackelford. CC: Spencer. A: Tanaka. T: Ward. TM. (2016).
Nonproliferative and proliferative lesions of the gastrointestinal tract, pancreas and salivary
glands of the rat and mouse [Review], J Toxicol Pathol 29(1 Suppl.): 1S-125S.
http://dx.doi.org/10.1293/tox.29.lS
Nordberg. GF: Fowler. BA: Nordberg. M. (2015). Handbook on the toxicology of metals. In GF
Nordberg; BA Fowler; M Nordberg (Eds.), (4th ed.). Burlington, MA: Elsevier Science.
http://dx.doi. org/10.1016/C2011 -0-07884-5
Norppa. H: Bonassi. S: Hansteen. IL: Hagmar. L: Stromberg. U: Rossner. P: Boffetta. P: Lindholm. C:
Gundy. S: Lazutka. 1: Cebulska-Wasilewska. A: Fabianova. E: Sram. RT: Knudsen. LE: Barale.
R: Fucic. A. (2006). Chromosomal aberrations and SCEs as biomarkers of cancer risk
[Review], MutatRes 600: 37-45. http://dx.doi.Org/10.1016/i.mrfmmm.2006.05.030
NRC (National Research Council). (1988). Health risks of radon and other internally deposited
alpha emitters: BEIRIV. Washington, DC.
http://books.nap.edu/openbook.php7record id=1026&page=Rl
This document is a draft for review purposes only and does not constitute Agency policy.
R-33 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
NRC (National Research Council). (2001). Standing operating procedures for developing acute
exposure guideline levels (AEGLs) for hazardous chemicals. Washington, DC: National
Academies Press, http://dx.doi.org/10.17226/10122
NRC (National Research Council). (2011). Review of the Environmental Protection Agency's draft
IRIS assessment of formaldehyde (pp. 1-194). Washington, DC: The National Academies
Press, http://dx.doi.org/10.17226/13142
Nriagu. TO. (1989). A global assessment of natural sources of atmospheric trace metals. Nature 338:
47-49. http://dx.doi.org/10.1038/338047a0
Nriagu. 10: Pacvna. TM. (1988). Quantitative assessment of worldwide contamination of air, water
and soils by trace metals. Nature 333: 134-139. http://dx.doi.org/10.1038/333134a0
NTP (National Toxicology Program). (1996a). Final report on the reproductive toxicity of potassium
dichromate (hexavalent) (CAS No. 7778-50-9) administered in dietto BALB/c mice [NTP],
Research Triangle Park: U.S. Department of Health and Human Services, Public Health
Service.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/PB97125363.xhtml
NTP (National Toxicology Program). (1996b). Final report on the reproductive toxicity of
potassium dichromate (hexavalent) (CAS No. 7778-50-9) administered in dietto SD rats
[NTP], Research Triangle Park, NC: U.S. Department of Health and Human Services, Public
Health Service.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/PB97125355.xhtml
NTP (National Toxicology Program). (1997). Final report on the reproductive toxicity of potassium
dichromate (CAS No. 7778-50-9) administered in dietto BALB/c mice [NTP], Research
Triangle Park, NC: U.S. Department of Health and Human Services, Public Health Service.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/PB97144919.xhtml
NTP. (2005). Immunotoxicity of hexavalent chromium in female B6C3F1 mice. Protocol: RF-HCR-
28-1 M-DW. RTP: NIEHS.
NTP. (2006a). Immunotoxicity of hexavalent chromium in female Fischer 344 rats. Protocol: RF-
HCR-28-2R-DW. RTP: NIEHS.
NTP. (2006b). Immunotoxicity of hexavalent chromium in female Sprague Dawley rats. Protocol:
RF-HCR-28-1R-DW. RTP: NIEHS.
NTP (National Toxicology Program). (2007). NTP technical report on the toxicity studies of sodium
dichromate dihydrate (CAS No. 7789-12-0) administered in drinking water to male and
female F344/N rats and B6C3F1 mice and male BALB/c and am3-C57BL/6 mice [NTP],
(Toxicity Report Series 72). Research Triangle Park, NC.
https://ntp.niehs.nih.gov/ntp/htdocs/st rpts/tox072.pdf
NTP (National Toxicology Program). (2008). Toxicology and carcinogenesis studies of sodium
dichromate dihydrate (Cas No. 7789-12-0) in F344/N rats and B6C3F1 mice (drinking
water studies) (pp. 1-192). (NTP TR 546; NIH Publication No. 08-5887).
NTP (National Toxicology Program). (2010). NTP toxicology and carcinogenesis studies of
chromium picolinate monohydrate (CAS No. 27882-76-4) in F344/N rats and B6C3F1 mice
(feed studies). (NTP TR 556; NIH Publication No. 10-5897). Research Triangle Park, NC: U.S.
Department of Health and Human Services, Public Health Service, National Institutes of
Health, National Toxicology Program, https: //ntp.niehs.nih.gov/go/tr556abs
NTP (National Toxicology Program). (2011). Report on carcinogens. Twelfth Edition, 2011 (12th
ed., pp. 507). Research Triangle Park, NC: U.S. Department of Health and Human Services,
Public Health Service, National Institutes of Health.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/PB2011111646.xhtml
NTP (National Toxicology Program). (2020). Organ sites with neoplasia: number of test articles
associated with site-specific neoplasia that produced positive, clear or some evidence of
carcinogenicity. Available online at https://manticore.niehs.nih.gov/organsites/
This document is a draft for review purposes only and does not constitute Agency policy.
R-34 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Nudler. SI: Ouinteros. FA: Miler. EA: Cabilla. TP: Ronchetti. SA: Duvilanski. BH. (2009). Chromium VI
administration induces oxidative stress in hypothalamus and anterior pituitary gland from
male rats. Toxicol Lett 185: 187-192. http://dx.doi.Org/10.1016/i.toxlet2009.01.003
O'Brien. KO: Ru. Y. (2017). Iron status of North American pregnant women: An update on
longitudinal data and gaps in knowledge from the United States and Canada [Review], Am J
Clin Nutr 106: 1647S-1654S. http: / /dx.doi.org/10.3945 /aicn.l 17.155986
O'Brien. TT: Cervak. S: Patierno. S. R. (2003). Complexities of chromium carcinogenesis: role of
cellular response, repair and recovery mechanisms [Review], MutatRes 533: 3-36.
http://dx.doi.Org/10.1016/j.mrfmmm.2003.09.006
O'Brien. TT: Ding. H: Suh. M: Thompson. CM: Parsons. BL: Harris. MA: Winkelman. WA: Wolf. TC:
Hixon. TG: Schwartz. AM: Myers. MB: Haws. LC: Proctor. DM. (2013). Assessment of K-Ras
mutant frequency and micronucleus incidence in the mouse duodenum following 90-days of
exposure to Cr(VI) in drinking water. MutatRes 754: 15-21.
http://dx.doi.Org/10.1016/i.mrgentox.2013.03.008
O'Flahertv. El. (1993). A pharmacokinetic model for chromium. Toxicol Lett 68: 145-158.
http://dx.doi.org/10.1016/0378-4274r93190127-T
O'Flahertv. EI. (1996). A physiologically based model of chromium kinetics in the rat. Toxicol Appl
Pharmacol 138: 54-64. http://dx.doi.org/10.1006/taap.1996.0Q97
O'Flahertv. EI: Kerger. BP: Hays. SM: Paustenbach. DT. (2001). A physiologically based model for the
ingestion of chromium(III) and chromium(VI) by humans. Toxicol Sci 60: 196-213.
http://dx.doi.Org/10.1093/toxsci/60.2.196
O'Flahertv. EI: Radike. Ml. (1991). Pharmacokinetic modeling of trivalent and hexavalent chromium
based on ingestion and inhalation of soluble chromium compounds. (AL-TR-1991-0162).
Wright-Patterson AFB, Ohio: Armstrong Laboratory.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?searchOuery=ADA256238
O'Neil. M: Heckelman. P: Koch. C. (2006). Chromium and compounds. In The Merck index.
Whitehouse Station, NJ: Merck & Co. Inc.
O'Hara. KA: Klei. LR: Vaghjiani. RT: Barchowskv. A. (2004). Selective signaling pathways for
chromium(VI) induced patterns of transcription factor binding in exposed airway epithelial
cells [Abstract], Toxicologist78: 239.
OAOPS (EPA Office of Air Quality Planning and Standards). (2012). Residual risk assessment for the
chromic acid anodizing, decorative chromium electroplating, and hard chromium
electroplating source categories. Research Triangle Park, NC: U.S. Environmental Protection
Agency.
Ochieng. 1: Chaudhuri. G. (2010). Cystatin superfamily [Review], J Health Care Poor Underserved 21:
51-70. http://dx.doi.Org/10.1353 /hpu.0.0257
OECD (Organisation for Economic Co-operation and Development). (2016a). Test no. 474:
Mammalian erythrocyte micronucleus test [Standard], In OECD guidelines for the testing of
chemicals Section 4, health effects. Paris, France: OECD Publishing.
http: / /dx.doi.org/10.1787/9789264264762-en
OECD (Organisation for Economic Co-operation and Development). (2016b). Test no. 478: Rodent
dominant lethal test. In OECD guidelines for the testing of chemicals, Section 4: Health
effects. Paris, France, http://dx.doi.org/10.1787/9789264264823-en
OECD (Organisation for Economic Co-operation and Development). (2020). Test no. 488:
Transgenic rodent somatic and germ cell gene mutation assays. In OECD Guidelines for the
Testing of Chemicals, Section 4: Health Effects. Paris, France.
http://dx.doi.org/10.1787/9789264203907-en
Ohvama. W: Gonda. M: Mivaiima. H: Kondo. K: Noguchi. T: Yoshida. 1: Hatakevama. S: Watabe. E:
Ueno. Y: Havashi. M: Tokumitsu. T. (2002). Collaborative validation study of the in vivo
This document is a draft for review purposes only and does not constitute Agency policy.
R-35 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
micronucleus test using mouse colonic epithelial cells. Mutat Res 518: 39-45.
http://dx.doi.org/10.1016/sl383-5718r02100061-x
Okada. E: Fuiiishi. Y: Narumi. K: Kado. S: Ohvama. W. (2019). Evaluation of a 28-day repeated-dose
micronucleus test in rat glandular stomach, colon, and liver using gastrointestinal tract-
targeted genotoxic-carcinogens and non-carcinogens. Mutat Res 844: 62-68.
http://dx.doi.Org/10.1016/i.mrgentox.2019.05.008
Olaguibel. TM: Basomba. A. (1989). Occupational asthma induced by chromium salts [Review],
Allergol Immunopathol 17: 133-136.
Oregon DEO (Oregon Department of Environmental Quality). (2014). Update: Former Ken Foster
Farm cleanup in Sherwood [Fact Sheet], Portland, OR.
http://www.deq.state.or.us/lq/cu/nwr/kenfoster/KenFosterFarmsFactSheetpdf
Oregon DEO (Oregon Department of Environmental Quality). (2016a). Bullseye Glass area-wide soil
sampling report Portland, OR.
Oregon DEO (Oregon Department of Environmental Quality). (2016b). Portland Metals air
monitoring data. Portland, OR.
Oregon DEO (Oregon Department of Environmental Quality). (2016c). Uroboros Glass area-wide
soil sampling report Portland, OR.
http://www.oregon.gov/deq/FilterDocs/soilsamplingUro.pdf
OSHA (Occupational Safety & Health Administration). (2006). Occupational exposure to hexavalent
chromium. Final rule. Fed Reg 71: 10099-10385.
Ottenwaelder. H: Wiegand. HI: Bolt. HM. (1988). Uptake of 51Cr(VI) by human erythrocytes:
Evidence for a carrier-mediated transport mechanism. Sci Total Environ 71: 561-566.
http://dx.doi.org/10.1016/0048-9697r88190237-9
Oze. C: Bird. DK: Fendorf. S. (2007). Genesis of hexavalent chromium from natural sources in soil
and groundwater. Proc Natl Acad Sci USA 104: 6544-6549.
http://dx.doi.Org/10.1073/pnas.0701085104
Oze. C: Fendorf. S: Bird. DK: Coleman. RG. (2004). Chrominium geochemistry of serpentine soils.
International Geology Review 46: 97-126. http://dx.doi.Org/10.2747/0020-6814.46.2.97
Pacvna. 1: Nriagu. 1. (1988). Atmospheric emissions of chromium from natural and anthropogenic
sources. In JO Nriagu; E Nieboer (Eds.), Chromium in the natural and human environments
(pp. 105-123). New York, NY: Wiley, https: //www.wiley.com/en-
us/Chromium+in+the+Natural+and+Human+Environments+-p-9780471856436
Page. BT: Loar. GW. (2004). Chromium compounds. In Kirk-Othmer encyclopedia of chemical
technology. New York, NY: John Wiley & Sons.
Paixao. P: Bermejo. M: Hens. B: Tsume. Y: Dickens. I: Shedden. K: Salehi. N: Koenigsknecht. Ml:
Baker. TR: Hasler. WL: Lionberger. R: Fan. I: Wvsocki. 1: Wen. B: Lee. A: Frances. A: Amidon.
GE: Yu. A: Benninghoff. G: Lobenberg. R: Talattof. A: Sun. D: Amidon. GL. (2018). Gastric
emptying and intestinal appearance of nonabsorbable drugs phenol red and paromomycin
in human subjects: A multi-compartment stomach approach. Eur J Pharm Biopharm 129:
162-174. http: //dx.doi.org/10.1016/i.eipb.2018.05.033
Pan. CH: Teng. HA: Lai. CH. (2017). Biomarkers of oxidative stress in electroplating workers exposed
to hexavalent chromium. J Expo Sci Environ Epidemiol 28: 76-83.
http://dx.doi.org/10.1038/jes.2016.85
Park. S: Zhang. X: Li. C: Yin. C: Li. 1: Fallon. IT: Huang. W: Xu. D. (2017). Single-cell RNA sequencing
reveals an altered gene expression pattern as a result of CRISPR/cas9-mediated deletion of
Gene 3 3/Mig6 and chronic exposure to hexavalent chromium in human lung epithelial cells.
Toxicol Appl Pharmacol 330: 30-39. http://dx.doi.Org/10.1016/i.taap.2017.07.003
Parker. D: Turk. TL: Drossier. K. (1984). Comparison of DNFB and K2Cr207 'flare-up' reactions in
the guinea pig. International Arch Allergy Appl Immunol 73: 123-128.
http://dx.doi.org/10.1159/000233451
This document is a draft for review purposes only and does not constitute Agency policy.
R-36 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Parrott. N: Lukacova. V: Fraczkiewicz. G: Bolger. MB. (2009). Predicting pharmacokinetics of drugs
using physiologically based modeling--application to food effects [Review], AAPS J11: 45-
53. http://dx.doi.org/10.1208/sl2248-008-9079-7
Parsons. BL: Marchant-Miros. KE: Delongchamp. RR: Verkler. TL: Patterson. TA: Mckinzie. PB: Kim.
LT. (2010). ACB-PCR quantification of K-RAS codonl2 GAT and GTT mutant fraction in
colon tumor and non-tumor tissue. Cancer Invest 28: 364-375.
http://dx.doi.org/10.3109/07357901003630975
Pascal. LE: Tessier. DM. (2004). Cytotoxicity of chromium and manganese to lung epithelial cells in
vitro. Toxicol Lett 147: 143-151. http://dx.doi.Org/10.1016/i.toxlet2003.ll.004
Patlolla. AK: Barnes. C: Hackett. D: Tchounwou. PB. (2009). Potassium dichromate induced
cytotoxicity, genotoxicity and oxidative stress in human liver carcinoma (HepG2) cells. Int J
Environ Res Public Health 6: 643-653. http://dx.doi.org/10.3390/iierph6020643
Paustenbach. DT: Hays. SM: Brien. BA: Dodge. DG: Kerger. BP. (1996). Observation of steady state in
blood and urine following human ingestion of hexavalent chromium in drinking water. J
Toxicol Environ Health 49: 453-461. http://dx.doi.org/10.1080/009841096160682
Pellerin. C: Booker. SM. (2000). Reflections on hexavalent chromium: Health hazards of an
industrial heavyweight Environ Health Perspect 108: a402-a407.
http://dx.doi.org/10.1289/ehp.108-a402
Peng. N: Zhao. X. (2014). Comparison of K-ras mutations in lung, colorectal and gastric cancer.
Oncol Lett 8: 561-565. http://dx.doi.org/10.3892/ol.2014.2205
Peng. Y: Hu. 1: Li. Y: Zhang. B: Liu. W: Li. H: Zhang. H: Hu. C: Chen. X: Xia. W: Lu. S: Xu. S. (2018).
Exposure to chromium during pregnancy and longitudinally assessed fetal growth: Findings
from a prospective cohort Environ Int 121: 375-382.
http://dx.doi.Org/10.1016/j.envint.2018.09.003
Peterson-Roth. E: Reynolds. M: Ouievrvn. G: Zhitkovich. A. (2005). Mismatch repair proteins are
activators of toxic responses to chromium-DNA damage. Mol Cell Biol 25: 3596-3607.
http://dx.doi.Org/10.1128/MCB.25.9.3596-3607.2005
Petrilli. FL: De Flora. S. (1982). Interpretations on chromium mutagenicity and carcinogenicity. In M
Sorsa; H Vainio (Eds.), Mutagens in our environment: Proceedings of the twelfth annual
meeting of the European Environmental Mutagen Society (pp. 453-464). New York, NY: A. R.
Liss.
Petrilli. FL: Rossi. GA: Camoirano. A: Romano. M: Serra. D: Bennicelli. C: De Flora. A: De Flora. S.
(1986). Metabolic reduction of chromium by alveolar macrophages and its relationships to
cigarette smoke. J Clin Invest 77: 1917-1924. http://dx.doi.org/10.1172/TCI112520
PHS (U.S. Public Health Service). (1953). Health of workers in chromate producing industry. (No.
192). Washington, DC.
Polisak. B: Gazdag. Z: Tenko-Brinovec. S: Fuis. S: Pesti. M: Belagvi. I: Plesnicar. S: Raspor. P. (2005).
Pro-oxidative vs antioxidative properties of ascorbic acid in chromium(VI)-induced
damage: an in vivo and in vitro approach. J Appl Toxicol 25: 535-548.
http://dx.doi.org/10.1002/iat.1093
Pool-Zobel. BL: Lotzmann. N: Knoll. M: Kuchenmeister. F: Lambertz. R: Leucht. U: Schroder. HG:
Schmezer. P. (1994). Detection of genotoxic effects in human gastric and nasal mucosa cells
isolated from biopsy samples. Environ Mol Mutagen 24: 23-45.
http://dx.doi.Org/10.1002/em.2850240105
Popper. HH: Grvgar. E: Ingolic. E: Wawschinek. 0. (1993). Cytotoxicity of chromium-III and
chromium-VI compounds. I. In vitro studies using different cell culture systems. Inhal
Toxicol 5: 345-369. http://dx.doi.org/10.3109/08958379308998392
Potten. CS: Booth. C: Pritchard. DM. (1997). The intestinal epithelial stem cell: the mucosal governor
[Review], Int J Exp Pathol 78: 219-243. http://dx.doi.Org/10.1046/i. 1365-
2613.1997.280362.x
This document is a draft for review purposes only and does not constitute Agency policy.
R-37 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Potten. CS: Gandara. R: Mahida. YR: Loeffler. M: Wright. NA. (2009). The stem cells of small
intestinal crypts: where are they? Cell Prolif 42: 731-750.
http://dx.doi.Org/10.llll/i.1365-2184.2009.00642.x
Pournourmohammadi. S: Khazaelp. P: Eslamizad. S: Taivar. A: Mohammadirad. A: Abdollahi. M.
(2008). Study on the oxidative stress status among cement plant workers. Hum Exp Toxicol
27: 463-469. http://dx.doi.org/10.1177/0960327108094956
Proctor. DM: Panko. TP: Liebig. EW: Paustenbach. DT. (2004). Estimating historical occupational
exposure to airborne hexavalent chromium in a chromate production plant: 1940--1972. J
Occup Environ Hyg 1: 752-767. http://dx.doi.org/10.1080/15459620490523294
Proctor. DM: Suh. M: Avlward. LL: Kirman. CR: Harris. MA: Thompson. CM: Giirleviik. H: Gerads. R:
Haws. LC: Hays. SM. (2012). Hexavalent chromium reduction kinetics in rodent stomach
contents. Chemosphere 89: 487-493.
http://dx.doi.Org/10.1016/j.chemosphere.2012.04.065
Proctor. DM: Suh. M: Mittal. L. iz: Hirsch. S: Salgado. RV: Bartlett. C: Van Landingham. C: Rohr. A:
Crump. K. (2016). Inhalation cancer risk assessment of hexavalent chromium based on
updated mortality for Painesville chromate production workers. J Expo Sci Environ
Epidemiol 26: 224-231. http://dx.doi.org/10.1038/ies.2015.77
Pukkala. E: Martinsen. 1: Lvnge. E: Gunnarsdottir. H: Sparen. P: Tryggvadottir. L: Weiderpass. E:
Kiaerheim. K. (2009). Occupation and cancer - follow-up of 15 million people in five Nordic
countries. Acta Oncol 48: 646-790. http://dx.doi.org/10.1080/02841860902913546
Oayyum. S: Ara. A: Usmani. TA. (2012). Effect of nickel and chromium exposure on buccal cells of
electroplaters. Toxicol Ind Health 28: 74-82.
http://dx.doi.org/10.1177/0748233711407237
Oian. 0. in: Li. P: Wang. T: Zhang. 1. i: Yu. S: Chen. T: Yan. L. ei: Song. Y: Liu. X: Gu. Y: Wang. Y. un: Tia.
G. (2013). Alteration of Thl/Th2/Thl7 cytokine profile and humoral immune responses
associated with chromate exposure. Occup Environ Med 70: 697-702.
http://dx.doi.org/10.1136/oemed-2013-101421
Oin. 0: Xie. H: Wise. SS: Browning. CL: Thompson. KN: Holmes. AL: Wise. TP. (2014). Homologous
recombination repair signaling in chemical carcinogenesis: prolonged particulate
hexavalent chromium exposure suppresses the rad51 response in human lung cells. Toxicol
Sci 142: 117-125. http://dx.doi.org/10.1093/toxsci/kful75
Ouansah. R: Taakkola. 11. (2009). Paternal and maternal exposure to welding fumes and metal dusts
or fumes and adverse pregnancy outcomes. Int Arch Occup Environ Health 82: 529-537.
http://dx.doi.Org/10.1007/s00420-008-0349-6
Ouante. M: Tu. SP: Tomita. H: Gonda. T: Wang. SS: Takashi. S: Baik. GH: Shibata. W: Diprete. B: Betz.
KS: Friedman. R: Varro. A: Tvcko. B: Wang. TC. (2011). Bone marrow-derived
myofibroblasts contribute to the mesenchymal stem cell niche and promote tumor growth.
Cancer Cell 19: 257-272. http://dx.doi.Org/10.1016/i.ccr.2011.01.020
Ouante. M: Varga. 1: Wang. TC: Greten. FR. (2013). The gastrointestinal tumor microenvironment
[Review], Gastroenterology 145: 63-78. http://dx.doi.Org/10.1053/i.gastro.2013.03.052
Ouante. M: Wang. TC. (2008). Inflammation and stem cells in gastrointestinal carcinogenesis
[Review], Physiology (Bethesda) 23: 350-359.
http://dx.doi.Org/10.1152/physiol.00031.2008
Ouante. M: Wang. TC. (2009). Stem cells in gastroenterology and hepatology. Nat Rev Gastroenterol
Hepatol 6: 724-737. http://dx.doi.org/10.1038/nrgastro.20Q9.195
Ouievrvn. G: Goulart. M: Messer. I: Zhitkovich. A. (2001). Reduction of Cr (VI) by cysteine:
Significance in human lymphocytes and formation of DNA damage in reactions with variable
reduction rates. Mol Cell Biochem 222: 107-118.
http://dx.doi.Org/10.1023/A:1017923609175
This document is a draft for review purposes only and does not constitute Agency policy.
R-38 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
Ouievrvn. G: Messer. 1: Zhitkovich. A. (2002). Carcinogenic chromium(VI) induces cross-linking of
vitamin C to DNA in vitro and in human lung A5 49 cells. Biochemistry 41: 3156-3167.
http://dx.doi.org/10.1021/bi011942z
Ouievrvn. G: Peterson. E: Messer. 1: Zhitkovich. A. (2003). Genotoxicity and mutagenicity of
chromium(VI)/ascorbate-generated DNA adducts in human and bacterial cells.
Biochemistry 42: 1062-1070. http://dx.doi.org/10.1021/bi0271547
Ouinteros. FA: Machiavelli. LI: Miler. EA: Cab ilia. TP: Duvilanski. BH. (2008). Mechanisms of
chromium (Vl)-induced apoptosis in anterior pituitary cells. Toxicology 249:109-115.
http://dx.doi.Org/10.1016/i.tox.2008.04.012
Ouinteros. FA: Poliandri. AH: Machiavelli. LI: Cabilla. TP: Duvilanski. BH. (2007). In vivo and in vitro
effects of chromium VI on anterior pituitary hormone release and cell viability. Toxicol Appl
Pharmacol 218: 79-87. http://dx.doi.Org/10.1016/i.taap.2006.10.017
Oureshi. IZ: Mahmood. T. (2010). Prospective role of ascorbic acid (vitamin C) in attenuating
hexavalent chromium-induced functional and cellular damage in rat thyroid. Toxicol Ind
Health 26: 349-359. http://dx.doi.org/10.1177/0748233710371109
Radke. EG: Glenn. B: Galizia. A: Persad. A: Nachman. R: Bateson. T: Wright. TM: Navas-Acien. A:
Arrovave. WD: Puett. RC: Harville. EW: Pollack. AZ: Burns. IS: Lynch. CD: Sagiv. SK: Stein. C:
Cooper. GS. (2019). Development of outcome-specific criteria for study evaluation in
systematic reviews of epidemiology studies. Environ Int 130: 104884.
http://dx.doi.Org/10.1016/j.envint.2019.05.078
Radtke. F: Clevers. H. (2005). Self-renewal and cancer of the gut: two sides of a coin [Review],
Science 307: 1904-1909. http://dx.doi.org/10.1126/science.1104815
Rafael. AI: Almeida. A: Santos. P: Parreira. I: Madeira. VM: Alves. R: Cabrita. AM: Alpoim. MC. (2007).
A role for transforming growth factor-beta apoptotic signaling pathway in liver injury
induced by ingestion of water contaminated with high levels of Cr(VI). Toxicol Appl
Pharmacol 224: 163-173. http://dx.doi.Org/10.1016/i.taap.2007.07.004
Rager. IE: Ring. CL: Fry. RC: Suh. M: Proctor. DM: Haws. LC: Harris. MA: Thompson. CM. (2017).
High-throughput screening data interpretation in the context of in vivo transcriptomic
responses to oral Cr(VI) exposure. Toxicol Sci 158: 199-212.
http://dx.doi.org/10.1093/toxsci/kfx085
Rager. IE: Suh. M: Chappell. GA: Thompson. CM: Proctor. DM. (2019). Review of transcriptomic
responses to hexavalent chromium exposure in lung cells supports a role of epigenetic
mediators in carcinogenesis [Review], Toxicol Lett 305: 40-50.
http://dx.doi.Org/10.1016/i.toxlet2019.01.011
Rahman. MM: Abe. SK: Rahman. MS: Kanda. M: Narita. S: Bilano. V: Ota. E: Gilmour. S: Shibuva. K.
(2016). Maternal anemia and risk of adverse birth and health outcomes in low- and middle-
income countries: Systematic review and meta-analysis [Review], Am J Clin Nutr 103: 495-
504. http://dx.doi.org/10.3945/aicn.115.107896
Rao. MY: Chawla. SL: Sharma. SR. (2009). Protective role of vitamin E on nickel and/or chromium
induced oxidative stress in the mouse ovary. Food Chem Toxicol 47: 1368-1371.
http://dx.doi.Org/10.1016/i.fct.2009.03.018
Rasool. M: Zaigham. K: Malik. A: Naseer. MI: Umm-E-Habiba. MI: Manan. A: Oazi. MH: Asif. M.
(2014). Potential reproductive health effects and oxidative stress associated with exposure
to potassium dichromate (K2Cr207) and magnesium sulphate (MgS04) in male mice.
Pakistan Journal of Medical Sciences online 30: 819-823.
Rawla. P: Barsouk. A. (2019). Epidemiology of gastric cancer: Global trends, risk factors and
prevention [Review], Prz Gastroenterol 14: 26-38.
http://dx.doi.org/10.5114/pg.2018.80001
This document is a draft for review purposes only and does not constitute Agency policy.
R-39 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Re my. LL: Bvers. V: Clay. T. (2017). Reproductive outcomes after non-occupational exposure to
hexavalent chromium, Willits California, 1983-2014. Environ Health 16: 18.
http://dx.doi.org/10.1186/sl2940-017-0222-8
Ren. K: Li. Y: Lu. H: Li. Z: Han. X. (2017). miR-3940-5p Functions as a Tumor Suppressor in Non-
Small Cell Lung Cancer Cells by Targeting Cyclin D1 and Ubiquitin Specific Peptidase-28.
Translational Oncology 10: 80-89. http://dx.doi.Org/10.1016/i.tranon.2016.ll.004
Ren. R: Li. Z: Zhang. Z. (2003). [Investigation on the effect of occupational chromium on workers'
fertility], Zhongguo Gonggong Weisheng Guanli/Chinese Journal of Public Health
Management (1): 82-83.
Reynolds. M: Armknecht. S: Tohnston. T: Zhitkovich. A. (2012). Undetectable role of oxidative DNA
damage in cell cycle, cytotoxic and clastogenic effects of Cr(VI) in human lung cells with
restored ascorbate levels. Mutagenesis 27: 437-443.
http://dx.doi.org/10.1093/mutage/ger095
Reynolds. M: Zhitkovich. A. (2007). Cellular vitamin C increases chromate toxicity via a death
program requiring mismatch repair but not p53. Carcinogenesis 28: 1613-1620.
http://dx.doi.org/10.1093/carcin/bgm031
Richardson. DB: Langholz. B. (2012). Background stratified Poisson regression analysis of cohort
data. Radiat Environ Biophys 51: 15-22. http://dx.doi.org/10.1007/s00411-011-0394-5
Ritchie. RF: Palomaki. GE: Neveux. LM: Navolotskaia. 0: Ledue. TB: Craig. WY. (2004). Reference
distributions for complement proteins C3 and C4: A practical, simple and clinically relevant
approach in a large cohort. J Clin Lab Anal 18: 1-8. http://dx.doi.org/10.1002/icla.1010Q
Riva. L: Pandiri. AR: Li. YR: Droop. A: Hewinson. 1: Quail. MA: Iyer. V: Shepherd. R: Herbert. RA:
Campbell. PI: Sills. RC: Alexandrov. LB: Balmain. A: Adams. DT. (2020). The mutational
signature profile of known and suspected human carcinogens in mice. Nat Genet 52: 1189-
1197. http://dx.doi.Org/10.1038/s41588-020-0692-4
Rovle. H. (1975). Toxicity of chromic acid in the chromium plating industry(2). Environ Res 10:
141-163. http://dx.doi.org/10.1016/0013-935ir75190079-l
Rudnvkh. AA: Zasukhina. GD. (1985). DNA repair synthesis in lymphocytes of people contacting
chromium compounds in industry. Cytology and Genetics 19: 68-69.
Russell. TL: Berardi. RR: Barnett. TL: Dermentzoglou. LC: Tarvenpaa. KM: Schmaltz. SP: Pressman. TB.
(1993). Upper gastrointestinal pH in seventy-nine healthy, elderly, North American men
and women. Pharm Res 10: 187-196. http: //dx.doi.Org/10.1023 /A:1018970323716
Russo. P: Catassi. A: Cesario. A: Imperatori. A: Rotolo. N: Fini. M: Granone. P: Dominioni. L. (2005).
Molecular mechanisms of hexavalent chromium-induced apoptosis in human
bronchoalveolar cells. Am J Respir Cell Mol Biol 33: 589-600.
http://dx.doi.org/10.1165/rcmb.2005-02130C
Saleh. FY: Parkerton. TF: Lewis. RV: Huang. TH: Dickson. KL. (1989). Kinetics of chromium
transformations in the environment. Sci Total Environ 86: 25-41.
Saloga. 1: Knop. I: Kolde. G. (1988). Ultrastructural cytochemical visualization of chromium in the
skin of sensitized guinea pigs. Arch Dermatol Res 280: 214-219.
Samuel. TB: Stanley. TA: Roopha. DP: Vengatesh. G: Anbalagan. 1: Banu. SK: Aruldhas. MM. (2011).
Lactational hexavalent chromium exposure-induced oxidative stress in rat uterus is
associated with delayed puberty and impaired gonadotropin levels. Hum Exp Toxicol 30:
91-101. http:/ /dx.doi.org/10.1177/0960327110364638
Samuel. TB: Stanley. TA: Sekar. P: Princess. RA: Sebastian. MS: Aruldhas. MM. (2012a). Persistent
hexavalent chromium exposure impaired the pubertal development and ovarian
histoarchitecture inwistar rat offspring. Environ Toxicol 29: 814-828.
http://dx.doi.org/10.1002/tox.21810
Samuel. TB: Stanley. TA: Vengatesh. G: Princess. RA: Muthusami. S: Roopha. DP: Suthagar. E: Kumar.
KM: Sebastian. MS: Aruldhas. MM. (2012b). Ameliorative effect of vitamin C on hexavalent
This document is a draft for review purposes only and does not constitute Agency policy.
R-40 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
chromium-induced delay in sexual maturation and oxidative stress in developing Wistar rat
ovary and uterus. Toxicol Ind Health 28: 720-733.
http://dx.doi.org/10.1177/0748233711422728
Sanchez-Martin. FT: Fan. Y: Carreira. V: Ovesen. TL: Vonhandorf. A: Xia. Y: Puga. A. (2015). Long-term
coexposure to hexavalent chromium and B[a]P causes tissue-specific differential biological
effects in liver and gastrointestinal tract of mice. Toxicol Sci 146: 52-64.
http://dx.doi.org/10.1093/toxsci/kfv070
Sanchez. LM: Lewicki. M: De Lucca. RC: Ubios. Am. (2015). Impairment of bony crypt development
associated with hexavalent chromium exposure during tooth eruption. Acta Odontol
Latinoam 28: 203-209.
Sanchez. LM: Ubios. Am. (2020). Alterations in odontogenesis and tooth eruption resulting from
exposure to hexavalent chromium in suckling animals. Int J Paediatr Dent 30: 35-41.
http://dx.doi.org/10.llll/ipd.12573
Sanchez. LM: Ubios. Am. (2021). Hexavalent chromium exposure alters bone remodeling in the
developing tooth alveolus and delays tooth eruption. 34: 91-97.
http://dx.doi.Org/10.54589/aol.34/2/091
Saraswathv. CP: Usharani. MY. (2007). Monitoring of cellular enzymes in the serum of
electroplating workers at Coimbatore. J Environ Biol 28: 287-290.
Sarkar. D: Sharma. A: Talukder. G. (1993). Differential protection of chlorophyllin against
clastogenic effects of chromium and chlordane in mouse bone marrow in vivo. Mutat Res
301: 33-38. http://dx.doi.org/10.1016/0165-7992C93190053-X
Sarkar. D: Sharma. A: Talukder. G. (1996). Protection afforded by crude spinach-beet leaf extract
and equivalent amounts of chlorophyll and chlorophyllin against clastogenic effects of
potassium dichromate in mice in vivo. Int J Pharmacogn 34: 58-63.
http://dx.doi.Org/10.1076/phbi.34.l.58.13174
Sarto. F: Cominato. I: Bianchi. V: Levis. AG. (1982). Increased incidence of chromosomal aberrations
and sister chromatid exchanges in workers exposed to chromic acid (Cr03) in electroplating
factories. Carcinogenesis 3: 1011-1016. http://dx.doi.Org/10.1093/carcin/3.9.1011
Sarto. F: Tomanin. R: Giacomelli. L: Iannini. G: Cupiraggi. AR. (1990). The micronucleus assay in
human exfoliated cells of the nose and mouth: Application to occupational exposures to
chromic acid and ethylene oxide. Mutat Res 244: 345-351. http: //dx.doi.org/10.1016/Q165-
7992C90190083-V
Sasso. AF: Schlosser. PM. (2015). An evaluation of in vivo models for toxicokinetics of hexavalent
chromium in the stomach. Toxicol Appl Pharmacol 287: 293-298.
http://dx.doi.Org/10.1016/j.taap.2015.06.016
Sawicka. E: Dlugosz. A. (2017). The role of 17(3-estradiol metabolites in chromium-induced
oxidative stress. Advances in Clinical and Experimental Medicine 26: 215-221.
http://dx.doi.org/10.17219/acem/62217
Sazakli. E: Villanueva. CM: Kogevinas. M: Maltezis. K: Mouzaki. A: Leotsinidis. M. (2014). Chromium
in Drinking Water: Association with Biomarkers of Exposure and Effect. Int J Environ Res
Public Health 11: 10125-10145. http://dx.doi.org/10.3390/iierphlll010125
SCAOMD (South Coast Air Quality Management District). (2016). Summary of efforts in paramount:
significant improvements since October 2016.
Schiestl. RH: Aubrecht. 1: Khogali. F: Carls. N. (1997). Carcinogens induce reversion of the mouse
pink-eyed unstable mutation. Proc Natl Acad Sci USA 94: 4576-4581.
http://dx.doi.Org/10.1073/pnas.94.9.4576
Schlesinger. RB: Lippmann. M. (1978). Selective particle deposition and bronchogenic carcinoma.
Environ Res 15: 424-431.
This document is a draft for review purposes only and does not constitute Agency policy.
R-41 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Schlosser. PM: Sasso. AF. (2014). A revised model of ex-vivo reduction of hexavalent chromium in
human and rodent gastric juices. Toxicol Appl Pharmacol 280: 352-361.
http://dx.doi.Org/10.1016/i.taap.2014.08.010
Schneeberger. HW: Forck. G. (1974). Experimental chromium allergy of animals: The influence of
valences and chemical properties of chromium ions. 249: 71-81.
Schroeder. WH: Dobson. M: Kane. DM: Tohnson. ND. (1987). Toxic trace elements associated with
airborne particulate matter: Areview [Review], J Air Waste Manag Assoc 37: 1267-1285.
http://dx.doi.Org/10.1080/08940630.1987.10466321
Schwitalla. S: Fingerle. AA: Cammareri. P: Nebelsiek. T: Goktuna. SI: Ziegler. PK: Canli. 0: Heiimans.
1: Huels. DT: Moreaux. G: Rupee. RA: Gerhard. M: Schmid. R: Barker. N: Clevers. H: Lang. R:
Neumann. 1: Kirchner. T: Taketo. MM: van den Brink. GR: Sansom. 01: Arkan. MC: Greten. FR.
(2013). Intestinal tumorigenesis initiated by dedifferentiation and acquisition of stem-cell-
like properties. Cell 152: 25-38. http://dx.doi.Org/10.1016/i.cell.2012.12.012
Scott. P: Anderson. K: Singhania. M: Cormier. R. (2020). Cystic Fibrosis, CFTR, and Colorectal Cancer
[Review], International Journal of Molecular Sciences 21.
http://dx.doi.org/10.3390/iims21082891
Sedman. RM: Beaumont. 1: Mcdonald. TA: Reynolds. S: Krowech. G: Howd. R. (2006). Review of the
evidence regarding the carcinogenicity of hexavalent chromium in drinking water [Review],
J Environ Sci Health C Environ Carcinog Ecotoxicol Rev 24: 155-182.
http://dx.doi.org/10.1080/10590500600614337
Sellappa. S: Prathvumnan. S: Kevan. KS: Toseph. S: Vasudevan. BS: Sasikala. K. (2010). Evaluation of
DNA damage induction and repair inhibition in welders exposed to hexavalent chromium.
Asian Pac J Cancer Prev 11: 95-100.
Sengupta. T: Chattopadhyav. D: Ghosh. N: Das. M: Chatterjee. GC. (1990). Effect of chromium
administration on glutathione cycle of rat intestinal epithelial cells. Indian J Exp Biol 28:
1132-1135.
Seoane. AI: Dulout. FN. (1999). Contribution to the validation of the anaphase-telophase test:
Aneugenic and clastogenic effects of cadmium sulfate, potassium dichromate and nickel
chloride in Chinese hamster ovary cells. Genet Mol Biol 22: 551-555.
Seoane. AI: Dulout. FN. (2001). Genotoxic ability of cadmium, chromium and nickel salts studied by
kinetochore staining in the cytokinesis-blocked micronucleus assay. Mutat Res 490: 99-106.
http://dx.doi.org/10.1016/SI 383-5718r00100145-5
Seoane. AI: Giierci. AM: Dulout. FN. (2002). Malsegregation as a possible mechanism of aneuploidy
induction by metal salts in MRC-5 human cells. Environ Mol Mutagen 40: 200-206.
http://dx.doi.org/10.1002/em.10110
Sharma. N: Chauhan. S: Varma. S. (2003). Fatal potassium dichromate ingestion [Letter], J Postgrad
Med 49: 286-287.
Sharma. P: Bihari. V: Agarwal. SK: Verma. V: Kesavachandran. CN: Pangtev. BS: Mathur. N: Singh. KP:
Srivastava. M: Goel. SK. (2012). Groundwater contaminated with hexavalent chromium [Cr
(VI)]: a health survey and clinical examination of community inhabitants (Kanpur, India).
PLoS ONE 7: e47877. http://dx.doi.org/10.1371/iournal.pone.0047877
Sharma. S: Kelly. TK: Tones. PA. (2010). Epigenetics in cancer [Review], Carcinogenesis 31: 27-36.
http://dx.doi.org/10.1093/carcin/bgp220
Shaw Environmental (Shaw Environmental, Inc). (2006). Final report Industry profile, exposure
profile, technological feasibility evaluation, and environmental impact for industries
affected by a revised OSHA standard for hexavalent chromium (General industry,
construction, and maritime sectors). (OSHA-H054A-2006-0064-2537). Washington, DC: U.S.
Department of Labor, Occupational Safety and Health Administration.
https://www.regulations.gov/document?D=OSHA-H054A-2006-0064-2537
This document is a draft for review purposes only and does not constitute Agency policy.
R-42 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Shi. H: Hudson. LG: Liu. KT. (2004). Oxidative stress and apoptosis in metal ion-induced
carcinogenesis [Review], Free Radic Biol Med 37: 582-593.
http://dx.doi.Org/10.1016/i.freeradbiomed.2004.03.012
Shih. GL: Brensinger. C: Katzka. DA: Metz. DC. (2003). Influence of age and gender on gastric acid
secretion as estimated by integrated acidity in patients referred for 24-hour ambulatory pH
monitoring. Am J Gastroenterol 98: 1713-1718. http://dx.doi.org/10.1016/S00Q2-
9270C03100435-0
Shih. IM: Wang. TL: Traverso. G: Romans. K: Hamilton. SR: Ben-Sasson. S: Kinzler. KW: Vogelstein. B.
(2001). Top-down morphogenesis of colorectal tumors. Proc Natl Acad Sci USA 98: 2640-
2645. http://dx.doi.org/10.1073/pnas.051629398
Shindo. Y: Tovoda. Y: Kawamura. K: Kurebe. M: Shimada. H: Hattori. C: Satake. S. (1989).
Micronucleus test with potassium chromate(VI) administered intraperitoneally and orally
to mice. MutatRes 223: 403-406.
Shobana. N: Aruldhas. MM: Tochhawng. L: Loganathan. A: Balaii. S: Kumar. MK: Banu. LAS: Navin.
AK: Mavilvanan. C: Ilangovan. R: Balasubramanian. K. (2017). Transient gestational
exposure to drinking water containing excess hexavalent chromium modifies insulin
signaling in liver and skeletal muscle of rat progeny. Chem Biol Interact 277:119-128.
http://dx.doi.Org/10.1016/i.cbi.2017.09.003
Shobana. N: Kumar. MK: Navin. AK: Akbarsha. MA: Aruldhas. MM. (2020). Prenatal exposure to
excess chromium attenuates transcription factors regulating expression of androgen and
follicle stimulating hormone receptors in Sertoli cells of prepuberal rats. Chem Biol Interact
328: 109188. http://dx.doi.org/10.1016/i.cbi.2020.109188
Shoeb. M: Kodali. VK: Farris. BY: Bishop. LM: Meighan. TG: Salmen. R: Eye. T: Friend. S: Schwegler-
Berrv. D: Roberts. TR: Zeidler-Erdelv. PC: Erdelv. A: Antonini. TM. (2017). Oxidative stress,
DNA methylation, and telomere length changes in peripheral blood mononuclear cells after
pulmonary exposure to metal-rich welding nanoparticles. Nanolmpact 5: 61-69.
http://dx.doi.Org/10.1016/i.impact.2017.01.001
Shrivastava. R: Srivastava. S: Upreti. RK: Chaturvedi. UC. (2005a). Effects of dengue virus infection
on peripheral blood cells of mice exposed to hexavalent chromium with drinking water.
Indian J Med Res 122: 111-119.
Shrivastava. R: Upreti. RK: Chaturvedi. UC. (2003). Various cells of the immune system and intestine
differ in their capacity to reduce hexavalent chromium. FEMS Immunol Med Microbiol 38:
65-70.
Shrivastava. R: Upreti. RK: Chaturvedi. UC. (2005b). Effects of dengue virus infection on the spleen
of male mice given hexavalent chromium with drinking water. Toxicol Mech Meth 15: 323-
329. http://dx.doi.org/10.1080/153765291009732
Shubin. AY: Demidvuk. IV: Komissarov. AA: Rafieva. LM: Kostrov. SV. (2016). Cytoplasmic
vacuolization in cell death and survival [Review], Onct 7: 55863-55889.
http://dx.doi.org/10.18632/oncotarget.10150
Shumilla. TA: Barchowskv. A. (1999). Inhibition of protein synthesis by chromium(VI) differentially
affects expression of urokinase and its receptor in human type II pneumocytes. Toxicol Appl
Pharmacol 158: 288-295. http://dx.doi.org/10.1006/taap.1999.87Q4
Siegenthaler. U: Laine. A: Polak. L. (1983). Studies on contact sensitivity to chromium in the guinea
pig The role of valence in the formation of the antigenic determinant. J Invest Dermatol 80:
44-47.
Singh. CBP: Sharma. A: Talukder. G. (1990). Effects of chromium on cellular systems in animals.
Nucleus 33: 84-106.
Singhal. VK: Deswal. BS: Singh. BN. (2015). Study of skin and mucous membrane disorders among
workers engaged in the sodium dichromate manufacturing industry and chrome plating
This document is a draft for review purposes only and does not constitute Agency policy.
R-43 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
industry. Indian J Occup Environ Med 19: 129-133. http://dx.doi.org/10.4103/0019-
5278.173994
Sitalakshmi. R: Saikumar. P: Tevachandran. P: Manoharan. P: Thangavel. P: Thomas. T. (2016). Civil
construction work: The unseen contributor to the occupational and global disease burden.
Indian J Occup Environ Med 20: 103-108. http://dx.doi.org/10.4103 /0019-5278.197542
Sivakumar. KK: Stanley. TA: Arosh. TA: Pepling. ME: Burghardt. RC: Banu. SK. (2014). Prenatal
exposure to chromium induces early reproductive senescence by increasing germ cell
apoptosis and advancing germ cell cyst breakdown in the F1 offspring. Dev Biol 388: 22-34.
http://dx.doi.Org/10.1016/j.ydbio.2014.02.003
Sivakumar. KK: Stanley. TA: Behlen. TC: Wuri. L: Dutta. S: Wu. 1: Arosh. TA: Banu. SK. (2022).
Inhibition of Sirtuin-1 hyperacetylates p53 and abrogates Sirtuin-l-p53 interaction in
Cr(VI)-induced apoptosis in the ovary. Reprod Toxicol 109: 121-134.
http://dx.doi.Org/10.1016/j.reprotox.2022.03.007
Smith. AH. (2008). Hexavalent chromium, yellow water, and cancer: a convoluted saga.[comment]
[Letter], Epidemiology 19: 24-26.
Smith. DB: Cannon. WF: Woodruff. LG: Solano. F: Kilburn. IE: Fey. PL. (2013). Geochemical and
mineralogical data for soils of the conterminous United States. (Data Series 801). Reston,
VA: U.S. Department of the Interior, U.S. Geological Survey, http: / /pubs.usgs.gov/ds /801/
Smith. MT: Guvton. KZ: Gibbons. CF: Fritz. TM: Portier. CI: Rusvn. I: DeMarini. DM: Caldwell. TC:
Kavlock. RT: Lambert. PF: Hecht. SS: Bucher. TR: Stewart. BW: Baan. RA: Cogliano. VT: Straif.
K. (2016). Key characteristics of carcinogens as a basis for organizing data on mechanisms
of carcinogenesis [Review], Environ Health Perspect 124: 713-721.
http://dx.doi.org/10.1289/ehp.15Q9912
Snyder. CA: Udasin. I: Waterman. ST: Toioli. E: Gochfeld. M. (1996). Reduced IL-6 levels among
individuals in Hudson County, New Jersey, an area contaminated with chromium. Arch
Environ Health 51: 26-28. http://dx.doi.org/10.1080/00039896.1996.9935989
Snyder. CA: Valle. CD. (1991). Immune function assays as indicators of chromate exposure. Environ
Health Perspect 92: 83-86. http://dx.doi.org/10.1289/ehp.919283
Spares. ME: Vieira. E: Bastos. M. (2010). Chromium speciation analysis in bread samples. J Agric
Food Chem 58: 1366-1370. http://dx.doi.org/10.1021/if903118v
Sobaszek. A: Edme. TL: Boulenguez. C: Shirali. P: Mereau. M: Robin. H: Haguenoer. TM. (1998).
Respiratory symptoms and pulmonary function among stainless steel welders. J Occup
Environ Med 40: 223-229.
Song. Y: Zhang. 1: Yu. S: Wang. T: Cui. X: Du. X: Tia. G. (2012). Effects of chronic chromium(vi)
exposure on blood element homeostasis: an epidemiological study. Metallomics 4: 463-472.
http://dx.doi.org/10.1039/c2mt20051a
Sorahan. T: Burges. DC: Hamilton. L: Harrington. TM. (1998). Lung cancer mortality in
nickel/chromium platers, 1946-95. Occup Environ Med 55: 236-242.
http://dx.doi.Org/10.1136/oem.55.4.236
Sorahan. T: Burges. DCL: Waterhouse. TAH. (1987). A mortality study of nickel/chromium platers.
Br JInd Med 44: 250-258.
Soudani. N: Bouaziz. H: Sefi. M: Chtourou. Y: Boudawara. T: Zeghal. N. (2013). Toxic effects of
chromium (VI) by maternal ingestion on liver function of female rats and their suckling
pups. Environ Toxicol 28:11-20. http: //dx.doi.org/10.1002 /tox.20692
Soudani. N: Ibtissem Ben Amara. N: Troudi. A: Bouaziz. H: Boudawara. T: Zeghal. N. (2011a).
Oxidative stress induced by chromium (VI) in bone of suckling rats. Toxicol Ind Health 27:
724-734. http://dx.doi.org/10.1177/0748233710395992
Soudani. N: Sefi. M: Bouaziz. H: Chtourou. Y: Boudawara. T: Zeghal. N. (2011b). Nephrotoxicity
induced by chromium (VI) in adult rats and their progeny. Hum Exp Toxicol 30: 1233-1245.
http://dx.doi.org/10.1177/0960327110387454
This document is a draft for review purposes only and does not constitute Agency policy.
R-44 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
Speer. RM: Tovoda. TH: Croom-Perez. TT: Liu. KT: Wise. TP. (2021). Particulate hexavalent chromium
inhibits E2F1 leading to reduced RAD51 nuclear foci formation in human lung cells. Toxicol
Sci 181: 35-46. http: / /dx. doi. o r g /10.10 9 3/toxsci/kfab019
Spiteri. NT. (1982). Circadian patterning of feeding, drinking and activity during diurnal food access
in rats. Physiol Behav 28: 139-147. http://dx.doi.org/10.1016/0031-9384r82190115-9
Squier. CA: Kremer. Ml. (2001). Biology of oral mucosa and esophagus [Review], J Natl Cancer Inst
Monogr 29: 7-15.
Standeven. AM: Wetterhahn. KE. (1989). Chromium-vi toxicity uptake reduction and dna damage. J
Am Coll Toxicol 8: 1275-1284. http://dx.doi.org/10.3109/10915818909009118
Standeven. AM: Wetterhahn. KE. (1991). Ascorbate is the principal reductantof chromium (VI) in
rat liver and kidney ultrafiltrates. Carcinogenesis 12: 1733-1737.
http://dx.doi.Org/10.1093/carcin/12.9.1733
Standeven. AM: Wetterhahn. KE. (1992). Ascorbate is the principal reductantof chromium(VI) in
rat lung ultrafiltrates and cytosols, and mediates chromium-DNA binding in vitro.
Carcinogenesis 13: 1319-1324. http://dx.doi.Org/10.1093/carcin/13.8.1319
Stanley. TA: Lee. 1: Nithv. TK: Arosh. TA: Burghardt. RC: Banu. SK. (2011). Chromium-VI arrests cell
cycle and decreases granulosa cell proliferation by down-regulating cyclin-dependent
kinases (CDK) andcyclins and up-regulating CDK-inhibitors. Reprod Toxicol 32: 112-123.
http://dx.doi.Org/10.1016/i.reprotox.2011.05.007
Stanley. TA: Sivakumar. KK: Arosh. TA: Burghardt. RC: Banu. SK. (2014). Edaravone mitigates
hexavalent chromium-induced oxidative stress and depletion of antioxidant enzymes while
estrogen restores antioxidant enzymes in the rat ovary in F1 offspring. Biol Reprod 91: 12.
http://dx.doi.org/10.1095/biolreprod.113.113332
Stanley. TA: Sivakumar. KK: Nithv. TK: Arosh. TA: Hover. PB: Burghardt. RC: Banu. SK. (2013).
Postnatal exposure to chromium through mother's milk accelerates follicular atresia in F1
offspring through increased oxidative stress and depletion of antioxidant enzymes. Free
Radic Biol Med 61: 179-196. http://dx.doi.Org/10.1016/i.freeradbiomed.2013.02.006
Stearns. DM. (2000). Is chromium a trace essential metal? [Review], BioFactors 11: 149-162.
http://dx.doi.Org/10.1002/biof.5520110301
Stearns. DM: Kennedy. LI: Courtney. KD: Giangrande. PH: Phieffer. LS: Wetterhahn. KE. (1995).
Reduction of chromium(VI) by ascorbate leads to chromium-DNA binding and DNA strand
breaks in vitro. Biochemistry 34: 910-919. http: //dx.doi.org/10.1021/bi000Q3a025
Stearns. DM: Wetterhahn. KE. (1994). Reaction of chromium(VI) with ascorbate produces
chromium(V), chromium(IV), and carbon-based radicals. Chem Res Toxicol 7: 219-230.
http://dx.doi.Org/10.1021/tx00038a016
Stearns. DM: Wetterhahn. KE. (1997). Intermediates Produced in the Reaction of Chromium(VI)
with Dehydroascorbate Cause Single-Strand Breaks in Plasmid DNA. Chem Res Toxicol 10:
271-278. http://dx.doi.org/10.1021/tx9601521
Steinhoff. D: ShC. G: Hatfield. GK: Mohr. U. (1986). Carcinogenicity study with sodium dichromate in
rats. Exp Toxicol Pathol 30: 129-141. http://dx.doi.org/10.1016/S0232-1513C86180085-8
Stern. AH: Yu. CH: Black. K: Lin. L: Liov. PI: Gochfeld. M: Fan. ZH. (2010). Hexavalent chromium in
house dust-a comparison between an area with historic contamination from chromate
production and background locations. Sci Total Environ 408: 4993-4998.
http://dx.doi.Org/10.1016/i.scitotenv.2010.07.035
Stift. A: Friedl. I: Laengle. F. (1998). Liver transplantation for potassium dichromate poisoning
[Letter], N Engl J Med 338: 766-767. http://dx.doi.org/10.1056/NETM199803123381115
Stift. A: Friedl. I: Langle. F: Berlakovich. G: Steininger. R: Miihlbacher. F. (2000). Successful
treatment of a patient suffering from severe acute potassium dichromate poisoning with
liver transplantation. Transplantation 69: 2454-2455.
This document is a draft for review purposes only and does not constitute Agency policy.
R-45 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Stout. MP: Nvska. A: Collins. BT: Witt. KL: Kissling. GE: Malarkev. DE: Hooth. Ml. (2009). Chronic
toxicity and carcinogenicity studies of chromium picolinate monohydrate administered in
feed to F344/N rats and B6C3F1 mice for 2 years. Food Chem Toxicol 47: 729-733.
http://dx.doi.Org/10.1016/i.fct.2009.01.006
Subramanian. S: Raiendiran. G: Sekhar. P: Gowri. C: Govindaraiulu. P: Aruldhas. MM. (2006).
Reproductive toxicity of chromium in adult bonnet monkeys (Macaca radiata Geoffrey).
Reversible oxidative stress in the semen. Toxicol Appl Pharmacol 215: 237-249.
http://dx.doi.Org/10.1016/i.taap.2006.03.004
Sudha. S: Kripa. SK: Shibilv. P: Toseph. S: Balachandar. V. (2011). Biomonitoring of genotoxic effects
among shielded manual metal arc welders. Asian Pac J Cancer Prev 12: 1041-1044.
Sugden. KD: Campo. CK: Martin. BP. (2001). Direct oxidation of guanine and 7,8-dihydro-8-
oxoguanine in DNA by a high-valent chromium complex: a possible mechanism for
chromate genotoxicity. Chem Res Toxicol 14: 1315-1322.
Sugden. KD: Stearns. DM. (2000). The role of chromium(V) in the mechanism of chromate-induced
oxidative DNA damage and cancer [Review], J Environ Pathol Toxicol Oncol 19: 215-230.
Sugivama. M: Tsuzuki. K: Ogura. R. (1991). Effect of ascorbic acid on DNA damage, cytotoxicity,
glutathione reductase, and formation of paramagnetic chromium in Chinese hamster V-79
cells treated with sodium chromate(VI). J Biol Chem 266: 3383-3386.
Suh. M: Thompson. CM: Kirman. CR: Carakostas. MC: Haws. LC: Harris. MA: Proctor. DM. (2014).
High concentrations of hexavalent chromium in drinking water alter iron homeostasis in
F344 rats and B6C3F1 mice. Food Chem Toxicol 65: 381-388.
http://dx.doi.Org/10.1016/i.fct.2014.01.009
Suh. M: Wikoff. D: Lipworth. L: Goodman. M: Fitch. S: Mittal. L: Ring. C: Proctor. D. (2019).
Hexavalent chromium and stomach cancer: a systematic review and meta-analysis. Crit Rev
Toxicol 49: 140-159. http://dx.doi.org/10.1080/10408444.2019.1578730
Suzuki. Y. (1990). Synergism of ascorbic acid and glutathione in the reduction of hexavalent
chromium in vitro. Ind Health 28: 9-19.
Suzuki. Y: Fukuda. K. (1990). Reduction of hexavalent chromium by ascorbic acid and glutathione
with special reference to the rat lung. Arch Toxicol 64: 169-176.
http://dx.doi.org/10.1007/BF02010721
Takahashi. Y: Kondo. K: Hirose. T: Nakagawa. H: Tsuvuguchi. M: Hashimoto. M: Sano. T: Ochiai. A:
Monden. Y. (2005). Microsatellite instability and protein expression of the DNA mismatch
repair gene, hMLHl, of lung cancer in chromate-exposed workers. Mol Carcinog 42: 150-
158. http://dx.doi.org/10.1002/mc.20073
Talattof. A: Amidon. GL. (2018). Pulse packet stochastic model for gastric emptying in the fasted
state: A physiological approach. Mol Pharm 15: 2107-2115.
http://dx.doi.org/10.1021/acs.molpharmaceut7b01077
Tanigawa. T: Araki. S: Sata. F: Nakata. A: Araki. T. (1998). Effects of smoking, aromatic amines, and
chromates on CD4+ and CD8+ T lymphocytes in male workers [Comment], Environ Res 78:
59-63. http://dx.doi.org/10.1006/enrs.1997.3827
Taniguchi. K: Karin. M. (2018). NF-kB, inflammation, immunity and cancer: coming of age. Nat Rev
Immunol309-324. http://dx.doi.org/10.1038/nri.2017.142
Tarrant. TM. (2010). Blood cytokines as biomarkers of in vivo toxicity in preclinical safety
assessment: Considerations for their use [Review], Toxicol Sci 117: 4-16.
http://dx.doi.org/10.1093/toxsci/kfql34
TCEO (Texas Commission on Environmental Quality). (2014). Development support document for
hexavalent chromium (particulate compounds). Austin, TX.
Than. BL: Linnekamp. IF: Starr. TK: Largaespada. DA: Rod. A: Zhang. Y: Bruner. V: Abrahante. 1:
Schumann. A: Luczak. T: Walter: Niemczvk. A: O'Sullivan. MG: Medema. TP: Fiineman. RT:
Meiier. GA: Van Den Broek. E: Hodges. CA: Scott. PM: Vermeulen. L: Cormier. RT. (2016).
This document is a draft for review purposes only and does not constitute Agency policy.
R-46 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
CFTR is a tumor suppressor gene in murine and human intestinal cancer. Oncogene 35:
4179-4187. http://dx.doi.org/10.1038/onc.2015.483
Thermo Fisher (Thermo Fisher Scientific Inc.). (2009). Safety data sheet: Potassium dichromate.
Thomann. RV: Snyder. CA: Squibb. KS. (1994). Development of a pharmacokinetic model for
chromium in the rat following subchronic exposure: I. The importance of incorporating
long-term storage compartment Toxicol Appl Pharmacol 128: 189-198.
http://dx.doi.org/10.1006/taap.1994.1197
Thompson. CM: Donahue. DA: Hobbs. C: Costecalde. Y: Franzen. A: Suh. M: Proctor. DM: Harris. MA.
(2020). Exposure to environmentally-relevant concentrations of hexavalent chromium does
notinduce ovarian toxicity in mice. Regul Toxicol Pharmacol 116: 104729.
http: / /dx.doi. org/10.1016/i .vrtph. 2020.104729
Thompson. CM: Fedorov. Y: Brown. DP: Suh. M: Proctor. DM: Kuriakose. L: Haws. LC: Harris. MA.
(2012a). Assessment of Cr(VI)-induced cytotoxicity and genotoxicity using high content
analysis. PLoS ONE 7: e42720. http://dx.doi.org/10.1371/iournal.pone.0042720
Thompson. CM: Kirman. CR: Proctor. DM: Haws. LC: Suh. M: Hays. SM: Hixon. TG: Harris. MA. (2014).
A chronic oral reference dose for hexavalent chromium-induced intestinal cancer. J Appl
Toxicol 34: 525-536. http://dx.doi.org/10.1002/iat2907
Thompson. CM: Proctor. DM: Haws. LC: Hebert. CD: Grimes. SD: Shertzer. HG: Kopec. AK: Hixon. TG:
Zacharewski. TR: Harris. MA. (2011). Investigation of the mode of action underlying the
tumorigenic response induced in B6C3F1 mice exposed orally to hexavalent chromium.
Toxicol Sci 123: 58-70. http://dx.doi.org/10.1093/toxsci/kfrl64
Thompson. CM: Proctor. DM: Suh. M: Haws. LC: Hebert. CD: Mann. IF: Shertzer. HG: Hixon. TG:
Harris. MA. (2012b). Comparison of the effects of hexavalent chromium in the alimentary
canal of F344 rats and B6C3F1 mice following exposure in drinking water: implications for
carcinogenic modes of action. Toxicol Sci 125: 79-90.
http://dx.doi.org/10.1093/toxsci/kfr28Q
Thompson. CM: Proctor. DM: Suh. M: Haws. LC: Kirman. CR: Harris. MA. (2013). Assessment of the
mode of action underlying development of rodent small intestinal tumors following oral
exposure to hexavalent chromium and relevance to humans [Review], Crit Rev Toxicol 43:
244-274. http:/ /dx.doi.org/10.3109/10408444.2013.768596
Thompson. CM: Seiter. 1: Chappell. MA: Tappero. RV: Proctor. DM: Suh. M: Wolf. TC: Haws. LC: Vitale.
R: Mittal. L: Kirman. CR: Hays. SM: Harris. MA. (2015a). Synchrotron-based imaging of
chromium and y-H2AX immunostaining in the duodenum following repeated exposure to
Cr(VI) in drinking water. Toxicol Sci 143: 16-25. http://dx.doi.org/10.1093 /toxsci/kfu206
Thompson. CM: Wolf. TC: Elbekai. RH: Paranipe. MG: Seiter. TM: Chappell. MA: Tappero. RV: Suh. M:
Proctor. DM: Bichteler. A: Haws. LC: Harris. MA. (2015b). Duodenal crypthealth following
exposure to Cr(VI): Micronucleus scoring, y-H2AX immunostaining, and synchrotron X-ray
fluorescence microscopy. MutatRes GenetToxicol Environ Mutagen 789-790: 61-66.
http://dx.doi.Org/10.1016/j.mrgentox.2015.05.004
Thompson. CM: Young. RR: Dinesdurage. H: Suh. M: Harris. MA: Rohr. AC: Proctor. DM. (2017).
Assessment of the mutagenic potential of hexavalent chromium in the duodenum of big
blue® rats. Toxicol Appl Pharmacol 330: 48-52.
http://dx.doi.Org/10.1016/j.taap.2017.07.002
Thompson. CM: Young. RR: Suh. M: Dinesdurage. HR: Elbekai. RH: Harris. MA: Rohr. AC: Proctor.
DM. (2015c). Assessment of the mutagenic potential of Cr(VI) in the oral mucosa of Big
Blue((R)) transgenic F344 rats. Environ Mol Mutagen 56: 621-628.
http://dx.doi.org/10.10Q2/em.21952
Tian. M: Guo. F: Sun. Y: Zhang. W: Miao. F: Liu. Y: Song. G: Ho. CL: Yu. X: Sun. TZ: Wong. WY. (2014). A
fluorescent probe for intracellular cysteine overcoming the interference by glutathione. Org
Biomol Chem 12: 6128-6133. http://dx.doi.org/10.1039/c4ob00382a
This document is a draft for review purposes only and does not constitute Agency policy.
R-47 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
Tielemans. E: Burdorf. A: Te Velde. ER: Weber. RFA: Van Kooij. RT: Veulemans. H: Heederik. PIT.
(1999). Occupationally related exposures and reduced semen quality: A case-control study.
Fertil Steril 71: 690-696.
Toebak. Ml: Pohlmann. PR: Sampat-Sardioepersad. SC: von Blomberg. BM: Bruvnzeel. DP: Scheper.
RT: Rustemever. T: Gibbs. S. (2006). CXCL8 secretion by dendritic cells predicts contact
allergens from irritants. Toxicol In Vitro 20: 117-124.
http://dx.doi.Org/10.1016/i.tiv.2005.06.039
Tong. Z: Zhang. C. (2003). [Study on liver toxic damage of occupational chromium exposure],
Anquan Jiankang he Huanjing/Safety Health and Environment (3): 21.
Travlos. GS: Mirris. RW: Elwell. MR: Duke. A: Rosenblum. S: Thompson. MB. (1996). Frequency and
relationships of clinical chemistry and liver and kidney histopathology findings in 13-week
toxicity studies in rats. Toxicology 107: 17-29. http://dx.doi.org/10.1016/030Q-
483XC95103197-N
Trivedi. B: Saxena. DK: Murthv. RC: Chandra. SV. (1989). Embryotoxicity and fetotoxicity of orally
administered hexavalent chromium in mice. Reprod Toxicol 3: 275-278.
http://dx.doi.org/10.1016/0890-6238r89190022-l
Tsao. DA: Tseng. WC: Chang. HR. (2011). The expression of RKIP, RhoGDI, galectin, c-Myc andp53
in gastrointestinal system of Cr(VI)-exposed rats. J Appl Toxicol 31: 730-740.
http://dx.doi.org/10.1002/iat.1621
Tsukamoto. T: Mizoshita. T: Tatematsu. M. (2007). Animal models of stomach carcinogenesis
[Review], Toxicol Pathol 35: 636-648. http://dx.doi.org/10.1080/01926230701420632
Tullv. DB: Collins. BT: Overstreet. ID: Smith. CS: Dinse. GE: Mumtaz. MM: Chapin. RE. (2000). Effects
of arsenic, cadmium, chromium, and lead on gene expression regulated by a battery of 13
different promoters in recombinant HepG2 cells. Toxicol Appl Pharmacol 168: 79-90.
http://dx.doi.Org/10.1006/taap.2000.9014
Turk. TL: Parker. D. (1977). Sensitization with Cr, Ni, and Zr salts and allergic type granuloma
formation in the guinea pig. J Invest Dermatol 68: 341-345.
U.S. EPA (U.S. Environmental Protection Agency). (1986a). Effect of chlorine on chromium
speciationin tap water [EPA Report], (EPA/600/M-86/015). Washington, D.C.: U.S.
Enviromental Protection Agency.
U.S. EPA (U.S. Environmental Protection Agency). (1986b). Guidelines for carcinogen risk
assessment [EPA Report] (pp. 33993-34003). (EPA/630/R-00/004). Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/iris/basic-information-about-integrated-risk-information-
svstem#risk
U.S. EPA (U.S. Environmental Protection Agency). (1986c). Guidelines for mutagenicity risk
assessment (pp. 1-17). (ISSN 0097-6326
EISSN 2167-2520
EPA/630/R-98/003). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
Forum, https://www.epa.gov/risk/guidelines-mutagenicity-risk-assessment
U.S. EPA (U.S. Environmental Protection Agency). (1988). Recommendations for and documentation
of biological values for use in risk assessment [EPA Report], (EPA600687008). Cincinnati,
OH. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=34855
U.S. EPA (U.S. Environmental Protection Agency). (1991). Guidelines for developmental toxicity risk
assessment Fed Reg 56: 63798-63826.
U.S. EPA (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
reference concentrations and application of inhalation dosimetry [EPA Report],
(EPA600890066F). Research Triangle Park, NC.
This document is a draft for review purposes only and does not constitute Agency policy.
R-48 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
https://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=71993&CFID=51174829&CFTOKE
N=25006317
U.S. EPA (U.S. Environmental Protection Agency). (1996a). Guidelines for reproductive toxicity risk
assessment (pp. 1-143). (EPA/630/R-96/009). Washington, DC: U.S. Environmental
Protection Agency, Risk Assessment Forum.
https://www.epa.gov/sites/production/files/2014-
11/documents/guidelines repro toxicity.pdf
U.S. EPA (U.S. Environmental Protection Agency). (1996b). Guidelines for reproductive toxicity risk
assessment Fed Reg 61: 56274-56322.
U.S. EPA (U.S. Environmental Protection Agency). (1996c). Proposed guidelines for carcinogen risk
assessment: notice of availability and opportunity to comment. Fed Reg 61: 17960-18011.
U.S. EPA (U.S. Environmental Protection Agency). (1998a). Guidelines for neurotoxicity risk
assessment [EPA Report] (pp. 1-89). (ISSN 0097-6326
EISSN 2167-2520
EPA/630/R-95/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
Forum, http://www.epa.gov/risk/guidelines-neurotoxicity-risk-assessment
U.S. EPA (U.S. Environmental Protection Agency). (1998b). Health effects test guidelines OPPTS
870.7800 immunotoxicity. (EPA 712-C-98-351). Washington, DC: Prevention, Pesticides and
Toxic Substances, U.S. Environmental Protection Agency.
http://www.epa.gov/ocspp/pubs/frs/publications/Test Guidelines/series870.htm
U.S. EPA (U.S. Environmental Protection Agency). (1998c). Toxicological review of hexavalent
chromium [EPA Report] (pp. 80). (630R98009). Washington, DC.
https: //cfpub.epa.gov/ncea/iris /iris documents/documents/toxreviews /0144tr.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and
reference concentration processes. (EPA630P02002F). Washington, DC.
https://www.epa.gov/sites/production/files/2014-12/documents/rfd-final.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2005a). Guidelines for carcinogen risk
assessment [EPA Report], (EPA630P03001F). Washington, DC.
https://www.epa.gov/sites/production/files/2013-
09/documents/cancer guidelines final 3-25-05.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2005b). Supplemental guidance for assessing
susceptibility from early-life exposure to carcinogens [EPA Report], (EPA/630/R-03/003F).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/risk/supplemental-guidance-assessing-susceptibility-early-life-
exposure-carcinogens
U.S. EPA (U.S. Environmental Protection Agency). (2006a). 2006 community water system survey-
Volume I: Overview [EPA Report], (EPA 815-R-09-001). Washington, DC: U.S.
Environmental Protection Agency, Office of Water.
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=Pl 009TTI.txt
U.S. EPA (U.S. Environmental Protection Agency). (2006b). 2006 community water system survey -
Volume II: Detailed tables and survey methodology [EPA Report], (EPA 815-R-09-002).
Washington, DC: U.S. Environmental Protection Agency, Office of Water.
https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P1009USA.txt
U.S. EPA (U.S. Environmental Protection Agency). (2006c). Approaches for the application of
physiologically based pharmacokinetic (PBPK) models and supporting data in risk
assessment (Final Report) [EPA Report] (pp. 1-123). (EPA/600/R-05/043F). Washington,
DC: U.S. Environmental Protection Agency, Office of Research and Development, National
Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 57668
This document is a draft for review purposes only and does not constitute Agency policy.
R-49 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
U.S. EPA (U.S. Environmental Protection Agency). (2006d). A framework for assessing health risk of
environmental exposures to children (pp. 1-145). (EPA/600/R-05/093F). Washington, DC:
U.S. Environmental Protection Agency, Office of Research and Development, National Center
for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 58363
U.S. EPA (U.S. Environmental Protection Agency). (2008). Reregistration eligibility decision for
chromated arsenicals. (EPA 739-R-08-006). Office of Prevention, Pesticides And Toxic
Substances; U.S. Environmental Protection Agency.
U.S. EPA (U.S. Environmental Protection Agency). (2009). Risk assessment guidance for superfund
volume I: Human health evaluation manual (PartF, supplemental guidance for inhalation
risk assessment): Final [EPA Report], (EPA/540/-R-070/002). Washington, DC.
https://www.epa.gov/risk/risk-assessment-guidance-superfund-rags-part-f
U.S. EPA (U.S. Environmental Protection Agency). (2011a). Exposure factors handbook: 2011
edition [EPA Report], (EPA/600/R-090/052F). Washington, DC: U.S. Environmental
Protection Agency, Office of Research and Development, National Center for Environmental
Assessment. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100F20S.txt
U.S. EPA (U.S. Environmental Protection Agency). (2011b). (Fourth) Five-year review report for
United Chrome Products ORD 009043001 Corvallis, Benton County, Oregon Draft August
2011 revision 5. Seattle, WA: U.S. Environmental Protection Agency Region 10.
U.S. EPA (U.S. Environmental Protection Agency). (2011c). Recommended use ofbody weight 3/4
as the default method in derivation of the oral reference dose. (EPA100R110001).
Washington, DC. https://www.epa.gov/sites/production/files/2013-
09/documents/recommended-use-of-bw3 4.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2011d). Toxicological review of
trichloroethylene (CAS No. 79-01-6) in support of summary Information on the Integrated
Risk Information System (IRIS) [EPA Report], (EPA635R09011F). Washington, DC.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100CB6V.txt
U.S. EPA (U.S. Environmental Protection Agency). (2012a). Benchmark Dose Software (BMDS).
Version 2.2. Retrieved from http://www.epa.gov/NCEA/bmds/index.html
U.S. EPA (U.S. Environmental Protection Agency). (2012b). Benchmark dose technical guidance
[EPA Report], (EPA100R12001). Washington, DC: U.S. Environmental Protection Agency,
Risk Assessment Forum, https: //www.epa.gov/risk/benchmark-dose-technical-guidance
U.S. EPA (U.S. Environmental Protection Agency). (2012c). The third Unregulated Contaminant
Monitoring Rule (UCMR 3). Presentation presented at US EPA Webinar, September, 2012,
OGWDW, SRMD Technical Support Center Cincinnati Ohio.
U.S. EPA (U.S. Environmental Protection Agency). (2012d). Toxicological review of
tetrachloroethylene (Perchloroethylene) (CASRN 127-18-4) in support of summary
information on the Integrated Risk Information System (IRIS) [EPA Report],
(EPA635R08011F). Washington, DC: National Center for Environmental Assessment
https: //cfpub.epa.gov/ncea/iris /iris documents/documents/toxreviews /0106tr.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2014a). Discharge monitoring report (DMR)
pollutant loading tool. Washington, DC. Retrieved from
https://cfpub.epa.gov/dmr/index.cfm
U.S. EPA (U.S. Environmental Protection Agency). (2014b). Preliminary materials for the Integrated
Risk Information System (IRIS) toxicological review of hexavalent chromium part 1:
Experimental animal studies [CASRN) 18540-29-9] [EPAReport], (EPA/635/R-14/094).
Research Triangle Park, NC. https://www.epa.gov/iris/iris-bimonthly-public-meeting-iun-
2014
U.S. EPA (U.S. Environmental Protection Agency). (2014c). Preliminary materials for the Integrated
Risk Information System (IRIS) toxicological review of hexavalent chromium part 2: Human,
This document is a draft for review purposes only and does not constitute Agency policy.
R-50 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
toxicokinetic, and mechanistic Studies [CASRN) 18540-29-9] [EPA Report], (EPA/635/R-
14/218). Research Triangle Park, NC. https://www.ep a. go v /ir is /ir is-bimonthlv-public-
meeting-oct-2014
U.S. EPA (U.S. Environmental Protection Agency). (2014d). The third Unregulated Contaminant
Monitoring Rule (UCMR 3) occurrence data. Retrieved from
http://water.epa.gOv/lawsregs/rulesregs/sdwa/ucmr/data.cfm#ucmr2013
U.S. EPA (U.S. Environmental Protection Agency). (2014e). Toxicological review of libby amphibole
asbestos: In support of summary information on the Integrated Risk Information System
(IRIS) [EPA Report], (EPA/635/R-11/002F). Washington, DC: Integrated Risk Information
System, National Center for Environmental Assessment, Office of Research and
Development
https: //cfpub.epa.gov/ncea/iris /iris documents/documents/toxreviews /1026tr.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2016a). 2014 National Emissions Inventory
(NEI) data (Version 2). Washington, DC. Retrieved from https://www,epa.gov/air-
emissions-inventories/2014-national-emissions-inventory-nei-data
U.S. EPA (U.S. Environmental Protection Agency). (2016b). Evaluation of the inhalation
carcinogenicity of ethylene oxide (CASRN 75-21-8): In support of summary information on
the Integrated Risk Information System (IRIS) [EPA Report], (EPA635R16350Fa).
Washington, DC: U.S. Environmental Protection Agency, Office of Research and
Development, National Center for Environmental Assessment
https: //cfpub.epa.gov/ncea/iris /iris documents/documents/toxreviews /1025tr.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2016c). US EPA's report on the environment
(ROE). Available online at https://cfpub.epa.gov/roe/index.cfm
U.S. EPA (U.S. Environmental Protection Agency). (2018). Toxic Release Inventory (TRI). Retrieved
from https://www.epa.gov/enviro/tri-search
U.S. EPA (U.S. Environmental Protection Agency). (2019). Systematic Review Protocol for the
Hexavalent Chromium IRIS Assessment (Preliminary Assessment Materials) [EPA Report],
(EPA/635/R-18/155).
U.S. EPA (U.S. Environmental Protection Agency). (2020a). Benchmark Dose Software (BMDS).
Version 3.2: User guide [EPA Report], (EPA/600/R-20/216). Washington, DC: U.S.
Environmental Protection Agency, Office of Research and Development.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P10103T2.txt
U.S. EPA (U.S. Environmental Protection Agency). (2020b). 0RD staff handbook for developing IRIS
assessments (public comment draft) [EPAReport], (EPA/600/R-20/137). Washington, DC:
U.S. Environmental Protection Agency, Office of Research and Development, Center for
Public Health and Environmental Assessment
https://cfpub.epa.gov/ncea/iris drafts/recordisplay.cfm?deid=350086
U.S. EPA (U.S. Environmental Protection Agency). (2021a). Model log files for the IRIS toxicological
review of hexavalent chromium.
U.S. EPA (U.S. Environmental Protection Agency). (2021b). National Emissions Inventory (NEI).
Available online at https://www.epa.gov/air-emissions-inventories/national-emissions-
inventorv-nei (accessed February 10, 2021).
U.S. EPA (U.S. Environmental Protection Agency). (2021c). TRI Toxics Tracker. Available online at
https://edap.epa.gov/public/extensions/TRIToxicsTracker/TRIToxicsTracker.html
Uddin. AN: Burns. FT: Rossman. TG: Chen. H: Kluz. T: Costa. M. (2007). Dietary chromium and nickel
enhance UV-carcinogenesis in skin of hairless mice. Toxicol Appl Pharmacol 221: 329-338.
http://dx.doi.Org/10.1016/i.taap.2007.03.030
Urbano. AM: Ferreira. LMR: Alpoim. MC. (2012). Molecular and cellular mechanisms of hexavalent
chromium-induced lung cancer: An updated perspective [Review], Curr Drug Metab 13:
284-305.
This document is a draft for review purposes only and does not constitute Agency policy.
R-51 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
USGS (U.S. Geological Survey). (1995). Chromium life cycle study. (IC 9411). Washington, D.C.: U.S.
Department of the Interior, Bureau of Mines.
Vaglenov. A: Nosko. M: Georgieva. R: Carbonell. E: Creus. A: Marcos. R. (1999). Genotoxicity and
radioresistance in electroplating workers exposed to chromium. Mutat Res 446: 23-34.
Valentine. CC: Young. RR: Fielden. MR: Kulkarni. R: Williams. LN: Li. T: Minocherhomji. S: Salk. 11.
(2020). Direct quantification of in vivo mutagenesis and carcinogenesis using duplex
sequencing. Proc Natl Acad Sci USA 117: 33414-33425.
http://dx.doi.org/10.1073/pnas.2013724117
van Heerden. PV: Tenkins. IR: Woods. WP: Rossi. E: Cameron. PP. (1994). Death by tanning-a case
of fatal basic chromium sulphate poisoning [Review], Intensive Care Med 20: 145-147.
van Pud Alblas. AB: van Furth. R. (1979). Origin, Kinetics, and characteristics of pulmonary
macrophages in the normal steady state. J Exp Med 149: 1504-1518.
Vandamme. D: Herrero. A: Al-Mulla. F: Kolch. W. (2014). Regulation of the MAPK pathway by Raf
kinase inhibitory protein [Review], CritRev Oncog 19: 405-415.
http://dx.doi.org/10.1615/critrevoncog.2014011922
Vanhauwaert. A: Vanparvs. P: Kirsch-Volders. M. (2001). The in vivo gut micronucleus test detects
clastogens and aneugens given by gavage. Mutagenesis 16: 39-50.
http://dx.doi.Org/10.1093/mutage/16.l.39
Vengosh. A: Covte. R: Karr. 1: Harkness. IS: Kondash. AT: Ruhl. LS: Merola. RB: Dvwer. GS. (2016).
Origin of hexavalent chromium in drinking water wells from the piedmont aquifers of North
Carolina. Environ Sci Technol Lett 3: 409-414.
http://dx.doi.org/10.1021/acs.estlett.6b00342
Verschoor. MA: Bragt. PC: Herber. RFM: Zielhuis. RL: Zwennis. WCM. (1988). Renal function of
chrome-plating workers and welders. Int Arch Occup Environ Health 60: 67-70.
http://dx.doi.Org/10.1007/BF00409381
Videla. LA: Rodrigo. R: Orellana. M: Fernandez. V: Tapia. G: Ouinones. L: Varela. N: Contreras. 1:
Lazarte. R: Csendes. A: Roias. I: Maluenda. F: Burdiles. P: Diaz. TC: Smok. G: Thielemann. L:
Poniachik. 1. (2004). Oxidative stress-related parameters in the liver of non-alcoholic fatty
liver disease patients. Clin Sci (Lond) 106: 261-268.
http://dx.doi.org/10.1042 /CS20030285
Viechtbauer. W. (2010). Conducting meta-analyses in R with the metafor package. J Stat Softw 36: 1-
48.
Vigliani. EC: Zurlo. N. (1955). Erfahrungen der Clinica del Lavoro mit einigen maximalen
Arbeitsplatzkonzentrationen (MAK) von Industriegiften. Int Arch Occup Environ Health 13:
528-534. http://dx.doi.org/10.100 7 /BFOO 312 75 6
Vincent. 1. (2013). Chromium: is it essential, pharmacologically relevant, or toxic? In Interrelations
between Essential Metal Ions and Human Diseases: Springer.
http://dx.doi.Org/10.1007/978-94-007-7500-8 6
Vincent. IB. (2017). New evidence against chromium as an essential trace element. J Nutr 147:
2212-2219. http://dx.doi.org/10.3945 /in.l 17.255901
Vreeburg. KT: de Groot. K: van Hoogstraten. IM: von Blomberg. BM: Scheper. RT. (1991). Successful
induction of allergic contact dermatitis to mercury and chromium in mice. International
Arch Allergy Appl Immunol 96: 179-183. http: //dx.doi.org/10.1159 /000235491
Wang. BT: Guo. YL: Chang. HY: Sheu. HM: Pan. MH: Lee. YH: Wang. YT. (2010a). N-acetylcysteine
inhibits chromium hypersensitivity in coadjuvant chromium-sensitized albino guinea pigs
by suppressing the effects of reactive oxygen species. Exp Dermatol 19: el91-e200.
http://dx.doi.Org/10.llll/i.1600-0625.2009.01045.x
Wang. BT: Sheu. HM: Guo. YL: Lee. YH: Lai. CS: Pan. MH: Wang. YT. (2010b). Hexavalent chromium
induced ROS formation, Akt, NF-kappaB, and MAPK activation, and TNF-alpha and IL-lalpha
This document is a draft for review purposes only and does not constitute Agency policy.
R-52 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Toxicological Review ofHexavalent Chromium
production in keratinocytes. Toxicol Lett 198: 216-224.
http://dx.doi.Org/10.1016/i.toxlet2010.06.024
Wang. L: Oiu. TG: He. 1: Liu. WT: Ge. X: Zhou. FM: Huang. YX: Tiang. BH: Liu. LZ. (2019). Suppression of
miR-143 contributes to overexpression of IL-6, HIF-la and NF-kB p65 in Cr(VI)-induced
human exposure and tumor growth. Toxicol Appl Pharmacol 378: 114603.
http://dx.doi.Org/10.1016/i.taap.2019.114603
Wang. TC: Tia. G: Song. Y. anS: Zhang. 1. i: Ma. Y. anHui: Feng. W. eiYue: Leu. L. anZ: Zhang. N: Zhou.
TW. en: Yan. L. ei: Zhen. S. en: Oian. 0. in. (2012a). Effects of chronic chromate exposure on
human serum prostate specific antigen: A cross sectional study. Ind Health 50: 95-102.
http://dx.doi.org/10.2486/indhealth.MS1278
Wang. TC: Song. YS: Wang. H: Zhang. 1: Yu. SF: Gu. YF: Chen. T: Wang. Y: Shen. HO: Tia. G. (2012b).
Oxidative DNA damage and global DNA hypomethylation are related to folate deficiency in
chromate manufacturing workers. J Hazard Mater 213-214: 440-446.
http://dx.doi.Org/10.1016/i.ihazmat.2012.02.024
Wang. X: Oin. 0: Xu. X: Xu. 1: Wang. 1: Zhou. 1: Huang. S: Zhai. W: Zhou. H: Chen. 1. (1994). Chromium-
Induced Early Changes in Renal Function among Ferrochromium-Producing Workers.
Toxicology 90: 93-101. http://dx.doi.org/10.1016/0300-483Xr94190208-9
Wang. XF: Lou. XM: Shen. Y: Xing. ML: Xu. LH. (2010c). Apoptotic-related protein changes induced
by hexavalent chromium in mice liver. Environ Toxicol 25: 77-82.
http://dx.doi.org/10.1002/tox.20478
Wang. Y: Wu. W: Yao. C: Lou. I: Chen. R: Tin. L: Wu. N: Gao. M: Song. P: Tan. Y: Liu. K. (2015). Elevated
tissue Cr levels, increased plasma oxidative markers, and global hypomethylation of blood
DNA in male Sprague-Dawley rats exposed to potassium dichromate in drinking water.
Environ Toxicol, http: / /dx.doi.org/10.1002 /tox.22117
Wang. Z: Peng. H: Zhang. R: Tiang. X: Chen. S: Guo. P: Xiao. Y: Zeng. X: Wang. 0: Aschner. M: Chou.
WC: Li. D: Chen. W: Chen. L. (2022). Assessment of intestinal injury of hexavalent chromium
using a modified in vitro gastrointestinal digestion model. Toxicol Appl Pharmacol 436:
115880. http://dx.doi.Org/10.1016/i.taap.2022.115880
Wang. Z: Yang. C. (2019). Metal carcinogen exposure induces cancer stem cell-like property through
epigenetic reprograming: A novel mechanism of metal carcinogenesis [Review], Semin
Cancer Biol 57: 95-104. http://dx.doi.Org/10.1016/i.semcancer.2019.01.002
Ward. FW: Coates. ME. (1987). Gastrointestinal pH measurement in rats: Influence of the microbial
flora, diet and fasting. Lab Anim 21: 216-222.
http://dx.doi.org/10.1258/002367787781268693
Wear. HM: Mcpike. Ml: Watanabe. KH. (2016). From primordial germ cells to primordial follicles: A
review and visual representation of early ovarian development in mice [Review], Journal of
Ovarian Research 9:36. http://dx.doi.org/10.1186/sl3048-016-Q246-7
Welling. R: Beaumont. II: Petersen. ST: Alexeeff. GV: Steinmaus. C. (2015). Chromium VI and stomach
cancer: a meta-analysis of the current epidemiological evidence [Review], Occup Environ
Med 72: 151-159. http://dx.doi.org/10.1136/oemed-2014-102178
Westbrook. A. vaM: Szakmarv. A: Schiestl. RH. (2010). Mechanisms of intestinal inflammation and
development of associated cancers: Lessons learned from mouse models [Review], Mutat
Res Rev Mutat Res 705: 40-59. http: //dx.doi.Org/10.1016/i.mrrev.2010.03.001
Whalan. IE. (2000). A guide to clinical pathology in animals. Whalan, JE.
Whalan. TE. (2015). Clinical Chemistry. 67-94. http://dx.doi.org/10.1007/978-3-319-15853-2 6
WHO (World Health Organization). (2003). Chromium in drinking water.
(WHO/SDE/WSH/03.04/04). Geneva, Switzerland.
http://www.who.int/water sanitation health/dwq/chemicals/chromium.pdf
This document is a draft for review purposes only and does not constitute Agency policy.
R-53 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
WHO (World Health Organization). (2010). WHO laboratory manual for the examination and
processing of human semen. In WHO laboratory manual for the examination and processing
of human semen (5th ed.). Geneva, Switzerland.
Wiegand. HI: Ottenwalder. H: Bolt. HM. (1985). Fast uptake kinetics in vitro of 51Cr (VI) by red
blood cells of man and rat Arch Toxicol 57: 31-34. http: //dx.doi.org/10.1007/BF0Q286571
Wiegand. HI: Ottenwalder. H: Bolt. HM. (1986). Disposition of a soluble chromate in the isolated
perfused rat liver. Xenobiotica 16: 839-844.
http://dx.doi.Org/10.3109/00498258609038965
Williams. AB: Schumacher. B. (2016). p53 in the DNA-Damage-Repair Process [Review], Cold Spring
Harb Perspect Med 6. http: //dx.doi.org/10.1101 /cshperspect.aO26070
Wisconsin DHS (Wisconsin Department of Health Services). (2010). Chromium. Available online at
http://www.dhs.wisconsin.gov/eh/ChemFS/fs/chromium.htm
Wise. TP: Young. TL: Cai. I: Cai. L. (2022). Current understanding of hexavalent chromium [Cr(VI)]
neurotoxicity and new perspectives [Review], Environ Int 158: 106877.
http://dx.doi.Org/10.1016/i.envint.2021.106877
Wise. SS: Aboueissa. AE: Martino. I: Wise. TP. (2018). Hexavalent chromium-induced chromosome
instability drives permanent and heritable numerical and structural changes and a dna
repair-deficientphenotype. Cancer Res 78: 4203-4214. http://dx.doi.org/10.1158/00Q8-
5472.CAN-18-0531
Wise. SS: Holmes. AL: Liou. L: Adam. RM: Wise. TP. (2016). Hexavalent chromium induces
chromosome instability in human urothelial cells. Toxicol Appl Pharmacol 296: 54-60.
http://dx.doi.Org/10.1016/i.taap.2016.02.015
Wise. SS: Wise. TP. Sr. (2010). Aneuploidy as an early mechanistic event in metal carcinogenesis
[Review], Biochem Soc Trans 38: 1650-1654. http://dx.doi.org/10.1042 /BST0381650
Witt. KL: Stout. MP: Herbert. RA: Travlos. GS: Kissling. GE: Collins. BT: Hooth. Ml. (2013).
Mechanistic Insights from the NTP Studies of Chromium. Toxicol Pathol 41: 326-342.
http://dx.doi.org/10.1177/0192623312469856
Wong. V: Armknecht. S: Zhitkovich. A. (2012). Metabolism of Cr(VI) by ascorbate but not
glutathione is a low oxidant-generating process. 26: 192-196.
http://dx.doi.Org/10.1016/i.itemb.2012.04.016
Wu. FY: Wu. WY: Kuo. HW: Liu. CS: Wang. RY: Lai. IS. (2001). Effect of genotoxic exposure to
chromium among electroplating workers in Taiwan. Sci Total Environ 279: 21-28.
Wultsch. G: Nersesvan. A: Kundi. M: Takse. R: Beham. A: Wagner. KH: Knasmueller. S. (2014). The
sensitivity of biomarkers for ge no toxicity and acute cytotoxicity in nasal and buccal cells of
welders. Int J Hyg Environ Health 217: 492-498.
http://dx.doi.Org/10.1016/i.iiheh.2013.09.005
Wultsch. G: Nersesvan. A: Kundi. M: Misik. M: Setavesh. T: Waldherr. M: Vodicka. P: Vodickova. L:
Knasmiiller. S. (2017). Genotoxic and cytotoxic effects in exfoliated buccal and nasal cells of
chromium and cobalt exposed electroplaters. J Toxicol Environ Health A 80: 1-10.
http: //dx.doi.org/10.1080/15287394.2017.1286918
Xia. W: Hu. 1: Zhang. B: Li. Y: Wise. TP: Bassig. BA: Zhou. A: Savitz. DA: Xiong. C: Zhao. 1: Du. X: Zhou.
Y: Pan. X: Yang. T: Wu. C: Tiang. M: Peng. Y: Oian. Z: Zheng. T: Xu. S. (2016). A case-control
study of maternal exposure to chromium and infant low birth weight in China.
Chemosphere 144: 1484-1489. http://dx.doi.Org/10.1016/i.chemosphere.2015.10.006
Xiao. F: Feng. X: Zeng. M: Guan. L: Hu. 0: Zhong. C. (2012a). Hexavalent chromium induces energy
metabolism disturbance and p53-dependent cell cycle arrest via reactive oxygen species in
L-02 hepatocytes. Mol Cell Biochem 371: 65-76. http://dx.doi.org/10.1007/sll010-012-
1423-7
Xiao. F: Li. Y: Dai. L: Deng. Y: Zou. Y: Li. P: Yang. Y: Zhong. C. (2012b). Hexavalent chromium targets
mitochondrial respiratory chain complex I to induce reactive oxygen species-dependent
This document is a draft for review purposes only and does not constitute Agency policy.
R-54 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Toxicological Review ofHexavalent Chromium
caspase-3 activation in L-02 hepatocytes. Int J Mol Med 30: 629-635.
http://dx.doi.org/10.3892/iimm.2012.1031
Xiao. F: Li. Y: Luo. L: Xie. Y: Zeng. M: Wang. A: Chen. H: Zhong. C. (2014). Role of mitochondrial
electron transport chain dysfunction in Cr(VI)-induced cytotoxicity in L-02 hepatocytes. Cell
Physiol Biochem 33: 1013-1025. http://dx.doi.org/10.1159/000358672
Xiao. Y: Zeng. M: Yin. L: Li. N: Xiao. F. (2019). Clusterin increases mitochondrial respiratory chain
complex I activity and protects against hexavalent chromium-induced cytotoxicity in L-02
hepatocytes. Toxicology Research 8: 15-24. http://dx.doi.org/10.1039/c8tx00231b
Xiaohua. L: Yanshuang. S: Li. W: Yuhui. L: Ti. Z: Yanhui. M: Yun. W: Weniun. M: Lei. Y: Guang. 1.
(2012). Evaluation of the correlation between genetic damage and occupational chromate
exposure through BNMN frequencies. J Occup Environ Med 54: 166-170.
http://dx.doi.Org/10.1097/TOM.0b013e31823d86b4
Xie. Y: Xiao. F: Luo. L: Zhong. C. (2014). Activation of autophagy protects against ROS-mediated
mitochondria-dependent apoptosis in L-02 hepatocytes induced by Cr(VI). Cell Physiol
Biochem 33: 705-716. http://dx.doi.org/10.1159/000358646
Xu. 1: Zhao. M: Pei. L: Zhang. R: Liu. X: Wei. L: Yang. M: Xu. 0. (2018). Oxidative stress and DNA
damage in a long-term hexavalent chromium-exposed population in North China: a cross-
sectional study. BMJ Open 8: e021470. http://dx.doi.org/10.1136/bmiopen-2017-021470
Yamate. 1: Izawa. T: Kuwamura. M. (2016). Histopathological analysis of rat hepatotoxicity based on
macrophage functions: In particular, an analysis for thioacetamide-induced hepatic lesions
[Review], Food Saf 4: 61-73. http://dx.doi.org/10.14252/foodsafetvfsci.2016012
Yang. F: Zhao. L: Mei. D: Tiang. L: Geng. C: Li. 0: Yao. X: Liu. Y: Kong. Y: Cao. 1. (2017). HMGA2 plays
an important role in Cr (Vl)-induced autophagy. Int J Cancer 141: 986-997.
http://dx.doi.org/10.1002/ijc.30789
Yang. L. (2006). Incidence and mortality of gastric cancer in China [Review], World J Gastroenterol
12: 17-20. http://dx.doi.org/10.3748/wig.vl2.il.17
Yazar. S: Yildirim. IY. (2018). Oxidative DNA damage in electroplaters co-exposed to chromium and
nickel. Fresen Environ Bull 27: 6831-6836.
Yi. X: Xiao. F: Zhong. X: Duan. Y: Liu. K: Zhong. C. (2017). A Ca(2+) chelator ameliorates chromium
(Vl)-induced hepatocyte L-02 injury via down-regulation of voltage-Dependent anion
channel 1 (VDAC1) expression. Environ Toxicol Pharmacol 49: 27-33.
http://dx.doi.Org/10.1016/i.etap.2016.ll.007
Yi. X: Zhang. Y: Zhong. C: Zhong. X: Xiao. F. (2016). The role of STIM1 in the Cr(vi)-induced [Ca(2+)]i
increase and cell injury in L-02 hepatocytes. Metallomics 8: 1273-1282.
http://dx.doi.org/10.1039/c6mt00204h
Yoisungnern. T: Das. 1: Choi. YI: Parnpai. R: Kim. TH. (2015). Effect of hexavalent chromium-treated
sperm on in vitro fertilization and embryo development. Toxicol Ind Health.
http://dx.doi.org/10.1177/0748233715579805
Younan. S: Sakita. GZ: Coluna. TGY: Rufino. MN: Keller. R: Bremer-Neto. H. (2019). Probiotic
mitigates the toxic effects of potassium dichromate in a preclinical study: A randomized
controlled trial. J Sci FoodAgric99: 183-190. http://dx.doi.org/10.1002 /isfa.9159
Younossi. ZM. (2019). Non-alcoholic fatty liver disease - A global public health perspective
[Review], J Hepatol 70: 531-544. http://dx.doi.Org/10.1016/i.ihep.2018.10.033
Yousef. MI: El-Demerdash. FM: Kamil. KI: Elaswad. FA. (2006). Ameliorating effect of folic acid on
chromium(VI)-induced changes in reproductive performance and seminal plasma
biochemistry in male rabbits. Reprod Toxicol 21: 322-328.
http://dx.doi.Org/10.1016/i.reprotox.2005.09.005
Yu. A: Tackson. T: Tsume. Y: Koenigsknecht. M: Wvsocki. I: Marciani. L: Amidon. GL: Frances. A:
Baker. TR: Hasler. W: Wen. B: Pai. A: Sun. D. (2017). Mechanistic Fluid Transport Model to
This document is a draft for review purposes only and does not constitute Agency policy.
R-55 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Toxicological Review ofHexavalent Chromium
Estimate Gastrointestinal Fluid Volume and Its Dynamic Change Over Time. AAPS J19:
1682-1690. http://dx.doi.org/10.1208/sl2248-017-0145-x
Yuan. TH. (1993). Modeling blood/plasma concentrations in dosed feed and dosed drinking water
toxicology studies. Toxicol Appl Pharmacol 119:131-141.
Yuan. WM: Lu. YO: Wei. Z: Chen. M: He. SL: Mai. WY: Zhang. MB. (2016). An epistaxis emergency
associated with multiple pollutants in elementary students. Biomed Environ Sci 29: 893-
897. http:/ /dx.doi.org/10.3967/hes2016.119
Yuan. Y: Ming. Z: Gong-Hua. H: Lan. G: Lu. D: Peng. L: Feng. 1: Cai-Gao. Z. (2012a). Cr(VI) induces the
decrease of ATP level and the increase of apoptosis rate mediated by ROS or VDAC1 in L-02
hepatocytes. Environ Toxicol Pharmacol 34: 579-587.
http://dx.doi.0rg/lO.lOl6/i.etap.2Oi2.O6.Ol6
Yuan. Y: Peng. L: Gong-Hua. H: Lu. D: Xia-Li. Z: Yu. Z: Cai-Gao. Z. (2012b). Chi-square analysis of the
reduction of ATP levels in L-02 hepatocytes by hexavalent chromium. 45: 482-487.
http://dx.doi.Org/10.1590/S0100-879X2012007500040
Zabulvte. D: Uleckiene. S: Drebickas. V: Semaska. V: Tonauskiene. I. (2009). Investigation of
combined effect of chromium (VI) and sodium fluoride in experiments on rats. Trace Elem
Electroly 26: 67-71.
Zabulvte. D: Uleckiene. S: Kalibatas. I: Paltanaviciene. A: Tuozulvnas. A: Gocentas. A. (2006).
Investigation of combined effect of chromium (VI) and nitrate in experiments on rats. Trace
Elem Electroly 23: 287-291.
Zahid. ZR: Al-hakkak. ZS: Kadhim. AHH: Elias. EA: Al-iumailv. IS. (1990). Comparative effects of
trivalent and hexavalent chromium on spermatogenesis of the mouse. Toxicol Environ
Chem 25: 131-136. http://dx.doi.org/10.1080/02772249009357512
Zendehdel. R: Shetab-Boushehri. SV: Azari. MR: Hosseini. V: Mohammadi. H. (2014). Chemometrics
models for assessment of oxidative stress risk in chrome-electroplating workers. Drug
Chem Toxicol 38: 1-6. http://dx.doi.org/10.3109/01480545.2014.922096
Zeng. M: Xiao. F: Zhong. X: Tin. F: Guan. L: Wang. A: Liu. X: Zhong. C. (2013). Reactive oxygen species
play a central role in hexavalent chromium-induced apoptosis in Hep3B cells without the
functional roles ofp53 and caspase-3. Cell Physiol Biochem 32: 279-290.
http://dx.doi.org/10.1159/00Q354436
Zev. IN: Lucas. C. (1985). Health hazard evaluation report no. HETA-82-358-1558, United Catalysts,
Inc. South Plant, Louisville, Kentucky. (HETA-82-358-1558). Cincinnati, OH: National
Institute for Occupational Safety and Health.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults.xhtml?search0uery=PB86116951
Zhang. B: Zhou. Y: Lin. N: Lowdon. RF: Hong. C: Nagaraian. RP: Cheng. IB: Li. D: Stevens. M: Lee. HI:
Xing. X: Zhou. 1: Sundaram. V: Elliott. G: Gu. I: Shi. T: Gascard. P: Sigaroudinia. M: Tlsty. TP:
Kadlecek. T: Weiss. A: O'Geen. H: Farnham. PI: Maire. CL: Ligon. KL: Madden. PA: Tam. A:
Moore. R: Hirst. M: Marra. MA: Zhang. B: Costello. IF: Wang. T. (2013). Functional DNA
methylation differences between tissues, cell types, and across individuals discovered using
the M&M algorithm. Genome Res 23: 1522-1540. http://dx.doi.org/10.1101/gr.156539.113
Zhang. ID: Li. S. (1997). RETRACTED: Cancer mortality in a Chinese population exposed to
hexavalent chromium in water. J Occup Environ Med 39: 315-319.
http://dx.doi.org/10.1097/00043764-199704000-00008
Zhang. ID: Li. XL. (1987a). [Chromium pollution of soil and water in Jinzhou], Zhonghua Yufang
Yixue Zazhi 21: 262-264.
Zhang. ID: Li. XL. (1987b). Investigation and Study of Chromium Pollution in Jinzhou. Zhonghua
Yufang Yixue Zazhi 21.
Zhang. R: Xiang. Y: Ran. 0: Deng. X: Xiao. Y: Xiang. L: Li. Z. (2014). Involvement of calcium, reactive
oxygen species, and ATP in hexavalent chromium-induced damage in red blood cells. Cell
Physiol Biochem 34: 1780-1791. http: //dx.doi.org/10.1159 /000366378
This document is a draft for review purposes only and does not constitute Agency policy.
R-56 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Toxicological Review ofHexavalent Chromium
Zhang. XH: Zhang. X: Wang. XC: Tin. LF: Yang. ZP: Tiang. CX: Chen. 0: Ren. XB: Cao. TZ: Wang. 0: Zhu.
YM. (2011). Chronic occupational exposure to hexavalent chromium causes DNA damage in
electroplating workers. BMC Public Health 11: 224. http://dx.doi.org/10.1186/1471-2458-
11-224
Zhang. Y: Su. Z: Hu. G: Hong. S: Long. C: Zhang. 0: Zheng. P: Wang. T: Yu. S: Yuan. F: Zhu. X: Tia. G.
(2022). Lung function assessment and its association with blood chromium in a chromate
exposed population. Sci Total Environ 818: 151741.
http://dx.doi.Org/10.1016/i.scitotenv.2021.151741
Zhang. Y: Xiao. F: Liu. X: Liu. K: Zhou. X: Zhong. C. (2017). Cr(VI) induces cytotoxicity in vitro
through activation of ROS-mediated endoplasmic reticulum stress and mitochondrial
dysfunction via the PI3K/Akt signaling pathway. Toxicol In Vitro 41: 232-244.
http://dx.doi.Org/10.1016/i.tiv.2017.03.003
Zhang. Y: Zhang. Y: Xiao. Y: Zhong. C: Xiao. F. (2019). Expression of Clusterin suppresses Cr(VI)-
induced premature senescence through activation of PI3K/AKT pathway. Ecotoxicol
Environ Saf 183: 109465. http://dx.doi.Org/10.1016/i.ecoenv.2019.109465
Zhang. Y: Zhang. Y: Zhong. C: Xiao. F. (2016). Cr(VI) induces premature senescence through R0S-
mediated p53 pathway in L-02 hepatocytes. Sci Rep 6(34578): 1-10.
http://dx.doi.org/10.1038/srep34578
Zhao. L: Song. Y: Pu. 1: Guo. 1: Wang. Y: Chen. Z: Chen. T: Gu. Y: Tia. G. (2014). Effects of repeated
Cr(VI) intratracheal instillation on club (Clara) cells and activation of nuclear factor-kappa B
pathway via oxidative stress. Toxicol Lett 231: 72-81.
http://dx.doi.Org/10.1016/i.toxlet2014.09.011
Zheng. W: Ge. F: Wu. K: Chen. X: Li. X: Chen. Y: Lv. Y: Lian. 0: Ge. RS. (2018). Inutero exposure to
hexavalent chromium disrupts rat fetal testis development Toxicol Lett 299: 201-209.
http://dx.doi.Org/10.1016/i.toxlet2018.10.010
Zhitkovich. A. (2005). Importance of chromium-DNA adducts in mutagenicity and toxicity of
chromium(VI) [Review], Chem Res Toxicol 18: 3-11. http://dx.doi.org/10.1021/tx049774+
Zhitkovich. A. (2011). Chromium in drinking water: sources, metabolism, and cancer risks [Review],
Chem Res Toxicol 24: 1617-1629. http: //dx.doi.org/10.1021/tx200251t
Zhitkovich. A: Voitkun. V: Costa. M. (1996). Formation of the amino acid-DNA complexes by
hexavalent and trivalent chromium in vitro: importance of trivalent chromium and the
phosphate group. Biochemistry 35: 7275-7282.
Zhong. X: de Cassia da Silveira E Sa. R: Zhong. C. (2017a). Mitochondrial biogenesis in response to
chromium (VI) toxicity in human liver cells. International Journal of Molecular Sciences 18.
http://dx.doi.org/10.3390/ijmsl8091877
Zhong. X: Yi. X: da Silveira E Sa. RC: Zhang. Y: Liu. K: Xiao. F: Zhong. C. (2017b). CoQIO Deficiency
May Indicate Mitochondrial Dysfunction in Cr(VI) Toxicity. International Journal of
Molecular Sciences 18. http: //dx.doi.org/10.3390/iimsl8040816
Zhong. X: Zeng. M: Bian. H: Zhong. C: Xiao. F. (2017c). An evaluation of the protective role of vitamin
C in reactive oxygen species-induced hepatotoxicity due to hexavalent chromium in vitro
and in vivo. J Occup Med Toxicol 12: 12. http: //dx.doi.org/10.1186/sl2995-017-0161-x
Zhu. L: Li. X: Yuan. Y: Dong. C: Yang. M. (2021). APC promoter methylation in gastrointestinal cancer
[Review], Front Oncol 11: 653222. http://dx.doi.org/10.3389/fonc.2021.653222
Zuo. Z: Cai. T: Li. I: Zhang. D: Yu. Y: Huang. C. (2012). Hexavalent chromium Cr(VI) up-regulates
C0X-2 expression through an NFicB/c-Jun/AP-l-dependent pathway. Environ Health
Perspect 120: 547-553. http://dx.doi.org/10.1289/ehp.1104179
This document is a draft for review purposes only and does not constitute Agency policy.
R-57 DRAFT-DO NOT CITE OR QUOTE
------- |