Chromiumvi Erd Toxreview 9-30-10

DRAFT - DO NOT CITE OR QUOTE EPA/635/R-10/004A
www.epa.gov/iris
TOXICOLOGICAL REVIEW
OF
HEXAVALENT CHROMIUM
(CAS No. 18540-29-9)
In Support of Summary Information on the

Integrated Risk Information System (IRIS)
September 2010
(Note: This document is a reassessment of the noncancer and cancer health effects
associated with the oral route of exposure only.)
NOTICE
This document is an External Review draft. This information is distributed solely for the
purpose of pre-dissemination peer review under applicable information quality guidelines. It has
not been formally disseminated by EPA. It does not represent and should not be construed to
represent any Agency determination or policy. It is being circulated for review of its technical
accuracy and science policy implications.
U.S. Environmental Protection Agency

Washington, DC
DISCLAIMER
This document is a preliminary draft for review purposes only. This information is
distributed solely for the purpose of pre-dissemination peer review under applicable information
quality guidelines. 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.
ii DRAFT - DO NOT CITE OR QUOTE

CONTENTS—TOXICOLOGICAL REVIEW OF HEXAVALENT CHROMIUM
(CAS No. 18540-29-9)
LIST OF TABLES ......................................................................................................................... vi

LIST OF FIGURES .........................................................................................................................x
LIST OF ABBREVIATIONS AND ACRONYMS ...................................................................... xi
FOREWORD ............................................................................................................................... xiii
AUTHORS, CONTRIBUTORS, AND REVIEWERS ............................................................... xiv
1. INTRODUCTION ......................................................................................................................1
2. CHEMICAL AND PHYSICAL INFORMATION ....................................................................3
2.1. ENVIRONMENTAL SOURCES AND OCCURRENCE ................................................. 5
2.2. ENVIRONMENTAL CHEMISTRY, SPECIATION, AND BIOACCESSIBILITY ....... 7
2.3. ANALYTICAL METHODS ............................................................................................. 15
3. TOXICOKINETICS .................................................................................................................19
3.1. BIOACCESSIBILITY OF INGESTED HEXAVALENT CHROMIUM ....................... 19
3.1.1. In Vitro and Ex Vivo Studies of Bioaccessibility ............................................... 19
3.1.2. In Vivo Studies of Bioaccessibility ..................................................................... 24
3.2. BIOVAILABILITY, DISPOSITION, AND ELIMINATION OF INGESTED
HEXAVALENT CHROMIUM ...................................................................................... 25
3.2.1. In Vitro and Ex Vivo Studies of Bioavailability, Disposition, and Elimination . 26
3.2.2. In Vivo Studies of Bioavailability, Disposition, and Elimination ....................... 27
3.2.2.1. Laboratory Animal Studies .......................................................................... 27
3.2.2.2. Human Studies ............................................................................................. 41
3.3. BIOCHEMISTRY OF INTRACELLULAR HEXAVALENT CHROMIUM ................ 46
3.4. TOXICOKINETIC CONSIDERATIONS FOR THE MODE OF ACTION OF
INGESTED HEXAVALENT CHROMIUM ................................................................. 49
3.5. BIOLOGICALLY-BASED MODEL OF INGESTED CHROMIUM COMPOUNDS
IN RATS AND HUMANS ............................................................................................. 52
3.6. CONSIDERATION OF CHROMIUM ESSENTIALITY VERSUS TOXICITY........... 66
4. HAZARD IDENTIFICATION .................................................................................................67
4.1. ORAL STUDIES IN HUMANS ...................................................................................... 67
4.1.1. Acute Exposure ................................................................................................... 67
4.1.2. Environmental Exposure ..................................................................................... 68
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS—ORAL ....................................................................................................... 80
4.2.1. Subchronic Oral Exposure ................................................................................... 81
4.2.2. Chronic Oral Exposure ...................................................................................... 101
4.3. REPRODUCTIVE AND DEVELOPMENTAL TOXICITY STUDIES—ORAL ....... 122
4.3.1. Effects on Reproductive Tissues and Mating Behavior .................................... 123
4.3.2. Effects on Reproductive Outcomes ................................................................... 134
4.3.3. Effects of Pregestational Exposure on Reproductive Outcome and Fetal
Development ..................................................................................................... 139
4.3.4. Effects of Gestational and/or Lactational Exposure on Reproductive
Outcome and Fetal Development ...................................................................... 142
4.4. SUMMARY OF HUMAN AND ANIMAL STUDIES BY ROUTES OF
EXPOSURE OTHER THAN ORAL............................................................................ 149
iii DRAFT - DO NOT CITE OR QUOTE

4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE
OF ACTION ................................................................................................................. 150
4.5.1. Genotoxicity Studies ......................................................................................... 150
4.5.1.1. Genotoxicity Assays in Experimental Systems ......................................... 152
4.5.1.2. Genotoxicity Studies in Humans ............................................................... 178
4.5.2. Intracellular Reduction ...................................................................................... 183
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS—ORAL ................................... 191
4.7. EVALUATION OF CARCINOGENICITY .................................................................. 199
4.7.1. Summary of Overall Weight of Evidence ......................................................... 199
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence ........................ 200
4.7.3. Mode-of-Action Information ............................................................................. 202
4.7.3.1. Hypothesized Mode of Action ................................................................... 202
4.7.3.2. Experimental Support for the Hypothesized Mode of Action ................... 203
4.7.3.3. Other Possible Modes of Action ................................................................ 211
4.7.3.4. Conclusions About the Hypothesized Mode of Action ............................. 212
4.7.3.5. Mutagenic Across All Routes of Exposure ............................................... 213
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES .............................................. 214
4.8.1. Possible Childhood Susceptibility ..................................................................... 214
4.8.2. Possible Gender Differences ............................................................................. 214
5. DOSE-RESPONSE ASSESSMENTS ....................................................................................215
5.1. ORAL REFERENCE DOSE (RfD) ............................................................................... 215
5.1.1. Choice of Principal Study and Critical Effect—with Rationale and
Justification ....................................................................................................... 215
5.1.1.1. Subchronic Studies .................................................................................... 216
5.1.1.2. Chronic Studies .......................................................................................... 217
5.1.2. Methods of Analysis—Including Models (PBPK, BMD, etc.) ......................... 219
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs)............ 222
5.1.4. Previous RfD Assessment ................................................................................. 222
5.2. UNCERTAINTIES IN THE ORAL REFERENCE DOSE ........................................... 223
5.3. ORAL CANCER ASSESSMENT ................................................................................. 224
5.3.1. Choice of Study/Data—with Rationale and Justification .................................. 224
5.3.2. Dose-Response Data .......................................................................................... 224
5.3.3. Dose Adjustments and Extrapolation Method(s) .............................................. 228
5.3.4. Oral Slope Factor ............................................................................................... 229
5.3.5. Application of ADAFs ...................................................................................... 229
5.3.6. Uncertainties in Cancer Risk Values ................................................................. 231
5.3.7. Previous Cancer Assessment ............................................................................. 234
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE..................................................................................................................................235
6.1. HUMAN HAZARD POTENTIAL ................................................................................ 235
6.2. DOSE RESPONSE ........................................................................................................ 239
6.2.1. Noncancer—Oral ............................................................................................... 239
6.2.2. Cancer—Oral ..................................................................................................... 240
7. REFERENCES .......................................................................................................................241
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS
AND DISPOSITION........................................................................................ A-1
iv DRAFT - DO NOT CITE OR QUOTE

APPENDIX B. BENCHMARK DOSE CALCULATIONS ......................................................B-1
B.1. DETAILS OF BMD ANALYSIS FOR THE RFD ....................................................... B-1
B.2. DETAILS OF BMD ANALYSIS FOR THE ORAL SLOPE FACTOR .................... B-15
v DRAFT - DO NOT CITE OR QUOTE

LIST OF TABLES
Table 2-1. Industrial uses of hexavalent chromium compounds .................................................... 6

Table 2-2. Physical properties of selected hexavalent chromium compounds ............................... 8
Table 2-3. Standard potentials for oxidation-reduction equilibria among chromium valence
states ........................................................................................................................... 11
Table 2-4. Detection limits for methods commonly used in the analysis of chromium in
water and soil extracts ................................................................................................ 16
Table 2-5. Biomonitoring options for assessing chromium exposure .......................................... 18
Table 3-1. Effect of gastric juice composition on binding and in vitro uptake by rat
intestinal rings of trivalent (Cr51Cl3) or hexavalent (Na2Cr51O4 ) radiolabeled
chromium compounds ................................................................................................ 21
Table 3-2. Estimates of the overall chromium(VI) reducing capacity of human fluids, cells
and tissues .................................................................................................................. 23
Table 3-3. In vitro kinetic parameters of hexavalent chromium uptake in RBCs of rats and
humans ....................................................................................................................... 26
Table 3-4. Distribution and retention of chromium in the rat after a single oral dose.................. 28
Table 3-5. Ratios (intestine:stomach) of chromium concentration in whole blood, plasma,
and RBCs after a single oral dose .............................................................................. 29
Table 3-6. Ratios of chromium concentration in RBCs and plasma in the stomach and
intestine for fasted and nonfasted animals after a single oral dose ............................ 29
Table 3-7. Terminal tissue chromium levels in rats ingesting potassium chromate
(K2CrO4) in drinking water for 1 year ....................................................................... 30
Table 3-8. Time course of chromium tissue concentrations in male Sprague-Dawley rats
ingesting 12.9 mg Cr/L of trivalent chromic chloride-hexahydrate or
hexavalent sodium dichromate in drinking water for 40 days ................................... 34
Table 3-9. Chromium in tissues (μg/g wet tissue or μg/mL blood) of mice and rats after
ingesting K2CrO7 in drinking water (8 mg hexavalent chromium/kg-day) for 4
or 8 weeks .................................................................................................................. 35
Table 3-10. Tissue concentrations of chromium in male F344/N rats and female B6C3F1
mice in the 2-year NTP drinking water study of sodium dichromate dihydrate........ 40
Table 3-11. Kinetic parameters of hexavalent chromium reduction in human liver
microsomes from five individuals ............................................................................. 48
Table 3-12. Chemical-specific parameters in the rat and human chromium models.................... 60
Table 3-13. Parameters for mouse developed by Campbell et al. (2009) ..................................... 63
Table 4-1. Data pertaining to hexavalent chromium concentrations in drinking water in
five villages along a path of groundwater contamination from an alloy plant in
western JinZhou, China from 1965 to 1979............................................................... 71
vi DRAFT - DO NOT CITE OR QUOTE

Table 4-2. Results pertaining to cancer mortality rates in five villages along path of
groundwater contamination from alloy plant and other comparison areas,
western JinZhou, China from 1970 to 1978, based on analyses by Beaumont et
al. (2008) and Kerger et al. (2009) ............................................................................. 73
Table 4-3. Risk ratios comparing cancer mortality rates in five villages along a path of
groundwater contamination from an alloy plant and other comparison areas in
western JinZhou, China from 1970 to 1978............................................................... 75
Table 4-4. Hematological effects in male and female F344/N rats exposed to sodium
dichromate dihydrate in drinking water for up to 3 months ...................................... 83
Table 4-5. Clinical chemistry effects in male and female F344/N rats exposed to sodium
dichromate dihydrate in drinking water for 3 months ............................................... 85
Table 4-6. Selected organ weights in male and female F344/N rats exposed to sodium
Table 4-7. Incidence of nonneoplastic lesions observed in male and female F344/N rats
exposed to sodium dichromate dihydrate in drinking water for 3 months ................ 88
Table 4-8. Selected organ weights in male and female B6C3F1 mice exposed to sodium
Table 4-9. Incidence of nonneoplastic lesions observed in male and female B6C3F1 mice
Table 4-10. Hematological effects in male B6C3F1, BALB/c, and am3-C57BL/6 mice
Table 4-11. Incidence of nonneoplastic lesions observed in male B6C3F1, BALB/c, and
am3-C57BL/6 mice exposed to sodium dichromate dihydrate in drinking water
for 3 months ............................................................................................................... 96
Table 4-12. Hematological effects in male F344/N rats exposed to sodium dichromate
dihydrate in drinking water for up to 12 months ..................................................... 102
Table 4-13. Serum ALT activity in male F344/N rats exposed to sodium dichromate
Table 4-14. Incidence of nonneoplastic lesions observed in male and female F344/N rats
exposed to sodium dichromate dihydrate in drinking water for 2 years .................. 106
Table 4-15. Incidence of neoplastic lesions observed in the oral cavity of male and female
F344/N rats exposed to sodium dichromate dihydrate in drinking water for 2
years ......................................................................................................................... 109
Table 4-16. Neoplastic lesions in other tissues (e.g., nonoral cavity) in F344/N rats
Table 4-17. Hematological effects in female B6C3F1 mice exposed to sodium dichromate
Table 4-18. Incidence of nonneoplastic lesions observed in male and female B6C3F1 mice
vii DRAFT - DO NOT CITE OR QUOTE

Table 4-19. Incidence of neoplastic lesions observed in the small intestine of male and
female B6C3F1 mice exposed to sodium dichromate dihydrate in drinking
water for 2 years....................................................................................................... 117
Table 4-20. Evidence of mutagenicity of hexavalent chromium compounds in
experimental systems ............................................................................................... 151
Table 4-21. In vitro genotoxicity studies of hexavalent chromium in nonmammalian cells ...... 153
Table 4-22. In vitro genotoxicity studies of hexavalent chromium in mammalian cells ............ 159
Table 4-23. In vivo genotoxicity studies of hexavalent chromium in rats and mice .................. 167
Table 4-24. In vivo genotoxicity studies of hexavalent chromium in D. melanogaster ............. 174
Table 4-25. In vivo genotoxicity studies in humans exposed to hexavalent chromium ............. 179
Table 4-26. Observed effects and corresponding NOAELs and LOAELs for subchronic,
chronic, and reproductive toxicity studies following oral exposure to
hexavalent chromium ............................................................................................... 193
Table 5-1. Incidence data for lesions in female F344/N rats and male and female
B6C3F1 mice exposed to sodium dichromate dihydrate in drinking water for 2
years ......................................................................................................................... 219
Table 5-2. Summary of BMD10 and BMDL10 from the best fitting models for lesions of the
liver, duodenum, mesenteric lymph nodes, and pancreas in female rats and
male and female mice after exposure to sodium dichromate dihydrate in
drinking water for 2 years (NTP, 2008) ................................................................... 221
Table 5-3. Incidences of squamous cell papillomas or carcinomas in the oral cavity of
male F344/N rats exposed to sodium dichromate dihydrate in drinking water
for 2 years ................................................................................................................ 226
Table 5-4. Incidences of squamous cell papillomas or carcinomas in the oral cavity of
female F344/N rats exposed to sodium dichromate dihydrate in drinking water
for 2 years ................................................................................................................ 226
Table 5-5. Incidences of adenomas and carcinomas combined in the small intestine of
male B6C3F1 mice exposed to sodium dichromate dihydrate in drinking water
for 2 years ................................................................................................................ 227
Table 5-6. Incidences of adenomas and carcinomas combined in the small intestine of
water for 2 years....................................................................................................... 228
Table 5-7. Application of ADAFs for a 70-year exposure to 0.0001 mg hexavalent
chromium/kg-day from ages 0 to 70 ........................................................................ 230
Table 5-8. Summary of uncertainties in the cancer risk assessment for hexavalent
chromium ................................................................................................................. 232
Table B-1. Incidence data for nonneoplastic lesions from all treatment groups of female
F344/N rats and male and female B6C3F1 mice exposed to sodium dichromate
dihydrate in drinking water for 2 years (NTP, 2008)............................................... B-1
viii DRAFT - DO NOT CITE OR QUOTE

Table B-2. BMD10 and BMDL10 values and goodness-of-fit statistics from models fit to
incidence data for chronic inflammation of the liver in female rats exposed to
sodium dichromium dihydrate in drinking water for 2 years .................................. B-2
incidence data for diffuse epithelial hyperplasia in the duodenum in male mice
exposed to sodium dichromium dihydrate in drinking water for 2 years ................ B-4
incidence data for histiocytic cellular infiltration in mesenteric lymph nodes of
male mice exposed to sodium dichromium dihydrate in drinking water for
2 years ...................................................................................................................... B-6
incidence data for diffuse epithelial hyperplasia in the duodenum of female
mice exposed to sodium dichromium dihydrate in drinking water for 2 years ....... B-8
incidence data for histiocytic cellular infiltration in mesenteric lymph nodes of
female mice exposed to sodium dichromium dihydrate in drinking water for 2
years ....................................................................................................................... B-10
incidence data for histiocytic cellular infiltration in the liver of female rats
exposed to sodium dichromium dihydrate in drinking water for 2 years .............. B-11
incidence data for pancreas: acinus, cytoplasmic alteration in female mice
exposed to sodium dichromium dihydrate in drinking water for 2 years .............. B-13
ix DRAFT - DO NOT CITE OR QUOTE

LIST OF FIGURES
Figure 2-1. Definitions of bioaccessibility and biovailability used in this document .................... 4
Figure 2-2. Schematic of possible reaction processes (grey ovals) that determine
disposition and speciation of chromium contamination in the environment ............. 12
Figure 2-3. Stability diagram showing aqueous speciation of chromium at various
equilibrium potential (Eh, volts) and pH ................................................................... 13
Figure 3-1. Estimates of Cr(VI) sequestration or reduction by organs, cell populations and
fluids in the human body relevant to portal of entry uptake or effects on remote
distribution kinetics.................................................................................................... 22
Figure 3-2. Chromium excretion in urine and feces of Sprague-Dawley rats .............................. 32
Figure 3-3. Chromium concentrations in male and female F344 rats following chronic
drinking water consumption of Cr(VI) ...................................................................... 37
Figure 3-4. Chromium absorption and elimination in human volunteers after ingestion of a
single bolus dose in drinking water ........................................................................... 43
Figure 3-5. Biovailability and elimination half-life for chromium ingested by human
volunteers as a single bolus dose in drinking water ................................................... 43
Figure 3-6. Schematic of ingested Cr(VI) to internal dose in GI tissue and blood....................... 51
Figure 3-7. Observed and simulated plasma chromium concentrations predicted by the
PBPK model for a human subject ingesting Cr(VI) dissolved in orange juice
(CrVI-OJ) in the study of Kerger et al. (1996) .......................................................... 57
Figure 3-8. Schematic of structure for PBPK model of chromium in the rat and human ............ 59
Figure 3-9. Extended PBPK model structure to predict chromium distribution in the oral
cavity and GI tract tissue for rats and mice................................................................ 65
Figure 4-1. Ternary DNA adduct formation by chromium......................................................... 185
Figure B-1. Predicted and observed incidence of chronic inflammation of the liver in
female rats exposed to sodium dichromium dihydrate in drinking water for 2
years ......................................................................................................................... B-3
Figure B-2. Predicted and observed incidence of diffuse epithelial hyperplasia in the
duodenum of male mice exposed to sodium dichromium dihydrate in drinking
water for 2 years....................................................................................................... B-5
Figure B-3. Predicted and observed incidence of diffuse epithelial hyperplasia in the
duodenum of female mice exposed to sodium dichromium dihydrate in
drinking water for 2 years ........................................................................................ B-9
Figure B-4. Predicted and observed incidence of histiocytic cellular infiltration in the
livers of female mice exposed to sodium dichromium dihydrate in drinking
water for 2 years..................................................................................................... B-12
Figure B-5. Predicted and observed incidence of pancreas: acinus, cytoplasmic alteration
in female mice exposed to sodium dichromium dihydrate in drinking water for
2 years .................................................................................................................... B-14
x DRAFT - DO NOT CITE OR QUOTE

LIST OF ABBREVIATIONS AND ACRONYMS
3β-Δ5-HSH 3β-Δ5-hydroxysteroid dehydrogenase

AAS atomic absorption spectrometry
AcP acid phosphatase
ADAF age-dependent adjustment factor
AIC Akaike’s Information Criterion
ALT alanine aminotransferase
AP alkaline phosphatase
AST aspartate aminotransferase
ATSDR Agency for Toxic Substances and Disease Registry
BER base excision repair
BMD benchmark dose
BMDL 95% lower confidence limit of benchmark dose
BMDS benchmark dose software
BMR benchmark response
CalEPA California Environmental Protection Agency
CASRN Chemical Abstracts Service Registry Number
CCA chromated copper arsenate
CI confidence interval
Cr chromium
CSF cancer slope factor
DNA deoxyribonucleic acid
ED Enumeration Districts
EDTA ethylenediaminetetraacetic acid
EPR electron paramagnetic resonance
ER excision repair
FSH follicle-stimulating hormone
GD gestation day
GFR glomerular filtration rate
GH growth hormone
GI gastrointestinal
GSH glutathione
GST glutathione S-transferase
H&E haematoxylin and eosin
HCl hydrochloric acid
Hct hematocrit
Hgb hemoglobin
HPLC high-performance liquid chromatography
ICP inductively coupled plasma
ICP-AES inductively coupled plasma atomic emission spectrophotometry
ICP-OES inductively coupled plasma optical emission spectrometry
ICP-MS inductively coupled plasma mass spectrometry
i.m. intramuscular
i.p. intraperitoneal
i.t. intratracheal
i.v. intravenous
IRIS Integrated Risk Information System
xi DRAFT - DO NOT CITE OR QUOTE

Km Michaelis constant
LH luteinizing hormone
LOAEL lowest-observed-adverse-effect level
LOD limit of detection
MCH mean cell hemoglobin
MCHC mean cell hemoglobin concentration
MCV mean cell volume
MMAD mass mean aerodynamic diameter
MMR mismatch repair
NER nucleotide excision repair
NJDEP New Jersey Department of Environmental Protection
NJDOH New Jersey Department of Health
NOAEL no-observed-adverse-effect level
NRC National Research Council
NTP National Toxicology Program
OPP Office of Pesticide Programs
PAM pulmonary alveolar macrophages
PBPK physiologically based pharmacokinetic
PEG polyethylene glycol
PND postnatal day
POD point of departure
RBC red blood cell
RfC reference concentration
RfD reference dose
RNA ribonucleic acid
SDH sorbitol dehydrogenase
SMART somatic mutation and recombination test
TEM transmission electron microscopy
UF uncertainty factor
U.S. EPA U.S. Environmental Protection Agency
Vmax maximum velocity of enzyme reaction
WBC white blood cell
xii DRAFT - DO NOT CITE OR QUOTE

FOREWORD
The purpose of this Toxicological Review is to provide the scientific support and
rationale for the hazard identification and dose-response assessments in IRIS pertaining to
chronic exposure to hexavalent chromium via ingestion. It is not intended to be a comprehensive
treatise on the chemical or toxicological nature of hexavalent chromium. This document is a
reassessment of the noncancer and cancer health effects associated with the oral route of
exposure. A reassessment of the noncancer and cancer health effects of hexavalent chromium
associated with the inhalation route of exposure will be conducted at a later date.
Section 5, Dose-Response Assessments, is based largely on the work of four independent
groups that have recently evaluated the toxicity of hexavalent chromium via ingestion: (1) U.S.
EPA’s Office of Pesticide Programs (OPP), (2) the New Jersey Department of Environmental
Protection (NJDEP), (3) the California Environmental Protection Agency (CalEPA), and (4) the
Agency for Toxic Substances and Disease Registry (ATSDR). Section 5.1 relies, in part, on
work conducted by ATSDR and CalEPA, and the reference dose (RfD) generally relies on
ATSDR’s analysis for chronic oral exposure to hexavalent chromium. Section 5.3 was
developed, in part, based on work conducted by CalEPA and NJDEP, and the oral cancer slope
factor (CSF) generally relies on NJDEP’s analysis for cancer potency.
The intent of Section 6, Major Conclusions in the Characterization of Hazard and Dose
Response, is to present the major conclusions reached in the derivation of the reference dose,
reference concentration and cancer assessment, where applicable, and to characterize the overall
confidence in the quantitative and qualitative aspects of hazard and dose response by addressing
the quality of data and related uncertainties. The discussion is intended to convey the limitations
of the assessment and to aid and guide the risk assessor in the ensuing steps of the risk
assessment process.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA’s IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
[email protected] (e-mail address).
xiii DRAFT - DO NOT CITE OR QUOTE

AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER/AUTHOR
Ted Berner, M.S.

U.S. EPA, ORD/NCEA
1200 Pennsylvania Ave., NW
Washington, DC 20460
AUTHORS
Catherine Gibbons, Ph.D

U.S. EPA, ORD/NCEA
Glinda Cooper, Ph.D

U.S. EPA, ORD/NCEA
CONTRIBUTING AUTHORS
Jenny Li, Ph.D

U.S. EPA, ORD/NCEA
Teneille Walker, Ph.D

U.S. EPA, ORD/NCEA
CONTRACTOR SUPPORT
Julie Klotzbach, Ph.D.

Michael H. Lumpkin, Ph.D.
Daniel J. Plewak, B.S.
SRC, Inc.
North Syracuse, NY
REVIEWERS
This document has been provided for review to EPA scientists and interagency reviewers
from other federal agencies and White House offices.
xiv DRAFT - DO NOT CITE OR QUOTE

1. INTRODUCTION
This document presents background information and justification for the Integrated Risk
Information System (IRIS) Summary of the hazard identification and dose-response assessments
of ingested hexavalent chromium. IRIS Summaries may include oral reference dose (RfD) and
inhalation reference concentration (RfC) values for chronic and other exposure durations, and a
carcinogenicity assessment. This Toxicological Review of Hexavalent Chromium provides
documentation for oral toxicity values only (i.e., RfD and oral cancer slope factor). A
reassessment of the noncancer and cancer health effects of hexavalent chromium associated with
the inhalation route of exposure will be conducted at a later date.
The RfD and RfC, if derived, provide quantitative information for use in risk assessments
for health effects known or assumed to be produced through a nonlinear (presumed threshold)
mode of action. 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. The inhalation RfC (expressed in units of mg/m3) is
analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The
inhalation RfC considers toxic effects for both the respiratory system (portal-of-entry) and for
effects peripheral to the respiratory system (extrarespiratory or systemic effects). Reference
values are generally derived for chronic exposures (up to a lifetime), but may also be derived for
acute (≤24 hours), short-term (>24 hours up to 30 days), and subchronic (>30 days up to 10% of
lifetime) exposure durations, all of which are derived based on an assumption of continuous
exposure throughout the duration specified. Unless indicated otherwise, the RfD and RfC are
derived for chronic exposure durations.
The carcinogenicity assessment provides information on the carcinogenic hazard
potential of the substance in question and quantitative estimates of risk from oral and inhalation
exposure may be derived. The information includes a weight-of-evidence judgment of the
likelihood that the agent is a human carcinogen and the conditions under which the carcinogenic
effects may be expressed. Quantitative risk estimates may be derived from the application of a
low-dose extrapolation procedure. If derived, the oral slope factor is a plausible upper bound on
the estimate of risk per mg/kg-day of oral exposure. Similarly, an inhalation unit risk is a
plausible upper bound on the estimate of risk per μg/m3 air breathed.
Development of these hazard identification and dose-response assessments for hexavalent
chromium has followed the general guidelines for risk assessment as set forth by the National
Research Council (NRC, 1983). U.S. Environmental Protection Agency (U.S. EPA) Guidelines
and Risk Assessment Forum Technical Panel Reports that may have been used in the
development of this assessment include the following: Guidelines for the Health Risk
1 DRAFT - DO NOT CITE OR QUOTE

Assessment of Chemical Mixtures (U.S. EPA, 1986a), Guidelines for Mutagenicity Risk
Assessment (U.S. EPA, 1986b), Recommendations for and Documentation of Biological Values
for Use in Risk Assessment (U.S. EPA, 1988), Guidelines for Developmental Toxicity Risk
Assessment (U.S. EPA, 1991), Interim Policy for Particle Size and Limit Concentration Issues in
Inhalation Toxicity (U.S. EPA, 1994a), Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994b), Use of the
Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995), Guidelines for
Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for Neurotoxicity Risk
Assessment (U.S. EPA, 1998), Science Policy Council Handbook: Risk Characterization (U.S.
EPA, 2000a), Benchmark Dose Technical Guidance Document (U.S. EPA, 2000b),
Supplementary Guidance for Conducting Health Risk Assessment of Chemical Mixtures (U.S.
EPA, 2000c), A Review of the Reference Dose and Reference Concentration Processes (U.S.
EPA, 2002), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), Supplemental
Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA,
2005b), Science Policy Council Handbook: Peer Review (U.S. EPA, 2006a), and A Framework
for Assessing Health Risks of Environmental Exposures to Children (U.S. EPA, 2006b).
The literature search strategy employed for this compound was based on the Chemical
Abstracts Service Registry Number (CASRN) and at least one common name. Any pertinent
scientific information submitted by the public to the IRIS Submission Desk was also considered
in the development of this document. The relevant literature was reviewed through September
2010.

2. CHEMICAL AND PHYSICAL INFORMATION
This toxicological review restricts its focus to oral exposure to hexavalent chromium
compounds. Hexavalent chromium compounds are a group of substances that contain chromium
in the hexavalent or +6 oxidation state. Hexavalent chromium compounds discussed in this
document include the following: chromium(VI) oxide, chromic acid, and selected salts of the
chromate (CrO42-) and dichromate (Cr2O72-) anions.
This section discusses sources of chromium in the environment and the physicochemical
properties of chromium compounds that determine their environmental bioaccessibility. It is
recognized that the definition of bioaccessibility and bioavailability has differed across
disciplines and can cause confusion especially in a regulatory context (Semple et al., 2004). For
the purposes of this toxicological review, bioaccessibility is defined as the ability of chromium to
be released (e.g., environmental solubilization and reduction or extracellular digestion and
reduction) from the environmental matrix to which it is bound, i.e., the fraction of the dose
ingested that becomes freely available for absorption via crossing a cellular membrane. As
shown in Figure 2-1, this would encompass processes A through C. Key determinants of
environmental bioaccessibility include the following: 1) type of contaminant, 2) contamination
level, 3) type of soil, 4) pH of soil, 5) aging of soil, and 6) metal speciation. Processes that
control the environmental chemistry of chromium include redox transformation,
precipitation/dissolution, and adsorption/desorption reactions.

Adapted from: Semple et al. (2004) and NRC (2002).
Figure 2-1. Definitions of bioaccessibility and biovailability used in this

document.
Processes that determine exposure to a contaminant include release of soil-bound contaminant (A), and
subsequent transport (B), or transport of bound contaminant (C), uptake across a physiological membrane
(D), and incorporation into a living system (E). For the purposes of this toxicological review,
environmental and extracellular processes A through C are defined as bioaccessibility. Environmental
bioaccessibility is discussed in this section. Extracellular bioaccessibility processes that also may occur
within an organism, for example in the GI lumen, are discussed in Section 3 on toxicokinetics. Processes D
and E, represented by the dashed lines, are defined as bioavailability and are also discussed in Section 3.
Processes A through C can also occur internal to an organism, and the same determinants
of bioaccessibility function in that internal environment, e.g., pH in different sections of the
gastrointestinal (GI) lumen. Both environmental and internal bioaccessibility processes are
extracellular. For the purposes of this Toxicological Review, however, this latter internal,
extracellular bioaccessibility is discussed in Section 3 on toxicokinetics.
Bioavailability (processes D and E in Figure 2-1) is defined in this document as the
potential for chromium to cross cellular boundaries, i.e., the degree to which it becomes available
to the target tissue after administration. Key mechanisms of bioavailability that determine
internal tissue dose include the following: uptake through cell membranes, intracellular

distribution, and binding to cellular macromolecules. These bioavailability processes are also
discussed in Section 3. The toxicodynamics of responses to that internal tissue dose are
discussed in Section 4.
2.1. ENVIRONMENTAL SOURCES AND OCCURRENCE

Elemental chromium was first discovered and characterized by the French chemist
Nicolas-Louis Vauquelin in Siberian red lead ore (crocoite) in 1797 (Katz and Salem, 1993;
Costa and Klein, 2008). It is a naturally occurring element present in the earth’s crust.
Chromium ranks 21st among elements in crustal abundance (Krausopf, 1979), and is found in
virtually all phases including air, water, soil and biota (Losi et al., 1994). The average chromium
concentration in the continental crust has been commonly observed to range from 80 to 200
mg/kg (NAS, 1974).
The chromium content of soil is largely dependent on the parent materials, with an
average reported as 40 mg/kg (Bertine and Goldberg, 1971). The average concentration of
chromium in freshwater is 1.0 µg/L (range: 0.1−6.0 µg/L), while the average concentration of
chromium in seawater is 0.3 µg/L (range: 0.2−50 ug/L) (Bowen, 1979 as cited in Losi et al.,
1994). Drainage water from contaminated areas may have higher concentrations. The aqueous
concentrations of chromium and its mobility in different geologic environments are dependent on
its oxidation state (Rai et al., 1989).
Hexavalent chromium compounds are well-known as laboratory reagents and as
manufacturing intermediates. Selected industrial uses of hexavalent chromium compounds are
provided in Table 2-1. The major industries using chromium are the metallurgical, chemical,
and refractory brick industries (Langard, 1980 as cited in Losi et al., 1994). Major uses of
hexavalent chromium compounds include metal plating, manufacture of pigments and dyes,
corrosion inhibitors, chemical synthesis, refractory production, leather tanning, and wood
preservation (Blade et al., 2007; Shanker et al., 2005; Page and Loar, 2004). Sodium chromate,
sodium dichromate, and chromium(VI) oxide are obtained directly from chromite ore through an
oxidative alkaline roasting process (Anger et al., 2005; Page and Loar, 2004). Sodium chromate
and sodium dichromate are the starting materials for the production of most other chromium
compounds (Anger et al., 2005; Page and Loar, 2004). Hexavalent chromium compounds are
classified as oxidizing agents (Anger et al., 2005; Cotton et al., 1999). Chromium(VI) oxide and
ammonium dichromate can react explosively when brought into contact with organic materials
(Lewis, 2007; O’Neil, 2006).

Table 2-1. Industrial uses of hexavalent chromium compounds
Name Formula Uses

Chromium oxide Cr2O3 Various metallurgical uses including production of
ferroalloys used to make stainless steel and nonferrous
alloys or superalloys, such as those used in jet engines
Barium oxide BaCrO4 Pyrotechnics, high-temperature batteries
Cadmium chromate CdCrO4 Catalysts, pigments
Cadmium dichromate CdCr2O7:H2O Metal finishing
Calcium chromate CaCrO4 Metal primers, corrosion inhibitors, high-temperature
batteries
Copper dichromate CuCr2O7:2H2O Wood preservatives, catalysts
Magnesium chromate MgCrO4:5H2O Corrosion inhibitor in photoengraving, ceramics
Mercuric chromate HgCrO4 Antifouling formulation
Pyridine dichromate (C5H5NH)2:Cr2O7 Photosensitizer in photoengraving, cermics
Strontium chromate SrCrO4 Corrosion-inhibiting pigment, plating additive
Source: Hartford (YEAR) as cited in Nriagu (1988).
Alloys with iron, nickel or cobalt are prepared from metallurgical grade ore (60% ≥
chromic oxide) for use in the production of a wide variety of steels including stainless steel,
austenite steel, and high-speed and high-temperature steels, and in other nonferrous alloys. The
chemical industry generally uses a lower grade chromite ore (≈ 45% chromic oxide) to
synthesize sodium chromate and dichromate from which most other chromium products are
prepared, including products used in pigment manufacture, plating/metal finishing, corrosion
inhibition, organic synthesis, leather tanning, and wood preservation. Other important
anthropogenic sources of chromium in the environment include fuel combustion, cement
production, and sewage sludge incineration/deposition (U.S. EPA, 1984).
Chromated copper arsenate (CCA) wood preservatives to prevent fungal decay and
infestations by wood-boring insects contain hexavalent chromium. The preservatives are used in
the industrial vacuum-impregnation of timber and supplied as pastes or water-based concentrates
that are diluted to between 1 and 10% w/w total salts (Cocker et al., 2006). Wood preservatives
such as CCA are regulated under the Control of Pesticide Regulations 1986, with only approved
compounds placed on the market. Cocker et al. (2006) showed that workers exposed to CCA
wood preservatives have urinary chromium levels significantly higher than nonoccupationally
exposed populations, but below occupational biological exposure indices for urinary chromium
based on inhalation exposures. Balasoiu et al. (2000) evaluated the influence of soil composition
and physicochemical characteristics on the retention and partitioning of chromium in nine CCA
artificially contaminated soils using a statistical mixture design to arrive at the different soil
compositions. Sequential extraction and modified solvent extraction were used to assess the
partitioning. Average metal retention in mineral soils was low (23%), but increased dramatically

in highly organic soils (78%). Levels of chromium in soluble or exchangeable form were very
low in highly organic soils. Conversely, 18% of chromium in mineral soils was found in
exchangeable form. Thus, chromium in moderately and highly organic contaminated soils was
present in less mobile and less bioavailable forms; in mineral soils the labile fraction was higher.
Food is also a major source of exposure to chromium. Daily oral intake rates for
populations in the United Kingdom (U.K.) were estimated by Rowbotham et al. (2000) for
infants (1 year), children (11 years) and adults to be 33–45, 123–171, and 246–343
μg/person/day, respectively. Plessi and Monzani (1990) showed that chromium content in whole
cereals differed substantially and was mostly concentrated in pericarps. Variations occurred not
only among different types of cereals, but also within cereals of the same type depending on the
area of origin. Kumulainen (1991) showed that food processing may increase food chromium
content, depending on the process. Processes such as meat grinding and homogenization using
stainless-steel equipment increased the chromium content of foods. Acidic fruit juices in contact
with steel cans are high in chromium as well.
2.2. ENVIRONMENTAL CHEMISTRY, SPECIATION, AND BIOACCESSIBILITY

Chromium is a first series transition element for group VIA on the periodic table. In its
elemental form, it is a hard, white, lustrous, and brittle metal with a high melting point (2000°C)
(Costa and Klein, 2006). Chemical structures and selected physical and chemical properties of
selected hexavalent chromium compounds are presented in Table 2-2, and the standard potential
for the oxidation-reduction equilibria among the valence states are summarized in Table 2-3.

Table 2-2. Physical properties of selected hexavalent chromium compounds
Name Chromium(VI) oxidea Chromic acida,b Sodium chromate Sodium dichromate Sodium dichromate, dihydrate
CASRN 1333-82-0 7738-94-5 (H2CrO4); 7775-11-3 10588-01-9 7789-12-0
13530-68-2 (H2Cr2O7)
Synonyms Chromium oxide; Chromic(VI) acid; chromium Sodium chromate(VI); Sodium dichromate(VI); Dichromic acid, disodium salt,
(ChemID hexavalent chromium hydroxide oxide; dichromic acid chromium disodium sodium bichromate; dihydrate
Plus, oxide; chromic (H2Cr2O7) oxide; disodium dichromic acid, disodium
2008) trioxide; chromic chromate; rachromate; salt; bichromate of soda
anhydride; chromic chromic acid, disodium
acid anhydride (Anger salt; chromate of soda
et al., 2005)
Structure 2Na+ O OO 2-
2Na+ O OO 2-
(ChemID O O
Cr Cr Cr Cr 2H2O
Plus, O O
O O O O
2008)
Molecular 99.994 (Lide, 2008) 118.010 (H2CrO4) (Lide, 2008); 161.974 (Lide, 2008) 261.968 (Lide, 2008) 297.999 (Lide, 2008)
weight 218.001 (H2Cr2O7) (ChemID Plus)
Molecular CrO3 (ChemID Plus, H2CrO4; H2Cr2O7 (ChemID Plus, Na2CrO4 (ChemID Na2Cr2O7 (ChemID Plus, Na2Cr2O7•2H2O (ChemID Plus,
formula 2008) 2008) Plus, 2008) 2008) 2008)
Form Dark red, deliquescent Exists only as an aqueous solution Yellow, orthorhombic Light brown to orange- Orange-red, monoclinic,
bipyramidal prismatic (Lide, 2008); yellow to orange-red crystals (Anger et al., red plates (Anger et al., translucent needles (Anger et al.,
crystals, flakes, or (Anger et al., 2005) 2005) 2005) 2005)
granular powder
(O’Neil, 2006)
Stability/ Deliquescent; Strong oxidizing agent (Anger et al., Hygroscopic (Anger et Strongly hygroscopic; Very hygroscopic, deliquesces
reactivity decomposition begins 2005) al., 2005) decomposes above 400°C in air; decomposes above 85°C;
above 198°C (Anger et (Lide, 2008); strong strong oxidizing agent in acid
al., 2005); powerful oxidizing agent (Anger et solution (Lide, 2008; Anger et
oxidizer (O’Neil, 2006) al., 2005) al., 2005)
Melting 197°C (Lide, 2008) Not applicable 794°C (Lide, 2008) 357°C (Lide, 2008) Decomposes prior to melting
point (Lide, 2008)
Density 2.7 g/cm3 (Lide, 2008) Not applicable 2.72 g/cm (Lide, 2008) 2.52 g/cm (Anger et al., 2.35 g/cm3 (Lide, 2008)
3 3
2005)
Water 169 g/100 g H2O at Not applicable 87.6 g/100 g H2O at 187 g/100 g H2O at 25°C 272.9 g/100 g H2O (73.18 wt%)
solubility 25°C (Lide, 2008) 25°C (Lide, 2008) (Lide, 2008) at 20°C (Anger et al., 2005)
Other Soluble in alcohol and Not applicable Slightly soluble in Not available Soluble in acetic acid (Lide,
solubility mineral acids (Lewis, ethanol (Lide, 2008) 2008)
2007)

Name Potassium chromate Potassium dichromate Calcium chromate Ammonium dichromate Zinc chromate Lead chromate
CASRN 7789-00-6 7778-50-9 13765-19-0 7789-09-5 13530-65-9 7758-97-6
Synonyms Potassium Potassium dichromate(VI); Calcium Ammonium bichromate; Zinc chromate(VI) Lead chromate(VI);
(ChemID chromate(VI); bichromate of potash; chromate(VI); ammonium dichromate(VI); hydroxide; buttercup phoenicochroite;
Plus, bipotassium chromate; potassium bichromate; calcium chrome diammonium dichromate; yellow; chromic plumbous
2008) dipotassium chromate; dipotassium bichromate; yellow; calcium chromic acid, diammonium acid, zinc salt; zinc chromate; chromic
chromate of potash; dipotassium dichromate; monochromate; salt chrome yellow; zinc acid, lead salt;
tarapacaite; dipotassium dichromium gelbin; yellow teraoxychromate chrome yellow
chromic acid, heptaoxide; lopezite; ultramarine; chromic (O’Neil, 2006)
dipotassium salt dichromic acid, acid, calcium salt
dipotassium salt
Structure 2K+ O OO 2-
2NH4+ O OO 2-
(ChemID O O
Cr Cr Cr Cr
Plus, O O
O O O O
2008)
Molecular 194.191 (Lide, 2008) 294.185 (Lide, 2008) 156.07 (Lide, 2008) 252.065 (Lide, 2008) 181.403 (Lide, 323.2 (Lide, 2008)
weight 2008)
Molecular K2CrO4 (ChemID Plus, K2Cr2O7 (ChemID Plus, CaCrO4 (ChemID (NH4)2Cr2O7 (ChemID Plus, ZnCrO4 (ChemID PbCrO4 (ChemID
formula 2008) 2008) Plus, 2008) 2008) Plus, 2008) Plus, 2008)
Form Lemon yellow prisms Tabular or prismatic, bright Yellow monoclinic Large, bright, orange-red Yellow prisms Yellow-orange
(Anger et al., 2005) orange-red triclinic crystals or rhombic crystals crystals (Anger et al., 2005) (Lide, 2008) monoclinic crystals
(Anger et al., 2005) (O’Neil, 2006) (Lide, 2008)
Stability/ Nonhygroscopic Nonhygroscopic; Decomposes at Flammable; nonhygroscopic; Not available Not available
reactivity (Anger et al., 2005) decomposes at 500°C 1,000°C (Lide, decomposition begins upon
(Anger et al., 2005; Lide, 2008); oxidizing heating at 180°C (O’Neil,
2008) agent (Lewis, 2007) 2006). Strong oxidizing agent,
may explode in contact with
organic materials (Lewis,
2007)
Melting 974°C (Lide, 2008) 398°C (Lide, 2008) Decomposes prior to Decomposes prior to melting 316°C (Lide, 2008) 844°C (Lide, 2008)
point melting (Lide, 2008) (Lide, 2008)
Density 2.73 g/cm3 (Lide, 2.68 g/cm3 (Lide, 2008) 3.12 g/cm3 (Anger et 2.155 g/cm3 (Lide, 2008) 3.40 g/cm3 (Lide, 6.12 g/cm3 (Lide,
2008) al., 2005) 2008) 2008)

Name Potassium chromate Potassium dichromate Calcium chromate Ammonium dichromate Zinc chromate Lead chromate
Water 65.0 g/100 g H2O at 15.1 g/100 g H2O at 25°C 4.5 g/100 g H2O 35.6 g/100 g H2O at 20°C 3.08 g/100 g H2O 0.000017 g/100 g
solubility 25°C (Lide, 2008) (Lide, 2008) (4.3 wt%) at 0°C (Lide, 2008) (Lide, 2008) H2O at 20°C (Lide,
(Anger et al., 2005) 2008)
Other Insoluble in alcohol Insoluble in alcohol Soluble in dilute Soluble in alcohol (Lewis, Dissolves readily in Insoluble in acetic
solubility (O’Neil, 2006) (Lewis, 2007) acids; practically 2007) acids (Anger et al., acid; soluble in
insoluble in alcohol 2005); insoluble in solutions of fixed
(O’Neil, 2006) acetone (Lide, 2008) alkali hydroxides;
soluble in dilute
nitric acid (O’Neil,
2006)
a
Chromic 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 in reference to 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.
b
Chromic acid exists in solution as both H2CrO4 and H2Cr2O7 (Anger et al., 2005; Page and Loar, 2004; Cotton et al., 1999). H2CrO4 is the main species in basic
solutions (pH >6) while H2Cr2O7 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).

Table 2-3. Standard potentials for oxidation-reduction equilibria among
chromium valence states
Half-cell reaction E0 (V) Change

Cr2O7 + H2O + 2e  2CrO43- + 2H+
2- -
0.55 VI to V
Cr2O72- + 6H+ + 4e-  2CrO2 + 3H2O 0.95 VI to IV
Cr2O72- + 14H+ + 6e-  2Cr3+ + 7H2O 1.38 VI to III
CrO43- + 4H+ + e-  CrO2 + 2H2O 1.34 V to IV
CrO43- + 8H+ + 2e-  Cr3+ + 4H2O 1.72 V to III
CrO2 + 4H+ + e-  Cr3+ + 2H2O 2.10 IV to III
Cr3+ + e-  Cr2+ -0.42 III to II
Cr3+ + 3e-  Cr -0.74 III to 0
Cr2+ + 2e-  Cr -0.90 II to 0
CrO42- + 4H2O + 3e-  [Cr(OH)4]- + 4OH- -0.72 VI to III
CrO42- + 4H2O + 3e-  Cr(OH)3 + 5OH- -0.11 VI to III
[Cr(OH)4]- + 3e-  Cr + 4OH- -1.33 III to 0
Cr(OH)3 + 3e-  Cr + 3OH- -1.33 III to 0
Source: Emsley (1989) as cited in Katz and Salem (1993).
The environmental chemistry of chromium is complex. Figure 2-2 illustrates the possible
fates of chromium in soil/water systems. Chromium is known to undergo various chemical and
biological reactions in natural systems that govern its speciation, and in turn, environmental
behavior. Important reactions include oxidation/reduction, precipation/dissolution, and
adsorption/desorption. Both oxidation of Cr(III) and reduction of Cr(VI) can occur in geologic
and aquatic environments. Hexavalent chromium is a strong oxidizing agent and is readily
reduced in the presence of appropriate electron donors as shown here:
−
HCrO4 + 7 H + + 3e − ⇔ Cr 3+ + 4 H 2 O (2-1)

Source: Adapted from Rai et al. (1989).
Figure 2-2. Schematic of possible reaction processes (grey ovals) that

determine disposition and speciation of chromium contamination in the
environment.
Chromium is a redox active soil contaminant with dramatic alterations in its mobility and
toxicity with changes in oxidation state (Fendorf et al., 2004; Rai et al., 1989). Chromium can
exist in oxidation states ranging from -2 to +6, but only +3 and +6 are typically found within the
range of pH and redox potential common in environmental systems (Shupack, 1991). Figure 2-3
depicts a generalized scheme of equilibrium potentials versus pH for chromium.

Source: Rai et al. (1989).
Figure 2-3. Stability diagram showing aqueous speciation of chromium at

various equilibrium potential (Eh, volts) and pH.
Hexavalent chromium is a strong oxidizer and as a result, exists only in oxygenated

species that are very soluble and pH-dependent according to the following equilibria (Nieboer
and Jusys, 1988 as cited in Losi et al., 1994):
−
H 2 CrO4 ⇔ H + + HCrO4 K a1 = 10 0.6 (2-2)
− 2−
HCrO4 ⇔ H + + CrO4 K a 2 = 10 −5.9 (2-3)
Neuss and Rieman (YEAR, as cited in Katz and Salem, 1993) evaluated the acid
functions and arrived at Ka1 and Ka2 values of 1.8 x 10-1 and 3.2 x 10-7. From equations (2-2) and
(2-3) it can be observed that at very low pH values (pH < 1), H2CrO4 is the predominant species,
while between pH 0 and 5.9, the HCrO4- and Cr2O72- anions prevail (Shupack, 1991). At pH 6 or
above, CrO42- prevails. Thus, H2CrO4 and CrO42- should be most abundant in natural systems.

Further, at concentrations greater than 0.01M (520 mg/L) dimerization of the chromate
ion occurs which yields the dichromate ion (Whitten and Gailey, 1984 as cited in Losi et al.,
1994):
− 2−
2CrO4 + 2 H + ⇔ Cr2 O7 + H 2O K c = 1014.6 (2-4)
[Cr O ]
2 7
2−
= 1014.6
[CrO ] [H ]
(2-5)
2− 2 + 2
4
This reaction is pH sensitive as well, with dichromate favored at lower pH (Losi et al., 1994).
Solving equation (2-4) using values for chromate concentration and pH that would likely be
encountered in contaminated groundwater (5.2 mg/L and 7.0), the ratio of dichromate to
chromate would be 0.04. Thus, hexavalent chromium chemistry in environmental systems is
largely confined to that of the chromate anion.
Speciation of hexavalent (CrVI) and trivalent (CrIII) chromium will generally depend on
a variety of environmental parameters including: pH, concentration, and the ligands available in
the matrices (Katz, 1991). In most natural systems, Cr(VI) will be present as CrO42- and major
trivalent species may include hydroxides and various organic complexes (Losi et al., 1994). The
acid anhydride CrO3 and acid chloride CrO2Cl2, and a wide variety of metal chromates MCrO4
and metal dichromates MCr2O7 are typical hexavalant compounds (Katz and Salem, 1993).
Chromium(VI) may also form other species, including HCr2O7- and CrO42-, but their formation
requires Cr(VI) concentrations >10-2 M, which are not found commonly in natural waters (Baes
and Mesmer, 1986 as cited in Rai et al., 1989).
Natural occurrence of hexavalent chromium is rare as it is readily reduced by organic
matter in the environment (Ashley et al., 2003; Barceloux, 1999; U.S. EPA, 1984). Industrial
releases of hexavalent chromium compounds to surface water and soil can result in the transport
and leaching of these substances into groundwater, provided these substances remain under
oxidizing conditions (Loyuax-Lawniczak et al., 2001; Pellerin and Booker, 2000; James et al.,
1997). Hexavalent chromium compounds released to the environment by anthropogenic sources
may persist in natural waters and soils that contain low amounts of organic matter (Johnson et
al., 2006; Loyaux-Lawniczak et al., 2001; U.S. EPA, 1984). Hexavalent chromium compounds
are considered to be more soluble in water and have greater mobility in soil than other types of
chromium compounds (Loyuax-Lawniczak et al., 2001; James et al., 1997).
Whereas reduction of Cr(VI) is likely to occur in environmental systems, oxidation of
Cr(III) is not frequently observed with the exception of some oxidation in the presence of Mn4+
(Losi et al., 1994). Thus, Cr(III) is considered more stable in most natural systems. Trivalent

chromium is also less mobile than Cr(VI) in most soil/water systems due to its relative
insolubility at relevant pH values (>5).
Consideration of these environmental chemistry factors determining speciation,
solubility, and mobility are of critical importance in assessing potential environmental hazards,
relevance of available risk estimates, and remediation strategies for sites with high levels of
chromium. The behavior of both hexavalent and trivalent chromium and the interconversion
between these forms must be understood when attempting to characterize bioaccessibility and
environmental contamination with chromium. Further, accurate data for describing the dominant
reactions in each process must be available for characterization of a particular site (Rai et al.,
1989).
2.3. ANALYTICAL METHODS

Analysis of chromium in either environmental or biological samples must consider the
thermodynamic stability of each oxidation state as a function of pH and the kind of species, i.e.,
the anions, cations, and polymeric ions that can form for each oxidation state as a function of the
sample such as pH, lability, equilibria, temperature, and the presence of other oxidants and
reductants (Katz, 1991; Shupack, 1991). Metal speciation is important because the
biogeochemical behavior, environmental mobility, bioaccessibility, bioavailability, and
subsequent toxicity all depend on it (Fytianos, 2001). Thus, the level of detection and the ability
to detect speciation versus total chrome in various environmental and biological compartments is
a critical consideration when evaluating the relevance and reliability of any given environmental
or toxicological data. While the topic is discussed briefly in this section with respect to
environmental bioaccessibility, it is equally relevant to discussions on bioavailability and to
considerations of experimental design in the toxicological studies in subsequent sections.
Attention should also be paid to the limit of detection (LOD) for the analytical method used.
Due to its redox sensitivity described above, any attempt at analysis of samples of soil
and water for chromium content should address speciation as a goal, especially due to the
differences in toxicity between Cr(VI) and Cr(III) compounds. Critical considerations include
sample treatment and storage, extraction, and preparation in the case of soils, and the actual
analytical method used. Analysis of chromium in soils and sediments is a special challenge. It is
important that samples be field-moist, sieved (4 mm), well-mixed and stored at 4°C (Bartlett,
1991 as cited in Losi et al., 1994). Quality assurance and quality control exercises are also
recommended (Katz, 1991).
Approximate limits of detection for several analytical methods commonly used for
chromium are provided in Table 2-4. While quantification of chromium in certain biological
compartments may require extremely sensitive detection methodology (see below), analysis of
chromium in environmental samples is typically done by either atomic absorption spectrometry
(AAS) and inductively coupled plasma (ICP) spectrometry for total chromium, and colorimetry

(diphenylcarbazide method) and high-performance liquid chromatography (HPLC) for
quantifying Cr(VI). The difference between the total chromium and Cr(VI) measurements is
assumed to be Cr(III). All of these methods are subject to interferences that must be evaluated
through the use of appropriate matrix spikes. Comparative discussions across studies must take
into account differences in analytical LOD and speciation. Further, translation of dose-response
estimates based on speciated chromium concentrations must be rectified with the sampling
methods used for exposure assessment as part of risk characterization and management efforts.
Vitale et al. (1997) have reviewed the challenges of chromium speciation as it relates to several
of the most current recommended EPA methods, including EPA Method 218.6 for water and
SW-846 methods for soils, sediments, and wastes.
Table 2-4. Detection limits for methods commonly used in the analysis of
chromium in water and soil extracts
Approximate detection limit

Method Species detected (µg/L)
Flame atomic absorption spectrometry Total chromium 500a
Graphite furnace atomic absorption spectrometry Total chromium 1a
Inductively coupled plasma mass spectrometry Total chromium 6–10a
Colorimetric (diphenylcarbazide) Cr(VI) 50b
High-performance liquid chromatography (single- Cr(VI) 92c
column ion-chromatography
a
Bartlett (YEAR), see Gochfeld (1991)
b
Mehra and Frankenberger (1989)
c
MISSING
Source: Losi et al. (1994).
Chromium in biological samples traditionally has been determined by graphite furnace

atomic absorption spectrometry, neutron activation analysis, or inductively coupled plasma
optical emission spectrometry (ICP-OES). More recently, inductively coupled plasma mass
spectrometry (ICP-MS) has become popular as a method due to its low detection limits and
multi-element analysis capabilities. Levine et al. (2007) developed and validated an ICP-MS
method after a rapid, open-vessel microwave digestion, to determine total chromium in the
tissues of the F344 rats used in the National Toxicology Program (NTP) studies of orally
ingested sodium dichromate dihydrate and chromium picolinate monohydrate described in (see
Sections 4 and 5). Performance of the method was evaluated using kidney tissue across a range
of 0.50 to 5.00 µg Cr/g tissue. Feces samples were analyzed by ICP-OES because of the
relatively high levels of chromium in this matrix (Levine et al., 2007). Matrix recovery for the
ICP-MS method ranged from 76.6 to 103%; mean recovery was 97.4%. Percent relative
standard deviations for both intra- and inter-assay preparations ranged from 0.88 to 13%. The
LOD, calculated as three times the SD of these matrix samples, was determined to be 0.01 ng/L,

equivalent to 0.02 µg Cr/g kidney tissue. Cross-validation with B6C3F1 mouse kidney tissue
was also demonstrated.
Because a majority of absorbed chromium is excreted in the urine, occupational
biomonitoring of urinary concentrations of water soluble chromium compounds (as total
chromium) has been successfully used to assess whether significant inhalation exposure to
chromium has occurred (ACGIH, 2004). The New Jersey Department of Health (NJDOH) has
also suggested that urinary chromium concentrations might serve as a useful indicator of
environmental exposures, but an expert panel cited concerns for differentiating environmental
contributions from dietary sources (Anderson et al., 1993). Food known to contain moderate
levels of Cr(III) include breads, cereals, spices, fresh vegetables, meat, fish, vitamin
supplements, Brewer’s yeast, and beer (Gargas et al., 1994), and although comparison of urinary
chromium concentration to the administered dose (exposure) provides an estimate of the amount
of chromium systemically absorbed, it does not provide an indication of the valence state in
which it was absorbed. Therefore, elevated urinary chromium levels should be carefully
considered and may be misleading if interpreted as a biomarker for Cr(VI) exposure or toxicity
(Kerger et al., 1996a). Typical background urinary chromium concentrations vary from ≈0.24 to
1.8 µg Cr/L in healthy individuals (IARC, 1990).
Atomic absorption spectrometry with a graphite furnace is currently recommended for
urinary biomonitoring programs (Paustenbach et al., 1997). Guidelines developed by Veillon et
al. (1982) show that a LOD of 0.05 µg/L is achievable for undiluted samples using this method,
whereas commercial laboratories typically report a LOD of 0.2 µg/L. As with other
biomonitoring evaluations, controlling for confounding variables is essential to be able to
understand their correlation with physiological effects. Confounding variables specifically
associated with elevated urinary chromium levels include diet, regular exercise, smoking habits,
beer consumption, past employment in chromium-related occupations, and health status
(Paustenbach et al., 1997). Pregnancy and stress have been shown to enhance losses of
chromium (Anderson et al., 1989), and diabetics typically show chromium levels twofold higher
than nondiabetics (Bukowski et al., 1991).
In addition to urinary biomonitoring, other biological matrices have been used to measure
recent chromium exposures, including whole blood, plasma, and hair (IARC, 1990).
Concentration of total chromium in RBCs is considered to be a more specific biomarker for
characterizing recent exposure to Cr(VI) (Korallus, 1986; Lewalter et al., 1985). The
biochemistry of chromium inside cells is discussed in Section 3.
Factors to be considered when implementing a monitoring program or when evaluating
results of studies using these matrices include the species of exposure, duration and dose of
exposure, and the analytical techniques and achievable LOD. Advantages and disadvantages are
associated with measuring chromium in each matrix and should be considered carefully,
especially when evaluating the dose-response behavior of different exposures and dosing

regimens across various toxicological studies. Negative and positive attributes of each
biomonitoring technique are provided in Table 2-5.
Table 2-5. Biomonitoring options for assessing chromium exposure
Biological matrix Advantages Disadvantages

Urine Easy sample collection, noninvasive, can Samples easily contaminated, difficulty in
evaluate high-level recent (within 48- distinguishing Cr levels from background,
hour) occupational exposure, good analytical limit of detection of 0.2 µg/L (typical
correlation for inhalation exposuresa background range), inability to distinguish between
Cr(III) and Cr(VI) exposure, seriously affected by
confounding variablesb
Red blood cell (RBC) Chromium detected up to 120 days Invasive technique, trained professionals required
(lifetime of RBC) following exposure, to collect samples, collection and analysis without
Cr(III) and Cr(VI) can be differentiated, contamination are difficult, costly to study large
analytical limit of detection of 0.09 µg/L populations
blood, good correlation for Cr(VI) oral
and inhalation exposures;c Cr(VI)
exposures may be identified 120 days post
exposure.
White blood cell (WBC) Animal studies show good correlation to Does not accumulate Cr(III), invasive technique,
Cr(VI) via oral and intravenous (i.v.) trained professionals required to collect samples,
exposure, accumulates Cr(V) exclusively, difficult to study large populations
accumulates Cr(VI) to a greater extent
than RBCd
Plasma Only Cr(III) confined to plasma Cr(VI) only detected up to 2 hours following
compartmente exposure, reduction of Cr(VI) in plasma is
significant, invasive technique, trained
professionals required to collect samples, difficult
to study large populationse
Hair Easy sample collection, noninvasive, Can not distinguish between Cr(III) and Cr(VI)
occupational exposure to Cr(VI) [in the exposure, inability to correlate time of exposuref
absence of Cr(III)] via inhalation has been
correlatedf
a
Mutti et al. (1979) and Tola et al. (1977) as cited in Paustenbach et al. (1997); bAnderson (1983), Gargas et al. (1994),
and Wiegand et al. (1988) as cited in Paustenbach et al. (1997); cGray and Sterling (1950), Wiegand et al. (1988), and
Kerger et al. (1996) as cited in Paustenbach et al. (1997); dCoogan et al. (1991) as cited in Paustenbach et al. (1997);
e
Wiegand et al. (1988) as cited in Paustenbach et al. (1997); fSaner et al. (1984) as cited in Paustenbach et al. (1997).
Source: Paustenbach et al. (1997)

3. TOXICOKINETICS
The internal environment of the GI tract, similar to the external environment discussed in
Section 2, can also affect the digestion, solubilization, and speciation of chromium compounds,
and thus impact their internal bioaccessibility (defined as the ability of chromium to be released
from the environmental matrix to which it is bound, i.e., the fraction of the dose ingested that
becomes freely available for absorption via crossing a cellular membrane), bioavailability
(defined as the potential for chromium to cross cellular boundaries, i.e., the degree to which it
becomes available to the target tissue after administration), and toxicokinetics within the body.
This section of the toxicological review discusses key determinants of these internal processes,
and why insights into how these processes vary across species and tissues are a critical
consideration for the mode of carcinogenic action of hexavalent chromium, Cr(VI).
As depicted in Figure 2-1 in the previous section, both environmental and internal
bioaccessibility processes are extracellular. Section 3.1 will discuss those physiological
processes, internal to the organism (but still extracellular), that impact bioaccessibility.
Bioavailability (processes D and E in Figure 2-1) will be discussed in Section 3.2. The available
data on the biochemistry of intracellular reactions involving chromium will be summarized in
Section 3.3. In Section 3.4, the impact of bioaccessibility and bioavailability on the toxicokinetic
component of the cancer mode of action for Cr(VI) will be discussed. These toxicokinetic
considerations are an important aspect of the mode of action and provide insights into the
evaluation of the tissue responses discussed in Section 4. Section 3.5 includes an evaluation and
discussion of available model structures to describe internal dosimetry of Cr(VI). Finally,
considerations of chromium essentiality versus toxicity are discussed in Section 3.6.
3.1. BIOACCESSIBILITY OF INGESTED HEXAVALENT CHROMIUM

3.1.1. In Vitro and Ex Vivo Studies of Bioaccessibility
A variety of in vitro and ex vivo studies have attempted to estimate the extent of
reduction of Cr(VI) to Cr(III) that is achieved in the GI lumen. The in vitro studies typically
develop surrogate formulations of physiological fluids, such as gastric juices, or employ different
beverages or diets in order to evaluate their reductive potential. Ex vivo studies of
bioaccessibility, on the other hand, evaluate the reductive potential of actual samples of saliva or
gastric juices.
Gammelgaard et al. (1999) used an artificial gastric juice to evaluate the degree of
reduction of Cr(III) inorganic compounds, chromium(III) picolinate, and hexavalent chromium,
Cr(VI). The artificial gastric juice was prepared from 2.0 g sodium chloride and 3.2 g pepsin
(1:10,000) dissolved in 4 ml of concentrated hydrochloric acid, diluted to 1 L with Milli-q water
with the pH adjusted to 1.2 with hydrochloric acid (HCl). All analyses were performed in

duplicate using atomic absorption spectrometry. The vessel volume was 1 L and the surrounding
water bath was kept constant at 37 ± 0.5° C. A volume of 900 ml of this artificial gastric juice
was added to potassium dichromate for an initial concentration of 100 µg/L and the Cr(VI)
concentration was observed for 4 hours. Chromium picolinate at a concentration of 100 µg/L
was studied in a similar fashion; experiments were repeated six times. Simultaneous
determination of Cr(VI) and Cr(III) was performed by ion chromatography with
chemiluminescence; the detection limit for both Cr(III) and Cr(VI) was 0.1 µg/L. The conversion
of Cr(VI) to Cr(III) in gastric juice followed first-order kinetics and a half-life (t1/2) of 23
minutes was calculated. In contrast, the organic chromium picolinate was unchanged.
Donaldson and Barreras (1966) performed ex vivo studies using everted sections of
female albino rat intestine. Post sacrifice, the intestines were immediately lavaged in situ with
saline, everted on glass rods, and cut into 1/8-inch sections. Rings from the proximal half of the
bowel were randomly distributed in Erlenmeyer flasks containing a range of concentrations (0.1
to 5.0 µg/ml) of either trivalent Cr51Cl3 or hexavalent Na2Cr51O4 in 10 ml of Krebs-Ringer
bicarbonate solution. Incubations were conducted for 60 minutes at 95%O2:5%CO2. Tissues
were then rinsed and digested with H2SO4 prior to assay for radioactivity. In vitro intestinal
uptake of the hexavalent Na2Cr51O4 was linear over the concentration range and significantly
greater than uptake of the trivalent Cr51Cl3.
The effect of gastric juice composition on uptake in these GI tissue samples was
investigated by mixing 25 µg of the labeled chromium compounds with 8 ml of either 0.1 N HCl,
human gastric juice (pH 1.4), or previously neutralized gastric juice (Donaldson and Barrera,
1966). After 30 minutes of incubation at room temperature, these mixtures were neutralized (to
pH 7.0) with 0.1 N NaOH and diluted to 25 ml with NaCl. Everted intestinal rings were then
incubated as described previously with 10 ml of the neutralized mixture and 1 µg/l of either
labeled chromium compound. To determine the proportion of the radioactivity bound to gastric
juice macromolecules, duplicate 5-ml aliquots were dialyzed in Viking cellulose bags against
repeated changes of NaCl for 48 hours. Results for these studies are shown in Table 3-1.

Table 3-1. Effect of gastric juice composition on binding and in vitro
uptake by rat intestinal rings of trivalent (Cr51Cl3) or hexavalent
(Na2Cr51O4 ) radiolabeled chromium compounds
Binding by Uptake
gastric juice by intestinal rings
Tissue preparation (µg/ml)1 (µg/gm)1
Trivalent Cr51Cl3 without gastric juice – 0.9 ± 0.1
Plus gastric juice pH 7.0 1.8 ± 0.2 0.2 ± 0.1
Hexavalent Na2Cr51O4 without gastric juice – 2.7 ± 0.4

1
Mean result of 4 experiments ± SD
Source: Donaldson and Barrera (1966).
DeFlora and collaborators (DeFlora et al., 1997; DeFlora and Wetterhahn, 1989; De Flora
et al., 1987; Petrilli and DeFlora, 1982) performed a series of studies to estimate the ability of
various human physiological fluids and tissues to reduce or sequester Cr(VI). The summary data
shown in Table 3-2 and depicted in Figure 3-1 represent a synthesis of data from studies in that
laboratory with anatomical and physiological parameters as described in detail in the source
reference (DeFlora et al., 1997). These parameters were used to arrive at estimates of the overall
Cr(VI) reducing or sequestering capacity of human body compartments relative to oral and
inhalation exposures. The general term “sequestration” was used to connote when intact cells
were tested and the term “reduction” was used when cell homogenates or their subfractions were
tested in the presence of an exogenous NADPH-generating system, an S9 mix. Estimates of
overall Cr(VI) reducing or sequestering capacity were calculated by multiplying the specific
reducing activity of a given organ, cell population, or fluid expressed as µg Cr(VI) reduced per
unit of weight, volume, or number, by the average content of the same organ, cell population, or
fluid in the human body (De Flora et al., 1997). De Flora (2000) proposed that these reduction
capacities account for the limited toxicity for Cr(VI) after oral ingestion due to efficient
detoxication by saliva, gastric juice and intestinal bacteria; similarly, lung cancer is only induced
when Cr(VI) doses overwhelm the reductive capacity of the fluid of the epithelial lining,
pulmonary alveolar macrophages, and bronchial tree and peripheral lung parenchyma cells. De
Flora (2000) also suggested that efficient uptake and reduction of Cr(VI) in RBC explains the
lack of carcinogenicity at sites remote to the portal of entry.

Source: De Flora et al. (1997).
Figure 3-1. Estimates of Cr(VI) sequestration or reduction by organs, cell populations and
fluids in the human body relevant to portal of entry uptake or effects on remote
distribution kinetics. See Table 3-2 and text for details on calculations.

Table 3-2. Estimates of the overall chromium(VI) reducing capacity of human fluids, cells and tissues
Organ, cell Weight of organs, Chromium(VI) Overall chromium(VI) reducing

population, or body number/weight/volume of cells, reduction or sequestration or sequestering capacity per
fluid or volume of body fluid (mean ± SD) individual
Saliva 500–1500 ml/day 1.4 ± 0.2 µg/ml 0.7–2.1 mg/day
Gastric juice 1000–15000 ml/day (fasting) 8.3 ± 4.3 µg/ml 8.3–12.5 mg/day during
interdigestive periods
+ 800 ml/meal 31.4 ± 6.7 µg/ml + 25.1 mg/meal
3400–3900 ml/day (3 meals) 84–88 mg/day (3 meals)
Intestinal bacteria 2.9–6.3 g eliminated daily with feces 3.8 ± 1.7 µg/109 bacteria 11–24 mg eliminated daily with feces
Liver 1500 g 2.2 ± 0.9 mg/g liver homogenate 3300 mg
Whole blood 4490 ml (males) 52.1 ± 5.9 µg/ml 234 mg (males)
3600 ml (females) intact blood 187 mg (females)
Red blood cells (RBC) 2030 ml (males) 63.4 ± 8.1 µg/ml 128 mg (males)
1470 (females) RBC lysate soluble fraction 93 mg (females)
Epithelial lining fluid 37.5–75 ml 23.7 ± 15.9 µg/ml 0.9–1.8 mg
(ELF)
Pulmonary alveolar 23 x 109 PAM 4.4 ± 3.9 µg/106 PAM 136 mg
macrophages (PAM) S9 fraction
Peripheral lung 1300 g 0.2 ± 0.07 mg/g lung 260 mg
parenchyma S12 fraction
Source: De Flora et al. (1997).

It should be noted, however, that because of the ability of many organic molecules to
reduce Cr(VI) in highly acidic solutions, the values reported by DeFlora and colleagues (DeFlora
et al., 1997; DeFlora, 2000) should be considered with some caution. The DeFlora studies relied
on the direct measurements of residual Cr(VI) in the calorimetric reaction with
diphenylcarbazide in the presence of 8% sulfuric acid, conditions that likely overestimated the
reducing capacities of these biological systems (Zhitkovich, 2005). More accurate
determinations of Cr(VI) reductive activities require the removal of organic molecules by
charcoal or other means prior to the addition of the reagent.
3.1.2. In Vivo Studies of Bioaccessibility

Donaldson and Barreras (1966) also performed in vivo studies of bioaccessibility in both
female albino rats and human volunteers. Female albino rats were fasted overnight and then
administered 1 ng of either radiolabeled trivalent Cr51Cl3 or hexavalent Na2Cr51O4 in 1 ml 0.9%
saline with an internal standard spike. Feces and urine were collected separately in individual
metabolic cages for 7 days post dosing. Excretion was expressed as percentage of administered
activity. Two weeks later, 1 ng of either radiolabeled trivalent Cr51Cl3 or hexavalent Na2Cr51O4
in 1 ml 0.9% saline was injected in anesthesized rats through a laparoscopic incision into the
lumen of the jejunem near the ligament of Treitz. The intestine proximal to the incision site was
compressed to prevent retrograde flow of the fluid. Rats were allowed to recover and urine and
feces collected as previously described. Fecal recovery of both forms of chromium was nearly
complete after oral administration in water (98 ± 4.2%, for Cr51Cl3 and 97.7 ± 2.5% for
Na2Cr51O4); whereas urinary recovery was 1.4 ± 0.7% and 0.8 ± 0.4% for Cr51Cl3 and
Na2Cr51O4. Intrajejunal administration resulted in significant absorption, as indicated by
increased urinary and decreased fecal recoveries of 16.5 ± 5.6% and 76.4 ± 8.9%, for the
hexavalent Na2Cr51O4, whereas only a slight increase in absorption for the trivalent Cr51Cl3 (91.6
± 3.4 % in feces and 4.3 ± 1.7%) occurred.
Donaldson and Barreras (1966) reported that adult volunteers hospitalized for either
obesity or pernicious anemia were administered 20 ng of trivalent Cr51Cl3 or hexavalent
Na2Cr51O4 either orally in drinking water or by an intestinal perfusion technique. The obesity
volunteers were otherwise healthy and the pernicious anemia patients were not anemic, but still
had histamine-fast achlorydria and vitamin B12 malabsorption when studied. For the ingestion
studies, urine was collected for 24 hour. The injection tubing was inserted under fluoroscopic
control into the small intestine so that its lumen opened at the ligament of Treitz as in the rats.
The proximal collection tube was located 15 cm distal to that site, and the distal collection tube
was 45 cm beyond. The infusion was performed at a constant rate of 10 ml/min after a 30-
minute equilibration period. The perfusion fluid contained 1% polyethylene glycol (PEG) as a
“nonabsorbable” reference marker and 2.5% xylose for calculation of absorption along with the
administered chromium. Fluid was collected for 1 hour and then fresh fluid was used for a 2nd

1-hour collection period. Fecal samples were collected for 6 days in a single container in which
the sample was homogenized. The majority of radioactivity was recovered in the feces when
either the trivalent or hexavalent compound was administered orally, but absorption was slighty
greater for the hexavalent form with recovery at 99.6 ± 1.8% for Cr51Cl3 and 89.4 ± 2.6% for
Na2Cr51O4. Urinary excretion was also slightly higher for the hexavalent Na2Cr51O4 (2.1 versus
1.5%). After intraduodenal administration, fecal and urinary excretion again indicated a low
absorption of the trivalent Cr51Cl3, whereas approximately 50% of the hexavalent Na2Cr51O4
appeared to be absorbed based on fecal excretion and approximately 10% appeared in the urine.
Calculation of the absorption using PEG confirmed a high absorption for the hexavalent
chromium compound and a minimal amount for the trivalent chromium compound.
In an extension of this perfusion study using Na2Cr51O4, Donaldson and Barreras (1966)
incubated the perfusion fluid with 0.1 N HCL for 30 minutes prior to neutralization with 0.1 N
NaOH and then used this fluid in the procedure described above with another five subjects. This
treatment did not result in impairment of absorption. They then incubated Na2Cr51O4 with 10 ml
of gastric juice at pH 1.4 for 30 minutes and proceeded with the procedure as before. This
pretreatment with gastric juice dramatically decreased the absorption to almost complete
inhibition. Absorption of xylose remained constant in these studies. Since it appeared that
absorption was significantly decreased by exposure to gastric juice, the researchers stratified
their analysis to evaluate control subjects versus those with pernicious anemia, and found a
significant decrease in the absorption for the latter group demonstrated by decreased fecal
excretion and increased urinary excretion.
3.2. BIOVAILABILITY, DISPOSITION, AND ELIMINATION OF INGESTED

HEXAVALENT CHROMIUM
The processes and factors that determine the ability of chromium to cross cellular
boundaries is defined as bioavailability. Key mechanisms of bioavailability that determine
internal tissue dose include the following: uptake through cell membranes, intracellular
distribution, and binding to cellular macromolecules. The potential for a chromium compound to
cross cellular boundaries is primarily affected by its solubility and valence state, which are, as
with environmental chemistry of chromium described in Section 2, dependent upon solubility
and pH. This section will describe the key events of chromium uptake and biochemistry within
cells as critical to its subsequent disposition in and elimination from the body. Studies that
provide a general understanding of the bioavailability and toxicokinetics of ingested Cr(VI) will
first be discussed here.
The extent of absorption of ingested hexavalent chromium from the GI tract is
determined by both the solubility of the hexavalent chromium compound ingested and how
rapidly hexavalent chromium is reduced to trivalent chromium, as trivalent chromium does not
diffuse as readily across cell membranes. Hexavalent chromium, on the other hand, can easily

cross cell membranes due to its ability to use existing nonspecific sulfate and phosphate anion
transport mechanisms.
Generally, absorbed hexavalent chromium is distributed throughout the body, but the
blood, liver, kidney, and spleen are the primary sites of distribution in addition to either the
respiratory or GI tract as the portal of entry. Bone is also a site of distribution, which may
contribute to the long-term retention kinetics of chromium. Absorbed chromium can be
transferred to fetuses through the placenta and to infants via breast milk. Chromium can also be
eliminated in hair, nails, and breast milk. There does not appear to be a gender difference in the
toxicokinetics of hexavalent chromium, and inter-individual variability in extracellular reduction
and subsequent absorption and elimination may be primarily driven by differences in gastric
contents and intervals between meals.
Quantitative descriptions of the pathways and mechanisms for this distribution, however,
have been constrained by detection limits and costs of the analytical methods and would be
especially informed by time course studies of speciated chromium content. Much of what is
considered generally known now about chromium kinetics has been inferred from studies
comparing Cr(III) to Cr(VI) compounds, but could be refined by studies with speciated
chromium measurements. Additionally, not all tissues have been routinely evaluated so that
much of the understanding is also based on blood, urine and fecal excretion studies. Finally,
there are limited studies comparing kinetics across species. Only two studies, Kargacin et al.,
1993 and the NTP study (Stout et al., 2009) have evaluated both rats and mice. The NTP (2007)
subchronic study also evaluated Guinea pigs.
3.2.1. In Vitro and Ex Vivo Studies of Bioavailability, Disposition, and Elimination

Wiegand et al. (1985) described the in vitro uptake kinetics of hexavalent chromium in
erythrocytes of rats and humans. No large species differences were observed, with both species
exhibiting Michaelis-Menten uptake kinetics, including an initial fast uptake rate (Table 3-3).
Table 3-3. In vitro kinetic parameters of hexavalent chromium uptake in

RBCs of rats and humans
Hexavalent chromium uptake Human Rat

Half-time (whole blood)
Initial phase 22.7 sec 6.9 sec
Second phase 10.4 min 10.1 min
Initial transport capacity (CrO42-/erythrocyte/min) 3.1 × 108 2.5 × 108
Whole blood kinetics
Vmax (μmol/mL/min) 2.8 3.0
Michaelis constant (Km) (mM/L blood) 20.9 14.1
Source: Wiegand et al. (1985).

K2Cr2O7, a hexavalent chromium compound, introduced into plasma and reconstituted
whole blood (stabilized with ethylenediaminetetraacetic acid [EDTA]) from three individuals
was readily reduced to trivalent chromium in the concentration range of 100–1,000 μg
hexavalent chromium/L (Corbett et al., 1998). Hexavalent chromium was detected in plasma
when spiked at concentrations of 2,000 and 10,000 μg hexavalent chromium/L, but not at
1,000 μg hexavalent chromium/L. Furthermore, the plasma to erythrocyte ratio of total
chromium decreased with increasing hexavalent chromium concentration. The variability
between subjects in the ratio of plasma to erythrocyte total chromium diminished by
approximately 1 order of magnitude as the hexavalent chromium concentration increased from
200 to 1,000 μg hexavalent chromium/L. Corbett et al. (1998) noted that these data suggest that
the reductive capacity of erythrocytes is much greater than plasma, and that the reduction rate of
hexavalent chromium in erythrocytes is greater than the rate of uptake from the plasma.
The partitioning of hexavalent chromium from plasma into erythrocytes is significant. It
has been used as a biomonitoring endpoint (Kerger et al., 1996; Minoia and Cavelleri, 1988) and
is responsible for the observed residence time of chromium in whole blood (Paustenbach et al.,
1996; Langard et al., 1978).
3.2.2. In Vivo Studies of Bioavailability, Disposition, and Elimination

Most of the in vivo studies in both laboratory animals and humans that provide data on
tissue uptake, disposition, and elimination of chromium evaluated total chromium and were
limited to blood, urine and feces. Some studies did evaluate target tissues such as liver and
kidney, and a few have evaluated uptake by the GI tract. As before, consideration of the
analytical methods and the constraint on inferences to be drawn from total chromium
measurements is an important consideration for evaluation of the data reliability and utility for
risk assessment.
3.2.2.1. Laboratory Animal Studies

In rats gavaged with a single dose of Na251CrO4, approximately 99% of the administered
dose was eliminated in the feces, while 0.8% was eliminated in the urine, both within 4 days
(Sayato et al., 1980). Rats given 0.138 µmol/day hexavalent chromium (approximately 7 µg/day
as Na251CrO4) for 3 days exhibited GI absorption of about 15% (Febel et al., 2001).
Approximately 81 and 2.17% was eliminated in the feces and urine, respectively (Febel et al.,
2001).
MacKenzie et al. (1959) performed several studies with radiolabeled chromium
(delivered as Na251CrO4) to evaluate absorption determinants and tissue distribution. In the first
study, a single oral gavage dose of 57 µg radiolabeled Na251CrO4 was administered to Sprague-
Dawley rats. The chromium solutions were adjusted to pH 7.5 prior to administration. Half of

the animals were fasted prior to chromium administration, the other half were fed a stock diet ad
libitum. Chromium concentrations in liver, kidney, stomach (plus contents), intestine (plus
contents), blood, spleen, brain, lung, submaxillary gland, urine, and feces were analyzed at 1, 7,
and 14 days post exposure. Table 3-4 summarizes the results; about 6% absorption occurred in
the fasted animals and 3% absorption in the nonfasted animals. Significant initial distribution in
the intestine was noted. The liver showed a maximal uptake of about 1% whereas the kidney,
blood, and spleen had a maximal content of 0.1 to 0.2%. The splenic level persisted unchanged
for 2 weeks post exposure. Most of the delivered dose was excreted in the feces.
Table 3-4. Distribution and retention of chromium in the rat after a single oral dose
Percent of the administered dose/total tissue

Post exposure day 1 Post exposure day 7 Post exposure day 14
Tissue Fasted Nonfasted Fasted Nonfasted Fasted Nonfasted
Stomach 1.964 2.22 0.02 <0.02 <0.02 <0.02
Intestine 26.78 18.2 0.07 0.04 0.05 0.03
Blood 0.17 0.03 0.05 0.03 0.05 0.03
Liver 1.03 0.21 0.14 0.03 0.07 0.03
Kidney 0.14 0.03 <0.02 0.02 — —
Spleen 0.02 <0.02 0.02 <0.02 0.02 <0.02
Urine 2.9 0.63 5.3 2.3 5.7 2.8
Feces 64.1 73.5 90.2 95.3 94.7 96.4
Total, % 97.1 94.8 95.8 97.8 100.6 99.3
Source: MacKenzie et al. (1959).
In the second part of the same study, MacKenzie et al. (1959) administered 131 µg
chromium (as Na251CrO4) and blood was removed from the heart at 4 hour post exposure for
analysis of chromium in RBCs and plasma. In a third experiment, MacKenzie et al. (1959) also
evaluated the role of the stomach wall in absorption. Using the same dosing, the influence of the
stomach was bypassed by injecting chromium directly into the intestine, about 4 cm below the
stomach. The RBC and plasma were measured for radioactivity 4 hour post exposure as in the
second part of the study. Table 3-5 shows that the concentrations of chromium in whole blood,
plasma, and RBCs were greater after administration to the intestine than to the stomach (0
indicates no significant count above background).

Table 3-5. Ratios (intestine:stomach) of chromium concentration in whole
blood, plasma, and RBCs after a single oral dose
Whole blood Plasma RBC

Cr(VI), fasted 1:3.94 1:2.10 1:6.03
Cr(VI), nonfasted 1:3.34 1:2.68 1:7.46
Cr(III), fasted 1:1.48 1:1.46 0:0
Cr(III), nonfasted 1:1.11 1:0.90 0:0
Since RBC chromium is assumed to be in the hexavalent form and plasma chromium is assumed
to be in the trivalent form, it appeared that there was some reduction of the hexavalent form in
both fasted and nonfasted animals. Ratios for the counts in RBC to plasma (RBC:plasma) are
shown in Table 3-6.
Table 3-6. Ratios of chromium concentration in RBCs and plasma in the

stomach and intestine for fasted and nonfasted animals after a single oral
dose
Organ and condition RBC:plasma

Stomach, fasted 1:4.76
Stomach, nonfasted 1:8.8
Intestine, fasted 1:1.66
Intestine, nonfasted 1:3.17
MacKenzie et al. (1958) administered potassium chromate (K2CrO4) in drinking water to

Sprague-Dawely albino rats (n = 4/sex/group or n = 5/sex/control) at concentrations of 0, 0.45,
2.2, 4.5, 7.7 and 11 ppm (chromate ion) for one year. Two other groups were given water
containing 25 ppm of either K2CrO4 or trivalent chromic chloride (CrCl3) for the same period.
Chromium was analyzed in liver, kidneys, and femurs at 6 months and in these tissues plus
spleen after one year by Saltzman’s diphenylcarbazide method of permanganate oxidation. No
changes in weight gain or food consumption were reported, but water intake was decreased to
77% in females and 84% in males. Resultant tissue concentrations for the rats receiving K2CrO4
are presented in Table 3-7. The order of chromium concentrations in tissues was as follows:
spleen > bone > kidney > liver. No gender-specific differences in chromium tissue accumulation
were observed. An appreciable increase in all tissues examined occurred when the animals
received between 5 and 10 ppm K2CrO4. Concentrations in these tissues were approximately

ninefold higher in the group given K2CrO4 than those receiving the trivalent compound, again
suggesting the greater bioavailability of hexavalent chromium.
Table 3-7. Terminal tissue chromium levels in rats ingesting potassium

chromate (K2CrO4) in drinking water for 1 year
K2Cr2O4 Liver Kidney Bone Spleen

concentration (μg/g) (μg/g) (μg/g) (μg/g)
(mg/L) Male Female Male Female Male Female Male Female
Controls 0 0 0 0.25 ± 0.02 0 0.72 ± 0.8 0 0
0.45 0.02 ± 0.002 0.08 ± 0.007 0.14 ± 0.007 0.39 ± 0.04 0.58 ± 0.04 0.76 ± 0.04 0.95 0.91 ± 0.11
2.2 0.08 ± 0.017 0.17 ± 0.03 0.29 ± 0.02 0.48 ± 0.07 1.27 ± 0.06 1.48 ± 0.04 0.68 ± 0.18 1.14 ± 0.1
4.5 0.15 ± 0.04 0.47 ± 0.06 0.45 ± 0.17 1.09 ± 0.13 2.14 ± 0.25 2.44 ± 0.25 3.41 ± 0.44 4.48 ± 0.71
7.7 0.70 ± 0.04 0.55 ± 0.06 3.30 ± 0.03 2.39 ± 0.09 3.43 ± 0.83 5.10 ± 0.35 5.24 ± 0.20 4.73 ± 0.8
11.2 1.22 ± 0.06 1.62 ± 0.14 4.40 ± 0.36 3.98 ± 0.32 3.84 ± 0.49 6.06 ± 0.58 9.91 ± 0.83 11.1 ± 0.86
Coogan et al. (1991a) exposed male F344 rats to hexavalent chromium as potassium
chromate (K2CrO4) dissolved in their drinking water at concentrations of 100 and 200 ppm for 3
or 6 weeks. Total chromium concentrations were measured in lung, liver, kidney, and blood by
AAS after acid digestion. Drinking water consumption was reduced at both concentrations for
the first 3 weeks and then only at the higher concentration the second 3 weeks. Chromium was
not detected in any lung samples. At both concentrations and durations, the order of tissue
chromium concentrations was as follows: kidney > liver > blood. Although a general trend of
increasing chromium content as a function of exposure duration existed for the liver, kidney, and
blood samples analyzed, only the kidney samples were significantly different between the 3- and
6-week sacrifices (p < 0.05). Blood chromium levels were not significantly different at either
sacrifice or at either concentration.
Witmer et al. (1991, 1989) performed several studies to evaluate the toxicity of
chromium contaminated soil samples from sites in Jersey City, NJ. In all studies, the organs
evaluated for chromium content included the liver, lung, spleen, kidney, muscle, brain, and
testes. An aliquot of blood from the abdominal aorta was also analyzed. Chromium was
identified in samples by the Baird ICP method (urine and feces) or by atomic absorption using a
graphite furnace (other tissues). In the initial pilot study, male Sprague Dawley rats were dosed
by gavage with 0, 20, 40 and 100 µmole/kg hexavalent chromium as Na2CrO4 · 4H2O in distilled
water for 7 days. Recovery of chromium was low, tissue burdens at the highest dose represented
only 1.7% of the amount administered. When tissue burden was expressed on a µg/g basis, the
kidney contained the highest amount with liver and blood also containing greater amounts than
the other samples. In the next experiment, doses of 120 µmole Cr/kg were administered to
Sprague-Dawley rats (n = 3/group) using four sources of chromium: 1) Na2CrO4, 2) CaCrO4, 3)

Pacific Avenue Fines [PAC] soil sample obtained from the NJDEP, and 4) a mixture of soil and
calcium chromate containing 60 µmole Cr/kg each. Recovery was again low whether expressed
as percent of total administered dose or based on the last dose before sacrifice, with the sampled
tissues accounting for less than 2% of the total administered dose, suggesting minimal tissue
storage; whatever the source, the kidney had the highest tissue concentrations. The absorption
from the soluble sodium salt was generally higher in all tissues studied than from the calcium
salt, soil, or mixture. In a third study, groups of male Sprague Dawley rats (n = 6/group) were
orally gavaged with chromium in corn oil and at a higher concentration (240 µmol Cr/kg) from
the same four sources for a longer duration (14 days). The total tissue levels again represented a
small percentage of the chromium administered, but this time, blood levels were higher than
those found in the kidney. Thus, to determine if the major portion of the orally administered
chromium was rapidly excreted, urine and feces were collected in the last experiment in which
rats (n = 3/group) were gavaged with 240 µmol Cr/kg as the calcium salt or in contaminated soil.
Dosing was carried out once daily at the same time each day for eight days. Urine and feces
were collected at 6, 12, and 24 hour after dosing on days 1 and 2 and on days 7 and 8; chromium
in the samples was measured by the ICP method. Results for urine and feces on the different
days are shown in Figure 3-2. The data indicate that chromium is not excreted under these
conditions of corn oil dosing to any appreciable extent in the urine, but some significant
excretion occurs in the feces. The patterns also show that more of the chromium from soil was
excreted at both sample periods in the urine and feces than from the rats treated with CaCrO4.

Source: Witmer et al. (1991, 1989).
Figure 3-2. Chromium excretion in urine and feces of Sprague-Dawley rats. Rats were
gavaged with corn oil (control), calcium chromate, or chromium-contaminated soil at a level of
240 µmole Cr/kg daily for either 2 days (upper two panels) or 8 days (lower two panels). Urine
and feces were collected at 6, 12, and 24 hour after each dosing period. Chromium levels shown
are for each time period as well as for the total 24-hour period.

As part of the development of the physiologically-based pharmacokinetic (PBPK) model
discussed in Section 3.5, O’Flaherty and Radike (1991) dosed male Sprague-Dawley rats with
either trivalent chromic chloride-hexahydrate (CrCl3 · 6H2O) or hexavalent sodium dichromate
(Na2Cr2O7 · 2H2O) administered either in drinking water or by inhalation. The inhalation
concentration (for 6 hour/day) chosen (200 µg Cr/m3) was based on the lack of toxicity observed
at that same concentration in a study conducted in rats by Glaser et al. (1985). The control group
was exposed to filtered air for the inhalation study and received deionized water for the ingestion
study. Chromium aerosols were generated from solutions by ultrasonic jet nebulizers and
chromium concentrations were measured daily by AAS; specific Cr(VI) analysis was performed
weekly by the diphenylcarbazide method (NIOSH method 7600). The mass mean aerodynamic
diameter (MMAD) for the Cr(III) aerosol was 0.9 µm ± 0.28, 80% respirable; and the Cr(VI)
MMAD was 1.0 µm ± 0.24, 63% respirable. The concentration in drinking water (deionized) of
12.9 mg Cr/L was based on the dose calculated to be delivered by the inhalation route in a 200 g
rat. Each exposure group contained 36 rats; 6 rats from each group were sacrificed following
days 2, 5, 10, 20, and 40 of exposure. The 36 rats remaining after the 40-day exposure were
allowed to live untreated 20 days longer (day 60 of experiment) to ascertain if there were
clearance differences among treatment groups. Blood, urine, and fecal chromium levels were
monitored on days 2, 5, 10, 20, and 40 for the 40-day exposure period, and again at 20 days post
exposure. Chromium content in blood and urine were measured by ICP-MS. Tissues were
digested in acid, ashed, and chromium content determined by AAS. Tissue total chromium
content was determined in kidney, liver, muscle, intestine, lung, and carcass. The summary of
the experimental data for the ingestion route is provided in Table 3-8.

Table 3-8. Time course of chromium tissue concentrations in male Sprague-Dawley rats1 ingesting 12.9 mg Cr/L of
trivalent chromic chloride-hexahydrate or hexavalent sodium dichromate in drinking water for 40 days
Day of Lung Liver Intestine Kidney Muscle Blood Urine Feces

study µg Cr/g µg Cr/g µg Cr/g µg Cr/g µg Cr/g ng Cr/ml µg Cr/day mg Cr/day
Control
2 ND2 ND 0.65 1.58 Trace 1.5 0.017 ND
5 ND ND 0.83 ND Trace 1.6 ND 0.002
10 ND ND 0.56 ND ND 4.2 0.003 ND
20 ND ND 0.85 ND Trace 3.4 ND 0.013
40 ND 0.035 0.68 ND Trace 6.8 0.010 ND
60 ND 0.032 0.72 ND 0.038 2.5 ND ND
Trivalent chromic chloride hexahydrate
2 ND 0.042 18.3 ND ND 2.48 0.227 0.821
5 ND Trace 17.2 ND ND 3.11 0.065 0.729
10 ND 0.034 20.6 ND ND 16.8 0.040 1.20
20 ND ND 26.8 ND ND 5.60 0.075 1.07
40 ND ND 7.15 ND ND 4.72 0.017 1.12
60 ND Trace 0.83 ND ND 5.52 2.01 ND
Hexavalent sodium dichromate
2 ND 0.209 15.5 0.249 ND 9.0 0.622 0.997
5 ND 0.372 22.7 0.588 ND 11.8 1.79 0.835
10 ND 0.585 14.4 1.60 ND 18.5 2.01 0.949
20 1.17 1.18 29.0 1.71 0.077 48.9 3.08 0.977
40 0.65 1.50 6.8 1.909 0.103 58.3 2.19 1.51
60 0.65 0.509 0.83 0.634 0.070 11.3 0.217 ND
1
n = 6 per time point per exposure group
2
Non detect
Source: O’Flaherty and Radike (1991).

Kargacin et al. (1993) examined the species differences in distribution of total chromium
in male C57BI/6J mice and F344 rats exposed to 130 ppm potassium chromate (K2CrO7) in
drinking water (8 mg hexavalent chromiujm/kg-day) for 4 or 8 weeks. Total concentrations in
the tissues listed in Table 3-9 were measured by AAS after acid digestion. Regardless of
duration, chromium accumulated primarily in the spleen, liver, kidney, and bone of mice and
rats, with mouse liver, spleen, and bone burdens being, on average, several-fold higher than in
rats (Table 3-9). The reason for the higher accumulation of chromium in mouse liver is
unknown, but may result from greater reduction of hexavalent to trivalent chromium in the rat
gut limiting uptake of chromium from the GI tract. Alternatively, the mouse liver may have a
higher hexavalent chromium reduction capacity than rats, causing more reduced trivalent
chromium to be sequestered in hepatocytes.
Table 3-9. Chromium in tissues (μg/g wet tissue or μg/mL blood) of mice
and rats after ingesting K2CrO7 in drinking water (8 mg hexavalent
chromium/kg-day) for 4 or 8 weeks
Controls 4-Week exposure 8-Week exposure

Mice
Liver 0.22 ± 0.14 10.92 ± 5.48 13.83 ± 6.06
Kidney 0.24 ± 0.14 3.77 ± 0.99 4.72 ± 0.68
Spleen 0.53 ± 0.38 5.04 ± 1.45 10.09 ± 2.50
Femur 0.90 ± 0.48 7.43 ± 1.03 12.55 ± 2.99
Lung 0.24 ± 0.12 0.99 ± 0.10 1.08 ± 0.26
Heart 0.32 ± 0.15 0.80 ± 0.23 1.02 ± 0.20
Muscle 0.32 ± 0.23 1.12 ± 0.37 0.60 ± 0.25
Blood 0.14 ± 0.05 0.71 ± 0.07 0.42 ± 0.04
Rats
Liver 0.19 ± 0.14 3.32 ± 0.93 3.59 ± 0.73
Kidney 0.34 ± 0.20 8.62 ± 2.40 9.49 ± 4.38
Spleen 0.43 ± 0.20 3.65 ± 1.87 4.38 ± 0.84
Femur 1.00 ± 0.46 1.85 ± 0.46 1.78 ± 0.99
Lung 0.39 ± 0.43 1.10 ± 0.38 0.67 ± 0.24
Heart 0.38 ± 0.22 0.52 ± 0.12 1.05 ± 0.19
Muscle 0.24 ± 0.14 0.19 ± 0.10 0.17 ± 0.10
Blood 0.19 ± 0.17 0.73 ± 015 0.58 ± 0.13
Source: Kargacin et al. (1993).
Sutherland et al. (2000) determined tissue concentrations of chromium in male and

female F344 rats (n = 7/sex/group) that drank hexavalent potassium chromate (K2CrO4) ad
libitum at 0, 0.5, 3.0 and 10.0 ppm for 44 weeks. Solutions were prepared in deionized water
weekly; consumption rates were recorded weekly. Rats were switched to deionized water only 4

to 6 days prior to sacrifice to ensure tissue concentrations did not reflect recent exposure. Rats
were euthanized by i.p. injection of sodium pentobarbital and exsanguinated prior to tissue
harvest. Kidneys (0.6–1.4 g) were digested in 1 N HNO3 and chromium content measured by
atomic absorption spectrophotometry with deuterium background correction. Liver (0.7–1.2 g),
brain (0.8–1.1 g), bone (0.5–1.0 g), whole blood (1.5–2 ml), testes (0.7–1.3 g), and ovarian (0.1–
0.2 g) samples were digested in a low trace metal reagent grade HNO3/HCLO4 mixture and
analyzed by inductively coupled plasma (optical) – atomic emission spectrophotometry (ICP-
AES) or inductively coupled plasma – mass spectrophotometry (ICP-MS). Because bone
samples were optically saturated with calcium and phosphorous, they were reanalyzed with ICP-
MS. ICP-MS was also used for the testis and ovarian samples due to their small size and the
better detection limit for this method. Detection limits of the ICP-AES and ICP-MS methods
were 5 and 2.5 ppb chromium in solution. Analytical accuracy was assessed by measuring total
chromium in liver, bone, brain, and testis samples spiked with a known amount of Cr(VI).
Recovery of chromium in these spiked samples was 101.7 ± 2.9% (range = 92.3–116.4%).
Additionally, spikes of chromium from a National Institute of Standards and Technology (NIST)
source independent of instrument calibration standards were also analyzed as unknowns and
average recovery was 95.7 ± 1.3% (range 86.5–100.6%). Tissue (bone, kidney, liver) and total
body burdens are shown in Figure 3-3. Chromium was most concentrated in kidney and bone in
this study, results consistent with those of MacKenzie et al. (1966) and Witmer et al. (1985).
Despite consuming more water than the males, the females did not have significantly higher
body burdens (Panel D, Figure 3-3) than males. The lack of difference between the sexes may
be due to the contribution of testicular burdens to the total, or perhaps the chromium was either
less bioavailable in females or cleared more effectively. Significant tissue accumulation
occurred at the 3 and 10 ppm exposure levels, with the effect most pronounced at 10 ppm,
indicating that a portion of the Cr(VI) escaped extracellular reduction in the GI tract and became
bioavailable for systemic distribution. An alternate mechanism proposed for the findings was
that Cr(III) formed in the GI tissues and absorbed was not cleared in the kidneys and was taken
up by the cells. Male rats showed a greater concentration in kidneys at the 3 ppm level than did
the females, of note since the females consumed more Cr per gram body weight. The liver was
the only tissue in female rats in which significantly elevated concentrations of chromium could
be found after ingestion of Cr(VI) at the 3 ppm level. Testicular concentrations were slightly
elevated in rats that drank 10 ppm Cr(VI). Brain, ovarian, and whole-blood concentrations were
below detection limits in all exposed groups. The lack of concentrations in whole-blood was
attributed to rapid delivery of Cr to tissues and clearance of plasma Cr.

Source: Sutherland et al. (2000).
Figure 3-3. Chromium concentrations in male and female F344 rats following chronic
drinking water consumption of Cr(VI). Bone (upper left), renal (upper right), and liver (lower
left) tissue burdens or total body burden (lower right) are mean ± SE. Means that do not share a
common superscript are significantly different (p ≤ 0.05).

The NTP conducted a comparative absorption subchronic study in rats, mice, and guinea
pigs prior to conducting the 2-year bioassay (NTP, 2007; Appendix G). Guinea pigs were
chosen for the study because they more closely resemble humans in that they do not have a
forestomach and require a reducing agent (Vitamin C) in their diet. Sodium dichromate
dihydrate was administered ad libitum in drinking water to male F344/N rats, B6C3F1 mice, and
Hartley guinea pigs (n = 4/species/dose) at dose concentrations of 0, 2.87, 8.62, 28.7, 86.2, 287,
and 862 mg Cr(VI)/L (equivalent to 0, 1, 3, 10, 30, 100, and 300 mg Cr/L) for 21 days, followed
by 2 days of drinking water alone. Animals were sacrificed on day 24, and total chromium
concentrations in blood, kidney, and femur (rats only) were determined. Blood and renal tissue
concentrations increased in all three species with dose. A statistically significant and apparently
nonlinear uptake in blood and kidney at the two highest concentrations was evident in all species
and in guinea pigs at 30 ppm. Uptake in guinea pigs did not appear to generally differ from that
of rodents, suggesting that the lack of a forestomach did not alter GI uptake appreciably. Values
for the rats and mice were in agreement with those previously published by Sutherland et al.
(2000).
As part of the 2-year NTP cancer bioassay (NTP; Stout et al., 2009) discussed in Section
4, a total chromium tissue distribution study was also conducted (NTP, 2008; Appendix J).
Groups of 40 male F344 /N rats and female B6C3F1 mice were randomly assigned to the tissue
distribution study at the beginning of the 2-year bioassay and treated identically to the core study
groups. Animals were exposed to sodium dichromate dehydrate (Na2Cr2O7 · 2H2O) in drinking
water at concentrations of 0, 14.3, 57.3, 172, or 257.4 mg/L for 53 weeks. Equivalent Cr(VI)
concentrations based on the percent mass of Cr in sodium dichromate dehydrate are 0, 5, 20, 60
and 190 mg/L. Dose formulations were prepared approximately every two weeks in tap water.
On days 4, 11, 180, and 369, up to 10 animals/dose group were removed from treatment and
placed in individual metabolism cages to allow for separate collection of urine and feces for 48
hours. Two collections of urine and feces were made to include the intervals from 0 to 24 and 24
to 48 hour; measured values were combined to yield the reported 48-hour values. At the end of
48 hours, animals were anesthetized with CO2/O2 for retroorbital sinus sampling of blood. Blood
was separated into cells and plasma. While the animals remained anesthetized the aorta was
severed and the abdominal wall opened to obtain liver, kidneys, and stomach (separated into
glandular and forestomach). Only plastic, ceramic, Teflon, or tungsten carbide instruments were
used to avoid chromium contamination from stainless steel. Tissue samples were digested in
concentrated nitric acid. Chromium content of the experimental samples was analyzed by ICP-
MS using spiked internal standards; with calibration performed prior to each analysis. Fecal and
urinary chromium concentrations expressed as µg were significantly elevated relative to controls
at all concentrations and at all sacrifice times in both the male rats and female mice, with the
majority of the chromium in the feces. Plasma and RBC chromium concentrations (µg/g) were

consistently significantly elevated relative to controls at the two highest exposure levels. Tissue
chromium concentrations were significantly elevated relative to control at all concentrations
(Shirley’s test; p < 0.05 at lowest exposure level and p < 0.01 at all others) in both liver and
kidney samples, with liver concentrations higher than kidney values in both the rats and mice.
The tissue chromium concentrations for the glandular and forestomach samples in both species
are presented in Table 3-10. As for the other tissues sampled, chromium concentration in the
glandular stomach and forestomach were significantly elevated relative to controls at the two
highest concentrations. Stout et al. (2009) calculated the average daily dose in mg/kg for
animals on test in the main study using body weight and water consumption data as follows:
Sodium dichromate dihydrate (mg/L): 0 14.3 57.3 172 516

Average daily dose (mg/kg):
Male rats 0 0.6 2.2 6 17
Female mice 0 1.1 3.9 9 25
Ratio (mice:rats) 0 1:1.83 1:1.77 1:1.50 1:1.47
Assuming a similar consumption and weight pattern in these satellite animals, it can be seen that
the mice generally consumed 1.5 to 1.8 more chromium as an average daily dose (mg/kg), which
may account to some degree for the larger tissue concentrations in this species observed in Table
3-10.

Table 3-10. Tissue concentrations of chromium in male F344/N rats and female B6C3F1 mice in the 2-year NTP
drinking water study of sodium dichromate dihydrate
0 mg/L 14.3 mg/L 57.3 mg/L 172 mg/L 516 mg/L

1
Tissue (µg/g) Rats
Glandular stomach
Day 6 0.076 ± 0.003 0.143 ± 0.008* 0.333 ± 0.040** 0.773 ± 0.105** 1.967 ± 0.109**
Day 13 0.095 ± 0.008 0.254 ± 0.073* 0.310 ± 0.049* 1.331 ± 0.060** 1.762 ± 0.042**
Day 182 0.197 ± 0.031 0.414 ± 0.012* 1.043 ± 0.081** 4.300 ± 0.367** 9.886 ± 0.354**
Day 371 0.253 ± 0.066 0.334 ± 0.029 1.038 ± 0.115* 4.801 ± 0.345** 14.643 ± 0.121**
Forestomach
Day 6 0.098 ± 0.024 0.076 ± 0.004 0.122 ± 0.008 0.294 ± 0.029 0.285 ± 0.283
Day 13 0.091 ± 0.015 0.102 ± 0.034 0.171 ± 0.050 0.221 ± 0.055* 0.593 ± 0.159**
Day 182 0.089 ± 0.018 0.099 ± 0.003 0.338 ± 0.022* 0.574 ± 0.171* 1.654 ± 0.244**
Day 371 0.090 ± 0.015 0.118 ± 0.008 0.328 ± 0.081* 1.338 ± 0.444** 2.849 ± 0.975**
Mice
Glandular stomach
Day 6 0.306 ± 0.056 0.645 ± 0.253 1.258 ± 0.290* 2.450 ± 0.266** 5.785 ± 0.131**
Day 13 0.207 ± 0.053 0.324 ± 0.030 2.614 ± 0.190* 7.048 ± 1.751** 13.130 ± 2.604**
Day 182 0.305 ± 0.078 0.644 ± 0.035* 3.659 ± 0.547** 11.520 ± 3.017** 52.673 ± 12.310**
Day 371 0.731 ± 0.306 0.676 ± 0.104 2.807 ± 0.330* 9.994 ± 1.079* 49.867 ± 12.251**
Forestomach
Day 6 0.328 ± 0.132 0.683 ± 0.262 1.308 ± 0.553 1.102 ± 0.373 1.286 ± 0.116
Day 13 0.201 ± 0.094 0.288 ± 0.056 0.400 ± 0.044 2.030 ± 0.532* 3.849 ± 1.811*
Day 182 0.173 ± 0.064 0.444 ± 0.099 1.033 ± 0.102* 2.141 ± 0.643** 9.624 ± 3.638**
Day 371 0.320 ± 0.049 0.381 ± 0.077 1.271 ± 0.300* 1.812 ± 0.208* 7.442 ± 0.764**
1
n = 3; mean ± SE. Statistical tests performed on unrounded data.
*significantly different from the control group (p ≤ 0.05) and ** (p ≤ 0.01) by Shirley’s test
Source: NTP (2008; Appendix J).

Hexavalent chromium is capable of crossing the placenta. Pregnant mice given a single
intravenous injection of 10 mg hexavalent chromium/kg (as Na251CrO4) on gestation
day (GD) 13 exhibited total embryo chromium levels that were 12% of maternal blood levels
(Danielsson et al., 1982). Intraperitoneal injection of 10 mg trivalent chromium/kg (as 51CrCl3)
in pregnant mice on GD 8 resulted in approximately equal [51Cr] activity in the embryo and
maternal blood (Iijima et al., 1983). While these studies demonstrate placental transfer of
chromium, they are of limited use for assessing embryonic exposure to chromium as a result of
maternal oral exposures to hexavalent chromium.
3.2.2.2. Human Studies

Most quantitative studies of the GI absorption of both hexavalent and trivalent chromium
in humans have estimated the absorption fraction to be <10% of the ingested dose. In general,
these studies suggest that the absorbed fraction of soluble hexavalent chromium compounds
(e.g., K2Cr2O7) is higher than insoluble forms; and soluble hexavalent chromium compounds are
absorbed to a greater extent than soluble trivalent chromium compounds (e.g., CrCl3).
Absorption and elimination of trivalent and hexavalent chromium, following ingestion by
human volunteers of either single or multiple drinking water doses, were evaluated in a series of
studies (Finley et al., 1996, 1997; Kerger et al., 1996, 1997; Paustenbach et al., 1996).
Collectively, these studies illustrate absorption and elimination kinetics for human volunteers
that provide critical data for the interpretation of biomonitoring in blood and urine, but the
disposition of chromium and in particular its speciation in different tissue compartments, can
only be inferred by comparison of these same rates (i.e., absorption and elimination) between
chromium (III) and chromium (VI) compounds. Nevertheless, they can provide a basis for
human model development (see Section 3.5) and a comparison for the laboratory animal studies.
Considerable variability across the human volunteers was noted in these studies, and may reflect
interindividual differences that influence gastric reduction such as time since last meal or volume
of the contaminated water ingested, but may also reflect different genetic capacities for Cr(VI)
reduction.
Of note, a human use committee reviewed the protocols prior to initiation and concluded
that the study design was adequate to meet the objectives of the study; that the dosing would not
pose a health risk to the participants, and that the participants were properly informed of
potential risks.
Kerger et al. (1996) examined the absorption (first 2 hours) and elimination (up to 14
days) kinetics of three different chromium compounds: 1) chromic chloride (Cr(III) as CrCl3);
2) potassium dichromate reduced with orange juice (Cr(VI)-OJ); and 3) potassium dichromate
(Cr(VI) as K2Cr2O7). In each experiment, three or four adult male volunteers ingested a single
bolus dose of 5 mg Cr (0.5 L of 10 ppm chromium solution in deionized water). One volunteer
was common to each of the experiments. Their diet was not controlled, but participants were

prohibited from ingesting vitamin supplements containing Vitamin C or chromium. A detailed
log of all intakes (food, beverage, and dietary supplementation) was maintained throughout the
duration of the study. Urine voids were collected beginning with the 1st morning void through
the last day of the study. Blood samples were drawn, centrifuged and pooled for analysis of
RBC and plasma at various intervals. All urine samples were analyzed for total chromium, total
urine volume, specific gravity and creatinine. Total chromium was measured in urine, plasma
and RBC samples by graphite furnace AAS according to EPA Method 218.2 as modified by
Pautenbach et al. (1996); the limit of detection was 0.5 µg Cr/L of urine, 0.3–0.5 µg Cr/L of
plasma, and 0.2 µg Cr/L of blood. The absorption was highly dependent on the form of
chromium ingested. CrCl3 was poorly absorbed (estimated 0.13% bioavailability) and rapidly
eliminated in urine (excretion half-life, T1/2, ~10 hour), whereas Cr(VI)-OJ was absorbed more
efficiently (0.60% bioavailability) and eliminated more slowly (T1/2 ~17 hour). Cr(VI) had the
highest bioavailability (6.9%) and longest T1/2 (~39 hour). All three compounds caused
temporary elevation in RBC and chromium concentrations with the magnitude and duration of
elevation exhibiting a clear trend: Cr(VI) > Cr(VI)-OJ > Cr(III). Figure 3-4 shows a comparison
of the absorption/elimination profiles for the three chromium compounds. Figure 3-5 shows a
comparison of the percent bioavailability and elimination T1/2. Plasma chromium concentrations
peaked within 90 minutes after the Cr(III) dosing and averaged 2.8 µg Cr/L (range 1.3–3.7).
Potassium dichromate reduced in orange juice resulted in a lesser elevation, with an average of
2.2 µg Cr/L; whereas potassium dichromate resulted in an average peak concentratin of 26 µg
Cr/L (range 5.1–57). The peak chromium concentrations in RBC mirrored those in plasma, with
the peak average concentrations for the Cr(III), Cr(VI)-OJ and Cr(VI) reported as 7.5 (range 5.1–
14), 5.5 (range 5.1–6.1), and 17.6 (range 13.5–24) µg Cr/L, respectively. Because the Cr(VI)
increases in RBC provide a specific indication of chromium in the hexavalent state, these data
suggest that at the low doses tested, there was predominant reduction of the ingested Cr(VI) in
the stomach and small intesting followed by systemic uptake. Distribution and excretion was
postulated to be as Cr(III) organic complexes. Because the increases in chromium content were
always similar for plasma and RBC, an equilibrium between the cellular and noncellular
compartments of the blood was suggested and considered to be consistent with the kinetic
behavior of absorbed Cr(III). The higher bioavailability of the Cr(VI)-OJ compound could
represent the formation of highly soluble chromium complexes (e.g., ligands with acorbate or
sulfhydryl proteins) that were rapidly cleared from plasma via tissue uptake or kidney filtration.
Alternatively, it is possible that a small fraction was absorbed as Cr(VI) where it encountered
additional reducing agents and potential complexation ligands that produced a variety of
chromium complexes with different kinetic patterns. Although it is unlikely that the reducing
capacity of the blood was overwhelmed at the absorbed doses in this study, elucidation of the
mechanisms would require additional investigations and be greatly informed by analytical
techniques capable of differentiating the speciated state in biological samples.

Source: Paustenbach et al. (1997).
Figure 3-4. Chromium absorption and elimination in human volunteers after

ingestion of a single bolus dose in drinking water.
Source: Paustenbach et al. (1997).
Figure 3-5. Biovailability and elimination half-life for chromium ingested by

human volunteers as a single bolus dose in drinking water.

Kerger et al. (1997) next explored the hexavalent chromium absorption, distribution, and
excretion following oral exposure in human volunteers (n = 5 adults) to 5 or 10 mg Cr(VI)/L in
drinking water administered as either a single bolus dose (0.5 L swallowed in 2 minutes) for a
total dose of 5 or 10 mg; or for 3 days at a dosage of 1 L/day (3 doses of 0.33 L each day, at 6-
hour intervals) for a total dose of 15 or 30 mg. The source of Cr(VI) was potassium dichromate
(K2Cr2O7) for the single dose experiment and potassium chromate (K2CrO4) for the multiple-
dose experiment. Other experimental methods and total chromium analyses in the samples were
as in Kerger et al. (1996); a study of the stability of the dosing solution confirmed no measurable
reduction of Cr(VI) occurred throughout the study. The percent uptake of total chromium
measured in the urine was 5.7% for the bolus dose at 10 mg Cr(VI)/L, and 1.7 and 3.4% for the
multiple doses at 5 and 10 mg Cr(VI)/L, respectively. Plasma and RBC total chromium levels
were consistent in timing and magnitude, with a temporary elevation about 60 minutes after
bolus dosing. The average peak concentration of total chromium after the multiple dosing
regimen in plasma and RBC showed dose trends and timing generally consistent with the bolus
dose experiment, but with greater interindividual variability. Urine and plasma samples taken
during day 1 of the single bolus experiment were also assayed for Cr(VI) according to EPA
method 218.6, using ion chromatographic separation with post-column color development and
spectroscopic detection, with a detection limit of 1–2 µg Cr(VI)/L. All samples were found to
contain no traces of Cr(VI) at any time point, including during the rapid uptake phase. These
observations are consistent with the interpretations of Kerger et al. (1996) regarding the
reduction, uptake, distribution and elimination of Cr(VI) – that Cr(VI) is reduced to Cr(III) with
multiple doses at 5 and 10 mg Cr(VI)/L. Additionally, the data were considered consistent with
published kinetic models of Cr(III) behavior in animals (Aitio et al., 1988; Lim et al., 1983).
Aitio et al. (1988) developed a compartmental model of the half-life of Cr(III) in humans that
was based on distribution and elimination rates from three separate compartments: the fast-
elimation (T1/2 = 7 hour), the moderate elimination (T1/2 = 15 days), and slow elimination (T1/2 =
3 years) compartments. Lim et al. (1983) described a similar compartmental model, suggesting
Cr(III) half-lives of 0.5–12 hour in blood (fast compartment), 1–14 days in storage organs such
as the liver and spleen (medium compartment), and 3 – 12 months in other solid tissues (slow
compartment).
Finley et al. (1996) evaluated urinary chromium clearance in six healthy, adult (ages 25–
39 years; four males weighing 79–97 kg, two females weighing 56–62 kg) human volunteers.
The entire chromium ingestion and urine collection period covered 18 days and was divided into
five separate, but consecutive, phases as described below:
Days 1 – 7: Chromium picolinate ingestion (loading dose)

Days 8 – 10: Cr(VI) ingestion

Days 11 – 13: No-dose period
Days 14 – 16: Cr(III) ingestion
Days 17 – 18: Post-dose period
The dosing regimen was designed to achieve a steady-state urine concentration during dosing
and to ensure a return to baseline between dosing with Cr(VI) and Cr(III). Chromium picolinate
was delivered at a dose of 0.2 mg for the first seven days, considered a loading dose for this
dietary supplement. Cr(VI) was delivered as potassium dichromate (K2Cr2O4) on days 8 through
10 in a capsule at a dose of 0.005 mg Cr(VI), the U.S. EPA RfD for Cr(VI) at that time. Cr(III)
was then ingested as chromic oxide (Cr2O3) at 1.0 mg Cr(III)/kg/day, the U.S. EPA RfD level for
Cr(III) at that time, on days 14 through 16. Urine samples and measurements were performed as
previously described for Kerger et al. (1996, 1997). The ingestion of chromium picolinate
resulted in significantly elevated urine concentrations such that participants routinely exceeded
background. Ingestion of Cr(VI) yielded individual mean total urinary chromium levels of 1.2–
23 µg/L, and a pooled mean value of 2.4 µg/L. Ingestion of the Cr(III) compound yielded no
significant increases in urinary chromium concentrations, suggesting negligible absorption.
Paustenbach et al. (1996) evaluated uptake and elimination in a male, Caucasian
volunteer (age 44 years) who ingested deionized water containing potassium dichromate
(K2Cr2O4) in five daily portions (400 mL, 2 mg Cr(VI)/L each) for 17 days. Methods described
for the Kerger et al. (1996) study were used to sample and quantify chromium in blood and
urine. Bioavailability was estimated at 2% and the plasma elimination half-life at 36 hour, both
consistent with the previous studies. The time to achieve steady-state concentrations in urine and
blood was 7 days. Both plasma and RBC chromium concentrations returned rapidly to
background levels within a few days, again suggesting that concentrations of 10 mg Cr(VI)/L or
less in drinking water of humans appears to be completely reduced to Cr(III) prior to systemic
distribution.
Finley et al. (1997) extended this work by evaluating chromium kinetics in human
volunteers following repeated oral exposure to Cr(VI) concentrations ranging from 0.1 to 10.0
mg/L. Five healthy, adult (age 30 to 54 years), male Caucasian volunteers ingested a liter (in
three volumes of 333 ml at approximate 6-hour intervals) of deionized water containing Cr(VI)
concentrations of 0.1, 0.5, 1.0, 5.0 and 10.0 mg Cr(VI)/L. Potassium chromate (K2CrO4) was
used as the source of soluble Cr(VI). Other experimental methods and total chrome analyses in
the samples were as in Kerger et al. (1996); a study of the stability of the dosing solution
confirmed no measurable reduction of Cr(VI) occurred throughout the study. Each of the five
subjects demonstrated an increase in the amount of urinary chromium excreted, ranging from a
mean of 1.7% at the low dose (0.1 mg Cr(VI)/L) to 3.5% at the high dose (10.0 mg Cr(VI)/L). A
dose-related increase in plasma chromium began at the 5 mg Cr(VI)/day dose, with 2 subjects
not clearly increasing above baseline at either 5 or 10 mg Cr(VI)/day. The RBC chromium

profiles generally mirrored those in plasma. The RBC chromium levels began to decrease within
days following cessation of the 10 mg Cr(VI)/L dose and exhibited a rapid and then slow decline
which were also similar to that for plasma, with an approximate 50% decrease within 7 days post
exposure.
3.3. BIOCHEMISTRY OF INTRACELLULAR HEXAVALENT CHROMIUM

At physiological pH, hexavalent chromium exists in the form of an oxyanion with an
overall -2 charge that resembles sulfate and phosphate, and as such, this anion is taken up by
cells. This transport system, along with intracellular reduction reactions, allows the anion to
accumulate in cells at much higher concentrations than extracellularly. Once hexavalent
chromium enters the cell, it is reduced by various intracellular reductants, including ascorbate,
glutathione, and cysteine. Low molecular weight thiols (glutathione and cysteine) and ascorbate
are believed to be primarily responsible for the intracellular reduction (Suzuki and Fukuda, 1990;
Standeven et al., 1991, 1992; Quivryn et al., 2001). Studies on the reduction of Cr(VI) by
extracts of rat liver, lung, or kidney have found that ascorbate accounted for at least 80% of
Cr(VI) metabolism in these tissues (Standeven et al., 1991,1992). Ascorbate is also the fastest
reducer in the in vitro reactions, and its rate of reduction at 1 mM exceeds that of cysteine and
glutathione by approximately 13 and 61 times, respectively (Zhitkovich, 2005; Quivryn et al.,
2001). This intracellular reduction yields reactive intermediates, chromium(V) and
chromium(IV). These reactive intermediates are formed along with oxygen radicals generated
via Fenton-like and other possible reactions that occur during intracellular reduction. Depending
on the nature of the reducing agent and its concentration, this process can generate various
amounts of unstable Cr(V) and Cr(IV) intermediates (Stearns et al., 1994).
Hexavalent chromium taken up by RBCs undergoes reduction to the trivalent form and
complexes with Hgb and other intracellular proteins that are sufficiently stable to retain
chromium for a substantial fraction of the RBC lifetime. GSH appears to dominate the reduction
of hexavalent chromium within RBCs (Wiegand et al., 1984). In RBC suspensions, the addition
of GSH results in intracellular reduction of hexavalent chromium to trivalent chromium. The
role of GSH was confirmed by decreased chromium binding (from 100 to 40%) in the RBCs
following pretreatment with diethylmaleate, a GSH depletion agent (Aaseth et al., 1982).
Incubation of human RBCs with K251Cr2O7 resulted in depletion of the RBC GSH content to
about 10% of normal. Subsequent analysis of RBC lysates suggested that chromium-GSH
complexes are formed and that approximately 97% of [51Cr] is bound to Hgb (Wiegand et al.,
1984). Excess trivalent chromium in the RBC is sequestered until cell death (Kerger et al., 1997;
Aaseth et al., 1982). Over time, the RBC-associated chromium appears to be transferred to the
spleen as a result of scavenging of aging RBCs from the blood. Trivalent chromium in plasma
does not readily diffuse into RBCs. This explains the observation of lower chromium plasma to
RBC ratios following exposure to hexavalent chromium.

Within parenchymal and phagocytic cells, hexavalent chromium may be reduced in the
cytosolic and microsomal compartments (De Flora and Wetterhahn, 1989). Isolated liver
perfusion in rats suggests that the majority of hexavalent chromium reduction is cytosolic, as 60,
14, 9, and 2% of [51Cr] activity was found in the cytosolic, mitochondrial, microsomal, and
nuclear fraction, respectively (Wiegand et al., 1987). Caution should be used in interpreting cell
culture data, as the cell culture medium could play a role in hexavalent chromium reduction,
confounding the extent of intracellular hexavalent chromium reduction. For example, Dulbeco’s
Modified Eagle’s Medium reduces hexavalent chromium to chromium(V) in the absence of cells
(Borthiry et al., 2008). In human bronchial epithelial cells (BEAS-2B), Na2CrO4, and to a lesser
extent, insoluble Zn2CrO4, were reduced to two reactive chromium(V) species; one appeared to
be mediated by a thiol-independent NADP(H) reductase, and the other possibly via a hexavalent
chromium-GSH intermediate (Borthiry et al., 2008). Electron paramagnetic resonance (EPR)
studies of hexavalent chromium reacting with GSH revealed the generation of two reactive
chromium(V) intermediates and a GSH thiyl radical (Aiyar et al., 1991). Pulmonary alveolar
macrophages (PAMs) also reduce hexavalent chromium via an NADP(H)-dependent reductase
and GSH (Petrilli et al., 1986). PAMs in smokers had approximately twice the reductive ability
than cells from nonsmokers, ostensibly due to reductase induction by cigarette smoke (Petrilli et
al., 1986).
The predominant mechanism for intracellular hexavalent chromium reduction via
microsomal enzymes has been extensively described. Incubation of K2Cr2O7 with rat liver
microsomes or NADP(H) alone resulted in very little hexavalent chromium reduction (Jennette,
1982; Gruber and Jennette, 1978). However, incubation with microsomes and NADP(H)
resulted in essentially complete disappearance of hexavalent chromium. Within seconds,
hexavalent chromium (as K2C2O7) incubated with rat liver microsomes and NADP(H) was
reduced to chromium(V), presumably via one-electron transfer from cytochrome P450 (Jennette,
1982).
In contrast to rat liver microsomes, human lung and liver microsomes do not reduce
hexavalent chromium via cytochrome P450. Pratt and Myers (1993) showed that human liver
and lung microsomes reduced hexavalent chromium via an NADP(H) reductase-dependent
system that was not perturbed by the addition of five different P450 inhibitors. The system was,
however, inhibited by the addition of TlCl3, indicating the involvement of flavoproteins,
specifically cytochrome c reductase. The Vmax and Michaelis-Menten constant (Km) values for
liver microsomal reduction of hexavalent chromium were 5.03 nmol/minute/mg protein and
1.04 mM, respectively. The human microsomal Km was 1–3 orders of magnitude lower than
those measured in rat liver microsomes (16–34 μM [Mikalsen et al., 1989] to 1.6 mM [Garcia
and Jenette, 1981]). Another striking difference between rat and human hexavalent chromium
microsomal reduction is the relative insensitivity to O2 in human microsomes (Pratt and Myers,
1993). While rat microsomal hexavalent chromium reduction was markedly inhibited in the

presence of 0.1% O2, human microsomal reduction was diminished by only 34–56% in the
presence of ambient (21%) O2. These results suggest two things about the spatial distribution of
microsomal hexavalent chromium reduction in rats and humans. First, P450-dependent
hexavalent chromium reduction is likely to be confined to the centrilobular region of the rat liver,
since an O2 tension of only 1 mm Hg exists there. Secondly, the insensitivity to O2 of human
microsomes makes it possible for enzymatic reduction to occur in highly aerated tissues, such as
the lung.
Myers and Myers (1998) verified and extended the description of enzymatic hexavalent
chromium reduction in human liver microsomes. Liver microsomes from five individuals were
incubated with Na2CrO4 to determine reduction kinetics. Using a series of P450 inhibitors and
TlCl3, the authors showed that hexavalent chromium reduction was mediated by flavoproteins,
NADP(H)-dependent P450 reductase, and cytochrome b5. Parameters for reduction kinetics in
these five individuals are shown in Table 3-11. The range of Vmax and Km values was very
similar across subjects. Lung microsomes from one individual exhibited Vmax and Km values that
were 0.66- and 2.8-fold lower than liver microsome values. Finally, the addition of iron to the
liver microsomal system revealed that hexavalent chromium reduction could be stimulated by
iron levels that were 3- to 26-fold lower than the hexavalent chromium levels, suggesting that the
iron may have a catalytic role in the enzymatic reduction of hexavalent chromium.
Table 3-11. Kinetic parameters of hexavalent chromium reduction in

human liver microsomes from five individuals
Parameter Observation
Vmax 10.4–10.7
Km 1.04–1.68
Inhibition by O2 26–37%
Inhibition by TlCl3 96–100%
Inhibition by P450 inhibitors
Carbon monoxide None
Piperonyl butoxide None
Aminopyrine None
Source: Myers and Myers (1998).
Proteoliposomes composed of recombinant human P450 reductase and cytochrome b5

were used to verify that electrons from NADP(H) could be transferred to cytochrome b5 during
the reduction of hexavalent chromium (Jannetto et al., 2001). Markedly less hexavalent
chromium reduction occurred in proteoliposomes devoid of cytochrome b5. Further, hexavalent
chromium reduction in proteoliposomes was almost identical to human liver microsomes when
corrected for the cytochrome b5 concentration.

The available data in human and animal studies do not suggest a significant gender
difference in metabolism of hexavalent chromium. Further, human liver microsome studies did
not identify marked variability in enzymatic rates of hexavalent chromium reduction (Myers and
Myers, 1998), although samples were examined from a small number of individuals.
3.4. TOXICOKINETIC CONSIDERATIONS FOR THE MODE OF ACTION OF

INGESTED HEXAVALENT CHROMIUM
Consideration of the bioaccessibility, bioavailability, and biochemistry of ingested Cr(VI)
are critically important as key toxicokinetic determinants of its mode of action (MOA). Figure
3-6 provides a schematic of the salient features of these processes, and the ensuing text discusses
their role in arriving at inferences for the MOA.
As discussed in Sections 2 and 3.1, external reduction in the environment, or
bioaccessibility, is a critical factor determining the location, amount, and speciation of a given
exposure. The type of compound ingested plays one of the most important roles dictating
subsequent toxicity, having a dramatic effect on the digestion, solubilization and speciation of
chromium compounds; thus it is a major influence on its internal bioaccessibility, biovailability
and toxicokinetics in the body.
Internal bioaccessibility involves factors dictating the extracellular reduction in the GI
tract lumen. Physiological fluids in the GI lumen such as gastric juice, and constituents of the
diet such as beverages like orange juice, both diminish the uptake of Cr(VI) compounds via their
capacities to provide reduction of Cr(VI) to Cr(III). Further, the disposition of chromium in
either the trivalent or hexavalent form is strongly dependent on the both the chemical
characteristics as well as on the solubility of the chromium compound and its method of
administration. Intraindividual variability due to differences in this reduction capacity have been
noted in human studies (Kerger et al., 1996; Finley et al., 1997; Paustenbach et al., 1996;
O’Flaherty et al., 2001).
The accuracy of any dose-response analysis would be improved by greater rigor afforded
the characterization of the reduction capacities, and this may be especially important for
interspecies extrapolation. This will likely need to involve a more physiologically-based
description of GI uptake as reduction is a function of physiological factors affecting lag time
(peristalsis) and spatial distribution in the GI tract. Environmental and internal bioaccessibility
processes are both extracellular, in which reduction of Cr(VI) to Cr(III) can occur with the
resulting Cr(III) not being able to be actively transported into cells (Levine et al., 2009). Any
comprehensive risk modeling must take into account the role of these processes associated with
extracellular reduction of Cr(VI) (Zhitkovich, 2005).
Hexavalent chromium that is not extracellularly reduced is readily transported by sulfate /
phosphate anion channels into the GI lumen and due to its reactivity, undergoes
biotransformation. As discussed in Section 3.2, Cr(VI) distributes to other tissues, notably the

blood, kidney and liver. The uptake and intracellular reactions depicted and described here for
the GI tissue are also applicable to those cells as well.
As described in Section 3.3, Cr(VI) once in the cell, is reduced by various intracellular
reductants, including ascorbate, glutathione, and cysteine, lipoic acid, hydrogen peroxide,
NAD(P)H-dependent enzymes, fructose and ribose (LeVina et al., 2003, 2007; Nickens et al.,
2010). The end product of Cr(VI) intracellular reduction in all biological systems is always
Cr(III), but the reduction process can also generate variable amounts of Cr(V), Cr(IV), and
organic radicals depending on the nature of the reducing agent, its concentration, and the ratio of
reactants (Stearns et al., 1994; Salnikov and Zhitkovich, 2009). The resultant Cr(III) from the
intracellular reduction of Cr(VI) forms stable adducts with macromolecules and other cellular
constituents.
While the presence of small amounts of short-lived Cr(V) at higher ratios of ascorbate to
Cr(VI) can not be unequivocally ruled out due to the technical limitations of the analytical
methods employed (electron spin resonance spectroscopy), it is doubtful that environmental
levels of Cr(VI) will be sufficient to produce significant quantities of Cr(V) in cells with
millimolar ascorbate concentrations (Salnikov and Zhitkovich, 2009). Thus, the reduction
reactions to Cr(III)-ligands are depicted as the dominant (solid line) in Figure 3-6, as is the
subsequent formation of Cr(III)-DNA adducts, as they have been determined to be mutagenic by
the various pathways depicted (as discussed more fully in Section 4). The contribution of
oxidative stress and single-strand DNA breaks are depicted to occur at high doses, but the
quantitative magnitude for designating “low” versus “high” cannot be established based on the
available laboratory animal data.
An additional important note on these biotransformations regards the interpretation and
reliability of data from in vitro assays. In vivo, the intracellular levels of ascorbate are quite high
(about 1 mM). In contrast, the levels of ascorbate in tissue culture media are quite low since
generally it is not added to the media so that the only source is supplemented fetal bovine serum
(FBS). With 10% FBS, the level of ascorbate in tissue cultured cells is only about 50 µM which
is 20 times lower than that which is found in vivo (Zhitkovich, 2005). Therefore, experiments on
mutagenesis and other toxic effects of hexavalent chromium in tissue culture may underestimate
its mutagenic, genotoxic, and cell-transforming activities (Zhitkovich, 2005; Costa and Klein,
2006). Attempts to address this concern by delivering ascorbic acid into cultured cells introduce
other difficulties because the oxidized dihydro form must be added in order for it to enter the cell
(Zhitkovich, 2005).

Source: Adapted from Beyersmann and Hartwig (2008), Salnikow and Zhitkovich (2008), and
Holmes (2008).
Figure 3-6. Schematic of ingested Cr(VI) to internal dose in GI tissue and blood.
Intracellular molecular mechanisms of biological disposition in GI tissue are expected in
other target tissues such as respiratory tract, liver, and kidney. See text for discussion.

3.5. BIOLOGICALLY-BASED MODEL OF INGESTED CHROMIUM COMPOUNDS
IN RATS AND HUMANS
Physiologically based pharmacokinetic (PBPK) models are mathematical representations
of biological systems in animals and humans that are relevant to the quantitative determination of
internal doses of toxic moieties of xenobiotics resulting from external doses or exposures and
thereby facilitate interspecies extrapolation (Krishnan et al., 1994). By employing chemical- and
species-specific parameter values for tissue volumes, process rates, and reaction kinetics, PBPK
models are used to extrapolate internal dosimetry of chemicals across routes of exposure, dose
ranges, and species. In risk assessment, the use of PBPK models quantitatively reduces
uncertainties in these extrapolations, thus partially or completely obviating the need to apply
uncertainty factors (UFs) in the derivation of exposure limits protective of cancer and noncancer
effects (Clewell and Andersen, 1985).
The development of PBPK models occurs in four sequential steps: (1) conceptual
representation of the body into discrete compartments, (2) parameterization of the model,
(3) exercise of the model by simulating one or more exposures and comparing model predictions
against empirical observations, and (4) verification of the ability of the model to adequately
predict empirical data not used for model exercising (Krishnan and Andersen, 1994). Recent
regulatory applications have extended these concepts to use a family approach to evaluate hazard
and arrive at risk estimates using a four-step framework for organizing and evaluating toxicity
data: 1) exposure, 2) tissue dosimetry, 3) mode of action, and 4) response (Barton et al., 2000).
This expansion of the traditional exposure-response analysis has been increasingly utilized in
risk assessment and represents advancement for maximal use of designing experiments and
maximal end use of toxicity data. The kinetics of a group of metabolically related compounds,
i.e., a family, can often be described by development of a template model structure that may only
need some refinement in specific parameters to be able to address specific members of that
family. The development of the chromium model actually represents such a process and
illustrates a distinct advantage of PBPK models, i.e., they can be adaptable to different
conditions (e.g., across routes, species, and for both chromium compounds) and provide greater
confidence that the physiological basis of the processes are adequately understood and accurately
described. This section will briefly describe the development and features of the model
described in greater detail elsewhere (O’Flaherty and Radike, 1991; O’Flaherty 1991a,b;
O’Flaherty, 1993, 1995, 1996; O’Flaherty et al., 2001), and discuss its potential utility for
internal dose descriptions for the assessment of ingested hexavalent chromium.
O’Flaherty published a model for chromium kinetics (both Cr(III) and Cr(VI)) in rats
(O’Flaherty, 1996) and humans (O’Flaherty et al., 2001) that was based on a general structure
that had been developed for lead kinetics in rats (O’Flaherty, 1991a,b) and humans (O’Flaherty
1993, 1995). Once developed to describe the kinetics of chromium in rats, the model structure
was then extended to describe the kinetics of humans using the rat chromium model and the

human lead model. Like lead, chromium exchanges between plasma and the bone surfaces in
contact with plasma, and also like lead, although with much lower efficiency, chromium can
become incorporated into actively mineralizing bone. Both processes are included in the model.
Parallel absorption and disposition schemes for Cr(III) and Cr(VI) are linked in the model by
reduction processes occurring throughout the body as well as in the GI tract and in the lungs as
points of entry.
Development of the rat model for chromium began with general physiological
parameters, as well as those parameters related to body growth and to bone growth and other
tissues and organs, as previously defined and assigned in the lead rat model (O’Flaherty,
1991a,b). It was adapted to chromium by first considering the disposition of Cr(III) following an
intravenous administration and then introducing other exposure routes in increasing order of
kinetic complexity. Incorporation of Cr(VI) kinetics followed a similar strategy. Initial values
for each added set of exposure, reduction, or distribution parameters were estimated based on
literature data, and model simulations were visually optimized to the appropriate data set at each
step in the development process. New model parameters specific to chromium are as follows:
rate constant for movement past the GI absorption region, the GI absorption rate constants for
both Cr(III) and Cr(VI), clearance constants for passage into and out of tissues including the
RBC, the fractional rates of depositon with forming bone, rate constants for the reduction of
Cr(VI) to Cr(III) in the GI tract and in tissues, and concentration-dependent urinary clearance
consistent with parallel renal processes.
The steps listed below summarize the strategy for development of the rat model and
provide an appreciation for the complex considerations of study design and biological processes
that were necessary to arrive at the final model structure:
1. General model structure taken from the existing PBPK model for lead (O’Flaherty,
1991a,b) with exclusion of any slow exchange in bone compartment. Relative
magnitudes of rapid surface exchange at bone surfaces and formation/resorption of bone
in juvenile and mature rats based on visual fit of model to data of Hopkins (1965), a study
in which Cr51Cl3 was administered by i.v. and radiolabel monitored for 72 days following
injection. The declining body burden data were fit with a three-term sum of
exponentials; the third term was presumed to be most closely related to loss of chromium
from bone, with bone data reported at 0.25, 4.0 and 24 hours.
2. Extension of time frame of model predictions. Starting value for plasma Cr(III) clearance
estimated from chromium body burden data of Hopkins (1965), the second term in that
data (described in Step 1) corresponding to a half-life of 5.9 days, results in a clearance
rate of 0.025 L Cr/day. Scaling by body weight0.75 resulted in whole-body clearance of
Cr(III) of 0.055 L Cr/day/kg. This value was later calibrated to be 0.065 to be more
consistent with other in vivo data sets after drinking water exposure (e.g., MacKenzie et
al., 1959).

3. Refinement of Cr(III) distribution parameters by comparison with data of Visek et al.
(1953) on Cr(III) i.v. administration reported at 4 and 42 days.
4. Total clearance of Cr(III) was fractionated into clearances into bile, urine, and GI tract as
1%, 90% of the remainder, and 10% of the remainder, respectively; primarily based on
the data of Cikrt and Bencko (1979) who administered Cr(III) salt by i.v. Further work
on the model showed that the data of Bragt and Van Dura (1983) and Edel and Sabbioni
(1985) were not compatible with significant transintestinal and biliary excretion, so these
two fractions were subsequently set at 0 in final form.
5. Addition of RBC compartment in communication with arterial blood. The first-order rate
constants for loss of Cr(III) from RBC to plasma and transfer of Cr(III) from plasma to
the RBC were fixed on the basis of measured half-life of chromium association with RBC
(Bishop and Surgenor, 1964). These were not changed in further model development.
Estimates of the rate constant for transfer of Cr(VI) from plasma to RBC based on in
vitro data on human RBC suspended in either saline (Gray and Sterling, 1950) or blood
plasma (Weigand et al., 1985). The same half-life was initially used for transfer of
Cr(VI) between plasma and peripheral tissues, but these were changed for poorly
perfused tissues in order to fit the data of Weber (1983).
6. Total excretion clearance for Cr(VI) set and changed as above in Step 4 for Cr(III).
7. Percent of chromium dose excreted in the urine in 24 hour following i.v. injection of a
soluble salt of either Cr(III) or Cr(VI) in rats was reported by Cikrt and Bencko (1979) to
be independent of oxidation state of administered compound. Initial estimate of
excretion clearance of Cr(VI) was set equal to Cr(III) and the two values remained equal
in the course of model development.
8. Link of Cr(VI) model to Cr(III) model by reduction processes in all tissues. A single
value of the first-order reduction rate constant was provisionally assigned to the Cr(VI)
pools in the RBCs, peripheral tissues, and lung. This simplification proved satisfactory
and was retained in the final model. The value of the reduction rate constant was
determined by fitting the tissue concentration data of Weber (1983) in accordance with
the results of studies in which little or no reduction was observed in human plasma in
vitro (Gray and Sterling, 1950; Korallus et al., 1984).
9 Expansion of i.v. model to include GI uptake. First-order rate constant set at 0.01 per day
for GI absorption of Cr(III) estimated from the single bolus dose data of MacKenzie et al.
(1959) at 1.8%, shown to be in agreement with the estimate of Mertz et al. (1965) at 2–
3% for oral administration. The same study found 85% of an orally administered soluble
Cr(VI) salt had been reduced before it could reach the intestinal absorption site. Model
was run with inclusion of reduction pathway and the first-order reaction rate constant set
to give 85% reduction and 10% absorption.
10. Final structural development step was to expand the composite model to allow absorption
and elimination of chromium from the lung. Pulmonary clearance of Cr(VI) salts is not
dose-dependent within a reasonable dose range (Weber, 1983; Bragt and van Dura,
1983). Mucociliary transfer to the GI tract identified as second route of chromium
clearance. Both Cr(III) and Cr(VI) assigned to two lung pools in the model. Chromium

from either intratracheal (i.t.) or inhalation exposures enters the first pool, from which it
can be systemically absorbed, transferred to the second pool, or cleared by mucociliary
action. Chromium in the second pool can be only cleared by mucociliary transport.
11. Adjustment of initial model parameter values by visual optimization to fit data from
O’Flaherty and Radike (1991). Published parameter estimates, especially for tissue
uptake and loss were refined by optimization of model predictions to the data set of
Weber (1983) in conjunction with the data sets of Bragt and Van Dura (1983) and Edel
and Sabbioni (1985). All were studies of radiolabeled chromium following i.t.
administration of soluble 51Cr(VI), or in one instance 51Cr(III), salts. The Weber (1983)
data set consisted of time course data in several tissues, whole blood, plasma, GI tract,
and carcass for 42 days after a single i.t. dose. The Bragt and Van Dura and Edel and
Sabbioni data sets had only limited tissue concentration measurements, but extensive
measurements of cumulative excretion in urine and feces.
To test the generalizeability of the final form of the model, the inhalation study of
Langard et al. (1978) was simulated. The study of Langard et al. (1978) consisted of a series of
inhalation exposures of rats to zinc chromate dust, 6 hour/day for 4 days followed by 4 days
during which urinary excretion was monitored and 37 days during which blood chromium levels
were monitored. Agreement of the O’Flaherty model against these data was reasonably good
despite the route extrapolation required. However, the model only fit the data of the MacKenzie
et al. (1958) drinking water study moderately well, as the nonlinear kidney and liver
oncentrations were over predicted. Modification of the uptake description to include a
Michaelis-Menton term may address this issue (see below).
The human model (O’Flaherty, 2001) was developed based on the rat using appropriate
scaling of physiological parameters and by estimating specific chromium-related parameters
using several studies in adult human volunteers administered chromium in drinking water
(described in Section 3.2 [Finley et al., 1997; Kerger et al., 1996; Paustenbach et al., 1996]).
Default values, determined by gender and age, were used for their body weights. The studies of
Kerger et al. (1996) and Finley et al. (1997) were used for calibration of the chromium-specific
parameters (e.g., clearance constants) in the model structure and the Paustenbach et al. (1996)
study was used for verification. In the absence of data on the magnitude of the rate of deposition
of Cr(III) or Cr(VI) in human bone, the fractional rates of deposition were assigned the values of
5 for Cr(III) and 15 for Cr(VI), the same values used in the rat model.
The human model generated reasonable time profiles to the data of Paustenbach et al.
(1996) despite the variability of the urinary excretion rates. Figure 3-7 illustrates the dependence
of the urinary excretion on the form of the chromium administered. As discussed for the data of
Kerger et al. (1996), when chromium has been administered as Cr(III) citrate rather than as the
inorganic salts of CrCl3 or K2Cr2O7, transfer of chromium from the blood to a compartment from
which excretion occurs is favored relative to other tissues. As a result, when the CrCl3 clearance
curve is applied and the excretion rates were optimized to fit urinary excretion rates, plasma

concentrations in that same subject were greatly over-predicted; if the model had instead been
optimized to plasma concentrations, urinary excretion rates would have been underpredicted.
O’Flaherty revised the clearance expression to address the administration in orange juice. The
alternate expression gives clearances of 2 L/day at chromium concentrations within the
background concentration range of 0.05–0.15 µg/L and is represented by the dotted line in
Figure 3-7. The fit to the data for the model prediction of the kinetic behavior of chromium in
the plasma for the Cr(VI)-OJ salt was greatly improved.

Source: O’Flaherty et al. (2001).
Figure 3-7. Observed and simulated plasma chromium concentrations predicted by

the PBPK model for a human subject ingesting Cr(VI) dissolved in orange juice
(CrVI-OJ) in the study of Kerger et al. (1996). The solid line is the simulation using
the general model expression for urinary clearance. The dotted line represents a
simulation using an alternate expression for that clearance (see text).

Another key finding of the model simulations was that based on total urinary excretion, a
consistently greater percentage of Cr(VI) than of Cr(III) is absorbed. This implies that some
Cr(VI) escaped reduction in the stomach and entered portal venous blood. This greater
absorption of Cr(VI) versus Cr(III) does not imply that the reduction capacity of gastric juice
was exceeded, but rather that absorption from the GI tract is so rapid that it is able to compete
effectively with reduction in the stomach (O’Flaherty, 1996). The rapidity of absorption is
further supported by the rate at which chromium appears in the blood.
The model schematic is provided in Figure 3-8 and the key parameters for both the rat
and human models are provided in Table 3-12. The model accounts for both oral and inhalation
exposures, plasma and RBC distribution, distribution in key target tissues such as liver and
kidney, and elimination via urine and feces. Additional target tissues could be defined from the
well- and poorly-perfused tissues if required; likewise, a more complex respiratory or GI tract
description could be incorporated as needed (see below).
As with all models of this type, certain parameter values are highly correlated. For
several constellations of constants that define parallel and competing processes, the relative
values of the constants are more important than are their absolute values in order to achieve a
good fit to the concentration data. These groups of constants include those determining the
simultaneous rapid excretion of Cr(VI) and its uptake into tissue, those determining loss of
Cr(III) from tissues, and those determining the relative rates of Cr(VI) reduction in the RBC and
release from the RBC. Uptake of Cr(III) into tissues has relatively little impact on the
predictions of the model (O’Flaherty, 1996).

POOL B
INHALATION
LUNG
EXPOSURE
POOL A
Cr(VI)→Cr(III)
WELL-PERFUSED
Cr(VI)→Cr(III)
POORLY-PERFUSED
Cr(VI)→Cr(III)
Cr(VI) →Cr(III)
RED CELLS
PLASMA
PLASMA
BONE
Cr(VI)→Cr(III)
ORAL
LIVER GI TRACT EXPOSURE
Cr(VI)→Cr(III) Cr(VI)→Cr(III)
KIDNEY
Cr(VI)→Cr(III)
FECAL EXCRETION
URINARY
EXCRETION
URINARY PATH
Source: O’Flaherty (1996).
Figure 3-8. Schematic of structure for PBPK model of chromium in the rat
and human.

Table 3-12. Chemical-specific parameters in the rat and human chromium
models
Rat Human
a
Parameter Cr(III) Cr(VI) Cr(III) Cr(VI) Definition
Absorption
KGI 0.01 0.04 0.25 2.5 First-order rate constant for absorption from the GI tract (Da-1)
KLU 0.2 2.0 NA NA First-order rate constant for absorption from the bioavailable
lung pool (pool A) (Da-1)
KMUCOA 0.8 0.8 NA NA First-order rate constant for mucociliary clearance from pool
A to the GI tract (Da-1)
KMUCOB 0.025 0.025 NA NA First-order rate constant for mucociliary clearance from the
nonbioavailable lung pool (pool B) to the GI tract (Da-1)
KLUAB 1.2 1.2 NA NA First-order rate constant for transfer from pool A to pool B
(Da-1)
FRLUNG NA NA 0.3 0.3 Fraction of inhaled chromium absorbed to blood
FRTRGI NA NA 0.7 0.7 Fraction of inhaled chromium transferred to GI tract
Distribution
b b
CR 5.0 15.0 NA NA Relative clearance of chromium into mineralizing bone
(L blood plasma cleared/L new bone formed)
KINRBC 0.0003 1.5 12.0 NA Clearance from plasma to red cell (L/Da)
KDIN 0.007 1.5 3.0 30.0 Clearance from plasma to kidney (L/Da)
LDIN 0.0001 1.5 3.0 30.0 Clearance from plasma to liver (L/Da)
WDIN 0.0001 1.5 3.0 30.0 Clearance from plasma to other well-perfused tissues (L/Da)
PDIN 0.0001 0.01 3.0 30.0 Clearance from plasma to poorly-perfused tissues (L/Da)
BDIN 0.0001 0.01 NAb NAb Clearance from plasma to bone (L/Da)
CR NA NA 5.0 15.0 Fraction deposition from blood to forming bone
KOUTRBC 0.0003 10.0 12.0 NA Clearance from red cell to plasma (L/Da)
KDOUT 0.001 10.0 3.0 30.0 Clearance from kidney to plasma (L/Da)
LDOUT 0.0003 10.0 3.0 30.0 Clearance from liver to plasma (L/Da)
WDOUT 0.001 10.0 3.0 30.0 Clearance from other well-perfused tissues to plasma (L/Da)
PDOUT 0.003 10.0 3.0 30.0 Clearance from poorly perfused tissues to plasma (L/Da)
BDOUT 0.003 10.0 NAb NAb Clearance from bone to plasma (L/Da)
Excretion
KFX 1.5 1.5 14.0 14.0 First-order rate constant for loss of chromium from intestinal
tract contents to the feces (Da-1)
QEC 0.065 0.065 NAc NAc Excretion clearance from the plasma (urinary clearance)
(L/kg/Da)
CLEARc NA NA 12.0 12.0 Parameter in expression for clearance from blood plasma to
urine (L/d)
MAXc NA NA 0.008 0.008 Parameter in expression for clearance from blood plasma to
urine (mg/d)
KMc NA NA 0.0008 0.0008 Parameter in expression for clearance from blood plasma to
urine (mg/L)
FB 0.0 0.0 NA NA Fraction of body burden secreted in the bile
FI 0.0 0.0 NA NA Fraction of body burden excreted via the GI tract

Table 3-12. Chemical-specific parameters in the rat and human chromium
models
Rat Human
a
Parameter Cr(III) Cr(VI) Cr(III) Cr(VI) Definition
Reduction
KREDRC NA 0.7 NA 7.0 First-order rate constant for reduction of hexavalent chromium
to Cr(III) in the red cell (Da-1)
KREDBP NA NA NA 0.2 First-order rate constant for reduction of hexavalent chromium
to Cr(III) in blood plasma (Da-1)
KREDKL NA NA NA 500.0 First-order rate constant for reduction of hexavalent chromium
to Cr(III) in kidney (Da-1)
KREDGI NA 10.0 NA 100.0 First-order rate constant for reduction of hexavalent chromium
to Cr(III) in GI tract contents (Da-1)
KRED NA 0.5 NA 5.0 First-order rate constant for reduction of hexavalent chromium
to Cr(III) in all other tissues and in lung contents (Da-1)
Lag time for excretion of urine
FRHOLD 0.7 0.7 NA NA Fraction of urinary chromium not excreted immediately; that
is, temporarily held in pool
KHOLD 0.05 0.05 NA NA First-order rate constant for excretion from the retained urine
pool (Da-1)
FR 0.10 0.10 NA NA Fraction of chromium in retained urine that is associated with
the kidney
a
Parameter names are those for human model in cases where the reported rat and human parameter names were not
identical.
b
Exchanges between blood plasma and cortical and trabecular bone are simulated as functions of bone formation
and resorption rates.
c
MAX
QE = CLEAR −
KM + CBP , where QE is clearance from blood plasma to urine (L/d) and CBP is plasma
concentration of chromium (mg/L).
NA = not applicable
Sources: O’Flaherty (1996) (rat parameters); O’Flaherty et al. (2001) (human parameters).

The model behavior in several respects suggests that it reflects critical physiological
behavior with reasonable accuracy. For example, although uptake of Cr(VI) into tissues is
clearly more rapid than that of Cr(III), it is not possible to fit Cr(VI) clearance disposition data
satisfactorily using an excretion clearance greater than that assigned to Cr(III). That is, the
mechanism of tissue uptake is fundamentally different than that for excretion, which would be
expected if one was carrier-mediated and the other not. Additionally, the magnitudes of the
observed tissue chromium concentrations require that a significant portion of tissue uptake be
chromium as Cr(VI), which is reduced and trapped as Cr(III), especially in the RBC. Cr(III)
would not have entered the RBC sufficiently rapidly to achieve the observed concentraions.
Further, the transient early chromium concentration peaks seen in certain tissues (e.g., liver)
would not be adequately described by assuming only that some Cr(VI) taken up by the tissue is
subsequently lost as Cr(VI).
As illustrated above, O’Flaherty (1996, 2001) noted for both models that a key
uncertainty in the models is that there is limited information regarding the factional absorption
from environmental sources. Since even the soluble salts of Cr(III) and Cr(VI) are not
particularly well absorbed from either the GI tract or respiratory tract, bioaccessibility of
chromium to absorption processes will prove to be the most important single characteristic
determining its potential absorption and toxicity. The urinary clearance observed in studies in
which inorganic Cr(VI) or Cr(III) were administered differ from the urinary clearance of Cr(III)
administered as a complex with citric acid anions. The latter are consistent with chromium
clearances reported in the general population. The disposition of chromium in either the trivalent
or hexavalent form was strongly dependent on both the chemical characteristics as well as on the
solubility of the chromium compound and its method of administration. Studies in which the
chromium compound was given by i.v., intramuscular (i.m.), or i.p. injection do not generate
data on which an understanding of the kinetics of chromium can be based and thus are not
suitable for setting chromium-exposure standards.
A second uncertainty noted by O’Flaherty (1996) was that little is known about the
importance of bone as a reservoir and continuing source of internal exposure. Dependence of
bone chromium uptake on age and physiologic status may be important features to complete any
comprehensive model application to chromium kinetics in populations relevant to risk
assessments.
Despite these limitations, the human model in its present form could be useable for the
generation of rough estimates with a urinary clearance set to a constant value of 1–2 L/day and
the rate constants of GI absorption set at 0.25/day for Cr(III) and 2.5/day for Cr(VI), i.e., 10
times the value of Cr(III). However, neither the rat nor human version of the model in its present
form has been subjected to formal computerized optimization of parameter values. As suggested
by O’Flaherty and Radike (1991), its use for predictive applications in risk assessment would be
greatly enhanced by both a sensitivity analysis and optimization of key parameters, with

attention to the variety, reliability, and utility of the kinetic data available. Newer data sets
published since the model was published may be particularly informative to this exercise,
including those of Kargacin et al. (1993) in mice and rats, Sutherland et al. (2000) in rats, and the
NTP (2008) bioassay in rats and mice.
Because of the toxicity and tumors observed in the buccal cavity of rats and the
duodenum in mice of the recent NTP 2-year oral bioassay with sodium dichromate dihydrate in
drinking water (NTP, 2008; Stout et al., 2009), new effort has been devoted to extending the
O’Flaherty model to predict kinetics in those tissues for those species (Campbell et al., 2009).
The O’Flaherty rat model was extrapolated to the mouse using known physiological parameter
values and allometric scaling of chromium kinetic parameters (Table 3-13).
Table 3-13. Parameters for mouse developed by Campbell et al. (2009)
Parameter Mouse Rat

QCC 16.5 15.0 Cardiac output (L/h/kg)
Fraction of cardiac output
QLC 0.02 0.25 Liver
QKC 0.17 0.17 Kidney
QIC 0.141 0.17 GI tract
QBC 0.03 0.03 Bone
QOCC 0.01 0.10 Oral cavity
Oral cavity
SAOC 2.5 5 Surface areas (cm2)
VLOC 1.2 2.4 Volumes (mL)
WLO 0.006 0.006 Lumen thickness (cm)
WTO 0.01 0.01 Tissue thickness (cm)
WXO 0.02 0.02 Submucosal thickness (cm)
FRACS 0.43 0.43 Saliva flow (fraction of ingestion rate)
Tissue volume (fraction of body weight)
VLC 0.055 0.04 Liver
VKC 0.0167 0.01 Kidney
VIC 0.0422 0.03 GI tract
VRBCC 0.03 0.03 RBC
VWC 0.2 0.2 Well perfused (less liver, kidney, and GI tract)
VPC 0.7 0.7 Slowly perfused (less bone)
Source: Campbell et al. (2009).

An oral mucosa compartment was added to the rat model and compartments for the stomach,
small intestine, and large intestine were added to the model description as shown in the
schematic of the model in Figure 3-9. This structure is likely superior to the simplified first-
order process used by O’Flaherty in that it addresses proximal to distal movement of chromium
through the GI tract. Using the NTP (2007) subchronic data, uptake in the model was fit to the
kidney data and then used to predict the blood uptake and GI tissue distribution of total
chromium in the NTP subchronic kinetic studies for the rats and mice. The apparent nonlinear,
dose-dependent uptake of chromium from drinking water in these studies could be adequately
described using Michaelis-Menten kinetics (Campbell et al., 2009). The model predictions were
shown to adequately fit the kidney and blood data. Further development, optimization, and
publication of the model against additional data sets with richer tissue information such as
Kargacin et al. (1993), Sutherland et al. (2000), and the NTP (2008) is warranted. Advantages of
this type of PBPK modeling is that the nonlinearities and spatial properties in uptake process can
be described more physiologically, quantitative extrapolations can be performed across species,
and predictions of target tissue dose can provide a more accurate description of dose-response.

Pool B
Inhalation
Pool A
Lung
CrVI →CrIII
Well Perfused
Red Blood Cells
CrVI →CrIII
CrVI →CrIII
Plasma
Poorly Perfused
CrVI →CrIII
Drinking
Water
Bone
CrVI →CrIII
Buccal Cavity
Liver CrVI →CrIII
CrVI →CrIII
GI Tract
Kidney
CrVI →CrIII
CrVI →CrIII
Urinary
Excretion Fecal
Retained Excretion Oral
Urine Exposure
Oral Cavity GI Tract
Feces
Lumen Cr Solution
+ Saliva
Cr(VI) Reduction Cr(III)
?
Epithelial
Tissue Reduction
Cr(VI) Cr(III)
?
Submucosal
Tissue Cr(VI) Cr(III)
Blood Flow
Source: Campbell et al. (2009).
Figure 3-9. Extended PBPK model structure to predict chromium

distribution in the oral cavity and GI tract tissue for rats and mice.

3.6. CONSIDERATION OF CHROMIUM ESSENTIALITY VERSUS TOXICITY
Further research is needed on this important area for characterizing the beneficial versus
toxic effects of chromium (Vincent, 2004; Costa and Klein, 2006). Chromium trivalent
supplements are among the most highly absorbed forms of Cr(III). Chromium (III) has been
purported to have beneficial effects on insulin, but that role is currently debated (Costa and
Klein, 2006; Stearns, 2000). It is now believed that “nutritional supplement” may be a better
term for describing the biological role for Cr(III) since the generation of chromium deficiency in
humans is difficult to define (Mertz, 1998; Stearns, 2000; Costa and Klein, 2004). As evidenced
in the preceding sections, any discussion of the potential toxicity of Cr(III) should also include
an evaluation of Cr(VI) as the biological disposition of the two are related (Stearns, 2000). The
Reference Daily Intake (RDI) for chromium of 50–200 µg was revised to a Dietary Reference
Intake (DRI) of 35 µg. DRI values are the most recent set of dietary recommendations
established by the Food and Nutrition Board of the Institute of Medicine, 1997-2001. They
replace previous RDAs, and may be the basis for eventually updating the RDIs.

4. HAZARD IDENTIFICATION
4.1. ORAL STUDIES IN HUMANS

The human health effects observed following oral ingestion of hexavalent chromium
usually come from individuals accidentally or intentionally ingesting hexavalent chromium
compounds or from human populations unknowingly consuming food or drinking water
contaminated with hexavalent chromium.
4.1.1. Acute Exposure

Several case reports have been published on clinical signs and symptoms in individuals
following acute accidental or intentional ingestion of high doses (fatal or near fatal) of
hexavalent chromium compounds, including chromic acid (Loubieres et al., 1999; Saryan and
Reedy, 1988; Fristedt et al., 1965), potassium dichromate (Hantson et al., 2005; Clochesy, 1984;
Iserson et al., 1983; Sharma et al., 1978; Kaufman et al., 1970; Partington, 1950; Goldman and
Karotkin, 1935), and ammonium dichromate (Hasan, 2007; Reichelderfer, 1968). Clinical
presentation of patients following acute, high-dose exposure was similar, regardless of the
specific hexavalent chromium compound ingested, and included the following: abdominal pain,
nausea, and vomiting; hematemesis and bloody diarrhea; caustic burns of mouth, pharynx,
esophagus, stomach, and duodenum and GI hemorrhage; anemia, decreased blood Hgb,
abnormal erythrocytes, and intravascular hemolysis; hepatotoxicity (hepatomegaly, jaundice,
elevated blood bilirubin, and liver enzymes activities); renal failure (oliguria and anuria);
cyanosis; and metabolic acidosis, hypotension, and shock. Findings on tissue biopsies included
hepatic fatty degeneration and necrosis and renal tubular degeneration and necrosis (Loubieres et
al., 1999; Sharma et al., 1978; Kaufman et al., 1970; Reichelderfer, 1968). Based on estimated
amounts of hexavalent chromium ingested, the range of lethal doses for hexavalent chromium in
humans is from approximately 4.1 to 357 mg hexavalent chromium/kg body weight (Loubieres
et al., 1999; Saryan and Reedy, 1988; Clochesy, 1984; Iserson et al., 1983; Kaufman et al.,
1970).
A series of acute and short-term repeated (17-day) ingestion studies were conducted on
volunteers to evaluate hexavalent chromium pharmacokinetics (Corbett et al., 1997; Finley et al.,
1997; Kerger et al., 1997, 1996b; Kuykendall et al., 1996; Paustenbach et al., 1996). With the
exception of Paustenbach et al. (1996), these studies reported that study protocols were reviewed
and approved by a human use committee comprised of three board-certified occupational
physicians and one board-certified toxicologist. In each case, the committee determined that
participants were properly informed of the reported adverse health effects associated with
hexavalent chromium exposure. The study by Paustenbach et al. (1996) involved a single male
volunteer. The methods section of this study noted that “The volunteer had a PhD in toxicology,

and the test protocol was approved by a human use committee.” As part of these studies,
standard clinical evaluations were performed that included blood cell counts, blood clinical
chemistry (SMA-20), and urinalysis (volume, specific gravity, creatinine). In the longest
duration exposure, a single subject ingested 2 L/day of a solution of containing 2 mg hexavalent
chromium/L (as potassium dichromate in water) for 17 consecutive days (approximately 0.06 mg
hexavalent chromium/kg-day, assuming a 70-kg body weight) (Paustenbach et al., 1996). In
shorter duration studies, 3–5 subjects ingested 1 L/day of solutions containing 0.1–10 mg
hexavalent chromium/L in water (approximately 0.001–0.14 mg hexavalent chromium/kg-day,
assuming a 70-kg body weight) for 1–3 days (Finley et al., 1997; Kerger et al., 1997, 1996b;
Kuykendall et al., 1996). Data from the clinical evaluations were not reported; however, results
were described in general terms that suggested that values for clinical chemistry endpoints were
“similar” when measured prior to, during, or following dosing (Paustenbach et al., 2003, 1996).
4.1.2. Environmental Exposure

Human studies of possible associations between oral exposures to environmental
hexavalent chromium and health outcomes include several epidemiology studies in which health
outcomes (primarily cancer) were evaluated among populations that resided near sources of
industrial waste containing hexavalent chromium compounds in Liaoning Province, China
(Kerger et al., 2009; Beaumont et al., 2008; Zhang and Li, 1997, 1987, 1980), Kings County/San
Bernardino County, California (Fryzek et al., 2001), Nebraska (Bednar and Kies, 1991), and
Glasgow, United Kingdom (Eizaguirre-Garcia et al., 2000, 1999). In addition to these studies,
two cases of Hodgkin’s disease in residents of Hinkley, California, where hexavalent chromium
was used as a cooling additive at a local gas plant, were described in a case report by Bick et al.
(1996).
Liaoning Province, China (Kerger et al., 2009; Beaumont et al., 2008; Zhang and Li, 1997,
1987)
In 1987, Zhang and Li published a paper describing the soil and water contamination by
chromium in the vicinity of an alloy plant where chromium was smelted in the JinZhou area of
Liaoning Province, China (Zhang and Li, 1987). This paper was based on an earlier unpublished
report (Zhang and Li, 1980). A more detailed mortality analysis, which included variation in
cancer mortality rates among the five villages along the Nuer River, was published in 1997
(Zhang and Li, 1997) in the Journal of Occupational and Environmental Medicine. This study
has had a controversial history that culminated in the retraction, in 2006, of the latest report
(Zhang and Li, 1997) by the editors of the Journal of Occupational and Environmental Medicine
because “financial and intellectual input to the paper by outside parties was not disclosed”
(Brandt-Rauf, 2006). The financial and intellectual input in question were those from a
consulting firm that had (or may have had) financial ties with industry clients potentially liable

for chromium wastes in the United States (Smith, 2008). Two reanalyses of data compiled by
Zhang and Li have also been reported (Kerger et al., 2009; Beaumont et al., 2008). The
following presentation of the studies begins with a description of the geographic area, industrial
operations, and resulting chromium dispersion in the surrounding communities, with information
obtained from the most recent reports (Kerger et al. 2009; Beaumont et al., 2008) and from
earlier published and unpublished reports (Zhang and Li, 1986, 1980; JinZhou Antiepidemic
Station, 1979). The commonalities and differences in the reanalyses by Kerger et al. (2009) and
Beaumont et al. (2008) are then described.
The study area is west of JinZhou, a city in Liaoning Province in northeastern China.
This area was described by Zhang and Li (1987) as being primarily agricultural with some
pockets of industries. One of the industrial plants is the JinZhou ferrochromium alloy plant,
located near the Nuer River. The town of TangHeZi was developed around the plant (Zhang and
Li, 1980). A series of five small rural villages (JinChangBao, Nuer River Village, YangXing,
ShiLiTai, and WenJiaTun) are located approximately 1–5 km to the east of the plant along the
Nuer River. The alloy plant began trial smelting of chromium in 1959, small-scale production in
1961, and mass production in 1965 (Zhang and Li, 1987). Liquid wastes from the production
process were released to a dry river bed (the “Old Nuer River”) near the plant. The amount of
hexavalent chromium in the wastewater was considerable (estimated as 20 mg/L at the end of the
discharge pipe) (Zhang and Li, 1986). Solid wastes (>300,000 tons by 1986) were stored in
outdoor piles and were subject to leaching to surface water and groundwater. These piles of ore
residue were the main long-term source of underground water contamination. Hexavalent
chromium was also released into the air through the various production and waste processes,
with a northeast prevailing wind pattern. An additional source of chromium exposure was from
food grown in areas using contaminated well water for irrigation.
In 1964, residents in the Nuer River Village noticed a yellowing of the color of their
drinking water. The local health department (referred to as the “JinZhou Disease Control and
Prevention Station”, “JinZhou Health and Anti-epidemic Station”, or “JinZhou Antiepidemic
Station” depending on the translation) initiated testing of well water samples in each of the five
villages in 1965. Chromium was found in 75 (28%) of the first set of samples from 266 wells in
JinChangBao and Nuer River Village, with levels up to 10 mg/L. By the end of 1965, the zone
of underground water contamination had spread, following a path eastward from the plant. In
JinChangBao, 41% of the wells contained hexavalent chromium, as did 96% of the wells in Nuer
River Village. The highest concentration (5 mg/L) was found in YangXing and Nuer Railway
Station, which are east of JinChangBao and Nuer River Village. In 1966, hexavalent chromium
was detected in the Nanshan reservoir (supplying drinking water to JinZhou), 9 km from the
alloy plant. Monitoring of well water continued, and the expansion of the contamination zone
appeared to peak in 1979 (Zhang and Li, 1986). A variety of efforts to reduce the chromium run-
off were undertaken in 1965–1967.

Table 4-1 includes a compilation of the available data from the 1965 water sampling
studies (based on Table 2 from Beaumont et al., 2008, with the addition of the distance from the
plant and average chromium levels in the well water samples from Kerger et al., 2009). The
analytical methods used to quantify chromium were not reported, but these values (and all other
values for chromium concentrations noted below) were reported as hexavalent chromium;
Beaumont et al. (2008) note that other forms would not be expected to be water soluble.
Beaumont et al. (2008) and Kerger et al. (2009) are in general agreement regarding their
interpretation of the 1965 water testing data. There is disagreement, however, as to what can be
established regarding levels in later years (Table 4-1), and the stability of the relative levels
among the villages. Beaumont et al. (2008) do not consider the available data to be adequate to
classify the individual villages with respect to a relative ranking of exposure, given the lack of
information regarding the selection of wells sampled, lack of information regarding use of
specific wells by individuals within the villages, paucity of data from later years, and rapid
changes in chromium concentrations in various areas due to the groundwater movement as well
as the efforts to curtail the chromium contamination. Kerger et al. (2009), however, use the 1965
well water sample data to derive two measures of exposure (average chromium concentration
and percent of wells >0.05 mg/L) that they applied to each of the five villages for an exposure-
response analysis of cancer risk.

Table 4-1. Data pertaining to hexavalent chromium concentrations in
drinking water in five villages along a path of groundwater contamination
from an alloy plant in western JinZhou, China from 1965 to 1979
Village (km from alloy plant)

JinChanBao Nuer River Village YangXing ShiLiTai WenJiaTun
Yr (1.4) (1.5) (3.0) (3.5) (5.0)
Early 1965a Hexavalent chromium detected in
75 (28%) of 265 wells sampled in
JinChanBao and Nuer River Village;
73 of the 75 wells were in Nuer
River Village; 41 (15%) were >2.0
mg/L (range 0.6–10.0 mg/L).
Later in 1965a,b
Number of wells sampleda,b 123 170 50 21 33
Hexavalent chromium (mg/L)a Number of wells (%)
<0.001 73 (59) 7 (4) 14 (28) 2 (10) 27 (82)
0.001–<0.05 35 (28) 1 (1) 16 (32) 19 (90) 6 (18)
0.05–<0.1 7 (6) 5 (3) 5 (10) 0 (0) 0 (0)
0.01–<0.5 8 (7) 27 (16) 12 (24) 0 (0) 0 (0)
0.5–<1.0 0 (0) 17 (10) 2 (4) 0 (0) 0 (0)
1.0–<5.0 0 (0) 76 (45) 1 (2) 0 (0) 0 (0)
≥5.0 0 (0) 37 (22) 0 (0) 0 (0) 0 (0)
Maximum (mg/L)a,b 0.4 20.0 <5 <0.05 <0.05
Average (mg/L)b 0.031 2.6 0.18 0.02 0.004
1966c 0.002–20.0
1967b <0.05 <0.05 <0.05
1972b <0.05
1974 10.5d 0.01–0.05c
1979c 0.06–4.33 0.001–0.03 0.003–0.004
a
As reported by Beaumont et al. (2008).
b
As reported by Kerger et al. (2009).
c
As reported by Zhang and Li (1986), number of samples not stated.
d
Zhang and Li (1986) report this concentration as 70.5 mg/L, but Zhang and Li (1987), Beaumont et al. (2008), and
Kerger et al. (2009) report a concentration of 10.5 mg/L. The total number of samples and the range in
concentrations were not specified.
A mortality study was described first by Zhang and Li (1980) in an unpublished report
for the JinZhou health department, and later published in a Chinese journal (Zhang and Li,
1987). Mortality records for the period 1970–1978 were obtained from local police stations for
the five villages along the Nuer River, the district surrounding the ferrochromium alloy plant
(TangHeZi), and three other areas to the west (YaoTangHeZi) and north (North ThangHeZi,
North Nuer River) of the plant. TangHeZi and the other three areas were not affected by the
groundwater chromium contamination, and these areas serve as one of the comparison groups in
the analyses. Cause of death was abstracted by trained study staff and reviewed by Dr. Zhang
(Kerger et al., 2009). A study interview was also conducted (with unspecified surrogates), but

the content of the interview was not described in detail (Zhang and Li, 1980). The mortality
analysis indicated that the lung cancer rate was relatively high in TangHeZi (the industrial town
surrounding the ferrochromium alloy plant), but decreased in areas further to the north (Zhang
and Li, 1980). In the areas to the east of TangHeZi (JinChangBao, Neur River Village, ShiLiTai,
YangZing, and WenJiaTun), total cancer mortality rates (71.9–92.7 per 100,000 person-years)
were high relative to the region (65.4 per 100,000 person-years). Similar elevations were seen
for lung cancer mortality (13.2–21.4 compared with 11.2 per 100,000 person-years in the eastern
villages and comparison region, respectively, and stomach cancer mortality rates (27.7–55.2 in
the eastern villages; comparison rates not given in the report, but the authors state these rates are
“higher than the district as a whole”) (Zhang and Li, 1987).
A subsequent paper by Zhang and Li (1997) expanded their work to include an analysis
of variation in cancer rates among the five villages in the contamination zone in relation to
distance from the plant and other exposure measures. This analysis is also included in the
Kerger et al. (2009) report, described below.
The mortality data described in the reports by Zhang and Li (1987, 1980) are the basis for
the subsequent analyses by Beaumont et al. (2008) and Kerger et al. (2009). The reanalyses by
Beaumont et al. (2008) and Kerger et al. (2009) provide very similar estimates of person-years.
Beaumont et al. (2008) used 1982 census data for the study areas and estimated annual growth
rates from 1970 to 1982 for the Liaoning Province to estimate yearly population counts for each
of the nine study areas; the summation of these figures from 1970 to 1978 represents the person-
years for the study period. Kerger et al. (2009) based the population figures on the estimated
populations in 1974 and multiplied these numbers by 9 (number of years of follow-up) to
estimate person-years for each of the study regions. TangHeZi, the industrial area surrounding
the ferrochromium alloy plant (1975 population approximately 17,500), is approximately 3–
10 times bigger than the other study areas (Table 4-2).

Table 4-2. Results pertaining to cancer mortality rates in five villages along path of groundwater contamination from
alloy plant and other comparison areas, western JinZhou, China from 1970 to 1978, based on analyses by Beaumont et
al. (2008) and Kerger et al. (2009)
Rate per 100,000 person-yrs

All cancer Stomach cancer Lung cancer
Age-adjusted Age-adjustment Estimated age- Estimated age-
Area (population or person-yrs)a rate influenceb Crude rate adjusted rateb Crude rate adjusted rateb
Areas in contamination zone
JinChanBao (2,900) 83.6 0.97 36.7 35.5 13.2 12.8
Nuer River Village (2,800) 71.9 0.98 28.0 Missingb 15.0 14.7
ShiLiTai (2,600) 93.0 0.94 55.2 51.7 Missing Missing
YangXing (1,100) 76.8 0.94 36.5 34.5 21.4 20.2
WenJiaTun (1,700) 91.1 0.94 27.7 26.0 20.8 19.5
Group average (~98,700)c 81.3 34.9 35.3 17.1 16.9
Comparison areas
TangHeZi (17,500) 71.3 0.86 16.9 14.5 21.4 18.3
North TangHeZi (3,600) 81.8 0.84 26.4d 22.1 8.8 7.4
North Nuer River (5,800) 71.8 1.05 30.5 31.9 7.6 8.0
Yao TangHeZi (1,500) 61.3 0.90 26.6 23.8 20.0 17.9
Group average—all (~252,500)e 72.1 19.4 14.7
Group average—without TangHeZi (96,826)f 73.7 28.6 9.7
a
Area population figures are based on approximate 1975 data from Beaumont et al. (2008); group values are total person-yrs for the combined area.
b
As calculated by Beaumont et al. (2008). Nuer River Village stomach cancer rate was not included in the primary analysis by Beaumont et al. (2008) because it was
missing in the original (1980) report; an additional analysis used a rate of 28 as reported by Zhang and Li (1987).
c
Beaumont et al. (2008) estimate was 98,458 and Kerger et al. (2009) estimate was 98,850.
d
Beaumont et al. (2008) report this value as 26.14 in Table 2, but based on the calculation of the estimated age-adjusted rate, it appears that a value close to 26.3 was
used; Kerger et al. (2009) report this value as 26.4.
e
Beaumont et al. (2008) estimate was 252,277 and Kerger et al. (2009) estimate was 253,282.
f
As reported by Kerger et al. (2009).

The numbers of total cancer deaths, lung cancer deaths, and stomach cancer deaths were
used in combination with estimated person-years at risk as the basis of the calculation of area-
specific mortality rates in the analyses by Zhang and Li (1997, 1987, 1980), Beaumont et al.
(2008), and Kerger et al. (2009). Because the results of Zhang and Li (1997) are repeated in the
presentation by Kerger et al. (2009), only the more recent of these analyses is described in more
detail below.
There are two differences between the analyses of the cancer mortality data presented by
Beaumont et al. (2008) and Kerger et al. (2009). One of the minor differences is the value used
for stomach cancer mortality for one of the villages in the contamination zone, Nuer River
Village. Beaumont et al. (2008) do not include an estimate of stomach cancer mortality for Nuer
River Village in their primary analysis because it was missing from the original (1980)
unpublished report (Zhang and Li, 1980) and Dr. Zhang indicated in a faxed communication with
the study authors that the estimated rate of 28 per 100,000 per year (reported in Zhang and Li,
1997) was of uncertain accuracy. Beaumont et al. (2008) did repeat their analysis using the 28
per 100,000 rate for stomach cancer mortality in Nuer River Village, and found that this
inclusion had very little effect on their estimates. Kerger et al. (2009) used 28 per 100,000 per
year as the stomach cancer rate for Nuer River Village. The second relatively minor difference
is in the estimation of age-adjusted mortality rates. The original analyses by Zhang and Li
(1987) presented age-adjusted rates for all cancer mortality, but not for stomach cancer or lung
cancer mortality. Kerger et al. (2009) do not attempt to make an age adjustment for lung or
stomach cancer because “small numbers of site-specific deaths in the villages would have
precluded the calculation of relatable direct standardized site-specific rates in the current study.”
Beaumont et al. (2008) addressed this issue by calculating the ratio of unadjusted to adjusted
total cancer rates for each study area, which they term the “age-adjustment influence” ratio. This
ratio ranged from 0.84 to 1.05. The area-specific lung and stomach cancer unadjusted rates were
multiplied by the respective area-specific age-adjustment influence ratio to create estimated age-
adjusted lung and stomach cancer rates (Table 4-2).
One of the major differences between the analyses by Beaumont et al. (2008) and Kerger
et al. (2009) was described previously: Kerger et al. (2009) use the 1965 exposure data for
exposure-response modeling of the variation in cancer rates among the five villages in the
chromium contamination zone, and Beaumont et al. (2008) do not believe that the available data
are adequate for this purpose. The other major difference between the analyses is the inclusion
of TangHeZi, the industrial district surrounding the ferrochromium alloy plant, in the comparison
group. Kerger et al. (2009) considered this district to be too different from the smaller villages in
terms of urban-rural lifestyles and other exposures that could affect cancer risk (specifically
stomach cancer and lung cancer), and therefore, did not include it in their comparison group.
Beaumont et al. (2008) include TangHeZi, presumably because it was part of the original study
design. They do not explicitly address the comparability issue with respect to stomach cancer

risk factors, although they do note the potential for occupational chromium exposure to
contribute to a relatively high lung cancer rate in TangHeZi.
Table 4-3 presents the measures of association between chromium exposure and cancer
mortality, based on the five villages in the contamination zone and the various comparison
groups used by Beaumont et al. (2008) and Kerger et al. (2009). These risk ratios are based on
comparison of the rates shown in Table 4-2, using a Poisson distribution for calculation of 95%
confidence intervals (CIs). With respect to stomach cancer, the primary site of interest from the
standpoint of drinking water contamination, Beaumont et al. (2008) report an association using
the four comparison areas (TangHeZi, North TangHeZi, North Nuer River, and YaoTangHeZi)
that were the basis for the original analysis (risk ratio = 1.82, 95% CI 1.11–2.91) and using rates
from all of Liaoning Province as a comparison (risk ratio = 1.69, 95% CI 1.12–2.44). Kerger et
al. (2009) excluded the most populous area, TangHeZi from the comparison group, and reported
a risk ratio = 1.22 (95% CI 0.74–2.01), which they interpret as being evidence of no association.
In the lung cancer analyses, Beaumont et al. (2008) report relatively little difference between the
rates in the contamination zone and the comparison area (risk ratio = 1.15, 95% CI 0.62–2.07),
but a stronger association using Liaoning Province as a comparison (risk ratio = 1.78, 95%
CI 1.03–2.87). Kerger et al. (2009) observed higher lung cancer rates in the five villages in the
contamination zone compared with the three rural areas they included in the comparison group
(risk ratio = 1.76, 95% CI 0.78–3.98), and slightly reduced risk when compared to TangHeZi
(risk ratio = 0.80, 95% CI 0.44–1.47).
Table 4-3. Risk ratios comparing cancer mortality rates in five villages
along a path of groundwater contamination from an alloy plant and other
comparison areas in western JinZhou, China from 1970 to 1978
All cancers Stomach cancer Lung cancer

a
Comparison group Risk ratio 95% CI Risk ratio 95% CI Risk ratio 95% CI
All four areasb 1.13 0.86–1.46 1.82 1.11–2.91 1.15 0.62–2.07
Excluding TangHeZic 1.10 0.80–1.51 1.22 0.74–2.01 1.76 0.78–3.98
Liaoning Provinceb 1.23 0.97–1.53 1.69 1.12–2.44 1.78 1.03–2.87
a
TangHeZi, North TangHeZi, North Nuer River, and YaoTangHeZi.
b
Reported by Beaumont et al. (2008).
c
Reported by Kerger et al. (2009).
Kerger et al. (2009) also presented results of analyses of variation in cancer rates within
the five villages in the chromium contamination zone, using three measures of exposure
potential: distance from the plant, average hexavalent chromium concentrations in 1965, and
percent of wells with >0.05 mg/L hexavalent chromium in 1965 (these measures can be found in
Table 4-1). The analysis was based on Poisson regression of the log-transformed cancer rate in
relation to the exposure measures (separate models run for each measure). For the distance

measure, a negative value for the coefficient indicates an increased cancer rate with closer
proximity to the plant, and for the other exposure measures, a positive coefficient indicates an
increased cancer rate with higher exposure. The results for all cancer mortality (given as the
regression coefficient and p-value) were 0.04 (p = 0.61), -0.07 (p = 0.54), and -0.24 (p = 0.45)
for the distance, average hexavalent chromium concentration in 1965, and percent of wells
>0.05 mg/L hexavalent chromium in 1965 measures, respectively. For stomach cancer mortality,
the coefficients were 0.01 (p = 0.93), -0.11 (p = 0.50), and -0.32 (p = 0.51) for the distance,
average hexavalent chromium concentration in 1965, and percent of wells >0.05 mg/L
hexavalent chromium in 1965 measures, respectively, and for lung cancer, the coefficients were
0.12 (p = 0.50), -0.06 (p = 0.79), and -0.11 (p = 0.88) for the distance, average hexavalent
chromium concentration in 1965, and percent of wells >0.05 mg/L hexavalent chromium in 1965
measures, respectively. As described previously, Beaumont et al. (2008) did not include this
type of exposure-response analysis because they believed that the inherent limitations of the
exposure data precluded a meaningful analysis.
In addition to the cancer mortality study, the JinZhou health department also collected
data pertaining to symptoms in 1965 in Nuer River Village, which was one of the highly
contaminated areas at that time (well water hexavalent chromium levels of 0.1–20.0 mg/L)
(Zhang and Li, 1987, 1986). Among 156 residents surveyed, 51 (33%) had oral ulcers, 20 (17%)
had diarrhea, 48 (31%) had abdominal pain, 26 (17%) had dyspepsia, 81 (30%) had stomach
pain, and 20 (17%) had vomiting (JinZhou Antiepidemic Station, 1979). The authors state that
“no such symptoms were found among the residents whose water wells were not contaminated.”
A similar study of 158 people in ShiLiTai in 1971 found a similar pattern of symptoms, with
92 (58%) reporting oral ulcers, 48 (30%) diarrhea, and 36 (23%) abdominal pain. In 1974,
another study of children in WenJiaTun and Sandaohao, at the eastern edge of the contamination
zone, also found similar symptoms (data not shown in the 1979 report). The authors speculate
that the symptoms may have been due to the increased concentrations of sulfates (>300 mg/L) in
the drinking water in these areas in 1974, rather than the relatively low concentrations of
hexavalent chromium (0.003–0.05 mg/L).
Zhang and Li (1987, 1986) also conducted hematological assessments of 12 individuals
in 1965, and another study of 93 individuals (time not specified). The exact location of the
participants was not specified, but they were said to be from “highly polluted” or “high density
contamination” areas. White blood cell counts were elevated in the first study, and the number
of neutrophilic granulocytes and what was termed “juvenile cells” among these granulocytes was
elevated in the second study.
Kings County/San Bernardino County, California (Fryzek et al., 2001)

A study of areas in Kings County and San Bernardino County, California, compared
cancer mortality in locations near natural gas compressor plants with areas not located near the

plants (Fryzek et al., 2001). Hexavalent chromium compounds had been used as anti-corrosion
additives in cooling tower water at the gas plants during the period 1950 to approximately 1980.
Waste material was released to surface ponds and was subject to percolation to groundwater.
Cooling tower water was also aerosolized and transported to the ground surface where it may
have contacted soil, crops, and surface water. Thus, exposures to hexavalent chromium may
have occurred by several routes (i.e., inhalation, ingestion, and dermal contact). Mortality
records for zip codes for the cities of Kettleman City (in Kings County), and Hinkley and
Topock (in San Bernadino County), in which natural gas compressor plants were located, were
compared to records from zip codes in Kings County and San Bernadino County, other than
those encompassing these three cities. The study included mortality records for the period 1989–
1998, during which time 2,226,214 deaths were recorded. Age-adjusted cancer mortality rate
ratios (rate in areas near the plant/rate in comparison areas) were 1.03 (95% CI 0.90–1.17) for
lung cancer death, 0.93 (95% CI 0.87–1.00) for all cancer deaths, and 0.98 (95% CI 0.95–1.02)
for all deaths. Rate ratios for stomach cancer were not reported. This study found no significant
difference between mortality or cancer mortality among residents from zip codes in which gas
plants that used hexavalent chromium additives in cooling tower water were located compared to
residents of other nearby areas without such plants. An important limitation of this study is that
exposure assignment was based on zip code, rather than on individual-level data, which is likely
to result in significant exposure misclassification.
Nebraska (Bednar and Kies, 1991)

Bednar and Kies (1991) compared levels of chromium (and other chemicals) in drinking
water in Nebraska counties with death rates in these same areas. Data on chromium in drinking
water were obtained for each of 453 communities (all incorporated communities of Nebraska)
for the period 1986–1987, and mortality data for each Nebraska county was obtained for the year
1986 (both compiled by the Nebraska Department of Health). Mean total chromium
concentration in drinking water for the 453 communities was 0.002 mg chromium/L (range
<0.001–0.01 mg chromium/L); the study report did not indicate valence state of chromium
detected in these drinking water samples. Possible associations between chromium exposure and
health outcomes were assessed by linear correlation (Pearson) of mortality rates (at the county
level) and chromium concentrations in drinking water (presumably aggregated from community
data to represent counties). Correlations were reportedly explored for mortality from cancer,
cerebrovascular disease, heart disease, pneumonia, and chronic lung disease; however, only one
chromium correlation coefficient was reported to be statistically significant, that for death from
chronic lung disease, and the correlation was negative (-0.101, p = 0.03). As with the other
studies of this design, a major limitation is that exposures to chromium cannot be estimated for
individual subjects in the study and may not be accurately represented by the drinking water
chromium measurements. For example, the 1986–1987 drinking water data do not necessarily

represent long-term exposure patterns, and an individual represented in a county death record
does not necessarily mean that the individual resided in the county for their lifetime or any
significant fraction of their lifetime.
Glasgow, United Kingdom (Eizaguirre-Garcia et al., 2000, 1999)

Eizaguirre-Garcia et al. (2000, 1999) examined risk of leukemia and birth defects in
people residing near the site of a former chromium processing facility in Glasgow, United
Kingdom. The factory was in operation for >100 years and ceased operations in 1967. A survey
conducted in 1991 found average soil concentrations at the site of operations of 8,164 mg/kg for
total chromium and 848 mg/kg for hexavalent chromium. Soil concentrations of total chromium
and hexavalent chromium approximately 2–3 km from the factory site were reported as
“approximately half” of those at the site; no additional information on soil levels off-site were
reported (Eizaguirre-Garcia et al., 2000, 1999). Reported cases of leukemia for the period 1975–
1989 were obtained from the Scottish Cancer Registration, during which 1,205 cases of leukemia
were reported in a population of 873,643 (Eizaguirre-Garcia et al., 1999). Leukemia cases were
aggregated at the level of Enumeration Districts (EDs) (approximately 350–500 individuals per
district). When stratified by distance of the EDs from the plant (out to 9–10 km), relative risks of
leukemia (0–2 km as reference) were unrelated to distance. When other influential variables
were included in a Poisson regression model (gender, socioeconomic status, and age) in addition
to distance of EDs from the plant (0–4, 4–9, 9–10 km), relative risk was significant (1.29, 95%
CI 1.07–1.56) for EDs 4–9 km from the plant (relative to 0–4 km), but not for EDs 9–10 km
from the plant. These results suggest that leukemia risk increased with distance from the plant
(i.e., 4–9 > 0–4 km) and then declined with further distance (i.e., 9–10 = 0–4 km). This pattern
does not strongly implicate the plant as a major contributor to leukemia risk.
A similar study of risk of birth defects was conducted on the same population
(Eizaguirre-Garcia et al., 2000). In this study, data on number of births and congenital
malformations were collected for the period 1982–1989. Case definitions (not reported)
followed those of the European-wide EURCAT network (http://www.eurocat.ulster.ac.uk/). The
study included 2,778 cases from a population of 81,057 births; cases were aggregated at the level
of EDs. When distance from the plant (0–1, 2–4, 4–10 km) and socioeconomic status were
included in a Poisson regression model, relative risk was significant for the EDs in the 2–4 km
category (1.47, 95% CI 1.2–1.7) and the 4–10 km category (1.25, 95% CI 1.05–1.49); however,
both distance categories were associated with higher risks than the closest distance category, 0–
1 km. Similar to the results for leukemia, this pattern does not strongly implicate the plant as a
major contributor to risk of congenital anomalies. Not taken into consideration in this study
were several other potentially influential variables on developmental outcomes; for example,
maternal age and health, smoking, and alcohol consumption.

Summary
The Liaoning Province studies provide the most detailed analysis of all of the
epidemiological studies that have been conducted with respect to chromium and cancer mortality
(specifically stomach cancer or other cancers of the digestive system). These studies are
important in that they examined a population exposed to very high levels of chromium in
drinking water wells (i.e., sufficient to impart a visible yellow color to the water). Sources of
exposure include the drinking water, food grown in contaminated soil, and possibly air. Levels
up to 20 mg/L in well water were documented in the first surveys done in 1965 in the two
villages closest to the source of exposure (a ferrochromium alloy plant). The contamination
began sometime between 1959 and 1964; the reporting of a yellowing of the water by local
residents in 1964 is what led to the investigation and identification of this contamination by the
local health department.
The interpretation of the mortality data originally collected by Zhang and Li (1980)
depends in large part on the choice of referent group. That choice depends on many factors,
including the perceived comparability and the size of the populations. 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. Smaller areas (e.g., a neighboring community)
offer the advantage of potentially greater similarities in ethnic background, socioeconomic
status, and occupational and lifestyle factors that may affect cancer risk. However, small
comparison groups are likely to produce imprecise estimates, and the issue of over-controlling
may arise, for example, if the comparison population shares the specific exposure of interest (for
example, with the selection of friends or co-workers in case-control studies). The associations
presented by Beaumont et al. (2008) using Liaoning Province as the comparison group provide
evidence of an excess risk in the villages in the contamination zone of mortality from stomach
cancer (rate ratio 1.69, 95% CI 1.12–2.44) and lung cancer (rate ratio 1.78, 95% CI 1.03–2.87),
with a small increase also suggested in total cancer mortality (rate ratio 1.23, 95% CI 0.97–1.53).
The association with stomach cancer mortality is also seen when using the four adjacent areas as
the referent group (rate ratio 1.82, 95% CI 1.11–2.91), but is weaker when the industrial area
surrounding the plant, TangZeHi, is removed from the comparison group (rate ratio 1.22, 95%
CI 0.74–2.01). Kerger et al. (2009) believe that the relatively urban environment of TangHeZi
makes it an inappropriate comparison group for the villages in the contamination zone. With
respect to stomach cancer, historical trends show clear decreases in the incidence of this cancer
in a variety of geographical areas, with improvements that come with economic development and
urbanization (e.g., sanitation, refrigeration) contributing to this decline. 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) (Yang, 2006). However, this same study
reported little difference between urban and rural 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. Thus, EPA does not consider the exclusion of TangZeHi from the
comparison group to be warranted.
Another issue regarding the interpretation of the mortality data is the validity of analyses
of the variability in cancer rates among the five villages in the contamination zone in relation to
the available exposure measures (distance from the plant, average concentration in wells in 1965,
and percent of wells with hexavalent chromium levels >0.05 mg/L in 1965). There are
considerable limitations to these measures, including the lack of individual-level data on use of
water from specific wells over time and the changes in exposure due to efforts to treat the water
in the most contaminated areas with treatment wells built in 1967. Based on these limitations,
EPA concluded that the exposure-response analyses presented by Zhang and Li (1997) and
Kerger et al. (2009) are not based on the quality of data that is needed to support a conclusion
regarding the presence or absence of a dose-response among the observed cancer rates in these
villages.
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN

ANIMALS—ORAL
The effects of subchronic oral exposure to hexavalent chromium have been evaluated in
rats (NTP, 2007; Quinteros et al., 2007; Rafael et al., 2007; Acharya et al., 2001; Chopra et al.,
1996; Vyskocil et al., 1993) and mice (NTP, 2007; Asmatullah and Noreen 1999), and the effects
of chronic oral exposure to hexavalent chromium have been evaluated in rats (NTP, 2008;
MacKenzie et al., 1958), mice (NTP, 2008; Borneff et al., 1968), and dogs (Anwar et al., 1961).
The studies conducted by the National Toxicology Program (NTP, 2008, 2007) provide dose-
response data on the effects of oral hexavalent chromium exposure based on a comprehensive
assessment of toxicological endpoints. Effects associated with oral exposure to hexavalent
chromium as reported in the NTP (2007) subchronic study included hematological effects,
hepatotoxicity, alterations in lipid metabolism, and histopathological changes in GI tissues and
pancreatic and mesenteric lymph nodes. The most sensitive hexavalent chromium-induced
effects were microcytic, hypochromic anemia, increased serum liver enzyme activities, and
histopathological changes to the duodenum and pancreatic lymph nodes in rats; and
histopathological changes in the duodenum in mice. The most sensitive noncancer effects in the
NTP (2008) 2-year toxicology and carcinogenicity study were nonneoplastic histopathological
changes to the liver, duodenum, and mesenteric lymph nodes in rats and the duodenum,
mesenteric lymph nodes, and liver in mice. In addition, based on findings of squamous cell
neoplasms of the oral cavity in rats and neoplasms of the small intestine in mice, NTP (2008)
concluded that results of this study provide clear evidence of carcinogenic activity of sodium
dichromate dihydrate.
Several other oral exposure studies (i.e., Quinteros et al., 2007; Rafael et al., 2007;
Asmatullah and Noreen, 1999; Vyskocil et al., 1993; Anwar et al., 1961) do not provide suitable

data for identifying NOAELs or LOAELs because comprehensive toxicological endpoints were
not evaluated in these studies. LOAELs identified by EPA in studies by Acharya et al. (2001)
and Chopra et al. (1996) were based on evaluation of a limited number of liver endpoints. In
addition, interpretation of results from these studies was limited due to the small number of
animals evaluated, lack of dose-response data, or inadequate reporting for estimation of doses in
mg hexavalent chromium/kg-day. However, results of these studies are useful for identification
of potential adverse effects of oral hexavalent chromium exposure. Subchronic and chronic
studies are summarized in Section 4.5, Synthesis of Major Noncancer Effects—Oral, Table 4-26.
4.2.1. Subchronic Oral Exposure

NTP, 2007
NTP (2007) conducted a 3-month toxicology study of sodium dichromate dihydrate in
drinking water in rats and mice. This study was divided into three separate studies evaluating
effects of treatment in: (1) male and female F344/N rats, (2) male and female B6C3F1 mice, and
(3) three strains of male mice (B6C3F1, BALB/c, and am3-C57BL/6). In the 3-month study in
F344/N rats, groups of 10 males and 10 females (“core” study animals) were exposed to sodium
dichromate dihydrate in drinking water at concentrations of 0, 62.5, 125, 250, 500, or 1,000 mg
sodium dichromate dihydrate/L (equivalent to 0, 21.8, 43.6, 87.2, 174.5, or 348 mg hexavalent
chromium/L, respectively) for 3 months. Based on water consumption monitored throughout the
study, NTP (2007) calculated average daily doses over the 3-month treatment duration of
approximately 0, 5, 10, 17, 32, or 60 mg sodium dichromate dihydrate/kg-day (equivalent to 0,
1.7, 3.5, 5.9, 11.2, or 20.9 mg hexavalent chromium/kg-day, respectively) for both males and
females. An additional 10 male and 10 female rats (“clinical pathology” animals) were exposed
to the same concentrations of sodium dichromate dihydrate for 4 weeks. “Core” study animals
were observed twice daily for mortality and clinical signs of toxicity; water consumption and
body weights were recorded weekly. Blood was collected from “clinical pathology” animals on
treatment days 5 and 23 and from “core” study animals at study termination for comprehensive
hematology and clinical chemistry endpoints. Urine was collected from “clinical pathology”
animals on day 16 and analyzed for comprehensive urinalytic endpoints. At study termination,
necropsies were performed on all “core” study animals, with organ weights recorded for heart,
right kidney, liver, lung, spleen, right testis, and thymus. Microscopic examinations of
comprehensive tissues were conducted in all “core” study animals in the control and 20.9 mg
hexavalent chromium/kg-day (high-dose) groups and on six “core” study animals from each of
the other treatment groups. In addition, all tissues identified as target organs in the 20.9 mg
hexavalent chromium/kg-day (high-dose) group were examined in lower dose groups until a no-
effect level was identified or all animals were examined.
No mortalities were observed in male or female rats exposed to sodium dichromate
dihydrate in drinking water for 3 months (NTP, 2007). Final body weights in male rats were

significantly decreased by 5 and 11% in the 11.2 and 20.9 mg hexavalent chromium/kg-day
groups, respectively, compared to controls. In females, final body weight was significantly
decreased by 9% in the 20.9 mg hexavalent chromium/kg-day group compared to controls. In
males and females in the ≥5.9 mg hexavalent chromium/kg-day groups, water consumption was
decreased (statistical significance not reported). Data on food consumption were not reported.
No treatment-related signs of clinical toxicity were observed throughout the study.
Results of hematology analyses show that exposure of male and female rats to sodium
dichromate dihydrate in drinking water produced microcytic, hypochromic anemia, characterized
by decreases in mean cell volume (MCV), hematocrit (Hct), hemoglobin (Hgb), and mean cell
hemoglobin (MCH) (NTP, 2007). In general, the severity of this anemia exhibited a dose
response, with the more pronounced effects observed at the 23-day time point versus the 3-
month time point (Table 4-4). After 5 days of exposure, small changes were observed in several
hematological parameters; however, decreases in all treatment groups were ≤5% compared to
controls. More severe, dose-related effects were observed after 23 days of treatment, with
changes observed in all treatment groups in males and females. Similar effects were observed
after 3 months of treatment, although severity at 3 months was generally less than that observed
at 23 days, indicating a compensatory hematopoietic response. Some of the hematological
parameters, e.g., hematocrit and erythrocyte counts of male rats treated for 3 months, although
decreased significantly, were within normal ranges. In female rats, the direction of the changes
in mean cell hemoglobin and erythrocyte count at 23 days differed from those at 3 months.
Blood smears showed evidence of erythrocyte injury or increased turnover, including erythrocyte
fragments, keratocytes, and blebbing (incidence data not reported). Increased reticulocyte counts
and nucleated erythrocytes, indicative of a compensatory hematopoietic response, were also
observed in both sexes at 23 days and 3 months; however, these increases did not exhibit a
consistent pattern of dose- or duration-dependence. Dose-dependent increases in platelet counts
occurred at 23 days in all treatment groups compared to controls; however, severity was
decreased at 3 months (Table 4-4). NTP (2007) stated that increased platelet counts are
consistent with compensatory hematopoiesis or an iron deficiency. Increased neutrophil and
monocyte counts were observed at higher doses (≥5.9 and ≥3.5 mg hexavalent chromium/kg-day
in males and females, respectively) and were considered by NTP (2007) to reflect an
inflammatory response related to the inflammatory gastric lesions. Results of hematological
analyses showed that exposure of rats to sodium dichromate dihydrate in drinking water at daily
doses ≥1.7 mg hexavalent chromium/kg-day produced microcytic, hypochromic anemia, but that
severity decreased slightly as exposure duration increased from 23 days to 3 months.

Table 4-4. Hematological effects in male and female F344/N rats exposed to
sodium dichromate dihydrate in drinking water for up to 3 months
Hematological Time on Treatment group (mg hexavalent chromium/kg-d)

parameter treatment 0 1.7 3.5 5.9 11.2 20.9
Males
Hct (%) 23 D 48.0 ± 0.5a 44.7 ± 0.7b 39.8 ± 0.8b 36.2 ± 1.0b 34.4 ± 0.5b 32.3 ± 1.1b
(93.1) (82.9) (75.4) (71.7) (67.3)
b
3 Mo 45.7 ± 0.2 45.2 ± 0.4 45.2 ± 0.3 44.8 ± 0.7 42.9 ± 0.4 36.9 ± 0.8b
(98.9) (98.9) (98.0) (93.9) (80.7)
b b b b
Hgb (g/dL) 23 D 15.9 ± 0.1 14.2 ± 0.2 12.0 ± 0.3 10.9 ± 0.3 10.3 ± 0.3 9.2 ± 0.3b
(89.3) (75.5) (68.6) (64.8) (57.9)
3 Mo 15.3 ± 0.1 15.2 ± 0.1 15.0 ± 0.1 14.4 ± 0.2b 13.3 ± 0.2b 10.9 ± 0.3b
(99.3) (98.0) (94.1) (86.9) (71.2)
b b b b
MCV (fL) 23 D 61.1 ± 0.5 53.6 ± 0.6 48.0 ± 0.4 46.4 ± 0.6 46.2 ± 0.3 46.4 ± 0.5b
(87.7) (78.6) (75.9) (75.6) (75.9)
3 Mo 51.8 ± 0.1 50.3 ± 0.2b 49.0 ± 0.1b 44.4 ± 1.0b 39.7 ± 0.5b 36.0 ± 0.4b
(97.1) (94.6) (85.7) (76.6) (69.5)
MCH (ρg) 23 D 20.1 ± 0.2 16.9 ± 0.2 b
17.2 ± 0.7 b
18.2 ± 0.4 19.7 ± 0.3 20.7 ± 0.6
(84.0) (85.6) (90.5) (98.1) (103.0)
b b b b
3 Mo 17.3 ± 0.1 16.9 ± 0.1 16.2 ± 0.1 14.2 ± 0.4 12.3 ± 0.2 13.0 ± 0.5b
(97.7) (93.6) (82.1) (71.1) (75.1)
Erythrocyte 23 D 7.94 ± 8.38 ± 0.11 7.13 ± 0.35c 6.0 ± 0.28b 5.25 ± 0.19b 4.54 ± 0.33b
count (106/µL) 0.10 (105.5) (89.8) (75.6) (66.1) (57.2)
3 Mo 8.88 ± 9.04 ± 0.09 9.25 ± 0.07 10.15 ± 0.22 10.87 ± 0.07 8.52 ± 0.45b
c b b b
0.05 (101.8) (104.2) (114.3) (122.4) (95.9)

Platelet count 23 D 745.2 ± 1,065.3 ± 2,768.6 ± 3,504.7 ± 4,226.0 ± 4,688.8 ±
(106/µL) 22.2 67.9b 328.5b 235.0b 204.5b 242.7b
(143) (372) (470) (567) (629)
b
3 Mo 618.6 ± 736.1 ± 11.5 604.3 ± 24.5 909.8 ± 119.1 1,743.1 ± 5,123.0 ±
20.0 (119) (98) (147) 178.0b 638.9b
(282) (828)
Females
a
Hct (%) 23 D 48.0 ± 0.4 46.6 ± 0.9 42.9 ± 0.8b 39.2 ± 0.7b 37.2 ± 0.7b 33.4 ± 0.6b
(97.1) (89.4) (81.7) (79.6) (69.6)
3 Mo 44.6 ± 0.4 45.2 ± 0.1 44.1 ± 0.3 42.9 ± 0.2b 42.6 ± 0.5b 38.3 ± 0.5b
(101.3) (98.9) (96.2) (95.5) (85.9)
Hgb (g/dL) 23 D 15.9 ± 0.1 14.7 ± 0.3b 13.0 ± 0.3b 11.8 ± 0.3b 10.9 ± 0.2b 9.7 ± 0.2b
(92.5) (81.8) (74.2) (68.6) (61.0)
3 Mo 15.2 ± 0.1 15.4 ± 0.1 14.9 ± 0.1 14.3 ± 0.1b 14.1 ± 0.2b 12.0 ± 0.2b
(101.3) (98.0) (94.1) (92.8) (78.9)
b
MCV (fL) 23 D 61.1 ± 0.4 53.9 ± 0.5 48.8 ± 0.5b 46.6 ± 0.6b 45.7 ± 0.4b 46.5 ± 0.5b
(88.2) (79.9) (76.3) (74.8) (76.1)
3 Mo 53.3 ± 0.1 53.3 ± 0.1 52.4 ± 0.2b 50.5 ± 0.3b 48.0 ± 0.9b 40.0 ± 0.7b
(100) (98.3) (94.7) (90.1) (75.0)
MCH (ρg) 23 D 20.4 ± 0.1 17.3 ± 0.2 18.0 ± 0.3 18.9 ± 0.7 21.0 ± 0.6 23.1 ± 0.5
(84.8) (88.2) (92.6) (102.9) (113.2)
3 Mo 18.4 ± 0.1 17.9 ± 0.1b 17.8 ± 0.1b 16.9 ± 0.1b 15.9 ± 0.4b 12.5 ± 0.3b
(97.3) (96.7) (91.8) (86.4) (67.9)

Table 4-4. Hematological effects in male and female F344/N rats exposed to
sodium dichromate dihydrate in drinking water for up to 3 months

Erythrocyte 23 D 7.82 ± 8.52 ± 0.14 7.22 ± 0.19 6.32 ± 0.36b 5.27 ± 0.23b 4.21 ± 0.16c
count (106/µL) 0.09 (109.0) (92.3) (80.8) (67.4) (53.8)
3 Mo 8.30 ± 8.60 ± 0.05 8.40 ± 0.04c
b
8.47 ± 0.04c 8.93 ± 0.11b 9.62 ± 0.10b
0.06 (103.6) (101.2) (102.0) (107.6) (115.9)
Platelet count 23 D 611.5 ± 1,156.3 ± 2,808.8 ± 3,295.0 ± 4,318.4 ± 5,132.8 ±
(106/µL) 43.7 76.4b 198.5b 349.7b 234.9b 247.0b
(189) (459) (539) (706) (839)
3 Mo 588.9 ± 605.8 ± 17.1 574.8 ± 21.3 528.2 ± 14.1 619.3 ± 55.4 1,524.9 ±
17.1 (103) (98) (90) (105) 193.3b
(259)
a
Values are means ± SE; values in parenthesis are percent of control; n = 10 rats/group, with the following
exceptions: 1.7 mg hexavalent chromium/kg-d group females on d 23 and mo 3 (n = 9), 3.5 mg hexavalent
chromium/kg-d group females on d 23 (n = 8), 5.9 mg hexavalent chromium/kg-d group females on d 23 (n = 9),
and 20.9 mg hexavalent chromium/kg-d group females on d 23 and mo 3 (n = 9).
b
Significantly different (p ≤ 0.01) from the control group by Dunn’s or Shirley’s test.
c
Source: NTP (2007).
Results of clinical chemistry analyses in male and female rats exposed to sodium
dichromate dihydrate in drinking water showed treatment-related increases in serum liver
enzyme activities, bile acids, and serum creatine kinase activity and alterations in lipid
metabolism (Table 4-5) (NTP, 2007). Serum alanine aminotransferase (ALT) and sorbitol
dehydrogenase (SDH) activities were significantly increased compared to controls in all
treatment groups at 3 months, with less severe effects seen at 23 days. A consistent relationship
between severity and dose was not observed. In male rats, elevations of ALT and SDH activities
increased with increasing dose between 1.7 and 11.2 mg/kg-day, but less severe elevations were
observed at 20.9 mg/kg-day (Table 4-5). In females, increases in ALT and SDH activities were
generally indicative of a uniform effect across the dose range (Table 4-5). NTP (2007) suggested
that increases are consistent with hepatocellular injury or membrane leakage. At 3 months, bile
acids were significantly increased compared to controls at ≥11.2 mg hexavalent chromium/kg-
day in males and in all treatment groups (except 5.9 mg hexavalent chromium/kg-day) in
females; similar to serum liver enzymes, increases in bile acids were not consistently related to
dose. NTP (2007) suggested that increased bile acid was indicative of hepatic toxicity rather
than colestasis, as other markers of colestasis (e.g., alkaline phosphatase [AP] and 5N-
nucleotidase) were not affected by treatment. At 3 months, decreased serum cholesterol and
triglycerides, indicative of altered lipid metabolism, were observed; however, a consistent
relationship between severity and dose was not observed. At 3 months, dose-related increases in
serum creatine kinase activity, indicative of muscle damage, were observed in males and females

at ≥5.9 mg hexavalent chromium/kg-day. Urinalysis showed dose-related decreased volume and
increased specific gravity, consistent with decreased water intake. NTP (2007) suggested that
decreased water intake was due to decreased palatability. Other changes in clinical chemistry
and urinalysis parameters were transient, with no apparent relationship to treatment. Results of
clinical chemistry analyses indicated that exposure of rats to sodium dichromate dihydrate in
drinking water induced hepatocellular membrane damage or cytotoxicity (both sexes) and
increased bile acids (females) at doses ≥1.7 mg hexavalent chromium/kg-day (both sexes).
Table 4-5. Clinical chemistry effects in male and female F344/N rats
exposed to sodium dichromate dihydrate in drinking water for 3 months
Clinical chemistry Time on Treatment group (mg hexavalent chromium/kg-d)

Males
a
ALT (IU/L) 3 Mo 98 ± 6 274 ± 30c 461 ± 102c 447 ± 121c 740 ± 81c 191 ± 17c
(280) (470) (456) (755) (195)
SDH (IU/L) 3 Mo 31 ± 2 55 ± 5c 110 ± 24c 102 ± 24c 173 ± 20c 59 ± 6c
(177) (355) (329) (558) (190)
Bile acids (µmol/L) 3 Mo 22.0 ± 2.2 24.0 ± 3.4 34.5 ± 7.0 32.6 ± 5.3 45.3 ± 2.8c 28.1 ± 2.0c
(109) (157) (148) (206) (128)
Cholesterol (mg/dL) 3 Mo 89 ± 2 95 ± 2 86 ± 4 65 ± 2c 86 ± 3b 71 ± 2c
(107) (97) (73) (97) (80)
Triglycerides 3 Mo 170 ± 9 169 ± 8 172 ± 15 170 ± 13 164 ± 12 98 ± 8c
(mg/dL) (99) (101) (100) (96) (57)
Creatine kinase 3 Mo 214 ± 26 286 ± 32 291 ± 36 364 ± 23c 413 ± 16c 374 ± 44c
(IU/L) (134) (136) (170) (193) (175)
Females
ALT (IU/L) 3 Mo 64 ± 5a 437 ± 68c 218 ± 27c 245 ± 30c 246 ± 37c 248 ± 22c
(683) (340) (383) (384) (387)
c
SDH (IU/L) 3 Mo 22 ± 2 101 ± 17 65 ± 10c 81 ± 13c 96 ± 20c 103 ± 12c
(459) (295) (368) (436) (468)
Bile acids (µmol/L) 3 Mo 19.7 ± 2.5 50.4 ± 6.0c 39.9 ± 4.3c 35.3 ± 3.5 45.3 ± 5.6c 38.7 ± 3.2b
(256) (203) (179) (230) (196)
Cholesterol (mg/dL) 3 Mo 95 ± 2 111 ± 4 94 ± 2 87 ± 2 83 ± 2b 79 ± 2c
(117) (99) (92) (87) (83)
Triglycerides 3 Mo 139 ± 18 116 ± 10 98 ± 9 81 ± 4c 76 ± 7c 59 ± 6c
(mg/dL) (93) (70) (58) (55) (42)
Creatine kinase 3 Mo 197 ± 23 311 ± 94 265 ± 23 296 ± 24c 359 ± 23c 432 ± 48c
(IU/L) (158) (135) (150) (182) (219)
a
exceptions: control group males (n = 9), 1.7 mg hexavalent chromium/kg-d group females (n = 9), and 20.9 mg
hexavalent chromium/kg-d group females (n = 9).
b
c
Source: NTP (2007).

Changes in organ weights in rats exposed to sodium dichromate dihydrate in drinking
water for 3 months are summarized in Table 4-6 (NTP, 2007). Treatment-related effects were
generally observed at doses ≥11.2 mg hexavalent chromium/kg-day. In males, decreases were
observed in absolute and relative liver weights and in absolute and relative spleen weights; in
females, relative right kidney weights and relative spleen weights were increased. Changes in
weights of other organs were considered by NTP (2007) to be secondary to changes in body
weight rather than due to adverse effects of treatment.
Table 4-6. Selected organ weights in male and female F344/N rats exposed
to sodium dichromate dihydrate in drinking water for 3 months
Treatment group (mg hexavalent chromium/kg-d)

Organ 0 1.7 3.5 5.9 11.2 20.9
Males
a
Liver, absolute weight 10.89 ± 0.42 10.30 ± 0.28 11.45 ± 0.38 10.51 ± 0.18 9.20 ± 0.17b 8.88 ± 0.18b
Liver, relative weightc 32.91 ± 0.65 31.91 ± 0.61 33.98 ± 0.75 31.90 ± 0.54 29.15 ± 0.53d 29.80 ± 0.35b
Spleen, absolute weight 0.64 ± 0.02 0.60 ± 0.01 0.62 ± 0.02 0.60 ± 0.02 0.53 ± 0.01b 0.60 ± 0.01b
Spleen, relative weightc 1.94 ± 0.03 1.85 ± 0.03 1.83 ± 0.04 1.81 ± 0.05d 1.69 ± 0.02b 2.00 ± 0.03
Females
a
Right kidney, relative weight 3.34 ± 0.09 3.32 ± 0.04 3.55 ± 0.05 3.55 ± 0.07 3.58 ± 0.10d 3.63 ± 0.09d
Spleen, relative weightc 2.12 ± 0.05 2.04 ± 0.03 2.16 ± 0.05 2.22 ± 0.03 2.25 ± 0.05d 2.39 ± 0.03d
a
Values are means ± SE; n = 10 rats/group.
b
Significantly different (p ≤ 0.01) from the control group by Williams’s or Dunnett’s test.
c
Relative weight = mg organ weight/g body weight.
d
Source: NTP (2007).
Gross and microscopic examinations of male and female rats exposed to sodium
dichromate dihydrate in drinking water for 3 months showed nonneoplastic lesions of the
duodenum, glandular stomach, pancreatic lymph nodes, liver (females only), and bone marrow
(females only) (NTP, 2007); incidence data are summarized in Table 4-7. The incidence of
minimal-to-mild duodenal histiocytic cellular infiltration was increased in males and females at
≥3.5 and ≥1.7 mg hexavalent chromium/kg-day, respectively, compared to controls; incidence
increased with dose. Histiocytic cellular inflammation appeared as multifocal, randomly
scattered, small clusters of enlarged macrophages with pale foamy cytoplasm. Incidences of
nonneoplastic lesions of the glandular stomach (ulcer, focal regenerative hyperplasia, and focal
squamous hyperplasia, all mild-to-moderate in severity) were increased in rats in the highest
dose group. Microscopically, ulcers were characterized by complete loss of the lining of the
mucosal epithelium with necrosis, often extending through to the submucosa, and muscle layers;
mild to marked chronic inflammation (infiltrates of neutrophils, macrophages, lymphocytes, and

eosinophils); and proliferation of fibrous connective tissue through the submucosa. Lesions were
not observed in the forestomach. Microscopic examinations of the oral mucosa and tongue in a
retrospective analysis conducted by NTP revealed nonneoplastic lesions in 4 rats that were not
considered by NTP to be treatment-related. In males, significant increases in the incidence of
minimal histiocytic cellular infiltration of pancreatic lymph nodes were observed at 1.7, 5.9,
11.2, and 20.9 mg hexavalent chromium/kg-day, without a clear dose-response relationship,
whereas significantly increased incidences and severity of pancreatic lymph node sinusoidal
ectasia and lymphoid hyperplasia were only seen at the highest dose; in females, significant
increases in the incidences and severity of nonneoplastic lesions of pancreatic lymph nodes were
only observed at the highest dose. Microscopically, lymphoid hyperplasia was characterized by
minimal-to-mild lymphocyte proliferation, and sinusoid ectasia was characterized by minimal-to-
mild dilatation of the subcapsular or medullary sinuses; histiocytic cellular infiltration was
similar to that observed in the duodenum. Minimal-to-mild histiocytic cellular infiltration was
observed in all groups including control animals. The increase in the incidences in the lower dose
groups was not statistically significant. In the liver of females, a dose-dependent increase in the
incidence of histiocytic cellular infiltration was observed at ≥3.5 mg hexavalent chromium/kg-
day. Minimal chronic inflammation in female liver was observed in three of the control animals
and in a few of the animals at doses up to 11.2 mg/kg-day. The incidence of chronic
inflammation in female rats was not dose dependent; however, the incidence was increased
significantly in the highest dose group. Although serum liver enzymes were statistically
significantly increased in all treatment groups (discussed above), significant histopathological
changes to the livers of male rats were not observed. The incidence of bone marrow hyperplasia
was significantly increased in high-dose females. This observation is consistent with an increase
in hematopoiesis in response to hexavalent chromium-induced microcytic, hypochromic anemia.

Table 4-7. Incidence of nonneoplastic lesions observed in male and female
F344/N rats exposed to sodium dichromate dihydrate in drinking water for
3 months

Tissue (lesion type) 0 1.7 3.5 5.9 11.2 20.9
Males
Duodenum (histiocytic cellular 0/10a 0/10 7/10b 9/10b 8/10b 7/10b
infiltration) (1.1) (1.2) (1.4) (1.4)
Stomach, glandular (ulcer) 0/10 0/10 0/10 0/10 0/10 8/10b
(3.0)
Stomach, glandular (focal 0/10 0/10 0/10 0/10 0/10 10/10b
regenerative hyperplasia) (2.2)
squamous hyperplasia) (2.0) (2.6)
Pancreatic lymph node (ectasia) 0/10 0/10 0/10 0/10 1/10 10/10b
(1.0) (1.7)
Pancreatic lymph node (lymphoid 0/10 0/10 0/10 3/10 3/10 6/10b
hyperplasia) (1.0) (1.0) (2.7)
Pancreatic lymph node (histiocytic 0/10 5/10c 2/10 4/10c 5/10c 9/10b
cellular infiltration) (1.0) (1.0) (1.0) (1.0) (1.9)
Females
Duodenum (histiocytic cellular 0/10a 1/10 5/10c 7/10b 8/10b 10/10b
infiltration) (1.0) (1.0) (1.4) (1.6) (1.7)
Stomach, glandular (ulcer) 0/10 0/10 0/10 0/10 0/10 10/10b
(3.5)
regenerative hyperplasia) (2.0)
squamous hyperplasia) (2.4)
Pancreatic lymph node (ectasia) 0/10 0/10 0/10 0/10 1/10 10/10b
(1.0) (1.8)
Pancreatic lymph node (lymphoid 0/10 0/10 2/10 0/10 0/10 10/10b
hyperplasia) (1.5) (2.1)
Pancreatic lymph node (histiocytic 4/10 8/10 7/10 7/10 7/10 9/10c
cellular infiltration) (1.0) (1.4) (1.7) (1.3) (1.7) (1.9)
Liver (histiocytic cellular 0/10 3/10 6/10b 6/10b 9/10b 8/10b
infiltration) (1.3) (1.0) (1.0) (1.2) (1.0)
Liver (chronic focal inflammation) 3/10 5/10 2/10 7/10 2/10 10/10b
(1.0) (1.0) (1.0) (1.0) (1.0) (1.0)
Bone marrow (hyperplasia) 0/10 0/10 0/10 0/10 0/10 4/10c
(1.0)
a
Number of animals with lesion/number of animals examined; parenthesis indicate average severity grade, with
1 = minimal; 2 = mild; 3 = moderate; 4 = severe.
b
Significantly different (p ≤ 0.01) from the control group by Fisher’s exact test.
c
Source: NTP (2007).

In conclusion, the NTP (2007) 3-month study in F344/N rats exposed to sodium
dichromate dihydrate in drinking water identified several effects of subchronic oral hexavalent
chromium exposure, including changes in hematological endpoints (microcytic, hypochromic
anemia), hepatotoxicity (increased serum enzyme activities, increased serum bile acids, and
histopathological changes), alterations in lipid metabolism (decreased serum cholesterol and
triglycerides), possible muscle damage (increased serum creatine kinase activity), and
histopathological changes in GI tissues (duodenum and glandular stomach) and pancreatic lymph
nodes. EPA used the results of this study to identify a LOAEL in male and female rats of 1.7 mg
hexavalent chromium/kg-day; a NOAEL was not identified. In males, the LOAEL was based on
observations of microcytic, hypochromic anemia (decreased Hct, Hgb, MCV, MCH) observed
after 23 days and 3 months of exposure, increased serum liver enzyme activities (ALT and
SDH), and histopathological changes to pancreatic lymph nodes (histiocytic cellular infiltration),
all observed at daily doses ≥1.7 mg hexavalent chromium/kg-day. In females, the LOAEL was
based on observations of microcytic, hypochromic anemia (decreased Hgb, MCV, MCH)
observed after 23 days and 3 months of exposure, and increased serum liver enzyme activities
(ALT and SDH) and bile acids, all observed at daily doses ≥1.7 mg hexavalent chromium/kg-
day.
In the 3-month study in B6C3F1 mice, groups of 10 males and 10 females were exposed
to sodium dichromate dihydrate in drinking water at concentrations of 0, 62.5, 125, 250, 500, or
1,000 mg sodium dichromate dihydrate/L (equivalent to 0, 21.8, 43.6, 87.2, 174.5, or 348 mg
hexavalent chromium/L, respectively) for 3 months (NTP, 2007). Based on water consumption
monitored throughout the study, NTP (2007) calculated average daily doses over the 3-month
treatment duration of approximately 0, 9, 15, 26, 45, or 80 mg sodium dichromate dihydrate/kg-
day (equivalent to 0, 3.1, 5.3, 9.1, 15.7, or 27.9 mg hexavalent chromium/kg-day, respectively)
for both males and females. Mice were subjected to the same evaluations and procedures as
those described above for “core” study rats (NTP, 2007), except that blood was not analyzed for
clinical chemistry as the study in mice did not include a group of “clinical pathology” animals
for evaluation after exposure durations of 5 and 23 days.
No mortalities were observed in male or female mice exposed to sodium dichromate
dihydrate in drinking water for 3 months (NTP, 2007). Dose-related significant decreases were
observed in final body weights in male mice, with decreases reaching 20% (compared with
control values) in the 27.9 mg hexavalent chromium/kg-day group; in females, dose-related
decreases in final body weight were observed at ≥5.3 mg hexavalent chromium/kg-day, with
decreases reaching 13% in the 27.9 mg hexavalent chromium/kg-day group. Drinking water
consumption was reduced in males at ≥5.3 mg hexavalent chromium/kg-day and in females at
27.9 mg hexavalent chromium/kg-day (statistical significance not reported). Data on food
consumption were not reported. No treatment-related signs of clinical toxicity were observed
throughout the study.

Results of hematological analyses showed that mice exposed to sodium dichromate
dihydrate in drinking water for 3 months developed mild erythrocyte microcytosis (NTP, 2007);
however, compared to hematological effects observed in rats (described above), effects in mice
were less severe. In male mice, MCV and MCH were significantly decreased in all treatment
groups, with maximum decreases of approximately 8%, compared to controls, in the highest dose
group. In females, MCV and MCH were significantly reduced at ≥3.1 and ≥5.2 mg hexavalent
chromium/kg-day, respectively, with maximum decreases of approximately 9 and 10%,
respectively, compared to controls, in the highest dose group. Although statistically significant
(p < 0.05) decreases in MCV were observed in males and females in the 3.1 mg hexavalent
chromium/kg-day group, decreases were very small (1–2%, compared to controls); at doses up to
9.1 mg hexavalent chromium/kg-day, decreases in MCV were ≤5%, compared with controls.
Thus, EPA did not consider the mild microcytosis observed at ≥9.1 mg hexavalent chromium/kg-
day to represent a clinically significant effect. Erythrocyte counts were slightly increased (≤6%,
compared with controls) at ≥5.2 mg hexavalent chromium/kg-day in females, but not in males.
Changes in organ weights in mice exposed to sodium dichromate dihydrate in drinking
water for 3 months are summarized in Table 4-8 (NTP, 2007). In males, absolute liver and right
kidney weights were decreased at ≥9.1 mg hexavalent chromium/kg-day, although the only
significant change in relative organ weight was an increase in relative kidney weight at 27.9 mg
hexavalent chromium/kg-day. In females, absolute liver weight was decreased at doses
≥15.7 mg hexavalent chromium/kg-day. Changes in weights of other organs were considered by
NTP (2007) to be secondary to changes in body weight rather than due to adverse effects of
treatment.
Table 4-8. Selected organ weights in male and female B6C3F1 mice exposed
to sodium dichromate dihydrate in drinking water for 3 months

Organ 0 3.1 5.3 9.1 15.7 27.9
Males
Right kidney, absolute weight 0.28 ± 0.01a 0.28 ± 0.01 0.26 ± 0.01 0.26 ± 0.01b 0.24 ± 0.01c 0.26 ± 0.01c
Right kidney, relative weightd 7.25 ± 0.11 7.68 ± 0.29 7.43 ± 0.35 7.75 ± 0.20 7.76 ± 0.30 8.18 ± 0.07c
Liver, absolute weight 1.60 ± 0.08 1.54 ± 0.05 1.50 ± 0.05 1.40 ± 0.05b 1.33 ± 0.06c 1.34 ± 0.04c
Females
d
Liver, absolute weight 1.15 ± 0.03 1.14 ± 0.04 1.06 ± 0.02 1.11 ± 0.04 1.04 ± 0.02b 0.99 ± 0.02c
a
a
Values are means ± SE; n = 10 mice/group.
b
c
d
Relative weight = mg organ weight/g body weight.
Source: NTP (2007).

Gross and microscopic examinations of male and female mice exposed to sodium
dichromate dihydrate in drinking water for 3 months showed nonneoplastic lesions of the
duodenum and mesenteric lymph nodes (NTP, 2007); incidence data are summarized in
Table 4-9. In the duodenum, dose-related increases were observed in the incidence of minimal-
to-mild histiocytic cellular infiltration in males and females in all treatment groups and in the
incidence of minimal-to-mild epithelial hyperplasia in males and females at ≥5.3 mg hexavalent
chromium/kg-day; a slight dose-related increase in severity was observed. The duodenum had
short, thick duodenal villi, elongated crypts with diffuse hyperplasia, and hyperplastic epithelial
cells with swollen, vacuolated cytoplasm, and increased numbers of “mitotic figures” (incidence
data not reported). NTP (2007) stated that the duodenal lesions were indicative of regenerative
hyperplasia subsequent to epithelial cell injury. Minimal histiocytic cellular infiltration,
morphologically similar to that observed in rats (discussed above), was observed in mesenteric
lymph nodes in male and female mice at ≥5.3 mg hexavalent chromium/kg-day.

B6C3F1 mice exposed to sodium dichromate dihydrate in drinking water for
3 months

Tissue (lesion type) 0 3.1 5.3 9.1 15.7 27.9
Males
Duodenum (histiocytic cellular 0/10a 4/10b 5/10c 10/10c 10/10c 10/10c
infiltration) (1.0) (1.0) (1.3) (1.7) (1.9)
c c
Duodenum (epithelial hyperplasia) 0/10 0/10 8/10 10/10 10/10c 10/10c
(1.3) (1.8) (2.1) (1.8)
b c
Mesenteric lymph node (histiocytic 0/10 0/9 4/9 6/8 3/8 8/10c
cellular infiltration) (1.0) (1.0) (2.0) (1.3)
Females
Duodenum (histiocytic cellular 0/10a 7/10c 8/9c 10/10c 10/10c 10/10c
infiltration) (1.0) (1.3) (1.3) (1.4) (1.7)
Duodenum (epithelial hyperplasia) 0/10 0/10 9/9c 10/10c 10/10c 10/10c
(1.1) (1.1) (1.5) (1.4)
Mesenteric lymph node (histiocytic 0/10 0/10 6/10c 6/10c 4/9b 9/10c
a
b
c
Source: NTP (2007).
In conclusion, the NTP (2007) 3-month study in B6C3F1 mice exposed to sodium
dichromate dihydrate in drinking water identified adverse treatment-related hematological effects
(erythrocytic microcytosis) and histopathological changes to the small intestine (duodenal
epithelial hyperplasia and cellular histiocytic infiltration) and mesenteric lymph nodes (cellular
histiocytic infiltration). Based on histopathological changes (i.e., histiocytic cellular infiltration)
in the duodenum, EPA identified a LOAEL of 3.1 mg hexavalent chromium/kg-day for male and
female mice; in both sexes, a NOAEL was not identified because the effects were observed at the
lowest dose tested. Although a statistically significant decrease in MCV also was observed at
3.1 mg hexavalent chromium/kg-day in males and females, hematological effects (e.g.,
microcytosis) were not considered as the basis of the LOAEL, since decreases in MCV were
small (1–2%) at the lowest dose tested.
Finally, NTP (2007) conducted a comparative study in three strains of mice (B6C3F1,
BALB/c, and am3-C57BL/6) on the effects of exposure to sodium dichromate dihydrate in
drinking water for 3 months. This comparative study was conducted to investigate possible
strain differences in mice based on results of an earlier study reporting hepatotoxicity
(hepatocellular vacuolization) in BALB/c mice fed 32 mg hexavalent chromium/kg-day in the

diet as potassium dichromate (NTP 1996a); no evidence of hepatotoxicity (including
histopathological changes) was observed in male or female B6C3F1 mice exposed for 3 months
to sodium dichromate dihydrate in drinking water at doses up to 20.9 mg hexavalent
chromium/kg-day (NTP, 2007; results summarized above). In the “core” study, groups of
10 male B6C3F1, 10 male BALB/c, and 5 male am3-C57BL/6 mice were exposed to sodium
dichromate dihydrate in drinking water at concentrations of 0, 62.5, 125, or 250 mg/L
(equivalent to 0, 21.8, 43.6, or 87.2 mg hexavalent chromium/L, respectively) for 3 months. The
am3-C57BL/6 strain of mice is transgenic for a gene that is sensitive to forward and reverse
mutation, and an additional five males of this strain were exposed to the same concentrations of
sodium dichromate dihydrate in order to conduct a mutagenesis study exploiting this capability.
However, this study was not conducted due to technical problems, although blood collected from
these animals was still analyzed for hematology and clinical chemistry. Based on water
consumption monitored throughout the study, NTP (2007) calculated average daily doses over
the 3-month treatment duration of approximately 0, 8, 15, or 25 mg sodium dichromate
dihydrate/kg-day (equivalent to 0, 2.8, 5.2, or 8.7 mg hexavalent chromium/kg-day, respectively)
for all strains. Animals were observed twice daily for mortality and clinical signs of toxicity;
body weights were recorded weekly and water consumption was recorded at least every 4 days.
Blood was collected at the end of the 3-month treatment period and analyzed for hematology and
clinical chemistry, as described above for “core” study rats (NTP, 2007). At study termination,
necropsies were performed on all mice, with organ weights recorded for heart, right kidney, liver
(except B6C3F1 mice), lung, spleen, right testis, and thymus. Microscopic examination was
conducted on all gross lesions and masses and selected tissues (liver, forestomach, glandular
stomach, duodenum, pancreas, kidney, and mesenteric and pancreatic lymph nodes). Sperm
count and motility were assessed in all study animals, including spermatids per testis and per mg
testis, spermatids per cauda and per mg cauda, sperm motility, and weights of left cauda, left
epididymis, and left testis.
No mortalities were observed in male B6C3F1, BALB/c, or am3-C57BL/6 mice exposed
to sodium dichromate dihydrate in drinking water for 3 months (NTP, 2007). In the 5.2 and
8.7 mg hexavalent chromium/kg-day groups, final body weights were significantly decreased
(compared to controls) by 9 and 12%, respectively, in B6C3F1 mice and by 7 and 11%,
respectively, in BALB/c mice. Final body weight was reduced in all treatment groups in
am3-C57BL/6 mice, with decreases reaching 44% in the 8.7 mg hexavalent chromium/kg-day
group. Water consumption was reduced at 8.7 mg hexavalent chromium/kg-day in all three
strains. Data on food consumption were not reported. No treatment-related signs of clinical
toxicity were observed in B6C3F1 or am3-C57BL/6 mice. In BALB/c mice, ruffled fur was
observed at 8.7 mg hexavalent chromium/kg-day.
Results of hematology analyses show that male B6C3F1, BALB/c, and am3-C57BL/6
mice exposed to sodium dichromate dihydrate in drinking water for 3 months developed mild

erythrocyte microcytosis (e.g., MCV) and small decreases in MCH, with changes observed in
most treatment groups (Table 4-10) (NTP, 2007). In the 2.8 and 5.2 mg hexavalent
chromium/kg-day groups, decreases in MCV and MCH were ≤7%, compared with controls, with
slightly greater decreases at 8.7 mg hexavalent chromium/kg-day. Erythrocyte counts were
significantly increased in B6C3F1 mice (7% at 8.7 mg hexavalent chromium/kg-day) and in
BALB/c mice (2 and 5% at 5.2 and 8.7 mg hexavalent chromium/kg-day, respectively), but not
in am3-C57BL/6 mice. Hgb and Hct were decreased by approximately 5%, in am3-C57BL/6
mice at 8.7 mg hexavalent chromium/kg-day, compared with controls, but not in B6C3F1 or
BALB/c mice. Compared with hematological effects observed in rats (described previously),
effects in mice were much less severe. Clinical chemistry analysis showed small increases
(1.2- to 1.3-fold) in ALT at ≥5.2 mg hexavalent chromium/kg-day in BALB/c mice and a
1.9-fold increase in ALT in am3-C57BL/6 mice; in B6C3F1 mice, no increases in serum liver
enzyme activities were observed. Decreases in various absolute and relative organ weights were
observed at ≥5.2 mg hexavalent chromium/kg-day. NTP (2007) considered all changes to be
related to decreased body weight, except for a significant decrease (29% compared with controls;
p ≤ 0.05) in absolute thymus weight in B6C3F1 mice in the 8.7 mg hexavalent chromium/kg-day
group; however, relative thymus weight was not different from controls in any treatment group.
No treatment-related effects were observed for reproductive tissue evaluations or other
reproductive parameters, except for a significant decrease (12.4% compared to controls;
p ≤ 0.01) in absolute left testis weight in am3-C57BL/6 mice at 8.7 mg hexavalent chromium/kg-
day; NTP (2007) stated that this change was related to decreased body weight.

Table 4-10. Hematological effects in male B6C3F1, BALB/c, and
am3-C57BL/6 mice exposed to sodium dichromate dihydrate in drinking
water for 3 months

0 2.8 5.2 8.7
B6C3F1 mice
MCV (fL) 47.7 ± 0.2a 46.6 ± 0.2b 46.4 ± 0.2 44.7 ± 0.1
(97.7) (97.3) (93.7)
MCH (ρg) 15.3 ± 0.1 14.9 ± 0.1b 14.7 ± 0.1b 14.2 ± 0.0b
(93.1) (96.1) (92.8)
BALB/c mice
MCV (fL) 44.8 ± 0.2a 43.8 ± 0.2b 42.9 ± 0.2 42.6 ± 0.2
(97.8) (95.8) (95.1)
MCH (ρg) 15.0 ± 0.1 14.5 ± 0.1b 14.2 ± 0.1b 14.0 ± 0.1b
(96.7) (94.7) (93.3)
am3-C57BL/6 mice
MCV (fL) 45.8 ± 0.2 a 44.2 ± 0.4 43.7 ± 0.3b 40.5 ± 0.3
(96.5) (95.4) (88.4)
MCH (ρg) 14.4 ± 0.1 14.1 ± 0.1b 13.8 ± 0.1b 13.5 ± 0.2b
(97.9) (95.8) (98.8)
a
Values are means ± SE; values in parenthesis are percent of control; n = 10 mice/group, with the following
exceptions: in B6C3F1 mice, controls (n = 7), 2.8 mg hexavalent chromium/kg-d group (n = 9), and 5.2 mg
hexavalent chromium/kg-d group (n = 9); in am3-C57BL/6 mice, 8.7 mg hexavalent chromium/kg-d group (n = 9).
b
Source: NTP (2007).
Microscopic examinations of gross lesions and masses and of selected tissues in male
B6C3F1, BALB/c, and am3-C57BL/6 mice exposed to sodium dichromate dihydrate in drinking
water for 3 months showed changes to the duodenum, liver, pancreas, and mesenteric lymph
nodes (NTP, 2007); incidence data are summarized in Table 4-11. Dose-related increases in the
incidences of hepatic glycogen depletion and pancreatic secretory depletion were also observed;
NTP (2007) stated that these lesions were likely due to depressed food consumption, which is
frequently observed when water consumption is decreased. The incidence of minimal-to-mild
histiocytic cellular infiltration of mesenteric lymph nodes was increased at 8.7 mg hexavalent
chromium/kg-day in am3-C57BL/6 mice, but not in B6C3F1 or BALB/c mice.
In the duodenum, dose-related increases in the incidences of minimal-to-mild histiocytic
cellular infiltration and epithelial hyperplasia were observed in all strains, with histopathological
changes of the duodenum observed in all exposure groups; severity increased with dose.
Microscopically, lesions were similar to those described above for male and female B6C3F1
mice. The incidences of histiocytic cellular infiltration and epithelial hyperplasia in the
duodenum, however, were smaller in the initial 3-month study in B6C3F1 mice than in this study.
For example, the incidences of histiocytic cellular infiltration in the initial study with male

B6C3F1 mice were 4/10 and 5/10 at the 3.1 and 5.3 mg/kg-day levels, respectively; the incidence
of epithelial hyperplasia was 0/10 at 3.1 mg/kg-day (Table 4-9). However, as shown in Table 4-
11, at comparable dose levels of 2.8 and 5.2 mg/kg-day, the incidences of histiocytic infiltration
in male B6C3F1 mice in the second study were 8/10 and 10/10, and the incidence of epithelial
hyperplasia at 2.8 mg/kg-day was 4/10. The basis for these inconsistencies in the magnitude of
the duodenal lesions across studies is not known.
Table 4-11. Incidence of nonneoplastic lesions observed in male B6C3F1,

BALB/c, and am3-C57BL/6 mice exposed to sodium dichromate dihydrate in
drinking water for 3 months

Tissue (lesion type) 0 2.8 5.2 8.7
B6C3F1 mice
a
Duodenum (histiocytic cellular 0/10 8/10b 10/10b 10/10b
infiltration) (1.0) (1.4) (2.0)
Duodenum (epithelial hyperplasia) 0/10 4/10c 10/10b 10/10b
(1.0) (1.1) (1.6)
b
Liver (glycogen depletion) 1/10 2/10 9/10 10/10b
(1.0) (1.5) (1.4) (2.2)
b
Pancreas (secretory depletion) 0/10 2/10 7/10 9/10b
(1.0) (1.0) (1.0)
BALB/c mice
a
Duodenum (histiocytic cellular 0/10 4/10c 8/10b 10/10b
infiltration) (1.0) (1.8) (1.7)
Duodenum (epithelial hyperplasia) 0/10 2/10 10/10b 10/10b
(1.0) (1.1) (1.4)
Pancreas (secretory depletion) 0/10 6/10b 9/10b 10/10b
(1.0) (1.3) (1.5)
am3-C57BL/6 mice
Duodenum (histiocytic cellular 0/5a 2/5 5/5b 4/5c
infiltration) (1.0) (1.4) (1.8)
Duodenum (epithelial hyperplasia) 0/5 5/5b 5/5b 5/5b
(1.0) (1.2) (1.8)
Liver (glycogen depletion) 0/5 4/5c 5/5b 5/5b
(2.0) (1.6) (3.8)
Pancreas (secretory depletion) 0/5 3/5 4/5c 5/5b
(1.0) (1.0) (1.6)
Mesenteric lymph node (histiocytic 0/5 0/5 0/5 4/5c
cellular infiltration) (1.5)
a
b
c
Source: NTP (2007).

In conclusion, the comparative 3-month drinking water study on sodium dichromate
dihydrate in male B6C3F1, BALB/c, and am3-C57BL/6 mice showed similar effects in the three
strains (NTP, 2007). A LOAEL of 2.8 mg hexavalent chromium/kg-day was identified by EPA
based on histopathological changes in the duodenum in B6C3F1 mice (i.e., histiocytic cellular
infiltration and epithelial hyperplasia), BALB/c mice (i.e., histiocytic cellular infiltration), and
am3-C57BL/6 mice (i.e., epithelial hyperplasia); a NOAEL was not identified because effects
were observed at the lowest doses tested. Mild erythrocyte microcytosis was not considered as
the basis for the LOAEL because the magnitude of decreases in MCV and MCH in the 2.8 mg
hexavalent chromium/kg-day group was ≤7% compared to controls.
Quinteros et al., 2007

Quinteros et al. (2007) showed that subchronic oral exposure of rats to hexavalent
chromium in drinking water decreased circulating prolactin levels. Groups of 15 male Wistar
rats were exposed to drinking water containing 0 or 500 mg hexavalent chromium/L as
potassium dichromate for 30 days. Based on water intake and body weights measured over the
course of the study, Quinteros et al. (2007) calculated a daily dose of 73.05 mg hexavalent
chromium/kg-day. At the end of the treatment period, blood was collected for analysis of
prolactin and luteinizing hormone (LH), and the pituitary gland and hypothalamus were analyzed
for chromium content. At the end of the 30-day treatment period, water consumption and body
weight in hexavalent chromium-treated rats were decreased by 30.5 and 11.5% compared to
controls. Serum prolactin levels in treated rats were decreased by approximately 59% (p <
0.001) compared to controls; serum levels of LH were comparable in control and treatment
groups. NOAEL and LOAEL values for this study could not be identified because only one dose
was evaluated and effects on other potential hexavalent chromium target tissues were not
assessed.
Rafael et al., 2007

Adverse hepatic effects were reported in rats following subchronic oral exposure to
hexavalent chromium, but further details of this study were not available (Rafael et al., 2007).
Male Wistar rats (9 control and 19 treated) were administered drinking water containing 0 or
20 mg hexavalent chromium/L (chromium compound not reported) for 10 weeks. According to
the investigators, no clinical signs of toxicity or changes in body weight were observed (data not
reported). Data on drinking water consumption were not reported, and the report did not indicate
if drinking water consumption was similar between control and treatment groups; thus, given this
uncertainty, daily hexavalent chromium doses cannot be estimated from this study. At the end of
the treatment period, serum glucose was decreased by 45% (p = 0.0002) and serum ALT activity
was increased by 153% (p = 0.039), compared with controls. Serum levels of total protein,
gamma glutamyl transferase, AP, cholesterol, and total bilirubin were not affected by treatment.

Microscopic examination of livers of treated mice showed increased intracellular space, “little”
focal necrosis, and degenerative alteration with vascularization; fibrosis was not observed.
NOAEL and LOAEL values could not be identified from this study because only one dose was
evaluated and limited exposure information was provided.
Acharya et al., 2001

Acharya et al. (2001) explored whether Wistar rats demonstrated sex-specific responses
to exposures to chromium and chromium plus ethanol using a study design similar to Chopra et
al. (1996), but exposing male rather than female Wistar rats. Acharya et al. (2001) exposed 1.5-
month-old male Wistar rats (5 or 6/group) to potassium dichromate in drinking water for 22
weeks at concentrations of 0 or 25 ppm. These dosed groups were part of a larger study to
evaluate the interactive effects of ethanol and chromium. The authors reported that food and
water consumption were monitored daily and each animal was weighed once a week, although
these results were not reported. Using reference values for body weight and drinking water
consumption (0.217 kg and 0.032 L/day, respectively) for male Wistar rats (U.S. EPA, 1988),
doses of 0 and 1.5 mg hexavalent chromium/kg-day were estimated by EPA. At study
termination, animals were sacrificed and blood samples were collected for analysis of serum
enzyme activities. Liver and kidney tissues were examined for histopathological changes, and
liver homogenates were used to measure total triglycerides, total cholesterol, glycogen, and total
GSH.
Serum aspartate aminotransferase (AST) and ALT levels were statistically significantly
elevated (approximately twofold) in chromium-treated rats compared to controls. Serum
succinate dehydrogenase, AP, and acid phosphatase (AcP) in chromium-treated rats were not
significantly different from the controls. Liver total triglyceride and liver glycogen levels were
significantly reduced in chromium-treated rats (by approximately 40 and 20%, respectively).
There was a significant increase in liver total cholesterol levels (approximately 10%) in
chromium-treated rats. Liver GSH levels in chromium-treated rats were similar to controls.
Histopathological examination of the livers of chromium-treated animals showed altered
hepatic architecture in the periportal area, with increased sinusoidal space, vacuolation, and
necrosis. Histopathological examination of the kidneys in chromium-treated rats revealed
vacuolation in glomeruli, degeneration of the basement membrane, and renal tubular epithelial
degeneration. No information regarding the number of animals examined or the number of
animals displaying histopathology was provided. The only dose tested in this study, 1.5 mg
hexavalent chromium/kg-day, was identified by EPA as a LOAEL. A NOAEL was not
identified because effects were seen at the only dose tested.

Asmatullah and Noreen, 1999
Asmatullah and Noreen (1999) studied the effects of subchronic exposure to hexavalent
chromium on growth rate and hepatic histological structure in mice. Groups of male albino
Swiss mice (nine per group) were exposed to drinking water containing 0, 500, 750, 1,000,
1,500, or 2,000 mg potassium dichromate/L (equivalent to 0, 177, 265, 353, 530, or 706 mg
hexavalent chromium/L, respectively) for 8 weeks. Data on drinking water consumption were
not reported; based on findings of other studies (NTP, 2008, 2007) showing decreased drinking
water consumption and body weight in animals treated with drinking water containing ≥30 mg
hexavalent chromium/L, daily doses of hexavalent chromium cannot be accurately estimated for
this study. Body weights and feed consumption were recorded weekly. At the end of the
treatment period, organ weights were determined for liver, heart, and kidney, and microscopic
examination of the liver was conducted. During the last 2 weeks of treatment, body weights
were decreased in all treatment groups, with decreases ranging from 9 to 29%, compared with
controls; decreases in body weight were accompanied by similar decreases in feed intake in all
treatment groups. After 8 weeks of treatment, absolute wet and dry weights of liver and heart
were increased in all treatment groups, although the magnitude of these increases did not exhibit
dose-dependence. No consistent pattern of change was observed for wet or dry weight of the
heart. Relative organ weights were not reported. Histopathological changes in the liver were
observed, with severity increasing with dose (but incidence data were not reported). At 265 mg
hexavalent chromium/L, an increase in the sinusoidal space was observed; at 353 and 530 mg
hexavalent chromium/L, hepatic cirrhosis and increased sinusoidal space were observed, with
severity increasing with dose; and at 706 mg hexavalent chromium/L, increased sinusoidal space,
cirrhosis and nuclear pyknosis (a marker for apoptosis) were observed. Results of microscopic
examination of the liver in mice treated with 177 mg hexavalent chromium/L were not reported.
A NOAEL or LOAEL could not be identified by EPA from this study.
Chopra et al., 1996

Chopra et al. (1996) exposed 50-day-old female Wistar rats (5 or 6/group) to potassium
dichromate in drinking water for 22 weeks at concentrations of 0 or 25 ppm. These dose groups
were part of a larger study designed to evaluate the interactive effects of ethanol and chromium.
The authors reported that food and water consumption were monitored daily and each animal
was weighed once a week, although these results were not reported. Using reference values for
body weight and drinking water consumption (0.156 kg and 0.025 L/day, respectively) for
female Wistar rats (U.S. EPA, 1988), doses of 0 and 1.4 mg hexavalent chromium/kg-day were
estimated by EPA. At study termination, animals were sacrificed and blood samples were
collected for analysis of serum enzyme activities and serum triglycerides, cholesterol, and
glucose. A kidney homogenate was used to measure GSH, and a liver homogenate was used to

measure triglycerides, cholesterol, glycogen, GSH, and lipid peroxidation. Liver and kidney
tissues were examined for histopathological changes.
Terminal body weights in chromium-treated rats were not significantly different from the
controls. The liver to body weight ratio in chromium-treated rats was statistically significantly
increased (approximately twofold) over the controls. Serum SDH levels were significantly lower
(by approximately 20%) in chromium-treated rats compared to the controls, whereas AST, ALT,
AP, and AcP were statistically significantly increased (approximately two- to threefold). Serum
triglycerides and glucose were statistically significantly increased (approximately threefold) in
chromium-treated rats; serum cholesterol was significantly reduced (approximately twofold).
Analysis of liver homogenates revealed that chromium treatment resulted in reduced liver
glycogen (by approximately twofold); levels of liver cholesterol, GSH, and lipid peroxidation (as
measured by diene conjugation) did not differ from the controls. Kidney GSH in chromium-
treated rats was statistically significantly lower than the controls (approximately 2.5-fold).
Histopathological examination of the liver of chromium-treated animals showed altered
hepatic architecture in the periportal area, with increased sinusoidal space, vacuolation, and
necrosis. Histopathological examination of the kidneys in chromium-treated rats revealed
significant damage to renal tubules and the Bowman’s capsule and degeneration of the basement
membrane. No information regarding the number of animals examined or the number of animals
displaying histopathology was provided. The only dose tested in this study, 1.4 mg hexavalent
chromium/kg-day, was identified as a LOAEL by EPA. A NOAEL was not identified because
effects were observed at the only dose tested.
Vyskocil et al., 1993

Alterations in renal function, as assessed by urinalysis, were observed in rats exposed to
oral potassium chromate for up to 6 months (Vyskocil et al., 1993). Groups of Wistar rats
(20/sex/group) were exposed to drinking water containing 0 or 25 mg hexavalent chromium/L.
Based on water consumption, which was comparable between control and treatment groups,
Vyskocil et al. (1993) calculated average daily hexavalent chromium doses of 2.18 and 2.47 mg
hexavalent chromium/kg-day in males and females, respectively, during the first 3 months of
exposure, and 1.40 and 1.76 mg hexavalent chromium/kg-day in males and females,
respectively, during the second 3 months of exposure. After 3 or 6 months of exposure, urine
was collected from 10 rats/sex/group and analyzed for total protein, albumin, β2-microglobulin,
β-N-acetyl-D-glucosamine, and lactate dehydrogenase and lysozyme activities, and body and
kidney weights were determined. Water consumption was monitored throughout the study. No
effects on body weight gain or kidney weight were observed. In male rats, results of urinalysis
did not show any treatment-related effects. In females, urinary albumin excretion, a marker of
glomerular function, was significantly increased by approximately twofold (p < 0.05), compared
to controls, at both 3 months and 6 months. Urinary β2-microglobulin, a marker of renal tubular

dysfunction, was increased by twofold (p < 0.05) at 3 months and by 1.4-fold at 6 months (not
statistically significant) compared to controls. Gross or microscopic examinations of kidneys
were not conducted. NOAEL and LOAEL values from this study could not be identified because
only one dose was evaluated and effects on other potential hexavalent chromium target tissues
were not assessed.
4.2.2. Chronic Oral Exposure

NTP, 2008
NTP (2008) conducted a 2-year toxicology and carcinogenicity study of sodium
dichromate dihydrate in drinking water in rats and mice. Groups of F344/N rats (“core” study
animals; 50/sex/group) were exposed to sodium dichromate dihydrate in drinking water at
concentrations of 0, 14.3, 57.3, 172, or 516 mg sodium dichromate dihydrate/L (equivalent to 0,
5, 20, 60, or 180 mg hexavalent chromium/L, respectively). Based on water consumption
measured throughout the study, NTP (2008) calculated average daily doses over the 2-year
treatment duration of approximately 0, 0.6, 2.2, 6, or 17 mg sodium dichromate dihydrate/kg-day
for males (equivalent to 0, 0.21, 0.77, 2.1, or 5.9 mg hexavalent chromium/kg-day, respectively)
and 0.7, 2.7, 7, and 20 mg sodium dichromate dihydrate/kg-day for females (equivalent to 0,
0.24, 0.94, 2.4, or 7.0 mg hexavalent chromium/kg-day, respectively). Animals were observed
twice daily for mortality and clinical signs of toxicity; after 5 weeks of treatment, clinical signs
were recorded at 4-week intervals. Body weights were recorded weekly for the first 13 weeks,
and then at 4-week intervals for the duration of the study. Water consumption was recorded
weekly for the first 13 weeks of treatment and then every 4 weeks. At the end of the 2-year
treatment period, complete necropsies and microscopic examinations of comprehensive tissues
were performed on all “core” study animals. An additional “special study” group of male rats
(10/group) was exposed to the same drinking water concentrations as “core” animals for up to
53 weeks. For the “special study” rats only, blood was collected on days 4 and 22 and at 3, 6,
and 12 months for hematology (i.e., Hct; Hgb concentration; erythrocyte, reticulocyte, and
platelet counts; erythrocyte and platelet morphology; MCV; MCH; mean cell hemoglobin
concentration [MCHC]; and leukocyte count and differentials) and clinical chemistry (i.e., urea
nitrogen, creatinine, total protein, albumin, ALT, AP, creatine kinase, sorbitol dehydrogenase,
bile acids) analyses. At the end of the 53-week treatment period, “special study” animals were
evaluated for chromium tissue distribution (see Section 3.2 for the results of this study).
Survival rates of exposed “core” study rats were similar to controls (NTP, 2008).
Throughout the study, water consumption was decreased in the two highest dose groups
compared to controls. During the second year of the study, water consumption in the two highest
dose groups in males was decreased by 15 and 22%, respectively, and by 15 and 27%,
respectively, in females (statistical significance not reported). No data on food consumption
were reported. At the end of the 2-year treatment period, body weight was decreased in males

and females in the highest dose group by 12 and 11%, respectively, compared with controls
(statistical significance not reported). NTP (2008) suggested that decreased body weights in the
highest dose group may have been partially due to decreased water consumption (due to
decreased palatability), rather than an adverse effect of sodium dichromate dihydrate. No
treatment-related signs of clinical toxicity were observed throughout the study.
Results of hematologic analyses in “special study” male rats showed that exposure to
sodium dichromate dihydrate in drinking water produced microcytic, hypochromic anemia,
characterized by decreases in MCV, Hct, Hgb, MCH, and MCHC (NTP, 2008). The severity of
microcytic, hypochromic anemia exhibited duration- and dose-dependence, with peak effects
occurring at 22 days (Table 4-12). After 4 days of exposure, small changes were observed in
several hematological parameters; however, decreases in all treatment groups were ≤5%,
compared to controls. More severe effects were observed after 22 days of treatment, with
significant decreases in MCV, Hct, and Hgb at ≥0.77 mg hexavalent chromium/kg-day. At
5.9 mg hexavalent chromium/kg-day, MCV, Hct, and Hgb decreased to approximately 76, 73,
and 65% of control values, respectively; reticulocyte and nucleated erythrocyte counts were
increased by approximately 66% (p ≤ 0.01) and 600% (p ≤ 0.01), respectively, compared to
controls, indicating compensatory hematopoiesis. Blood smears showed evidence of erythrocyte
injury or increased turnover, including poikilocytes, erythrocyte fragments, and keratocytes
(incidence data not reported). Similar effects were observed after 3 months of treatment,
although severity at 3 months was generally less than that observed at 22 days. Severity was
further decreased after 6 and 12 months of exposure; at 12 months, affected parameters were
generally only decreased by ≤5%, compared to controls. Results of hematological analyses show
that exposure of rats to sodium dichromate dihydrate in drinking water produced microcytic,
hypochromic anemia at subchronic exposure durations (22 days to 3 months), but that severity
decreased with increasing exposure duration (6–12 months).
Table 4-12. Hematological effects in male F344/N rats exposed to sodium

dichromate dihydrate in drinking water for up to 12 months

parameter treatment 0 0.21 0.77 2.1 5.9
a b b
MCV (fL) D 22 59.5 ± 0.4 58.6 ± 0.5 54.9 ± 0.5 47.4 ± 0.4 45.0 ± 0.7b
(98.5) (92.3) (80.0) (75.6)
Mo 3 48.6 ± 0.2 48.3 ± 0.2 47.3 ± 0.2b 45.7 ± 0.2b 39.2 ± 0.6b
(99.4) (97.3) (94.0) (80.7)
b b
Mo 6 49.8 ± 0.1 49.5 ± 0.1 48.6 ± 0.1 47.8 ± 0.2 45.4 ± 0.5b
(99.4) (97.6) (96.0) (91.2)
b
Mo 12 52.6 ± 0.2 52.4 ± 0.2 51.9 ± 0.3 51.4 ± 0.3 49.9 ± 0.2b
(99.6) (98.7) (97.7) (94.9)
Hct (%) D 22 46.0 ± 1.1 44.4 ± 0.4 43.2 ± 0.6c 38.7 ± 0.6b 33.5 ± 0.8b
(96.5) (93.9) (84.1) (72.8)

Table 4-12. Hematological effects in male F344/N rats exposed to sodium

Mo 3 45.3 ± 0.4 44.5 ± 0.3 44.5 ± 0.4 44.1 ± 0.5 41.0 ± 0.5b
(98.2) (98.2) (97.4) (90.5)
Mo 6 45.9 ± 0.4 45.7 ± 0.5 45.5 ± 0.4 45.5 ± 0.5 45.0 ± 0.3
(99.6) (99.1) (99.1) (98.0)
Mo 12 47.6 ± 0.5 46.6 ± 0.4 47.4 ± 0.5 47.7 ± 0.4 47.3 ± 0.4
(97.9) (99.6) (100.2) (99.4)
Hgb (g/dL) D 22 15.5 ± 0.3 15.1 ± 0.2 14.2 ± 0.2b 12.0 ± 0.3b 10.1 ± 0.2b
(97.4) (91.6) (77.4) (65.2)
Mo 3 15.1 ± 0.1 14.9 ± 0.1 14.9 ± 0.2 14.6 ± 0.2c 12.9 ± 0.2b
(98.7) (98.7) (96.7) (85.4)
Mo 6 15.2 ± 0.1 15.2 ± 0.2 15.0 ± 0.2 14.9 ± 0.1 14.5 ± 0.1b
(100) (98.7) (98.0) (95.4)
Mo 12 15.8 ± 0.2 15.4 ± 0.2 15.6 ± 0.2 15.6 ± 0.2 15.3 ± 0.1c
(97.5) (98.7) (98.7) (96.8)
MCH (ρg) D 22 19.8 ± 0.1 19.5 ± 0.2 17.7 ± 0.2b 14.8 ± 0.2b 16.3 ± 0.5b
(98.5) (89.4) (74.7) (82.3)
Mo 3 16.2 ± 0.1 16.2 ± 0.1 15.7 ± 0.0b 15.0 ± 0.1b 11.9 ± 0.3b
(100) (96.9) (92.6) (73.5)
Mo 6 16.3 ± 0.1 16.1 ± 0.1 15.7 ± 0.1b 15.3 ± 0.1b 14.3 ± 0.2b
(98.8) (96.3) (93.9) (87.7)
Mo 12 17.0 ± 0.1 16.8 ± 0.1 16.6 ± 0.1c 16.2 ± 0.1b 15.7 ± 0.1b
(98.8) (97.6) (95.3) (92.4)
MCHC (g/dL) D 22 33.3 ± 0.1 33.3 ± 0.1 32.2 ± 0.2 31.2 ± 0.2b 36.2 ± 0.8
(100) (96.7) (93.7) (108.7)
Mo 3 33.4 ± 0.1 33.5 ± 0.2 33.2 ± 0.1 32.7 ± 0.1b 30.2 ± 0.3b
(100.3) (99.4) (97.9) (90.4)
Mo 6 32.7 ± 0.1 32.5 ± 0.1 32.3 ± 0.1c 32.1 ± 0.1b 31.6 ± 0.2b
(99.4) (98.8) (98.2) (96.6)
Mo 12 32.3 ± 0.2 32.1 ± 0.3 32.0 ± 0.2 31.6 ± 0.2c 31.5 ± 0.2c
(99.4) (99.1) (97.8) (97.5)
Erythrocyte count D 22 7.80 ± 0.13 7.74 ± 0.15 8.06 ± 0.16 8.10 ± 0.14 6.21 ± 0.13b
(106/µL) (99.2) (103.3) (103.8) (79.6)
Mo 3 9.28 ± 0.05 9.24 ± 0.06 9.46 ± 0.11 9.75 ± 0.11b 10.93 ± 0.16b
(99.6) (101.9) (105.1) (117.7)
Mo 6 9.34 ± 0.06 9.43 ± 0.08 9.54 ± 0.11 9.71 ± 0.08b 10.15 ± 0.13b
(101.0) (102.1) (104.0) (108.7)
Mo 12 9.27 ± 0.10 9.17 ± 0.07 9.40 ± 0.12 9.61 ± 0.11 9.74 ± 0.08b
(98.9) (101.4) (103.7) (105.1)
a
exceptions: control group on d 4 (n = 9), 0.77 mg hexavalent chromium/kg-d group on d 4 (n = 9), and 2.1 mg
hexavalent chromium/kg-d group in mo 12 (n = 8).
b
c
Source: NTP (2007).

Results of clinical chemistry analyses in “special study” male rats (clinical chemistry was
not assessed in female rats) showed that exposure to sodium dichromate dihydrate in drinking
water produced dose-dependent increases in serum ALT activity (NTP, 2008). Significant
increases in serum ALT activity were observed at 4 days and 6 months in rats treated with
≥2.1 mg hexavalent chromium/kg-day and at 22 days and 3 and 12 months at ≥0.77 mg
hexavalent chromium/kg-day (Table 4-13). Serum ALT enzyme activity reached maximum
increases (approximately 170–260% of control values) in rats treated for 3–12 months at daily
doses of ≥2.1 mg hexavalent chromium/kg-day. In rats treated for 12 months with 2.1 and
5.9 mg hexavalent chromium/kg-day, serum SDH activity was 164 and 173% of control values,
respectively; however, no increases in SDH activity were observed at other doses or time points.
No increases in serum AP activity were observed in any treatment group throughout the
12-month treatment period. Increased serum ALT activity is consistent with histopathological
findings of minimal chronic inflammation of the liver observed in “core” study animals
(discussed below); however, because other clinical chemistry markers of hepatic damage were
not observed, NTP (2008) suggested that increased serum ALT activity may reflect enzyme
induction rather than hepatocellular damage. Changes in other clinical chemistry outcomes were
generally <5% compared to controls and did not exhibit dose- or duration-dependence.
Table 4-13. Serum ALT activity in male F344/N rats exposed to sodium

Time on treatment 0 0.21 0.77 2.1 5.9
D4 54 ± 2a 53 ± 2 60 ± 3 68 ± 1b 70 ± 2b
(98) (113) (126) (130)
b
D 22 45 ± 1 46 ± 1 58 ± 2 75 ± 3b 73 ± 4b
(102) (129) (167) (162)
c
Mo 3 82 ± 4 82 ± 12 135 ± 18 176 ± 13b 216 ± 21b
(100) (165) (215) (263)
Mo 6 122 ± 15 114 ± 9 150 ± 12 238 ± 2b 210 ± 12b
(93) (123) (195) (172)
c
Mo 12 102 ± 6 107 ± 8 135 ± 10 261 ± 23b 223 ± 15b
(105) (132) (256) (219)
a
exception of 0.77 mg hexavalent chromium/kg-d group on d 4 (n = 9). Note: clinical chemistry was not assessed in
female rats.
b
c
Source: NTP (2008).

Gross and microscopic examinations of “core” study rats exposed to sodium dichromate
dihydrate in drinking water for 2 years showed nonneoplastic lesions of the small intestine
(duodenum), liver, and lymph nodes in both sexes, nonneoplastic lesions of the salivary gland in
females, and neoplastic lesions of the oral cavity in both sexes (NTP, 2008). Incidence data for
nonneoplastic lesions are summarized in Table 4-14. The incidence of minimal-to-mild cellular
histiocytic infiltration of the duodenum was significantly increased in males and females at
≥0.77 and ≥2.4 mg hexavalent chromium/kg-day, respectively, compared with controls;
increases in both sexes were dose-related. Duodenal histiocytic infiltrate was characterized by
single or clusters of macrophages in the lamina propria of the duodenal villi. Based on incidence
data, males appeared more sensitive than females to hexavalent chromium-induced
nonneoplastic changes to the small intestine.

2 years

Tissue (lesion type) 0 0.21 0.77 2.1 5.9
Males
Liver (histiocytic cellular infiltration) 1/50a 0/50 2/49 5/50 34/49b
(1.0) (1.0) (1.4) (1.4)
c
Liver (chronic inflammation) 19/50 25/50 21/49 28/50 26/49
(1.1) (1.2) (1.3) (1.1) (1.3)
c c
Liver (basophilic focus) 22/50 28/50 29/49 32/50 30/49
c b
Small intestine, duodenum (histiocytic 0/48 0/48 6/47 36/46 47/48b
cellular infiltration) (1.2) (1.1) (1.5)
Lymph node, mesenteric (histiocytic 13/49 11/50 30/49 39/50b 41/49 b
c c
Lymph node, mesenteric (hemorrhage) 2/49 7/50 9/49 8/50 17/49b
(1.5) (1.1) (1.3) (1.1) (1.3)
Females
a
Liver (histiocytic cellular infiltration) 1/50 5/50 21/50b 42/50b 47/50b
(1.0) (1.0) (1.3) (2.0) (2.6)
Liver (chronic inflammation) 12/50 21/50c 28/50b 35/50b 39/50b
(1.3) (1.2) (1.3) (1.6) (2.1)
Liver (fatty change) 3/50 7/50 10/50c 13/50b 16/50b
(3.3) (3.6) (2.5) (2.5) (2.8)
Liver (clear cell focus) 7/50 5/50 7/50 20/50b 7/50
Small intestine, duodenum (histiocytic 0/46 0/49 1/48 30/46b 47/50b
Lymph node, mesenteric (histiocytic 21/50 18/50 27/50 36/50b 42/50b
Lymph node, mesenteric (hemorrhage) 11/50 13/50 16/50 14/50 21/50c
(1.1) (1.3) (1.3) (1.1) (1.3)
Lymph node, pancreatic (histiocytic 17/29 20/36 23/30 32/34b 27/33
Salivary gland (atrophy) 9/50 7/50 10/50 17/50c 17/50
(1.3) (1.4) (1.2) (1.4) (2.1)
a
b
Significantly different (p ≤ 0.01) from the control group by the Poly-3 test.
c
Source: NTP (2008).

Significant findings in the liver included histiocytic cellular infiltration, chronic
inflammation, fatty change, basophilic foci, and clear cell foci. The incidence of histiocytic
cellular inflammation, which was mild-to-moderate in severity and characterized by clusters of
macrophages in parenchymal and portal areas, was significantly increased in males and females
at 5.9 and ≥0.94 mg hexavalent chromium/kg-day, respectively (Table 4-14); in females,
increases in incidence and severity were dose-dependent. Increased minimal-to-mild hepatic
inflammation was observed in males at 2.1 mg hexavalent chromium/kg-day and in females in
all treatment groups, with dose-dependent increases in incidence and severity in females.
NTP (2008) noted that chronic inflammation is a typical hepatic lesion observed in aged rats;
however, exposure to sodium dichromate dihydrate appeared to enhance development of this
lesion. An increase in the incidence of mild-to-moderate fatty change was observed only in
females at ≥0.94 mg hexavalent chromium/kg-day. Morphologically, fatty change was
characterized by hepatocytes with fat-containing cytoplasmic vacuoles. The incidence of
basophilic foci was increased in males only at 0.77 and 2.1 mg hexavalent chromium/kg-day,
and the incidence of clear cell foci was increased in females at 2.4 mg hexavalent chromium/kg-
day. Based on the dose-response data for histopathological changes of the liver, female rats
appear more sensitive to hexavalent chromium than male rats to hepatic effects of sodium
dichromate dihydrate.
In lymph nodes, lesions were observed in mesenteric lymph nodes (histiocytic cellular
infiltration and hemorrhage) in both sexes and in pancreatic lymph nodes (histiocytic cellular
infiltration) in females only. The incidence of histiocytic cellular infiltration in mesenteric
lymph nodes was significantly elevated in both sexes at the two highest doses (2.1 and 5.9 mg
hexavalent chromium/kg-day in males and 2.4 and 7.0 mg hexavalent chromium/kg-day in
females). The incidence of mesenteric lymph node hemorrhage was significantly increased in
males at ≥0.77 mg hexavalent chromium/kg-day (i.e., the three highest doses) and in females at
7.0 mg hexavalent chromium/kg-day (i.e., the highest dose). In males, the severity of histiocytic
cellular infiltration and hemorrhage of mesenteric lymph nodes was minimal-to-mild in all
groups, but severity of histiocytic cellular infiltration was slightly increased at ≥2.4 mg
hexavalent chromium/kg-day. The incidence of cellular histiocytic infiltration of pancreatic
lymph nodes was significantly increased in females in the 2.4 mg hexavalent chromium/kg-day
group only, with severity increased at ≥0.94 mg hexavalent chromium/kg-day group.
Morphologically, histiocytic cellular infiltrate of the lymph nodes was similar to that observed in
the liver, with random clusters of macrophages located in the cortex and medullary sinuses; in
mesenteric lymph nodes, some clusters merged to form sheets that replaced the parenchyma.
NTP (2008) suggested that mesenteric lymph node hemorrhage may have resulted from
histiocytic infiltration. A significant increase in the incidence of minimal-to-mild salivary gland
atrophy, appearing as single focal lesions, was observed in females in the 2.4 mg hexavalent
chromium/kg-day group only, compared with controls. NTP (2008) noted that atrophy is an age-

related change commonly observed in rats and that the biological significance of salivary atrophy
in female rats chronically treated with 2.4 mg hexavalent chromium/kg-day group is unknown.
Incidence data for neoplastic lesions of the oral cavity in male and female rats exposed to
sodium dichromate dihydrate in drinking water for 2 years are summarized in Table 4-15 (NTP,
2008). Neoplasms observed in the oral cavity of treated rats were squamous cell carcinoma of
the oral mucosa (both sexes), squamous cell papilloma of the oral mucosa (males only),
squamous cell carcinoma of the tongue (both sexes), and squamous cell papilloma and carcinoma
of the tongue (both sexes). The incidences of squamous cell carcinoma of the oral mucosa
(13.6%) and of combined squamous cell papilloma or carcinoma (15.7%) of the oral mucosa
were significantly increased in male rats treated with 5.9 mg hexavalent chromium/kg-day,
compared with controls. The incidences of squamous cell carcinoma of the oral mucosa (23.9%)
and of combined squamous cell carcinoma of the oral mucosa or tongue (23.9%) were
significantly increased in females treated with 7.0 mg hexavalent chromium/kg-day, compared
with controls. The incidences of other neoplastic lesions of the oral cavity were not significantly
increased in any treatment group in males or females compared with controls, although the
incidence of squamous cell carcinoma of the oral mucosa in female rats in the 2.4 mg hexavalent
chromium/kg-day group (4.6%) exceeded that of historical controls (0/300 in drinking water
studies; 5/1,400 by all routes). Other neoplasms observed in treated rats included pancreatic
acinar adenoma and benign pheochromocytomas in males and mononuclear cell leukemia in
females (Table 4-16). However, the incidence of these neoplasms did not exhibit dose-
dependence. Thus, NTP (2008) concluded that the relationship of neoplastic changes in other
tissues (e.g., not of the oral cavity) to exposure to sodium dichromate dihydrate was uncertain.

Table 4-15. Incidence of neoplastic lesions observed in the oral cavity of
male and female F344/N rats exposed to sodium dichromate dihydrate in
drinking water for 2 years

Neoplasm type 0 0.21 0.77 2.1 5.9
Males
Oral mucosa, squamous cell papilloma
Overall ratea,b 0/50 0/50 0/49 0/50 1/49
(0%) (0%) (0%) (0%) (2%)
Oral mucosa, squamous cell carcinoma
Overall ratea 0/50 0/50 0/49 0/50 6/49
(0%) (0%) (0%) (0%) (12%) [543]
Adjusted ratec 0% 0% 0% 0% 13.6%
p < 0.001 p = 0.015
Tongue, squamous cell papilloma
Overall ratea,b 0/50 0/50 0/49 0/50 1/49
(0%) (0%) (0%) (0%) (2%)
Tongue, squamous cell carcinoma
Overall ratea 0/50 1/50 0/49 0/50 0/49
(0%) (2%) (0%) (0%) (0%)
Oral mucosa or tongue, squamous cell papilloma or carcinoma
Overall ratea 0/50 1/50 0/49 0/50 7/49
(0%) (2%) [729T] (0%) (0%) (14.5%) [543]
Adjusted ratec 0% 2.4% 0% 0% 15.7%
p < 0.001 p = 0.007
Neoplasm type 0 0.24 0.94 2.4 7.0
Females
Oral mucosa, squamous cell carcinoma
Overall ratea 0/50 0/50 0/50 2/50 11/50
(0%) (0%) (0%) (4%) [646] (22%) [506]
Adjusted ratec 0% 0% 0% 4.6% 23.9%
p < 0.001 p < 0.001
Tongue, squamous cell papilloma
Overall ratea,b 1/50 1/50 0/50 0/50 0/50
(2%) (2%) (0%) (0%) (0%)
Tongue, squamous cell carcinoma
Overall ratea,b 0/50 0/50 0/50 1/50 0/50
(0%) (0%) (0%) (2%) (0%)

Table 4-15. Incidence of neoplastic lesions observed in the oral cavity of
male and female F344/N rats exposed to sodium dichromate dihydrate in

Neoplasm type 0 0.24 0.94 2.4 7.0
Females
Oral mucosa or tongue, squamous cell papilloma or carcinoma
Overall ratea 1/50 1/50 0/50 2/50 11/50
(2%) [618] (2%) [729T] (0%) (4%) [646] (22%) [506]
Adjusted ratec 2.2% 2.3% 0% 4.6% 23.9%
p < 0.001 p = 0.002
a
Overall rate: number of animals with lesion/number of animals examined; parenthesis are the percent of animals
examined with lesion; brackets are days to first incidence; T: observed at terminal sacrifice. p-Value under
treatment group incidence data indicates statistically significant Poly-3 test for pairwise comparison between
control and exposed group. Statistical analysis using overall rates was only conducted if adjusted rates were not
determined.
b
Adjusted rate not reported.
c
Adjusted rate: Poly-3 estimated neoplasm incidence (expressed as percent of animals with neoplasm) adjusted for
intercurrent mortality. p-Value under control group indicates statistically significant positive Poly-3 trend test.
p-Value under treatment group incidence data indicates statistically significant Poly-3 test for pairwise comparison
between control and exposed groups, using adjusted rates.
Source: NTP (2008).
Table 4-16. Neoplastic lesions in other tissues (e.g., nonoral cavity) in

2 years

Neoplasm type 0 0.21 0.77 2.1 5.9
Males
Pancreatic acinar adenoma 1/50a,b 2/50 6/49c 2/50 2/49
a,b c
Benign pheochromocytoma (adrenal medulla) 6/49 13/50 14/49c 5/50 4/49
Neoplasm type 0 0.24 0.94 2.4 7.0
Females
Mononuclear cell leukemia 8/50a,b 18/50c 13/50 7/50 11/50
a
Number of animals with lesion/number of animals examined.
b
Not statistically significant for positive trend (p > 0.05) by the Poly-3 test.
c
Significantly different from controls by the Poly-3 test (p < 0.05).
Source: NTP (2008).
In conclusion, from the NTP (2008) 2-year drinking water toxicology and carcinogenicity
study on sodium dichromate dihydrate, EPA identified NOAEL and LOAEL values for
noncancer effects in male rats of 0.21 and 0.77 mg hexavalent chromium/kg-day, respectively,

based on increased incidences of nonneoplastic histopathological changes to the liver (basophilic
foci), duodenum (histiocytic cellular infiltrate), and mesenteric lymph nodes (histiocytic cellular
infiltrate and hemorrhage). Although hematological effects indicative of microcytic,
hypochromic anemia were observed in male rats exposed to ≥0.77 mg hexavalent chromium/kg-
day from 4 days to 6 months, the severity of effects decreased over time, such that only small
changes (<5%) were observed at ≥2.1 mg hexavalent chromium/kg-day after 12 months of
exposure; therefore, hematological effects were not considered by EPA as the basis for the
chronic NOAEL in male rats. In female rats, a LOAEL for noncancer effects of 0.24 mg
hexavalent chromium/kg-day was identified by EPA based on the increased incidence of chronic
inflammation of the liver (observed in all treatment groups); a NOAEL was not identified
because these liver effects were seen at the lowest dose tested. In addition to noncancer effects,
exposure of rats to sodium dichromate dihydrate in drinking water for 2 years resulted in a
significant increase in squamous epithelial neoplasms of the oral mucosa and tongue at the
highest exposure level (average daily doses of 5.9 and 7.0 mg hexavalent chromium/kg-day in
males and females, respectively), but not at the three lower exposure levels. NTP (2008)
concluded that results from this study provide clear evidence of carcinogenic activity of sodium
dichromate dihydrate in male and female F344/N rats based on increased incidences of
squamous cell neoplasms of the oral cavity.
B6C3F1 mice were exposed to sodium dichromate dihydrate in drinking water for up to
2 years (NTP, 2008). Groups of 50 male mice (male “core” study animals) were exposed to
sodium dichromate dihydrate in drinking water at concentrations of 0, 14.3, 28.6, 85.7, or
257.4 mg sodium dichromate dihydrate/L (equivalent to 0, 5, 10, 30, or 90 mg hexavalent
chromium/L, respectively). Based on water consumption measured throughout the study,
NTP (2008) calculated average daily doses for males over the 2-year treatment duration of
approximately 0, 1.1, 2.6, 7, or 17 mg sodium dichromate dihydrate/kg-day (equivalent to 0,
0.38, 0.91, 2.4, or 5.9 mg hexavalent chromium/kg-day, respectively). Groups of 50 female mice
(female “core” study animals) were exposed to sodium dichromate dihydrate in drinking water at
concentrations of 0, 14.3, 57.3, 172, or 516 mg sodium dichromate dihydrate/L (equivalent to 0,
5, 20, 50, or 190 mg hexavalent chromium/L, respectively). Based on water consumption
measured throughout the study, NTP (2008) calculated average daily doses for females over the
2-year treatment duration of approximately 0, 1.1, 3.9, 9, or 25 mg sodium dichromate
dihydrate/kg-day (equivalent to 0, 0.38, 1.4, 3.1, or 8.7 mg hexavalent chromium/kg-day,
respectively). “Core” study mice were subjected to the same evaluations and procedures as those
described above for “core” study rats (NTP, 2008). An additional “special study” group of
female mice (10/group) were exposed to the same drinking water concentrations of sodium
dichromate dihydrate as “core” animals for up to 53 weeks. For the “special study” mice only,
blood was collected on day 22 and at 3, 6, and 12 months for hematologic analyses only (i.e.,
Hct; Hgb concentration; erythrocyte, reticulocyte, and platelet counts; erythrocyte and platelet

morphology; MCV; MCH; MCHC; and leukocyte count and differentials). At the end of the
53-week treatment period, “special study” animals were evaluated for chromium tissue
distribution (see Section 3.2 for the results of this study).
Survival rates of “core” study mice exposed to sodium dichromate dihydrate were similar
to controls (NTP, 2008). Throughout the study, water consumption by males and females was
decreased in the two highest dose groups compared with controls. During the second year of the
study, water consumption in the two highest dose groups was decreased by 15 and 35%,
respectively, in males and by 25 and 32%, respectively, in females (statistical significance not
reported). No data on food consumption were reported. At the end of the 2-year treatment
period, body weight in males in the highest dose group was decreased by 6% compared with
controls (statistical significance not reported), and body weight in females in the two highest
dose groups was decreased by 8 and 15%, respectively. NTP (2008) suggested that decreased
body weights in the highest dose groups may have been partially due to reduced water
consumption because of poor drinking water palatability, rather than an adverse effect of sodium
dichromate dihydrate exposure. No treatment-related signs of clinical toxicity were observed
throughout the study.
Results of hematology analyses in “special study” female mice (hematology was not
assessed in male mice) showed that exposure to sodium dichromate dihydrate in drinking water
produced hypochromic microcytosis (NTP, 2008), characterized by dose-related decreases in
MCV and MCH and increases in erythrocyte counts (Table 4-17); the magnitude of change in
other hematological parameters was small (≤5% compared with controls). The pattern of dose-
and duration-related severity in female mice was similar to that observed in male “special study”
rats (as described above); however, severity in mice was less than in rats. Thus, exposure of
female mice to sodium dichromate dihydrate in drinking water produced microcytosis at
subchronic exposure durations (22 days to 3 months), with decreased severity at 6–12 months.

Table 4-17. Hematological effects in female B6C3F1 mice exposed to sodium

a b c c
MCV (fL) D 22 48.8 ± 0.2 48.3 ± 0.1 47.8 ± 0.2 47.0 ± 0.2 46.8 ± 0.2c
(90.0) (98.0) (96.3) (95.9)
Mo 3 47.2 ± 0.1 46.9 ± 0.3 46.7 ± 0.1 45.1 ± 0.2c 43.7 ± 0.3c
(99.4) (98.9) (95.6) (92.6)
b c
Mo 6 45.8 ± 0.2 45.5 ± 0.3 45.1 ± 0.2 44.6 ± 0.2 42.8 ± 0.3c
(99.3) (98.5) (97.4) (93.4)
c
Mo 12 46.9 ± 0.3 46.9 ± 0.3 46.3 ± 0.3 45.2 ± 0.2 43.9 ± 0.5c
(100) (98.7) (96.4) (93.6)
MCH (ρg) D 22 16.4 ± 0.1 16.2 ± 0.0 b
15.9 ± 0.1 c
15.7 ± 0.1 c
15.5 ± 0.1c
(98.8) (97.0) (95.7) (94.5)
c c
Mo 3 15.8 ± 0.0 15.7 ± 0.1 15.6 ± 0.0 14.9 ± 0.1 14.3 ± 0.1c
(99.4) (98.7) (88.6) (90.5)
Mo 6 15.3 ± 0.1 15.2 ± 0.1 15.1 ± 0.1 14.9 ± 0.1c 14.1 ± 0.1c
(99.3) (98.7) (97.4) (92.2)
b
Mo 12 15.5 ± 0.1 15.7 ± 0.2 15.5 ± 0.1 15.1 ± 0.1 14.4 ± 0.2c
(101.3) (100) (97.4) (92.9)
Erythrocyte D 22 10.25 ± 0.15 10.20 ± 0.08 10.47 ± 0.19 10.77 ± 0.13 10.61 ± 0.13b
b
count (106/µL) (99.5) (102.1) (105.1) (103.5)

Mo 3 10.10 ± 0.16 10.66 ± 0.13 10.55 ± 0.17 10.95 ± 0.10 11.55 ± 0.16c
b b c
(105.5) (104.5) (108.4) (114.4)

Mo 6 10.56 ± 0.15 10.81 ± 0.10 10.60 ± 0.13 10.77 ± 0.20 11.50 ± 0.20c
(102.4) (100.4) (102.0) (108.9)
Mo 12 9.58 ± 0.10 9.72 ± 0.09 9.77 ± 0.10 9.95 ± 0.13b 10.30 ± 0.21c
(101.4) (102.0) (103.9) (107.5)
a
Values are means ± SE; values in parenthesis are percent of control; n = 10 mice/group, with the exception of
1.4 mg hexavalent chromium/kg-d group on mo 12 (n = 9).
b
c
Source: NTP (2008).
Gross and microscopic examinations of “core” study mice exposed to sodium dichromate
dihydrate in drinking water for 2 years showed nonneoplastic lesions of the small intestine, liver,
lymph nodes, and pancreas, and neoplastic lesions of the small intestine (NTP, 2008). Incidence
data for nonneoplastic lesions are summarized in Table 4-18. In the small intestine, statistically
significant increases in the incidences of minimal-to-mild diffuse epithelial hyperplasia of the
duodenum were observed in male and female mice in all treatment groups and of the jejunum in
females at 8.7 mg hexavalent chromium/kg-day, compared with controls. NTP (2008) noted that
diffuse epithelial hyperplasia is consistent with tissue regeneration following epithelial cell
damage. Incidences of minimal-to-mild histiocytic cellular infiltration of the duodenum were

increased at ≥2.4 and ≥3.1 mg hexavalent chromium/kg-day in males and females, respectively,
and of the jejunum at 8.7 mg hexavalent chromium/kg-day in females, compared with controls.
Moderate-to-severe focal epithelial hyperplasia was also observed in the duodenum in males and
females, although incidences were not significantly different from controls (the incidence did not
exceed 2/50 rats in any dose group) and did not exhibit dose-dependence. Due to its
morphological similarity to adenoma, focal epithelial hyperplasia was classified as a
preneoplastic lesion by NTP (2008).

B6C3F1 mice exposed to sodium dichromate dihydrate in drinking water for
2 years

Males
Small intestine, duodenum (diffuse 0/50a 11/50b 18/50b 42/50b 32/50b
epithelial hyperplasia) (2.0) (1.6) (2.1) (2.1)
b
Small intestine, duodenum (histiocytic 0/50 2/50 4/50 37/50 35/50b
b b b
Lymph node, mesenteric (histiocytic 14/47 38/49 31/49 32/49 42/46b
c
Lymph node, pancreatic (histiocytic 0/5 2/13 2/10 5/8 12/16c
Pancreas (cytoplasmic alteration) 0/49 1/49 1/50 9/49b 8/48b
(3.0) (3.0) (2.1) (2.6)
Females
a
Small intestine, duodenum (diffuse 0/50 16/50b 35/50b 31/50b 42/50b
epithelial hyperplasia) (1.6) (1.7) (1.6) (2.2)
Small intestine, duodenum (histiocytic 0/50 0/50 4/50 33/50b 40/50b
Small intestine, jejunum (diffuse epithelial 0/50 2/50 1/50 0/50 8/50b
hyperplasia) (2.0) (1.0) (1.9)
Small intestine, jejunum (histiocytic 0/50 0/50 0/50 2/50 8/50b
cellular infiltration) (1.0) (1.6)
Liver (histiocytic cellular infiltration) 2/49 15/50b 23/50b 32/50b 45/50b
(1.0) (1.1) (1.0) (1.0) (1.9)
Liver (chronic inflammation) 16/49 21/50 22/50 27/50b 24/50
(1.1) (1.1) (1.10 (1.1) (1.0)
Lymph node, mesenteric (histiocytic 3/46 29/48b 26/46b 40/50b 42/50b
Lymph node, pancreatic (histiocytic 0/14 1/12 2/15 7/14b 8/13b
Pancreas (cytoplasmic alteration) 0/48 6/50c 6/49c 14/50b 32/50b
(2.5) (2.0) (2.4) (2.6)
a
b
c
Source: NTP (2008).
In the liver of female mice, dose-dependent increases were observed in the incidences of
histiocytic infiltration at all doses and of chronic inflammation in the 3.1 mg hexavalent

chromium/kg-day group, both lesions were minimal to mild in severity. Significant decreases in
the incidences of clear cell and eosinophilic foci were observed in the liver of males at 5.9 mg
hexavalent chromium/kg-day and of eosinophilic foci in the liver of females at ≥3.1 mg
hexavalent chromium/kg-day; NTP (2008) indicated that the biological significance of these
decreases is uncertain.
Dose-dependent increases in the incidences and severity (minimal-to-mild) of histiocytic
cellular infiltration of the mesenteric lymph nodes were observed in males and females in all
treatment groups and of the pancreatic lymph nodes in males and females at ≥2.4 and ≥3.1 mg
hexavalent chromium/kg-day, respectively, compared with controls.
In the pancreas, the dose-dependent increases in the incidences and severity (mild-to-
moderate) of cytoplasm alterations, characterized by depletion of cytoplasm zymogen granules,
were observed at ≥2.4 mg hexavalent chromium/kg-day in males and in all treatment groups in
females. NTP (2008) stated that the biological significance of this finding is uncertain.
Incidence data for neoplastic lesions of the small intestine in male and female mice
exposed to sodium dichromate dihydrate in drinking water for 2 years are summarized in
Table 4-19 (NTP, 2008). In male mice, incidences of combined small intestine (duodenum,
jejunum, and ileum) adenoma or carcinoma were significantly increased at ≥2.4 mg hexavalent
chromium/kg-day and incidences of duodenal adenoma, small intestine adenoma, and small
intestine carcinoma were significantly increased at 5.9 mg hexavalent chromium/kg-day. In
addition, significant positive dose-related trends were observed for the incidences of duodenal
adenoma, duodenal carcinoma, jejunal adenoma, small intestine adenoma, small intestine
carcinoma, and combined small intestine adenoma or carcinoma. In female mice, significant
increases in the incidences of duodenal adenoma, small intestine adenoma, and combined small
intestine adenoma or carcinoma were observed at ≥3.1 mg hexavalent chromium/kg-day and
incidences of duodenal carcinoma, jejunal adenoma, and small intestine carcinoma were
significantly increased at 8.7 mg hexavalent chromium/kg-day. Significant positive dose-related
trends were observed for duodenal adenoma, duodenal carcinoma, jejunal adenoma, small
intestine adenoma, small intestine carcinoma, and combined small intestine adenoma or
carcinoma. No other statistically or biologically significant neoplasms were observed in other
tissues.

Table 4-19. Incidence of neoplastic lesions observed in the small intestine of
male and female B6C3F1 mice exposed to sodium dichromate dihydrate in

Tissue and lesion type 0 0.38 0.91 2.4 5.9
Males
Duodenum, adenoma
Overall ratea,b 0/50 0/50 0/50 0/50 6/50
(0%) (0%) (0%) (0%) (12%)
p ≤ 0.05
Duodenum, all adenoma (includes multiple adenomas)
Overall ratea 1/50 0/50 1/50 5/50 15/50
(2%) [665] (0%) (2%) [729T] (10%) [729T] (30%) [451]
Adjusted ratec 2.2% 0% 2.3% 10.8% 32.9%
p < 0.001 p < 0.001
Duodenum, carcinoma
Overall ratea 0/50 0/50 0/50 2/50 3/50
(0%) (0%) (0%) (4%) [729T] (6%) [729T]
Adjusted ratec 0% 0% 0% 4.3% 6.8%
p < 0.011
Jejunum, adenoma
Overall ratea 0/50 0/50 0/50 0/50 3/50
(0%) (0%) (0%) (0%) (6%) [714]
Adjusted ratec 0% 0% 0% 0% 6.8%
p = 0.002
Jejunum, multiple carcinoma
Overall ratea,b 0/50 1/50 0/50 0/50 0/50
Jejunum, all carcinoma (includes multiple)
Overall ratea,b 0/50 2/50 0/50 1/50 2/50
All small intestined, adenoma
Overall ratea 1/50 1/50 1/50 5/50 17/50
(2%) [665] (2%) [729T] (2%) [729T] (10%) [729T] (34%) [451]
Adjusted ratec 2.2% 2.3% 2.3% 10.8% 37.2%
p < 0.001 p < 0.001
All small intestined, carcinoma
Overall ratea 0/50 2/50 1/50 3/50 5/50
(0%) (4%) [729T] (2%) [729T] (6%) [729T] (10%) [729T]
Adjusted ratec 0% 4.5% 2.3% 6.5% 11.4%
p = 0.014 p = 0.028
All small intestined, adenoma or carcinoma
Overall ratea 1/50 3/50 2/50 7/50 20/50
(2%) [665] (6%) [729T] (4%) [729T] (14%) [729T] (40%) [451]
Adjusted ratec 2.2% 6.8% 4.6% 15.1% 43.8%
p < 0.001 p = 0.032 p < 0.001


Tissue and lesion type 0 0.38 1.4 3.1 8.7
Females
Duodenum, multiple adenoma
Overall ratea,b 0/50 0/50 0/50 1/50 6/50
(0%) (0%) (0%) (2%) (12%)
p ≤ 0.05
Duodenum, all adenoma (includes multiple)
Overall ratea 0/50 0/50 2/50 13/50 12/50
(0%) (0%) (4%) [729T] (25%) [729T] (24%) [693]
Adjusted ratec 0% 0% 4.2% 27.8% 25.2%
p < 0.001 p < 0.001 p < 0.001
Duodenum, carcinoma
Overall ratea 0/50 0/50 0/50 1/50 6/50
(0%) (0%) (0%) (2%) [729T] (12%) [625]
Adjusted ratec 0% 0% 0% 2.1% 12.6%
p < 0.001 p = 0.019
Jejunum, multiple adenomas
Overall ratea,b 0/50 0/50 0/50 0/50 1/50
(0%) (0%) (0%) (0%) (2%)
Jejunum, all adenomas (including multiple)
Overall ratea 0/50 1/50 0/50 2/50 5/50
(0%) (2%) [729T] (0%) (4%) [729T] (10%) [729T]
Adjusted ratec 0% 2.2% 0% 4.3% 10.6%
p = 0.002 p = 0.035
Jejunum, carcinoma
Overall ratea,b 1/50 0/50 2/50 2/50 1/50
(2%) (0%) (4%) (4%) (2%)
All small intestined, adenoma
Overall ratea 0/50 1/50 2/50 15/50 16/50
(0%) (2%) [729T] (4%) [729T] (30%) [729T] (32%) [693]
Adjusted ratec 0% 2.2% 4.2% 32.0% 33.7%
p < 0.001 p < 0.001 p < 0.001
All small intestined, carcinoma
Overall ratea 1/50 0/50 2/50 3/50 7/50
(2%) [729T] (0%) (4%) [729T] (6%) [729T] (14%) [625]
Adjusted ratec 2.2% 0% 4.2% 6.4% 14.7%
p < 0.001 p = 0.037

All small intestined, adenoma or carcinoma

Overall ratea 1/50 1/50 4/50 17/50 22/50
(2%) [729T] (2%) [729T] (8%) [729T] (34%) [729T] (44%) [625]
Adjusted ratec 2.2% 2.2% 8.3% 36.3% 45.9%
p < 0.001 p < 0.001 p < 0.001
a
Overall rate: number of animals with lesion/number of animals examined; parentheses are the percent of animals
examined with lesion; brackets indicate the days to first incidence; T: observed at terminal sacrifice. p-Value
under treatment group incidence data indicates statistically significant Poly-3 test for pairwise comparison between
control and exposed group. Statistical analysis using overall rates were only conducted if adjusted rates were not
determined.
b
Adjusted rate not reported.
c
Adjusted rate: Poly-3 estimated neoplasm incidence (expressed as % of animals with neoplasm) adjusted for
intercurrent mortality. p-Value under control group indicates statistically significant positive Poly-3 trend test.
p-Value under treatment group incidence data indicates statistically significant Poly-3 test for pairwise comparison
between control and exposed groups, using adjusted rates.
d
Duodenum, jejunum, or ileum.
Source: NTP (2008).
In conclusion, from the NTP (2008) 2-year toxicology and carcinogenicity study on
sodium dichromate dihydrate, EPA identified a LOAEL for noncancer effects of 0.38 mg
hexavalent chromium/kg-day in male and female B6C3F1 mice; a NOAEL value was not
identified because effects were seen at the lowest dose administered. In males, the LOAEL was
based on increased incidences of histopathological changes to the duodenum (diffuse epithelial
hyperplasia) and mesenteric lymph nodes (histiocytic cellular infiltration); in females, the
LOAEL was based on increased incidences of histopathological changes to the duodenum
(diffuse epithelial hyperplasia), mesenteric lymph nodes (histiocytic cellular infiltration), liver
(histiocytic cellular infiltration), and pancreas (depletion of cytoplasmic zymogen granules).
Although mild microcytic, hypochromic anemia was observed in female mice at ≥0.38 mg
hexavalent chromium/kg-day after 22 days of exposure, the severity of these effects decreased
over time, such that only small changes (<5%) were observed at ≥3.1 mg hexavalent
chromium/kg-day after 12 months of exposure; therefore, hematological effects were not
considered by EPA as the basis for the chronic LOAEL value in female mice. In addition to
noncancer effects, exposure of B6C3F1 mice to sodium dichromate dihydrate in drinking water
for 2 years resulted in significant increases in the incidences of neoplasms of the small intestine
in males and females at doses ≥2.4 and ≥3.1 mg hexavalent chromium/kg-day, respectively.
NTP (2008) concluded that results of this study provide clear evidence of carcinogenic activity
of sodium dichromate dihydrate in male and female B6C3F1 mice based on increased incidences
of neoplasms of the small intestine.

Although water consumption was reduced in both male and female rats and mice at the
two highest doses in the NTP (2008) study, the NTP concluded that the animals in this two-year
bioassay were not suffering from dehydration, and thus this reduced water consumption had little
impact on the study results. More specifically, the NTP stated the following regarding this
potential dehydration issue in their final Technical Report (NTP, 2008):
The lower body weights observed in male and female rats and mice exposed to
the two highest exposure concentrations were partly attributed to poor palatability
of the dosed water and consequent reductions in water consumption. However,
several lines of evidence suggest that the animals were not dehydrated. When
water consumption is adjusted for body weight (data not shown), dosed male and
female rats and female mice drank approximately the same quantities of water per
gram of body weight as the controls after the first 20 weeks on study. Male mice
exposed to 257.4 mg/L drank less water per gram of body weight than did the
controls throughout the study. Although mean body weights and water
consumption were reduced in the higher exposure concentration groups, the
average daily doses (mg sodium dichromate dihydrate per kilogram body weight)
were in the same proportions as the drinking water concentrations (mg/L) for
male and female rats and mice. Clinical observations related to dehydration
including loss of skin turgor, dry mucous membranes, retraction of eyes,
hypoactivity, and poor hair coats were not observed in rats or mice in the 2-year
studies of sodium dichromate dihydrate. Abnormalities in hematology and clinical
chemistry parameters that typically indicate dehydration include increases in
hematocrit, urine specific gravity, and serum concentrations of albumin, total
protein, and urea nitrogen. In the current 2-year studies, hematology and clinical
chemistry parameters were measured in male rats on days 4 and 22 and at months
3, 6, and 12. Significant decreases in hematocrit and serum concentrations of
albumin and total protein were observed in males exposed to 516 mg/L. Taken
together, these data suggest that the neoplastic and nonneoplastic effects of
sodium dichromate dihydrate were not associated with dehydration.
Borneff et al., 1968

Borneff et al. (1968) conducted a long-term animal cancer bioassay of hexavalent
chromium administered in drinking water. Using a three-generation study design, Borneff et al.
(1968) treated 120 female and 10 male NMRI mice with 1 mg potassium chromate/day
(500 ppm) in drinking water (containing 3% household detergent). A control group of animals
received drinking water (3% detergent) only. An outbreak of mousepox (ectromelia) virus
occurred during the 8th month of the experiment, and within 3 months, the majority (512) of the
animals died. All animals received a mousepox vaccination 2 months after the outbreak, and this
effectively ended the epidemic and the study continued. Two carcinomas of the stomach were
observed in female mice exposed to potassium chromate. No malignant stomach tumors were
found in control mice. Nine benign stomach tumors were observed in female mice exposed to
potassium chromate. The combined incidence of malignant and benign stomach tumors (11/66)

in potassium chromate-exposed-female mice was significantly different than the combined
incidence of tumors in control female mice (2/79; Fisher’s exact test, p < 0.05). This increase in
tumors, however, was seen almost exclusively in the generation most affected by the epidemic,
and it is likely that the observed increase in tumors was due, at least in part, to the infection.
Because of the unknown impact of the mousepox infection on the results of this study, as well as
other methodological problems, EPA chose not to identify a NOAEL or LOAEL from this study.
Furthermore, this study is considered unsuitable for the assessment of the oral carcinogenicity of
hexavalent chromium.
Anwar et al., 1961

The effects of chronic oral exposure to hexavalent chromium were evaluated in dogs by
Anwar et al. (1961). Dogs (one control dog and one to two dogs/treatment group) were exposed
to potassium chromate in drinking water at concentrations of 0, 0.45, 2.25, 4.5, 6.75, or 11.2 mg
hexavalent chromium/L for 4 years. Several different breeds of dogs (German shepherds,
poodles, and beagles) were used and body weights of animals were not reported; thus, daily
hexavalent chromium doses cannot be accurately estimated. Throughout the exposure period,
animals were evaluated for clinical signs of toxicity, and food consumption and growth rate were
recorded (frequency of observations not reported). At monthly intervals, blood was obtained for
evaluation of hematology (i.e., erythrocyte counts, total and differential leukocyte counts, and
Hgb), and at 6-month intervals, urine was analyzed for albumin, acetone, bile pigments, glucose,
erythrocytes, and specific gravity. At the end of the 4-year treatment period, weights of the liver,
kidney, and spleen were recorded, and microscopic examination was conducted on selected
tissues of major organs. No chromium-related effects were observed. Interpretation of the study
results is limited by the small number of animals evaluated and the inability to estimate daily
doses of hexavalent chromium received by the treated animals. A NOAEL or LOAEL could not
be identified from this study by EPA.
MacKenzie et al., 1958

MacKenzie et al. (1958) conducted two experiments in which Sprague-Dawley rats were
administered hexavalent chromium in drinking water for 1 year. In the first experiment, groups
of rats (10 per sex in the control group and 8 per sex in the treatment groups) were exposed to
drinking water containing potassium chromate at concentrations of 0, 0.45, 2.2, 4.5, 7.7, or
11 mg hexavalent chromium/L. In the second experiment, groups of 12 male and 9 female rats
were exposed to drinking water containing potassium chromate at concentrations of 0 or 25 mg
hexavalent chromium/L. For experiment 1, MacKenzie et al. (1958) reported that drinking water
consumption and body weights in the treatment groups were comparable to controls, although
data were not reported. Using reference values for body weight (males: 0.523 kg; females:
0.338 kg) and daily drinking water intake (males: 0.062 L/day; females: 0.045 L/day) for adult

male and female Sprague-Dawley rats (U.S. EPA, 1988), doses of 0.05, 0.26, 0.53, 0.91, or
1.3 mg hexavalent chromium/kg-day for males and 0.06, 0.29, 0.60, 1.0, or 1.5 mg hexavalent
chromium/kg-day for females exposed to drinking water containing 0.45, 2.2, 4.5, 7.7, or 11 mg
hexavalent chromium/L, respectively, were estimated by EPA. For experiment 2, drinking water
consumption was decreased by 16 and 27% in male and female rats, respectively. Thus, using
reference values for body weight and daily drinking water intake for adult male and female
Sprague-Dawley rats (listed above; U.S. EPA, 1988) and assuming decreases in water
consumption of 16 and 27% in males and females, respectively, average daily doses of 2.8 and
2.4 mg hexavalent chromium/kg-day in males and females, respectively, were estimated by EPA.
Throughout the treatment period in both experiments, animals were examined for clinical signs
of toxicity, and weight gain and food and water consumption were recorded (frequency of
observations not reported). At monthly intervals, blood was analyzed for Hgb, erythrocyte
counts, and total and differential leukocyte counts. At the end of treatment, microscopic
examinations of selected tissues (kidney, adrenal gland, liver, spleen, heart, brain, stomach,
duodenum, ileum, colon, and bone marrow) were conducted (as described by Decker et al.,
1958). No treatment-related clinical signs of toxicity, effects on food consumption, body weight
gain, or histopathological findings were observed. Therefore, EPA identified a NOAEL of 2.8
mg/kg-day.
4.3. REPRODUCTIVE AND DEVELOPMENTAL TOXICITY STUDIES—ORAL

Studies evaluating the potential reproductive effects of oral exposure to hexavalent
chromium compounds have been conducted in monkeys (Aruldhas et al., 2006, 2005, 2004;
Subramanian et al., 2006), rats (Bataineh et al., 2007, 1997; Elsaieed and Nada, 2002; Li et al.,
2001; Kanojia et al., 1998, 1996; NTP, 1996b; Chowdhury and Mitra, 1995;), mice (Al-Hamood
et al., 1998; Elbetieha and Al-Hamood, 1997; NTP, 1997; 1996a; Junaid et al., 1996a, b, 1995;
Murthy et al., 1996; Zahid et al., 1990; Trivedi et al., 1989), and rabbits (Yousef et al., 2006). In
addition, several studies have specifically evaluated the potential effects of pregestational,
gestational, or lactational exposure on fetal development in rats (Banu et al., 2008; Elsaieed and
Nada, 2002; Kanojia et al., 1998, 1996) and mice (Al-Hamood et al., 1998; Junaid et al.,
1996a, b, 1995; Trivedi et al., 1989). Studies conducted by NTP (1997, 1996a,b) and Zahid et al.
(1990) evaluated dietary exposure; all other studies evaluated animals exposed to hexavalent
chromium in drinking water or by gavage. In general, studies that evaluated developmental
effects of hexavalent chromium were conducted at higher exposure levels than those that
evaluated reproductive effects.
Collectively, the available studies provide evidence that oral exposure of laboratory
animals to hexavalent chromium compounds produces adverse reproductive effects, including
histopathological changes to reproductive organs in males (Aruldhas et al., 2006, 2005, 2004;
Chowdhury and Mitra, 1995; Li et al., 2001; Zahid et al., 1990) and females (Murthy et al.,

1996); alterations in sperm, including decreased count, decreased motility, and abnormal
morphology (Subramanian et al., 2006; Yousef et al., 2006; Li et al., 2001; Zahid et al., 1990);
decreased plasma testosterone levels (Yousef et al., 2006; Chowdhury and Mitra, 1995);
increased estrous cycle length (Kanojia et al., 1998, 1996; Murthy et al., 1996); changes in
mating behavior and decreased fertility in males (Bataineh et al., 1997); and adverse
reproductive outcomes, including decreased numbers of live fetuses and implantations, and
increased numbers of resorptions and pre- and postimplantation losses (Bataineh et al., 2007;
Elsaieed and Nada, 2002; Elbetieha and Al-Hamood, 1997; Junaid et al., 1996a,b, 1995; Kanojia
et al., 1998, 1996; Trivedi et al., 1989). Developmental effects observed have included
decreased fetal weight and length (Elsaieed and Nada, 2002; Kanojia et al., 1998; Junaid et al.,
1996a, b, 1995; Trivedi et al., 1989); external (subdermal hemorrhage and tail malformations)
and skeletal abnormalities (decreased ossification) (Elsaieed and Nada, 2002; Junaid et al.,
1996a, b, 1995; Kanojia et al., 1998, 1996; Trivedi et al., 1989); and delayed sexual maturation
and function in female offspring (Banu et al., 2008; Al-Hamood et al., 1998). In contrast to
results of the above studies, adverse effects were not observed in dietary exposure studies
conducted by NTP that investigated the potential for hexavalent chromium to produce adverse
effects on male reproductive organs in rats and mice (NTP, 1996a,b) and on reproductive
outcomes in a continuous breeding study in mice (NTP, 1997).
The following review of available reproductive and developmental studies is organized as
follows: (1) studies evaluating effects on reproductive tissues and mating behavior, (2) studies
evaluating effects on reproductive outcomes, (3) studies evaluating pregestational exposure on
reproductive outcomes and fetal development, and (4) studies evaluating gestational and/or
lactational exposure on reproductive outcomes and fetal development. The results of these
studies are summarized in Section 4.5, Synthesis of Major Noncancer Effects, Table 4-26.
4.3.1. Effects on Reproductive Tissues and Mating Behavior

Aruldhas et al., 2006, 2005, 2004; Subramanian et al., 2006
In a series of studies conducted by the same research group, adverse effects on male
reproductive organs were observed in monkeys exposed to hexavalent chromium in drinking
water (Aruldhas et al., 2006, 2005, 2004; Subramanian et al., 2006). All of these studies
followed the same exposure protocol; adult male bonnet monkeys (6–8 years old) were exposed
to drinking water containing 0, 100, 200, or 400 mg potassium dichromate/L in Aruldhas et al.
(2006, 2005, 2004) or 0, 50, 100, 200, or 400 mg potassium dichromate/L in Subramanian et al.
(2006) for 180 days; two of the studies included a 180-day post-treatment recovery period
(Aruldhas et al., 2006; Subramanian et al., 2006). Aruldhas et al. (2004) noted that 400 mg
potassium dichromate/L was selected as the maximum concentration tested because exposure to
higher concentrations resulted in decreased food and drinking water consumption and death
within 3 months. At the beginning of the treatment period, body weights of monkeys were

reported as 7–8 kg by Aruldhas et al. (2005) and 7–9 kg by Subramanian et al. (2006). Although
body weights were not reported by Aruldhas et al. (2006, 2004), it is assumed that initial body
weights were similar in all studies. The study authors did not report body weights or drinking
water consumption over the course of treatment or calculate daily doses of hexavalent chromium.
For this review, daily doses of 0, 1.0, 2.1, 4.1, and 8.3 mg hexavalent chromium/kg-day for the 0,
50, 100, 200, or 400 potassium dichromate/L groups, respectively, were estimated using the
allometric equation for drinking water consumption for primates (0.09 × body weight0.7945; U.S.
EPA, 1988) and an average reported initial body weight of 8 kg (Subramanian et al., 2006;
Aruldhas et al., 2005); however, these dose estimates are uncertain due to the absence of data on
body weight and drinking water consumption over the course of the 6-month treatment period.
In the following discussions, the three treatment groups evaluated in the Aruldhas et al. (2006,
2005, 2004) studies (i.e., 100, 200, and 400 mg potassium dichromate/L, approximately
equivalent to 2.1, 4.1, and 8.3 mg hexavalent chromium/kg-day, respectively) are referred to as
the low-, mid-, and high-dose groups, respectively; the four treatment groups evaluated in the
Subramanian et al. (2006) study (i.e., 50, 100, 200, and 400 mg potassium dichromate/L,
approximately equivalent to 1.0, 2.1, 4.1, and 8.3 mg hexavalent chromium/kg-day, respectively)
are referred to as the lowest-, low-, mid-, and high-dose groups, respectively.
Aruldhas et al. (2004) conducted histological assessments of testes and epididymides
from monkeys (three monkeys/group) following 180 days of treatment. Testes and epididymides
were evaluated by light microscopy (resin-embedded slices) and transmission electron
microscopy (TEM). In the three treatment groups, epididymal damage and the development of
microcanals in the cauda epididymal epithelium were observed; severity of ductal damage
increased with dose. In the low-dose group, the cauda epididymal epithelium appeared
pseudostratified; degeneration of principal cells and epithelial rupture, with the lumen occluded
by principal cells, were observed. In the mid-dose group, the occluded lumen appeared packed
with immature germ cells and macrophages. In the high-dose group, hypertrophy of the caudal
epithelium and “obliteration” of the ductal lumen were observed. The development of two
morphologically distinct microcanals was observed in all treatment groups. Arulhhas et al.
(2004) proposed that microcanal development was an adaptive response to provide passage for
spermatozoa around the obstructed ducts and to entrap spermatozoa that had been released into
the epithelium due to the epithelial rupture. Appearance of tissues from the control group was
not reported. Additional TEM evaluations of testes from monkeys (three monkeys/group) in the
three hexavalent chromium treatment groups showed a dose-related accumulation of basal cells
along the basal lamina of the epididymis, giving the epithelium a pseudostratified appearance,
and intraepithelial macrophages (Aruldhas et al., 2006). In addition, cells showed an
accumulation of sperm-derived lipofuscin material, indicative of phagocytosis and processing of
sperm. In contrast, these findings were not observed in testes from control monkeys.

Aruldhas et al. (2005) evaluated the effects of hexavalent chromium exposure in male
monkeys at the completion of the 180-day treatment period (three monkeys/group) and following
an additional 180-day recovery period (three monkeys/group); assessments included plasma
chromium concentration, absolute and relative testicular weights, and microscopic (light and
TEM) evaluations of testes. At the end of the treatment period, chromium plasma concentration
was significantly (p < 0.05) increased in the three treatment groups, with increases reaching
almost ninefold in the high-dose group compared to controls. Relative testicular weight was
significantly (p < 0.05) decreased by 23, 35, and 34% in the low-, mid-, and high-dose groups,
respectively; absolute testicular weight was not affected by treatment (data not reported).
Following the recovery period, chromium plasma concentrations and relative testicular weight in
treatment groups were comparable to controls. Light microscopic evaluations of testes in control
monkeys showed seminiferous tubules and Leydig cells with normal appearance and cellular
organization. In the three hexavalent chromium treatment groups, seminiferous tubules appeared
disorganized, with decreased diameters, epithelial degeneration, and lumens filled with
prematurely released germ cells and cellular debris; depletion of germ cells, hyperplasia of
Leydig cells, and Sertoli cell fibrosis were also observed. TEM examination of testes from the
three treatment groups showed morphological changes in spermatids (granulation of chromatin
and vacuolization) and spermatocytes (fragmented chromatin and swollen mitochondria) and the
presence of macrophages containing phagocytosed sperm; effects were more severe in the high-
dose group. Following the recovery period, no histopathological findings were observed in
testes of hexavalent chromium-treated monkeys, with the exception of “a few” prematurely
released germ cells in the seminiferous tubular lumen (treatment group for this observations was
not specified).
Subramanian et al. (2006) evaluated sperm count and sperm straight-line velocity
at monthly intervals during the 180-day treatment period; the same evaluations were
conducted monthly in monkeys in the high-dose group during a 180-day recovery period. In the
lowest-dose group, no effects were observed on sperm count or straight-line velocity. Sperm
count was significantly decreased in the low-, mid-, and high-dose groups, compared with
controls; decreases were dose- and duration-dependent. For example, in the low-dose group,
significant (p < 0.05) decreases in sperm count were first observed after 4 months (11%
decrease), with a maximum decrease of 25% after 6 months; in the high-dose group, sperm
counts were significantly decreased by 13% after 2 months, with a 30% reduction after 6 months.
Similar effects were observed for sperm straight-line velocity. In the low-dose group, velocity
was significantly (p < 0.05) decreased by 10 and 25% after 4 and 6 months of treatment,
respectively; in the high-dose group, velocity was significantly decreased by 12% after 2 months
and by 35% after 6 months. Effects on sperm count and straight-line velocity were reversible
following withdrawal from treatment. During the first month of the recovery period (high-dose
monkeys only), sperm count was significantly increased compared with that observed at the end

of the treatment period, with counts returning to pretreatment levels by month 3 of the recovery
period; sperm velocity returned to pretreatment levels by month 3 of the recovery period.
Results of these four studies (Aruldhas et al., 2006, 2005, 2004; Subramanian et al.,
2006) indicate that exposure of monkeys to hexavalent chromium as potassium dichromate in
drinking water produced reversible changes to male reproductive organs, including disruption of
spermatogenesis. Effects on sperm count and velocity and histopathological changes were
observed in the low-, mid-, and high-dose groups (≥2.1 mg hexavalent chromium/kg-day), but no
effects on sperm count or velocity were observed in monkeys in the lowest treatment group
(1 mg hexavalent chromium/kg-day). This dose cannot be considered a NOAEL, however,
because microscopic evaluations were not conducted in monkeys from this group. For this
reason, NOAEL and LOAEL values from these studies were not identified by EPA. Although
group sizes in these studies were small, the results provide evidence of adverse male
reproductive effects in nonhuman primates exposed to hexavalent chromium in drinking water at
concentrations as low as 35.3 mg hexavalent chromium/L (2.1 mg hexavalent chromium/kg-
day).
Chowdhury and Mitra, 1995

Effects of oral exposure to hexavalent chromium on male reproductive organs was
evaluated in mature (age not reported) male Charles Foster rats that were administered 0, 20, 40,
or 60 mg hexavalent chromium/kg-day as sodium dichromate in saline by gavage for 90 days
(Chowdhury and Mitra, 1995). Although Chowdhury and Mitra (1995) stated that the control
and exposure groups included 10 animals per group, conflicting summaries of the actual group
sizes are presented in the report. Body weights were recorded twice weekly. At the end of the
treatment period, testes were excised, weighed, and prepared for histological or biochemical
evaluations, and serum testosterone activity was determined. For biochemical analyses, fresh
tissue was homogenized and assayed for total cholesterol, activities of succinic dehydrogenase
and 3β-Δ5-hydroxysteroid dehydrogenase (3β-Δ5-HSH), and total protein, deoxyribonucleic acid
(DNA), and ribonucleic acid (RNA). For microscopic evaluations, testes were fixed in Bouin’s
fluid, embedded in paraffin, and stained with haematoxylin and eosin (H&E).
Final body weight was significantly reduced by approximately 27% compared to controls
in the mid- and high-dose groups (statistical significance not reported); absolute testis weights
were significantly reduced by 28% (p < 0.05) and 35% (p < 0.001) in the mid- and high-dose
groups, respectively, compared with controls. Serum testosterone levels were decreased by 31%
in the low- (p < 0.05) and mid-dose (p < 0.001) groups and by 47% (p < 0.001) in the high-dose
group. Biochemical analysis of testes showed significant decreases in total cholesterol by 2%
(p < 0.05) and 25% (p < 0.001) in the mid- and high-dose groups, respectively, and significant
(p < 0.001) decreases in succinic dehydrogenase activity by 35 and 45% in the mid- and high-
dose groups, respectively. In all treatment groups, 3β-Δ5-HSH was significantly decreased by

25% (p < 0.05), 28% (p < 0.05), and 52% (p < 0.001) in the low-, mid-, and high-dose groups,
respectively. Dose-related decreases in total testicular protein were observed, with decreases
reaching 46% (p < 0.001) in the high-dose group. Testicular DNA and RNA levels were
significantly decreased in the mid- and high-dose groups, with decreases reaching 45%
(p < 0.001) and 37% (p < 0.001), respectively, in the high-dose group. Microscopic evaluation
of testicular tissue showed adverse effects in the mid- and high-dose groups including
disintegration of peritubular membranes, detachment of seminiferous cellular components from
basement membranes, and accumulation of cellular debris in the mid-dose group, and cellular
degeneration and complete disruption of the epithelium with fibrous tissue in the high-dose
group; reduction in seminiferous tubular diameter, decreased number of Leydig cells, and Leydig
cell degeneration were observed in the mid- and high-dose groups. No change in the number of
spermatogonia were observed, although the number of pachytene spermatocytes and stage 7
spermatids were decreased in the mid- and high-dose groups and resting spermatocytes were
decreased in the high-dose group. No treatment-related histopathological effects were observed
in the testes of rats in the low-dose group, although histochemical evaluations of testes showed
dose-related loss of 3β-Δ5-HSH activity in all treatment groups.
Results of histological and biochemical analyses show that oral exposure of male rats to
hexavalent chromium for 90 days produced adverse effects on male reproductive tissues,
including decreased spermatogenic and steroidogenic activities. Based on decreased serum
testosterone levels and loss of 3β-Δ5-HSH activity in testes observed in all treatment groups,
EPA identified a LOAEL of 20 mg hexavalent chromium/kg-day from this gavage study of male
Charles Foster rats,.
Bataineh et al., 1997

Effects of oral hexavalent chromium administration on mating behavior, aggression, and
fertility were assessed in male rats by Bataineh et al. (1997). Adult (age not specified) male
Sprague-Dawley rats (n = 12 or 13) were administered drinking water containing 0 or 1,000 mg
potassium dichromate/L (equivalent to 353 mg hexavalent chromium/L) for 12 weeks. No data
on drinking water consumption were included in the study report. Based on findings of other
studies (NTP, 2008, 2007) showing decreased drinking water consumption and body weight at
drinking water concentrations ≥30 mg hexavalent chromium/L, it is likely that drinking water
consumption was decreased in the chromium treatment group; thus, daily doses of hexavalent
chromium cannot be accurately estimated from this study. Following the treatment period,
assessments were conducted for sexual behavior in the presence of females in estrous (number of
mounts without penile intromission, time to first mount, time from presentation of female to first
intromission, number of penile intromissions, time from first intromission to ejaculation, and
time from ejaculation to next intromission), aggressive behavior in the presence of a second
untreated male (number of lacerations given, boxing bouts, fights, and ventral presenting),

fertility following a 10-day mating period with untreated females (numbers of pregnant females,
viable fetuses, and resorptions), body weight, and weights of reproductive organs (paired testes,
seminal vesicles, and preputial glands). Histopathological evaluations of tissues were not
conducted.
All rats “appeared healthy” throughout the treatment period. Assessment of mating
behavior in hexavalent chromium-treated rats showed significant decreases in number of mounts
(35% decrease; p < 0.001) and percentage of males ejaculating (79% decreases; p < 0.005), and
increases in the time from first intromission to ejaculation (59% increase; p < 0.001) and time
from ejaculation to next intromission (37% increases, p < 0.001), compared with controls. All
measures of aggressive behavior were decreased in rats treated with potassium dichromate. All
measures of fertility were comparable between control and treatment groups. Treatment resulted
in significant (p < 0.001) decreases in body weight (19% decrease) and absolute weights of testes
(24% decrease), seminal vesicles (15% decrease), and preputial gland (23% decrease); however,
for relative weights of reproductive tissues, only relative testes weight was significantly
decreased (6% decrease, p < 0.05) compared to controls.
EPA identified a LOAEL of 353 mg hexavalent chromium/L as potassium dichromate in
drinking water based on adverse effects on mating and aggressive behaviors; a NOAEL was not
identified because effects were observed at the only dose tested. Because drinking water
consumption and body weight data over the course of the study were not provided by the
investigators, a LOAEL, expressed in mg hexavalent chromium/kg-day, could not be derived
from this study.
Li et al., 2001
Oral exposure of male rats to chromium(VI) oxide for 6 days resulted in adverse
reproductive effects, including reduced epididymal sperm counts and increased abnormal sperm
(Li et al., 2001). Groups of 8–11 male Wistar rats (60 days old) were administered
chromium(VI) oxide by gavage at doses of 0, 10, or 20 mg chromium(VI) oxide/kg-day
(equivalent to 0, 5.2, or 10.4 mg hexavalent chromium/kg-day, respectively) for 6 days. After
6 weeks, rats were sacrificed; testes and epididymis were removed and analyzed for epididymal
sperm count and abnormal sperm; and testes were prepared (fixed in formaldehyde, embedded in
paraffin, sliced, and stained with H&E) for histological evaluations of morphological
abnormalities and diameter of seminiferous tubules. Epididymal sperm counts were significantly
(p < 0.05) decreased by 76 and 80%, and the percentage of abnormal sperm was significantly
(p < 0.01) increased by 143 and 176% in the 5.2 and 10.4 mg hexavalent chromium/kg-day
groups, respectively. Treatment-related histopathological findings included decreased diameter
of seminiferous tubules and disruption of germ cell arrangement within seminiferous tubules in
both treatment groups. Based on decreased sperm counts and histopathological changes to the
testes, 5.2 mg hexavalent chromium/kg-day was identified by EPA as a LOAEL for male rats

exposed to gavage doses of chromium(VI) oxide for 6 days; a NOAEL was not identified
because effects were seen at the lowest dose administered.
Zahid et al., 1990

Zahid et al. (1990) reported adverse effects on the male reproductive system in mice fed
diets containing potassium dichromate. However, other research groups (NTP 1997, 1996a, b;
Finley et al., 1993) have questioned the validity of the Zahid et al. (1990) study due to concerns
regarding study methods and reporting inconsistencies (as discussed below). Zahid et al. (1990)
fed male weanling BALB/c albino Swiss mice diets containing 0, 100, 200, or 400 mg potassium
dichromate/kg diet (equivalent to 0, 35.3, 70.6, or 141.2 mg hexavalent chromium/kg diet,
respectively) for 35 days. Although Zahid et al. (1990) stated that the control and exposed
groups included seven animals/group, conflicting summaries of the actual group sizes are
presented throughout the report. Body weights were recorded weekly and food consumption was
recorded every 48 hours. The study report stated that body weight gain and food consumption in
treatment groups were comparable to the control group (data not reported); however, Zahid et al.
(1990) did not calculate daily doses of hexavalent chromium. Since treatment did not affect
body weight gain or food consumption, doses of 0, 6.4, 12.7, or 25.5 mg hexavalent
chromium/kg-day for the 0, 35.3, 70.6, or 141.2 mg hexavalent chromium/kg diet groups,
respectively, were estimated for this review using reference values for body weight (0.0316 kg)
and daily food intake (0.0057 kg food/day) for subchronic exposure of male B6C3F1 mice (U.S.
EPA, 1988). After 35 days, testes and epididymis were weighed and then minced in buffered
formalin. Sperm counts were then subsequently determined and sperm were examined for
morphological abnormalities. Testes were fixed with Bouin’s fluid for 1 week, embedded in
paraffin, and subsequently sectioned to 0.6 micron thickness and stained with H&E for
histological examination. Ten sections were chosen randomly from the anterior, middle, and
posterior parts of each testis and studied. One seminiferous tubule was chosen and examined to
determine the cellular stages of spermatogenesis and the number of degenerated tubules.
Statistical analyses of the data were conducted using either a t-test or a 2 × 2 contingency χ2 test.
Adverse effects observed in the male mouse testes included ambiguous levels of degeneration in
the outermost cellular layers of the seminiferous tubules, reduced (or absent) spermatogonia per
tubule, accumulation of germ cells in the resting spermatocytes stage, reduced sperm count in the
epididymis, and increased percentage of morphologically abnormal sperm. Effects were
observed in all hexavalent chromium groups and the severity of effects appeared to increase with
dose for percentage of degenerated tubules, percentage of tubules that were not degenerated but
were without spermatogonia, percentage of abnormal sperm, and number of spermatogonia.
Based on these findings, the lowest dietary concentration tested (100 mg potassium
dichromate/kg diet or approximately 6.4 mg hexavalent chromium/kg-day) was identified as the
LOAEL by EPA.

Other research groups (NTP, 1997, 1996a, b; Finley et al., 1993) have questioned the
validity of the Zahid et al. (1990) study due to concerns regarding study design and methods.
Finley et al. (1993) noted the following three concerns: (1) use of immersion fixatives (such as
Bouin’s fluid and paraffin embedding), which can introduce artifacts, such as grains and
shrinkage, that can mimic tubular or spermatogenic pathology, (2) use of staining methods that
were unable to detect the acrosome (i.e., the part of the sperm that releases enzymes to penetrate
the egg) of developing spermatids, and (3) uncertainties regarding the actual groupings of
animals used, the small number of animals assessed per group, and inappropriate statistical
analysis of the data. NTP (1997, 1996a, b) concluded that the methods utilized by Zahid et al.
(1990) were insufficient to identify spermatogonia, were likely to have generated
nonreproducible counts of epididymal sperm, and resulted in the biologically implausible
conclusion of reduction in spermatogonia numbers concurrent with unchanged spermatocyte and
spermatid numbers.
Murthy et al., 1996

Effects on ovarian function were investigated in adult Swiss albino mice (90 days old;
mean initial body weight of 30 g) exposed to drinking water containing potassium dichromate for
20 or 90 days (Murthy et al., 1996). For the 20-day study, groups of 30 female mice were
exposed to drinking water containing 0, 250, 500, or 750 mg hexavalent chromium/L; the 20-day
exposure period was selected as it coincides with one folliculogenesis cycle. For the 90-day
study, groups of 10 female mice were administered drinking water containing 0, 0.05, 0.5, or
5 mg hexavalent chromium/L. The study report states that mice in both studies were evaluated
daily for clinical signs of toxicity, body weight, and water and food consumption; however, no
data for these outcomes were reported. Based on findings of other studies (NTP, 2008, 2007)
showing decreased drinking water consumption and body weight at drinking water
concentrations ≥30 mg hexavalent chromium/L, it is likely that drinking water consumption and
body weight were decreased in all treatment groups in the 20-day study; thus, daily doses of
hexavalent chromium cannot be accurately estimated from this study. For the 90-day study, the
concentrations of hexavalent chromium in drinking water were very low and not likely to affect
drinking water consumption or body weight. Thus, using reference values for body weight
(0.035 kg) and daily drinking water (0.0084 L/day) intake for mature female B6C3F1 mice (U.S.
EPA, 1988), doses of 0, 0.01, 0.12, or 1.2 mg hexavalent chromium/kg-day were estimated by
EPA for female mice exposed to drinking water containing 0, 0.05, 0.5, or 5 mg hexavalent
chromium/L, respectively. In the 20-day study, three types of assessments were conducted at the
end of the treatment period (each in 10 mice/group): (1) ovaries were evaluated by light
microscopy and the number of follicles at each development stage, based on size (small,
medium, large) and structural maturity, were determined, (2) superovulation was induced (by
administration of gonadotropin) and the number of released ova were counted, and (3) estrous

cycle length was assessed (by vaginal smears) for 12 consecutive estrous cycles following
treatment. In the 90-day study, all mice were sacrificed at the end of the treatment period and
ovaries were evaluated by electron microscopy for ultrastructural changes.
In mice exposed for 20 days, significant (p < 0.05) changes in follicular development
were observed in all treatment groups, with dose-related decreases in the number of small
follicles in the mid- and high-dose groups and medium and large follicles in all treatment groups.
In the high-dose group, the numbers of small, medium, and large follicles were reduced by 36,
53, and 72%, respectively, compared with controls. Ovarian response to gonadotropin was
affected in the mid- and high-dose groups, with reductions in the number of ova released of
30 and 90%, respectively, compared with controls. Estrous cycle length was significantly
increased (p < 0.05) by 1.7-fold in the high-dose group, compared with controls.
Histopathological evaluation of ovaries after 20 days of treatment showed changes in the mid-
dose (i.e., proliferated, dilated, and congested blood vessels, pyknotic nuclei in follicular cell of
mature follicles) and high-dose (i.e., undeveloped follicles with degenerative cumulus cells
containing dense pyknotic nuclei, neovascularization and karyorrhexis of follicular cells,
erythrocytes located within stromal spaces) groups; histopathological changes were not observed
in ovaries from control and low-dose mice. In mice treated for 90 days, ultrastructural changes
(i.e., disintegrated cell membranes in two-layered follicular cells and alterations in mitochondria
in thecal cells, which are cells of the corpus luteum that secrete estrone, estradiol, and
progesterone) were observed in the high-dose group; the study report did not provide any
information on ultrastructural evaluations in the low- and mid-dose groups. Murthy et al. (1996)
concluded that hexavalent chromium may induce changes in ovarian function and ovulation.
Due to inadequate reporting of the 20-day study (i.e., no information on effects of treatment on
body weight or drinking water consumption) and the 90-day study (i.e., lack of information on
ultrastructural evaluations in the low- and mid-dose groups), a LOAEL from these studies could
not be identified by EPA.
Yousef et al., 2006

Adverse effects on male reproductive tissues were observed in rabbits exposed to
potassium dichromate for 10 weeks (Yousef et al., 2006). Groups of six male New Zealand
white rabbits (7 months old) were administered 0 or 5 mg potassium dichromate/kg-day by
gavage (vehicle not specified) for 10 weeks. Yousef et al. (2006) reported that the dose of 5 mg
potassium dichromate/kg-day was equivalent to 3.6 mg hexavalent chromium/kg-day. During
the treatment period, food intake and body weights were recorded weekly. Semen was collected
weekly and analyzed for pH and sperm count, motility, and morphology. Blood was collected
every 2 weeks and analyzed for testosterone. At the end of the treatment period, animals were
sacrificed and relative testes and epididymis weights were determined. At sacrifice, seminal

plasma was collected and analyzed for AST, ALT, AP, AcP, and glutathione S-transferase (GST)
activities. Histopathological evaluations of tissues were not conducted.
No clinical signs of toxicity were observed throughout the study. Mean body weight over
the 10-week treatment period was significantly (p < 0.05) decreased by 9% compared to controls,
although average food intake over the 10-week period was not affected by treatment; final body
weight was not reported. After treatment for 10 weeks, relative testes and epididymis weights
were significantly decreased by 22% (p < 0.05). The 10-week mean plasma testosterone level in
treated rabbits was decreased by 21% (p < 0.05) compared with controls. In treated rabbits
compared with controls, mean values of the following sperm-related characteristics were
significantly (p < 0.05) decreased after 10 weeks: (1) packed sperm volume (10% decrease),
(2) sperm concentration (18% decrease), (3) total sperm output (26% decrease), (4) sperm
motility (5% decrease), (5) total motile sperm per ejaculation (34% decrease), (6) total functional
sperm fraction (37% decrease), and (7) normal sperm (4% decrease). Both percentage of dead
sperm (24% increase) and seminal fluid pH (4% increase) were increased; no effect was
observed on semen ejaculate volume. Seminal fluid activities of GST, AST, and AcP were
significantly (p < 0.05) decreased at the end of the treatment period, although decreases were
small (≤12%) compared with controls.
The results indicate that exposure of rabbits to oral potassium dichromate gavage doses
of 3.6 mg hexavalent chromium/kg-day for 10 weeks produced adverse effects on male
reproductive tissues, including decreased testes and epididymis weight and decreased sperm
output. Thus, EPA identified a LOAEL for hexavalent chromium of 3.6 mg/kg-day from this
study.
NTP, 1996a,b
The NTP conducted studies to investigate the potential effects of dietary hexavalent
chromium as potassium dichromate on male reproductive organs in Sprague-Dawley rats (NTP,
1996b) and BALB/c mice (NTP, 1996a). The NTP studies were designed to expand or replicate
the Zahid et al. (1990) study (described above), and thereby provide data to either refute or
confirm findings of adverse male reproductive effects from hexavalent chromium exposure.
Groups of 24 male and 48 female Sprague-Dawley rats were exposed to diets containing
0, 15, 50, 100, or 400 mg potassium dichromate/kg diet (equivalent to 0, 5.3, 17.6, 35.3, or
141.2 mg hexavalent chromium/kg diet, respectively) daily for 9 weeks followed by an 8-week
recovery period (NTP, 1996b). Based on food consumption measured during the 9-week
treatment period, NTP (1996a, b) calculated average daily doses of 0, 1, 3, 6, or 24 mg potassium
dichromate/kg-day (equivalent to 0, 0.35, 1.1, 2.1, or 8.5 mg hexavalent chromium/kg-day) in
males and 0, 1, 3, 7, or 28 mg hexavalent chromium/kg-day (equivalent to 0, 0.35, 1.1, 2.5, or
9.9 mg hexavalent chromium/kg-day) in females for the 0, 15, 50, 100, or 400 mg potassium
dichromate/kg diet groups, respectively. Animals were examined twice daily for mortality and

clinical signs of toxicity. Physical examinations and measurement of body weight and food and
water consumption were conducted weekly. After 3, 6, or 9 weeks of treatment or after the full
recovery period, 6 males and 12 females were sacrificed; necropsies were performed; blood was
obtained for hematology (i.e., Hgb, Hct, MCV, MCH, MCHC, mean platelet volume, and
erythrocyte, leukocyte, and platelet counts); organ weights (not specified, but including right and
left testes) were recorded; microscopic examinations were conducted on liver, kidney, ovary, and
testes (testes and epididymis were examined for Sertoli nuclei and preleptotene spermatocyte
counts in Stage X or XI tubules); and sperm were collected and analyzed for chromatin structure.
No mortalities or treatment-related clinical signs of toxicity were observed in rats in any
treatment group (NTP, 1996b). Body weights and food and drinking water consumption were
comparable between controls and treatment groups. Results of hematological analyses showed a
slight erythrocyte microcytosis in the highest dose group, as indicated by small, but significant,
decreases in MCV in females exposed for 3 weeks (3% decrease; p < 0.05) and in males exposed
for 9 weeks (6% decrease; p < 0.05), compared with controls; at 9 weeks, MCV in females was
decreased by 3%, but the change was not statistically significant. No changes in MCV were
observed in rats exposed for 6 weeks or at the end of the 8-week recovery period. After 9 weeks
of treatment, MCH was decreased by approximately 6% in males and females (statistical
significance not reported). No treatment-related findings were observed on necropsy or on
microscopic examination of the liver, kidney, ovary, testes, epididymis, or sperm. In conclusion,
no adverse effects on reproductive organs were observed in male or female rats exposed to
dietary potassium dichromate at doses of 8.5 and 9.9 mg hexavalent chromium/kg-day,
respectively, for up to 9 weeks. Based on slight erythrocyte microcytosis, the study authors
identified respective NOAELs and LOAELs of 2.1 and 8.5 mg hexavalent chromium/kg-day in
male Sprague-Dawley rats, and 2.5 and 9.9 mg hexavalent chromium/kg-day in females.
Groups of 24 male and 48 female BALB/c mice were exposed to diets containing 0, 15,
50, 100, or 400 mg potassium dichromate/kg diet (equivalent to 0, 5.3, 17.6, 35.3, or 141.2 mg
hexavalent chromium/kg diet, respectively) daily for 9 weeks followed by an 8-week recovery
period (NTP, 1996a). Based on food consumption measured during the 9-week treatment period,
the study authors calculated average daily doses of 0, 3, 10, 21, or 92 mg potassium
dichromate/kg-day (equivalent to 0, 1.1, 3.5, 7.4, or 32.5 mg hexavalent chromium/kg-day) in
males and 0, 5, 16, 34, or 137 mg hexavalent chromium/kg-day (equivalent to 0, 1.8, 5.6, 12.0, or
48.4 mg hexavalent chromium/kg-day) in females for the 0, 15, 50, 100, or 400 mg potassium
dichromate/kg diet groups, respectively. This study followed the same protocol and conducted
the same evaluations as described in the NTP (1996b) study in rats (described above).
Mortalities occurred in five male mice, but they were deemed not related to treatment,
and no treatment-related findings were observed on necropsy. The number of deaths were one,
one, two, one, and none in the control through high-dose male groups, respectively. All females
survived to study completion. No treatment-related clinical signs of toxicity were observed. At

most weekly evaluations, body weight was decreased by 5–9% in males in the highest dose
group and by 2–4% in females in the two highest dose groups (statistical significance not
reported); body weights in these groups remained depressed during the post-treatment recovery
period in high-dose males and in females at 12 mg hexavalent chromium/kg-day (but not high-
dose females). Feed consumption was generally increased (5–34%, relative to controls) in all
treatment groups in males, although changes were not statistically significant; in females, feed
consumption was increased in all dose groups (1–37%), with changes of statistical significance
in most dose groups during treatment weeks 5 and 6. Water consumption in males and females
was decreased through the first 3 weeks of treatment and comparable to controls for the
remainder of the exposure period. Hematological analyses showed a slight erythrocyte
microcytosis. In high-dose male and female mice, MCV was decreased by 2–4% (p < 0.05) at
weeks 3, 6, and 9; MCV was also slightly decreased (<2%) at 12 mg hexavalent chromium/kg-
day in females at 6 weeks. Changes in MCV were generally accompanied by small decreases in
MCH. At the end of the recovery period, a small increase in MCV (2.8%; p < 0.05) was
observed in males; in females, MCV in all treatment groups was comparable to controls. No
other effects on hematological parameters were observed. Microscopic evaluations revealed a
treatment-related increase in the incidence of cytoplasmic vacuolization of hepatocytes in male
and female mice at the end of the 9-week treatment period. Vacuoles were demarked and
appeared small and clear; NTP (1996a) noted that vacuoles were consistent with lipid
accumulation. Incidences of hepatic cytoplasmic vacuolization in the control through high-dose
groups were 0/6, 0/6, 1/6, 2/6, and 2/5 in males and 1/12, 0/12, 3/12, 2/12, and 4/12 in females,
respectively; lesion severity and statistical significance were not reported. No other treatment-
related histopathological findings were observed.
In conclusion, no adverse effects on reproductive organs were observed in male or female
mice exposed to dietary potassium dichromate at doses up to 32.5 and 48.4 mg hexavalent
chromium/kg-day, respectively, for 9 weeks. Based on histopathological changes to the liver
(cytoplasmic vacuolization), the study authors identified a NOAEL of 1.1 mg hexavalent
chromium/kg-day for males and 1.8 mg hexavalent chromium/kg-day for females.
4.3.2. Effects on Reproductive Outcomes

Elbetieha and Al-Hamood, 1997
Reproductive effects of drinking water containing 1,000–5,000 mg potassium
dichromate/L (equivalent to 353–1,765 mg hexavalent chromium/L) were evaluated in Swiss
mice in a series of three experiments (Elbetieha and Al-Hamood, 1997). No data on drinking
water consumption were included in the study report. Based on findings of other studies (NTP,
2008, 2007) showing decreased drinking water consumption and body weight at drinking water
concentrations ≥30 mg hexavalent chromium/L, it is likely that drinking water consumption was
decreased in all chromium treatment groups; thus, daily doses of hexavalent chromium cannot be

accurately estimated for this study. In the first experiment, sexually mature (i.e., 50 days old)
male Swiss mice were exposed to drinking water containing 0 (20 males), 1,000 (19 males),
2,000 (11 males), 4,000 (9 males), or 5,000 (13 males) mg potassium dichromate/L (equivalent
to 0, 353, 706, 1,412, or 1,765 mg hexavalent chromium/L, respectively) for 12 weeks. After
12 weeks, males were mated with untreated sexually mature females for 10 days; 1 week after
completion of the mating period, females were sacrificed and evaluated for the number of
pregnant females, viable fetuses, resorptions, and dead fetuses. Histopathological evaluations of
tissues were not conducted. No data on body weights were reported. Exposure of male mice to
hexavalent chromium did not affect the percentage of pregnant females. The numbers of
implantations and viable fetuses were significantly reduced from 33% in controls to 20%
(p < 0.01) and 16% (p < 0.05) in the 706 and 1,412 mg hexavalent chromium/L groups,
respectively; in the 1,765 mg hexavalent chromium/L group, the numbers of implantation and
viable fetuses were reduced to 19%, although this reduction did not reach statistical significance.
No resorptions or dead fetuses were observed in the control, 706, or 1,412 mg potassium
dichromate/L groups, but three resorptions were observed at 353 mg hexavalent chromium/L and
six resorptions and six dead fetuses were observed at 1,765 mg hexavalent chromium/L
(statistical significance not reported).
In the second experiment, sexually mature (i.e., 50 days old) female Swiss mice were
exposed to drinking water containing 0 (19 females), 2,000 (15 females), or 5,000 (11 females)
mg potassium dichromate/L (equivalent to 0, 706, or 1,765 mg hexavalent chromium/L,
respectively) for 12 weeks (Elbetieha and Al-Hamood, 1997). After 12 weeks, each female was
mated with an untreated sexually mature male for 10 days; 1 week after completion of the mating
period, females were sacrificed and evaluated for the numbers of pregnant females, viable
fetuses, and resorptions and dead fetuses. No data on body weights were reported. No
treatment-related effects were observed on the number of pregnant mice. The number of
implantations was significantly reduced from 17% in controls to 14% (p < 0.01) and 9%
(p < 0.05) in the 706 and 1,765 mg hexavalent chromium/L groups, respectively, and the number
of viable fetuses was significantly reduced from 17% in controls to 9% in the 706 (p < 0.05) and
1,765 (p < 0.01) mg hexavalent chromium/L groups, respectively. The number of mice with
resorptions was significantly increased from 11% in controls to 53% (p < 0.01) and 63%
(p < 0.005) in the 706 and 1,765 mg hexavalent chromium/L groups, respectively, and the total
number of resorptions was increased from 4 in controls to 36 and 14 in the 706 and 1,765 mg
hexavalent chromium/L groups, respectively (statistical significance not reported).
In the third experiment, sexually mature (i.e., 50 days old) mice were exposed to drinking
water containing 0 (10 males, 8 females), 2,000 (13 males, no females), or 5,000 (13 males,
10 females) mg potassium dichromate/L for 12 weeks (Elbetieha and Al-Hamood, 1997).
Following treatment, body weights and weights of reproductive organs (paired testes, seminal
vesicles, preputial glands, paired ovaries, and uteri) were determined. No mortalities or clinical

signs of toxicity were observed. Final body weights of males were significantly (p < 0.01)
reduced by approximately 10 and 12% in the 706 and 1,765 mg hexavalent chromium/L groups,
respectively; final mean body weights of treated females were similar to controls. Relative testes
weights were increased by approximately 18% (p < 0.01) and 22% (p < 0.05) in the 706 and
1,765 mg hexavalent chromium/L groups, respectively, and relative weights of seminal vesicles
and preputial gland were significantly (p < 0.001) decreased by approximately 27 and 34%,
respectively, in the 1,765 mg hexavalent chromium/L group. Relative ovary weight was
significantly increased by 54% in females in the 1,765 mg hexavalent chromium/L group,
although uterine weight was unaffected by treatment. Histopathological assessments of
reproductive tissues were not conducted.
In conclusion, results of the three experiments conducted by Elbetieha and Al-Hamood
(1997) show that exposure to potassium dichromate in drinking water affects reproductive
outcomes in exposed males and females. In female mice, decreased numbers of implantations
and viable fetuses and increased resorptions were observed at 2,000 mg potassium dichromate/L
(equivalent to 706 mg hexavalent chromium/L). In males, exposure for 12 weeks prior to mating
reduced the numbers of implantations and viable fetuses at 2,000 and 4,000 mg potassium
dichromate/L (equivalent to 706 and 1,412 mg hexavalent chromium/L, respectively), but not at
1,000 mg potassium dichromate/L (equivalent to 353 mg hexavalent chromium/L). In addition,
treatment-related changes in weights of male reproductive organs were observed at 2,000 and
5,000 mg potassium dichromate/L (equivalent to 706 and 1,412 mg hexavalent chromium/L,
respectively). Although reproductive performance was not affected at the lowest exposure level,
weights of male reproductive organs were not evaluated in male mice treated with 1,000 mg
potassium dichromate/L. Due to inadequate reporting (i.e., no information on effects of
treatment on body weight or drinking water consumption), EPA could not identify NOAEL or
LOAEL values from this study.
NTP, 1997
The potential reproductive toxicity of dietary potassium dichromate was evaluated in
BALB/c mice in a continuous breeding study (NTP, 1997). Groups of 20 male and female pairs
(F0) were exposed to dietary potassium dichromate at 0, 100, 200, and 400 mg potassium
dichromate/kg diet (equivalent to 0, 17.6, 35.3, or 141.2 mg hexavalent chromium/kg diet,
respectively) for 13 weeks (1 week prior to and 12 weeks during cohabitation). During exposure
of the F0 generation, animals were examined daily for mortality and clinical signs of toxicity;
body weights and food consumption were measured periodically (4–5 times). Litters produced
during the cohabitation period were evaluated (i.e., total pups, live and dead pups, and sex),
weighed on postnatal day (PND) 1, and euthanized with no additional assessments; pregnancy
index (number of litters/breeding pair) was also determined. After the cohabitation period,
F0 breeding pairs were separated and continued on study diets; litters born during the post-

separation period (F1 animals) were reared with the F0 dams until weaning (PND 21). Dam and
pup weights and dam food consumption were monitored during the lactational period. Upon
weaning, F0 animals were sacrificed and the following terminal evaluations were conducted:
necropsy; organ weights (liver, kidneys, right cauda epididymis, right epididymis, prostate,
seminal vesicles with coagulating glands, right testis, and ovaries); sperm evaluations (testicular
spermatid head count and epididymal sperm density, motility, and morphology); and
histopathology (liver and kidneys). Following weaning of F1 animals, animals were maintained
on the same study diets as their parents. During postlactational exposure of the F1 generation,
animals were examined daily for mortality and clinical signs of toxicity; body weights and food
consumption were measured periodically (3–4 times). At sexual maturity (approximately
74 days), groups of 20 F1 animals of each sex were selected as breeding pairs (avoiding sibling
matings), cohabitated for 7 days, and then separated. Reproductive endpoints (numbers of live
and dead pups, sexes of pups, and total pup weight by sex) were evaluated on PND 1 of the
F2 offspring; there was no further evaluation of the F2 pups. Estrous cycle (time spent in estrous
stages, cycle length, number of cycles, number of cycling females, and number of females with
regular cycles) was evaluated using 12-day vaginal smears beginning 4 days after the last
delivery. Terminal evaluations of F1 adults (time from separation to terminal sacrifice not
reported) were the same as those described above for F0 adults, with the addition of hematology
(i.e., Hgb, Hct, MCV, MCH, MCHC, mean platelet volume, erythrocyte morphology, and
erythrocyte, leukocyte, and platelet counts).
No treatment-related mortalities or clinical signs of toxicity were observed in
F0 generation BALB/c mice exposed to dietary potassium dichromate (NTP, 1997). Mortalities
occurred in eight animals (four low-dose males, one mid-dose male, and three mid-dose
females); however, since no mortalities were observed in the high-dose group, NTP (1997)
concluded that these deaths were not related to treatment. Terminal body weight of males in all
treatment groups was comparable to controls; mean body weight of females in the high-dose
groups was decreased by 7% (p < 0.05). In general, food consumption was increased in
treatment groups. Based on measured food consumption and body weights during the
cohabitation period, NTP (1997) calculated average daily doses in F0 males and females of 0,
19.4, 38.6, or 85.7 mg potassium dichromate/kg-day (equivalent to 0, 6.8, 13.6, or 30.3 mg
hexavalent chromium/kg-day, respectively). During lactation, sporadic decreases in body
weights of dams in the mid- and high-dose groups were observed, but body weights at the end of
lactation (PND 21) were similar to controls; food consumption during lactation was similar
between control and treatment groups. Based on measured food consumption and body weights,
NTP (1997) calculated average daily doses in lactating F0 females of 0, 32.8, 69.0 or 143.1 mg
potassium dichromate/kg-day (equivalent to 0, 11.6, 24.4, or 50.5 mg hexavalent chromium/kg-
day, respectively). At the terminal evaluations of F0 animals, absolute (but not relative) liver
weights were increased by 17% (p < 0.05) and 22% (p < 0.05) in high-dose males and females,

respectively, compared with controls. No other changes in organ weights were observed. No
treatment-related histopathological findings were observed in the F0 generation. Although
various hepatic lesions were observed, including cytoplasmic vacuolization, the study authors
concluded that these findings were not treatment-related, since incidence data did not show a
relationship with dose. Evaluations of male reproductive tissues did not reveal any treatment-
related effects. In the F0 generation, no treatment-related effects on reproductive outcomes,
including pregnancy index, mean cumulative time to litter, litter size, live and dead pups/litter,
live pup weight, and sex ratio, were observed.
Evaluations conducted on F1 pups during lactational exposure showed no effects on pup
survival (NTP, 1997). On PND 21, weight of high-dose male pups was decreased by 16%
compared with controls, but the decrease was not statistically significant. From weaning to
sexual maturity, two mortalities occurred (one control male and one high-dose male). No
treatment-related clinical signs of toxicity were observed. At the initiation of the F1 breeding
phase (approximately PND 74), mean body weights of mid-dose females were decreased by 6%
compared with controls and by 9% in high-dose F1 males and females (statistical significance not
reported). Food consumption was generally increased during the period from weaning to sexual
maturity. Based on measured food consumption and body weights, NTP (1997) calculated
average daily doses in F1 animals of 0, 22.4, 45.5 or 104.9 mg potassium dichromate/kg-day
(equivalent to 0, 7.9, 16.1, or 37.1 mg hexavalent chromium/kg-day, respectively).
Hematological analysis at terminal sacrifice of F1 adults revealed slight erythrocyte microcytosis
based on the following observations (comparisons to controls, statistical significance not
reported): MCV decreased by 3% in mid- and high-dose males and by 2, 3, and 4% in low-,
mid-, and high-dose females, respectively; MCH decreased by 3% in high-dose males; and Hgb
decreased by 5% in high-dose F1 females. No changes in erythrocyte morphology were
observed. Relative kidney weight was increased by 5% in mid-dose females, but no other organ
weight changes were observed. No treatment-related histopathological findings were observed.
Although various hepatic lesions were observed, including cytoplasmic vacuolization, NTP
(1997) concluded that findings were not treatment-related, since incidence data did not show a
relationship with dose. Evaluations of male reproductive tissues and female estrous cycle did not
reveal any treatment-related effects. In the F1 generation, no treatment-related effects on
reproductive outcomes, including pregnancy index, mean cumulative time to litter, gestation
length, litter size, live and dead pups/litter, and sex ratio, were observed. Live pup weight of
females in the high-dose group was decreased by 11% (p < 0.05) compared to controls, but no
decrease was observed for live pup weight of males or of combined males and females.
In conclusion, NTP (1997) identified a LOAEL for parental toxicity in the F1 generation
of 7.9 mg hexavalent chromium/kg-day in females exposed to potassium dichromate in the diet
based on erythrocyte microcytosis (slight decrease in MCH); a NOAEL for parental toxicity in
the F1 generation was not established because effects were seen at the lowest dose tested.

Although NTP (1997) did not specifically identify a NOAEL for reproductive effects, in the
absence of reproductive findings, the highest dose tested in this study is identified by EPA as a
free-standing NOAEL for effects of dietary hexavalent chromium exposure on fertility and on
male and female reproductive organ histology and weights (i.e., 30.3 mg hexavalent
chromium/kg-day in F0 mice and 37.1 mg hexavalent chromium/kg-day in F1 mice).
4.3.3. Effects of Pregestational Exposure on Reproductive Outcome and Fetal

Development
Kanojia et al., 1996
Kanojia et al. (1996) administered adult Swiss albino female rats (20/group) drinking
water containing 0, 250, 500, or 750 mg hexavalent chromium/L (as potassium dichromate) for
20 days prior to gestation. During the exposure and gestational periods, body weights and water
intake were recorded daily. At the end of the exposure period, rats were mated overnight with
untreated males. Following mating, the mating index (percentage of mated females) and the
fertility index (percentage of pregnant females) were determined. On GD 19, 10 rats/group were
sacrificed and the numbers of copora lutea, fetuses/litter, live and dead fetuses, and resorptions;
pre- and postimplantation losses; and fetal and placental weights were recorded and fetuses were
examined for internal abnormalities (one third of fetuses) and external and skeletal abnormalities
(remaining fetuses). In the remaining 10 rats/group, estrous cycle length was evaluated for
12 consecutive cycles. Based on drinking water consumption during the exposure period,
Konijia et al. (1996) reported daily hexavalent chromium intakes of 6.4, 12.2, and 15.3 mg
hexavalent chromium/rat-day. The study report did not include data on body weights over the
course of the 20-day treatment period, although it is likely that treatment-related effects on body
weight occurred during the exposure period, as significant decreases in gestational weight gain
were observed in all treatment groups (decreases of approximately 8, 14, and 21% in the low-,
mid-, and high-dose groups, respectively, compared to controls). Thus, in the absence of data on
the effect of treatment on body weights during the exposure period, daily doses of hexavalent
chromium in terms of body weight (e.g., mg hexavalent chromium/kg-day) cannot be accurately
estimated.
No mortalities or clinical signs of toxicity in dams were observed. Dose-related
decreases in mating and fertility indices were observed; in the high-dose group, mating and
fertility indices were decreased by 60 and 68%, respectively, compared to controls (statistical
significance not reported). In all treatment groups, the number of live fetuses was decreased, the
numbers of resorptions and postimplantation loss were increased, and placental weight was
increased. In the mid- and high-dose groups, numbers of corpora lutea and implantations were
decreased and preimplantation losses were increased. No treatment-related effects were
observed for fetal weight or crown-rump length. Examination of fetuses showed gross
abnormalities in the high-dose group, including patches of subdermal hemorrhage, kinky tail,

short tail, and dropping wrist. Skeletal abnormalities were also observed, including reduced
caudal ossification in mid- and high-dose groups and reduced parietal and inter-parietal
ossification in the high-dose group. No visceral abnormalities were observed. Postpartum
estrous cycle length was significantly increased by 37% (p < 0.05) in the high-dose group.
Results of this study show that 20-day pregestational exposure of Swiss albino rat dams
to hexavalent chromium adversely affected reproductive outcomes (decreased number of live
fetuses and increased number of resorptions and postimplantation loss) at the lowest drinking
water concentrations of potassium dichromate tested (≥250 mg hexavalent chromium/L or
≥6.4 mg hexavalent chromium/rat-day) and produced adverse developmental effects (gross and
skeletal abnormalities) at the highest drinking water concentrations tested (750 mg hexavalent
chromium/L or 15.3 mg hexavalent chromium/rat-day). Because of the lack of reporting of body
weight data over the course of the study, NOAELs and/or LOAELs, expressed in mg hexavalent
chromium/kg-day, could not be derived from this study by EPA.
Kanojia et al., 1998

Kanojia et al. (1998) administered adult Druckrey female rats (20/group; mean initial
body weight 80 g) drinking water containing 0, 250, 500, or 750 mg hexavalent chromium/L (as
potassium dichromate) for 3 months prior to gestation. This study was designed to follow the
same protocol as that used in the Kanojia et al. (1996) study (described above). However, at the
end of the 3-month exposure period, rats in all treatment groups were acyclic (persistent
diestrous phase). Therefore, since mating could not take place immediately following
completion of the exposure period, rats were held for an additional 15–20 days (treatment-free),
during which time the estrous cycle resumed.
During the exposure period, mortality occurred in 15 and 10% of rats in the mid- and
high-dose groups, respectively; no deaths occurred in the control or low-dose groups. Clinical
signs of toxicity observed during the exposure period in the mid- and high-dose groups included
hair loss and lethargic and aggressive behavior. At the end of the exposure period, body weight
was significantly (p < 0.05) decreased by approximately 18 and 24% in the mid- and high-dose
groups, respectively, compared with controls. Kanojia et al. (1998) reported average hexavalent
chromium intakes (based on water consumption) of 5.57, 10.18, and 13.56 mg hexavalent
chromium/rat-day in the low-, mid-, and high-dose groups, respectively. Using these daily
intake levels and the mean initial body weight of 80 g, daily doses of 70, 127, and 170 mg
hexavalent chromium/kg-day for the low-, mid-, and high-dose groups, respectively, were
estimated. During the post-exposure gestational period, maternal weight gain was significantly
(p < 0.05) decreased by 17 and 22% in the mid- and high-dose groups, respectively, compared
with controls. The mating index was decreased by 30, 40, and 60% and the fertility index was
decreased by 32, 41, and 49% in the low-, mid-, and high-dose groups, respectively, compared
with controls (statistical significance not reported). In all treatment groups, pre- and

postimplantation losses were significantly (p < 0.05) increased, with increases in the high-dose
group reaching 3.1- and 4.2-fold, respectively. In the mid- and high-dose groups, the numbers of
implantations, live fetuses, and resorptions were significantly (p < 0.05) increased. Assessments
of fetuses (on a per litter basis compared with controls) showed the following (significant
difference compared with controls; p < 0.05): decreased fetal weight (all treatment groups);
decreased crown-rump length (mid- and high-dose groups); gross external abnormalities,
including subdermal hemorrhagic patches and drooping wrists in all treatment groups and kinky
and short tail in mid- and high-dose groups; and skeletal abnormalities, including decreased
caudal ossification in all treatment groups and reduced parietal and interparietal ossification in
mid- and high-dose groups. No internal abnormalities in fetuses were observed. Postpartum
estrous cycle length was significantly (p < 0.05) increased in all treatment groups, with increases
reaching approximately 1.7-fold in the high-dose group.
Results of this study show that 3-month pregestational exposure of Druckrey rat dams to
hexavalent chromium as potassium dichromate adversely affected reproductive outcomes
(increased pre- and postimplantation losses) and produced adverse developmental effects
(decreased fetal weight and external and skeletal abnormalities) at all drinking water
concentrations tested (≥250 mg hexavalent chromium/L or approximately ≥70 mg hexavalent
chromium/kg-day). Thus, a LOAEL of 70 mg hexavalent chromium/kg-day was identified from
this study by EPA.
Junaid et al., 1996a

Junaid et al. (1996a) administered Swiss albino female mice drinking water containing 0,
250, 500, or 750 mg hexavalent chromium/L (as potassium dichromate) from days 6 to 14 of
gestation. The study followed the same protocol and conducted the same evaluations as those
reported in the study by Kanojia et al. (1996) (described above), except that estrous cycle length
was not evaluated. Evaluations on reproductive outcomes and developmental effects were
conducted in 10 mice/group.
No clinical signs of toxicity were observed in mice during the exposure period. In the
high-dose group, mortality occurred in 20% of animals; the cause of death was not established.
Based on drinking water consumption monitored during the exposure period, the study authors
reported daily hexavalent chromium intake levels of 1.9, 3.56, and 5.23 mg hexavalent
chromium/mouse-day in the low-, mid-, and high-dose groups, respectively. No treatment-
related effects were observed on body weight (data not reported); thus, using the reported mean
initial body weight of 30 g, daily doses of 63, 119, and 174 mg hexavalent chromium/kg-day for
the low-, mid-, and high-dose groups, respectively, were estimated. During the gestational
period, maternal weight gain in the low- and mid-dose groups was comparable to controls; no
weight gain was observed during gestation in high-dose group dams. In the low-dose group,
postimplantation loss was significantly (p < 0.05) increased compared with controls (control:

0%; low-dose group: 17.5%); no effects were observed for the numbers of corpora lutea,
implantations, live fetuses, or resorptions or for preimplantation loss. In the mid-dose group, the
numbers of implantation and live fetuses were significantly (p < 0.05) decreased and the
numbers of resorptions and pre- and postimplantation losses were significantly (p < 0.05)
increased; no effect on the number of corpora lutea was observed. In the high-dose group, no
litters were produced and implantation sites were completely absent; corpora lutea were present,
but numbers were decreased by 44% compared to controls. Assessments of fetuses (on a per
litter basis compared with controls) showed the following (significant difference compared to
controls; p < 0.05): decreased fetal weight and length in the low- and mid-dose groups; gross
(external) abnormalities, including subdermal hemorrhagic patches and short and kinky tail in
the mid-dose group; and skeletal abnormalities, including reduced caudal ossification in the low-
and mid-dose groups and reduced parietal and interparietal ossification in the mid-dose group.
No internal abnormalities in fetuses were observed.
Thus, at all drinking water concentrations of potassium dichromate tested (≥250 mg
hexavalent chromium/L or approximately ≥63 mg hexavalent chromium/kg-day), pregestational
exposure of Swiss albino female mice for 20 days produced adverse effects on reproductive
outcome (decreased fertility) and fetal development (decreased fetal body weight and delays in
skeletal development). Thus, EPA identified a LOAEL of 63 mg hexavalent chromium/kg-day
from this study.
4.3.4. Effects of Gestational and/or Lactational Exposure on Reproductive Outcome and

Fetal Development
Elsaieed and Nada, 2002
Effects of gestational exposure to hexavalent chromium were investigated in Wistar rats
(Elsaieed and Nada, 2002). Groups of 10 pregnant rats (mean initial body weight of 170 g) were
administered drinking water containing 0 or 50 mg hexavalent chromium/L as potassium
dichromate on GDs 6 through 15. During the exposure period, dams were evaluated for clinical
signs of toxicity, body weights, and food and drinking water consumption. One day before
delivery, rats were sacrificed and the following were evaluated: numbers of corpora lutea, pre-
and postimplantation losses, resorptions, and live and dead fetuses; fetal weight; and visceral and
skeletal anomalies.
No mortalities or clinical signs of toxicity were observed. Elsaieed and Nada (2002)
stated that food and drinking water consumption was comparable between control and treatment
groups, although data were not reported. Gestational weight gain was significantly (p < 0.05)
decreased by 40% in treated dams, compared with controls. Based on an average gestational
body weight of 177 g (average calculated using body weights at mating and at the end of
gestation) and the allometric equation for drinking water consumption for laboratory mammals
(0.10 × body weight0.7377; U.S. EPA, 1988), a daily dose of 7.9 mg hexavalent chromium/kg-day

was estimated by EPA. In this study, treatment of rats with hexavalent chromium resulted in
significant (p < 0.05) increases in preimplantation loss/litter (2.1 vs. 0 in control),
postimplantation loss/litter (1.5 vs. 0), resorptions/litter (1.2 vs. 0), and dead fetuses/litter (1.2 vs.
0) and decreases in live fetuses/litter (1.5 vs. 6.8 in control) and fetal weight (33% decrease). In
the exposed group, increased litters with fetal abnormalities or malformations were observed
including visceral (renal pelvis dilation: 2.1/litter) and skeletal (incomplete skull ossification:
1.0/litter) changes; no control fetuses showed these changes.
The results showed that exposure of pregnant Sprague-Dawley rats to drinking water
containing 50 mg hexavalent chromium/L as potassium dichromate (approximately 7.9 mg
hexavalent chromium/kg-day) on GDs 6–15 produced adverse effects on reproductive outcome
and fetal development. Thus, EPA identified a LOAEL of 7.9 mg hexavalent chromium/kg-day
from this study.
Bataineh et al., 2007

Reproductive outcomes were evaluated in adult female rats (age not specified) orally
exposed to potassium dichromate for 3 days following mating (Bataineh et al., 2007). Groups of
10 successfully mated female Sprague-Dawley rats were administered daily doses of 0 or 25 mg
potassium dichromate/rat (equivalent to 8.8 mg hexavalent chromium/rat-day or approximately
35 mg hexavalent chromium/kg-day, based on the average reported body weight of 245 g at
mating) in saline daily by gavage on GDs 1–3 or 4–6. On GD 20, rats were sacrificed and the
numbers of implantation sites, live fetuses, and resorptions along the uterine horns were
recorded; fetuses were not assessed for external, skeletal, or visceral abnormalities.
In rats treated with potassium dichromate on GDs 1–3, no pregnancies, implantations,
resorptions, or viable fetuses were observed, compared with 10/10 pregnancies, 8.2 implantations/
female, 8.2 live fetuses/female, and 0/82 resorptions in controls. In rats treated on GDs 4–6, the
numbers of pregnant rats and implantations/female were comparable to values in the control
group. However, the number of viable fetuses was decreased by 69% (p < 0.001) and the
percentage of resorptions per implantations was increased by 222% (p < 0.001). The study report
did not indicate if clinical signs of toxicity were observed in chromium-treated dams, and no
additional measures to assess systemic toxicity were reported.
The results indicate that short-term gavage exposure of Sprague-Dawley rat dams to
potassium dichromate at a dose of 35 mg hexavalent chromium/kg-day on GDs 1–3 completely
impaired implantation; exposure on GDs 4–6 markedly increased resorptions and decreased the
number of viable fetuses, compared with controls. Thus, a LOAEL of 35 mg hexavalent
chromium/kg-day was identified from this study by EPA.

Trivedi et al., 1989
Effects on reproductive outcome and fetal development were observed in ITRC-bred
albino mice administered hexavalent chromium in drinking water (Trivedi et al., 1989). Groups
of 10–13 pregnant mice (average initial body weight of 30 g) were administered drinking water
containing 0, 250, 500, or 1,000 mg hexavalent chromium/L (as potassium dichromate) during
the entire gestational period. Dams were observed daily for mortality, clinical signs of toxicity,
body weight, and water consumption. On GD 19, dams were sacrificed and the following were
recorded: numbers of corpora lutea, total implantations, live and dead fetuses, and pre-
implantation and postimplantations losses; placental weight; fetal weight and crown-rump
length; number of stunted fetuses; and sex ratio per litter. In addition, fetuses were examined for
external (all fetuses), internal (approximately one-third of fetuses), and skeletal (remaining
fetuses) anomalies.
No mortalities or clinical signs of toxicity were observed in the dams. In the low-dose
group, body weight gain was comparable to controls; however, body weight gain was
significantly decreased by 21% (p < 0.05) in the mid-dose group, and dams in the high-dose
group lost weight during treatment. Daily hexavalent chromium intakes were reported as 1.76,
3.6, and 7.03 mg hexavalent chromium/mouse-day in the low-, mid-, and high-dose groups,
respectively, based on measured drinking water consumption. Using average body weights for
the gestational period (36.8, 36.6, and 29.4 g in the low-, mid-, and high-dose groups,
respectively; calculated for this review using: [average initial body weight + body weight at the
end of gestation]/2) and reported daily chromium intakes, daily doses of 48, 98, and 239 mg
hexavalent chromium/kg-day were estimated. In low-dose mice, the percentages of resorptions
and postimplantation loss were significantly increased (p < 0.001) to 33 and 36%, respectively,
compared with 10 and 1.7%, respectively, in controls; the number of litters, litter size, number of
copora lutea, and placental weight in the low-dose group were comparable to controls. In the
mid-dose group, the percentages of resorptions and postimplantation losses were significantly (p
< 0.001) increased to 52 and 88%, respectively. In addition, in the mid-dose group, litter size
was significantly decreased by 44% (p < 0.01) compared with controls, and the percentage of
preimplantation loss was increased to 26.2% (p < 0.001), compared with 3.6% in controls. No
treatment-related effects on placental weight were observed in the low- or mid-dose groups. In
the high-dose group, no litters were produced and implantation sites were completely absent. In
the low- and mid-dose groups, mean fetal crown-rump lengths were decreased (p < 0.001) by 17
and 27%, respectively, and mean fetal weights were decreased (p < 0.001) by 31 and 44%,
respectively. Sex ratio was unaffected by treatment. Examination of fetuses for external
anomalies showed no effects in the low-dose group; in the mid-dose group, tail kinking and
subdermal hemorrhagic patches and streaks were observed. An increase in the incidence of
minor skeletal anomalies was observed in fetuses in the low-dose (reduced ossification of the
cranium) and mid-dose (reduced ossification of the cranium, forelimb, hindlimb, sternebrae, and

thoracic and caudal vertebrae and reduced number of ribs) groups. No internal anomalies were
observed.
From this study, EPA identified a LOAEL and NOAEL for maternal toxicity, and a
LOAEL for developmental effects. The LOAEL and NOAEL values for maternal toxicity, based
on decreased body weight gain in ITRC-bred albino mice exposed to potassium dichromate in
drinking water throughout gestation, were 98 and 48 mg hexavalent chromium/kg-day,
respectively. Based on increased resorptions and postimplantation loss, and decreased fetal
length and weight, the lowest concentration tested (250 mg hexavalent chromium/L or 48 mg
hexavalent chromium/kg-day) was identified as a LOAEL for developmental effects.
Junaid et al., 1996b

Junaid et al. (1996) evaluated the effects of oral exposure of pregnant mice to hexavalent
chromium on reproductive outcome and fetal development. Groups of 10 successfully mated
Swiss albino female mice (average initial body weight of 30 g) were administered drinking water
containing 0, 250, 500, or 750 mg hexavalent chromium/L (as potassium dichromate) on GDs 6
though 14. Throughout the exposure period, dams were evaluated daily for clinical signs of
toxicity, body weight, and drinking water consumption. On GD 19, dams were sacrificed and
evaluations of dams and fetuses were conducted as described by Trivedi et al. (1989)
(summarized above).
No mortalities or clinical signs of toxicity were observed in the dams. Gestational weight
gain was significantly (p < 0.05) decreased in the mid- and high-dose groups by 8 and 32%,
respectively, but was comparable to controls in the low-dose group. Daily hexavalent chromium
intakes were reported as 2.00, 3.75, or 5.47 mg chromium/mouse-day in the low-, mid-, and
high-dose groups, respectively, based on measured drinking water consumption. Using average
body weights for the gestational period (37.6, 37.2, and 35.9 g in the low-, mid-, and high-dose
groups, respectively; calculated for this report using: [average initial body weight + body weight
at the end of gestation]/2) and reported daily chromium intakes, daily doses of 53, 101, and
152 mg hexavalent chromium/kg-day in the low-, mid-, and high-dose groups, respectively, were
estimated. The number of resorptions was significantly (p < 0.05) increased in all treatment
groups, with increases reaching 7.7-fold in the high-dose group. In the mid- and high-dose
groups, significant (p < 0.05) decreases in the total number of fetuses and increases in the
numbers of dead fetuses and resorption sites were observed. Fetal weight was significantly
(p < 0.05) decreased by 13 and 19% in the mid- and high-dose groups, respectively; no
treatment-related effects were observed on fetal length. Gross external examination of fetuses
showed significant (p < 0.05) increases in the incidences of minor abnormalities (subdermal
hemorrhagic patches, drooping wrist, kinky and short tail) in the high-dose group. Examination
of fetuses for skeletal abnormalities showed significant (p < 0.05) increases in the incidences of
reduced caudal ossification in the mid- and high-dose groups and of reduced nasal, frontal,

parietal, interparietal, carpals, and tarsals ossification. No external or skeletal abnormalities were
observed in fetuses in the low-dose group. No visceral abnormalities were observed in any
treatment group.
Junaid et al. (1996b) concluded that oral exposure of dams during the organogenesis
phase of gestation produced adverse effects in embryos and during fetal development. EPA
identified LOAEL and NOAEL values for maternal toxicity of 101 and 53 mg hexavalent
chromium/kg-day, respectively, based on decreased body weight gain in Swiss albino mice
administered potassium dichromate in drinking water on GDs 6–14. Based on reduced number
of implantation sites, the lowest dose tested (approximately 53 mg hexavalent chromium/kg-day)
in this study was identified by EPA as a developmental LOAEL.
Junaid et al., 1995

The effects of late gestational exposure to hexavalent chromium on reproductive outcome
and fetal development were evaluated in mice (Junaid et al., 1995). Groups of 10 successfully
mated Swiss albino female mice (average initial body weight of 30 g) were administered
drinking water containing 0, 250, 500, or 750 mg hexavalent chromium/L (as potassium
dichromate) on GDs 14 though 19. Throughout the exposure period, dams were evaluated daily
for clinical signs of toxicity, body weight, and drinking water consumption. On GD 19, dams
were sacrificed and evaluations of dams and fetuses were conducted as described by Trivedi et
al. (1989) (summarized above).
No mortalities or clinical signs of toxicity were observed in the dams. Gestational weight
gain was significantly (p < 0.05) decreased in the mid- and high-dose groups by 11 and 26%,
respectively, but was comparable to controls in the low-dose group. No data on drinking water
consumption were reported; however, it is likely that daily doses were similar to those calculated
for the study by Junaid et al. (1996b) (e.g., approximately 53, 101, and 152 mg hexavalent
chromium/kg-day in the low-, mid-, and high-dose groups, respectively), which used the same
mouse strain and drinking water concentrations, and a similar study design. In the mid- and
high-dose groups, the numbers of dead fetuses and postimplantation losses were significantly (p
< 0.05) increased; the numbers of corpora lutea and total fetuses per litter were similar to
controls in all treatment groups. Fetal weight and length were significantly decreased in all
treatment groups, with decreases reaching approximately 47 and 29%, respectively, in the high-
dose group. Gross external examination of fetuses showed significant (p < 0.05) increases in the
incidences of minor abnormalities in the mid-dose (drooping wrists) and high-dose (drooping
wrists, subdermal hemorrhagic patches, kinky and short tail) groups. Examination of fetuses for
skeletal abnormalities showed significant (p < 0.05) increases in the incidences of reduced caudal
ossification in all treatment groups, reduced tarsals ossification in mid- and high-dose groups,
and reduced nasal, parietal, interparietal, carpals, and metatarsals ossifications in the high-dose
group. No visceral abnormalities were observed in any treatment group.

From this study, EPA identified a LOAEL and NOAEL for maternal toxicity, and a
LOAEL for developmental effects. The NOAEL and LOAEL values for maternal toxicity, based
on decreased body weight gain in Swiss albino mice administered potassium dichromate in
drinking water on GDs 14–19, were 53 and 101 mg hexavalent chromium/kg-day, respectively.
Based on reduced fetal weight and length and increased incidence of reduced caudal ossification
in all treatment groups, the lowest dose tested (approximately 53 mg hexavalent chromium/kg-
day) was identified as a developmental LOAEL.
Al-Hamood et al., 1998

The effects of gestational and lactational exposure of mice to hexavalent chromium on
sexual maturation and fertility in offspring were investigated by Al-Hamood et al. (1998). On
GD 12 through day 20 of lactation, groups of 25 pregnant Swiss strain BALB/c mice (mean
initial body weight of 25 g) were administered drinking water containing 0 or 1,000 mg
potassium dichromate/L (equivalent to 353 mg hexavalent chromium/L). Based on drinking
water consumption by dams, daily hexavalent chromium intakes of 2.1 and 1.7 mg hexavalent
chromium/mouse-day were calculated for the gestational and lactational periods, respectively.
No data on body weights of dams were reported; however, since other studies have shown
decreased maternal weight gain in pregnant mice exposed to drinking water containing
≥176 hexavalent chromium/L) (Junaid et al., 1996b, 1995), it is likely that treatment-related
decreases in maternal weight gain occurred. Therefore, given this uncertainty, daily hexavalent
chromium doses expressed in terms of body weight cannot be accurately estimated for this study.
At birth, litters were culled to eight pups per female and offspring were weaned on PND 21;
from weaning to day 60 of age, offspring received control drinking water. From PND 20 to the
onset of puberty, female offspring were examined for time to vaginal opening. Fertility in
offspring was assessed at day 60 of age; male offspring were mated with untreated females and
female offspring were mated with untreated males for 10 days. At completion of the mating
period, females were examined for numbers of pregnant females, implantations, viable fetuses,
and resorptions. Additional groups (n = 9–12) of offspring were sacrificed on day 50 of age, and
body weights and weights of reproductive organs (paired testes, seminal vesicles, and preputial
glands in males and paired ovaries and uteri in females) were determined.
In female offspring, time to vaginal opening was significantly (p < 0.001) increased from
24.6 days in controls to 27.1 days in treated mice. Mating studies in female offspring showed
decreased numbers of pregnant females (35% decrease; p < 0.025), implantations (12% decrease;
p < 0.05), and viable fetuses (14% decrease; p < 0.05). No treatment-related effects on female
body weight or relative weights of reproductive organs were observed. In male offspring, no
treatment-related effects were observed in mating studies or on body weights or weights of
reproductive organs.

The results indicate that gestational and lactational exposure of BALB/c mouse dams to
drinking water containing 353 mg hexavalent chromium/L as potassium dichromate resulted in
impaired reproductive development and function in female offspring. Because of the lack of
reporting of body weight data over the course of the study, EPA could not identify NOAEL or
LOAEL values, expressed in mg hexavalent chromium/kg-day, for this study.
Banu et al., 2008

Banu et al. (2008) investigated the effects of lactational exposure to hexavalent
chromium on sexual development of female rat offspring. Groups of 18 lactating Wistar rats
were administered drinking water containing 200 mg potassium dichromate (equivalent to
70.6 mg hexavalent chromium/L) on postpartum days 1 through 21. No specific assessments of
dams were conducted. Banu et al. (2008) noted that toxic effects in dams were not “significant,”
although no additional information regarding maternal toxicity or data on body weights or
drinking water consumption in dams were reported. As discussed above, exposure of laboratory
animals to hexavalent chromium in drinking water may result in decreased body weight and
drinking water consumption; thus, in the absence of data on body weight and drinking water
consumption in dams, daily doses of hexavalent chromium cannot be accurately estimated for
this study. At birth, litters were culled to four female pups per dam. Following weaning on
PND 21, pups were separated from dams. Pups (n = 24) were evaluated for the onset of puberty
by daily examination for vaginal opening. After the onset of puberty, the time spent in each
estrous cycle phase (proestrous, estrous, metestrous, and diestrous) was determined by analysis
of vaginal smears (n = 24). On PNDs 21, 45, and 65, pups (n = 24, at each time point) were
sacrificed; at each time point, blood was analyzed for hormones (estradiol, progesterone,
testosterone, LH, follicle-stimulating hormone [FSH], growth hormone [GH], and prolactin) and
ovaries were examined for the number of follicles and follicle development stage (primordial,
primary, secondary, and antral).
The onset of puberty was significantly (p < 0.05) increased from 33 days in control rats to
55 days in treated rats. Estrous cycle phase was also altered in treated rats, with the time spent in
diestrous significantly (p < 0.05) increased by approximately 1.4-fold compared with controls
(data presented graphically); time spent in other estrous phases was unaffected by treatment.
Evaluations of ovaries on PNDs 21 and 45 showed significant (p < 0.05) decreases in the
numbers of primordial, primary, secondary, and antral follicles in treated rats compared with
control rats; on PND 65, the numbers of primordial and primary follicles were also decreased in
treated rats. At the 21- and 45-day assessments in treated rats, plasma concentrations of
estradiol, progesterone, testosterone, GH, and prolactin were significantly (p < 0.05) decreased
(by approximately 40–60%) and concentrations of FSH were significantly increased (by
approximately 40%), compared with controls. Similar effects were observed at the 65-day

assessment, except that FSH concentrations in treatment and control groups were comparable.
Plasma LH concentration was not affected by treatment at any time point.
The results indicate that lactational exposure of Wistar rat dams to drinking water
containing 70.6 mg hexavalent chromium/L as potassium dichromate resulted in delayed onset of
puberty and follicular development and impaired ovarian steroidogenesis in female offspring;
male offspring were not assessed for possible effects on sexual maturation. Because of the lack
of reporting of body weight data over the course of the study, EPA could not identify NOAEL or
LOAEL values, expressed in mg hexavalent chromium/kg-day, for this study.
4.4. SUMMARY OF HUMAN AND ANIMAL STUDIES BY ROUTES OF EXPOSURE

OTHER THAN ORAL
Human exposure to chromium compounds by inhalation has been studied in the chromate
production, chrome plating and chrome pigment, ferrochromium production, gold mining,
leather tanning, and chrome alloy production industries. Results of occupational epidemiologic
studies of chromium-exposed workers have consistently demonstrated excess risks for lung
cancer with chromium exposure. Epidemiological studies of chromate production plants in
Japan, Great Britain, West Germany, and the United States have revealed a correlation between
occupational exposure to chromium (specific form not identified) and lung cancer (Mancuso,
1997; Davies, 1984; Watanabe and Fukuchi, 1975; Frentzel-Beyme, 1983; Langard and
Vigander, 1983; Korallus et al., 1982; Alderson et al., 1981; Haguenor et al., 1981; Satoh et al.,
1981; Hayes et al., 1979; Hill and Ferguson, 1979; Ohsaki et al., 1978; Sano and Mitohara, 1978;
Mancuso, 1975; Enterline, 1974; Taylor, 1966; Todd, 1962; Bidstrup and Case, 1956; Brinton et
al., 1952; Bidstrup, 1951; Mancuso and Hueper, 1951; Baejter, 1950a,b; Machle and Gregorius,
1948). Similarly, studies of chrome pigment workers in the United States (Hayes et al., 1989),
England (Davies, 1984, 1979, 1978), Norway (Langard and Vigander, 1983; Langard and
Norseth, 1975), and in the Netherlands and Germany (Frentzel-Beyme, 1983) have demonstrated
an association between occupational chromium exposure and lung cancer. Finally, several
studies of the chrome plating industry have demonstrated a positive relationship between cancer
and exposure to chromium compounds (Sorahan et al., 1987; Royle, 1975).
Animal data via the inhalation route of exposure are consistent with the findings of
human epidemiological studies of hexavalent chromium. Hexavalent chromium compounds
were carcinogenic in animal assays producing the following tumor types: lung tumors following
inhalation of aerosols of sodium chromate and pyrolized Cr(VI)/Cr(III) oxide mixtures in rats
(Glaser et al., 1986), lung tumors following intratracheal administration of sodium dichromate in
rats (Steinhoff et al., 1983), intramuscular injection site tumors for various Cr(VI) compounds in
Fischer 344 and Bethesda Black rats and in C57BL mice (Furst et al., 1976; Payne, 1960a);
intrapleural implant site tumors for various Cr(VI) compounds in Sprague-Dawley and Bethesda
Black rats (Hueper and Payne, 1962; Hueper, 1961; Payne, 1960b), and intrabronchial

implantation site tumors for various Cr(VI) compounds in Wistar rats (Levy and Martin, 1983;
Laskin et al., 1970; Levy, (as cited in NIOSH, 1975).
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF

ACTION
4.5.1. Genotoxicity Studies
The mutagenic potential of hexavalent chromium has been studied extensively. Although
study results vary with specific test systems, experimental conditions, and the type of hexavalent
chromium compound tested, results of in vitro and in vivo studies provide substantial evidence
for the mutagenic activity of hexavalent chromium. A general summary of this evidence is
provided in Table 4-20. As discussed in detail in Section 4.4.2 (Intracellular Reduction),
mutagenicity of hexavalent chromium is mediated through the generation of the highly reactive
chromium intermediates penta- and tetravalent chromium, reactive oxygen species, and trivalent
chromium formed during the intracellular reduction of hexavalent chromium. These chromium
and oxygen species can react with DNA, leading to oxidative DNA damage, chromium-DNA
adducts, DNA strand breaks, and chromosomal aberrations.

Table 4-20. Evidence of mutagenicity of hexavalent chromium compounds in experimental systems
In vitro studies (nonmammalian In vitro studies In vivo studies

cells) (mammalian cells) (Drosophila melanogaster or mammals)
DNA Chromosomal DNA Chromosomal
Chemical DNA damage Mutations damage Mutations damage damage Mutations damage
Ammonium chromate ND + ND ND ND ND ND ND
Calcium chromate ND + ND ND + ND + (D) ND
Chromic acid ND + ND ND + ND + (D) ND
Potassium chromate + + + + + + (M) + (D) + (M)
+ (M)
Potassium dichromate + + + + + + (M) + (D) + (M)
Sodium chromate ND + + ND + ND ND ND
Sodium dichromate ND + + ND + + (M) + (D) ND
Sodium dichromate ND + ND + ND ND ND + (M)
dihydrate
+ = positive results; (D) = study in D. melanogaster; (M) = study in laboratory mammal; ND = no data identified for this review

4.5.1.1. Genotoxicity Assays in Experimental Systems
The mutagenic activity of hexavalent chromium has been demonstrated in numerous
studies using both in vitro and in vivo experimental systems. In in vitro test systems (see
Tables 4-21 and 4-22 for studies in nonmammalian and mammalian cells, respectively),
hexavalent chromium compounds have mostly tested positive for gene mutations (including
reverse mutations, frame shift mutations, and base pair substitutions) and DNA damage
(including DNA-protein crosslinks) in bacterial cells (Salmonella typhimurium, Escherichia coli,
Bacillus subtilis). Reverse mutations were observed in multiple species and strains, including
those that are sensitive to frameshift mutagens (S. typhimurium TA97, TA98, TA1537, and
TA1538), G/C base-pair substitution mutagens (S. typhimurium TA100 and TA1535), and
A/T base-pair substitution mutagens caused by oxidizing and/or cross-linking agents
(S. typhimurium TA102; E. coli WP2uvrA and WP2uvrA/pKM101). Positive results were also
found for forward mutations and mitotic gene conversion in yeast (Saccharomyces cerevisiae),
and DNA damage (DNA strand breaks, fragmentation, DNA-protein crosslinks, DNA-DNA
crosslinks), chromosomal damage (sister chromatid exchanges and chromosomal aberrations),
and DNA synthesis inhibition in mammalian cell lines and primary cultures (including primary
cultures of human gastric mucosal cells, respiratory tract cells, and lymphocytes).

Table 4-21. In vitro genotoxicity studies of hexavalent chromium in nonmammalian cells
Results
Without
Endpoint Chemical form Test system activation With activation Reference
Reverse mutations Ammonium chromate S. typhimurium TA97, TA1538, NS De Flora et al., 1984
+
TA98, TA100
Reverse mutations Ammonium chromate S. typhimurium TA1537 ± NS De Flora et al., 1984
Reverse mutations Ammonium chromate S. typhimurium TA1535 – – De Flora et al., 1984
Reverse mutations Calcium chromate S. typhimurium TA97, TA1538, NS De Flora et al., 1984
+
TA98, TA100
Reverse mutations Calcium chromate S. typhimurium TA1537 ± NS De Flora et al., 1984
Reverse mutations Calcium chromate S. typhimurium TA1535 – – De Flora et al., 1984
Reverse mutations Calcium chromate S. typhimurium TA98 – ± Dunkel et al., 1984
Reverse mutations Calcium chromate S. typhimurium TA100, TA1535, – Dunkel et al., 1984
–
TA1537, TA1538
Reverse mutations Calcium chromate E. coli WP2 uvrA – ± Dunkel et al., 1984
Reverse mutations Chromic acid S. typhimurium TA102, TA2638 + ND Watanabe et al., 1998
Reverse mutations Chromic acid E. coli, WP2/pKM101, WP2 ND Watanabe et al., 1998
+
uvrA/pKM101
Reverse mutations Chromium trioxide S. typhimurium TA97, TA1538, NS De Flora et al., 1984
+
TA98, TA100
Reverse mutations Chromium trioxide S. typhimurium TA1537 ± NS De Flora et al., 1984
Reverse mutations Chromium trioxide S. typhimurium TA1535 – – De Flora et al., 1984
Reverse mutations Potassium chromate S. typhimurium TA102 + ND Marzin and Phi, 1985
Reverse mutations Potassium chromate S. typhimurium TA97, TA1538, NS De Flora et al., 1984
+
TA98, TA100
Reverse mutations Potassium chromate S. typhimurium TA1537 ± NS De Flora et al., 1984
Reverse mutations Potassium chromate S. typhimurium TA1535 – – De Flora et al., 1984
Reverse mutations Potassium chromate E. coli Hs30R + ND Nakamuro et al., 1978
Reverse mutations Potassium chromate E. coli WP2 hcr- try-, B/rWP2 + (WP2 hcr) ND Kanematsu et al., 1980
Reverse mutations Potassium chromate E. coli WP2(try-) + ND Venitt and Levy, 1974
Reverse mutations Potassium chromate E. coli WP2uvrA, CM571 + ND Seo and Lee, 1993

Results
Without
Reverse mutations Potassium dichromate S. typhimurium TA97, TA98, + Zeiger et al., 1992
+
TA100, TA1535, TA1537
Reverse mutations Potassium dichromate S. typhimurium TA102 + ND Marzin and Phi, 1985
Reverse mutations Potassium dichromate S. typhimurium TA100 + + Venier et al., 1982
Reverse mutations Potassium dichromate E. coli WP2 hcr- try-, B/rWP2 + (WP2 hcr) ND Kanematsu et al., 1980
Reverse mutations Potassium dichromate E. coli Hs30R + ND Nakamuro et al., 1978
Reverse mutations Potassium dichromate E. coli WP2, WP2uvrA, CM571 + ND Nishioka, 1975
Reverse mutations Potassium dichromate E. coli WP2uvrA, CM571 + ND Seo and Lee, 1993
Reverse mutations Potassium dichromate S. cerevisiae D7 + ND Singh, 1983
Reverse mutations Potassium dichromate S. typhimurium TA98 ± – Venier et al., 1982
Reverse mutations Potassium dichromate S. typhimurium TA1538 – – Venier et al., 1982
Reverse mutations Sodium chromate E. coli WP2(try-) + ND Venitt and Levy, 1974
Reverse mutations Sodium dichromate S. typhimurium TA102, TA2638 + ND Watanabe et al., 1998
Reverse mutations Sodium dichromate S. typhimurium TA102 + Bennicelli et al., 1983
+
Reverse mutations Sodium dichromate S. typhimurium TA100 + – De Flora, 1978

Reverse mutations Sodium dichromate S. typhimurium TA97 + NS De Flora et al., 1984
Reverse mutations Sodium dichromate S. typhimurium TA1537, TA1538, NS De Flora et al., 1984
±
TA98, TA100
Reverse mutations Sodium dichromate S. typhimurium TA1535 – – De Flora et al., 1984
Reverse mutations Sodium dichromate E. coli, WP2/pKM101, WP2 ND Watanabe et al., 1998
+
uvrA/pKM101
Reverse mutations Sodium dichromate dihydrate S. typhimurium TA102, TA2638a + – Ryden et al., 2000
Reverse mutations Sodium dichromate dihydrate S. typhimurium TA100, TA98 + + NTP, 2007
Reverse mutations Sodium dichromate dihydrate E. coli, WP2 uvrA/pKM101 + + NTP, 2007
Induction of SOS response Chromic acid E. coli AB1157, GC2375, ND Llagostera et al., 1986
+
UA4202, PQ30
Induction of SOS response Potassium chromate E. coli PQ37, PQ35 + – Olivier and Marzin, 1987

Results
Without
Induction of SOS response Potassium chromate E. coli AB1157, GC2375, ND Llagostera et al., 1986
+
UA4202, PQ30
Induction of SOS response Potassium dichromate E. coli AB1157, GC2375, ND Llagostera et al., 1986
+
UA4202, PQ30
Induction of SOS response potassium dichromate E. coli PQ37, PQ35 + – Olivier and Marzin, 1987
Mutations Ammonium chromate S. typhimurium TA1978 (rec+), + ND Gentile et al., 1981
TA1538 (rec - )
Mutations Ammonium chromate B. subtilis + ND Gentile et al., 1981
Mutations Calcium chromate S. typhimurium TA97, TA98, – Brams et al., 1987
–
TA100
Mutations Chromic acid S. typhimurium TA1978 (rec+), + ND Gentile et al., 1981
TA1538 (rec - )
Mutations Chromic acid B. subtilis + ND Gentile et al., 1981
Mutations Potassium chromate S. typhimurium TA98, TA100, ND Arlauskas et al., 1985
+
TA1537
Mutations Potassium chromate S. typhimurium TA100 + ND Arlauskas et al., 1985
Mutations Potassium chromate E. coli WP2 uvrA pKm 101 + ND Arlauskas et al., 1985
Mutations Potassium chromate B. subtilis H17 + ND Nishioka, 1975
Mutations Potassium chromate S. typhimurium TA1535, TA1538 – ND Arlauskas et al., 1985
Mutations Potassium dichromate S. typhimurium TA 1535 pSK1002 + + Yamamoto et al., 2002
Mutations Potassium dichromate S. typhimurium TA100, TA1025, ND Le Curieux et al., 1993
+
TA98
Mutations Potassium dichromate S. typhimurium TA1978 (rec+), + ND Gentile et al., 1981
TA1538 (rec-)
Mutations Potassium dichromate E. coli WP2uvrA + ND Venier et al., 1987
Mutations Potassium dichromate B. subtilis + ND Gentile et al., 1981
Mutations Sodium dichromate B. subtilis + ND Gentile et al., 1981
Mutations Potassium dichromate B. subtilis NIG45, NIG17 + ND Matsui, 1980
Mutations Potassium dichromate B. subtilis H17 + ND Nishioka, 1975

Results
Without
Mutations Sodium dichromate S. typhimurium TA1978 (rec+), + ND Gentile et al., 1981
TA1538 (rec - )
Frame shift mutations Calcium chromate S. typhimurium TA98, TA1537 + ND Haworth et al., 1983
Frame shift mutation Potassium chromate S. typhimurium TA1537 + ND La Velle, 1986
Frame shift mutation Potassium chromate E. coli 343/358, /415, /435, /477 + ND La Velle, 1986
Frame shift mutations Potassium dichromate S. typhimurium TA97a, TA98 + + Tagliari et al., 2004
Frame shift mutations Potassium dichromate S. typhimurium TA100, TA1537, – ND Kanematsu et al., 1980
TA1538
Frame shift mutations Sodium dichromate S. typhimurium TA97, TA1978 + ND Bennicelli et al., 1983
Frame shift mutations Sodium dichromate S. typhimurium TA1537, TA1538 – ND Bennicelli et al., 1983
Frame shift mutations, base Calcium chromate S. typhimurium TA1537, TA98, + + Petrilli and De Flora, 1977
pair substitutions TA100
Frame shift mutations, base Calcium chromate S. typhimurium TA1535 – – Petrilli and De Flora, 1977
pair substitutions
Frame shift mutations, base Chromic acid S. typhimurium TA1537, TA98, + + Petrilli and De Flora, 1977
Frame shift mutations, base Chromic acid S. typhimurium TA1535 – – Petrilli and De Flora, 1977
pair substitutions
Frame shift mutations, base Potassium chromate S. typhimurium TA1537, TA98, + + Petrilli and De Flora, 1977
Frame shift mutations, base Potassium chromate S. typhimurium TA1535 – – Petrilli and De Flora, 1977
pair substitutions
Frame shift mutations, base Potassium dichromate S. typhimurium TA98 TA100, + + Bianchi et al., 1983
substitutions TA1535, TA1538
Frame shift mutations, base Sodium dichromate S. typhimurium TA1537, TA98, + + Petrilli and De Flora, 1977
Frame shift mutations, base Sodium dichromate S. typhimurium TA1535 – – Petrilli and De Flora, 1977
pair substitutions
Base pair substitutions Ammonium chromate S. typhimurium TA1537, TA1538, NS De Flora et al., 1984; De Flora,
±
TA98, TA100 1981
Base pair substitutions Ammonium chromate S. typhimurium TA1535 – – De Flora, 1981

Results
Without
Base pair substitutions Calcium chromate S. typhimurium TA100, TA1535 + ND Haworth et al., 1983
Base pair substitutions Calcium chromate S. typhimurium TA1537, TA1538, NS De Flora et al., 1984; De Flora,
±
TA98, TA100 1981
Base pair substitutions Calcium chromate S. typhimurium TA1535 – – De Flora, 1981
Base pair substitutions Chromic acid S. typhimurium TA1537, TA1538, ± NS De Flora et al., 1984; De Flora,
TA98, TA100 1981
Base pair substitutions Chromic acid S. typhimurium TA1537, TA1538, ± NS De Flora et al., 1984; De Flora,
TA98, TA100 1981
Base pair substitutions Potassium chromate S. typhimurium TA1537, TA1538, ± NS De Flora et al., 1984; De Flora,
TA98, TA100 1981
Base pair substitutions Potassium chromate S. typhimurium TA1535 – – De Flora, 1981
Base pair substitutions Potassium dichromate S. typhimurium TA100, TA102 + + Tagliari et al., 2004
Base pair substitutions Potassium dichromate S. typhimurium TA100, TA1535 – ND Kanematsu et al., 1980
Base pair substitutions Sodium dichromate S. typhimurium TA100, TA102, + ND Bennicelli et al., 1983
TA92
Base pair substitutions Sodium dichromate S. typhimurium TA1537, TA1538, ± NS De Flora et al., 1984; De Flora,
TA98, TA100 1981
Base pair substitutions Sodium dichromate S. typhimurium TA1535 – – De Flora, 1981
Base pair substitutions Sodium dichromate S. typhimurium TA1535 – ND Bennicelli et al., 1983
Reverse mutation, Potassium dichromate S. cerevisiae D7 + ND Kharab and Singh, 1985
induction of gene
conversion
Forward mutation Potassium dichromate Schizosaccharomyces pombe ± ND Bonatti et al., 1976
972, h-
Mitotic cross-over Chromic acid S. cerevisiae D7 + ND Fukunaga et al., 1982
Mitotic gene conversions Chromic acid S. cerevisiae D7 + ND Singh, 1983; Fukunaga et al.,
1982
Mitotic gene conversion, Sodium chromate S. cerevisiae D7 + ND Bronzetti and Galli, 1989
point reverse mutation

Results
Without
Mitotic gene conversion at Chromic acid S. cerevisiae D7 + ND Vashishat and Vasudeva, 1987
trp5 locus, reverse mutation
of ilvl-92 allele
Mitotic gene conversion at Potassium dichromate S. cerevisiae D7 + ND Vashishat and Vasudeva, 1987
trp5 locus, reverse mutation
of ilvl-92 allele
Induction of disomic and Potassium dichromate S. cerevisiae D1S13 + ND Sora et al., 1986
diploid spores
umu gene expression Potassium dichromate S. typhimurium TA1535 ± – Nakamura et al., 1987
DNA damage Potassium dichromate E. coli PQ37 + ND Le Curieux et al., 1993
DNA-protein crosslinks Potassium chromate E. coli DNA – ND Fornace et al., 1981
DNA polymerase arrest Sodium dichromate PSV2neo-based plasmid DNA – + Bridgewater et al., 1998, 1994
+ = positive; ± = equivocal or weakly positive; – = negative; ND = no data; NS = not specified

Table 4-22. In vitro genotoxicity studies of hexavalent chromium in mammalian cells
Results
Without With
Endpoint Chemical form Test system activation activation Reference
DNA damage Potassium Human lymphocytes + ND Blasiak and Kowalik, 2000
dichromate
DNA damage Potassium Human gastric mucosa + ND Trzeciak et al., 2000
dichromate
DNA damage Potassium Human peripheral blood lymphocytes + ND Trzeciak et al., 2000
dichromate
DNA damage Potassium Human lymphocytes, human lymphoblastoid TK-6 + ND Cemeli et al., 2003
dichromate cells
DNA damage Sodium Human gastric mucosa cells, rat gastric mucosa cells + ND Pool-Zobel et al., 1994
dichromate
DNA adducts, [32P] Potassium Calf thymus DNA – – Adams et al., 1996
postlabeling chromate (+1 mM
H2O2)
DNA fragmentation Potassium Human bronchial epithelial cells + ND Fornace et al., 1981
chromate
DNA fragmentation Potassium Human embryonic lung fibroblasts (IMR-90) + ND Fornace et al., 1981
chromate
DNA fragmentation Potassium Mouse L1210 leukemia cells + ND Fornace et al., 1981
chromate
DNA fragmentation sodium Chinese hamster ovary cells + ND Blankenship et al., 1997
chromate
DNA strand breaks Potassium Vero kidney fibroblasts, Pam 212 keratinocytes + ND Flores and Perez, 1999
dichromate
DNA strand breaks Sodium Rat primary lymphocytes + ND Gealy et al., 2007
dichromate
DNA strand breaks Sodium Rat hepatocytes + ND Gao et al., 1993
dichromate
DNA strand breaks Potassium Human lymphocytes + ND Depault et al., 2006
chromate

Results
Without With
DNA strand breaks Potassium Human fibroblast + ND Fornace, 1982
chromate
DNA strand breaks Potassium Bacteriophage λ DNA + + Adams et al., 1996
chromate (+1mM
H2O2)
DNA strand breaks Sodium Rat primary lymphocytes + ND Elia et al., 1994
dichromate
DNA strand breaks Potassium Human lymphocytes, human gastric mucosa cells + ND Blasiak et al., 1999
dichromate
DNA-DNA crosslinks Sodium Human lung fibroblasts + ND Xu et al., 1996
chromate
DNA-protein crosslinks Potassium Human embryonic lung fibroblasts (IMR-90) + ND Fornace et al., 1981
chromate
DNA-protein crosslinks Potassium Human fibroblast + ND Fornace, 1982
chromate
DNA-protein crosslinks Potassium Chinese hamster cells (V79-UL) + ND Merk et al., 2000
chromate
DNA-protein crosslinks Potassium Mouse L1210 leukemia cells + ND Fornace et al., 1981
chromate
DNA-protein crosslinks Sodium Human HL-60 cells + ND Capellmann et al., 1995
chromate
Induced DNA methylation Potassium Chinese hamster V79 cells (hpr-1gpt+ transgenic cell + (T) ND Klein et al., 2002
chromate line G12)
Unscheduled DNA synthesis Sodium Rat hepatocytes + (T) ND Gao et al., 1993
dichromate
DNA synthesis inhibition Potassium HeLa S3 cells + ND Heil and Reifferscheid, 1992
chromate
DNA synthesis inhibition Potassium Mouse L cells + ND Nishio and Uyeki, 1985
dichromate

Results
Without With
DNA polymerase arrest Sodium Human lung fibroblasts + ND Xu et al., 1996
chromate
Mutations at the HGPRT Potassium Chinese hamster ovary cells (AT3-2) + ND Paschin et al., 1983
locus dichromate
Mutations at the HGPRT Potassium Chinese hamster cells (V79) + ND Paschin et al., 1983
locus dichromate
Forward mutation Calcium Mouse lymphoma cells (L5178Y tk+/tk-) + + McGregor et al., 1987
chromate
Forward mutation Calcium Mouse lymphoma cells (L5178Y tk+) + + Mitchell et al., 1988
chromate
Forward mutation Calcium Mouse lymphoma cells (L5178Y tk+) + + Myhr and Caspary, 1988
chromate
Forward mutation Calcium Mouse lymphoma cells (L5178Y tk+) + + Oberly et al., 1982
chromate
Morphological Calcium Syrian hamster embryo cells + ND Elias et al., 1991
transformation chromate
Morphological Sodium Syrian hamster cells + ND DiPaolo and Casto, 1979
transformation chromate
dihydrate
Cell transformation Calcium Balb/3T3, Syrian hamster embryo, R-MuLV-RE cells + ND Dunkel et al., 1981
chromate
Transformations Potassium Rat liver epithelial cells + ND Briggs and Briggs, 1988
chromate
Chromosomal damage Calcium Chinese hamster ovary cells + ND Levis and Majone, 1979
chromate
Chromosomal damage Chromic acid Chinese hamster ovary cells + ND Levis and Majone, 1979
Chromosomal damage Potassium Chinese hamster ovary cells + ND Levis and Majone, 1979
chromate
Chromosomal damage Potassium Chinese hamster ovary cells + ND Seoane and Dulout, 1999
dichromate

Results
Without With
Chromosomal damage Potassium Chinese hamster ovary cells + ND Levis and Majone, 1979
dichromate
Chromosomal damage Potassium Chinese hamster ovary cells + ND Majone and Levis, 1979
dichromate
Chromosomal damage Sodium Chinese hamster ovary cells + ND Levis and Majone, 1979
chromate
Chromosomal damage Sodium Chinese hamster ovary cells + ND Levis and Majone, 1979
dichromate
Chromosomal damage Sodium Chinese hamster ovary cells + ND Majone and Levis, 1979
dichromate
Chromosome aberrations Calcium Chinese hamster lung DON cells + ND Koshi and Iwasaki, 1983; Koshi,
chromate 1979
Chromosome aberrations Calcium Chinese hamster ovary cells (C3H10T1/2) + ND Sen et al., 1987
chromate
Chromosome aberrations Chromic acid BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
Chromosome aberrations Chromic acid Mouse mammary FM3A carcinoma cells + ND Umeda and Nishmura, 1979
Chromosome aberrations Chromic acid Chinese hamster lung DON cells + ND Koshi and Iwasaki, 1983; Koshi,
1979
Chromosome aberrations Potassium Human fibroblasts + ND MacRae et al., 1979
chromate
Chromosome aberrations Potassium Chinese hamster lung DON cells + ND Koshi and Iwasaki, 1983; Koshi,
chromate 1979
Chromosome aberrations Potassium Chinese hamster ovary cells + ND MacRae et al., 1979
chromate
Chromosome aberrations Potassium BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
chromate
Chromosome aberrations Potassium Mouse mammary FM3A carcinoma cells + ND Umeda and Nishmura, 1979
chromate
Chromosome aberrations Potassium Human fibroblasts + ND MacRae et al., 1979
dichromate

Results
Without With
Chromosome aberrations Potassium Chinese hamster ovary cells + ND MacRae et al., 1979
dichromate
Chromosome aberrations Potassium BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
dichromate
Chromosome aberrations Potassium Mouse mammary FM3A carcinoma cells + ND Umeda and Nishmura, 1979
dichromate
Chromosome aberrations Sodium Human primary bronchial fibroblasts + ND Wise et al., 2004, 2002
chromate
Chromosome aberrations Sodium Human bronchial fibroblasts (WTHBF-6 cells) + ND Holmes et al., 2006
chromate
Chromosome aberrations Sodium Human bronchial epithelial cells (BEP2D cells) + ND Wise et al., 2006a
chromate
Chromosome aberrations Sodium Chinese hamster ovary cells + ND Blankenship et al., 1997
chromate
Chromosome aberrations Sodium Chinese hamster ovary cells (AA8 (parental), EM9 + ND Grlickova-Duzevik, 2006
chromate (XRCC1 mutant), and H9T3
Chromosome aberrations Sodium BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
dichromate
Chromosome and chromatid Potassium Human lymphocytes + ND Imreh and Radulescu, 1982
aberrations dichromate
Sister chromatid exchanges Calcium Human lymphocytes + ND Gomez-Arroyo et al., 1981
chromate
Sister chromatid exchanges Calcium Chinese hamster lung DON cells + ND Koshi and Iwasaki, 1983; Koshi,
chromate 1979
Sister chromatid exchanges Calcium Chinese hamster ovary cells + ND Levis and Majone, 1979
chromate
Sister chromatid exchanges Chromic acid Chinese hamster ovary cells + ND Levis and Majone, 1979
Sister chromatid exchanges Chromic acid Chinese hamster cells DON + ND Ohno et al., 1982
Sister chromatid exchanges Chromic acid Chinese hamster lung DON cells + ND Koshi and Iwasaki, 1983; Koshi,
1979

Results
Without With
Sister chromatid exchanges Chromic acid BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
Sister chromatid exchanges Potassium Human fibroblasts + ND MacRae et al., 1979
chromate
Sister chromatid exchanges Potassium Chinese hamster lung DON cells + ND Koshi and Iwasaki, 1983; Koshi,
chromate 1979
Sister chromatid exchanges Potassium Chinese hamster ovary cells + ND Levis and Majone, 1979
chromate
Sister chromatid exchanges Potassium Chinese hamster ovary cells + ND MacRae et al., 1979
chromate
Sister chromatid exchanges Potassium Chinese hamster cells DON + ND Ohno et al., 1982
chromate
Sister chromatid exchanges Potassium BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
chromate
Sister chromatid exchanges Potassium Human lymphocytes + ND Gomez-Arroyo et al., 1981
dichromate
Sister chromatid exchanges Potassium Human lymphocytes + ND Imreh and Radulescu, 1982
dichromate
Sister chromatid exchanges Potassium Human fibroblasts + ND MacRae et al., 1979
dichromate
dichromate
dichromate
Sister chromatid exchanges Potassium Chinese hamster ovary cells + ND Majone and Levis, 1979
dichromate
Sister chromatid exchanges Potassium Chinese hamster ovary cells + ND MacRae et al., 1979
dichromate
Sister chromatid exchanges Potassium Chinese hamster cells DON + ND Ohno et al., 1982
dichromate

Results
Without With
Sister chromatid exchanges Potassium BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
dichromate
Sister chromatid exchanges Potassium Mouse blastocysts + ND Iijima et al., 1983
dichromate
Sister chromatid exchanges Sodium Chinese hamster ovary cells + ND Levis and Majone, 1979
chromate
Sister chromatid exchanges Sodium BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
chromate
Sister chromatid exchanges Sodium Chinese hamster ovary cells + ND Levis and Majone, 1979
dichromate
Sister chromatid exchanges Sodium Chinese hamster ovary cells + ND Majone and Levis, 1979
dichromate
Sister chromatid exchanges Sodium BHK and Chinese hamster ovary cells + ND Bianchi et al., 1980
dichromate
Disruption of mitosis Sodium Human bronchial fibroblasts (WTHBF-6 cells) + ND Wise et al., 2006b
chromate
+ = positive; ± = equivocal or weakly positive; – = negative; (T) = toxicity; ND = no data

In in vivo test systems (see Table 4-23), hexavalent chromium compounds have tested
positive for mutations in Drosophila melanogaster and for DNA damage (DNA-protein
crosslinks, DNA strand breaks), mutations (in mice exposed in utero, in mouse germ cells, and in
transgenic mice), chromosomal damage (sister chromatid exchanges, chromosomal aberrations,
and micronuclei), and DNA synthesis inhibition in rats and mice. The in vivo studies in
laboratory mammals have evaluated the mutagenic activity of hexavalent chromium following
exposure by the oral, parenteral, inhalation, and intratracheal routes.

Table 4-23. In vivo genotoxicity studies of hexavalent chromium in rats and mice
Endpoint Test system Test conditions Results Dosea Comments Reference

Oral exposures (drinking water): mice
DNA deletions Female C57BL/ Potassium dichromate was administered in drinking water at + 62.5 mg Dose-response; no signs of Kirpnick-Sobol
6Jpun/pun mouse concentrations of 0, 62.5, or 125 mg/L at 10.5 to 20.5 days Cr(VI)/L toxicity observed. et al., 2006
offspring postcoitum (average dose of 12.5 or 25 mg/kg-day). 20-day-
old offspring were harvested to visualize eyespots
corresponding to DNA deletions in their retinal pigment
epithelium (RPE).
Micronuclei Pregnant Swiss Potassium dichromate was administered in drinking water at – 10 mg No effect on PCE/NCE De Flora et al.,
albino mouse bone concentrations of 0, 5, or 10 mg hexavalent chromium/L Cr(VI)/L ratio (no cytotoxicity). 2006
marrow, dams; throughout the duration of pregnancy. Mice were sacrificed
fetal liver and on d 18 of pregnancy and bone marrow cells were collected
peripheral blood from dams and liver cells were collected from fetuses.
cells Sodium dichromate dihydrate was administered in drinking – 10 mg
water at concentrations of 0, 5, or 10 mg hexavalent Cr(VI)/L
chromium/L throughout the duration of pregnancy. Mice were
sacrificed on d 18 of pregnancy and bone marrow cells were
collected. Liver and peripheral blood samples were collected
from the fetuses.
BDF1 male mouse Potassium dichromate was administered in drinking water at 0, – 20 mg
bone marrow and 10, or 20 mg hexavalent chromium/L for 20 d. Cr(VI)/L
peripheral blood
cells
BDF1 mouse (male Sodium dichromate dihydrate was administered in drinking – 500 mg
and female) bone water at 0, 5, 50, and 500 mg hexavalent chromium/L for 210 Cr(VI)/L
marrow or d. Peripheral blood cells were collected on d 0, 14, 28, 56,
peripheral blood and 147; bone marrow cells were collected on d 210.
cells
Micronuclei Swiss-Webster Potassium dichromate was administered at concentrations of – 20 mg No effect on %PCEs. Mirsalis et al.,
mouse bone 0, 1, 5, or 20 mg hexavalent chromium/L in drinking water. Cr(VI)/L 1996
marrow cells One set of mice was allowed access to drinking water ad
libitum, for 48 hours, while a second group was administered
two bolus doses (20 mL/kg) of the same concentrations at 24
and 48 hours before sacrifice.


Micronuclei B6C3F1 (5/group), Sodium dichromate dihydrate was administered in drinking ±b 87.2 mg No effect on PCE/NCE NTP, 2007
BALB/c (5/group), water for 3 mo at concentrations of 0, 62.5, 125, or 250 mg/L Cr(VI)/L ratio; no clinical signs of
and am3-C57BL/6 (0, 21.8, 43.6, or 87.2 mg hexavalent chromium/L). NTP (B6C3F1) toxicity observed.
(10/group) male estimated average daily doses at 0, 2.8, 5.2, or 8.7 mg
mouse peripheral hexavalent chromium/kg. – 87.2 mg
red blood cells Cr(VI)/L
(BALB/c)
+ 43.6 mg
Cr(VI)/L
(am3-
C57BL/6)
B6C3F1 mouse Sodium dichromate dihydrate was administered in drinking – 349 mg
(10/sex/group) water for 3 mo at concentrations 0, 62.5, 125, 250, 500, or Cr(VI)/L
peripheral red 1,000 mg/L (0, 21.8, 43.6, 87.2, 174.5, or 349 mg hexavalent
blood cells chromium/L). NTP estimated daily doses at 0, 3.1, 5.2, 9.1,
15.7, or 27.9 mg hexavalent chromium/kg.
Oxidative DNA Female SKH-1 Sodium dichromate dihydrate was administered in drinking – 20 mg No measure of cytotoxicity; De Flora et al.,
damage, DNA hairless mouse water at concentrations of 0, 5, and 20 mg hexavalent Cr(VI)/L no weight changes in mice. 2008
protein forestomach, chromium/L for 9 mo. Using reference values for body
crosslinks glandular stomach, weight (0.0353 kg) and daily drinking water intake (0.0085
and duodenum L/d) for female B6C3F1 mice (U.S. EPA, 1988), doses of 1.20
cells and 4.82 mg hexavalent chromium/kg-d for the 5 and 20 mg
hexavalent chromium/L groups, respectively, were estimated.
DNA-protein crosslinks and oxidative DNA damage (8-oxo-
2’deoxyguanosine) were measured in forestomach, glandular
stomach, and duodenum cells.
Unscheduled Fischer 344 rat Potassium dichromate was administered at concentrations of – 20 mg No measure of cytotoxicity. Mirsalis et al.,
DNA synthesis hepatocytes 0, 1, 5, or 20 mg hexavalent chromium/L in drinking water ad Cr(VI)/L RDS not determined. 1996
libitum for 48 hours, while a second group was administered
single gavage doses (20 mL/kg) at the same concentrations.
Hepatocytes were collected from the rat livers and analyzed in
the in vivo-in vitro hepatocyte DNA repair assay.
Oral exposures (drinking water): rats
DNA-protein Male Fischer 344 Potassium chromate was administered in drinking water for 3 + 100 mg No cytotoxicity detected. Coogan et al.,
crosslinks rat liver and and 6 wks at 100 and 200 ppm hexavalent chromium. Liver Cr(VI)/L 1991
splenic and splenic lymphocytes were examined for DNA-protein
lymphocytes crosslinks; crosslinks were detected in liver, not lymphocytes.


Oral exposures (gavage): mice
DNA damage, Swiss albino Potassium dichromate was administered by single gavage + 0.21 mg Dose-response from 0.59- Devi et al.,
comet assay mouse leukocytes doses of 0, 0.59, 1.19, 2.38, 4.75, 9.5, 19, 38, or 76 mg/kg (0, Cr(VI)/kg 9.5 mg/kg. Peak response 2001
0.21, 0.42, 0.84, 1.68, 3.37, 6.7, 13.5, or 26.9 mg hexavalent at 48 h. No cytotoxicity
chromium/kg). Samples of whole blood were collected at 24, detected (trypan blue).
48, 72, and 96 hours, and 1 and 2 wks post-treatment for
alkaline SCGE comet assay analysis of leukocytes.
DNA damage, Swiss albino Potassium dichromate was administered by gavage at doses of + (T) 8.8 mg Dose-response; apoptosis Wang et al.,
comet assay mouse peripheral 0, 25, 50, and 100 mg/kg for 1 day or daily for 5 consecutive Cr(VI)/kg detected only in liver, not 2006
lymphocytes days (0, 8.8, 17.7, and 35.4 mg hexavalent chromium/kg). in kidney. No lipid
Statistically significant: DNA damage in lymphocytes, and peroxidation.
ROS, apoptosis, and suppression of catalase and SOD in liver
and not kidney, at 1 and 5 days. No suppression of MDA in
liver.
DNA damage, ddY mouse Potassium chromate was administered in single gavage doses + 85.7 mg One dose group. Effects Sekihashi et al.,
comet assay stomach, colon, of 0 or 320 mg/kg (0 or 85.7 mg hexavalent chromium/kg). Cr(VI)/kg subsided at 24 h; 3 h peak 2001
liver, kidney, Cells were collected 3, 8, and 24 hours after treatment and for bladder only. No
bladder, lung, analyzed for DNA damage using the comet assay. DNA clinical or microscopic
brain, and bone damage was found in stomach, colon, liver, kidney, bladder, signs of cytotoxicity.
marrow lung, and brain, but not in bone marrow.
Micronuclei BDF1 male mouse Potassium dichromate was given as a single gavage dose of 0 – 50 mg No effect on PCE/NCE De Flora et al.,
bone marrow cells or 50 mg hexavalent chromium/kg. Cr(VI)/kg ratio (no cytotoxicity). 2006
Micronuclei Male MS/Ae and Potassium chromate was administered by single gavage doses – (T) 113.1 mg Negative up to acutely Shindo et al.,
CD-1 mouse bone of 0, 10, 20, 40, 80, 160, or 320 mg/kg (0, 3.5, 7.1, 14.1, 28.3, Cr(VI)/kg toxic doses. 1989
marrow cells 56.6, or 113.1 mg hexavalent chromium/kg).
Parenteral exposures: mice
Mutation Female C57BL/6J Potassium chromate was administered by intraperitoneal (i.p.) + 2.7 mg Decline in number of Knudsen, 1980
mouse offspring injection at a dose of 0, 10, or 20 mg/kg on days 8, 9, and 10 Cr(VI)/kg surviving offspring with
of pregnancy in a mammalian spot test (0, 2.7, or 5.4 mg dose.
hexavalent chromium/kg). The offsprings’ fur was checked
for colored spots from wk 2 through 5 after birth.
Mutation Male lacZ Potassium chromate was administered as an i.p. dose of 0 or + 14.1 mg One dose group. Itoh and
transgenic MutaTM 40 mg/kg once a day for 2 consecutive days (0 or 14.1 mg Cr(VI)/kg Cytotoxicity not reported. Shimada, 1997
mouse liver and hexavalent chromium/kg). Only one sampling time at day 7
bone marrow cells after second treatment. Statistically significant increase in
mutation frequency in liver and not bone marrow.


Mutation Male lacZ Potassium chromate was administered as an i.p. dose of 0 or + 14.1 mg One dose group. Itoh and
transgenic MutaTM 40 mg/kg once a day for 2 consecutive days (0 or 14.1 mg Cr(VI)/kg Cytotoxicity not reported. Shimada, 1998
mouse liver and hexavalent chromium/kg). Two sampling times (1 and 7 days)
bone marrow cells after second treatment. Statistically significant increase in
mutation frequency in bone marrow on d 1 (not d 7) and in
liver on d 7 (not d 1).
Dominant CBA × C57Bl/6J Potassium dichromate was administered as a single i.p. + 7.1 mg Statistically significant Paschin et al.,
lethality hybrid male mouse injection of 0, 0.5, 1.0, 2.0, 10, or 20 mg/kg (0, 0.18, 0.35, Cr(VI)/kg decrease in embryo 1982
0.70, 3.5, or 7.1 mg hexavalent chromium/kg) or with (acute i.p. survival. Too few doses
intraperitoneal injections of 0, 1.0, or 2.0 mg/kg potassium injection) used to detect dose-
dichromate daily for 21 days (0, 0.35, 0.70 mg hexavalent response.
chromium/kg). Each male was mated with two untreated 0.7 mg
females for 7 days, and then replaced by two more females + Cr(VI)/kg
every 7 days for 4 consecutive wks. Pregnant dams were (repeated
sacrificed 12–14 days after conception. The frequency of i.p.
dominant lethal mutations in male mice was determined based injection)
on the postimplantation loss.
DNA damage, ddY mouse Potassium chromate was administered as a single i.p. dose of + 32.1 mg One dose group. Effects Sekihashi et al.,
comet assay stomach, colon, 0 or 120 mg/kg (0 or 32.1 mg hexavalent chromium/kg). Cells Cr(VI)/kg subsided at 24 h; peak at 3 2001
liver, kidney, were collected 3, 8, and 24 hours after treatment and analyzed h for bladder, lung, and
bladder, lung, for DNA damage using the comet assay. DNA damage was brain. No clinical or
brain, and bone detected in stomach, colon, bladder, lung, and brain, but not in microscopic signs of
marrow liver, kidney, or bone marrow. cytotoxicity.
DNA damage, Male albino mouse Potassium dichromate was administered as a single i.p. dose of + 20 mg Same pattern as Cr(V) Ueno et al.,
comet assay liver, kidney, 0 or 20 mg hexavalent chromium/kg. Organs were removed Cr(VI)/kg complexes. Cytotoxicity 2001
spleen, lung, and and cells were collected for DNA strand break analysis by not reported. DNA damage
brain single-cell gel electrophoresis. DNA damage was detected in reduced with deferoxamine
liver and kidney (and not in spleen, lung, or brain) at 15 min
post-injection; damage back to control levels at 3 h.
Micronuclei CBA × C57Bl/6J Potassium dichromate was administered as a single i.p. + 0.35 mg Increased response with Paschin and
hybrid mouse bone injection of 0, 1, 5, or 10 mg/kg (0.35, 1.77, or 3.54 mg Cr(VI)/kg dose and time; peak at 48 h. Toropzev, 1982
marrow hexavalent chromium/kg). Bone marrow was sampled 24, 48, No measure of cytotoxicity.
and 72 hours after treatment for the micronucleus test.
Micronuclei Slc:ddY mouse Potassium chromate was administered by i.p. injection once a + 10.6 mg Statistically significant Itoh and
bone marrow cells day for 2 consecutive days at doses of 0, 30, 40, and 50 mg/kg Cr(VI)/kg dose-response. %PCEs Shimada, 1996
(0, 10.6, 14.1, or 17.7 mg hexavalent chromium/kg). decreased in two highest
doses only.


Micronuclei NMRI mouse bone Potassium chromate was administered by 2 i.p. injections with + 13 mg Statistically significant, Wild, 1978
marrow 24 hours between each injection at doses of 0, 12.12, 24.25, or Cr(VI)/kg dose-related increase.
48.5 mg/kg (0, 3.2, 6.49, or 13.0 mg hexavalent Cytotoxicity not reported.
chromium/kg).
Micronuclei MS/Ae and CD-1 Potassium chromate was administered by single i.p. doses of + 14.1 mg Dose-response observed. Shindo et al.,
male mouse bone 0, 10, 20, 40, or 80 mg/kg (0, 3.5, 7.1, 14.1, or 28.3 mg Cr(VI)/kg %PCEs only decreased at 1989
marrow cells hexavalent chromium/kg). highest dose.
Micronuclei LacZ transgenic Potassium chromate was administered by an i.p. dose of 0 or + 14.1 mg One dose group. Itoh and
MutaTM male 40 mg/kg once a day for 2 consecutive days (0 or 14.1 mg Cr(VI)/kg Cytotoxicity not reported. Shimada, 1997
mouse peripheral hexavalent chromium/kg).
red blood cells
Micronuclei MS and ddY Potassium chromate was administered by single i.p. doses of + 17.7 mg Dose-response observed. Hayashi et al.,
mouse bone 0, 12.5, 25, or 50 mg/kg (0, 4.4, 8.8, or 17.7 mg hexavalent Cr(VI)/kg Cytotoxicity not reported. 1982
marrow cells chromium/kg).
Micronuclei BALB/c mouse Potassium dichromate was administered as a single i.p. + (T) 20.8 mg One dose group; Wronska-Nofer
bone marrow injection at a dose of 0 or 400 μmol (20.8 mg hexavalent Cr(VI)/kg significantly decreased et al., 1999
chromium/kg). %PCEs.
Micronuclei Pregnant Swiss Potassium dichromate was administered as a single i.p. + 50 mg No effect on PCE/NCE De Flora et al.,
albino mouse: injection at 0 or 50 mg hexavalent chromium/kg on day 17 of Cr(VI)/kg ratio (no cytotoxicity). 2006
bone marrow, pregnancy. Mice were sacrificed on day 18 of pregnancy.
dams; fetal liver Liver and peripheral blood samples were collected from the
and peripheral fetuses and bone marrow from the dams.
blood cells Sodium dichromate dihydrate was administered as a single i.p. + 50 mg
injection at 0 or 50 mg/kg on day 17 of pregnancy. Mice were Cr(VI)/kg
sacrificed on day 18 of pregnancy and bone marrow cells were
collected. Liver and peripheral blood samples were collected
from the fetuses.
BDF1 male mouse Potassium dichromate was administered as single i.p. doses of + 50 mg
bone marrow cells of 0 or 50 mg hexavalent chromium/kg. Cr(VI)/kg
Suppressed Mouse tubular Potassium dichromate was given as a single i.p. injection at a + NS No measure of cytotoxicity. Amlacher and
nuclear DNA renal cells concentration of 15–30% of the LD50 (unspecified) in a Rudolph, 1981
synthesis thymidine incorporation inhibiting screening system; an
intraperitoneal injection of [3H]-thymidine was administered
15 hours later.


Parenteral exposures: rats
DNA damage, Sprague-Dawley Potassium dichromate was administered i.p. at doses of 2.5, 5, + 0.88 mg Abstract only. Patlolla and
comet assay rat leukocytes 7.5, and 10 mg/kg-d for 5 days (0, 0.88, 1.77, 2.65, or 3.54 mg Cr(VI)/kg-d Tchounwou,
hexavalent chromium/kg-d). Whole blood was sampled at 24, 2006
48, 72, and 96 hours after treatment for analysis of leukocytes.
Chromosomal Sprague-Dawley Potassium dichromate was administered i.p. at doses of 2.5, 5, + 0.88 mg Statistically significant Patlolla et al.,
aberrations rat bone marrow 7.5, and 10 mg/kg-d for 5 days (0, 0.88, 1.77, 2.65, or 3.54 mg Cr(VI)/kg-d CAs and MI with positive 2008
hexavalent chromium/kg-d). Bone marrow cells were dose-response; increases in
harvested at the end of the exposure period and measured for MN not observed.
increases in chromosomal aberrations (CAs), micronuclei
(MN), and mitotic indices (MI).
DNA-protein Sprague-Dawley Sodium dichromate was given as a single i.p. injection of + 7 mg No measure of cytotoxicity. Tsapakos et al.,
crosslinks male rat lung, 20 or 40 mg/kg (7 or 14 mg hexavalent chromium/kg). No Cr(VI)/kg 1983
liver, and kidney control group was used in this study. Nuclei from the right
nuclei renal cortex, the front hepatic lobe, and the whole lung were
used for analysis.
Intratracheal instillation and inhalation exposures: mice
Mutations C57BL/6 Big Blue Potassium dichromate was given as single doses of 0 or 6.75 + 6.75 mg No clinical evidence of Cheng et al.,
mouse lung, mg hexavalent chromium/kg and allowed 4 wks for gene Cr(VI)/kg toxicity at doses ≤ 6.75 2000
kidney, and liver expression. Isolated DNA samples from lung, liver, and mg/kg.
kidney tissues were used for LacI gene mutagenesis assay.
Mutations were detected in mouse lung and kidney tissue, but
not in liver tissue. Depletion of tissue GSH by pretreatment
with buthionine sulfoximine decreased the mutagenic
response, suggesting that reduced GSH plays a role in
producing reactive intermediates during intracellular reduction
of chromium (VI).
Intratracheal instillation and inhalation exposures: rats
DNA Sprague-Dawley Sodium dichromate was administered as intratracheal + 0.09 mg No measure of cytotoxicity. Izzotti et al.,
alterations rat lung and liver instillations at doses of 0 or 0.25 mg/kg for 3 consecutive days Cr(VI)/kg Positive results in lung and 1998
(0 or 0.09 mg hexavalent chromium/kg). After the last not liver.
treatment, lung and livers were removed to analyze for DNA
fragmentation, DNA-protein crosslinks, and adducts by [32P]
postlabeling. DNA-protein crosslinks, DNA fragmentation,
and DNA adducts were detected in lung, but not liver.


3
Chromosomal Sprague-Dawley Rats were exposed to chromium fumes (valence state not + 1.84 mg/m Abstract only. Koshi et al.,
aberrations, rat bone marrow specified) generated from a plasma flame sprayer and (1-wk 1987
sister chromatid cells and chromium metal powders at a concentration of 1.84 mg/m3 for inhalation
exchange peripheral 1 wk (5 hours/d, 5 d/wk) or 0.55 mg/m3 for 2 mo (5 hours/d, 5 exposure)
lymphocytes d/wk). Cytogenetic analysis was performed 20 hours, 3 d, 7 d,
and 1 mo after the last exposure. Chromosome aberrations and 0.55 mg/m3
sister chromatid exchanges were detected in rat peripheral (2-mo
lymphocytes but not in bone marrow cells. inhalation
exposure)
+ = positive; ± = equivocal or weakly positive; – = negative; NS = not specified; (T) = toxicity

a
Lowest effective dose for positive results, highest dose tested for negative results.
b
NTP determined this result to be equivocal due to a trend test p-value very nearly significant (p = 0.031; α level = 0.025) and a significant response (p = 0.0193) in the
highest dose group of 87.2 mg/L.

Table 4-24. In vivo genotoxicity studies of hexavalent chromium in D. melanogaster
Endpoint Chemical Test system Test conditions Results Dosea Reference

Gene mutation Calcium chromate D. melanogaster 24-Hr-old males were fed calcium chromate for 72 hours + 500 ppm Zimmering et
at doses of 0, 500, or 750 ppm. The males were (in diet) al., 1985
removed and mated.
Gene mutation Chromic acid D. melanogaster 24–48-Hr-old males were treated by intraperitoneal + 100 ppm Rodriguez-
injection with 0, 100, 200, 300, and 400 ppm potassium (intraperitoneal Arnaiz and
dichromate or 0, 100, 200, and 300 ppm chromium injection) Martinez, 1986
trioxide. The F2 generation of flies was scored for sex-
linked recessive lethal.
Gene mutation Chromium oxide D. melanogaster 2–3-D-old larvae were fed potassium chromate or + 1 mM Graf and
(wing somatic chromium(VI) oxide for 3 d at concentrations of 0, 1, or (in diet) Wurgler, 1996
mutation) 5 mM.
Gene mutation Chromium oxide D. melanogaster 2–3-D-old larvae were fed potassium chromate or – 5 mM Graf and
(white-ivory eye chromium(VI) oxide for 2 d at concentrations of 0, 1, or (in diet) Wurgler, 1996
spot test) 5 mM.
Gene mutation Potassium D. melanogaster Larvae were fed the test substance in wing spot test at + 0.1 mM Amrani et al.,
chromate concentrations of 0, 0.1, 0.5, 1.0, and 2.5 mM for the (in diet) 1999
duration of their development. Surviving
transheterozygous (mwh/flr3) and inversion
heterozygous (mwh/TM3) flies were used.
Gene mutation Potassium D. melanogaster 2–3-D-old larvae were fed potassium chromate or + 1 mM Graf and
(wing somatic chromate chromium(VI) oxide for 3 d at concentrations of 0, 1, or (in diet) Wurgler, 1996
mutation) 5 mM.
Gene mutation Potassium D. melanogaster 3-D-old larvae were fed potassium chromate for 6 hours + 0.5 mM (48 hours) Spano et al.,
chromate at concentrations ranging from 0 to 100 mM or 48 hours (in diet) 2001
at concentrations ranging from 0 to 5.0 mM. Marker-
heterozygous and balancer-heterozygous wings from 5 mM (6 hours)
adult flies were then examined in the wing somatic (in diet)
mutation and recombination test (SMART).
Gene mutation Potassium D. melanogaster 2–3-D-old larvae were fed potassium chromate or – 5 mM Graf and
(white-ivory eye chromate chromium(VI) oxide for 2 d at concentrations of 0, 1, or (in diet) Wurgler, 1996
spot test) 5 mM.
Gene mutation Potassium D. melanogaster 3-D-old transheterozygous larvae were fed potassium + 0.5 mM Kaya et al.,
dichromate dichromate at 0 or 0.5 mM and analyzed for multiple (in diet) 2002
wing hair and flare gene mutations in the Drosophila
wing SMART.

Table 4-24. In vivo genotoxicity studies of hexavalent chromium in D. melanogaster
Endpoint Chemical Test system Test conditions Results Dosea Reference

Gene mutation Potassium D. melanogaster Larvae were fed the test substance in wing spot test at + 0.1 mM Amrani et al.,
dichromate concentrations of 0, 0.1, 0.5, 1.0, and 2.5 mM for the (in diet) 1999
duration of their development. Surviving
transheterozygous (mwh/flr3) and inversion
heterozygous (mwh/TM3) flies were used.
Gene mutation Potassium D. melanogaster 24–48-Hr-old males were treated by intraperitoneal + 100 ppm Rodriguez-
dichromate injection with 0, 100, 200, 300, and 400 ppm potassium (intraperitoneal Arnaiz and
dichromate or 0, 100, 200, and 300 ppm chromium injection) Martinez, 1986
trioxide. The F2 generation of flies was scored for sex-
linked recessive lethal.
Gene mutation Sodium dichromate D. melanogaster Larvae were treated on filter papers soaked with sodium + 2.34 mM Rasmuson,
dichromate at doses of 1.17 and 2.34 mM for 6 hours 1985
and then transferred to vials with substrate. Adult males
were checked for wild-type pigmented spots in the eyes.
+ = positive; ± = equivocal or weakly positive; – = negative; NS = not specified; (T) = toxicity

a
Lowest effective dose for positive results, highest dose tested for negative results.

Hexavalent chromium-induced mutagenicity has been demonstrated following oral
exposure. Oral exposure studies evaluating the mutagenicity of hexavalent chromium in tissues
from the GI tract are of particular relevance in light of the results of the NTP (2008) cancer
bioassay showing neoplasms of the oral cavity in rats (at 5.9–7.0 mg hexavalent chromium/kg-
day) and of the small intestine in mice (at 2.4–3.1 mg hexavalent chromium/kg-day)
administered sodium dichromate dihydrate in drinking water for 2 years. In studies involving
gavage administration of hexavalent chromium in mice and rats in vivo, DNA damage has been
observed in several tissues, including stomach, colon, liver, lung, kidney, bladder, brain, and
peripheral leukocytes (Wang et al., 2006; Devi et al., 2001; Sekihashi et al., 2001; Coogan et al.,
1991).
In ddY mice, positive results were reported for DNA damage as measured by the comet
assay in the stomach and colon following gavage administration of a single high dose of
hexavalent chromium (85.7 mg hexavalent chromium/kg) (Sekihashi et al., 2001). This dose is
at least 12-fold greater than chronic dosages associated with oral and GI neoplasms in rats and
mice (NTP, 2008), although no concurrent cytotoxicity was found. Data on the potential for
DNA damage in cells of the GI tract at lower oral doses (e.g., those in the range of the NTP
[2008] bioassay) are not available.
Devi et al. (2001) observed DNA damage via the comet assay in mouse leukocytes
following an oral dose as low as 0.21 mg/kg, an effect that increased with dose up to 9.5 mg/kg
and did not cause a decrease in cell viability. Similarly, Wang et al. (2006) found a dose-
dependent increase in DNA damage in peripheral lymphocytes using the comet assay that was
found to persist for 5 days post-exposure and was accompanied by a significant increase in
reactive oxygen species and apoptosis in the liver. Sekihashi et al. (2001) found comet damage
in mouse stomach, colon, liver, kidney, bladder, lung, and brain following a single gavage dose
of 85.7 mg/kg. These effects were not accompanied by cytotoxicity, although it is unknown
whether a response to dose would have occurred.
Three drinking water exposure studies of hexavalent chromium in mice and rats have
yielded positive results for the induction of chromosomal damage. Coogan et al. (1991)
observed DNA-protein crosslinks in rat liver following 3- and 6-week exposures that were not
accompanied by cytotoxicity. In another drinking water study in a strain of mice containing a
mutation allowing for visual representation of chromosome deletions in the form of eye spots in
the offspring of exposed pregnant females, exposures to 62.5 mg/L of hexavalent chromium
from 10.5 to 20.5 days postcoitum resulted in a significant level of DNA deletions in 20-day old
offspring that increased with dose and was not accompanied by cytotoxicity (Kirpnick-Sobol et
al., 2006). Statistically significant increases in chromosomal damage (as indicated by
micronuclei formation) with a significant dose-response was observed in peripheral RBCs of one
strain of mice (am3-C57BL/6) exposed to ≥ 43.6 mg hexavalent chromium/kg-day as sodium
dichromate dihydrate in drinking water for 3 months, but not in BALB/c mice at daily doses up

to 87.2 mg hexavalent chromium/kg-day (NTP, 2007). The B6C3F1 strain of mice used in the 2-
year NTP bioassay (NTP, 2008) was also tested for micronucleus formation; the first test was
negative, but the second test showed a nearly significant positive trend for micronucleus
formation. These results were considered equivocal overall because the trend test p-value of
0.031 was close in value to the designated alpha level of 0.025 used in this study (compared to
the typical alpha level of 0.05).
Other studies have reported negative results in bone marrow or peripheral blood cells
following oral exposures (NTP, 2007; De Flora et al., 2006; Mirsalis et al., 1996; Shindo et al.,
1989). One study investigated tissues in mice identified by the NTP bioassay (2008) as having
significant hexavalent chromium-induced tumors and reported negative results for oxidative
DNA damage and DNA-protein crosslinks in cells of the forestomach, glandular stomach, and
duodenum of female SKH-1 mice administered drinking water containing 5 or 20 mg hexavalent
chromium/L (approximately equivalent to 1.20 and 4.82 mg hexavalent chromium/kg-day,
respectively) as sodium dichromate dihydrate for 9 months (De Flora et al., 2008). It is worth
noting the absence of positive findings in De Flora et al. (2008) given that the highest dose
evaluated in this study is slightly less than chronic dosages associated with neoplasms of the oral
cavity in rats (5.9–7.0 mg hexavalent chromium/kg-day), and slightly greater than those
associated with neoplasms of the small intestine in mice (2.4–3.1 mg hexavalent chromium/kg-
day) (NTP, 2008). No oral exposure studies on the potential clastogenic activity of hexavalent
chromium in rat tumor target tissue (oral mucosa) were identified. Although the NTP (2007) 3-
month drinking water study evaluated micronuclei formation in peripheral RBCs of mice (with
positive results in the am3-C57BL/6 strain and equivocal results in the B6C3F1 strain),
mutagenic effects of hexavalent chromium exposure in GI tissues were not evaluated in this
study.
Results of parenteral exposure studies are uniformly positive for hexavalent chromium-
induced mutagenicity. Following parenteral exposure, DNA damage has been observed in
numerous tissues, including peripheral lymphocytes, stomach, colon, liver, kidney, bladder, lung,
and brain (Patlolla and Tchounwou, 2006; Sekihashi et al., 2001; Ueno et al., 2001); mutations
have been observed in liver (Knudsen, 1980); and chromosomal damage (micronuclei) has been
observed in peripheral RBCs and bone marrow (De Flora et al., 2006; Itoh and Shimada, 1997;
Shindo et al., 1989; Hayashi et al., 1982; Wild, 1978).
Mutagenic activity of hexavalent chromium has also been demonstrated in lung cells of
animals following intratracheal exposure. DNA damage (DNA fragmentation, DNA-protein
crosslinks, and DNA adducts) was reported in lung cells of Sprague-Dawley rats administered
0.09 mg hexavalent chromium/kg by intratracheal instillation for 3 days (Izzotti et al., 1998) and
mutations were reported in lung cells of C57BL/6 mice administered a single intratracheal dose
of 7.65 mg hexavalent chromium/kg. Results of these studies are relevant to occupational
exposure studies showing increased respiratory tract cancers in hexavalent chromium workers

(see Section 4.4.1.2). No inhalation or intratracheal exposure studies on the potential clastogenic
activity of hexavalent chromium in respiratory tract cells were identified. Chromosomal damage
(chromosome aberrations and sister chromatid exchange) was observed in peripheral
lymphocytes, but not bone marrow, of Sprague-Dawley rats exposed to chromium fumes for
1 week (1.84 mg/m3) or 2 months (0.55 mg/m3) (Koshi et al., 1987).
4.5.1.2. Genotoxicity Studies in Humans

In addition to mutagenicity evaluations in experimental systems, several studies have
evaluated mutagenicity in humans occupationally exposed to hexavalent chromium;
experimental details and citations are summarized in Table 4-25. Data from available
mutagenicity studies in exposed workers are limited to assessments of tissues with easy
accessibility (e.g., circulating lymphocytes and buccal and nasal mucosal cells). Data on
mutagenicity in cancer target tissues (e.g., lung and GI tract) are not available. Available data
provide some evidence of hexavalent chromium-induced mutagenicity in occupationally exposed
humans, although results of studies in workers have yielded mixed results. In general,
associations between hexavalent chromium exposure and mutagenicity in workers are uncertain
because exposure levels were often not quantified or estimated, past exposure history was not
well characterized in all studies, small numbers of workers were evaluated, and/or workers were
potentially co-exposed to other compounds with mutagenic activity.

Table 4-25. In vivo genotoxicity studies in humans exposed to hexavalent chromium
Exposure type
Endpoint (chemical form) Cell type Test conditions Results Exposure levela Reference
DNA strand Occupational, Human Nineteen chromium plating workers in Italy (mean employment of + NS Gambelunghe
breaks chromium plating peripheral 6.3 yrs) and two groups of control subjects (18 hospital workers et al., 2003
(chromic acid) lymphocytes and 20 university personnel) gave pre- and postshift urine samples
and blood samples for analysis in the comet assay. Duration of
employment ranged from 4 mo to 14 yrs with a mean duration of
6.3 yrs. Mean chromium concentrations in urine were determined
to be 5.29 µg/g creatinine (preshift) and 7.31 µg/g creatinine
(postshift). Mean erythrocyte and lymphocyte concentrations in
the exposed workers were 4.94 µg/L and 50.3 µg/1012cells,
respectively. Air concentrations of chromium were not reported.
DNA strand Occupational, Human Urine and blood samples were taken from 10 exposed workers and – 0.001–0.055 mg Gao et al., 1994
breaks, production of peripheral 10 nonexposed workers at the end of a workweek at a bichromate hexavalent
hydroxylation of dichromate lymphocytes production plant in England. The mean duration of exposure was chromium/m3
deoxyguanosine (included exposure 15 yrs. Chromium concentrations in the factory ranged from (measured
to chromic acid, 0.001 to 0.055 mg hexavalent chromium/m3 (obtained from exposure range)
potassium personal and area samplers). Mean chromium concentrations in
dichromate and urine (5.97 µg/g creatinine), whole blood (5.5 µg/L), plasma
sodium (2.8 µg/L), and lymphocytes (1.01 µg/1010 cells) of exposed
dichromate) workers were significantly higher than in nonexposed workers.
DNA-protein Experimental oral Human Four adult volunteers ingested a single bolus dose of 5,000 µg – 71 µg hexavalent Kuykendall et
crosslinks exposure peripheral hexavalent chromium as potassium dichromate (approximately chromium/kg al., 1996
(potassium lymphocytes equivalent to 71 µg hexavalent chromium/kg, assuming a body
dichromate) weight of 70 kg). Blood samples were collected at 0, 60, 120,
180, and 240 mins after ingestion. Preingestion background
DNA-protein crosslink levels for each individual served as the
controls.
Chromosome Occupational, Human Blood from seven chromium electroplating workers at a Chinese + 8.1 µg Deng et al.,
aberrations, sister chromium peripheral electroplating facility (mean employment period of 12.8 yrs) and chromium/mm3b 1988
chromatid electroplating lymphocytes 10 control subjects were analyzed. Air samples from the
exchanges (chemical not electroplating room were collected, along with stool and hair
specified) samples to determine exposure. The mean chromium (total) air
concentration (by random air collection) was 8.1 µg/mm3, the
mean chromium concentration in stool samples was 8.5 µg/g stool,
and the mean chromium concentration in hair was 35.68 µg/g.
The valence of chromium that workers were exposed to was
unspecified.

Exposure type
Sister chromatid Occupational, Human whole Thirty-five chromium electroplating factory workers employed at + 5.99 mg Wu et al., 2001
exchanges chromium blood cells three electroplating plants in Tawain and 35 control subjects gave hexavalent
electroplating blood samples to analyze the frequency of sister chromatid chromium/m3
(chemical not exchange. Exposure duration ranged from 2 to 14 yrs with a mean
specified) of 6.5 yrs. Mean chromium exposure (determined by personal
monitoring samplers) was 5.99 mg hexavalent chromium/m3. The
mean urinary chromium concentration of the exposed workers was
3.67 µg/g creatinine.
Chromosomal Occupational, Human Thirty-eight male chromium plating factory workers in Italy were + NS Sarto et al.,
aberrations, sister chromium plating peripheral examined for urinary concentrations of chromium and 1982
chromatid (chromic acid) lymphocytes chromosomal aberrations and sister chromatid exchanges.
exchanges Chromium exposure levels were not reported. There were
35 unexposed control individuals.
Sister chromatid Occupational, Human The frequency of sister chromatid exchanges was determined in + NS Stella et al.,
exchanges chromium plating peripheral lymphocytes from 12 chromium plating workers in Italy and 1982
(chromic acid lymphocytes 10 control subjects. Exposure durations ranged from 0.5 to 18 yrs
fumes) (mean exposure duration was not reported). Hexavalent
chromium exposure levels and blood concentrations were not
reported.
Sister chromatid Occupational, Human Thirty-five chromium electroplating factory workers in Taiwan + NS Wu et al., 2000
exchanges chromium peripheral and 35 control subjects (matched for age and gender) gave blood
electroplating lymphocytes samples to determine sister chromatid exchange frequency. The
(chemical not mean duration of employment was 6.5 yrs. Exposure
specified) concentrations were not reported.
Chromosomal Occupational, Human Blood samples and buccal mucosal cells from 15 Bulgarian – Results reported Benova et al.,
aberrations, sister chromium plating peripheral chromium platers occupationally exposed were taken; exposure for combined 2002
chromatid (chemical not lymphocytes was estimated with personal air samplers and in urine samples. groups (0.0075 and
exchanges specified) and buccal Control subjects were matched with exposed individuals. 0.0249 mg
mucosal cells Duration of exposure ranged from 2 to >20 yrs; mean duration of chromium/m3)
exposure was not reported. Mean air concentration of total
chromium was 0.0075 mg chromium/m3 in the low-exposure
group and 0.0249 mg chromium/m3 in the high-exposure group
(number of workers in each exposure group was not reported).
Mean concentrations of chromium in urine were 18.63 µg/L (low)
and 104.22 µg/L (high)

Exposure type
Sister chromatid Occupational, Human Venous blood and urine sample were collected from 12 male – NS Nagaya et al.,
exchanges chromium plating peripheral chromium platers in Japan over a 5-yr period. No control subjects 1991
(chemical not lymphocytes were used in this study. Employment duration ranged from 6.6 to
specified) 25.1 yrs, with mean employment duration of 15.5 yrs. Exposure
concentrations were not reported. Urinary chromium
concentrations ranged from 1.2 to 57.0 µg/g with a mean urinary
chromium concentration of 17.9 µg/g creatinine. Sister chromatid
exchange frequency in lymphocytes was determined in blood-
urine paired samples.
Sister chromatid Occupational, Human Venous blood and urine sample were collected from 24 male – NS Nagaya, 1986
exchanges chromium plating peripheral chromium platers in Japan and 24 control subjects. Duration of
(chemical not lymphocytes employment ranged from 0.5 to 30.5 yrs with a mean employment
specified) of 11.6 yrs. Exposure concentrations were not reported. The
mean concentration of chromium in the urine was 13.1 µg/L.
Micronuclei Occupational, Human Forty electroplating workers in Bulgaria and 18 control subjects + 0.043 and Vaglenov et al.,
chromium peripheral gave blood samples to analyze for the frequency of micronuclei. 0.083 mg 1999
electroplating lymphocytes The workers were split into two groups based on levels of chromium/m3
(chemical not exposure. Mean air chromium (total) concentrations were 43 and
specified) 83 µg/m3 in the low- and high-exposure groups, respectively.
Duration of employment ranged from 4 to 25 yrs with mean
durations of 10.44 and 11.63 yrs in the low- and high-exposure
groups, respectively. Mean chromium concentrations in
erythrocytes and urine of the low-exposure group were 4.31 and
3.97 µg/L, respectively. The mean chromium concentrations in
erythrocytes and urine of the high-exposure group were 8.4 and
5.0 µg/L, respectively.
Micronuclei Occupational, Human Blood samples and buccal mucosal cells from 15 Bulgarian + Positive results Benova et al.,
chromium plating peripheral chromium platers occupationally exposed were taken. Exposure reported for 2002
(chemical not lymphocytes was estimated with personal air samplers and in urine samples. combined groups
specified) and buccal Control subjects were matched with exposed individuals. (0.0075 and 0.0249
mucosal cells Duration of exposure ranged from 2 to >20 yrs; mean duration of mg chromium/m3)
exposure was not reported. Mean air concentration of total
chromium was 0.0075 mg chromium/m3 in the low-exposure
group and 0.0249 mg chromium/m3 in the high-exposure group.
Mean concentrations of chromium in urine were 18.63 (low) and
104.22 µg/L (high).

Exposure type
Micronuclei Occupational, Human Sixteen exposed Italian electroplating factory workers and – NS Sarto et al.,
chromium plating buccal and 27 unexposed control subjects gave samples of exfoliated buccal 1990
(chromic acid) nasal cells and nasal swabs. Duration of exposure ranged from 0.5 to 23 yrs
with a mean duration of 8 yrs. Urine samples were collected at the
end of work days to determine chromium exposure. Urinary
chromium concentrations ranged from 2.5 to 88 µg/g creatinine;
the mean urinary chromium concentration was not reported.
Chromium levels in air were not determined.
a
All exposure levels associated with positive results, highest exposure level for negative results.
b
The exposure level of 8.1 µg chromium/mm3 is as reported by Deng et al. (1988); however, this appears to be a reporting error, as this concentration is equivalent to
8,100,000 mg chromium/m3.
+ = positive; – = negative; NS = not specified

In a comet assay in Italian chrome platers, positive results were reported for DNA strand
breaks in peripheral lymphocytes; although urine chromium concentrations were determined,
hexavalent chromium exposure levels were not reported (Gambelunghe et al., 2003). However,
no DNA damage was observed in peripheral lymphocytes in dichromate production workers
exposed to 0.001–0.055 mg hexavalent chromium/m3 (Gao et al., 1994) or in volunteers
ingesting single oral doses of 71 µg hexavalent chromium/kg (Kuykendall et al., 1996). In
chrome electroplaters, chromosome aberrations and sister chromatid exchanges were observed in
whole blood of workers exposed to relatively high concentrations estimated at 5.99 mg
hexavalent chromium/m3 (Wu et al., 2001). However, chromosome aberrations and sister
chromatid exchanges in peripheral lymphocytes from chrome platers were not observed at lower
exposure levels (0.0075 and 0.0249 mg chromium[total]/m3) (Benova et al., 2002). Other studies
reporting positive (Wu et al., 2000; Sarto et al., 1982; Stella et al., 1982) or negative (Nagaya et
al., 1991; Nagaya, 1986) results for chromosome aberrations or sister chromatid exchanges in
peripheral lymphocytes of workers did not report hexavalent chromium exposure levels.
Micronuclei formation in peripheral lymphocytes was also observed in chrome platers at
exposure levels of 0.043–0.083 mg chromium(total)/m3 (Vaglenov et al., 1999) and 0.0075–
0.0249 mg chromium(total)/m3 (Benova et al., 2002). In buccal mucosal cells collected from
chrome platers, micronuclei formation was increased at exposure levels of 0.0075–0.0249 mg
chromium(total)/m3, although chromosome aberrations and sister chromatid exchanges were not
observed (Benova et al., 2002). Sarto et al. (1990) reported negative results for micronuclei in
buccal and nasal cells of chrome platers, but exposure levels were not reported.
In summary, results of available studies in hexavalent chromium-exposed workers
provide some evidence of the mutagenic activity of hexavalent chromium in occupationally
exposed humans, but results have not been consistent across studies and endpoints. For example,
associations with increased micronuclei in peripheral lymphocytes or buccal mucosal cells have
been reported in chrome platers at estimated exposure levels as low as 0.0075–0.0249 mg
chromium(total)/m3 (Benova et al., 2002; Vaglenov et al., 1999), although chromosome
aberrations and sister chromatid exchanges were not observed (Benova et al., 2002). In contrast,
increased frequencies of chromosome aberrations and sister chromatid exchanges were observed
in another group of chrome platers exposed to higher concentrations estimated at 5.99 hexavalent
chromium/m3 (Wu et al., 2001).
4.5.2. Intracellular Reduction

The mutagenic effects of hexavalent chromium are contingent upon its reduction within
the cell. Extracellularly, soluble hexavalent chromium exists as a chromate oxyanion. The
tetrahedral arrangement of the oxygen groups makes it structurally similar to phosphate and
sulfate, allowing it to easily be taken up by the nonspecific phosphate/sulfate anionic transporters
and cross the cell membrane (Zhitkovich, 2005). This method of cellular uptake also allows an

accumulation of chromium in the cell at concentrations much higher than that found
extracellularly (Zhang et al., 2002). Chromium in its hexavalent state is thermodynamically
stable in pure water, and is not reactive with DNA at physiological concentrations. However,
hexavalent chromium is a strong oxidizer, and once inside the cell, it can undergo rapid
reduction. This is most often mediated by the nonenzymatic reductants ascorbate (vitamin C)
and low molecular weight thiols including GSH and cysteine. Other potential reductants include
cytochrome P450 reductase, NAD(P)H-dependent flavoenzymes, and mitochondrial electron
transport complexes (O’Brien et al., 2003; Sugden and Stearns, 2000; Standeven and
Wetterhahn, 1989).
The hexavalent chromium-reductant substrate complexes that are formed upon
intracellular interaction of hexavalent chromium with these reductants are considered the first
step in the reduction process, although the actual mechanisms of how these reactions proceed are
unknown (Levina and Lay, 2005). There are two theorized pathways for the intracellular
reduction of hexavalent chromium. When reductants are present in abundance, the process can
occur with a two-electron reduction to tetravalent chromium, immediately followed by a one-
electron reduction to trivalent chromium. If lower levels of reductants are available, the first step
of this process will occur as two distinct one-electron transfers, producing the intermediates,
pentavalent and tetravalent chromium, and ultimately trivalent chromium (O’Brien et al., 2003).
Either process can produce oxidative states of chromium localized within the cell that are able to
damage DNA directly, forming DNA adducts and subsequent DNA breakage. These chromium
species can also indirectly cause genetic damage via associated radical species derived from the
reductants that can be involved in secondary DNA damage (Sugden and Stearns, 2000) and
disruption of DNA replication.
Final reduction product: trivalent chromium. Trivalent chromium is the ultimate
product of hexavalent chromium reduction within the cell. It contains six coordination sites,
allowing it to form stable complexes with amino acids, proteins, RNA, and DNA. In vitro
studies of the kinetics of chromium−DNA binding have shown that most of the DNA binding
occurs within 1 hour of incubation (Quievryn et al., 2003). When hexavalent chromium is
reduced by ascorbate or cysteine in the presence of the trivalent chromium chelator, EDTA, the
mutagenic response is all but eliminated and very little chromium−DNA binding is detected,
indicating that the trivalent state is the most DNA reactive of all the valence states of chromium
(O’Brien et al., 2003; Quievryn et al., 2003; Zhitkovich et al., 2001). Several types of
chromium−DNA adducts have been detected following the intracellular reduction of hexavalent
to trivalent chromium.
DNA−peptide/amino acid ligand−trivalent chromium crosslinks. Trivalent chromium
can form ternary DNA crosslinks with GSH, ascorbate, cysteine, and histidine. Although the
ascorbate−trivalent chromium−DNA adducts are recovered less frequently in vitro due to the low
concentrations of vitamin C present in commonly used tissue culture media (Zhitkovich, 2005),

these adducts have been shown to be the most mutagenic of all the ternary adducts (Quievryn et
al., 2003). These ternary adducts form by the attachment of trivalent chromium (in a binary
complex with the ligand) to phosphate groups in DNA (Zhitkovich et al., 1995), primarily
through coordinate covalent binding or electrostatic/ionic interactions (O’Brien et al., 2003)
(Figure 4-1). They have been detected in vitro in Chinese hamster ovary cells following
exposure to hexavalent chromium, and account for up to 50% of all chromium−DNA adducts.
The ternary adducts have been found to cause mutagenic and replication-blocking lesions in
human fibroblasts in vitro (Quievryn et al., 2003; Voitkun et al., 1998).
DNA
O
OH
L L
O
H
L Cr(III) O
O H N NH
7
O P O L
O N
O N NH2
OH
O P O
O DNA
Hexavalent chromium, when reduced intracellularly to trivalent chromium, can

form ternary DNA crosslinks with the peptide or amino acid ligand (L) involved
in the reduction. Here, chromium(III) directly coordinates to the 5’-phosphate in
the DNA backbone and forms a hydrogen bond with the N-7 of deoxyguanosine.
Source: Zhitkovich (2005).
Figure 4-1. Ternary DNA adduct formation by chromium.
DNA−trivalent chromium crosslinks. Reduction of hexavalent chromium in vitro

produces a large proportion of binary trivalent chromium−DNA adducts, but these have not been
detected in vivo. It has been theorized that the formation of the ternary adducts described above
occurs far more frequently due to the high concentration of ligands capable of complexing with
trivalent chromium before it can bind to DNA (Zhitkovich, 2005). In addition, these adducts

have been found to be less mutagenic than the ternary adducts in vitro (Quievryn et al., 2003;
Zhitkovich et al., 2001).
DNA−protein crosslinks. These bulky lesions have been detected in hexavalent
chromium-treated cells in vitro in Chinese hamster ovary cells (Costa, 1991) and in vivo in chick
embryos (Hamilton and Wetterhahn, 1986). They are not detected in the presence of the
trivalent chromium chelator, EDTA, indicating that trivalent chromium is the species involved in
their formation (Miller and Costa, 1989). It has been recently shown that the mechanism
forming DNA−protein crosslinks induced by hexavalent chromium requires intracellular
reduction to trivalent chromium, formation of DNA−trivalent chromium adducts, and subsequent
capture of proteins by the DNA bound to trivalent chromium (Macfie et al., 2010). Tests for the
mutagenicity of these crosslinks have proved inconclusive (reviewed in Macfie et al., 2010), but
the bulkiness of these lesions indicates the potential for genotoxicity resulting from replication
fork stalling (Costa, 1991).
DNA−DNA crosslinks. These inter- or intra-strand DNA crosslinks are likely formed by
oligomers of trivalent chromium. They have been detected following hexavalent chromium
exposure, although only when the reductants are ascorbate or cysteine, and not GSH (Zhitkovich,
2005). However, these adducts have only been detected in vitro and are not expected to form in
significant amounts in vivo; the high intracellular concentrations of ligands available to form
complexes with trivalent chromium make it unlikely that these oligomers would have a chance to
form (Salnikow and Zhitkovich, 2008).
Repair of chromium−DNA adducts. Repair processes have been shown to be effectively
carried out by excision repair (ER), a DNA repair mechanism responsible for removal of bulky
DNA lesions. Exposing nucleotide excision repair (NER)-deficient human cells to hexavalent
chromium was shown to induce apoptosis and clonogenic cell death. The most efficient
substrates for this repair process are lesions that create major distortions in the DNA structure.
Chromium−DNA adducts do not create major helix distortions, but their bulkiness makes them
adequate substrates for NER, although they are less efficiently removed than optimal NER
substrates such as UV light-induced lesions (Reynolds et al., 2004). Interestingly, it has also
been shown that proficient ER systems, including both nucleotide and base excision repair (NER
and BER), are involved in genomic instability resulting from hexavalent chromium exposure. In
a study by Brooks et al. (2008), cell lines deficient in these repair mechanisms were protected
from hexavalent chromium-induced chromosomal instability.
Another closely related repair mechanism, mismatch repair (MMR), is responsible for the
correction of errors in DNA replication. MMR enzymes recognize misincorporated bases during
DNA replication and homologous recombination, and repair single base mispairings and small
insertions or deletions. However, MMR has also been shown to be a causative factor in many of
the toxic and genotoxic effects of hexavalent chromium, when processing the repair of the bulky
lesions formed by chromium lead to the formation of DNA double-strand breaks (Peterson-Roth

et al., 2005). In this study, mouse and human cell lines deficient in MMR exposed to hexavalent
chromium had greatly increased clonogenic survival due to a diminished apoptotic response as
compared to MMR-proficient cells. The apoptotic response in the MMR-proficient cells was
preceded by a significant induction of DNA double-strand breaks, indicated by an increased
formation of gamma-H2AX foci. These discrete foci form when phosphorylation of this histone
H2A variant occurs in response to DNA double-strand breaks, and can be visualized and
quantified by immunofluorescence. This increase in gamma-H2AX foci was not detected at
significant levels until 6 hours postexposure to hexavalent chromium, suggesting that the DNA
double-strand breaks were not induced directly by hexavalent chromium, but rather from
processing of the damaged DNA. These foci also co-localized with cyclin B1 staining,
indicating the breaks occurred in the G2 phase of the cell cycle and providing evidence that
passage through S phase, where MMR would be taking place, was necessary for the induction of
this damage (Salnikow and Zhitkovich, 2008). The mechanism of this toxic response mediated
by MMR proteins is unknown, but has been theorized to involve the futile repair of damaged
bases or the initiation of a stress response (Peterson-Roth et al., 2005).
Cellular effects. Genomic instability, defined as an increased rate of acquisition of
alterations in the genome, is a hallmark of tumorigenesis and may be instrumental in hexavalent
chromium-induced carcinogenicity. The loss of MMR function leads to an unstable mutator
phenotype, in which replication errors, particularly those occurring in simple nucleotide repeat
sequences known as microsatellites, are not corrected, leading to an increase in mutation
frequency (Loeb et al., 2008). Further, chromosomal instability has been demonstrated in human
lung cells in vitro exposed to particulate hexavalent chromium. Following chronic exposures,
Holmes et al. (2010) found concentration- and time-dependent increases in aneuploidy as the
result of centrosome amplification and spindle assembly checkpoint bypass. Thus, genomic
instability may result after prolonged exposure to hexavalent chromium.
As mentioned above, apoptosis, or programmed cell death, has been observed in cells
exposed to hexavalent chromium as a response to extensive DNA damage that cannot be
adequately repaired by the cell. Ye et al. (1999) found hexavalent chromium induced apoptosis
in human lung epithelial cells exposed to doses ranging from 75 to 300 µM in vitro; the authors
theorized that this response involved reactive oxygen species formed both directly during the
process of hexavalent chromium reduction and indirectly through the induction of p53.
Similarly, Wang et al. (2006) measured significant levels of apoptosis in liver and not kidney in
mice exposed to daily gavage doses of 0, 25, 50, or 100 mg/kg hexavalent chromium for 1 or 5
days. The apoptosis was accompanied by an increase in ROS and dose-dependent increases in
DNA damage, SOD, and catalase. Flores and Perez (1999), using doses close to the IC50 values,
observed apoptosis concurrent with DNA interstrand crosslinks and DNA single-strand breaks in
murine keratinocytes transformed with the H-ras oncogene. These studies indicate that multiple

mechanisms induced by hexavalent chromium exposure, including oxidative stress and DNA
binding, can lead to cell death.
In addition to the effects involving DNA repair mechanisms is the finding that hexavalent
chromium, after intracellular reduction to the +3 oxidation state, can interfere with normal DNA
replication and transcription processes. Intracellular trivalent chromium has been shown to
inhibit the enzymatic activity of DNA polymerases, simultaneously increasing the rate of
replication and the processivity of the DNA polymerase, thereby decreasing its fidelity and
causing more frequent errors, with a dose-dependent increase in mutation frequency in vitro
(Snow, 1991). There can also be replication arrest as a result of the bulky chromium−DNA
lesions, creating a physical obstruction to the progression of DNA polymerases (Bridgewater et
al., 1998). These effects were recently confirmed in a study utilizing the DNA synthesome, an in
vitro DNA replication model system that is fully competent to carry out all phases of the DNA
replication process mediated by human cells (Dai et al., 2009). This study found a reduction of
the fidelity and an inhibition of DNA synthesis that led to a dose-dependent increase in mutation
frequency following intracellular exposure to trivalent chromium. Thus, hexavalent chromium
can lead to the disruption of DNA synthesis and gene transcription at multiple levels,
corresponding to an observable, dose-dependent increase in mutation frequency in human cells.
Epigenetic effects have also been observed following hexavalent chromium exposure.
Epigenetic modifications, defined as heritable changes in gene expression that occur without
altering the genetic material (Sharma et al., 2010), can drive malignant cellular transformation.
These modifications can effect methylation, phosphorylation, gene expression, and cell signaling
(Holmes et al., 2008). Cellular signaling involved in cell survival may be affected; a study
exposing human bronchial epithelial cells in vitro found that soluble hexavalent chromium could
inhibit apoptosis via NF-kB activation and inhibition of p53 (Wang et al., 2004). DNA repair
has also been shown to be sensitive to epigenetic modifications. In a study finding microsatellite
instability in hexavalent chromium-induced lung tumors of chromate-exposed workers (Hirose et
al., 2002), increased DNA methylation was observed in the promoter region of the tumor
suppressor gene p16 and the MMR gene hMLH1, indicating that chromium can induce
epigenetic effects (Kondo et al., 2006; Takahashi et al., 2005). Gene transcription has also been
shown to be affected by exposure to hexavalent chromium in vitro via epigenetic mechanisms.
Sun et al. (2009) found alterations in the levels of histone methylation in human lung A549 cells
exposed to hexavalent chromium, indicating the capability of these exposures to lead directly to
changes in gene expression. This evidence suggests that epigenetic effects may contribute to the
carcinogenicity of hexavalent chromium and its reduced valence states once inside the cell.
Reduction intermediates: pentavalent and tetravalent chromium. Depending on the
reductant involved and the concentration of hexavalent chromium present, various amounts of
the unstable intermediates pentavalent and tetravalent chromium can be generated prior to
reduction to the final stable oxidative +3 state. At lower levels of hexavalent chromium

exposure, intracellular concentrations of these reductants are sufficient to complete the reduction
of hexavalent chromium to its trivalent state. However, at higher hexavalent chromium
exposures, these levels are depleted, resulting in a higher yield of pentavalent chromium from the
one-electron reducing thiols, GSH, and cysteine, as well as tetravalent chromium from the two-
electron donating ascorbate. While pentavalent and tetravalent chromium can be short-lived
states of chromium within the cell, they are DNA reactive and can participate in redox reactions,
forming free radical species that can also damage DNA (Stearns and Wetterhahn, 1994).
Redox cycling of the chromium ions can occur intracellularly when they are formed
during reduction of hexavalent chromium. The process of hexavalent chromium reduction by
GSH is accompanied by the reduction of molecular oxygen, yielding superoxide radicals.
Reduction by GSH has been shown to involve the formation of GSH-derived thiyl radicals that
can directly damage DNA or react with other thiols to also generate superoxide radicals. These
radical species will react with hydrogen peroxide to produce hydroxyl radicals via Haber-Weiss
reactions (Shi et al., 1999). Both hydrogen peroxide and superoxide radicals can participate in
redox reactions involving both the pentavalent and tetravalent transition states of chromium that
can generate hydroxyl radicals via Fenton and Haber-Weiss reactions (Shi et al., 1999).
Hydroxyl radicals can directly react with genetic material, forming DNA-protein crosslinks, and
DNA adducts with proteins and amino acids, damaging DNA bases, and producing DNA single-
and double-strand breaks (reviewed in Kasprzak, 1996).
Although less frequent than the low molecular weight nonenzymatic reductants,
reduction of hexavalent chromium can also occur by NAD(P)H-dependent flavoenzymes,
including GSH reductase, lipoyl dehydrogenase, and ferredoxin-NADP+ oxidoreductase (Shi
and Dalal, 1990). These enzymes catalyze a one-electron reduction that can result in the
formation of stable pentavalent chromium−NADPH complexes that can react with hydrogen
peroxide to generate hydroxyl radicals (Shi et al., 1999). The ability to form complexes with
biological ligands allows stabilization of pentavalent, but not tetravalent, chromium
intermediates (Levina and Lay, 2005). These pentavalent chromium−NADPH complexes have
been shown to form in vitro in E. coli (Shi et al., 1991) and in vivo in mice (Liu et al., 1995).
Two other important nonenzymatic reducers of hexavalent chromium are ascorbate and
cysteine. Ascorbate and cysteine are present at lower concentrations intracellularly than GSH,
but they have kinetically faster rates of hexavalent chromium reduction. Ascorbate has been
shown to yield pentavalent and tetravalent chromium and radical species when the intracellular
ratio of ascorbate to chromium is <3:1 (Stearns and Wetterhahn, 1994). The precise nature of the
radical species relevant to DNA damage is not known, however, and the degree of damage
attributable to oxidative mechanisms is the subject of much debate. One study found an increase
in mutations and replication-blocking DNA lesions in human fibroblasts resulting from the
ascorbate-driven reduction of hexavalent chromium, but concluded that the mechanism
responsible did not involve oxidative radicals, in part because the DNA damage anticipated by

species including hydroxyl radicals and pentavalent chromium-peroxo complexes, namely abasic
sites and strand breaks was not observed (Quievryn et al., 2003). This study also found that no
mutagenesis occurred in the presence of a trivalent chromium chelator, indicating the
involvement of trivalent chromium−DNA adducts (see previous section). Similarly, studies of
the DNA damage resulting from the intracellular reduction of hexavalent chromium by cysteine
have shown that, while the intermediate species pentavalent and tetravalent chromium and thiyl
radicals were formed, they were not responsible for DNA damage; rather, the trivalent
chromium−DNA adducts were found to be the mutagenic species (Zhitkovich et al., 2001). The
same group also found an elimination of mutagenicity when GSH reduction of hexavalent
chromium occurred in the presence of phosphate ions that led to the sequestration of trivalent
chromium, preventing its binding to DNA (Guttmann et al., 2008).
The ability of these intermediate chromium species to generate damaging free radicals is
not in doubt, however, and there is evidence of reactive oxygen species generated by pentavalent
chromium causing DNA damage. A decrease in DNA strand breaks was observed when
hexavalent chromium reduction with GSH occurred in the presence of free radical scavengers
(Kortenkamp et al., 1990). In addition, DNA double-strand breaks in subcellular systems were
observed when ascorbate-mediated reduction of hexavalent chromium generated hydroxyl
radicals via a Fenton-like reaction (Shi et al., 1994).
In an attempt to explain these conflicting results, it has been theorized that the
responsible free radicals may be chromium-based and not oxygen-based radicals. This is due to
the observation that the mutational spectra observed by chromium-induced radicals differs from
that expected by damage due to reactive oxygen species that are generated following exposure to
hydrogen peroxide, x-rays, or ionizing radiation (Sugden and Stearns, 2000). Hexavalent
chromium has been shown to induce the formation of 8-oxo-deoxyguanosine adducts that are
known to be induced by oxidative damage (Sander et al., 2005), but these lesions have also been
shown to be induced directly by pentavalent chromium, with the subsequent addition of
molecular oxygen (Sugden and Martin, 2002). In addition, the oxidant-sensitive dyes used to
detect reactive oxygen species intracellularly can also be oxidized directly by pentavalent
chromium and chromium-based radicals (O’Brien et al., 2003). Therefore, the induction of
mutagenic lesions by the intracellular reduction of hexavalent chromium could be attributed to
nonoxygen-dependent mechanisms.
Pentavalent chromium has been detected using EPR spectroscopy following
intraperitoneal administration of hexavalent chromium in vivo, both in the liver and RBCs of
chick embryos (Liebross and Wetterhahn, 1992), and in mouse liver and blood (Liu et al., 1994).
In vitro, levels of DNA strand breaks were found to correlate with increasing levels of
pentavalent chromium in Chinese hamster V79 cells (Sugiyama et al., 1989). Another in vitro
study in human leukemic T-lymphocyte MOLT4 cells detected pentavalent chromium species
and hydroxyl radicals with EPR following exposure to hexavalent chromium (Mattagajasingh et

al., 2008). The same study also observed a dose-dependent increase in protein carbonyls and
malondialdehyde generated via protein oxidation and lipid peroxidation, respectively, although
the lipid peroxidation only occurred significantly at much higher exposures of chromate
(≥100 µM) compared with the protein oxidation, which was significant as low as 10 µM.
Tetravalent chromium has been more difficult to observe due to its unstable nature compared to
pentavalent chromium, but this species was shown to induce mitotic recombination in the
somatic wing spot assay in Drosophila (Katz et al., 2001). Both species caused an induction of
NF-kB, a nuclear transcription factor involved in the cellular response to oxidative damage, in
cultured Jurkat cells. This activation was enhanced by hydrogen peroxide and eliminated when
catalase was added to decompose hydrogen peroxide, indicating that hydroxyl radicals may have
had a role (Shi et al., 1999).
In summary, there are many potential mechanisms involved in the genotoxicity of
hexavalent chromium as a result of intracellular reduction. Intermediate valence states can react
directly and indirectly through coordinate complexes with DNA as well as form radical species,
and the final reduction product, trivalent chromium, can form various damaging DNA adducts.
Additionally, significant evidence points to the aberrant processing of DNA mismatches induced
by chromium−DNA adducts, leading to apoptosis of the damaged cells, or further promotion of
these mutagenic lesions as the DNA double-strand breaks generated are substrates for error-
prone repair processes such as nonhomologous end joining.
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS—ORAL

In humans, several case reports have been published on clinical signs and symptoms in
individuals following acute accidental or intentional ingestion of high doses (fatal or near fatal)
of hexavalent chromium compounds, including chromic acid, potassium dichromate, and
ammonium dichromate. Clinical presentation of patients following acute, high-dose exposure
was similar, regardless of the specific hexavalent chromium compound ingested, and included
the following: abdominal pain, nausea, and vomiting; hematemesis and bloody diarrhea; caustic
burns of mouth, pharynx, esophagus, stomach, and duodenum and GI hemorrhage; anemia,
decreased blood Hgb, abnormal erythrocytes, and intravascular hemolysis; hepatotoxicity
(hepatomegaly, jaundice, elevated blood bilirubin, and liver enzyme activities); renal failure
(oliguria and anuria); cyanosis; and metabolic acidosis, hypotension, and shock. Findings on
tissue biopsies included hepatic fatty degeneration and necrosis and renal tubular degeneration
and necrosis.
Information on chronic human health effects resulting from exposure to hexavalent
chromium comes from several studies of human populations unknowingly consuming food or
drinking water contaminated with hexavalent chromium over some extended time period. These
studies have been primarily focused on cancer. However, the noncancer effects that have been
recorded are consistent with the GI effects observed following acute exposures to hexavalent

chromium and have included oral ulcers, diarrhea, abdominal pain, dyspepsia, stomach pain, and
vomiting (JinZhou Antiepidemic Station, 1979).
Table 4-26 presents a summary of studies of the noncancer effects of hexavalent
chromium exposure from repeat-dose oral toxicity studies in experimental animals. The most
sensitive targets of toxicity identified in these studies included the blood, liver, and GI tract. The
effects seen in these target organs are more specifically discussed below.

Table 4-26. Observed effects and corresponding NOAELs and LOAELs for subchronic, chronic, and reproductive
toxicity studies following oral exposure to hexavalent chromium
Exposure NOAEL LOAEL

Species Sex Exposure levela duration (mg/kg-d) (mg/kg-d) Effects at the NOAEL/LOAEL Reference
Subchronic studies
F344/N rat F, M 0, 1.7, 3.5, 5.9, 3 mo F: ND 1.7 F: Microcytic, hypochromic anemia (decreased Hgb, NTP, 2007
11.2, or 20.9 M: ND 1.7 MCV, MCH), increased serum liver enzyme activities
mg/kg-d via (ALT and SDH) and bile acids, and histopathological
drinking water changes to the duodenum (histiocytic cellular infiltration).
M: Microcytic, hypochromic anemia (decreased Hct,

Hgb, MCV, MCH), increased serum liver enzyme
activities (ALT and SDH), and histopathological changes
to pancreatic lymph nodes (histiocytic cellular
infiltration).
B6C3F1 mouse F, M 0, 3.1, 5.3, 9.1, 3 mo F: ND 3.1 Histopathological changes (histiocytic cellular NTP, 2007
15.7, or M: ND 3.1 infiltration) in the duodenum.
27.9 mg/kg-d
via drinking
water
B6C3F1, M 0, 2.8, 5.2, or 3 mo ND 2.8 Histopathological changes in the duodenum in B6C3F1 NTP, 2007
BALB/c, and 8.7 mg/kg-d via mice (histiocytic cellular infiltration and epithelial
am3-C57BL/6 drinking water hyperplasia), BALB/c mice (histiocytic cellular
mouse infiltration), and am3-C57BL/6 mice (epithelial
hyperplasia).
Wistar rat M 0 or 73.05 mg/ 30 d ND ND Decreased serum prolactin levels. Data not adequate for Quinteros et al.,
kg-d via estimation of a NOAEL or LOAEL. 2007
drinking water
Wistar rat M 0 or 20 mg/L in 10 wks ND ND Liver histopathologic changes. Doses in mg hexavalent Rafael et al.,
drinking water chromium/kg-d could not be estimated. 2007
Wistar rat M 0 or 1.5 mg/kg-d 22 wks ND 1.5 Changes in serum enzymes; liver triglycerides, glycogen Acharya et al.,
via drinking and cholesterol; liver histopathologic changes. 2001
water
Swiss mouse M 0, 177, 265, 8 wks ND ND Liver histopathologic changes. Doses in mg hexavalent Asmatullah and
353, 530, or chromium/kg-d could not be estimated. Noreen, 1999
706 mg/L in
drinking water


Wistar rat F 0 or 1.4 mg/kg-d 22 wks ND 1.4 Changes in liver weight; serum enzyme levels, Chopra et al.,
via drinking triglycerides, glucose; liver glycogen; liver 1996
water histopathology.
Wistar rat F, M F: 0 or 1.76– 6 mo ND ND Changes in urinary markers of renal function. No Vyskocil et al.,
2.47 mg/kg-d histopathologic examination of the kidney. 1993
via drinking
water
M: 0 or 1.4–
2.18 mg/kg-d
via drinking
water
Chronic studies
F344/N rat F, M F: 0.24, 0.94, 2 yrs F: ND 0.24 F: Increased incidence of chronic inflammation of the NTP, 2008
2.4, or 7.0 mg/ M: 0.21 0.77 liver.
kg-d via
drinking water M: Increased incidences of nonneoplastic
histopathological changes to the liver (basophilic foci),
M: 0.21, 0.77, duodenum (histiocytic cellular infiltrate), and mesenteric
2.1, or lymph nodes (histiocytic cellular infiltrate and
5.9 mg/kg-d via hemorrhage).
drinking water
B6C3F1 mouse F, M F: 0.38, 1.4, 2 yrs F: ND 0.38 F: Increased incidences of histopathological changes to NTP, 2008
3.1, or M: ND 0.38 the duodenum (diffuse epithelial hyperplasia), mesenteric
8.7 mg/kg-d via lymph nodes (histiocytic cellular infiltration), liver
drinking water (histiocytic cellular infiltration), and pancreas (depletion
of cytoplasmic zymogen granules).
M: 0.38, 0.91,
2.4, or M: Increased incidences of histopathological changes to
5.9 mg/kg-d via the duodenum (diffuse epithelial hyperplasia) and
drinking water mesenteric lymph nodes (histiocytic cellular infiltration).
Dog Not 0, 0.45, 2.25, 4 yrs ND ND No effects were observed. Doses in mg hexavalent Anwar et al.,
specified 4.5, 6.75, or chromium/kg-d could not be estimated. 1961
11.2 mg/L in
drinking water


Sprague- F, M 0.05–2.8 mg/kg- 1 yr 2.4–2.8 ND No adverse effects observed at the highest dose tested. MacKenzie et
Dawley rat d via drinking al., 1958
water
Reproductive/developmental studies
Bonnet M 0, 1.0, 2.1, 4.1, 180 d ND 2.1 Reversible changes to male reproductive organs, Aruldhas et al.,
monkey and 8.3 mg/kg-d including disruption of spermatogenesis, effects on sperm 2006, 2005,
via drinking count and velocity, and histopathological changes. 2004;
water Subramanian et
al., 2006
Charles Foster M 0, 20, 40, or 90 d ND 20 Decreased serum testosterone levels and loss of Chowdhury and
rat 60 mg/kg-d via 3β-Δ5-HSH activity in testes. Mitra, 1995
gavage
Wistar rat M 0, 5.2, or 6d ND 5.2 Decreased sperm counts and histopathological changes to Li et al., 2001
10.4 mg/kg-d the testes.
via gavage
BALB/c mouse M 0, 6.4, 12.7, or 35 d ND 6.4 Increased percentage of degenerated tubules, Zahid et al.,
25.5 mg/kg-d undergenerated tubules without spermatogonia, abnormal 1990
via gavage sperm, and reduced number of spermatogonia.
New Zealand M 0 or 3.6 mg/kg-d 10 wks ND 3.6 Decreased testes and epididymis weight and decreased Yousef et al.,
White rabbit via gavage sperm output. 2006
Sprague- F, M F: 0, 0.25, 1.1, 9 wks F: 2.5 F: 9.9 F: Slight erythrocyte microcytosis. NTP, 1996b
Dawley rat 2.5, or 9.9 M: 2.1 M: 8.5
mg/kg-d via the M: Slight erythrocyte microcytosis.
diet
M: 0, 0.35, 1.1,
2.1, or
8.5 mg/kg-d via
the diet


BALB/c mouse F, M F: 0, 1.8, 5.6, or 9 wks F: 1.8 F: 5.6 F: Histopathological changes to the liver (cytoplasmic NTP, 1996a
12.0, M: 3.5 M: 7.4 vacuolization).
48.4 mg/kg-d
via the diet M: Histopathological changes to the liver (cytoplasmic
vacuolization).
M: 0, 1.1, 3.5,
7.4, or
32.5 mg/kg-d
via the diet
BALB/c mouse F 0, 7.9, 16.1, or Continuous ND 7.9 Erythrocyte microcytosis (slight decrease in MCH) in the NTP, 1997
37.1 mg/kg-d breeding study F1 generation.
via the diet
(F1 generation)
Druckrey rat F 0, 70, 127, or 3 mo ND 70 Dam: Increased pre- and postimplantation losses. Kanojia et al.,
170 mg/kg-d via 1998
drinking water Offspring: Decreased fetal weight and external and
skeletal abnormalities.
Swiss mouse F 0, 63, 119, or GDs 6 through ND 63 Dam: Decreased fertility. Junaid et al.,
174 mg/kg-d via 14 1996a
drinking water Offspring: Decreased fetal body weight and delays in
skeletal development.
Wistar rat F 0 or 7.9 mg/kg-d GDs 6 through ND 7.9 Dam: Increased preimplantation loss/litter, Elsaieed and
via drinking 15 postimplantation loss/litter, resorptions/litter, and dead Nada, 2002
water fetuses/litter and decreased live fetuses/litter.
Offspring: Decreased fetal weight and increased litters

with fetal abnormalities or malformations including
visceral and skeletal changes.
Sprague- F 0 or 35 mg/kg-d GDs 1–3 or 4–6 ND 35 Dam: Impaired implantation, increased resorptions, and Bataineh et al.,
Dawley rat via drinking decreased number of viable fetuses. 2007
water
ITRC-Bred F 0, 48, 98, or Entire Dam: 48 Dam: 98 Dam: Decreased body weight gain and increased Trivedi et al.,
mouse 239 mg/kg-d gestational Offspring: ND Offspring: 48 resorptions and postimplantation loss. 1989
via drinking period
water Offspring: Decreased fetal length and weight.


Swiss mouse F 0, 53, 101, or 20 days prior to Dam: 53 Dam: 101 Dam: Decreased body weight gain and reduced number Junaid et al.,
152 mg/kg-d via mating Offspring: ND Offspring: 53 of implantation sites. 1996b
drinking water
Swiss mouse F 0, 53, 101, or GDs 14 through Dam: 53 Dam: 101 Dam: Decreased body weight gain. Junaid et al.,
152 mg/kg-d via 19 Offspring: ND Offspring: 53 1995
drinking water Offspring: Reduced fetal weight and length and increased
incidence of reduced caudal ossification.
a
Unless otherwise noted, dose or concentration expressed as hexavalent chromium.
F = female; M = male; ND = not determined

In regard to hematological effects, NTP (2007) observed microcytic, hypochromic
anemia (i.e., decreased Hct, Hgb, MCV, and MCH) at a dose of 1.7 mg/kg-day of hexavalent
chromium in both male and female F344/N rats in a 3-month (subchronic) study. In this same
study, NTP (2007) also saw histopathological changes (i.e., histiocytic cellular infiltration) in the
pancreatic lymph nodes in male F344/N rats at 1.7 mg/kg-day of hexavalent chromium. Finally,
in a chronic (2-year) study, NTP (2008) observed histopathological changes (i.e., histiocytic
cellular infiltration) in the mesenteric lymph nodes in male F344/N rats at 0.77 mg/kg-day of
hexavalent chromium and male and female B6C3F1 mice at 0.38 mg/kg-day of hexavalent
chromium.
In the NTP (2007) subchronic study referenced above, liver effects were also observed at
1.7 mg/kg-day of hexavalent chromium and included increased serum liver enzyme activities
(i.e., ALT and SDH) in both males and females and increased bile acids in females. In their
2-year bioassay, NTP (2008) found an increased incidence of chronic inflammation of the liver
at 0.24 mg/kg-day of hexavalent chromium in female F344/N rats and increased incidences of
histopathological changes to the liver (i.e., basophilic foci) at 0.77 mg/kg-day of hexavalent
chromium in male F344/N rats. In this same bioassay, increased incidences of histopathological
changes to the liver (i.e., histiocytic cellular infiltration) were seen at 0.38 mg/kg-day of
hexavalent chromium in female B6C3F1 mice.
Effects of hexavalent chromium ingestion on the GI tract have been primarily observed in
the small intestine (duodenum). In a 3-month study, NTP (2007) saw histopathological changes
to the duodenum in male F344/N rats at 1.7 mg/kg-day of hexavalent chromium, in male and
female B6C3F1 mice at 5.3 mg/kg-day of hexavalent chromium, and in male BALB/c and
am3-C57BL/6 mice at 2.8 mg/kg-day of hexavalent chromium. These changes included diffuse
epithelial hyperplasia and histiocytic cellular infiltration. In their 2-year study, NTP (2008) also
found increased incidences of histopathological changes to the duodenum in male F344/N rats at
0.77 mg/kg-day of hexavalent chromium and in male and female B6C3F1 mice at 0.38 mg/kg-
day of hexavalent chromium. Similar to that observed in the subchronic study, these changes in
the duodenum included diffuse epithelial hyperplasia and histiocytic cellular infiltration.
Animal studies also provide evidence that oral exposure to hexavalent chromium
compounds produces reproductive effects, including histopathological changes to reproductive
organs in males (Aruldhas et al., 2006, 2005, 2004; Li et al., 2001; Chowdhury and Mitra, 1995;
Zahid et al., 1990) and females (Murthy et al., 1996); alterations in sperm, including decreased
count, decreased motility, and abnormal morphology (Subramanian et al., 2006; Yousef et al.,
2006; Li et al., 2001; Zahid et al., 1990); decreased plasma testosterone levels (Yousef et al.,
2006; Chowdhury and Mitra, 1995); increased estrous cycle length (Kanojia et al., 1998, 1996;
Murthy et al., 1996); changes in mating behavior and decreased fertility in males (Bataineh et al.,
1997); and adverse reproductive outcomes, including decreased numbers of live fetuses and

implantations, and increased numbers of resorptions and pre- and postimplantation losses
(Bataineh et al., 2007; Elsaieed and Nada, 2002; Kanojia et al., 1998, 1996; Elbetieha and Al-
Hamood, 1997; Junaid et al., 1996a, b, 1995; Trivedi et al., 1989). These reproductive toxicity
studies are summarized in Table 4-26.
Developmental effects observed in animal studies have included decreased fetal weight
and length (Elsaieed and Nada, 2002; Kanojia et al., 1998; Junaid et al., 1996a, b, 1995; Trivedi
et al., 1989); external (subdermal hemorrhage and tail malformations) and skeletal abnormalities
(decreased ossification) (Elsaieed and Nada, 2002; Kanojia et al., 1998, 1996; Junaid et al.,
1996a, b, 1995; Trivedi et al., 1989); and delayed sexual maturation and function in female
offspring (Banu et al., 2008; Al-Hamood et al., 1998). These effects were seen at hexavalent
chromium doses ranging from about 2 to 100 mg/kg-day. These studies and the developmental
effects observed are also summarized in Table 4-26.
In contrast to results of the above studies on reproductive toxicity, reproductive effects
were not observed in dietary exposure studies conducted by NTP that investigated the potential
effects of hexavalent chromium on male reproductive organs in rats and mice (NTP, 1996a,b)
and on reproductive outcomes in a continuous breeding study in mice (NTP, 1997). The reason
for the inconsistent results between the NTP studies and the other reproductive toxicity studies of
hexavalent chromium are not readily apparent, as daily dose ranges evaluated in the NTP studies
overlapped with those used in the other studies showing hexavalent chromium-induced
reproductive effects.
Based on a review of the NOAELs and LOAELs in Table 4-26, the most sensitive
hexavalent chromium-induced effects in rats were increased incidence of chronic inflammation
of the liver in females and increased incidences of nonneoplastic histopathological changes to the
liver (basophilic foci), duodenum (histiocytic cellular infiltrate), and mesenteric lymph nodes
(histiocytic cellular infiltrate and hemorrhage) in males. In mice, the most sensitive hexavalent
chromium-induced effects were increased incidences of histopathological changes to the
duodenum (diffuse epithelial hyperplasia), mesenteric lymph nodes (histiocytic cellular
infiltration), liver (histiocytic cellular infiltration), and pancreas (depletion of cytoplasmic
zymogen granules) in females and increased incidences of histopathological changes to the
duodenum (diffuse epithelial hyperplasia) and mesenteric lymph nodes (histiocytic cellular
infiltration) in males. All of these effects were observed in the 2-year chronic study by NTP
(2008), and in general, occurred at lower doses than the reproductive or developmental effects.
4.7. EVALUATION OF CARCINOGENICITY

4.7.1. Summary of Overall Weight of Evidence
Under the U.S. EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
hexavalent chromium is “likely to be carcinogenic to humans” via the oral route of exposure
based on a statistically significant increase in the incidence of tumors of the oral mucosa and

tongue of rats and of the small intestine of mice; and evidence of an association between oral
exposure to hexavalent chromium and stomach cancer in humans. Additionally, available
evidence indicates that chromium interacts with DNA, resulting in DNA damage and
mutagenesis. Thus, hexavalent chromium is proposed to induce carcinogenicity via a mutagenic
mode of action.
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence

Human studies in which health outcomes (primarily cancer) were evaluated among
populations that resided near sources of industrial waste containing hexavalent chromium
compounds and unknowingly consumed hexavalent chromium in drinking water provide some
evidence of possible associations between oral exposure to hexavalent chromium and cancer.
These epidemiological studies evaluated populations in Liaoning Province, China (Kerger et al.,
2009; Beaumont et al., 2008; Zhang and Li, 1997, 1987), Kings County/San Bernardino County,
California (Fryzek et al., 2001), Nebraska (Bednar and Kies, 1991), and Glasgow, United
Kingdom (Eizaguirre-Garcia et al., 2000, 1999) that unknowingly were exposed to hexavalent
chromium over some time period. Of these studies, the most detailed analyses were of data
collected from the JinZhou area of Liaoning Province, China, where groundwater, surface water,
and agricultural soils were contaminated with chromium derived from hexavalent chromium
production (e.g., 0.001–20 mg chromium/L in residential well water). This study found evidence
of an excess risk of mortality from stomach cancer from 1970 to 1978 in residents of the area,
relative to the reference populations (four other areas in Liaoning Province, and the total
population of the province) (Beaumont et al., 2008). The association with stomach cancer
mortality was weaker when an urban area was excluded from the reference population (Kerger et
al., 2009). However, there was little difference between stomach cancer rates in urban compared
to rural areas during this period, indicating no sound rationale for excluding this urban area from
the reference group. Studies of chromium-exposed populations in California and Nebraska
(Fryzek et al., 2001; Bednar and Kies, 1991) found no significant correlation between cancer
mortality and drinking water concentration, and the study of the population in Glasgow
(Eizaguirre-Garcia et al., 2000, 1999) found no correlation between leukemia risk and distance
from a former chromium processing facility (where elevated soil concentrations for hexavalent
chromium were measured). Interpretation of the findings from these three studies is limited by
the analysis of all cancer mortality (rather than individual cancer types) in the case of the
California and Nebraska studies and leukemia only in the case of the Glasgow study.
Evidence of carcinogenicity in animals was provided by the NTP (2008) bioassay
conducted in rats and mice. In this study, exposure of F344/N rats to sodium dichromate
dihydrate in drinking water for 2 years resulted in a statistically significant increase in the
incidence of squamous epithelial papillomas and carcinomas of the oral mucosa and tongue
(noted by NTP as rare when compared with historical controls) at the highest exposure level

(average daily doses of 5.9 and 7.0 mg hexavalent chromium/kg-day in males and females,
respectively), but not at the three lower exposure levels. NTP (2008) also exposed B6C3F1 mice
to sodium dichromate dihydrate in drinking water for 2 years and reported statistically significant
increases in the incidence of adenomas and carcinomas of the small intestine in males and
females at doses ≥2.4 and ≥3.1 mg hexavalent chromium/kg-day, respectively.
As discussed in detail in Section 4.6.3, hexavalent chromium is proposed to induce
carcinogenicity via a mutagenic mode of action. The key precursor events leading to
mutagenicity have been identified in animals and these events are anticipated to occur in humans
and progress to tumors.
The “likely to be carcinogenic to humans” descriptor is appropriate when the weight of
the evidence is adequate to demonstrate carcinogenic potential to humans but does not reach the
weight of evidence for the descriptor “carcinogenic to humans”. The database supports this
descriptor for hexavalent chromium exposure via the oral route. On the other hand, available
evidence to support the descriptor of “carcinogenic to humans” was also considered.
The “carcinogenic to humans” descriptor indicates strong evidence of human
carcinogenicity, and can be characterized by different combinations of evidence. One line of
evidence indicates this descriptor is appropriate when there is convincing epidemiologic
evidence of a causal association between human exposure and cancer (U.S. EPA, 2005a). This is
not the case for exposure to hexavalent chromium via ingestion. A moderately elevated risk of
stomach cancer mortality was seen in JinZhou (Liaoning Province, China), but this risk has not
been established in other populations exposed to drinking water contaminated with hexavalent
chromium. The epidemiologic data are not sufficient to establish a causal association between
exposure to hexavalent chromium by ingestion and cancer.
A second line of evidence under which this descriptor may be appropriate involves a
lesser weight of epidemiologic evidence that is strengthened by other information, including
strong evidence of an association between human exposure and either cancer or the key events of
the mode of action and extensive evidence of carcinogenicity in animals (U.S. EPA, 2005a). As
discussed above, the epidemiologic evidence for the oral route of hexavalent chromium exposure
is not considered strong. In addition, extensive evidence of the carcinogenicity of hexavalent
chromium in animals via ingestion does not exist. Only one multiple-dose chronic oral
carcinogenicity study of hexavalent chromium in animals is available (i.e., the 2-year bioassay in
rodents conducted by NTP [2008]). Taken together, these considerations do not provide a basis
for the characterization of hexavalent chromium as “carcinogenic to humans” via oral exposure.
Therefore, U.S. EPA concluded that, based on the available information, the descriptor “likely to
be carcinogenic to humans” is the most appropriate descriptor for the carcinogenic potential of
hexavalent chromium via ingestion.

4.7.3. Mode-of-Action Information
4.7.3.1. Hypothesized Mode of Action
The hypothesized mode of action for carcinogenicity induced by hexavalent chromium is
via mutagenesis. The hypothesis is that carcinogenicity can be induced directly by reduced
forms of chromium interacting with DNA to form adducts and crosslinks that can lead to DNA
breaks and mutations, and indirectly by free radical species generated during the reduction
process that can also lead to DNA breakage and mutagenesis.
Trivalent chromium is the ultimate product of the intracellular reduction of hexavalent
chromium. Trivalent chromium is capable of interacting directly with DNA, forming stable
coordination complexes with nucleic acids and peptides (Salnikow and Zhitkovich, 2008). In
particular, trivalent chromium is capable of forming ternary complexes with DNA and an
intracellular reducer, such as ascorbate, GSH, or cysteine (Zhitkovich et al., 1996; Salnikow et
al., 1992), as well as crosslinking DNA and proteins, and forming intrastrand DNA-DNA
crosslinks (Zhitkovich, 2005; Voitkun et al., 1998). These chromium-DNA complexes, as well
as DNA-protein and DNA-DNA crosslinks, have the capability of causing DNA single- and
double-strand breaks, which, if not adequately repaired, can lead to cell death, or if misrepaired,
can result in mutation.
Thus, once inside the cell, hexavalent chromium, through reduction to its pentavalent,
tetravalent, and trivalent forms, is capable of inducing a wide range of mutagenic and genotoxic
damage, including the formation of DNA adducts, DNA-protein and DNA-DNA crosslinks,
mutations, DNA single- and double-strand breaks, abasic sites, oxidized DNA bases,
chromosomal aberrations, sister chromatid exchanges, and micronuclei.
Key events.
1. Cellular uptake of hexavalent chromium. The first key event in hexavalent chromium-
induced carcinogenesis is cellular uptake of hexavalent chromium. Hexavalent
chromium readily enters cells via nonspecific sulfate and phosphate transporters, which
occurs due to the structural similarity of hexavalent chromium to these tetrahedral anions
(Bridges and Zalups, 2005). If hexavalent chromium is reduced before entering the cell,
very little chromium will be taken up, as cells are relatively impermeable to trivalent
chromium (Standeven and Wetterhahn, 1989).
2. Intracellular reduction of hexavalent chromium. Once inside the cell, hexavalent
chromium quickly undergoes a series of reduction reactions to yield pentavalent,
tetravalent, and ultimately the thermodynamically stable trivalent chromium. These
reactions are enabled by abundant nonenzymatic reductants within the cell, primarily
GSH, ascorbate, and cysteine (reviewed in McCarroll et al., 2009).
3. DNA damage via reduced chromium species. Following this intracellular reduction,
several possible mechanisms leading to mutagenicity can occur, since the products of
hexavalent chromium reduction within the cell (pentavalent, tetravalent, and trivalent

chromium) have all been shown to be DNA reactive (O’Brien et al., 2003). Hexavalent
chromium is reduced by GSH to yield pentavalent chromium and thiyl radicals, which
can react with other thiol molecules to produce superoxide radicals. Both pentavalent
and tetravalent chromium can participate in Fenton reactions, generating hydroxyl
radicals (Salnikow and Zhitkovich, 2008; Volko et al., 2006). Hence, the next key event
is the direct interaction of the reduced forms of chromium with DNA, leading to DNA
single- and double-strand breaks, base modifications, lipid peroxidation, and overall
genomic instability, which can lead to mutations if not adequately repaired.
4. Apoptosis and clonal expansion of mutated cells. Apoptosis induced by hexavalent
chromium exposure is initiated by several pathways, both genotoxic and nongenotoxic, to
eliminate the damaged cells from the population. In the process, cells that are resistant to
apoptosis (due either to mutations caused by hexavalent chromium or pre-existing
mutations) are selected for, allowing clonal expansion of cells that are capable of evading
apoptosis.
4.7.3.2. Experimental Support for the Hypothesized Mode of Action

Strength, consistency, and specificity of association. A large database of experimental
data exists on the mutagenic activity of hexavalent chromium compounds (these results are
summarized in Section 4.4.1 and in the corresponding tables). In vitro, positive results were
found in the majority of tests performed on hexavalent chromium compounds in bacterial test
systems (see Table 4-21). Similarly, in yeast (S. cerevisiae and S. pombe), all available studies
described positive results for the detection of gene mutations, mitotic gene conversion, and
mitotic crossing over.
In mammalian cell lines and primary cells, all studies using whole cells in vitro yielded
positive results (Table 4-22). Evidence of mutation induction was shown at the tk locus in the
mouse lymphoma assay, as well as at the HGPRT locus in Chinese hamster ovary cells (V79 and
AT3-2). In human cells, chromosome aberrations, DNA damage, and DNA-DNA and
DNA-protein crosslinks were detected in primary cultures and established cell lines originating
from target organs, including the gastric mucosa, bronchial epithelium, and fibroblasts from the
bronchial tubes and lung. Chromosome aberrations, sister chromatid exchanges, and DNA
damage were observed in primary human dermal fibroblasts and lymphocytes as well as
bronchial fibroblasts and epithelial cells. Chromosome aberrations and DNA damage were
found in mouse carcinogenic cell lines, and sister chromatid exchanges were detected in mouse
blastocysts. In rats, DNA damage and unscheduled DNA synthesis were observed in rat gastric
mucosal cells and hepatocytes as well as in primary lymphocytes, and transformation was
observed in rat liver epithelial cells upon exposure to hexavalent chromium. A number of
studies have been performed using cultured Chinese hamster ovary cells, showing chromosomal
aberrations and sister chromatid exchanges as well as DNA damage, DNA-protein crosslinks,

and induced DNA methylation, and three studies showed induced transformation in cultured
Syrian hamster embryo cells.
In vivo, most studies of the mutagenicity of hexavalent chromium compounds have
yielded positive results (Table 4-23). Somatic and germ cell mutations were detected in 3-day-
old D. melanogaster larvae fed potassium chromate, potassium dichromate, or calcium chromate
(Kaya et al., 2002; Spano et al., 2001; Amrani et al., 1999; Graf and Wurgler, 1996; Zimmering
et al., 1985). A number of in vivo oral exposure studies of the mutagenicity of hexavalent
chromium in mice and rats are available, with slightly differing results depending on the method
used. In the two studies in rats, Coogan et al. (1991) found DNA-protein crosslinks in liver and
not in splenic lymphocytes following 3- or 6-week exposures of 100 or 200 mg/L in drinking
water, but Mirsalis et al. (1996) did not find any evidence of DNA repair via unscheduled DNA
synthesis in rat hepatocytes following 48-hour exposures of up to 20 mg/L in drinking water or a
single gavage dose of 20 mL/kg at the same concentration. In other studies of mice exposed via
gavage, DNA damage, as measured by the comet assay, was found in peripheral leukocytes
(including isolated lymphocytes), stomach, colon, liver, kidney, bladder, lung, and brain (Wang
et al., 2006; Devi et al., 2001; Sekihashi et al., 2001), but neither DNA damage nor micronuclei
were found in bone marrow (De Flora et al., 2006; Sekihashi et al., 2001; Shindo et al., 1989).
Similarly, in studies of mice exposed via drinking water, De Flora et al. (2008, 2006) reported
negative results for the detection of micronuclei in the bone marrow of pregnant Swiss albino
mice and in the fetal polychromatic erythrocytes after exposures up to 20 mg/L and also in adult
BDF1 mice following 500 mg/L exposure for 210 days.
Interestingly, NTP (2007) investigated micronuclei induction in male mouse bone
marrow following a 3-month drinking water exposure and found differing results depending on
the strain of mouse used. In one phase of the study, results were negative in B6C3F1 mice
exposed to doses as high as 349 mg/L, while in another phase, following exposures of 0, 21.8,
43.6, or 87.2 mg/L hexavalent chromium, results were negative in BALB/c mice, equivocal in
B6C3F1 mice, and significantly positive at ≥43.6 mg/L exposures in am3-C57BL/6 mice, with a
statistically significant positive trend starting at 21.8 mg/L.
Somatic and germ cell mutations were detected in D. melanogaster treated
intraperitoneally with chromic acid or potassium dichromate (Rodriguez-Arnaiz and Martinez,
1986) or with sodium dichromate via filter paper (Rasmuson, 1985). Following parenteral
exposure in mice, DNA damage was detected in the stomach, colon, bladder, lung, brain, liver,
and kidney (Sekihashi et al., 2001; Ueno et al., 2001; Amlacher and Rudolph, 1981); mutations
were found in the liver of transgenic mice (Itoh and Shimada, 1998, 1997), in the germ cells of
hybrid male mice (Paschin et al., 1982), and in the offspring of exposed female mice (Knudsen,
1980); and micronuclei were increased in bone marrow and polychromatic erythrocytes (De
Flora et al., 2006; Wronska-Nofer et al., 1999; Itoh and Shimada, 1996; Hayashi et al., 1982;
Paschin and Toropzev, 1982; Wild, 1978), as well as in the liver and peripheral blood of mice

exposed prenatally (De Flora et al., 2006). In rats exposed parenterally, DNA damage was
detected in leukocytes (Patlolla and Tchounwou, 2006), and DNA-protein crosslinks were found
in lung, liver, and kidney (Tsapakos et al., 1983). Mutations were observed in the lung and
kidney from transgenic mice exposed intratracheally to hexavalent chromium (Cheng et al.,
2000); DNA-protein crosslinks and DNA fragmentation and adducts were found in the lung of
rats similarly exposed (Izzotti et al., 1998), while in rats exposed via inhalation, chromosomal
aberrations and sister chromatid exchanges were observed in peripheral lymphocytes (Koshi et
al., 1987).
In addition to the in vivo evidence in animals for the genotoxicity of hexavalent
chromium, several studies are available in humans (Table 4-24). In the only mutagenicity study
following oral doses, DNA-protein crosslinks were not detected in peripheral lymphocytes up to
4 hours after the four volunteers were given 71 µg hexavalent chromium/kg (Kuykendall et al.,
1996). Another study (Gao et al., 1994) failed to detect DNA damage in peripheral lymphocytes
of workers inhalationally exposed to 0.001–0.055 mg/m3. However, several studies of
occupational exposures via inhalation provide evidence of significant levels of chromium-
induced DNA damage (Gambelunghe et al., 2003), and the formation of micronuclei (Benova et
al., 2002; Vaglenov et al., 1999), chromosomal aberrations (Deng et al., 1988; Sarto et al., 1982),
and sister chromatid exchanges (Wu et al., 2001, 2000; Deng et al., 1988; Sarto et al., 1982;
Stella et al., 1982) in peripheral lymphocytes and/or buccal mucosal cells. These studies
detected genotoxicity in workers exposed to mean air concentrations as low as 0.0075 and
0.0249 mg/m3 (Benova et al., 2002). In addition, three studies found negative results for
micronuclei and sister chromatid exchange, but the exposure concentrations were not reported
(Nagaya et al., 1991; Sarto et al., 1990; Nagaya, 1986).
Dose-response concordance and temporal relationship. As noted above, hexavalent
chromium is hypothesized to induce carcinogenicity via a mutagenic mode of action. The initial
key events in the hypothesized mutagenic mode of action are the capability of the hexavalent
form of chromium to pass through the cell membrane and, once inside, to be reduced to
pentavalent, tetravalent, and trivalent chromium.
The available animal studies show that hexavalent chromium induces tumors in the
tongue, oral mucosa, and intestines of rodents (NTP, 2008). Studies of a human cohort in
Liaoning Province, China, exposed to 0.001–20 mg chromium/L in residential well water
(Beaumont et al., 2008; Zhang and Li, 1997, 1987) reported an excess risk of mortality from
stomach cancer in residents of the area. However, this risk has not been established in other
human populations exposed to drinking water contaminated with hexavalent chromium, and thus
the epidemiologic data are not sufficient to establish a casual association between exposure to
hexavalent chromium by ingestion and cancer.
NTP (2008) reported a statistically significant increase in the incidence of tumors of the
oral mucosa and tongue in rats exposed to hexavalent chromium for 2 years in drinking water at

average daily doses of 5.9 and 7.0 mg/kg-day for males and females, respectively, and tumors of
the small intestine in mice exposed to average daily doses of ≥2.4 and 3.1 mg/kg-day in males
and females, respectively. Correlating these data with mutagenicity testing by establishing
temporal and dose and/or site concordance can be difficult, as in vivo assays designed to detect
mutagenicity are conducted within a relatively short time after the exposure period has ended,
and tend to rely mainly on cells from tissues such as bone marrow and/or blood that are actively
replicating and therefore sensitive to mutagenic agents. There is evidence, however, that
hexavalent chromium can accumulate and induce mutagenicity in tissues at the site of entry and
systemically, at doses relevant to human exposures.
Following drinking water exposures, only one animal study has directly investigated
target tissue genotoxicity (De Flora et al., 2008). With regard to dose, the De Flora et al. (2008)
study tested levels (5 and 20 mg/L, or 1.2 and 4.82 mg/kg-day of hexavalent chromium) that
were just below those leading to murine intestinal (duodenum, jejunum, and ileum) tumors in the
2-year NTP study (30 and 50 mg/L for males and females, respectively). Negative results were
reported for DNA-protein crosslinks and DNA adducts when measuring the forestomach,
glandular stomach, and duodenum of mice exposed to hexavalent chromium for 9 months via
drinking water. However, the shorter study duration of DeFlora et al. (2008) makes a direct
comparison of these results to the duodenal tumors reported in the chronic NTP bioassay
infeasible.
Other studies have shown evidence of in vivo genotoxicity in nontarget tissues at early
time points following exposure. In three studies that used the comet assay to detect DNA
damage following gavage exposures in mice, Devi et al. (2001) found evidence of DNA damage
in leukocytes that peaked at 48 hours postexposure, Wang et al. (2006) detected DNA damage in
lymphocytes after 1- or 5-day consecutive exposures, and Seikihashi et al. (2001) detected DNA
damage in stomach, colon, liver, kidney, bladder, lung, and brain within 8 hours of dosing that
subsided by 24 hours posttreatment.
Devi et al. (2001) found positive dose-dependent results at >10-fold lower doses (0.21,
0.42, 0.84, 1.68, and 3.37 mg hexavalent chromium/kg). In fact, many of the positive in vivo
mutagenicity studies found a positive trend with dose, including oral exposures (Wang et al.,
2006; Devi et al., 2001) and parenteral exposures (Itoh and Shimada, 1996; Shindo et al., 1989;
Hayashi et al., 1982; Paschin and Toropzev, 1982; Knudsen, 1980; Wild, 1978) in rats (Patlolla
et al., 2008) and mice.
Therefore, the detection of DNA damage, a key event for the mutagenic mode of action
following oral exposure to hexavalent chromium, which exhibits dose-dependence and is
observed at time points prior to tumor development, strengthens the causal nature of this
association. Although DNA-protein crosslinks and DNA adducts were not detected in target
tissues following drinking water exposure in mice (De Flora et al., 2008), the lack of these

findings did not preclude the observation of mutations in other tissues and organs, considered to
be early events following hexavalent chromium exposure leading to carcinogenesis.
Biological plausibility and coherence. Mutagenicity as a mode of action for
carcinogenicity in humans is a biologically plausible mechanism for tumor induction.
Hexavalent chromium has been shown to be mutagenic in vitro and in vivo, across species and
tissue types. Human studies have shown induction of DNA damage, chromosomal aberrations,
and micronucleus induction following exposure to hexavalent chromium, and in vivo animal
studies show that hexavalent chromium induces DNA damage in rat blood, bone marrow, lung,
liver, and kidney, and in mouse blood, lung, liver, kidney, bladder, colon, and brain. Exposures
that induced a mutagenic response in these studies included doses within the range causing
tumors in rats and mice in a chronic exposure bioassay (NTP, 2008).
Only one study examined tumor target tissue for evidence of mutagenicity (De Flora et
al., 2008). De Flora et al. (2008) found negative results for DNA-protein crosslinks and DNA
adducts in the duodenum in mice following drinking water exposures. Other available drinking
water exposure studies of hexavalent chromium that measured mutagenicity in mice failed to
show evidence of micronucleus induction in the blood or bone marrow (De Flora et al., 2008,
2006; NTP, 2007; Mirsalis et al., 1996).
It has been postulated (De Flora et al., 2008) that the positive results for DNA damage
found in mice following gavage exposures (Wang et al., 2006; Devi et al., 2001; Sekihashi et al.,
2001) were the result of overwhelming the reductive capacity of the GI tract in mice, allowing
the accumulation and subsequent absorption of hexavalent chromium. This would indicate that
the comparatively lower concentrations of hexavalent chromium administered in the drinking
water studies (De Flora et al., 2008, 2006) are effectively reduced to trivalent chromium when
ingested, thereby inhibiting cellular uptake and subsequent DNA damage. While this is a
plausible explanation for the results following drinking water exposures, which are unusual in
that they represent the only component of the hexavalent chromium mutagenicity database that
does not show overwhelmingly positive results, there are inconsistencies with this explanation.
For example, although the doses administered in De Flora et al. (2008) were lower than those in
Wang et al. (2006) and Sekihashi et al. (2001), Devi et al. (2001) found positive results at doses
approximately sixfold lower than the lowest dose used by De Flora et al. (2008).
In addition, genetic differences have been implicated in predicting the severity of
genotoxic responses to hexavalent chromium exposure. In the 3-month NTP bioassay (2007),
three different strains of mice (B6C3F1, BALB/c, and am3-C57BL/6) were exposed to
hexavalent chromium in drinking water at concentrations of 21.8, 43.6, or 87.2 mg/L, and
different results for micronucleus induction in polychromatic erythrocytes were found among
strains. The BALB/c mice showed no micronucleus induction, and results in the B6C3F1 mice
were equivocal at the highest dose of 87.2 mg/L and in the trend test (p = 0.031). However, the
am3-C57BL/6 mice responded with a statistically significant overall positive trend, with the two

highest doses statistically significant, and the lowest dose nearly so. Based on the expected
reduction capacity of an average 50 g mouse, it does not appear that the reductive capacities
were overwhelmed in the NTP bioassay. The average rate of hexavalent chromium exposure for
all three strains of mice was estimated to have been 2.9 × 10-2 mg/hour at the highest dose (NTP,
2007). This rate is within the estimated reductive capacity of the mouse GI tract of 4.4 × 10-2
mg/hour that is based on an estimated 0.33 mL/hour rate of drinking water consumption.
However, the micronucleus results could reflect minor differences in the capacities of these three
strains of mice to reduce hexavalent chromium extracellularly, since the exact reductive capacity
of each mouse strain used is unknown.
The repair mechanisms in place for the resolution of DNA damage also appear to play a
role in the carcinogenicity of hexavalent chromium. While ER has been shown to prevent
hexavalent chromium-induced DNA damage (O’Brien et al., 2005), it has also been shown to be
responsible for the generation of DNA damage following exposures (Brooks et al., 2008).
Another DNA repair pathway important in resolving mismatched bases during DNA replication,
MMR, has recently been implicated in the genotoxic responses to hexavalent chromium
exposure. It has been shown that the processing of chromium-DNA adducts by the MMR
pathway is responsible for turning these lesions into frank DNA double-strand breaks (Peterson-
Roth et al., 2005). This group found that cells deficient in MMR were not subject to the same
toxic responses to hexavalent chromium as were cells with these repair processes intact. This
loss of MMR function leads to an unstable mutator phenotype, in which replication errors,
particularly those occurring in simple nucleotide repeat sequences known as microsatellites, are
not corrected, leading to an increase in mutation frequency (Loeb et al., 2008). Further, these
effects would be exacerbated by the physical and chemical interference with DNA replication
that occurs when trivalent chromium is present intracellularly (Eastmond et al., 2008).
There are several forms of cancer that exhibit microsatellite instability. For example,
microsatellite instability has been implicated as the cause of the majority of cases of hereditary
nonpolyposis colorectal cancer due to the inactivation of genes involved in the MMR pathway.
In an epidemiological study of chromate-exposed workers, microsatellite instability was reported
to occur in 79% of hexavalent chromium-induced lung tumors compared to only 15% in the
nonchromate lung cancer group (Hirose et al., 2002). The same group also reported finding
increased DNA methylation in the promoter region of the tumor suppressor gene p16 and the
MMR gene hMLH1 in human lung cancers in these chromate-exposed workers, indicating that
chromium can induce epigenetic effects (Kondo et al., 2006; Takahashi et al., 2005). These
findings reflect a loss of functional MMR capability that could be mechanistically involved in
chromate-induced lung cancer.
It was found that all four proteins responsible for MMR function were required for the
processing of chromium-DNA adducts into DNA double-strand breaks (Peterson-Roth et al.,
2005). The genes involved in MMR are known to be highly polymorphic in humans (Goode et

al., 2002), and given spontaneous background rates of mutation in human cells, it would not be
unexpected to find small populations of cells that have acquired mutations in one of these four
MMR genes. An inactivating mutation in any one of these would result in a growth advantage to
cells exposed to hexavalent chromium, allowing them to evade apoptotic responses to these
genotoxic lesions, as well as incurring further microsatellite instability, leading to a mutator
phenotype. Thus, a selective advantage upon chronic exposure to even low levels of hexavalent
chromium could translate into a clonal expansion of these MMR-deficient cells, leading to
further evasion of cell death and increasing mutation frequencies, resulting in a state of genomic
instability. This suggests that interindividual differences in the capacity and fidelity of DNA
repair processes could determine susceptibility to ingested hexavalent chromium.
In summary, DNA damage can occur following oral exposure to hexavalent chromium at
doses that should be within the reductive capacity of the organism. This DNA damage may be
repaired by error-prone mechanisms, resulting in DNA double-strand breaks and microsatellite
instability, and further exacerbated by both hexavalent chromium-induced epigenetic effects that
alter these DNA repair mechanisms and the interference of DNA replication processes by
hexavalent chromium. Genomic instability, or an increased rate of acquisition of genetic
alterations, may result not only in the form of microsatellite instability but also as chromosomal
instability and aneuploidy that have been shown to occur after prolonged exposure to hexavalent
chromium. Exposure also results in complex alterations in gene expression that can alter cell
survival pathways; this combined with apoptosis induced by both genotoxic and non-genotoxic
mechanisms induced by hexavalent chromium can lead to a deregulation of cellular proliferation,
resulting in the clonal expansion of cells that are resistant to apoptosis and eventually leading to
neoplastic transformation.
In addition, it is of note that among the available oral exposure studies in mice, all studies
that investigated DNA damage or micronucleus induction in bone marrow cells found negative
results, including the study by Sekihashi et al. (2001), which found DNA damage in every tissue
examined (liver, kidney, lung, brain, stomach, colon, and bladder) except for the bone marrow.
The reason for the negative findings in these assays is unknown, but the high turnover of cells in
the bone marrow may have allowed for more efficient repair of the damaged cells.
Bioavailability. As noted above, there is uncertainty surrounding the ability of
hexavalent chromium to induce mutagenicity and carcinogenicity in humans considering the
potential for reduced bioavailability. Intrinsic to the mutagenic and carcinogenic processes of
hexavalent chromium is its ability to reach relevant tissues prior to being reduced to pentavalent,
tetravalent, and trivalent chromium. When hexavalent chromium is reduced to the trivalent form
extracellularly, this reduction process effectively detoxifies hexavalent chromium, since trivalent
chromium is nearly impermeable to the cell.
Quantitative studies of GI absorption of hexavalent chromium in humans have estimated
that as much as 10% of an ingested dose of 5 mg is absorbed (Kuykendall et al., 1996),

indicating that not all hexavalent chromium is reduced by the gastric juices of the stomach. In
rats and mice, daily oral doses of 8 mg hexavalent chromium/day for 8 weeks resulted in
absorption and accumulation of chromium in the bone, spleen, liver, and kidney (Kargacin et al.,
1993); rats given 0.138 µmol hexavalent chromium/day for 3 days exhibited GI absorption of
about 16% (Febel et al., 2001), and the absorption of 4–10% of a single daily dose of 57 µg
hexavalent chromium (as Na51CrO4) was observed in rats, regardless of fasting state (MacKenzie
et al., 1959). Distribution studies have shown that hexavalent chromium, once absorbed,
distributes to nearly all tissues, particularly concentrating in the kidney, liver, bone, and RBCs.
Thus, at oral doses within human exposure ranges, hexavalent chromium was not completely
reduced by the GI tract, making available some portion of ingested hexavalent chromium to be
absorbed directly by the mucosal cells of the GI tract, or to be distributed to other tissues
throughout the body.
However, based on an understanding of chromium chemistry, as well as in vitro and in
vivo studies conducted by De Flora et al. (2008, 1997), the reduction of at least some portion of
ingested hexavalent chromium to trivalent chromium likely occurs in the GI tract (see
Chapter 3). No data are currently available on the capacity of the rodent stomach to reduce
hexavalent chromium. However, based on in vitro measurements, De Flora et al. (1997)
estimated that the reductive capacity of the human GI tract is sufficiently large to effectively
reduce even high doses of ingested hexavalent chromium to the less toxic trivalent form. Given
this assertion, it is appropriate to ask whether the observed effects at the doses employed in the
NTP (2008) study resulted from an exceedance of the reductive capacity of the rodent GI tract.
This is important because if the effects observed only occurred due to the reductive capacity of
the rodent GI tract being exceeded, these results may be less relevant to human risk at the lower
doses that humans are more likely to be exposed.
In discussing the results of the NTP (2008) study, the original NTP investigators, Stout et
al. (2009), specifically addressed this extracellular reduction issue. Qualitatively, Stout et al.
(2009) noted that, in the 2-year NTP study, the observed increases in neoplasms of the small
intestine of mice and the toxicity to the erythron, histiocytic infiltration, and uptake of hexavalent
chromium into the tissues of rats and mice suggested that, under the conditions of this study, at
least a portion of the administered hexavalent chromium was not reduced in the stomach.
Moreover, Stout et al. (2009) also pointed out the significant disparity in the oral toxicity and
carcinogenicity of hexavalent chromium versus trivalent chromium in rodents, including the
absence of increases in neoplasms or nonneoplastic lesions of the small intestine in rats or mice
exposed to chromium picolinate monohydrate, a trivalent chromium compound tested in an
earlier NTP bioassay. Stout et al. (2009) believe that these data provide additional evidence that
hexavalent chromium is not completely reduced in the stomach and is responsible for the
observed effects.

In addressing the De Flora et al. (2008) suggestion that increases in neoplasms of the
small intestine observed in mice are the result of a saturation of the gastric reduction capacity,
Stout et al. (2009) took a more quantitative approach. Stout et al. (2009) postulated that if the
threshold mechanism proposed by De Flora et al. (2008) actually existed, then the dose that
saturated the reduction capacity would likely represent an inflection point on a sublinear dose-
response curve, with doses above the inflection point demonstrating an increasing rate of
response per unit dose. To test this hypothesis, Stout et al. (2009) evaluated tissue concentration
and mouse small intestine neoplasm data for linearity and found that data that were statistically
nonlinear were supralinear (i.e., exhibited a decreasing rate of response per unit dose), which
does not support the presence of a reduction threshold.
Finally, De Flora et al. (1997) estimated the reductive capacity of human gastric juice to
be about 84–88 mg of hexavalent chromium/day. Similar data are not available for the reductive
capacity of mouse gastric juice. However, Stout et al. (2009) assumed that hexavalent chromium
reduction is equally effective in mice and humans and that gastric secretion scales across species
by body weight3/4. Then, they estimated the reductive capacity of the gastric juice from a 50-g
mouse to be approximately 0.4 mg/day (8 mg/kg-day). Stout et al. (2009) then pointed out that
this value is greater than all of the male mouse doses and is nearly equivalent to the average daily
dose of hexavalent chromium in the high-dose group of female mice in the NTP (2008) study.
Therefore, Stout et al. (2009) concluded from their analysis that the neoplasms in the small
intestine of mice occurred at dose levels that did not exceed the estimated hexavalent chromium
reduction capacity of the gastric juices in mice.
4.7.3.3. Other Possible Modes of Action

In the carcinogenic process, aberrant cell survival, proliferation, and tissue remodeling
are known contributors to the etiology of cancer (Hanahan and Weinberg, 2000). Evidence of
diffuse duodenal hyperplasia in mice in all exposure groups was observed in the 3-month NTP
(2007) study. The sites where hyperplasia was observed correlated with the site of tumors
observed in the 2-year bioassay (NTP, 2008). One mechanism of cellular proliferation known to
occur following exposures to xenobiotic agents involves that of toxicity causing cellular death
and consequent regenerative cellular proliferation, comprising a potential mode of action for
carcinogenesis. However, the study by NTP noted that no evidence of tissue damage or necrosis
was observed in these animals, and most of the available studies of hexavalent chromium-
induced genetic damage observed genotoxicity at doses below those inducing cytotoxicity,
particularly in studies showing dose-dependent genetic damage.
Another acquired capability during tumor development is resistance to apoptosis, a
hallmark of most if not all types of cancer (Hanahan and Weinberg, 2000). Apoptosis has been
shown to occur following exposure to hexavalent chromium exposure (Flores and Perez, 1999;
Ye et al., 1999; Singh et al., 1998). It is possible that the apoptotic cell death occurring after

hexavalent chromium exposure, initiated by ROS damage, altered cell signaling pathways, and
genotoxic damage, occurs at levels that would not result in visible pathology or a regenerative
response, but are significant enough to have an impact on the balance between cell survival and
death. It has been proposed that in cells exposed to hexavalent chromium, apoptosis occurring in
response to DNA damage, oxidative stress, or damage to mitochondria may serve as another key
event in the carcinogenesis of hexavalent chromium, in that it allows for selection of cells that
are resistant to apoptosis due to mutation and provides a means of clonal expansion for these
cells (Nickens et al., 2010). DNA damage leads to the induction of cell cycle checkpoints to
assess and repair the genetic damage; if the damage is too severe to repair, the cell will be
targeted for cell death. This removes a potentially mutagenic cell from the population, but other
cells deemed sufficient for repair could still exist and incur mutations following error-prone
repair processes. In this manner, cells that have mutations enabling them to elude apoptosis,
whether they were the result of hexavalent chromium mutagenesis or pre-existing, are conferred
a growth advantage. Cells may also avoid targeted death due to changes in gene expression that
lead to upregulation of pro-inflammatory and/or anti-apoptotic genes. These processes could be
temporally similar to those of DNA damage and mutation, and may serve to lay the groundwork
for the acquisition of other carcinogenic traits, including uncontrolled cell growth, leading to
tumor formation. Therefore, rather than an alternate mode of action per se, apoptosis induced by
hexavalent chromium exposure is considered here to be a key event in the carcinogenic process.
4.7.3.4. Conclusions About the Hypothesized Mode of Action

As noted above, hexavalent chromium is hypothesized to be carcinogenic by a mutagenic
mode of action. The key events in the hypothesized mutagenic mode of action are the uptake of
hexavalent chromium into the cell followed by intracellular reduction to pentavalent, tetravalent,
and trivalent chromium. These reduced forms of hexavalent chromium and the free radicals that
are formed during the reduction process are capable of directly interacting with cellular
components, giving rise to mutagenicity (including DNA adduct formation, DNA damage, gene
mutations, chromosomal aberrations, and micronuclei formation). Considering the database,
there is evidence that hexavalent chromium can accumulate and induce mutagenicity in various
tissues throughout the body at doses relevant to human exposures and, for oral exposures, within
the reductive capacity of the GI tract.

(1) Is the hypothesized mode of action sufficiently supported in the test animals? The
experimental evidence that hexavalent chromium is mutagenic, as presented in
Section 4.4.1, includes multiple adverse genetic effects including DNA adduct
formation, DNA damage, gene mutations, chromosomal aberrations, and the
formation of micronuclei. In addition to the evidence supporting a mutagenic mode
of action in test animals, alternative or additional hypothesized modes of action for
hexavalent chromium carcinogenicity have not been demonstrated.
(2) Is the hypothesized mode of action relevant to humans? Mutagenicity is a well-

established cause of carcinogenicity. The evidence discussed above demonstrates
that hexavalent chromium is a mutagen in bacteria, yeast, cultured rodent and human
cells, fruit flies, mice, and rats, supporting the presumption that it could also be a
mutagen in humans. Moreover, several studies of exposed workers provide direct
evidence of DNA damage by hexavalent chromium. In conclusion, the weight of
evidence supports a mutagenic mode of action for hexavalent chromium
carcinogenicity.
(3) Which populations or lifestages can be particularly susceptible to the hypothesized

mode of action? The mutagenic mode of action is considered relevant to all
populations and lifestages. According to U.S. EPA’s Supplemental Guidance (U.S.
EPA, 2005b), there may be increased susceptibility to early-life exposures for
carcinogens with a mutagenic mode of action. Therefore, because the weight of
evidence supports a mutagenic mode of action for hexavalent chromium
carcinogenicity and in the absence of chemical-specific data to evaluate differences in
susceptibility, early-life susceptibility should be assumed and the age-dependent
adjustment factors (ADAFs) should be applied, in accordance with the Supplemental
Guidance. In addition, individuals with genetic polymorphisms conveying
deficiencies in DNA repair capacity may have increased susceptibility to hexavalent
chromium carcinogenicity.
4.7.3.5. Mutagenic Across All Routes of Exposure

As summarized previously, following inhalation exposures, hexavalent chromium has
been shown to induce lung tumors in a number of human occupational studies, as well as tumors
at or near the site of entry in animal studies. Evidence also exists, however, that ingested
hexavalent chromium can reach the systemic circulation and affect tissues beyond those at or
near the site of entry. In addition to hexavalent chromium toxicity in the lungs, it can be
absorbed by the lung when inhaled and can then enter systemic circulation. Consistent with this
evidence, DNA damage, micronucleus induction, and sister chromatid exchanges have been
observed in circulating peripheral lymphocytes from workers exposed to inhalation
concentrations as low as 7.5 and 24.9 µg/m3 (Benova et al., 2002), and for durations of 4 months
to 14 years (Gambelunghe et al., 2003), 0.5–18 years (Stella et al., 1982), 2–>20 years (Benova
et al., 2002), or 4–25 years (Vaglenov et al., 1999). These studies indicate that, while tumor
incidence following inhalation exposure to hexavalent chromium occurs primarily in the lungs,
hexavalent chromium also has the capacity to damage DNA in other tissues at timepoints and
concentrations relevant to human exposures.

EPA has concluded that hexavalent chromium is carcinogenic by a mutagenic mode of
action. Considering the available oral and inhalation evidence for mutagenicity and subsequent
carcinogenicity and that these events are capable of occurring in all cells, this mode of action is
considered to be applicable to all routes of exposure.
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES

4.8.1. Possible Childhood Susceptibility
No studies are available that address the possible adverse effects of hexavalent chromium
in children. However, there is evidence that hexavalent chromium may act through a mutagenic
mode of action. In accordance with the Supplemental Guidance (U.S. EPA, 2005b), the
mutagenic mode of carcinogenic action for hexavalent chromium would indicate an increased
carcinogenic susceptibility for early-life exposures. In addition, developmental toxicity also is of
concern due to the mutagenicity of hexavalent chromium and the possibility for genetic damage
to the germ cells of the F1 generation that could be transmitted to the F2 generation. The
reproductive and developmental toxicity studies that have been conducted employing hexavalent
chromium suggest that the developing fetus may be a target of toxicity, as well as male and
female reproductive organs, which may result in a reduction in fertility.
4.8.2. Possible Gender Differences

The extent to which men and women differ in susceptibility to hexavalent chromium is
unknown. However, animal data exist that imply a difference between males and females in
their response to ingestion of hexavalent chromium. For example, in the NTP (2008) study, at
the highest concentration administered (516 mg/L), female rats exhibited a higher incidence of
tumors of the oral cavity than male rats (i.e., 11/48 [23%] vs. 7/50 [14%], respectively). The
biological significance of this finding at lower doses and for other species, including humans, is
unknown.

5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)

5.1.1. Choice of Principal Study and Critical Effect—with Rationale and Justification
Two types of studies are available that provide information on the toxicological effects of
ingested chromium in humans. The first type of study provides evidence of acute human health
effects in individuals who accidentally or intentionally ingested high (fatal or near-fatal) doses of
hexavalent chromium. The second type of study provides evidence of chronic human health
effects (primarily cancer) in populations exposed unintentionally to food or drinking water
containing high levels of hexavalent chromium over an extended time period. Because both
types of studies provide little information on dose-response relationships and because the second
type of study is primarily concerned with cancer as an outcome, these available human data are
not useful for quantifying the risk of noncancer effects resulting from chronic exposure to
hexavalent chromium.
In animals, the effects of subchronic oral exposure to hexavalent chromium have been
evaluated in rats (NTP, 2007; Quinteros et al., 2007; Rafael et al., 2007; Acharya et al., 2001;
Chopra et al., 1996; Vyskocil et al., 1993) and mice (NTP, 2007; Asmatullah and Noreen, 1999),
and the effects of chronic oral exposure to hexavalent chromium have been evaluated in rats
(NTP, 2008; MacKenzie et al., 1958), mice (NTP, 2008), and dogs (Anwar et al., 1961). In
particular, the subchronic and chronic studies conducted by NTP (2008, 2007) provide the most
useful dose-response data on the noncancer effects of oral hexavalent chromium exposure
because of their comprehensive assessments of numerous toxicological endpoints at multiple
dose levels. A number of other studies of reproductive and developmental toxicity of hexavalent
chromium have been conducted in rats, mice, and rabbits, but typically at higher doses and for
shorter durations than the NTP (2008, 2007) studies. All of these animal studies are summarized
in Table 4-26.
Results from the NTP (2007) subchronic (i.e., 90-day) study identified several hexavalent
chromium-induced noncancer effects, including hematological effects, hepatotoxicity, alterations
in lipid metabolism, and histopathological changes in GI tissues and pancreatic and mesenteric
lymph nodes. The most sensitive hexavalent chromium-induced noncancer effects were
microcytic, hypochromic anemia, increased serum liver enzyme activities, and histopathological
changes to the duodenum and pancreatic lymph nodes in rats; and histopathological changes in
the duodenum in mice. In the 2-year toxicology and carcinogenicity study by NTP (2008), the
most sensitive noncancer effects identified were histopathological changes to the liver,
duodenum, and mesenteric lymph nodes in rats; and in the duodenum, mesenteric lymph nodes,
and liver in mice. LOAELs of 1.7–3.1 mg hexavalent chromium/kg-day were identified by EPA

in the subchronic NTP (2007) study, and LOAELs of 0.24–0.77 mg hexavalent chromium/kg-
day were identified by EPA in the chronic NTP (2008) study.
Other subchronic and chronic oral exposure studies of hexavalent chromium compounds
do not provide suitable data for identifying points of departure (PODs) for RfD derivation
because comprehensive toxicological evaluations were not conducted in these studies. In
addition, interpretation of results from these studies was compromised because of the small
number of animals evaluated, lack of a dose-response relationship, or inadequate reporting of
results (see Table 4-26). Where LOAELs were identified based on examination of a limited set
of endpoints (e.g., Acharya et al., 2001; Chopra et al., 1996), the LOAELs were higher than
those identified in the chronic NTP (2008) bioassay.
Studies of reproductive and developmental toxicity indicate that hexavalent chromium
exposure can affect reproductive organs, increase pre- and postnatal implantation loss, and cause
reduced fetal weight and fetal abnormalities. In general, the NOAELs or LOAELs associated
with reproductive and developmental effects are higher than those identified in the subchronic
and chronic toxicity studies summarized in Table 4-26.
Thus, based on the comprehensive examination of endpoints and measurement of
sensitive endpoints of toxicity, the bioassays by NTP (2008, 2007) were deemed the best
candidates for use in deriving an oral RfD for hexavalent chromium. Specifically, five studies,
three subchronic (i.e., one in rats and two in mice) (NTP, 2007) and two chronic (i.e., one in rats
and one in mice) (NTP, 2008), were identified as candidate principal studies. The key results
from these five studies are summarized below.
5.1.1.1. Subchronic Studies

NTP (2007) 90-day studies in rats and mice
In F344/N rats, sodium dichromate dihydrate was administered in drinking water to
groups of males and females at five different concentrations for 90 days. Based on average
water consumption rates, the mean effective doses of hexavalent chromium were estimated by
NTP to be 0, 1.7, 3.5, 5.9, 11.2, and 20.9 mg/kg-day for both males and females. Results of this
study identified a LOAEL in male and female rats of 1.7 mg hexavalent chromium/kg-day; a
NOAEL was not identified because effects were observed at the lowest dose tested. This
LOAEL was based on observations of microcytic, hypochromic anemia, increased serum liver
enzyme activities, and histopathological changes to pancreatic lymph nodes (in males) and
histopathological changes to the duodenum (in females) at daily doses ≥1.7 mg hexavalent
chromium/kg-day.
In B6C3F1 mice, groups of males and females were exposed to sodium dichromate
dihydrate in drinking water for 90 days. Based on water consumption monitored throughout the
study, NTP calculated average daily doses over the 90-day treatment duration of approximately
0, 3.1, 5.3, 9.1, 15.7, and 27.9 mg hexavalent chromium/kg-day for both males and females.

Based on histopathological changes (histiocytic cellular infiltration) in the duodenum in both
sexes, a LOAEL of 3.1 mg hexavalent chromium/kg-day was identified for male and female
mice; a NOAEL was not identified because the effects observed were at the lowest dose tested.
In a comparative 90-day drinking water study in male B6C3F1, BALB/c, and
am3-C57BL/6 mice, groups of each strain were exposed to three different concentrations of
sodium dichromate dihydrate. Based on water consumption and body weights monitored
throughout the study, NTP calculated average daily doses over the 90-day treatment duration of
approximately 0, 2.8, 5.2, or 8.7 mg hexavalent chromium/kg-day for all strains. At the end of
the study, similar effects were observed in all three strains. A LOAEL of 2.8 mg hexavalent
chromium/kg-day was identified based on histopathological changes in the duodenum in B6C3F1
mice (histiocytic cellular infiltration and diffuse epithelial hyperplasia), BALB/c mice
(histiocytic cellular infiltration), and am3-C57BL/6 mice (diffuse epithelial hyperplasia); a
NOAEL was not identified because effects seen were at the lowest dose tested.
5.1.1.2. Chronic Studies

NTP (2008) 2-year studies in rats and mice
In F344/N rats, groups of 50 males and females were administered sodium dichromate
dihydrate in drinking water at four different concentrations for 2 years. Based on measured
water consumption rates and body weights in rats, NTP estimated that male rats received time-
weighted average doses of hexavalent chromium of 0.21, 0.77, 2.1, or 5.9 mg/kg-day, while
female rats received 0.24, 0.94, 2.4, or 7.0 mg/kg-day of hexavalent chromium. This study
identified NOAEL and LOAEL values for noncancer effects in male rats of 0.21 and 0.77 mg
hexavalent chromium/kg-day, respectively, based on increased incidences of nonneoplastic
histopathological changes to the liver (basophilic foci), duodenum (histiocytic cellular infiltrate),
and mesenteric lymph nodes (histiocytic cellular infiltrate and hemorrhage). In female rats, a
LOAEL for noncancer effects of 0.24 mg hexavalent chromium/kg-day was identified based on
the increased incidence of chronic inflammation of the liver (observed in all treatment groups); a
NOAEL was not identified because effects observed were at the lowest dose tested.
In B6C3F1 mice, groups of 50 males and females were administered sodium dichromate
dihydrate in drinking water at four different concentrations for 2 years. Based on measured
amounts of water consumption and body weights in mice, NTP estimated that male mice
received average doses of hexavalent chromium of 0.38, 0.91, 2.4, or 5.9 mg/kg-day, while
female mice received 0.38, 1.4, 3.1, or 8.7 mg/kg-day of hexavalent chromium. This study
identified a LOAEL for noncancer effects of 0.38 mg hexavalent chromium/kg-day in both male
and female B6C3F1 mice; a NOAEL value was not identified because effects seen were at the
lowest dose administered. In males, the LOAEL was based on increased incidences of
histopathological changes to the duodenum (diffuse epithelial hyperplasia) and mesenteric lymph
nodes (histiocytic cellular infiltration); in females, the LOAEL was based on increased

incidences of histopathological changes to the duodenum (diffuse epithelial hyperplasia),
mesenteric lymph nodes (histiocytic cellular infiltration), liver (histiocytic cellular infiltration),
and pancreas (depletion of cytoplasmic zymogen granules).
The NTP (2008) study was of chronic duration (i.e., 2 years), involved the use of multiple
dose groups, and included a comprehensive evaluation of multiple endpoints. Also, this bioassay
used lower doses than the subchronic (90-day) studies also conducted by NTP (2007), and thus
provided dose-response information at lower exposure levels than the 90-day studies.
Additionally, the chronic NTP (2008) study was more sensitive, yielding lower LOAELs than
the subchronic studies. Thus, the chronic NTP (2008) study was selected as the principal study.
As indicated, NTP (2008) observed several hexavalent chromium-induced noncancer
effects in their chronic studies in rats and mice. Based on a comparison of LOAELs in rats and
mice (see Table 4-26), the lowest LOAELs were observed for the following seven effects:
(1) Chronic liver inflammation in female rats,
(2) Histiocytic cellular infiltration in the liver of female mice,
(3) Diffuse epithelial hyperplasia in the duodenum of male mice,
(4) Diffuse epithelial hyperplasia in the duodenum of female mice,
(5) Histiocytic cellular infiltration in the mesenteric lymph nodes of male mice
(6) Histiocytic cellular infiltration in the mesenteric lymph nodes of female mice, and
(7) Cytoplasmic cellular alteration of acinar epithelial cells in the pancreas of female
mice.
All of these effects occurred at the lowest doses tested (i.e., 0.24 mg/kg-day in female
rats and 0.38 mg/kg-day in male and female mice), and were considered as possible critical
effects for derivation of the RfD for hexavalent chromium. The incidences of these seven effects
across all treatment groups in NTP (2008) are shown in Table 5-1.

Table 5-1. Incidence data for lesions in female F344/N rats and male and
water for 2 years
Dose
(mg hexavalent chromium/kg-d)
Endpoint 0 0.24 0.94 2.4 7.0
Female rats
Liver: chronic inflammation 12/50 21/50a 28/50b 35/50b 39/50b
Dose
0 0.38 0.91 2.4 5.9
Male mice
Duodenum: diffuse epithelial hyperplasia 0/50 11/50b 18/50b 42/50b 32/50a
Mesenteric lymph node: histiocytic cellular infiltration 14/47 38/47b 31/49b 32/49b 42/46a
Dose
0 0.38 1.4 3.1 8.7
Female mice
Duodenum: diffuse epithelial hyperplasia 0/50 16/50b 35/50b 31/50b 42/50b
Mesenteric lymph node: histiocytic cellular infiltration 3/46 29/48b 26/46b 40/50b 42/50b
Liver: histiocytic cellular infiltration 2/49 15/50b 23/50b 32/50b 45/50b
Pancreas: acinus, cytoplasmic alteration 0/48 6/50a 6/49a 14/50b 32/50b
a
b
Source: NTP (2008).
5.1.2. Methods of Analysis—Including Models (PBPK, BMD, etc.)

To determine the specific endpoint for use in derivation of the RfD, all available
dichotomous models in U.S. EPA’s Benchmark Dose Software (BMDS), version 1.4.1, were fit
to the incidence data for the seven selected endpoints (see Table 5-1) in female rats and male and
female mice administered sodium dichromate dihydrate in drinking water for 2 years (NTP,
2008). The incidence data employed in the benchmark dose (BMD) modeling of these seven
endpoints also are shown in Table 5-1. Doses (i.e., the benchmark dose [BMD10] and the 95%
lower confidence limit on the benchmark dose [BMDL10]) associated with a benchmark response
(BMR) of 10% extra risk were estimated by each model. In accordance with U.S. EPA’s
Benchmark Dose Technical Guidance Document (U.S. EPA, 2000b), a BMR of 10% is generally
used in the absence of information regarding what level of change is considered biologically
significant, and also to facilitate a consistent basis of comparison across assessments.
Details of the BMD modeling conducted for each endpoint presented in Table 5-1 are
provided in Appendix B. In general, model fit was assessed by a chi-square goodness-of-fit test
(i.e., models with p < 0.1 failed to meet the goodness-of-fit criterion) and the Akaike’s

Information Criterion (AIC) value (i.e., a measure of the deviance of the model fit that allows for
comparison across models for a particular endpoint). Of the models exhibiting adequate fit, the
model yielding the lowest AIC value was selected as the best-fit model (as long as the BMDL
estimates across the models exhibiting adequate fit were “sufficiently close”). If more than one
model shared the lowest AIC, BMDL10 values from these models were averaged to obtain a POD
(U.S. EPA, 2000b).
For chronic liver inflammation in female rats with all dose groups included, only the log-
logistic model provided an adequate fit, as assessed by the chi-square goodness-of-fit statistic,
yielding BMD10 and BMDL10 values of 0.22 and 0.14 mg hexavalent chromium/kg-day,
respectively. For diffuse epithelial hyperplasia in the duodenum of male mice with all dose
groups included, none of the dichotomous models in BMDS provided an adequate fit to the data
(i.e., χ2 p-value < 0.1). After dropping the high-dose group, the gamma, log-logistic, multistage,
log-probit, quantal linear, and Weibull models provided adequate fits to the data (i.e., χ2 p-value
≥ 0.1). As assessed by comparing AIC values, the multistage and quantal linear models provided
the best fit, yielding BMD10 and BMDL10 values of 0.16 and 0.13 mg hexavalent chromium/kg-
day, respectively. For diffuse epithelial hyperplasia in the duodenum of female mice with all
dose groups included, none of the dichotomous models in BMDS provided an adequate fit to the
data (i.e., χ2 p-value < 0.1). Only after dropping the two highest dose groups was an adequate fit
achieved for any model. In this instance, all of the dichotomous models in BMDS, except the
logistic and probit models, provided an adequate fit to the data (i.e., χ2 p-value ≥ 0.1). As
assessed by comparing the AIC values, the best fit was provided by several models (i.e., gamma,
multistage, quantal linear, and Weibull), yielding BMD10 and BMDL10 values of 0.12 and 0.09
mg hexavalent chromium/kg-day, respectively. For histiocytic cellular infiltration in the liver of
female mice with all dose groups included, only the log-logistic model provided an adequate fit
to the data (i.e., χ2 p-value ≥ 0.1), yielding BMD10 and BMDL10 values of 0.17 and 0.12 mg
hexavalent chromium/kg-day, respectively. For cytoplasmic alteration of acinar epithelial cells
of the pancreas in female mice, all of the dichotomous models in BMDS provided adequate fits
to the data (i.e., χ2 p-value ≥ 0.1). As assessed by comparing AIC values, the log-logistic model
produced the best fit, yielding BMD10 and BMDL10 values of 0.68 and 0.52 mg hexavalent
chromium/kg-day, respectively. Finally, for lesions of the mesenteric lymph nodes (i.e.,
histiocytic cellular infiltration) in both male and female mice, none of the available dichotomous
models in BMDS provided adequate fits to the data, even with the two highest doses dropped
from the analysis; thus, data sets for these lesions were considered to be unsuitable for BMD
modeling. Therefore, the LOAEL of 0.38 mg hexavalent chromium/kg-day for histiocytic
cellular infiltration of the mesenteric lymph nodes in male and female mice serves as the
candidate POD for this endpoint.

A summary of this BMD modeling information is presented in Table 5-2, and further
details of this modeling are contained in Appendix B-1.
Table 5-2. Summary of BMD10 and BMDL10 from the best fitting models for
lesions of the liver, duodenum, mesenteric lymph nodes, and pancreas in
female rats and male and female mice after exposure to sodium dichromate
dihydrate in drinking water for 2 years (NTP, 2008)
Number BMDa BMDLa

Endpoint Species/sex Model of doses (mg/kg-d) (mg/kg-d)
Liver: chronic Rat/female Log-logistic 5 0.22 0.14
inflammation
Duodenum: diffuse Mouse/male 1-Degree polynomial 4 0.16 0.13
epithelial hyperplasia multistage/quantal linear
Mesenteric lymph Mouse/male – – – –
node: histiocytic
cellular infiltrationb
Duodenum: diffuse Mouse/female Gamma/multistage/quantal 3 0.12 0.09
epithelial linear/Weibull
hyperplasia
Mesenteric lymph Mouse/female – – – –
node: histiocytic
cellular infiltrationb
Liver: histiocytic Mouse/female Log-logistic 5 0.17 0.12
cellular infiltration
Pancreas: acinus, Mouse/female Log-logistic 5 0.68 0.52
cytoplasmic
alteration
a
BMDs and BMDLs from dichotomous data are associated with a 10% extra risk; doses are in terms of mg
hexavalent chromium/kg-d.
b
None of the models provided an adequate fit to the data.
BMDL = lower confidence limit (95%) on the BMD
Source: ATSDR (2008).
The lowest BMDL10 value of 0.09 mg hexavalent chromium/kg-day, based on the

selection of the incidence of diffuse epithelial hyperplasia of the duodenum in female mice as the
critical effect, was identified as the POD from which to derive the RfD for hexavalent chromium.
As indicated in Section 4, due to its morphological similarity to adenoma, focal epithelial
hyperplasia was classified as a preneoplastic lesion by NTP (2008), and so the possibility exists
that diffuse epithelial hyperplasia may also represent a preneoplastic lesion. However, even
though this possibility exists and thus this lesion may progress to cancer (i.e., adenoma) in some
cases, the EPA considers the selection of this critical effect on which to base the derivation of the
RfD (a noncancer endpoint) to be appropriate because definitive data on the progression of this
particular lesion do not currently exist.

5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs)
The following UFs were applied to the POD of 0.09 mg/kg-day, based on the incidence
of diffuse epithelial hyperplasia of the duodenum in female mice from NTP (2008), to derive the
RfD for hexavalent chromium.
• A UF of 10 was used to account for uncertainty in extrapolating from laboratory

animals to humans (i.e., interspecies variability) because information was
unavailable to quantitatively assess toxicokinetic or toxicodynamic differences
between animals and humans.
• A UF of 10 was used to account for variation in susceptibility among members of

the human population (i.e., interindividual variability) because information is
unavailable to predict potential variability in human susceptibility.
• A UF was not needed to account for extrapolation from subchronic-to-chronic

exposure because a chronic study was used to derive the chronic RfD.
• A UF for LOAEL to NOAEL extrapolation was not used because the current
approach is to address this extrapolation as one of the considerations in selecting a
BMR for BMD modeling. In this case, a BMR represented by a 10% extra risk of
diffuse epithelial hyperplasia was selected under an assumption that it represents a
minimal biologically significant change.
• A UF of 1 was used to account for database deficiencies. The toxicity of ingested

hexavalent chromium has been extensively examined in a range of animal
toxicology studies. The database for oral toxicity includes a chronic drinking water
study in rats and mice, a chronic drinking water study in rats, a subchronic drinking
water study in rats and mice, and a number of reproductive/developmental toxicity
studies in monkeys, rabbits, rats, and mice. The reproductive toxicity database
includes a continuous breeding study (NTP, 1997), in which F0 and F1 generation
animals were exposed to hexavalent chromium in the diet, and the offspring of
F1 animals were evaluated on PND 21.
For this assessment, the RfD of 0.0009 or 9 × 10-4 mg/kg-day for hexavalent chromium
was derived by dividing the BMDL10 (or POD) of 0.09 mg/kg-day by a composite uncertainty
factor of 100 (10 for extrapolation from animals to humans and 10 for human variability).
5.1.4. Previous RfD Assessment

The previous RfD assessment for hexavalent chromium 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 hexavalent chromium (as potassium chromate) at a
dose of 2.5 mg/kg-day (MacKenzie et al., 1958). No toxicity was reported in these animals at
this dose, resulting in identification of a NOAEL of 2.5 mg/kg-day, the only dose administered in

the study, as the POD. 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, 1987)
were applied to this POD to yield an oral RfD of 3 × 10-3 mg/kg-day.
5.2. UNCERTAINTIES IN THE ORAL REFERENCE DOSE

The following discussion identifies uncertainties associated with the RfD for hexavalent
chromium. As presented above, an RfD of 9 × 10-4 mg/kg-day was derived based on the
incidence of diffuse epithelial hyperplasia of the duodenum in female mice from a 2-year
drinking water study (NTP, 2008). UFs were applied to the POD, a BMDL10 generated through
BMD modeling. Factors accounting for uncertainties associated with a number of steps in the
analyses were adopted to account for extrapolating from an animal bioassay to humans with
varying susceptibilities.
An adequate range of animal toxicology data is available for the hazard assessment of
hexavalent chromium via ingestion, as described previously in Chapter 4. The database of oral
toxicity studies includes a chronic drinking water study in rats and mice, a chronic drinking
water study in rats, a subchronic drinking water study in rats and mice, and several
reproductive/developmental toxicity studies in monkeys, rabbits, rats, and mice. Toxicity
associated with oral exposure to hexavalent chromium is observed in the liver, GI tract, and
reproductive organs, with the liver and GI tract being the most sensitive target organs.
Consideration of the available dose-response data to determine an estimate of oral
exposure that is likely to be without an appreciable risk of adverse health effects over a lifetime
led to the selection of the 2-year drinking water study in F344/N rats and B6C3F1 mice (NTP,
2008) and increased incidence of diffuse epithelial hyperplasia in the duodenum of female mice
as the principal study and critical effect, respectively, for deriving the RfD for hexavalent
chromium.
The selection of the BMD model for identifying the POD does not lead to significant
uncertainties since benchmark effect levels were within the range of the experimental data.
However, the selected models do not represent all possible models one might fit, and other
models could be selected to yield more extreme results, both higher and lower than those
included in this assessment.
Animal-to-human extrapolation yields further uncertainties. The effect and the
magnitude of this effect associated with the dose at the POD in mice are extrapolated to humans.
Pharmacokinetic models are useful to examine species differences in pharmacokinetic
processing; however, dosimetric adjustment using pharmacokinetic modeling was not possible
for the toxicity observed following oral exposure to hexavalent chromium. Information was
unavailable to quantitatively assess toxicokinetic or toxicodynamic differences between animals

and humans. Accordingly, a 10-fold UF was used to account for uncertainty in extrapolating
from laboratory animals to humans in the derivation of the RfD.
Heterogeneity among humans is another area of uncertainty. In the absence of
hexavalent chromium-specific data on variation in human response, a factor of 10 was used in
the derivation of the RfD. Human variation may be larger or smaller than this 10-fold factor;
however, hexavalent chromium-specific data to examine the potential magnitude of over- or
underestimation are unavailable.
5.3. ORAL CANCER ASSESSMENT

5.3.1. Choice of Study/Data—with Rationale and Justification
Several epidemiology studies have examined the association between oral exposure to
environmental hexavalent chromium and cancer in populations that resided near sources of
industrial waste containing hexavalent chromium compounds, including studies of populations in
Liaoning Province, China (Kerger et al., 2009; Beaumont et al., 2008; Zhang and Li, 1997, 1987,
1980), Kings County/San Bernardino County, California (Fryzek et al., 2001), Nebraska (Bednar
and Kies, 1991), and Glasgow, United Kingdom (Eizaguirre-Garcia et al., 2000, 1999). The
Liaoning Province studies provide some evidence of an excess risk of mortality from stomach
cancer; however, because of various limitations, including limited characterization of exposure,
the Liaoning Province studies are not considered adequate for dose-response analysis.
The NTP rodent bioassay, in which F344/N rats and B6C3F1 mice were administered
sodium dichromate dihydrate, a hexavalent chromium compound, in drinking water for 2 years
(NTP, 2008), was selected as the basis for deriving the oral cancer slope factor (CSF) for
hexavalent chromium. This bioassay was selected for dose-response assessment because it is a
well-conducted lifetime animal study of hexavalent chromium carcinogenicity via ingestion (see
detailed summary of the study in Section 4.2.2). No other adequate studies of hexavalent
chromium carcinogenicity by ingestion are available.
5.3.2. Dose-Response Data

The dose-response data considered in the derivation of the CSF for hexavalent chromium
were the incidences of benign and malignant tumors in rat oral mucosa and mouse small intestine
observed in the NTP (2008) bioassay.
Incidence data for neoplastic lesions of the oral cavity in male and female rats exposed to
sodium dichromate dihydrate in drinking water for 2 years are summarized in Table 4-15.
Neoplasms observed in the oral cavity of treated rats were squamous cell carcinoma of the oral
mucosa (both sexes), squamous cell papilloma of the oral mucosa (males only), squamous cell
carcinoma of the tongue (both sexes), and squamous cell papilloma of the tongue (both sexes).
The incidences of squamous cell carcinoma of the oral mucosa (13.6%) and of combined
squamous cell papilloma or carcinoma (15.7%) of the oral mucosa were statistically significantly

increased (at p < 0.05) in male rats treated with 5.9 mg/kg-day hexavalent chromium (the highest
dose tested) compared with controls. The incidences of squamous cell carcinoma of the oral
mucosa (23.9%) and of combined squamous cell carcinoma of the oral mucosa or tongue
(23.9%) were statistically significantly increased (at p < 0.05) in female rats treated with 7.0 mg
hexavalent chromium/kg-day (the highest dose tested) compared with controls. The incidences
of other neoplastic lesions of the oral cavity were not statistically significantly increased in any
treatment group in male or female rats compared with controls, although the incidence of
squamous cell carcinoma of the oral mucosa in female rats in the penultimate (2.4 mg/kg-day)
dose group (4.6%) exceeded that of historical controls (i.e., 0/300 in drinking water studies;
5/1,400 [0.4%] by all routes of exposure). Other neoplasms observed in treated rats included
pancreatic acinar adenomas and benign pheochromocytomas in males and mononuclear cell
leukemias in females (see Table 4-16); however, the incidence of these neoplasms did not exhibit
dose-dependence. Thus, NTP (2008) concluded that evidence of a relationship between
neoplastic changes in tissues other than the oral cavity and exposure to sodium dichromate
dihydrate was equivocal. In summary, exposure of rats to sodium dichromate dihydrate in
drinking water for 2 years resulted in a significant increase in squamous epithelial neoplasms of
the oral mucosa and tongue at the highest exposure levels (average daily doses of 5.9 and 7.0 mg
hexavalent chromium/kg-day in males and females, respectively), but not at the three lower
exposure levels. The incidences of squamous cell papillomas or carcinomas in the oral cavity of
male and female F344/N rats exposed to sodium dichromate dihydrate in drinking water for
2 years in the NTP (2008) study are presented in Table 5-3 (for male rats) and Table 5-4 (for
female rats).

Table 5-3. Incidences of squamous cell papillomas or carcinomas in the oral
cavity of male F344/N rats exposed to sodium dichromate dihydrate in
Sodium dichromate dihydrate Estimated daily intake of hexavalent

concentration chromiuma Incidence of squamous cell
(mg/L) (mg/kg-d) papillomas or carcinomasb
0 0 0/50 (0%)
14.3 0.21 1/50 (2%)
57.3 0.77 0/49 (0%)
172 2.1 0/50 (0%)
516 5.9 7/49 (14.5%)c
a
Intakes were reported by NTP (2008) based on drinking water intakes and mean body weights observed during the
study.
b
Number of animals with lesion/number of animals examined. Incidence estimates include all animals that were
examined for oral tumors unadjusted for survival.
c
Statistically significantly elevated above control at p < 0.05 using Fisher’s exact test.
Source: NTP (2008).
Table 5-4. Incidences of squamous cell papillomas or carcinomas in the oral

cavity of female F344/N rats exposed to sodium dichromate dihydrate in
Sodium dichromate dihydrate Estimated daily intake of hexavalent

concentration chromiuma Incidence of squamous cell
(mg/L) (mg/kg-d) papillomas or carcinomasb
0 0 1/50 (2%)
14.3 0.24 1/50 (2%)
57.3 0.94 0/50 (0%)
172 2.4 2/50 (4%)
516 7.0 11/50 (22%)c
a
Intakes were reported by NTP (2008) based on drinking water intakes and mean body weights observed during the
study.
b
Number of animals with lesion/number of animals examined. Incidence estimates include all animals that were
examined for oral tumors unadjusted for survival.
c
Source: NTP (2008).
Also from the NTP (2008) study, incidence data for neoplastic lesions of the small
intestine in male and female B6C3F1 mice exposed to sodium dichromate dihydrate in drinking
water for 2 years are summarized in Table 4-19. In male mice, statistically significant increases
(p < 0.05) were observed in the incidences of adenomas or carcinomas combined in the small
intestine (duodenum, jejunum, and ileum) at hexavalent chromium doses ≥2.4 mg/kg-day (i.e., at
the two highest doses tested). Furthermore, significant positive trends were observed in the
incidences of duodenal adenomas, duodenal carcinomas, jejunal adenomas, small intestine

adenomas, small intestine carcinomas, and small intestine adenomas or carcinomas combined in
male mice. In female mice, statistically significant increases (p < 0.05) were observed in the
incidences of duodenal adenomas, small intestine adenomas, and small intestine adenomas or
carcinomas combined at hexavalent chromium doses ≥3.1 mg/kg-day (i.e., at the two highest
doses tested). Furthermore, significant positive trends were observed in the incidences of
duodenal adenomas, duodenal carcinomas, jejunal adenomas, small intestine adenomas, and
small intestine adenomas or carcinomas combined in female mice. No other statistically or
biologically significant increases in neoplasms were observed in other tissues.
In summary, exposure of B6C3F1 mice to sodium dichromate dihydrate in drinking water
for 2 years resulted in statistically significant increases in the incidences of neoplasms of the
small intestine in males and females at hexavalent chromium doses ≥2.4 and ≥3.1 mg/kg-day,
respectively. The incidences of adenomas and carcinomas combined in the small intestine of
male and female B6C3F1 mice exposed to sodium dichromate dihydrate in drinking water for
2 years are summarized in Tables 5-5 and 5-6, respectively. In evaluating the tumor incidences
in rats and mice, the mouse was determined to be the most sensitive species because tumor
incidences were statistically significantly elevated at lower doses and a greater response was
exhibited by the mice at the two highest doses. Therefore, the mouse tumor incidence data were
used as the basis for the oral CSF derived employing BMD modeling.
Table 5-5. Incidences of adenomas and carcinomas combined in the small

intestine of male B6C3F1 mice exposed to sodium dichromate dihydrate in
Sodium dichromate dihydrate Estimated daily intake of

concentration hexavalent chromiuma Incidence of adenomas or
(mg/L) (mg/kg-d) carcinomasb
0 0 1/49 (2%)
14.3 0.38 3/49 (6.1%)
28.6 0.91 2/49 (4.1%)
85.7 2.4 7/50 (14%)c
257.4 5.9 20/48 (41.7%)c
a
Intakes were reported by NTP (2008) based on drinking water intakes and mean body weights observed during
the study.
b
Calculated from reported percentages of mice with adenomas or carcinomas. Incidence estimates included all
animals that were examined for intestinal tumors and survived for at least 451 days. In each of the control and
first two dose groups, one animal died prior to day 451. In the high-dose group, two animals died prior to
day 451. None of these animals were found to have intestinal adenomas or carcinomas at the time of death.
c
Source: NTP (2008).

Table 5-6. Incidences of adenomas and carcinomas combined in the small
intestine of female B6C3F1 mice exposed to sodium dichromate dihydrate
in drinking water for 2 years
Sodium dichromate dihydrate Estimated daily intake of

concentration hexavalent chromiuma Incidence of adenomas or
(mg/L) (mg/kg-d) carcinomasb
0 0 1/49 (2%)
14.3 0.38 1/50 (2%)
57.3 1.4 4/49 (8.2%)
172 3.1 17/49 (34.7%)c
516 8.7 22/49 (44.9%)c
a
Intakes were reported by NTP (2008) based on drinking water intakes and mean body weights observed during
the study.
b
Calculated from reported percentages of mice with adenomas or carcinomas. Incidence estimates included all
animals that were examined for intestinal tumors and survived for at least 451 days. In all of the dose groups
except the low-dose group, one animal died prior to day 451. None of these animals were observed to have
intestinal adenomas or carcinomas at the time of death.
c
Source: NTP (2008).
5.3.3. Dose Adjustments and Extrapolation Method(s)

U.S. EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend
that the method used to characterize and quantify cancer risk from a chemical be determined by
what is known about the mode of action of that chemical, and how this mode of action impacts
the shape of the dose-response curve at low doses. According to the Cancer Guidelines, the dose
response is generally considered to be linear in the low-dose range when evidence supports a
mutagenic mode of action for a chemical because of its DNA reactivity and direct mutagenic
activity. A linear low-dose extrapolation approach was used to estimate human carcinogenic risk
associated with hexavalent chromium exposure due to the mutagenic mode of action of this
chemical.
In order to derive an oral CSF, BMD modeling was carried out using U.S. EPA’s BMDS
(U.S. EPA, 2000b). U.S. EPA’s BMDS offers several possible mathematical dose-response
functions for use with dichotomous data including logistic, gamma, Weibull, quantal linear,
probit, and multistage models. For this assessment, EPA relied on the results obtained from the
multistage model only, as this is the model preferred by the Agency for conducting cancer dose-
response assessments. In applying the BMD approach to the derivation of a CSF, the standard
procedure is to calculate a lower 95% confidence bound on the dose corresponding to the BMR,
where the BMR is typically set at 10% extra risk. This lower confidence bound is referred to as
the BMDL10. The CSF is then calculated by dividing the BMR by the BMDL10 and then
converting this slope value to human equivalents.

In estimating the CSF, the incidence of neoplasms in the small intestine of mice was
employed, as this species was deemed to be more sensitive than the rat. Only animals that
survived for at least 451 days, the time until appearance of the first tumor, were considered at
risk for tumor development. Consequently, the incidence estimates included all animals that
were examined for intestinal tumors and survived for at least 451 days (see Tables 5-5 and 5-6).
The BMD modeling results for the incidence of neoplasms in the small intestine of male and
female mice are shown in Appendix B-2. For male mice, the two-stage multistage model
exhibited the best fit to the data yielding a slope of 0.09 (mg/kg-day)-1. For female mice, the
two-stage multistage model also exhibited the best fit to the data yielding a slope of 0.10 (mg/kg-
day)-1.
In order to estimate an oral CSF, these slopes were converted to human equivalents. For
this conversion, body weight to the ¾ power scaling was used, where the time-weighted average
male and female mouse body weights of controls (i.e., 50 and 53 g, respectively) were employed,
along with an assumed human body weight of 70 kg. The mouse body weights were taken from
the NTP (2008) study report. The following equation was then used to convert the slopes
derived from the BMD modeling to oral CSFs expressed in human equivalents:
Slope × (WH/WA)0.25 = CSF
where
WH = animal body weight (kg)
WA = human body weight (kg)
Using the above equation, the CSFs resulting from the fitting of the two-stage multistage
model in BMDS to the incidence of neoplasms in the small intestine of male or female mice
were 0.5 and 0.6 (mg/kg-day)-1, respectively, expressed in human equivalents.
5.3.4. Oral Slope Factor

The CSF values based on the incidence of small intestine tumors in male and female mice
are similar (i.e., 0.5 [mg/kg-day]-1 for males and 0.6 [mg/kg-day]-1 for females). Given the
poorer fit of the multistage model to the female mouse data, a CSF estimate based on the male
mouse data was considered to be associated with less uncertainty. Therefore, the CSF of
0.5 (mg/kg-day)-1, based on the incidence of neoplasms in the small intestine of male mice, was
selected as the most appropriate CSF for hexavalent chromium.
5.3.5. Application of ADAFs

Because a mutagenic mode of action for hexavalent chromium carcinogenicity is
sufficiently supported in laboratory animals and is relevant to humans (see Section 4.6.3.4), and
in the absence of chemical-specific data to evaluate differences in age-specific susceptibility,

increased early-life susceptibility to hexavalent chromium is assumed and ADAFs should be
applied, as appropriate, in accordance with the Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b). The oral slope
factor of 0.5 (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, 2005b). 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 hexavalent
chromium.
To illustrate the use of the ADAFs established in the Supplemental Guidance for
Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b), sample
calculations are presented for three exposure duration scenarios, including full lifetime, assuming
that the exposure rate to hexavalent chromium remains constant at an average daily dose of
0.0001 mg hexavalent chromium/kg-day (Table 5-7). This average daily dose of 0.0001 mg
hexavalent chromium/kg-day is being used here for illustrative purposes only to demonstrate
how to apply ADAFs. In practice, actual exposure information specific to the situation under
consideration should be used.
Table 5-7. Application of ADAFs for a 70-year exposure to 0.0001 mg

hexavalent chromium/kg-day from ages 0 to 70
Slope factor Average daily dose Duration

Age group ADAF (per mg/kg-d) (mg/kg-d) adjustment Partial risk
0–<2 yrs 10 0.5 0.0001 2 yrs/70 yrs 1 × 10-5
2–<16 yrs 3 0.5 0.0001 14 yrs/70 yrs 3 × 10-5
≥16 yrs 1 0.5 0.0001 54 yrs/70 yrs 4 × 10-5
Total risk 8 × 10-5
Note that the partial risk for each age group is the product of the values in columns 2–5
(e.g., 10 × 0.5 × 0.0001 × 2/70 = 0.00001 for exposures from age 0 to <2 years), and the total
risk is the sum of the partial risks. Thus, a 70-year risk estimate for a constant average daily
dose of 0.0001 mg/kg-day starting at birth is 0.00008 or 8 × 10-5.
If calculating the cancer risk for a 30-year exposure to a constant average daily dose of
0.0001 mg hexavalent chromium/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 0.00001, 0.00003, and 0.00001,
resulting in a total risk estimate of 0.00005 or 5 × 10-5.
If calculating the cancer risk for a 30-year exposure to a constant average daily dose of
0.0001 mg hexavalent chromium/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 0.00002, resulting in a
total risk estimate of 0.00002 or 2 × 10-5.
5.3.6. Uncertainties in Cancer Risk Values

As in most risk assessments, extrapolation of data from experimental animals to estimate
potential lifetime cancer risks to human populations from exposure to hexavalent chromium
yields uncertainties. Some of these uncertainties can be evaluated for their quantitative impact
on the final result, while for others, only their qualitative impact can be assessed. The principal
uncertainties in the assessment of the cancer risk from exposure to hexavalent chromium are
summarized below in Table 5-8, and discussed in more detail in the following text.

Table 5-8. Summary of uncertainties in the cancer risk assessment for
hexavalent chromium
Consideration/ Impact on oral slope

approach factor Decision Justification
Low-dose Alternatives could ↓ or Multistage model A linear-low-dose extrapolation approach was used
extrapolation ↑ CSF by an unknown used to determine to estimate human carcinogenic risk associated with
procedure extent POD, linear low- hexavalent chromium exposure consistent with a
dose extrapolation mutagenic mode of carcinogenic action.
from POD
Cross-species Alternatives could ↓ or BW3/4 (default In the absence of hexavalent chromium-specific
scaling ↑ CSF (e.g., sixfold ↓ approach) information on interspecies differences in
[scaling by BW] or ↑ toxicokinetics, the default scaling factor of BW3/4
twofold [scaling by was used to calculate equivalent cumulative
BW2/3]) exposures for estimating equivalent human risks
(U.S. EPA, 1992).
Statistical ↓ CSF 25% if BMDL (default Size of bioassay results in sampling variability;
uncertainty at maximum likelihood approach for lower bound is 95% CI on administered dose.
POD estimation (i.e., calculating
BMD10) used rather reasonable upper
than lower bound bound CSF)
(BMDL10) for POD
Species/gender Human risk could ↓ or Male mouse tumors It was assumed that humans are as sensitive as the
combination ↑, depending on (adenomas or most sensitive rodent gender/species tested; true
relative sensitivity carcinomas of the correspondence is unknown. The carcinogenic
small intestine) response occurs across species. Generally, direct
site concordance is not assumed; consistent with
this view, some human tumor types are not found in
rodents and rat and mouse tumor types also differ.
Human relevance Lack of human Tumors with Hexavalent chromium is judged to be carcinogenic
of rodent tumor relevance of tumor significant dose- through a mutagenic mode of action and is a
data data would ↓ CSF response considered multisite carcinogen in rodents; therefore, the
for estimating carcinogenicity observed in rodent studies is
potential human assumed to be relevant to human exposure.
cancer response
Human Low-dose risk ↑ or ↓ to Considered No data are available to support the range of human
population an unknown extent qualitatively variability/sensitivity to hexavalent chromium.
variability in
metabolism and
response/
sensitive
subpopulations
Choice of low-dose extrapolation approach. The mode of action is a key consideration in

clarifying how risks should be estimated for low-dose exposure. A linear, low-dose
extrapolation approach was used to estimate human carcinogenic risk associated with hexavalent
chromium exposure consistent with a hypothesized mutagenic mode of carcinogenic action of
hexavalent chromium (U.S. EPA, 2005a).

The multistage model was used to model the tumor incidence data because this is the
model preferred by the Agency for conducting cancer dose-response assessments; however, it is
unknown how well this model or the linear low-dose extrapolation predicts low-dose risks for
hexavalent chromium. The selected model does not represent all possible models one might fit,
and other models could conceivably be selected to yield more extreme results consistent with the
observed data, both higher and lower than those included in this assessment.
Cross-species scaling. The default cross-species scaling factor (BW3/4) was applied to
address toxicological equivalence of internal doses between rodent species and humans,
consistent with the 2005 Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a).
Because it is unknown whether there are differences in the pharmacokinetic pathways in animals
and humans following hexavalent chromium exposure, it is not possible to estimate the
magnitude of the uncertainty in the use of this default beyond that associated with other choices
for default cross-species scaling factors (e.g., BW2/3 or BW1).
Statistical uncertainty at the POD. Measures of statistical uncertainty require assuming
that the underlying model and associated assumptions are valid for the data under consideration.
For the multistage model applied to the incidence of male mice GI tract tumors, there is a
reasonably typical degree of uncertainty at the 10% extra risk level (the POD for linear low-dose
extrapolation). That is, the BMDL10 for male mice is approximately 25% lower than the BMD10.
Choice of species/gender. The oral CSF for hexavalent chromium was quantified using
the tumor incidence data for mice, which were thought to be more sensitive than rats to the
carcinogenicity of hexavalent chromium. While tumor responses in the mouse were higher than
those of rats at a comparable dose level, suggesting greater sensitivity of the mouse, it is
unknown whether this higher sensitivity would be maintained at lower exposures.
Relevance to humans. The Guidelines for Carcinogen Risk Assessment (U.S. EPA,
2005a) state that site concordance is not a prerequisite for evaluating the implications of animal
study results for humans. Chemicals that are mutagenic and cause tumors at multiple sites in
animals are likely relevant to human carcinogenesis. Hexavalent chromium is thought to be
carcinogenic through a mutagenic mode of action and is a multisite carcinogen in rodents.
Considering all of the available information, the carcinogenicity observed in rodent studies is
considered relevant to human exposure. In addition, the concordance of the alimentary system
tumors across rats and mice lends strength to the concern for human carcinogenic potential.
Human population variability. The extent of inter-individual variability in response to
hexavalent chromium is unknown. Although a mutagenic mode of action would indicate
increased early-life susceptibility, the data exploring whether there is differential sensitivity to
hexavalent chromium carcinogenicity across life stages are unavailable. This lack of
understanding about potential differences in metabolism and susceptibility across exposed
human populations thus represents a source of uncertainty. The uncertainties associated with this
lack of data and knowledge about human variability can, at present, only be considered in

qualitative terms; however, EPA has developed ADAFs to quantitatively account for some of the
potential differences in age-dependent response to carcinogens with a mutagenic mode of action.
ADAFs are to be applied to the CSF for hexavalent chromium when assessing cancer risks in
exposed populations composed of individuals <16 years old (U.S. EPA, 2005b). More specific
guidance in applying these ADAFs was provided in Section 5.3.5.
5.3.7. Previous Cancer Assessment

The previous IRIS assessment for hexavalent chromium was posted to the IRIS database
in 1998. In that assessment, EPA concluded that the oral carcinogenicity of hexavalent
chromium could not be determined (and was thus classified as Group D) because no data were
located in the available literature that suggested that hexavalent chromium is carcinogenic by the
oral route of exposure. Therefore, no oral CSF was derived.

6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE
6.1. HUMAN HAZARD POTENTIAL

Hexavalent chromium compounds are a group of substances that contain chromium in the
hexavalent or +6 oxidation state. As a class, hexavalent chromium compounds are strong
oxidizing agents, and thus, it is rare to find hexavalent chromium naturally occurring in the
environment because it is readily reduced to trivalent chromium (i.e., chromium in the +3
oxidation state) by organic matter. However, hexavalent chromium compounds released to the
environment by anthropogenic sources may persist in natural waters and soils that contain low
amounts of organic matter. Major uses or former uses of hexavalent chromium compounds
include metal plating, manufacture of pigments and dyes, corrosion inhibitors, chemical
synthesis, refractory production, leather tanning, and wood preservation. Individuals may be
exposed to hexavalent chromium compounds through ingestion of drinking water or contact with
soils or other media contaminated with these substances.
Toxicokinetic studies in humans, mice, and rats have examined the absorption,
distribution, metabolism, and elimination of hexavalent chromium compounds. Hexavalent
chromium can be absorbed via oral, inhalation, or dermal routes of exposure in humans and
laboratory animals. For this toxicological review, however, the focus is on the toxicokinetics of
hexavalent chromium following ingestion. Once ingested, hexavalent chromium compounds can
interact with endogenous fluids and other organic matter in the GI tract, resulting, to some
extent, in the reduction of hexavalent chromium to trivalent chromium. This process, whereby
hexavalent chromium is reduced to trivalent chromium in the GI tract, is termed “extracellular”
reduction. The extent of absorption of ingested hexavalent chromium appears to be determined
by both the solubility of the hexavalent chromium compound ingested and how rapidly
hexavalent chromium is reduced to trivalent chromium in the GI tract, as trivalent chromium
does not diffuse readily across cell membranes. Hexavalent chromium can easily cross cell
membranes due to its ability to use existing nonspecific sulfate and phosphate anion transport
mechanisms.
Ingested hexavalent chromium is distributed throughout the body. Liver, kidney, spleen,
and bone are the primary sites of chromium distribution. Once inside the cell, hexavalent
chromium is reduced to trivalent chromium, either enzymatically or nonenzymatically. This
process is called “intracellular” reduction to distinguish it from the extracellular process
described above. This intracellular reduction yields such reactive intermediates as chromium(V)
and chromium(IV). These reactive intermediates, along with oxygen radicals generated during
this intracellular reduction, can indirectly damage DNA. In addition, trivalent chromium, the

final product of the intracellular reduction of hexavalent chromium, can form adducts with a
number of macromolecules, including DNA.
Hexavalent chromium is eliminated primarily in the urine as trivalent chromium.
Chromium can also be eliminated in hair, nails, and breast milk. There does not appear to be a
gender difference in the toxicokinetics of hexavalent chromium, and inter-individual variability
in the presystemic reduction and subsequent absorption and elimination may be primarily driven
by differences in gastric contents and intervals between meals.
Two PBPK models have been developed for hexavalent and trivalent chromium in rats
and humans (O’Flaherty et al., 2001; O’Flaherty, 1996, 1993). The inclusion of trivalent
chromium in the model allows for the use of trivalent chromium exposure time course data to aid
in parameterization of chromium elimination and to evaluate the ability of the model to predict
elimination of hexavalent chromium as trivalent chromium. However, neither the rat nor human
version of the model in its present form has been subjected to formal computerized optimization
of parameter values.
Two types of studies provide information on the toxicological effects in humans resulting
from exposure to ingested hexavalent chromium. In the first type of study, acute human health
effects have been observed following oral ingestion of hexavalent chromium in individuals
accidentally or intentionally ingesting high (fatal or near-fatal) doses of hexavalent chromium.
These studies are not particularly useful for establishing dose-response relationships. In the
second type of study, chronic human health effects have been reported in human populations
exposed unintentionally to elevated levels of hexavalent chromium in food or drinking water
over an extended time period. Human studies of possible associations between oral exposure to
hexavalent chromium and cancer are limited to a few epidemiology studies in which health
outcomes (primarily cancer) were evaluated among populations that were exposed to drinking
water contaminated with hexavalent chromium in Liaoning Province, China (Kerger et al., 2009,
Beaumont et al., 2008; Zhang and Li, 1997, 1987), Kings County/San Bernardino County,
California (Fryzek et al., 2001; Bick et al., 1996), Nebraska (Bednar and Kies, 1991), and
Glasgow, United Kingdom (Eizaguirre-Garcia et al., 2000, 1999). Analyses of data collected
from the JinZhou area of Liaoning Province, China, where groundwater, surface water, and
agricultural soils were heavily contaminated with chromium derived from hexavalent chromium
production (e.g., 0.001–20 mg chromium/L in residential well water), provide evidence of an
excess risk of mortality from stomach cancer from 1970 to 1978 in residents of the area, relative
to the reference populations in the province (four other areas in Liaoning Province, and the total
population of the province) (Beaumont et al., 2008). EPA concluded that the exposure-response
analyses presented by Zhang and Li (1997), Beaumont et al. (2008), and Kerger et al. (2009) are
not based on the quality of data that is needed to support a conclusion regarding the presence or
absence of a dose-response among the observed cancer rates in these villages. The other
epidemiologic studies did not find a significant correlation between hexavalent chromium

concentrations in drinking water (or proximity to the source of hexavalent chromium soil
contamination) and cancer.
In animals, the effects of subchronic oral exposure to hexavalent chromium have been
evaluated in rats (NTP, 2007; Quinteros et al., 2007; Rafael et al., 2007; Acharya et al., 2001;
Chopra et al., 1996; Vyskocil et al., 1993) and mice (NTP, 2007; Asmatullah and Noreen 1999),
and the effects of chronic oral exposure to hexavalent chromium have been evaluated in rats
(NTP, 2008, MacKenzie et al., 1958), mice (NTP, 2008; Borneff et al., 1968), and dogs (Anwar
et al., 1961). The studies conducted by the NTP (2008, 2007) provide dose-response data on the
effects of oral hexavalent chromium exposure based on a comprehensive assessment of
toxicological endpoints. The EPA used NTP (2008, 2007) to identify LOAELs and NOAELs in
rats and mice for subchronic and chronic exposure durations. Results from the NTP (2007)
subchronic study identified several hexavalent chromium-induced noncancer effects, including
hematological effects, hepatotoxicity, alterations in lipid metabolism, and histopathological
changes in GI tissues and pancreatic and mesenteric lymph nodes. The most sensitive
hexavalent chromium-induced noncancer effects in rats were microcytic, hypochromic anemia,
increased serum liver enzyme activities, and histopathological changes to the duodenum and
pancreatic lymph nodes; in mice, the most sensitive noncancer effect was histopathological
changes in the duodenum. The most sensitive noncancer effects in the NTP (2008) 2-year
toxicology and carcinogenicity study were histopathological changes to the liver, duodenum, and
mesenteric lymph nodes in rats; and in the duodenum, mesenteric lymph nodes, and liver in
mice.
A number of animal studies have also evaluated the reproductive/developmental toxicity
of hexavalent chromium via the oral route of exposure. Collectively, these studies provide
evidence that oral exposure to hexavalent chromium compounds produces reproductive effects,
including histopathological changes to reproductive organs in males (Aruldhas et al., 2006, 2005,
2004; Li et al., 2001; Chowdhury and Mitra, 1995; Zahid et al., 1990) and females (Murthy et al.,
1996); alterations in sperm, including decreased count, decreased motility, and abnormal
morphology (Subramanian et al., 2006; Yousef et al., 2006; Li et al., 2001; Zahid et al., 1990);
decreased plasma testosterone levels (Yousef et al., 2006; Chowdhury and Mitra, 1995);
increased estrous cycle length (Kanojia et al., 1998, 1996; Murthy et al., 1996); changes in
mating behavior and decreased fertility in males (Bataineh et al., 1997); and adverse
reproductive outcomes, including decreased numbers of live fetuses and implantations, and
increased numbers of resorptions and pre- and postimplantation losses (Bataineh et al., 2007;
Elsaieed and Nada, 2002; Kanojia et al., 1998, 1996; Elbetieha and Al-Hamood, 1997; Junaid et
al., 1996a, b, 1995; Trivedi et al., 1989). Developmental effects observed have included
decreased fetal weight and length (Elsaieed and Nada, 2002; Kanojia et al., 1998; Junaid et al.,
1996a, b, 1995; Trivedi et al., 1989); external (subdermal hemorrhage and tail malformations)
and skeletal abnormalities (decreased ossification) (Elsaieed and Nada, 2002; Kanojia et al.,

1998, 1996; Junaid et al., 1996a, b, 1995; Trivedi et al., 1989); and delayed sexual maturation
and function in female offspring (Banu et al., 2008; Al-Hamood et al., 1998). In contrast to
results of the above studies, effects were not observed in dietary exposure studies conducted by
NTP that investigated the potential for hexavalent chromium to produce effects on male
reproductive organs in rats and mice (NTP, 1996a, b) and on reproductive outcomes in a
continuous breeding study in mice (NTP, 1997). The reasons for these inconsistent results are
not readily apparent, as daily dose ranges evaluated in the NTP studies overlapped with those
used in other studies showing hexavalent chromium-induced adverse reproductive effects. The
most sensitive noncancer effects observed in the 2-year chronic study by NTP (2008), in general,
occurred at lower doses than the reproductive or developmental effects.
In regard to carcinogenic effects, exposure of rats to sodium dichromate dihydrate in
drinking water for 2 years resulted in a significant increase in squamous epithelial neoplasms of
the oral mucosa and tongue at the highest exposure level (average daily doses of 5.9 and 7.0 mg
hexavalent chromium/kg-day in males and females, respectively), but not at the three lower
exposure levels (NTP, 2008). Exposure of B6C3F1 mice to sodium dichromate dihydrate in
drinking water for 2 years resulted in significant increases in the incidences of neoplasms of the
small intestine in males and females at doses ≥2.4 and ≥3.1 mg hexavalent chromium/kg-day,
respectively. NTP (2008) concluded that results from these studies provide clear evidence of
carcinogenic activity of sodium dichromate dihydrate in male and female F344/N rats based on
increased incidences of squamous cell neoplasms of the oral cavity and clear evidence of
carcinogenic activity of sodium dichromate dihydrate in male and female B6C3F1 mice based on
increased incidences of neoplasms of the small intestine.
The potential mutagenicity of hexavalent chromium has been studied extensively.
Although study results vary with specific test systems, experimental conditions, and hexavalent
chromium compounds tested, results of in vitro and in vivo studies provide substantial evidence
for mutagenic activity of hexavalent chromium compounds. The mutagenicity of hexavalent
chromium is mediated through the generation of highly reactive chromium intermediates (e.g.,
chromium(IV) and chromium(V)) and reactive oxygen species formed during the intracellular
reduction of hexavalent chromium. Reactive chromium intermediates and oxygen species react
with DNA, leading to oxidative DNA damage, chromium-DNA adducts, DNA strand breaks, and
chromosomal aberrations.
In in vitro test systems, hexavalent chromium compounds have mostly tested positive for
gene mutations (including reverse mutations, frame shift mutations, and base pair substitutions)
and DNA damage (including DNA-protein crosslinks) in bacterial cells (S. typhimurium, E. coli,
B. subtilis); for forward mutations and mitotic gene conversion in yeast (S. cerevisiae); and for
DNA damage (DNA strand breaks, fragmentation, DNA-protein crosslinks, DNA-DNA
crosslinks), chromosomal damage (sister chromatid exchanges and chromosomal aberrations),
and DNA synthesis inhibition in mammalian cell lines and primary cell cultures (including

primary cell cultures of human gastric mucosal cells, respiratory tract cells, and lymphocytes).
In in vivo test systems, hexavalent chromium compounds have tested positive for mutations in D.
melanogaster and for DNA damage (DNA-protein crosslinks, DNA strand breaks), mutations,
chromosomal damage (sister chromatid exchanges, chromosomal aberrations, and micronuclei),
and DNA synthesis inhibition in rats and mice. Thus, the mutagenic activity of hexavalent
chromium has been demonstrated in numerous studies using both in vitro and in vivo
experimental systems.
Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), hexavalent
chromium is “likely to be carcinogenic to humans” via the oral route of exposure based on a
statistically significant increase in the incidence of tumors of the oral mucosa and tongue of rats
and of the small intestine of mice, and evidence of an association between oral exposure to
hexavalent chromium and stomach cancer in humans. Additionally, available evidence indicates
that chromium interacts with DNA, resulting in DNA damage and mutagenesis. Based on the
weight of the available evidence, hexavalent chromium is proposed to act through a mutagenic
mode of carcinogenic action, and thus, ADAFs should be applied.
6.2. DOSE RESPONSE

6.2.1. Noncancer—Oral
NTP (2008), a 2-year animal bioassay that used multiple dose groups and included a
comprehensive assessment of endpoints, was selected as the principal study for derivation of the
RfD. Dose-response analysis using BMD methods was conducted for the following endpoints
from this study: histopathological changes in the liver (chronic inflammation in female rats and
histiocytic cellular infiltration in female mice), duodenum (diffuse epithelial hyperplasia in male
and female mice), mesenteric lymph node (histiocytic cellular infiltration in male and female
mice), and pancreas (cytoplasm cellular alteration of acinar epithelial cells in female mice).
All available dichotomous models in EPA’s BMDS were fit to the incidence data for the
selected endpoints, using 10% extra risk as the BMR in accordance with EPA’s Benchmark Dose
Technical Guidance Document (U.S. EPA, 2000b).
Based on the lowest BMDL10 value of 0.09 mg hexavalent chromium/kg-day, diffuse
epithelial hyperplasia of the duodenum in female mice was selected as the POD for derivation of
the RfD. The RfD of 0.0009 or 9 × 10-4 mg/kg-day for hexavalent chromium was derived by
dividing the BMDL10 (or POD) of 0.09 mg/kg-day by a composite uncertainty factor of 100
(10 for extrapolation from animals to humans and 10 for human variability).
Confidence in the principal study, NTP (2008), is high. NTP (2008) is a 2-year
toxicology and carcinogenicity study of sodium dichromate dihydrate in drinking water in rats
and mice that used a relevant route of exposure (drinking water) and a robust study design (i.e.,
50 animals/sex/group, a control and four exposed groups, measurements of drinking water
consumption, and a full suite of hematology, clinical chemistry, and gross and miscroscopic

tissue examinations). Confidence in the oral database is medium to high. The database includes
subchronic and chronic drinking water studies in rats and mice conducted by NTP (2008, 2007),
a second chronic drinking water study in rats, and several other subchronic oral toxicity studies
in rats and mice. Other than the NTP bioassays, however, the available subchronic and chronic
toxicity studies did not provide adequate characterization of both the NOAEL and LOAEL for
the hexavalent chromium compounds tested. Studies have been conducted to evaluate the effects
of ingested hexavalent chromium compounds on reproductive tissues, mating behavior,
reproductive outcomes, and fetal development. The reproductive toxicity database includes a
continuous breeding study by NTP. Overall confidence in the RfD is medium to high.
6.2.2. Cancer—Oral
The mode of action is a key consideration in clarifying how risks should be estimated for
low-dose exposure. A linear low-dose extrapolation approach was used to estimate human
carcinogenic risk associated with hexavalent chromium exposures. This approach is supported
by the evidence for genotoxicity and a mutagenic mode of action.
The CSF for hexavalent chromium is based on tumor incidence data from the NTP (2008)
animal bioassay. The incidence of neoplasms in the small intestine of mice was used to derive
the CSF. Only animals that survived for at least 451 days, the time until appearance of the first
tumor, were considered at risk for tumor development.
BMD modeling was carried out using the multistage model in EPA’s BMDS (U.S. EPA,
2000b) to identify a POD. In applying the BMD approach to the derivation of a CSF, the lower
95% confidence bound on the dose corresponding to the BMR (defined as 10% extra risk of
small intestine tumors) was calculated. This lower confidence bound is referred to as the
BMDL. The CSF was calculated by dividing the BMR by the BMDL, and then converting this
CSF to human equivalents using body weight to the ¾ power scaling.
The CSF resulting from the fitting of the multistage model in BMDS to the incidence of
neoplasms in the small intestine of male and female mice was 0.5 (mg/kg-day)-1 and 0.6 (mg/kg-
day)-1, respectively, expressed in human equivalents. Because of the poorer fit of the multistage
model to the female mouse data, the cancer potency estimate of 0.5 (mg/kg-day)-1 based on the
male mouse data was selected as the CSF for hexavalent chromium.

7. REFERENCES
Aaseth, J; Alexander, J; Norseth, T. (1982) Uptake of 51Cr-chromate by human erythrocytes-a role of glutathione.
Acta Pharmacol Toxicol (Copenh) 50(4):310–315.
ACGIH (American Conference of Governmental Industrial Hygienists). (2004) Threshold limit values for chemical
substances and physical agents and Biological exposure indices. Cincinnati, OH: American Conference of
Governmental Industrial Hygienists. ISBN 1-882417-54-2.
Acharya, S; Mehta, K; Krishnan, S; et al. (2001) A subtoxic interactive toxicity study of ethanol and chromium in
male Wistar rats. Alcohol 23:99–108.
Adams, SP; Laws, GM; Storer, RD; et al. (1996) Detection of DNA damage induced by human carcinogens in
acellular assays: potential application for determining genotoxic mechanisms. Mutat Res 368(3–4):235–248.
Aitio, A.; Järvisaleo, J.; Kiilunen, M.; Tossavainen, A.; Vaittinen, P. (1984) Urinary excretion of chromium as an
indicator of exposure to trivalent chromium sulphate in leather tanning. Int Arch Occup Environ Health 50:310 –
315.
Aitio, A.; Järvisaleo, J.; Kiilunen, M.; Kalliomaki, P.L.; Kalliomaki, K. (1988) Chromium: In Biological
monitoring of toxic metals. Eds. Clarkson, T.W.; Friberg, L.; Nordberg, G.F.; Sager, D.R. pp. 369 – 382. Plenum
Press, New York.
Aiyar, J; Berkovits, HJ; Floyd, RA. (1991) Reaction of hexavalent chromium with glutathione or with hydrogen
peroxide: identification of reactive intermediates and their role in hexavalent chromium-induced DNA damage.
Environ Health Perspect 92:53–62.
Alderson, MR; Rattan, NS; Bidstrup, L. (1981) Health of workmen in the chromate-producing industry in Britain.
Br J Ind Med 38:117-124.
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(2):179–
183.
Amlacher, E; Rudolph, C. (1981) The thymidine incorporation inhibiting screening system (TSS) to test
carcinogenic substances (a nuclear DNA synthesis suppressive short term test). Arch Geschwulstforsch 51(7):605–
610.
Amrani, S; Rizki, M; Creus, A; et al. (1999) Genotoxic activity of different chromium compounds in larval cells of
Drosophila melanogaster, as measured in the wing spot test. Environ Mol Mutagen 34(1):47–51.
Anger, G; Halstenberg, J; Hochgeschwender, K; et al. (2005) Chromium compounds. In: Ullmann's encyclopedia of
industrial chemistry. New York, NY: John Wiley & Sons, Inc., pp. 1–36.
Anwar, RA; Langham, CF; Hoppert, CA; et al. (1961) Chronic toxicity studies. III. Chronic toxicity of cadmium and
chromium in dogs. Arch Environ 3:456–460.
Arlauskas, A; Baker, RS; Bonin, AM; et al. (1985) Mutagenicity of metal ions in bacteria. Environ Res 36(2):379–
388.
Aruldhas, MM; Subramanian, S; Sekhar, P; et al. (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(1):127–137.
Aruldhas, MM; Subramanian, S; Sekar, P; et al. (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(10):2801–2813.

Aruldhas, MM; Subramanian, S; Sekhar, P; et al. (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(4 Suppl):1097–1105.
Ashley, K; Howe, AM; Demange, M; et al. (2003) Sampling and analysis considerations for the determination of
hexavalent chromium in workplace air. J Environ Monit 5:707–716.
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.
ATSDR (Agency for Toxic Substances and Disease Registry). (2008) Toxicological profile for chromium. Draft for
public comment. U.S. Department of Health and Human Services. U.S. Public Health Service, Atlanta, GA.
Baetjer, AM. (1950a) Pulmonary carcinoma in chromate workers. In: A review of the literature and report of cases.
Arch Ind Hyg Occup Med 2(5):487-504.
Baetjer, AM. (1950b) Pulmonary carcinoma in chromate workers. II. Incidence on basis of hospital records. Arch
Ind Hyg Occup Med 2(5):505-516.
Balasoiu, CF; Zagury, GJ; Deschenes, L. (2001) Partitioning and speciation of chromium, copper, and arsenic in
CCA-contaminated soils: Influence of soil composition. Sci Total Environ 280(1-3):239-55.
Banu, SK; Samuel, JB; Arosh, JA; et al. (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(2):180–189.
Baranowska-Dutkiewicz, B. (1981) Absorption of hexavalent chromium by skin in man. Arch Toxicol 47(1):47–50.
Barceloux, DG. (1999) Chromium. J Toxicol Clin Toxicol 37(2):173–194.
Barton, H.A.; Deisinger, P.J.; English, J.C.; Gearhart, J.M.; Faber, W.D.; Tyler, T.R.; Banton, M.I.; Teeguarden, J.;
Andersen, M.E. (2000) Family approach for estimating reference concentrations/doses for series of related organic
chemicals. Toxicol Sci 54:251 – 261.
Bataineh, H; Al-Hamood, MH; Elbetieha, A; et al. (1997) Effect of long-term ingestion of chromium compounds on
aggression, sex behavior and fertility in adult male rat. Drug Chem Toxicol 20(3):133–149.
Bataineh, HN; Bataineh, Z; Daradka, H. (2007) Short-term exposure of female rats to industrial metal salts: effect on
implantation and pregnancy. Reprod Med Biol 6(3):179–183.
Beaumont, JJ; Sedman, RM; Reynolds, SD; et al. (2008) Cancer mortality in a Chinese population exposed to
hexavalent chromium in drinking water. Epidemiology 19(1):12–23.
Bednar CM; Kies C. (1991) Inorganic contaminants in drinking water correlated with disease occurrence in
Nebraska. Water Resour Bull 27(4):631–635.
Bennicelli, C; Camoirano, A; Petruzzelli, S; et al. (1983) High sensitivity of Salmonella TA102 in detecting
hexavalent chromium mutagenicity and its reversal by liver and lung preparations. Mutat Res 122(1):1–5.
Benova, D; Hadjidekova, V; Hristova, R; et al. (2002) Cytogenetic effects of hexavalent chromium in Bulgarian
chromium platers. Mutat Res 514(1–2):29–38.
Beyersmann, D.; Hartwig, A. (2008) Carcinogenic metal compounds: Recent insights into molecular and cellular
mechanisms. Arch Toxicol 82:493 – 512.
Bianchi, V; Dal Toso, R; Debetto, P; et al. (1980) Mechanisms of chromium toxicity in mammalian cell cultures.
Toxicology 17(2):219–224.

Bianchi, V; Celotti, L; Lanfranchi, G; et al. (1983) Genetic effects of chromium compounds. Mutat Res 117(3–
4):279–300.
Bick, RL; Girardi, TV; Lack, WJ; et al. (1996) Hodgkin’s disease in association with hexavalent chromium
exposure. Int J Hematol 64(3–4):257–262.
Bidstrup, PL. (1951) Carcinoma of the lung in chromate workers. Br J Med 8:302-305.
Bidstrup, PL; Case, RAM. (1956) Carcinoma of the lung in workmen in the bichromates-producing industry in
Great Britain. Br J Ind Med 13:260-264.
Bishop, C.; Surgenor, M. Eds. (1964) The red blood cell: A comprehensive treatise. Academic Press, New York.
Blade, LM; Yencken, MS; Wallace, ME; et al. (2007) Hexavalent chromium exposures and exposure-control
technologies in American enterprise: results of a NIOSH field research study. J Occup Environ Hyg 4:595–618.
Blankenship, LJ; Carlisle, DL; Wise, JP; et al. (1997) Induction of apoptotic cell death by particulate lead chromate:
differential effects of vitamins C and E on genotoxicity and survival. Toxicol Appl Pharmacol 146(2):270–280.
Blasiak, J; Kowalik, J. (2000) A comparison of the in vitro genotoxicity of tri- and hexavalent chromium. Mutat
Res 469(1):135–145.
Blasiak, J; Trzeciak, A; Malecka-Panas, E; et al. (1999) DNA damage and repair in human lymphocytes and gastric
mucosa cells exposed to chromium and curcumin. Teratog Carcinog Mutagen 19:19–31.
Bonatti, S; Meini, M; Abbondandolo, A. (1976) Genetic effects of potassium dichromate in Schizosaccharomyces

pombe. Mutat Res 38(2):147–150.
Borges, K.M.; Boswell, J.S.; Liebross, R.H.; Wetterhahn, K.E. (1991) Activation of chromium(VI) by thiols results
in chromium(V) formation, chromium binding to DNA and altered DNA conformation. Carcinogenesis 12:551 –
561.
Borneff, J; Engelhardt, K; Griem, W; et al. (1968) [Carcinogens in water and soil. XXII. Mouse drinking water
experiments with 3,4-benzopyrene and potassium chromate]. Arch Hyg Bakteriol 152(1):45–53.
Borthiry, GR; Antholine, WE; Myers, JM; et al. (2008) Reductive activation of hexavalent chromium by human
lung epithelial cells: generation of Cr(V) and Cr(V)-thiol species. J Inorg Biochem 102(7):1449–1462.
Bragt, PC; van Dura, EA. (1983) Toxicokinetics of hexavalent chromium in the rat after intratracheal administration
of chromates of different solubilities. Ann Occup Hyg 27(3):315–322.
Brams, A; Buchet, JP; Crutzen-Fayt, MC; et al. (1987) A comparative study, with 40 chemicals, of the efficiency of
the Salmonella assay and the SOS chromotest (kit procedure). Toxicol Lett 38(1–2):123–133.
Brandt-Rauf, P. (2006) Editorial retraction. Cancer mortality in a Chinese population exposed to hexavalent
chromium in water. J Occup Environ Med 48(7):749.
Bridges, CC; Zalups, RK. (2005) Molecular and ionic mimicry and the transport of toxic metals. Toxicol Appl
Pharmacol 204(3):274–308.
Bridgewater, LC; Manning, FC; Woo, ES; et al. (1994) DNA polymerase arrest by adducted trivalent chromium.
Mol Carcinog 9(3):122–133.
Bridgewater, LC; Manning, FC; Patierno, SR. (1998) Arrest of replication by mammalian DNA polymerases alpha
and beta caused by chromium-DNA lesions. Mol Carcinog 23(4):201–206.
Briggs, JA; Briggs, RC. (1988) Characterization of chromium effects on a rat liver epithelial cell line and their
relevance to in vitro transformation. Cancer Res 48(22):6484–6490.

Brinton, HP; Frasier, ES; Koven AL. (1952) Morbidity and mortality experience among chromate workers. Public
Health Rep 67(9):835-847.
Bronzetti, GL; Galli, A. (1989) Influence of NTA on the chromium genotoxicity. Toxicol Environ Chem 23:101–
104.
Brooks, B; O’Brien, TJ; Ceryak, S; et al. (2008) Excision repair is required for genotoxin-induced mutagenesis in
mammalian cells. Carcinogenesis 29(5):1064-1069.
Bukowksi, JA. (1991) Biological markers in chromium exposure assessment: Confounding variables. Arch Environ
Health 46(4):230–236.
Campbell J.L.; Tan,Y.; Clewell, H.J. (2009) Development of a PBPK model for hexavalent chromium in rats and
mice to estimate exposure to oral mucosa and small intestine. Toxicologist 108(1):98 (Abstract) Poster ID # 108.
Capellmann, M; Mikalsen, A; Hindrum, M; et al. (1995) Influence of reducing compounds on the formation of
DNA-protein cross-links in HL-60 cells induced by hexavalent chromium. Carcinogenesis 16(5):1135–1139.
Cavalleri, A; Minoia, C; Richelmi, P; et al. (1985) Determination of total and hexavalent chromium in bile after
intravenous administration of potassium dichromate in rats. Environ Res 37(2):490–496.
CDPH (California Department of Public Health). (2007) Chromium-6 in drinking water: sampling results. Health,
CDoP. Available online at http://ww2.cdph.ca.gov/certlic/drinkingwater/pages/chromium6sampling.aspx (accessed
May 10. 2010).
Cemeli, E; Carder, J; Anderson, D; et al. (2003) Antigenotoxic properties of selenium compounds on potassium
dichromate and hydrogen peroxide. Teratog Carcinog Mutagen 23(Suppl 2):53–67.
ChemIDplus. (2008) Chromium and compounds. Bethesda, MD: National Library of Medicine. Available online at
http://chem.sis.nlm.nih.gov/chemidplus/ (accessed May 10, 2010).
Cheng, L; Sonntag, DM; de Boer, J; et al. (2000) Chromium(VI) induces mutagenesis in the lungs of big blue
transgenic mice. J Environ Pathol Toxicol Oncol 19(3):239–249.
Chopra, A; Pereira, G; Gomes, T; et al. (1996) A study of chromium and ethanol toxicity in female Wistar rats.
Toxicol Environ Chem 53:91–106.
Chowdhury, AR; Mitra, C. (1995) Spermatogenic and steroidogenic impairment after chromium treatment in rats.
Indian J Exp Biol 33(7):480–484.
Cikrt, M; Bencko, V. (1979) Biliary excretion and distribution of 51Cr(III) and 51Cr(VI) in rats. J Hyg Epidemiol
Microbiol Immunol 23:241–246.
Clewell, HJ, III; Andersen, ME. (1985) Risk assessment extrapolations and physiological modeling. Toxicol Ind
Health 1(4):111–131.
Clochesy, JM. (1984) Chromium ingestion: a case report. J Emerg Nurs 10(6):281–282.
Cocker, J; Morton, J; Warren, N; Wheeler, JP; Garrod, ANI. (2006) Biomonitoring for chromium and arsenic in
timber treatment plant workers exposed to CCA wood preservatives. Ann Occup Hyg 50(5):517-625.
Coogan, T.P.; Squibb, K.S.; Motz, J.; Kinney, P.L.; Costa, M. (1991a) Distribution of chromium within cells of the
blood. Toxicol Appl Pharmacol 108:157 – 166.
Coogan, TP; Motz, J; Snyder, CA; et al. (1991b) Differential DNA-protein crosslinking in lymphocytes and liver
following chronic drinking water exposure of rats to potassium chromate. Toxicol Appl Pharmacol 109(1):60–72.
Corbett, GE; Finley, BL; Paustenbach, DJ; et al. (1997) Systemic uptake of chromium in human volunteers
following dermal contact with hexavalent chromium (22 mg/L). J Expo Anal Environ Epidemiol 7(2):179–189.

Costa, M. (1991) DNA-protein complexes induced by chromate and other carcinogens. Environ Health Perspect
92:45–52.
Costa, M.; Klein, C.B. (2008) Toxicity and carcinogenicity of chromium compounds in humans. Crit Rev Toxicol
36(2):155 – 163.
Cotton, FA; Wilkinson, G; Murillo, CA; et al. (1999) The elements of the first transition series. In: Advanced
inorganic chemistry. New York, NY: John Wiley & Sons, Inc., pp. 736–752.
Dai, H; Liu, J; Malkas, LH; et al. (2009) Chromium reduces the in vitro and fidelity of DNA replication mediated by
the human cell DNA synthesome. Toxicol Appl Pharmacol 236:154–165.
Danielsson, BRG; Hassoun, E; Dencker, L. (1982) Embryotoxicity of chromium: distribution in pregnant mice and
effects on embryonic cells in vitro. Arch Toxicol 51:233–245.
Davies, JM. (1978) Lung-cancer mortality in workers making chrome pigments. Lancet 1:384.
Davies, JM. (1979) Lung cancer mortality of workers in chromate pigment manufacture: An epidemiological survey.
J Oil Chem Assoc 62:157-163.
Davies, JM. (1984) Lung cancer mortality among workers making lead chromate and zinc chromate pigments at
three English factories. Br J Ind Med 41:158-169.
Decker, LE; Byerrum, RU; Decker, CF; et al. (1958) Chronic toxicity studies. I. Cadmium administered to drinking
water in rats. AMA Arch Ind Health 18(3):228–231.
De Flora, S. (1978) Metabolic deactivation of mutagens in the Salmonella-microsome test. Nature 271(5644):455–
456.
De Flora, S. (1981) Study of 106 organic and inorganic compounds in the Salmonella/microsome test.
Carcinogenesis 2(4):283–298.
De Flora, S. (2000) Threshold mechanisms and site specificity in chromium(VI) carcinogenesis. Carcinogenesis
21(4):533 – 541.
De Flora, S; Wetterhahn, KE. (1989) Mechanisms of chromium metabolism and genotoxicity. Life Sci 7:169–244.
De Flora, S; Bennicelli, C; Zanacchi, P; et al. (1984) Metabolic activation and deactivation of mutagens by
preparations of human lung parenchyma and bronchial tree. Mutat Res 139(1):9–14.
De Flora, S; Badolati, GS; Serra, D; et al. (1987) Circadian reduction of chromium in the gastric environment.
Mutat Res 192(3):169–174.
De Flora, S; Camoirano, A; Bagnasco, M; et al. (1997) Etimates of the chromium(VI) reducing capacity in human
body compartments as a mechanism for attenuating its potential toxicity and carcinogenicity. Carcinogenesis
18(3):531–537.
De Flora, S; Iltcheva, M; Balansky, RM. (2006) Oral chromium (VI) does not affect the frequency of micronuclei in
hematopoietic cells of adult mice and of transplacentally exposed fetuses. Mutat Res 610(1–2):38–47.
De Flora, S; D’Agostini, F, Balansky, R; et al. (2008) Lack of genotoxic effects in hematopoietic and
gastrointestinal cells of mice receiving chromium(VI) with the drinking water. Mutat Res 659:60–67.
Deng, C; Lee, HH; Xian, H; et al. (1988) Chromosomal aberrations and sister chromatid exchanges of peripheral
blood lymphocytes in Chinese electroplating workers: effect of nickel and chromium. J Trace Elem Exper Med
1:57–62.

Depault, F; Cojocaru, M; Fortin, F; et al. (2006) Genotoxic effects of hexavalent chromium and cadmium(II) in
human blood lymphocytes using the electron microscopy in situ end-labeling (EM-ISEL) assay. Toxicol In Vitro
20(4):513–518.
Devi, KD; Rozati, R; Saleha Banu, B; et al. (2001) In vivo genotoxic effect of potassium dichromate in mice
leukocytes using comet assay. Food Chem Toxicol 39(8):859–865.
DiPaolo, JA; Casto, BC. (1979) Quantitative studies of in vitro morphological transformation of Syrian hamster
cells by inorganic metal salts. Cancer Res 39(3):1008–1013.
Donaldson, RM, Jr; Barreras, RF. (1966) Intestinal absorption of trace quantities of chromium. J Lab Clin Med
68(3):484–493.
Dunkel, VC; Pienta, RJ; Sivak, A; et al. (1981) Comparative neoplastic transformation responses of Balb/3T3 cells,
Syrian hamster embryo cells, and Rauscher murine leukemia virus-infected Fischer 344 rat embryo cells to chemical
compounds. J Natl Cancer Inst 67(6):1303–1312.
Dunkel, VC; Zeiger, E; Brusick, D; et al. (1984) Reproducibility of microbial mutagenicity assays: I. Tests with
Salmonella typhimurium and Escherichia coli using a standardized protocol. Environ Mutagen 6(Suppl 2):1–251.
Edel, J; Sabbioni, E. (1985) Pathways of Cr(III) and Cr(VI) in the rat after intratracheal administration. Hum
Toxicol 4(4):409–416.
Eizaguirre-Garcia, D; Rodriguez-Andres, C; Watt, GC; et al. (1999) A study of leukaemia in Glasgow in connection
with chromium-contaminated land. J Public Health Med 21(4):435–438.
Eizaguirre-Garcia, D; Rodriguez-Andres, C; Watt, GC. (2000) Congenital anomalies in Glasgow between 1982 and
1989 and chromium waste. J Public Health Med 22(1):54–58.
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(1–3):39–47.
Elia, MC; Storer, RD; McKelvey, TW; et al. (1994) Rapid DNA degradation in primary rat hepatocytes treated with
diverse cytotoxic chemicals: analysis by pulsed field gel electrophoresis and implications for alkaline elution assays.
Environ Mol Mutagen 24(3):181–191.
Elsaieed, EM; Nada, SA. (2002) Teratogenicity of hexavalent chromium in rats and the beneficial role of ginseng.
Bull Environ Contam Toxicol 68(3):361–368.
Enterline, PE. (1974) Respiratory cancer among chromate workers. J Occup Med 16:523-526.
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(2):203–209.
Fendorf, S; La Force, MJ; Li, G. (2004) Heavy metals in the environment: Temporal changes in soil partitioning of
and bioaccessibility of arsenic, chromium and lead. J Environ Qual 33:2049–2055.
Finley, BL; Johnson, EM; Holson, JF. (1993) Comment on “Comparative effects of trivalent and hexavalent
chromium on spermatogenesis of the mouse.” Toxicol Environ Chem 39:133–137.
Finley, BL; Scott, PK; Norton, RL; et al. (1996) Urinary chromium concentrations in humans following ingestion of
safe doses of hexavalent and trivalent chromium: implications for biomonitoring. J Toxicol Environ Health
48(5):479–499.
Finley, BL; Kerger, BD; Katona, MW; et al. (1997) Human ingestion of chromium (VI) in drinking water:
pharmacokinetics following repeated exposure. Toxicol Appl Pharmacol 142(1):151–159.
Flores, A; Perez, JM. (1999) Cytotoxicity, apoptosis, and in vitro DNA damage induced by potassium chromate.
Toxicol Appl Pharmacol 161(1):75–81.

Fornace, AJ, Jr. (1982) Detection of DNA single-strand breaks produced during the repair of damage by DNA-
protein cross-linking agents. Cancer Res 42(1):145–149.
Fornace, AJ, Jr.; Seres, DS; Lechner, JF; et al. (1981) DNA-protein cross-linking by chromium salts. Chem Biol
Interact 36(3):345–354.
Frentzel-Beyme, R. (1983) Lung cancer mortality of workers employed in chromate pigment factories. A
multicentric European epidemiological study. J Cancer Res Clin Oncol 105:183-188.
Fristedt, B; Lindqvist, B; Schuetz, A; et al. (1965) Survival in a case of acute oral chromic acid poisoning with acute
renal failure treated by haemodialysis. Acta Med Scand 177:153–159.
Fryzek, JP; Mumma, MT; McLaughlin, JK; et al. (2001) Cancer mortality in relation to environmental chromium
exposure. J Occup Environ Med 43(7):635–640.
Fukunaga, M; Kurachi, Y; Mizuguchi, Y. (1982) Action of some metal ions on yeast chromosomes. Chem Pharm
Bull 30:3017–3019.
Furst, A; Schlauder, M; Sasmore, DP. (1976) Tumorigenic activity of lead chromate. Cancer Res 36:1779-1783.
Fytianos, K. (2001) Speciation analysis of heavy metals in natural waters: A review. J AOAC Intl 84(6):1763–
1769.
Gambelunghe, A; Piccinini, R; Ambrogi, M; et al. (2003) Primary DNA damage in chrome-plating workers.
Toxicology 188(2–3):187–195.
Gammelgaard, B.; Jensen, K.; Steffansen, B. (1999) In vitro metabolism and permeation studies in rat jejunum:
Organic chromium compared to inorganic chromium. J Trace Elements Med Biol 13:82 – 88.
Gao, M; Levy, LS; Braithwaite, RA; et al. (1993) Monitoring of total chromium in rat fluids and lymphocytes
following intratracheal administration of soluble trivalent or hexavalent chromium compounds. Hum Exp Toxicol
12(5):377–382.
Gao, M; Levy, LS; Faux, SP; et al. (1994) Use of molecular epidemiological techniques in a pilot study on workers
exposed to chromium. Occup Environ Med 51(10):663–668.
Garcia, JD; Jennette, KW. (1981) Electron-transport cytochrome P-450 system is involved in the microsomal
metabolism of the carcinogen chromate. J Inorg Biochem 14(4):281–295.
Gargas, ML; Norton, RL; Paustenbach, DJ; et al. (1994) Urinary excretion of chromium by humans following
ingestion of chromium picolinate. Implications for biomonitoring. Drug Metab Dispos 22(4):522–529.
Gealy, R; Wright-Bourque, JL; Kraynak, AR; et al. (2007) Validation of a high-throughput in vitro alkaline
elution/rat hepatocyte assay for DNA damage. Mutat Res 629(1):49–63.
Gentile, JM; Hyde, K; Schubert, J. (1981) Chromium genotoxicity as influenced by complexation and rate effects.
Toxicol Lett 7(6):439–448.
Glaser, U.; Hochrainer, D.; Klöpper, H.; Kuhnen, H. (1985) Low level chromium (VI) inhalation effects on alveolar
macrophages and immune functions in Wistar rats. Arch Toxicol. 57(4):250 – 256.
Glaser, U; Hochrainer, D; Kloppel, H; et al. (1986) Carcinogenicity of sodium dichromate and chromium(VI/III)
oxide aerosols inhaled by male Wistar rats. Toxicology 42:219-232.
Goldman, M; Karotkin, RH. (1935) Acute potassium bichromate poisoning. Am J Med Sci 189:400–403.
Gomez-Arroyo, S; Altamirano, M; Villalobos-Pietrini, R. (1981) Sister-chromatid exchanges induced by some

chromium compounds in human lymphocytes in vitro. Mutat Res 90(4):425–431.

Gonzalez-Vergara, E.; de Gonzalez, B.C.; Hegenauer, J.; Saltman, P. (1980) Chromium coordination complexes of
pyridoxal and nicotinic acid: Synthesis, absorption and metabolism. Isr J Chem 21:18 – 22.
Goode, EL; Ulrich, CM; Potter, JD. (2002) Polymorphisms in DNA repair genes and associations with cancer risk.
Cancer Epidemiol Biomarkers Prev 11(12):1513–1530.
Goodgame, D.M.L.; Joy, M.A. (1998) EPR stud of the Cr(V) and radical species produced in the reduction of
Cr(VI) by ascorbate. Inorg Chem Acta 135:115 – 118.
Graf, U; Wurgler, FE. (1996) The somatic white-ivory eye spot test does not detect the same spectrum of genotoxic
events as the wing somatic mutation and recombination test in Drosophila melanogaster. Environ Mol Mutagen
27(3):219–226.
Gray, S.J.; Sterling, K. (1950) The tagging of red cells and plasma proteins with radioactive chromium. J Clin
Invest 29:1604 – 1613.
Grlickova-Duzevik, E; Wise, SS; Munroe, RC; et al. (2006) XRCC1 protects cells from chromate-induced
chromosome damage, but does not affect cytotoxicity. Mutat Res 610(1-2):31–37.
Gruber, JE; Jennette, KW. (1978) Metabolism of the carcinogen chromate by rat liver microsomes. Biochem
Biophys Res Commun 82(2):700–706.
Guttmann, D; Poage, G; Johnston, T; et al. (2008) Reduction with glutathione is a weakly mutagenic pathway in
chromium(VI) metabolism. Chem Res Toxicol 21(11):2188–2194.
Gylseth, B; Gundersen, N; Langard, S. (1977) Evaluation of chromium exposure based on a simplified method for
urinary chromium. Scand J Work Environ Health 3:28–31.
Haguenor, JM; Dubois, G; Frimat, P; et al. (1981) Mortality due to bronch-pulmonary >cancer in a factory
producing pigments based on lead and zinc chromates. In: Prevention of occupational cancer - International
symposium, occupational safety and health series 46. Geneva, Switzerland: International Labour Office, pp. 168-
176. (French).
Hamilton, JW; Wetterhahn, KE. (1986) Chromium(VI)-induced DNA damage in chick embryo liver and blood cells
in vivo. Carcinogenesis 7(12):2085–2088.
Hanahan, D; Weinberg, RA (2000) The hallmarks of cancer. Cell 100:57-70.
Hantson, P; Van Caenegem, O; Decordier, I; et al. (2005) Hexavalent chromium ingestion: biological markers of
nephrotoxicity and genotoxicity. Clin Toxicol (Phila) 43(2):111–112.
Hasan, A. (2007) A case report: ammonium dichromate poisoning. Biomed Res 18(1):35–37.
Haworth, S; Lawlor, T; Mortelmans, K; et al. (1983) Salmonella mutagenicity test results for 250 chemicals.
Environ Mutagen 5(Suppl 1):1–142.
Hayashi, M; Sofuni, T; Ishidate, M, Jr. (1982) High-sensitivity in micronucleus induction of a mouse strain (MS).
Mutat Res 105(4):253–256.
Hayes, RB; Lilienfeld, AM; Snell, LM. (1979) Mortality in chromium chemical production workers: a prospective
study. Int J Epidemiol 8(4):365-374.
Hayes, RB; Sheffet, A; Spirtas, R. (1989) Cancer mortality among a cohort of chromium pigment workers. Am J Ind
Med 16:127-133.
Heil, J; Reifferscheid, G. (1992) Detection of mammalian carcinogens with an immunological DNA synthesis-
inhibition test. Carcinogenesis 13(12):2389–2394.

Hill, WJ; Ferguson, WS. (1979) Statistical analysis of epidemiological data from chromium chemical manufacturing
plant. J Occup Med 21:103-106.
Hirose, T; Kondo, K; Takahashi, Y; et al. (2002) Frequent microsatellite instability in lung cancer from chromate-
exposed workers. Mol Carcinog 33(3):172–180.
Holmes AL, Wise SS, Sandwick SJ, et al. (2006) The clastogenic effects of chronic exposure to particulate and
soluble Cr(VI) in human lung cells. Mutat Res 610(1–2):8–13.
Holmes, A.; Wise, SS; Wise, Sr., JP (2008) Carcinogenicity of hexavalent chromium. Indian J Med Res 128:353 –
372.
Holmes, AL; Wise, SS; Pelsue, SC; et al. (2010) Chronic exposure to zinc chromate induces centrosome
amplification and spindle assembly checkpoint bypass in human lung fibroblasts. Chem Res Toxicol 23:386-395.
Hueper, WC. (1961) Environmental carcinogenesis and cancers. Cancer Res 21:842-857.
Hueper, WC; Payne, WW. (1962) Experimental studies in metal carcinogenesis: Chromium, nickel, iron, and
arsenic. Arch Environ Health 5:445-462.
Iijima, S; Spindle, A; Pedersen, RA. (1983) Developmental and cytogenetic effects of potassium dichromate on
mouse embryos in vitro. Teratology 27(1):109–115.
Imreh, S; Radulescu, D. (1982) Cytogenetic effects of chromium in vivo and in vitro. Mutat Res 97(3):192–193.
Iserson, KV; Banner, W; Froede, RC; et al. (1983) Failure of dialysis therapy in potassium dichromate poisoning. J
Emerg Med 1(2):143–149.
Itoh, S; Shimada, H. (1996) Micronucleus induction by chromium and selenium, and suppression by metallothionein
inducer. Mutat Res 367(4):233–236.
Itoh, S; Shimada, H. (1997) Clastogenicity and mutagenicity of hexavalent chromium in lacZ transgenic mice.
Itoh, S; Shimada, H. (1998) Bone marrow and liver mutagenesis in lacZ transgenic mice treated with hexavalent
chromium. Mutat Res 412(1):63–67.
Izzotti, A; Bagnasco, M; Camoirano, A; et al. (1998) DNA fragmentation, DNA-protein crosslinks, postlabeled
nucleotidic modifications, and 8-hydroxy-2’-deoxyguanosine in the lung but not in the liver of rats receiving
intratracheal instillations of hexavalent chromium. Chemoprevention by oral N-acetylcysteine. Mutat Res 400(1–
2):233–244.
James, BR; Petura, JC; Vitale, RJ; et al. (1997) Oxidation-reduction chemistry of chromium: relevance to the
regulation and remediation of chromate-contaminated soils. J Soil Contam 6(6):569–580.
Jannetto, PJ; Antholine, WE; Myers, CR. (2001) Cytochrome b5 plays a key role in human microsomal hexavalent
chromium reduction. Toxicology 159(3):119–133.
Jennette, KW. (1982) Microsomal reduction of the carcinogen chromate produces chromium(V). J Am Chem Soc
104:874–875.
Johnson, J; Schewel, L; Graedel, TE. (2006) The contemporary anthropogenic chromium cycle. Environ Sci
Technol 40:7060–7069.
Junaid, M; Murthy, RC; Saxena, DK. (1995) Chromium fetotoxicity in mice during late pregnancy. Vet Hum
Toxicol 37(4):320–323.
Junaid, M; Murthy, RC; Saxena, DK. (1996a) Embryotoxicity of orally administered chromium in mice: exposure
during the period of organogenesis. Toxicol Lett 84(3):143–148.

Junaid, M; Murthy, RC; Saxena, DK. (1996b) Embryo- and fetotoxicity of chromium in pregestationally exposed
mice. Bull Environ Contam Toxicol 57(2):327–334.
Kanematsu, N; Hara, M; Kada, T. (1980) Rec assay and mutagenicity studies on metal compounds. Mutat Res
77(2):109–116.
Kanojia, RK; Junaid, M; Murthy, RC. (1996) Chromium induced teratogenicity in female rat. Toxicol Lett
89(3):207–213.
Kanojia, RK; Junaid, M; Murthy, RC. (1998) Embryo and fetotoxicity of hexavalent chromium: a long-term study.
Kargacin, B; Squibb, KS; Cosentino, S; et al. (1993) Comparison of the uptake and distribution of chromate in rats
and mice. Biol Trace Elem Res 36:307–318.
Katz, AJ; Chiu, A; Beaubier, J; et al. (2001) Combining Drosophila melanogaster somatic-mutation-recombination
and electron-spin-resonance-spectroscopy data to interpret epidemiologic observations on chromium
carcinogenicity. Mol Cell Biochem 222(1–2):61–68.
Katz, SA. (1991) The analytical biochemistry of chromium. Environ Hlth Perspect 92:13–16.
Katz, SA; Salem, H. (1993) The toxicology of chromium with respet to its chemical speciation: A review. J Appl
Toxicol 13(3):217–224.
Kaufman, DB; DiNicola, W; McIntosh, R. (1970) Acute potassium dichromate poisoning. Treated by peritoneal
dialysis. Am J Dis Child 119(4):374–376.
Kaya, B; Creus, A; Velazquez, A; et al. (2002) Genotoxicity is modulated by ascorbic acid. Studies using the wing
spot test in Drosophila. Mutat Res 520(1–2):93–101.
Kerger, BD; Richter, RO; Chute, SM; et al. (1996a) Refined exposure assessment for ingestion of tap water
contaminated with hexavalent chromium: Consideration of exogenous and endogenous reducing agents. J Exposure
Anal Environ Epi 6(2):163–179.
Kerger, BD; Paustenbach, DJ; Corbett, GE; et al. (1996b) Absorption and elimination of trivalent and hexavalent
chromium in humans following ingestion of a bolus dose in drinking water. Toxicol Appl Pharmacol 141(1):145–
158.
Kerger, BD; Finley, BL; Corbett, GE; et al. (1997) Ingestion of hexavalent chromium in drinking water by human
volunteers: absorption, distribution, and excretion of single and repeated doses. J Toxicol Environ Health 50(1):67–
95.
Kerger, BD; Butler WJ; Paustenbach, DJ; et al. (2009) Cancer mortality in Chinese populations surrounding an alloy
plant with chromium smelting operations. J Toxicol Environ Health Part A 72(5):329–344.
Kharab, P; Singh, I. (1985) Genotoxic effects of potassium dichromate, sodium arsenite, cobalt chloride and lead
nitrate in diploid yeast. Mutat Res 155(3):117–120.
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(7):3480-3484.
Klein, CB; Su, L; Bowser, D; et al. (2002) Chromate-induced epimutations in mammalian cells. Environ Health
Perspect 110(Suppl 5):739–743.
Knudsen, I. (1980) The mammalian spot test and its use for the testing of potential carcinogenicity of welding fume
particles and hexavalent chromium. Acta Pharmacol Toxicol (Copenh) 47(1):66–70.

Kondo, K; Takahashi, Y; Hirose, Y; et al. (2006) The reduced expression and aberrant methylation of p16(INK4a)
in chromate workers with lung cancer. Lung Cancer 53(3):295–302.
Korallus, U. (1986) Chromium compounds: Occupational health, toxicological and biological monitoring aspects.
Toxicol Environ Chem 12:47–59.
Korallus, U; Lange H; Ness, A; et al. (1982) Relationships between precautionary measures and bronchial
carcinoma mortality in the chromate-producing industry. Arbeitsmedizin, Socialmedizin, Preventivmedizin.
17(7):159-167. (German - Eng. summary).
Kortenkamp, A; Oetken, G; Beyersmann, D. (1990) The DNA cleavage induced by a chromium(V) complex and by
chromate and glutathione is mediated by activated oxygen species. Mutat Res 232(2):155–161.
Koshi, K. (1979) Effects of fume particles from stainless steel welding on sister chromatid exchanges and
chromosome aberrations in cultured Chinese hamster cells. Ind Health 17:39–50.
Koshi, K; Iwasaki, K. (1983) Solubility of low-solubility chromates and their clastogenic activity in cultured cells.
Ind Health 21(2):57–65.
Koshi, K; Serita, F; Sawatari, K; et al. (1987) Cytogenetic analysis of bone marrow cells and peripheral blood
lymphocytes from rats exposed to chromium fumes by inhalation. Mutat Res 181:365.
Krishnan, K; Andersen, ME. (1994) Physiologically based pharmacokinetic modeling in toxicology. In: Hayes,
AW, ed. Principles and methods of toxicology. New York, NY: Raven Press Ltd., pp. 149–188.
Krishnan, K; Andersen, ME; Clewell, HJ; et al. (1994) Physiologically based pharmacokinetic modeling of chemical
structures. In: Yang, RSH, ed. Toxicology of chemical mixtures. New York, NY: Academic Press, pp. 399–437.
Kumpulainen, JT. (1992) Chromium content of foods and diets. Biol Trace Elem Res 32:9-18.
Kuykendall, JR; Kerger, BD; Jarvi, EJ; et al. (1996) Measurement of DNA-protein cross-links in human leukocytes
following acute ingestion of chromium in drinking water. Carcinogenesis 17(9):1971–1977.
Langård, S; Gundersen, N, Tsalev, DL; et al. (1978) Whole blood chromium and chromium excretion in the rat after
zinc chromate inhalation. Acta Pharmacol Toxicol (Copenh) 42(2):142–149.
Laskin, S; Kuschner, M; Drew, RT. (1970) Studies in pulmonary carcinogenesis. In: Hanna, Jr., MG;, Nettesheim,
P; and Gilbert, JR, eds.
LaVelle, JM. (1986) Potassium chromate potentiates frameshift mutagenesis in E. coli and S. typhimurium. Mutat
Res 171(1):1–10.
Lay, P.A.; Levina, A. (1998) Activation of molecular oxygen during the reactions of chromium(VI/V/IV) with
biological reductants: Implications for chromium-induced genotoxicities. J Am Chem Soc 120:6704 – 6714.
Le Curieux, F; Marzin, D; Erb, F. (1993) Comparison of three short-term assays: results on seven chemicals.
Potential contribution to the control of water genotoxicity. Mutat Res 319(3):223–236.
Levina, A.; Zhang, L.; Lay, P.A. (2003) Structure and reactivity of chromium(V) glutathione complex. Inorg Chem
42:767 – 784.
Levina, A; Lay, PA. (2005) Mechanistic studies of relevance to the biological activities of chromium. Coord Chem
Rev 249(3–4):281–298.
Levina, A.; Harris, H.H.; Lay, P.A. (2007) X-ray absorption and EPR spectroscopic studies of the
biotransformations of chromium(VI) in mammalian cells. Is chormodulin an artifact of isolation methods? J Am
Chem Soc 129:1065 – 1075.

Levis, AG; Majone, F. (1979) Cytotoxic and clastogenic effects of soluble chromium compounds on mammalian
cell cultures. Br J Cancer 40(4):523–533.
Levis, AG; Majone, F. (1981) Cytotoxic and clastogenic effects of soluble and insoluble compounds containing
hexavalent and trivalent chromium. Br J Cancer 44(2):219–235.
Levy, LS; Martin, PA. (1983) The effects of a range of chromium-containing materials on rat lung. Sponsored by
Dry Color Manufacturers' Association and others. (Unpublished)
Lewalter, J; Korallus, U; Harzdorf, C; Wiedmann, H. (1985) Chromium bond detection in isolated erythrocytes: A
new principle of biological monitoring of exposure to hexavalent chromium. Int Arch Occup Environ Health
55:305–318.
Lewis, RJ. (2007) Chromium and compounds. In: Hawley's condensed chemical dictionary. New York, NY: John
Wiley & Sons, pp. 68–69, 216, 297, 1029–1030.
Li, H; Chen, Q; Li, S; et al. (2001) Effect of Cr(VI) exposure on sperm quality: human and animal studies. Ann
Occup Hyg 45(7):505–511.
Lide, DR. (2008) Chromium and compounds. In: Lide, DR; ed. CRC handbook of chemistry and physics. Boca
Raton, FL: CRC Press, pp. 4-46, 44-54, 44-58, 44-59, 44-70, 44-82, 44-89, 44-100.
Liebross, RH; Wetterhahn, E. (1992) In vivo formation of chromium(V) in chick embryo liver and red blood cells.
Lim, T.H.; Sargent, T. III.; Jusubov, N. (1983) Kinetics of trace element chromium (III) in the body. Am J Physiol
244(4):R445 – R454.
Liu, KJ; Shi, X; Jiang, JJ; et al. (1995) Chromate-induced chromium(V) formation in live mice and its control by
cellular antioxidants: an L-band electron paramagnetic resonance study. Arch Biochem Biophys 323(1):33–39.
Llagostera, M; Garrido, S; Guerrero, R; et al. (1986) Induction of SOS genes of Escherichia coli by chromium
compounds. Environ Mutagen 8(4):571–577.
Losi, ME; Amrhein, C; Frankenberger, Jr, WT. (1994) Environmental biochemistry of chromium. Rev Environ
Contam Toxicol 136:91–121.
Loubieres, Y; de Lassence, A; Bernier, M; et al. (1999) Acute, fatal, oral chromic acid poisoning. J Toxicol Clin
Toxicol 37(3):333–336.
Loyaux-Lawniczak, S; Lecomte, P; Ehrhardt, JJ. (2001) Behavior of hexavalent chromium in a polluted

groundwater: redox processes and immobilization in soils. Environ Sci Technol 35:1350–1357.
Macfie, A; Hagan, E; Zhitkovich, A. (2010) Mechanism of DNA–protein cross-linking by chromium. Chem Res
Toxicol 23(2):341–347.
Machle, W; Gregorius, F. (1948) Cancer of the respiratory system in the United States chromate-producing industry.
Public Health Rep 63(35):1114-1127.
MacKenzie, RD; Byerrum, RU; Decker, CF; et al. (1958) Chronic toxicity studies. II. Hexavalent and trivalent
chromium administered in drinking water to rats. AMA Arch Ind Health 18(3):232–234.
MacKenzie, RD; Anwar, RA; Byerrum, RU; et al. (1959) Absorption and distribution of 51Cr in the albino rat. Arch
Biochem Biophys 79:200–205.
Macrae, WD; Whiting, RF; Stich, HF. (1979) Sister chromatid exchanges induced in cultured mammalian cells by
chromate. Chem-Biol Interact 26(3):281–286.

Majone, F; Levis, AG. (1979) Chromosomal aberrations and sister-chromatid exchanges in Chinese hamster cells
treated in vitro with hexavalent chromium compounds. Mutat Res 67(3):231–238.
Mancuso, TF; Hueper, WC. (1951) Occupational cancer and other health Hazards in a chromate plant: A medical
appraisal. In: Lung cancers in chromate workers. Ind Med Surg 20(8):358-363.
Mancuso, TF. (1975) Consideration of chromium as an industrial carcinogen.

International Conference on Heavy Metals in the Environment, Toronto, Ontario, Canada, October 27-31. pp. 343-
356.
Mancuso, TF. (1997) Chromium as an industrial carcinogen: Part 1. Am J Ind Med 31:129-139.
Marzin, DR; Phi, HV. (1985) Study of the mutagenicity of metal derivatives with Salmonella typhimurium TA102.
Mutat Res 155(1–2):49–51.
Matsui, S. (1980) Evaluation of a Bacillus subtilis rec-assay for the detection of mutagens which may occur in water
environments. Water Res 14(11):1613–1619.
McCarroll, N; Keshava, N; Chen, J; et al. (2010) An evaluation of the mode of action framework for mutagenic
carcinogens case study II: Chromium(VI). Environ Mol Mutagen 51:89–111.
McGregor, DB; Martin, R; Cattanach, P; et al. (1987) Responses of the L5178Y tk+/tk- mouse lymphoma cell
forward mutation assay to coded chemicals. I: Results for nine compounds. Environ Mutagen 9(2):143–160.
Merk, O; Reiser, K; Speit, G. (2000) Analysis of chromate-induced DNA-protein crosslinks with the comet assay.
Mutat Res 471(1–2):71–80.
Mertz, W; Roginski, EE; Feldman, FJ; et al. (1969) Dependence of chromium transfer into the rat embryo on the
chemical form. J Nutr 99(3):363–367.
Mertz, W. (1998) Chromium research from a distance: From 1959 to 1980. J Am College Nutrition 17(6):544 –
547.
Mikalsen, A; Alexander, J; Ryberg, D. (1989) Microsomal metabolism of hexavalent chromium. Inhibitory effect of
oxygen and involvement of cytochrome P450. Chem Biol Interact 69:175–192.
Miller, CA; Costa M. (1989) Analysis of proteins cross-linked to DNA after treatment of cells with formaldehyde,
chromate, and cis-diamminedichloroplatinum(II). Mol Toxicol 2:11–26.
Minoia, C; Cavalleri, A. (1988) Chromium in urine, serum and red blood cells in the biological monitoring of
workers exposed to different chromium valency states. Sci Total Environ 71(3):323–327.
Mirsalis, JC; Hamilton, CM; O’Loughlin, KG; et al. (1996) Hexavalent chromium 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(1):60–63.
Mitchell, AD; Rudd, CJ; Caspary, WJ. (1988) Evaluation of the L5178Y mouse lymphoma cell mutagenesis assay:
intralaboratory results for sixty-three coded chemicals tested at SRI International. Environ Mol Mutagen 12(Suppl
13):37–101.
Murthy, RC; Junaid, M; Saxena, DK. (1996) Ovarian dysfunction in mice following chromium (VI) exposure.
Myers, CR; Myers, JM. (1998) Iron stimulates the rate of reduction of hexavalent chromium by human microsomes.
Myhr, B; Caspary, W. (1988) Evaluation of the L5178Y mouse lymphoma cell mutagenesis assay; intralaboratory
results for sixty-three coded chemicals tested at Litton Bionetics, Inc. Environ Mol Mutagen 13(12):103–194.

Nagaya, T. (1986) No increase in sister-chromatid exchange frequency in lymphocytes of chromium platers. Mutat
Res 170(3):129–132.
Nagaya, T; Ishikawa, N; Hata, H; et al. (1991) Sister-chromatid exchanges in lymphocytes from 12 chromium
platers: a five-year follow-up study. Toxicol Lett 58(3):329–335.
Nakamura, SI; Oda, Y; Shimada, T; et al. (1987) SOS-inducing activity of chemical carcinogens and mutagens in
Salmonella typhimurium TA1535/pSK1002: examination with 151 chemicals. Mutat Res 192(4):239–246.
Nakamuro, K; Yoshikawa, K; Sayato, Y; et al. (1978) Comparative studies of chromosomal aberration and
mutagenicity of the trivalent and hexavalent chromium. Mutat Res 58(2–3):175–181.
National Institute for Occupational Safety and Health (NIOSH). (1975) Criteria for a recommended standard
occupational exposure to chromium (VI). U.S. Department of Health, Education, and Welfare, Washington, DC.
Nickens, K.P.; Patierno, S.R.; Ceryak, S. (2010) Chromium genotoxicity: A double-edged sword. Chemi Biol
Interact Nov 5;188(2):276-88. Epub 2010 Apr 27.
Nishio, A; Uyeki, EM. (1985) Inhibition of DNA synthesis by chromium compounds. J Toxicol Environ Health
15(2):237–244.
Nishioka, H. (1975) Mutagenic activities of metal compounds in bacteria. Mutat Res 31(3):185–189.
Norseth, T; Alexander, J; Aaseth, J; et al. (1982) Biliary excretion of chromium in the rat: a role of glutathione.
Acta Pharmacol Toxicol (Copenh) 51(5):450–455.
Nriagu, JO. (1988) Production and uses of chromium. In: Nriagu, J.O.; Nieboer, E. (eds) Chromium in natural and
human environments. John Wiley and Sons, New York, pp. 81–105.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press.
NRC. (2002) Bioavailability of contaminants in soils and sediments: Processes, tools and applications. National
Academies Press: Washington D.C.
NTP (National Toxicology Program). (1996a) Final report on the reproductive toxicity of potassium dichromate
(hexavalent) (CAS No. 7778-50-9) administered in diet to BALB/c mice. National Institute of Environmental
Health Sciences, National Toxicology Program. PB97125363.
NTP. (1996b) Final report on the reproductive toxicity of potassium dichromate (hexavalent) (CAS No. 7778-50-9)
administered in diet to SD rats. National Institute of Environmental Health Sciences, National Toxicology Program.
PB97125355.
NTP. (1997) Final report on the reproductive toxicity of potassium dichromate (CAS No. 7778-50-9) administered
in diet to BALB/c mice. National Institute of Environmental Health Sciences, National Toxicology Program.
PB97144919.
NTP. (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. Washington, DC: National Toxicology Program. Toxicity Report Series Number 72. Available
online at http://ntp.niehs.nih.gov/ntp/htdocs/ST_rpts/TOX72.pdf (accessed January 29, 2008).
NTP. (2008) NTP technical report on the toxicology and carcinogenesis studies of sodium dichromate dihydrate
(CAS No. 7789-12-0) in F344/N rats and B6C3F1 mice (drinking water studies). Washington, DC: National
Toxicology Program; NTP TR 546. Available online at http://ntp.niehs.nih.gov/files/546_web_FINAL.pdf
(accessed January 29, 2008).
Oberly, TJ; Piper, CE; McDonald, DS. (1982) Mutagenicity of metal salts in the L5178Y mouse lymphoma assay. J
Toxicol Environ Health 9(3):367–376.

O’Brien, P.; Wang, G.; Wyatt, P.B. (1992) Studies on the kinetics of the reduction of chromate by glutathione and
related thiols. Polyhedron 11:3211 – 3216.
O'Brien, TJ; Ceryak, S; Patierno, SR. (2003) Complexities of chromium carcinogenesis: role of cellular response,
repair and recovery mechanisms. Mutat Res 533(1–2):3–36.
O’Flaherty, E.J. (1991a) Physiologically based models for bone-seeking elements. I. Rat skeletal and bone growth.
Toxicol Appl Pharmacol 111:299 – 312.
O’Flaherty, E.J. (1991b) Physiologically based models for bone-seeking elements. II. Kinetics of lead disposition
in rats. Toxicol Appl Pharmacol 111:313 – 331.
O'Flaherty, EJ. (1993) Physiologically based models for bone-seeking elements. IV. Kinetics of lead disposition in
humans. Toxicol Appl Pharmacol 118(1):16–29.
O’Flaherty, EJ. (1995) Physiologically based models for bone-seeking elements. V. Lean absorption and disposition
in childhood. Toxicol Appl Pharmacol 131(2):297–308.
O’Flaherty, EJ. (1996) A physiologically based model of chromium kinetics in the rat. Toxicol Appl Pharmacol
138(1):54–64.
O’Flaherty, EJ; Kerger, BD; Hays, SM; et al. (2001) A physiologically based model for the ingestion of trivalent
chromium and hexavalent chromium by humans. Toxicol Sci 60(2):196–213.
O’Flaherty, E.J.; Radike, M.J. (1991) Pharmacokinetic modeling of trivalent and hexavalent chromium based on
ingestion and inhalation of soluble chromium compounds. Armstrong Laboratory, Wright-Patterson Air Force Base,
Dayton, OH. Final report. AL-TR-1991-0162.
Ohno, H; Hanaoka, F; Yamada, M. (1982) Inducibility of sister-chromatid exchanges by heavy-metal ions. Mutat
Res 104(1-3):141–145.
Ohsaki, Y; Abe, S; Kimura, K; et al. (1978) Lung cancer in Japanese chromate workers. Thorax 33:372-374.
Olivier, P; Marzin, D. (1987) Study of the genotoxic potential of 48 inorganic derivatives with the SOS chromotest.
Mutat Res 189(3):263–269.
O'Neil, MJ; Heckelman, PE; Koch, CB; et al. (2006) Chromium and compounds. In: O'Neil, MJ; Heckelman, PE;
Koch, CB; et al.; eds. The Merck index. Whitehouse Station, NJ: Merck & Co. Inc., pp. 86, 270, 371–372, 937,
1315–1316.
Quivryn, G.; Goulart, M.; Messer, J.; 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.
Page, BJ; Loar, GW. (2004) Chromium compounds. In: Kirk-Othmer encyclopedia of chemical technology. New
York, NY: John Wiley & Sons, pp. 526–571.
Partington, CN. (1950) Acute poisoning with potassium bichromate. Br Med J 2(4688):1097–1098.
Paschin, YV; Toropzev, SN. (1982) Chromosome damage induced in vivo by heavy metal ion detected by indirect
testing. Acta Biol Acad Sci Hung 33(4):419–422.
Paschin, YV; Zacepilova, TA; Kozachenko, VI. (1982) Induction of dominant lethal mutations in male mice by
potassium dichromate. Mutat Res 103(3-6):345–347.
Paschin, YV; Kozachenko, VI; Sal'nikova, LE. (1983) Differential mutagenic response at the HGPRT locus in V79
and CHO Chinese hamster cells after treatment with chromate. Mutat Res 122(3-4):361–365.

Patlolla, AK; Tchounwou, PB. (2006) In vivo genotoxic effect of hexavalent chromium in rat leukocytes using
comet assay. Toxicol Sci 90(1-S):106.
Patlolla, A; Armstrong, N; Tchounwou, PB. (2008) Cytogenetic evaluation of potassium dichromate toxicity in bone
marrow cells of Sprague-Dawley rats. In: Collery P; Maymard, I; Theophanides, T; et al., eds. Metal ions in biology
and medicine. Paris: John Libbey Eurotext, pp. 353–358.
Paustenbach, DJ; Hays, SM; Brien, BA; et al. (1996) Observation of steady state in blood and urine following
human ingestion of hexavalent chromium in drinking water. J Toxicol Environ Health 49(5):453–461.
Paustenbach, DJ; Panko, JM; Fredrick, MM; Finley, BL; Proctor, DM. (1997) Urinary chromium as a biological
maker of environmental exposure: What are the limitations? Reg Toxicol Pharm 26:S23–S34.
Paustenbach, DJ; Finley, BL; Mowat, FS; et al. (2003) Human health risk and exposure assessment of hexavalent
chromium in tap water. J Toxicol Environ Health A 66(14):1295–1339.
Payne, WW. (1960a) The role of roasted chromite ore in the production of cancer. Arch Environ Health 1:20-26.
Payne, WW. (1960b) Production of cancers in mice and rats by chromium compounds. Arch Ind Health 21:530-535.
Pechova, A.; Pavlata, L. (2007) Chromium is an essential nutrient: A review. Veterinari Medicina 52(1):1 – 18.
Pellerin, C; Booker, SM. (2000) Reflections of hexavalent chromium. Environ Health Perspect 108(9):A402–A407.
Peterson-Roth, E; Reynolds, M; Quievryn, G; et al. (2005) Mismatch repair proteins are activators of toxic
responses to chromium-DNA damage. Mol Cell Biol 25(9):3596–3607.
Petrilli, F.L.; de Flora, S. (1982) Interpretations on chromium mutagenecity and carcinogenicity. In: Sorsa, M.;
Vainio, H. (Eds.) Mutagens in our environment. New York, Alan R. Liss, Inc. pp. 453 – 464
Petrilli, FL; DeFlora, S. (1977) Toxicity and mutagenicity of hexavalent chromium on Salmonella typhimurium.
Appl Environ Microbiol 33:805–809.
Petrilli, FL; Rossi, GA; Camoirano, A; et al. (1986) Metabolic reduction of chromium by alveolar macrophages and
its relationships to cigarette smoke. J Clin Invest 77(6):1917–1924.
Plessi, M; Monzani, A. (1990) Survey of total and bioavailable chromium in grain and cereal by atomic absorption
spetrophotometry. J Assoc Off Anal Chem 73(5):7987-800.
Pool-Zobel, BL; Lotzmann, N; Knoll, M; et al. (1994) Detection of genotoxic effects in human gastric and nasal
mucosa cells isolated from biopsy samples. Environ Mol Mutagen 24(1):23–45.
Pratt, PF; Myers, CR. (1993) Enzymatic reduction of hexavalent chromium by human hepatic microsomes.
Quievryn, G; Peterson, E; Messer, J; et al. (2003) Genotoxicity and mutagenicity of chromium(VI)/ascorbate-

generated DNA adducts in human and bacterial cells. Biochemistry 42(4):1062–1070.
Quinteros, FA; Poliandri, AH; Machiavelli, LI; et al. (2007) In vivo and in vitro effects of chromium VI on anterior
pituitary hormone release and cell viability. Toxicol Appl Pharmacol 218(1):79–87.
Rafael, AI; Almeida, A; Santos, P; et al. (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(2):163–173.
Rai, D; Eary, LE; Zachara, JM. (1989) Environmental chemistry of chromium. The Science of the Total
Environment 86:15 –23.

Rasmuson, A. (1985) Mutagenic effects of some water-soluble metal compounds in a somatic eye-color test system
in Drosophila melanogaster. Mutat Res 157(2-3):157–162.
Reichelderfer, TE. (1968) Accidental death of an infant caused by ingestion of ammonium dichromate. South Med J
61:96–97.
Rodriguez-Arnaiz, R; Martinez, RFM. (1986) Genetic effects of potassium dichromate and chromium trioxide in
Drosophila melanogaster. Cytologia (Tokyo) 51:421–425.
Rowbotham, AL; Levy, LS; Shuker, LK. (2000) Chromium in the environment: An evaluation of exposure of the
UK general population and possible adverse health effects. J Toxicol Environ Health B Crit Rev 3(3):145-178.
Royle, H. (1975) Toxicity of chromic acid in the chromium plating industry. Environ Res 10:141-163.
Ryden, E; Ekstrom, C; Hellmer, L; et al. (2000) Comparison of the sensitivities of Salmonella typhimurium strains
TA102 and TA2638A to 16 mutagens. Mutagenesis 15(6):495–502.
Salnikow, K; Zhitkovich, A. (2008) Genetic and epigenetic mechanisms in metal carcinogenesis and
cocarcinogenesis: nickel, arsenic and chromium. Chem Res Toxicol 21:28–44.
Salnikow, K; Zhitkovich, A; Costa, M. (1992) Analysis of the binding sites of chromium to DNA and protein in
vitro and in intact cells. Carcinogenesis 13(12):2341–2346.
Salnikov, K.; Zhitkovich, A. (2008) Genetic and epigenetic mechanisms in metal carcinogenesis and
cocarcinogenesis: Nickel, arsenic, and chromium. Chem Res Toxicol 21(1):28 – 44.
Sander, M; Cadet, J; Casciano, DA; et al. (2005) Proceedings of a workshop on DNA adducts: biological
significance and applications to risk assessment Washington, DC, April 13-14, 2004. Toxicol Appl Pharmacol
208:1–20.
Sano, T; Mitohara, I. (1978) Occupational cancer among chromium workers. Jpn J Chest Dis 37(2):90-101.
Sarto, F; Cominato, I; Bianchi, V; et al. (1982) Increased incidence of chromosomal aberrations and sister chromatid
exchanges in workers exposed to chromic acid (CrO3) in electroplating factories. Carcinogenesis 3(9):1011–1016.
Sarto, F; Tomanin, R; Giacomelli, L; et al. (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.
Saryan, LA; Reedy, M. (1988) Chromium determinations in a case of chromic acid ingestion. J Anal Toxicol
12(3):162–164.
Satoh, K; Fukuda, Y; Torii, K; et al. (1981) Epidemiologic study of workers engaged in the manufacture of
chromium compounds. J Occup Med 23(12):835-838.
Sayato, Y; Nakamuro, K; Matsui, S; et al. (1980) Metabolic fate of chromium compounds. I. Comparative behavior
of chromium in rat administered with NA251Cr4 and 51CrCl3. J Pharmacobiodyn 3(1):17–23.
Sekihashi, K; Sasaki, T; Yamamoto, A; et al. (2001) A comparison of intraperitoneal and oral gavage administration
in comet assay in mouse eight organs. Mutat Res 493(1-2):39–54.
Semple, KT; Doick, KJ; Jones, KC; et al. (2004) Defining bioavailability and bioaccessibility of contaminated soil
and sediment is complicated. Environ Sci Technol. Jun 15; 38(12):228A–231A.
Sen, P; Conway, K; Costa, M. (1987) Comparison of the localization of chromosome damage induced by calcium
chromate and nickel compounds. Cancer Res 47(8):2142–2147.
Seo, KS; Lee, CS. (1993) Mutagenicity of metal compounds using Escherichia. Misaengmul Hakhoechi 31(6):527–
531.

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(4):551–555.
Shanker, AK; Cervantes, C; Loza-Tavera, H; et al. (2005) Chromium toxicity in plants. Environ Int 31:739–753.
Sharma, BK; Singhal, PC; Chugh, KS. (1978) Intravascular haemolysis and acute renal failure following potassium
dichromate poisoning. Postgrad Med J 54(632):414–415.
Sharma, S; Kelly, TK; Jones, PA (2010) Epigenetics in cancer. Carcinogenesis 31(1):27-36.
Shi, X; Dalal, NS. (1990) NADPH-dependent flavoenzymes catalyze one electron reduction of metal ions and
molecular oxygen and generate hydroxyl radicals. FEBS Lett 276(1–2):189–191.
Shi, X; Mao, Y; Knapton, AD; et al. (1994) Reaction of Cr(VI) with ascorbate and hydrogen peroxide generates
hydroxyl radicals and causes DNA damage: role of a Cr(IV)-mediated Fenton-like reaction. Carcinogenesis
15(11):2475–2478.
Shi, X; Ding, M; Ye, J; et al. (1999) Cr(IV) causes activation of nuclear transcription factor κB, DNA strand breaks
and dG hydroxylation via free radical reactions. J Inorg Biochem 75:37–44.
Shindo, Y; Toyoda, Y; Kawamura, K; et al. (1989) Micronucleus test with potassium chromate(VI) administered
intraperitoneally and orally to mice. Mutat Res 2223(4):403–406.
Shupack, SI. (1991) The chemistry of chromium and some resulting analytical problems. Environ Hlth Perspect
92:7–11.
Singh, I. (1983) Induction of reverse mutation and mitotic gene conversion by some metal compounds in
Saccharomyces cerevisiae. Mutat Res 117(1–2):149–152.
Smith, AH. (2008) Hexavalent chromium, yellow water, and cancer: a convoluted saga. Epidemiology 19(1):24–26.
Snow, ET. (1991) A possible role for chromium(III) in genotoxicity. Environ Health Perspect 92:75–81.
Sora, S; Agostoni Carbone, ML; Pacciarini, M; et al. (1986) Disomic and diploid meiotic products induced in
Saccharomyces cerevisiae by the salts of 27 elements. Mutagenesis 1(1):21–28.
Sorahan, T; Burgess, DC; Waterhouse, JA. (1987) A mortality study of nickel/chromium platers. Br J Ind Med
44:250-258.
Spano, MA; Frei, H; Wurgler, FE; et al. (2001) Recombinagenic activity of four compounds in the standard and
high bioactivation crosses of Drosophila melanogaster in the wing spot test. Mutagenesis 16(5):385–394.
Standeven, AM; Wetterhahn, KE. (1989) Chromium(VI) toxicity: uptake, reduction, and DNA damage. J Am Coll
Toxicol 8(7):1275–1283.
Standeven, A.M.; Wetterhahn, K.E. (1991) Ascorbate is the principal reductant of chromium(VI) in rat liver and
kidney ultrafiltrates. Carcinogenesis 12:1733 – 1737.
Standeven, A.M.; Wetterhahn, K.E. (1992) Ascorbate is the principal reductant of chromium(VI) in rat lung
ultrafiltrates and cytosols, and mediates chromium-DNA binding in vitro. Carcinogenesis 13:1319 – 1324.
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.
Stearns, DM; Courtney, KD; Giangrade, PH; et al. (1994) Hexavalent chromium reduction by ascorbate: role of
reactive intermediates in DNA damage in vitro. Environ Health Perspect 102(Suppl 3):21–25.
Stearns, D.M. (1999) Is chromium a trace essential metal? (2000) BioFactors 11:149 – 162.

Steinhoff, S; Gad, SC; Hatfield, GK; et al. (1983) Listing sodium dichromate and soluble calcium chromate for
carcinogenicity in rats. Bayer AG Institute of Toxicology. (Unpublished)
Stella, M; Montaldi, A; Rossi, R; et al. (1982) Clastogenic effects of chromium on human lymphocytes in vitro and
in vivo. Mutat Res 101(2):151–164.
Stout, MD; Herbert, RA; Kissling, GE; et al. (2009) Hexavalent chromium is carcinogenic to F344/N rats and
B6C3F1 mice after chronic exposure. Environ Health Perspect 117(5):716–722.
Subramanian, S; Rajendiran, G; Sekhar, P; et al. (2006) Reproductive toxicity of chromium in adult bonnet monkeys
(Macaca radiata Geoffrey). Reversible oxidative stress in the semen. Toxicol Appl Pharmacol 215(3):237–249.
Sugden, KD; Martin BD. (2002) Guanine and 7,8-dihydro-8-oxo-guanine-specific oxidation in DNA by
chromium(V). Environ Health Perspect 110(Suppl 5):725–728.
Sugden, KD; Stearns, DM. (2000) The role of chromium(V) in the mechanism of chromate-induced oxidative DNA
damage and cancer. J Environ Pathol Toxicol Oncol 19(3):215–230.
Sugiyama, M; Ando, A; Nakao, K; et al. (1989) Influence of vitamin B2 on formation of chromium(V), alkali-labile
sites, and lethality of sodium chromate(VI) in Chinese hamster V-79 cells. Cancer Res 49:6180–6184.
Sutherland, JE; Zhitkovich, A; Kluz, T; et al. (2000) Rats retain chromium in tissues following chronic ingestion of
drinking water containing hexavalent chromium. Biol Trace Elem Res 74:41–53.
Suzuki, Y; Fukuda, K. (1990) Reduction of hexavalent chromium by ascorbic acid and glutathione with special
reference to the rat lung. Arch Toxicol 93:9 – 14.
Tagliari, KC; Cecchini, R; Vargas, VMF. (2004) Mutagenicity of chromium (VI) using the Salmonella
microsuspension bioassay. Rev Bras Toxicol 17(2):45–50.
Takahashi, Y; Kondo, K; Hirose, T; et al. (2005) Microsatellite instability and protein expression of the DNA
mismatch repair gene, hMLH1, of lung cancer in chromate-exposed workers. Mol Carcinog 42(3):150–158.
Taylor, FH. (1966) The relationship of mortality and duration of employment as reflected by a cohort of chromate
workers. Am J Public Health 56(2):218-229.
Thompson, RC; Hollis, OL. (1958) Irradiation of the gastrointestinal tract of the rat by ingested ruthenium-106. Am
J Physiol 194(2):308–312.
Tkaczyk, C.; Huk, O.L.; Mwale, F.; Antoniou, J.; Zukor, D.J.; Petit, A.; Tabrizian, M. (2010) Investigation of the
binding of Cr(III) complexes to bovine and human serum proteins: A proteomic approach. J Biomed Mater Res
94A:214 – 222.
Todd, GE. (1962) Tobacco manufacturer's standing committee research papers. No. 1. Statistics of Smoking in the
United Kingdom, 3rd ed. Tobacco Research Council, London.
Trivedi, B; Saxena, DK; Murthy, RC; et al. (1989) Embryotoxicity and fetotoxicity of orally administered
hexavalent chromium in mice. Reprod Toxicol 3(4):275–278.
Trzeciak, A; Kowalik, J; Malecka-Panas, E; et al. (2000) Genotoxicity of chromium in human gastric mucosa cells
and peripheral blood lymphocytes evaluated by the single cell gel electrophoresis (comet assay). Med Sci Monit
6(1):24–29.
Tsapakos, MJ; Hampton, TH; Wetterhahn, KE. (1983) Hexavalent chromium-induced DNA lesions and chromium
distribution in rat kidney, liver, and lung. Cancer Res 43(12 Pt 1):5662–5667.
Ueno, S; Kashimoto, T; Susa, N; et al. (2001) Detection of dichromate(VI)-induced DNA strand breaks and
formation of paramagnetic chromium in multiple mouse organs. Toxicol Appl Pharmacol 170(1):56–62.

Umeda, M; Nishimura, M. (1979) Inducibility of chromosomal aberrations by metal compounds in cultured
mammalian cells. Mutat Res 67(3):221–229.
U.S. EPA (Environmental Protection Agency). (1984) Health assessment document for chromium. Final report.
Cincinnati, OH: Environmental Criteria and Assessment Office. EPA600883014F.
U.S. EPA. (1986a) Guidelines for the health risk assessment of chemical mixtures. Federal Register 51(185):34014-
34025. Available online at http://www.epa.gov/iris/backgrd.html (accessed April 20, 2010).
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register 51(185):34006-34012. Available
online at http://www.epa.gov/iris/backgrd.html (accessed April 20, 2010).
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk assessment.
Prepared by the Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment,
Cincinnati, OH for the Office of Solid Waste and Emergency Response, Washington, DC; EPA 600/6-87/008.
Available online at http://www.epa.gov/iris/backgrd.html (accessed April 20, 2010).
U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Federal Register 56(234):63798-63826.
U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation toxicity studies.
Federal Register 59(206):53799. Available online at http://www.epa.gov/iris/backgrd.html (accessed April 20,
2010).
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application of inhalation
dosimetry. Office of Research and Development, Washington, DC; EPA/600/8-90/066F. Available online at
http://www.epa.gov/iris/backgrd.html (accessed April 20, 2010).
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/R-94/007. Available online at
http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=42601 (accessed March 11, 2010).
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register 61(212):56274–56322.
U.S. EPA. (1998) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926–26954. Available
online at http://www.epa.gov/iris/backgrd.html (accessed April 20, 2010).
U.S. EPA. (2000a) Science policy council handbook: risk characterization. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA 100-B-00-002. Available online at
U.S. EPA. (2000b) Benchmark dose technical guidance document. External review draft. Risk Assessment Forum,
Washington, DC; EPA/630/R-00/001. Available online at http://www.epa.gov/iris/backgrd.html (accessed April 20,
2010).
U.S. EPA. (2000c) Supplementary guidance for conducting for health risk assessment of chemical mixtures. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/002. Available online at http://www.epa.gov/iris/backgrd.html
(accessed April 20, 2010).
U.S. EPA. (2002) A review of the reference dose and reference concentration processes. Risk Assessment Forum,
Washington, DC; EPA/630/P-02/0002F. Available online at http://www.epa.gov/iris/backgrd.html (accessed April
20, 2010).
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum, Washington, DC;
EPA/630/P-03/001B. Available online at http://www.epa.gov/iris/backgrd.html (accessed April 20, 2010).

U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure to carcinogens. Risk
Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available online at
U.S. EPA. (2006a) Science policy council handbook: peer review. Third edition. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA/100/B-06/002. Available online at
U.S. EPA. (2006b) A Framework for Assessing Health Risk of Environmental Exposures to Children. National
Center for Environmental Assessment, Washington, DC, EPA/600/R-05/093F. Available online at
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=158363 (accessed March 11, 2010).
Vaglenov, A; Nosko, M; Georgieva, R; et al. (1999) Genotoxicity and radioresistance in electroplating workers
exposed to chromium. Mutat Res 446(1):23–34.
Valko, M; Rhodes, CJ; Moncol, J; et al. (2006) Free radicals, metals and antioxidants in oxidative stress-induced
cancer. Chem Biol Interact 160:1–40.
Vashishat, RK; Vasudeva, M. (1987) Genotoxic potential of chromium salts in Saccharomyces cerevisiae. Ind J
Microbiol 27:35–36.
Venier, P; Montaldi, A; Majone, F; et al. (1982) Cytotoxic, mutagenic and clastogenic effects of industrial
chromium compounds. Carcinogenesis 3(11):1331–1338.
Venier, P; Gava, C; Zordan, M; et al. (1987) Interactions of chromium with nitrilotriacetic acid (NTA) in the
induction of genetic effects in bacteria. Toxicol Environ Chem 14(3):201–218.
Venitt, S; Levy, LS. (1974) Mutagenicity of chromates in bacteria and its relevance to chromate carcinogenesis.
Nature 250:493–495.
Vincent, J.B. (2004) Recent developments in the biochemistry of chromium(III). Biol Trace Elem Res 99(1 – 3):1
– 16.
Vitale, RJ; Mussoline, GR; Rinehimer, KA. (1997) Environmental monitoring of chromium in air, soil and water.
Regul Toxicol Pharmacol 26(1 Pt 2):S80-5.
Voitkun, V; Zhitkovich, A; Costa, M. (1998) Cr(III)-mediated crosslinks of glutathione or amino acids to the DNA
phosphate backbone are mutagenic in human cells. Nucleic Acids Res 26(8):2024–2030.
Vyskocil, A; Viau, C; Cizkova, M; et al. (1993) Kidney function in male and female rats chronically exposed to
potassium dichromate. J Appl Toxicol 13(5):375–376.
Wang, XF; Xing, ML; Shen, Y; et al. (2006) Oral administration of Cr(VI) induced oxidative stress, DNA damage
and apoptotic cell death in mice. Toxicology 228(1):16–23.
Wang, S; Chen, F; Zhang, Z; et al. (2004) NF-kB prevents cells from undergoing Cr(VI)-induced apoptosis. Mol
Cell Biochem 255:129-137.
Watanabe, K; Sakamoto, K; Sasaki, T. (1998) Comparisons on chemically-induced mutation among four bacterial
strains, Salmonella typhimurium TA102 and TA2638, and Escherichia coli WP2/pKM101 and WP2 uvrA/pKM101:
collaborative study II. Mutat Res 412(1):17–31.
Watanabe, S; Fukuchi, Y. (1975) An epidemiological survey on lung cancer in workers of a chromate-producing

industry in Hokkaido, Japan. Presented at International Congress on Occupational Health.
Weber, H. (1983) Long-term study of the distribution of soluble chromate-51 in the rat after a single intratracheal
administration. J Toxicol Environ Health 11(4-6):749–764.

Wiegand, HJ; Ottenwalder, H; Bolt, HM. (1984) The reduction of chromium (VI) to chromium (III) by glutathione:
an intracellular redox pathway in the metabolism of the carcinogen chromate. Toxicology 33(3–4):341–348.
Wiegand, HJ; Ottenwalder, H; Bolt, HM. (1985) Fast uptake kinetics in vitro of 51Cr(VI) by red blood cells of man
and rat. Arch Toxicol 57(1):31–34.
Wiegand, HJ; Ottenwalder, H; Bolt, HM. (1987) Bioavailability and metabolism of hexavalent chromium
compounds. Toxicol Environ Chem 14:263–275.
Wild, D. (1978) Cytogenetic effects in the mouse of 17 chemical mutagens and carcinogens evaluated by the
micronucleus test. Mutat Res 56(3):319–327.
Wise, JP, Sr.; Wise, SS; Little, JE. (2002) The cytotoxicity and genotoxicity of particulate and soluble hexavalent
chromium in human lung cells. Mutat Res 517(1–2):221–229.
Wise, SS; Holmes, AL; Ketterer, ME; et al. (2004) Chromium is the proximate clastogenic species for lead
chromate-induced clastogenicity in human bronchial cells. Mutat Res 560(1):79–89.
Wise, SS; Holmes, AL; Wise, JP, Sr. (2006a) Particulate and soluble hexavalent chromium are cytotoxic and
genotoxic to human lung epithelial cells. Mutat Res 610(1–2):2–7.
Wise, SS; Holmes, AL; Xie, H; et al. (2006b) Chronic exposure to particulate chromate induces spindle assembly
checkpoint bypass in human lung cells. Chem Res Toxicol 19(11):1492–1498.
Wise, SS; Holmes, AL; Wise, JP, Sr. (2008) Hexavalent chromium-induced DNA damage and repair mechanisms.
Rev Environ Health 23(1):39–57.
Witmer, C.M.; Harris, R/; Shupack, S.I. (1991) Oral bioavailability of chromium from a specific site. Environ
Health Perspect 92:105 – 110.
Witmer, C.M.; Park, H-S. (1989) Mutagenicity and disposition of chromium. Sci. Total Environ 86: 131 – 148.
Wronska-Nofer, T; Wisniewska-Knypl, J; Wyszyñska, K. (1999) Prooxidative and genotoxic effect of transition

metals (cadmium, nickel, chromium, and vanadium) in mice. Trace Elem Electrolytes 15(2):87–92.
Wu, FY; Tsai, FJ; Kuo, HW; et al. (2000) Cytogenetic study of workers exposed to chromium compounds. Mutat
Res 464(2):289–296.
Wu, FY; Wu, WY; Kuo, HW; et al. (2001) Effect of genotoxic exposure to chromium among electroplating workers
in Taiwan. Sci Total Environ 279(1–3):21–28.
Xu, J; Bubley, GJ; Detrick, B; et al. (1996) Hexavalent chromium treatment of normal human lung cells results in
guanine-specific DNA polymerase arrest, DNA-DNA cross-links and S-phase blockade of cell cycle.
Yamamoto, A; Kohyama, Y; Hanawa, T. (2002) Mutagenicity evaluation of forty-one metal salts by the umu test. J
Biomed Mater Res 59(1):176–183.
Ye, J; Wang, S; Leonard, SS; et al. (1999) Role of reactive oxygen species and p53 in chromium(VI)-induced
apoptosis. J Biol Chem 274(49):34974–34980.
Yousef, MI; El-Demerdash, FM; Kamil, KI; et al. (2006) Ameliorating effect of folic acid on hexavalent chromium-
induced changes in reproductive performance and seminal plasma biochemistry in male rabbits. Reprod Toxicol
21(3):322–328.
Zahid, ZR; Al-Hakkak, ZS; Kadhim, AHH; et al. (1990) Comparative effects of trivalent and hexavalent chromium
on spermatogenesis of the mouse. Toxicol Environ Chem 25:131–136.

Zeiger, E; Anderson, B; Haworth, S; et al. (1992) Salmonella mutagenicity tests: V. Results from the testing of 311
chemicals. Environ Mol Mutagen 19(Suppl 21):2–141.
Zhang, JD; Li, XL. (1987) Chromium pollution of soil and water in Jinzhou. Zhonghua Yu Fang Yi Xue Za Zhi
21(5):262–264. (Chinese)
Zhang, JD; Li, S. (1997) Cancer mortality in a Chinese population exposed to hexavalent chromium in water. J
Occup Environ Med 39(4):315–319.
Zhang, Q; Kluz, T; Salnikow, K; et al. (2002) Comparison of the cytotoxicity, cellular uptake, and DNA-protein
crosslinks induced by potassium chromate in lymphoblast cell lines derived from three different individuals. Biol
Trace Elem Res 86:11–22.
Zhitkovich, A. (2005) Importance of chromium-DNA adducts in mutagenicity and toxicity of chromium(VI). Chem
Res Toxicol 18(3):3–11.
Zhitkovich, A; Voitkun, V; Costa, M. (1995) Glutathione and free amino acids form stable complexes with DNA
following exposure of intact mammalian cells to chromate. Carcinogenesis 16(4):907–913.
Zhitkovich, A; Song, Y; Quievryn, G; et al. (2001) Non-oxidative mechanisms are responsible for the induction of
mutagenesis by reduction of Cr(VI) with cysteine: role of ternary DNA adducts in Cr(III)-dependent mutagenesis.
Biochemistry 40:549–560.
Zimmering, S; Mason, JM; Valencia, R; et al. (1985) Chemical mutagenesis testing in Drosophila. II. Results of 20
coded compounds tested for the National Toxicology Program. Environ Mutagen 7(1):87–100.

APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION
[page intentionally left blank]
A-1 DRAFT - DO NOT CITE OR QUOTE

APPENDIX B. BENCHMARK DOSE CALCULATIONS
B.1. DETAILS OF BMD ANALYSIS FOR THE RFD
Table B-1. Incidence data for nonneoplastic lesions from all treatment
groups of female F344/N rats and male and female B6C3F1 mice exposed to
sodium dichromate dihydrate in drinking water for 2 years (NTP, 2008)
Dose
0 0.24 0.94 2.4 7.0
Female rats
Liver: chronic inflammation 12/50 21/50a 28/50b 35/50b 39/50b
Dose
0 0.38 0.91 2.4 5.9
Male mice
Duodenum: diffuse epithelial 0/50 11/50b 18/50b 42/50b 32/50a
hyperplasia
Mesenteric lymph node: histiocytic 14/47 38/47b 31/49b 32/49b 42/46a
Dose
0 0.38 1.4 3.1 8.7
Female mice
Duodenum: diffuse epithelial 0/50 16/50b 35/50b 31/50b 42/50b
hyperplasia
Mesenteric lymph node: histiocytic 3/46 29/48b 26/46b 40/50b 42/50b
Liver: histiocytic cellular infiltration 2/49 15/50b 23/50b 32/50b 45/50b
Pancreas: acinus, cytoplasmic 0/48 6/50a 6/49a 14/50b 32/50b
alteration
a
b
Chronic inflammation of the liver in female rats. As assessed by the χ2 goodness-of-fit

statistic, only the log-logistic model provided an adequate fit (χ2 p-value ≥ 0.1) to the data
(Table B-2). Based on the log-logistic model, the BMD associated with a 10% extra risk was
0.22 mg hexavalent chromium/kg-day and its lower 95% confidence limit (BMDL) was 0.14 mg
hexavalent chromium/kg-day (Figure B-1).
B-1 DO NOT CITE OR QUOTE

Table B-2. BMD10 and BMDL10 values and goodness-of-fit statistics from
models fit to incidence data for chronic inflammation of the liver in female
rats exposed to sodium dichromium dihydrate in drinking water for 2 years
BMD10 BMDL10
Model (mg/kg-d) (mg/kg-d) χ2 p-value AIC
a
Gamma 0.51 0.37 0.04 317.97
Logistic 0.84 0.65 0.01 321.45
Log-logisticb 0.22 0.14 0.37 312.57
Multistagec 0.51 0.37 0.04 317.97
Probit 0.88 0.70 0.01 321.80
Log-probitb 0.89 0.61 0.01 320.86
Quantal linear 0.51 0.37 0.04 317.97
Weibulla 0.51 0.37 0.04 317.97
a
Restrict power ≥1.
b
Slope restricted to >1.
c
Restrict betas ≥0; lowest degree polynomial (up to n-2) with an adequate fit is reported; no degree of polynomial
provided a fit, a 3-degree polynomial is reported.
AIC = Akaike’s information criterion

Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
0.9
0.8
0.7
Fraction Affected
0.6
0.5
0.4
0.3
0.2
0.1 BMDL BMD
0 1 2 3 4 5 6 7
dose
13:34 04/08 2008
BMDs and BMDLs indicated are associated with a 10% extra risk, and are in
units of mg hexavalent chromium/kg-day.
Figure B-1. Predicted and observed incidence of chronic inflammation of the

liver in female rats exposed to sodium dichromium dihydrate in drinking
water for 2 years.
Diffuse epithelial hyperplasia of the duodenum in male mice. As assessed by the

χ goodness-of-fit statistic, none of the models provided an adequate fit (χ2 p-value ≥ 0.1) to the
2
full dataset (Table B-3). In order to achieve an adequately fitting model, the highest dose was
dropped. This is determined to be appropriate, as the area of concern is with the low-dose region
of the response curve. After dropping the highest dose, the gamma, log-logistic, multistage, log-
probit, quantal linear, and Weibull models provided adequate fits to the data (χ2 p-value > 0.1).
Comparing across models, a better fit is generally indicated by a lower Akaike’s Information
Criterion (AIC) (EPA, 2000b). As assessed by AIC, the 1-degree polynomial multistage model
provided the best fit to the data (Figure B-2). Based on the multistage model, the BMD
associated with a 10% extra risk was 0.16 mg hexavalent chromium/kg-day and its lower 95%
confidence limit (BMDL) was 0.13 mg hexavalent chromium/kg-day.

models fit to incidence data for diffuse epithelial hyperplasia in the
duodenum in male mice exposed to sodium dichromium dihydrate in
BMD10 BMDL10
All doses
Gammaa 0.31 0.25 0.00 270.99
Logistic 0.90 0.74 0.00 296.25
Log-logisticb 0.15 0.12 0.00 247.93
Multistagec 0.31 0.25 0.00 270.99
Probit 0.90 0.76 0.00 296.18
Log-probitb 0.48 0.36 0.00 274.38
Quantal linear 0.31 0.25 0.00 270.99
Weibulla 0.31 0.25 0.00 270.99
Highest dose dropped (four doses modeled)
Gammaa 0.22 0.14 0.43 167.67
Logistic 0.47 0.39 0.03 177.09
Log-logisticb 0.26 0.15 0.20 169.23
Multistaged 0.16 0.13 0.52 166.34
Probit 0.45 0.37 0.04 176.19
Log-probitb 0.28 0.23 0.33 167.41
Quantal linear 0.16 0.13 0.52 166.34
Weibulla 0.22 0.14 0.47 167.50
a
b
c
d
Restrict betas ≥0; lowest degree polynomial (up to n-2) with an adequate fit is reported; degree polynomial = 1.

Multistage Model with 0.95 Confidence Level
1
Multistage
0.8
0.6
Fraction Affected
0.4
0.2
0
BMDL BMD
0 0.5 1 1.5 2 2.5
dose
15:45 04/08 2008
Figure B-2. Predicted and observed incidence of diffuse epithelial

hyperplasia in the duodenum of male mice exposed to sodium dichromium
dihydrate in drinking water for 2 years.
Histiocytic cellular infiltration of the mesenteric lymph nodes in male mice. As assessed
by the χ2 goodness-of-fit statistic, none of the models provided an adequate fit (χ2 p-value ≥ 0.1)
to the full dataset (Table B-4). In order to achieve a statistically fit model, the highest dose was
dropped. This is determined to be appropriate, as the area of concern is with the low-dose region
of the response curve. Dropping the highest dose did not result in adequately fitting models, nor
did dropping the two highest doses. This dataset is considered not suitable for BMD modeling.

models fit to incidence data for histiocytic cellular infiltration in mesenteric
lymph nodes of male mice exposed to sodium dichromium dihydrate in
BMD10 BMDL10
All doses
Gammaa 0.38 0.26 0.00 285.94
Logistic 0.53 0.39 0.00 286.38
Log-logisticb 0.16 0.08 0.00 284.48
Multistagec 0.43 0.26 0.00 287.88
Probit 0.56 0.43 0.00 286.35
Log-probitb 0.83 0.52 0.00 289.36
Quantal linear 0.38 0.26 0.00 285.94
Weibulla 0.38 0.26 0.00 285.94
Gammaa 0.47 0.24 0.00 258.50
Logistic 0.61 0.35 0.00 259.04
Log-logisticb 0.21 0.08 0.00 256.81
Multistaged 0.47 0.24 0.00 258.50
Probit 0.63 0.37 0.00 259.08
Log-probitb 1.24 0.56 0.00 261.28
Quantal linear 0.47 0.24 0.00 258.50
Weibulla 0.47 0.24 0.00 258.50
Two highest doses dropped (three doses modeled)
Gammaa 0.11 0.07 0.00 187.77
Logistic 0.17 0.12 0.00 189.97
Log-logisticb 0.05 0.03 0.00 183.77
Multistagee 0.11 0.07 0.00 187.77
Probit 0.17 0.12 0.00 190.12
Log-probitb 0.17 0.11 0.00 190.37
Quantal linear 0.11 0.07 0.00 187.77
Weibulla 0.11 0.07 0.00 187.77
a
b
c
d
e
Diffuse epithelial hyperplasia of the duodenum in female mice. As assessed by the

χ goodness-of-fit statistic, none of the models provided an adequate fit (χ2 p-value ≥ 0.1) to the
2

data (Table B-5). In order to achieve a statistically fit model, the highest dose was dropped.
This is determined to be appropriate, as the area of concern is with the low-dose region of the
response curve. After dropping the highest dose, an adequate fit was still not achieved. After
dropping the two highest doses, all of the models except for the logistic and probit models
provided an adequate fit (χ2 p-value ≥ 0.1) to the data. Comparing across models, a better fit is
generally indicated by a lower AIC (EPA, 2000b). As assessed by AIC, the gamma, multistage,
quantal linear, and Weibull models generated identical goodness of fit statistics and BMD, as
these models all took the form of a 1-degree polynomial multistage model, which provides the
best fit (Figure B-3). Based on these models, the BMD associated with a 10% extra risk was
0.12 mg hexavalent chromium/kg-day and its lower 95% confidence limit (BMDL) was 0.09 mg
hexavalent chromium/kg-day.

models fit to incidence data for diffuse epithelial hyperplasia in the
duodenum of female mice exposed to sodium dichromium dihydrate in
BMD10 BMDL10
All doses
Gammaa 0.34 0.27 0.00 275.34
Logistic 0.88 0.72 0.00 293.17
Log-logisticb 0.12 0.09 0.04 245.54
Multistagec 0.34 0.27 0.00 275.34
Probit 0.93 0.78 0.00 294.03
Log-probitb 0.52 0.38 0.00 279.54
Quantal linear 0.34 0.27 0.00 275.34
Weibulla 0.34 0.27 0.00 275.34
Gammaa 0.20 0.16 0.00 213.41
Logistic 0.55 0.46 0.00 236.10
Log-logisticb 0.11 0.08 0.04 200.07
Multistaged 0.20 0.16 0.00 213.41
Probit 0.54 0.45 0.00 235.61
Log-probitb 0.29 0.24 0.00 220.04
Quantal linear 0.20 0.16 0.00 213.41
Weibulla 0.20 0.16 0.00 213.41
a
Gamma 0.12 0.09 0.87 126.06
Logistic 0.34 0.27 0.00 141.77
Log-logisticb 0.12 0.06 1.00 127.77
e
Multistage 0.12 0.09 0.87 126.06
Probit 0.32 0.26 0.00 140.65
Log-probitb 0.20 0.16 0.48 127.17
Quantal linear 0.12 0.09 0.87 126.06
Weibulla 0.12 0.09 0.87 126.06
a
b
c
d
e

Multistage Model with 0.95 Confidence Level
Multistage
0.8
0.7
0.6
Fraction Affected
0.5
0.4
0.3
0.2
0.1
0
BMDL BMD
0 0.2 0.4 0.6 0.8 1 1.2 1.4
dose
09:36 04/09 2008
Figure B-3. Predicted and observed incidence of diffuse epithelial

hyperplasia in the duodenum of female mice exposed to sodium dichromium
Histiocytic cellular infiltration of the mesenteric lymph nodes in female mice. As

assessed by the χ2 goodness-of-fit statistic, none of the models provided an adequate fit
(χ2 p-value ≥ 0.1) to the full dataset (Table B-6). In order to achieve a statistically fit model, the
highest dose was dropped. This is determined to be appropriate, as the area of concern is with
the low-dose region of the response curve. Dropping the highest dose did not result in
adequately fitting models, nor did dropping the two highest doses. This dataset is not suitable for
BMD modeling.

models fit to incidence data for histiocytic cellular infiltration in mesenteric
lymph nodes of female mice exposed to sodium dichromium dihydrate in
BMD10 BMDL10
All doses
Gammaa 0.41 0.30 0.00 282.46
Logistic 0.77 0.61 0.00 290.18
Log-logisticb 0.09 0.06 0.00 263.55
Multistagec 0.41 0.30 0.00 282.46
Probit 0.85 0.69 0.00 291.41
Log-probitb 0.68 0.47 0.00 285.85
Quantal linear 0.41 0.30 0.00 282.46
Weibulla 0.41 0.30 0.00 282.46
Gammaa 0.20 0.15 0.00 224.84
Logistic 0.40 0.33 0.00 230.81
Log-logisticb 0.07 0.05 0.00 215.19
Multistaged 0.20 0.15 0.00 224.84
Probit 0.40 0.34 0.00 230.85
Log-probitb 0.37 0.24 0.00 231.76
Quantal linear 0.20 0.15 0.00 224.84
Weibulla 0.20 0.15 0.00 224.84
a
Gamma 0.14 0.10 0.00 172.32
Logistic 0.31 0.24 0.00 178.99
Log-logisticb 0.07 0.04 0.00 164.47
Multistagee 0.14 0.10 0.00 172.32
Probit 0.30 0.23 0.00 178.74
Log-probitb 0.21 0.15 0.00 178.11
Quantal linear 0.14 0.10 0.00 172.32
Weibulla 0.14 0.10 0.00 172.32
a
b
c
d
e

Histiocytic cellular infiltration of the liver in female mice. As assessed by the
χ goodness-of-fit statistic, only the log-logistic model provided an adequate fit (χ2 p-value ≥
2
0.1) to the data (Table B-7). Based on the log-logistic model, the BMD associated with a 10%
extra risk was 0.17 mg hexavalent chromium/kg-day and its lower 95% confidence limit
(BMDL) was 0.12 mg hexavalent chromium/kg-day (Figure B-4).
models fit to incidence data for histiocytic cellular infiltration in the liver of
female rats exposed to sodium dichromium dihydrate in drinking water for
2 years
BMD10 BMDL10
a
Gamma 0.35 0.28 0.08 255.40
Logistic 0.85 0.70 0.00 267.56
Log-logisticb 0.17 0.12 0.44 251.36
Multistagec 0.35 0.28 0.08 255.40
Probit 0.88 0.75 0.00 268.64
Log-probitb 0.62 0.48 0.01 260.00
Quantal linear 0.35 0.28 0.08 255.40
Weibulla 0.35 0.28 0.08 255.40
a
b
c

1 Log-Logistic
0.8
Fraction Affected
0.6
0.4
0.2
0
BMDL BMD
0 1 2 3 4 5 6 7 8 9
dose
11:17 04/09 2008
Figure B-4. Predicted and observed incidence of histiocytic cellular

infiltration in the livers of female mice exposed to sodium dichromium
Cytoplasmic alteration of acinar epithelial cells of the pancreas in female mice. As

assessed by the χ2 goodness-of-fit statistic, all of the models provide adequate fits
(χ2 p-value ≥ 0.1) to the data (Table B-8). Comparing across models, a better fit is generally
indicated by a lower AIC (EPA, 2000b). As assessed by AIC, the log-logistic model provides
the best fit (Figure B-5). Based on the log-logistic model, the BMD associated with a 10% extra
risk was 0.68 mg hexavalent chromium/kg-day and its lower 95% confidence limit (BMDL) was
0.52 mg hexavalent chromium/kg-day.

models fit to incidence data for pancreas: acinus, cytoplasmic alteration in
female mice exposed to sodium dichromium dihydrate in drinking water for
2 years
BMD10 BMDL10
a
Gamma 0.92 0.72 0.13 206.82
Logistic 2.43 2.03 0.09 211.78
Log-logisticb 0.68 0.52 0.19 205.22
Multistagec 0.92 0.72 0.13 206.82
Probit 2.24 1.89 0.11 210.99
Log-probitb 1.77 1.40 0.11 209.99
Quantal linear 0.92 0.72 0.13 206.82
Weibulla 0.92 0.72 0.13 206.82
a
b
c
Restrict betas ≥0; lowest degree polynomial (up to n-2) with an adequate fit is reported; a 1-degree polynomial is
reported.

0.8 Log-Logistic
0.7
0.6
0.5
Fraction Affected
0.4
0.3
0.2
0.1
0
BMDL BMD
0 1 2 3 4 5 6 7 8 9
dose
11:41 04/09 2008
Figure B-5. Predicted and observed incidence of pancreas: acinus,

cytoplasmic alteration in female mice exposed to sodium dichromium

B.2. DETAILS OF BMD ANALYSIS FOR THE ORAL SLOPE FACTOR
The fit of the multistage model to the incidence of neoplasms in the small intestine of
male mice administered sodium dichromate dihydrate in drinking water for 2 years (NTP, 2008):
Source: NJDEP (2009)
====================================================================
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008)
Input Data File:
M:\ChromiumVI\msc_MALE_MICE_INTESTINAL_TUMORS_NTP_2008_Setting.(d)
Gnuplot Plotting File:
M:\ChromiumVI\msc_MALE_MICE_INTESTINAL_TUMORS_NTP_2008_Setting.plt
Fri Feb 05 09:42:31 2010
====================================================================
BMDS Model Run

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(

-beta1*dose^1-beta2*dose^2)]
The parameter betas are restricted to be positive

Dependent variable = Response
Independent variable = Dose
Total number of observations = 5

Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250

Relative Function Convergence has been set to: 1e-008
Parameter Convergence has been set to: 1e-008
Default Initial Parameter Values

Background = 0.0291151
Beta(1) = 0.0232273
Beta(2) = 0.0107072
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(1) Beta(2)
Background 1 -0.73 0.62
Beta(1) -0.73 1 -0.96
Beta(2) 0.62 -0.96 1
Parameter Estimates
95.0% Wald Confidence Interval

Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0.0287353 * * *
Beta(1) 0.024191 * * *
Beta(2) 0.0105146 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value

Full model -77.3728 5
Fitted model -77.8649 3 0.984149 2 0.6114
Reduced model -96.8272 1 38.9088 4 <.0001
AIC: 161.73

Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
------------------------------------------------------------------------
0.0000 0.0287 1.408 1.000 49 -0.349
0.3800 0.0391 1.915 3.000 49 0.800
0.9100 0.0581 2.848 2.000 49 -0.518
2.4000 0.1374 6.869 7.000 50 0.054
5.9000 0.4160 19.969 20.000 48 0.009
Chi^2 = 1.03 d.f. = 2 P-value = 0.5968
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 2.21769
BMDL = 1.16524
BMDU = 3.23024
Taken together, (1.16524, 3.23024) is a 90 % two-sided confidence

interval for the BMD
Multistage Cancer Slope Factor = 0.085819

The fit of the multistage model to the incidence of neoplasms in the small intestine of
female mice administered sodium dichromate dihydrate in drinking water for 2 years (NTP,
2008):
Source: NJDEP (2009)
====================================================================
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008)
Input Data File:
M:\ChromiumVI\msc_FEMALE_MICE_INTESTINAL_TUMORS_NTP_2008_Setting.(d)
Gnuplot Plotting File:
M:\ChromiumVI\msc_FEMALE_MICE_INTESTINAL_TUMORS_NTP_2008_Setting.plt
Fri Feb 05 09:54:51 2010
====================================================================
BMDS Model Run

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(

-beta1*dose^1-beta2*dose^2)]
The parameter betas are restricted to be positive
Dependent variable = Response

Independent variable = Dose

Total number of observations = 5
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250

Relative Function Convergence has been set to: 1e-008
Parameter Convergence has been set to: 1e-008
Default Initial Parameter Values

Background = 0.0398439
Beta(1) = 0.0695693
Beta(2) = 0
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(2)

have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background Beta(1)
Background 1 -0.62
Beta(1) -0.62 1
Parameter Estimates
95.0% Wald Confidence Interval

Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0.0140838 * * *
Beta(1) 0.0792034 * * *
Beta(2) 0 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value

Full model -88.9774 5
Fitted model -91.8504 2 5.74595 3 0.1246
Reduced model -117.047 1 56.1401 4 <.0001
AIC: 187.701

Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
------------------------------------------------------------------------
0.0000 0.0141 0.690 1.000 49 0.376
0.3800 0.0433 2.166 1.000 50 -0.810
1.4000 0.1176 5.761 4.000 49 -0.781
3.1000 0.2287 11.208 17.000 49 1.970
8.7000 0.5050 24.746 22.000 49 -0.785
Chi^2 = 5.90 d.f. = 3 P-value = 0.1164
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 1.33025
BMDL = 1.02757
BMDU = 1.93668
Taken together, (1.02757, 1.93668) is a 90 % two-sided confidence

interval for the BMD
Multistage Cancer Slope Factor = 0.0973173

Chromiumvi Erd Toxreview 9-30-10

Uploaded by

Copyright:

Available Formats

Chromiumvi Erd Toxreview 9-30-10

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chromiumvi Erd Toxreview 9-30-10

Uploaded by

Copyright:

Available Formats

DRAFT - DO NOT CITE OR QUOTE EPA/635/R-10/004A

In Support of Summary Information on the

U.S. Environmental Protection Agency

ii DRAFT - DO NOT CITE OR QUOTE

LIST OF TABLES ......................................................................................................................... vi

iii DRAFT - DO NOT CITE OR QUOTE

iv DRAFT - DO NOT CITE OR QUOTE

v DRAFT - DO NOT CITE OR QUOTE

Table 2-1. Industrial uses of hexavalent chromium compounds .................................................... 6

vi DRAFT - DO NOT CITE OR QUOTE

vii DRAFT - DO NOT CITE OR QUOTE

viii DRAFT - DO NOT CITE OR QUOTE

ix DRAFT - DO NOT CITE OR QUOTE

x DRAFT - DO NOT CITE OR QUOTE

3β-Δ5-HSH 3β-Δ5-hydroxysteroid dehydrogenase

xi DRAFT - DO NOT CITE OR QUOTE

xii DRAFT - DO NOT CITE OR QUOTE

xiii DRAFT - DO NOT CITE OR QUOTE

Ted Berner, M.S.

Catherine Gibbons, Ph.D

Glinda Cooper, Ph.D

Jenny Li, Ph.D

Teneille Walker, Ph.D

Julie Klotzbach, Ph.D.

xiv DRAFT - DO NOT CITE OR QUOTE

1 DRAFT - DO NOT CITE OR QUOTE

2 DRAFT - DO NOT CITE OR QUOTE

3 DRAFT - DO NOT CITE OR QUOTE

Figure 2-1. Definitions of bioaccessibility and biovailability used in this

4 DRAFT - DO NOT CITE OR QUOTE

2.1. ENVIRONMENTAL SOURCES AND OCCURRENCE

5 DRAFT - DO NOT CITE OR QUOTE

Name Formula Uses

Source: Hartford (YEAR) as cited in Nriagu (1988).

6 DRAFT - DO NOT CITE OR QUOTE

2.2. ENVIRONMENTAL CHEMISTRY, SPECIATION, AND BIOACCESSIBILITY

7 DRAFT - DO NOT CITE OR QUOTE

8 DRAFT - DO NOT CITE OR QUOTE

9 DRAFT - DO NOT CITE OR QUOTE

10 DRAFT - DO NOT CITE OR QUOTE

Half-cell reaction E0 (V) Change

Source: Emsley (1989) as cited in Katz and Salem (1993).

11 DRAFT - DO NOT CITE OR QUOTE

Figure 2-2. Schematic of possible reaction processes (grey ovals) that

12 DRAFT - DO NOT CITE OR QUOTE

Figure 2-3. Stability diagram showing aqueous speciation of chromium at

Hexavalent chromium is a strong oxidizer and as a result, exists only in oxygenated

13 DRAFT - DO NOT CITE OR QUOTE

14 DRAFT - DO NOT CITE OR QUOTE

2.3. ANALYTICAL METHODS

15 DRAFT - DO NOT CITE OR QUOTE

Approximate detection limit

Source: Losi et al. (1994).

Chromium in biological samples traditionally has been determined by graphite furnace

16 DRAFT - DO NOT CITE OR QUOTE

17 DRAFT - DO NOT CITE OR QUOTE

Table 2-5. Biomonitoring options for assessing chromium exposure

Biological matrix Advantages Disadvantages