DRAFT - DO NOT CITE OR QUOTE EPA/635/R-08/005D
www.epa.gov/iris
TOXICOLOGICAL REVIEW
OF
CARBON TETRACHLORIDE
(CAS No. 56-23-5)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
November 2009
NOTICE
This document is an Interagency Science Discussion 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
DRAFT - DO NOT CITE OR QUOTE
-------
DISCLAIMER
This document is a preliminary review 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ii DRAFT - DO NOT CITE OR QUOTE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 CARBON TETRACHLORIDE
(CAS No. 56-23-5)
LIST OF TABLES vi
LIST OF FIGURES viii
LIST OF ABBREVIATIONS AND ACRONYMS xii
FOREWORD xv
AUTHORS, CONTRIBUTORS, AND REVIEWERS xvi
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION 3
3. TOXICOKINETICS 6
3.1. ABSORPTION 6
3.1.1. Oral Exposure 6
3.1.2. Inhalation Exposure 7
3.1.3. Dermal Exposure 7
3.2. DISTRIBUTION 8
3.2.1. Oral Exposure 8
3.2.2. Inhalation Exposure 8
3.2.3. Dermal Exposure 10
3.2.4. Lactational Transfer 10
3.3. METABOLISM 11
3.4. ELIMINATION 14
3.5. PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS 17
4. HAZARD IDENTIFICATION 28
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL
CONTROLS 28
4.1.1. Oral Exposure 28
4.1.2. Inhalation Exposure 30
4.1.3. Dermal Exposure 38
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIO ASSAYS IN
ANIMALS—ORAL AND INHALATION 38
4.2.1. Oral Exposure 39
4.2.2. Inhalation Exposure 47
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND INHALATION. 64
4.3.1. Oral Exposure 64
4.3.2. Inhalation Exposure 67
4.4. OTHER DURATION-OR ENDPOINT-SPECIFIC STUDIES 68
4.4.1. Acute and Short-term Toxicity Data 68
4.4.2. Genotoxicity Studies 72
4.4.3. Initiation-promotion Studies 113
4.4.4. Neurotoxicity Studies 114
4.4.5. Immunotoxicity Studies 115
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION 118
iii DRAFT - DO NOT CITE OR QUOTE
-------
4.5.1. Metabolism is Required for Toxicity 119
4.5.2. Role of Free Radicals 120
4.5.3. Lipid Peroxidation 121
4.5.4. Depletion of Glutathione 126
4.5.5. Disruption of Calcium Homeostasis 127
4.5.6. Immunological and Inflammatory Effects 130
4.5.7. Changes in Gene Expression 133
4.5.8. Mechanisms of Kidney Toxicity 134
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS 135
4.6.1. Oral 135
4.6.2. Inhalation 140
4.6.3. Mode of Action Information 145
4.7. EVALUATION OF CARCINOGENICITY 146
4.7.1. Summary of Overall Weight-of-Evidence 146
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence 147
4.7.3. Mode of Action Information for Liver Tumors 150
4.7.4. Mode of Action Information for Pheochromocytomas 167
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES 168
4.8.1. Possible Childhood Susceptibility 168
4.8.2. Possible Effects of Aging 171
4.8.3. Possible Gender Differences 173
4.8.4. Nutritional Status 173
4.8.5. Disease Status 174
4.8.6. Exposure to Other Chemicals 175
5. DOSE-RESPONSE ASSESSMENTS 177
5.1. ORAL REFERENCE DOSE (RfD) 177
5.1.1. Choice of Principal Study and Critical Effect—with Rationale and
Justification 177
5.1.2. Methods of Analysis—Including Models 177
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs) 184
5.1.4. RfD Comparison Information 187
5.1.5. Previous RfD Assessment 191
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 191
5.2.1. Choice of Principal Study and Critical Effect—with Rationale and
Justification 191
5.2.2. Methods of Analysis—Including Models 194
5.2.3. RfC Derivation—Including Application of Uncertainty Factors 205
5.2.4. RfC Comparison Information 207
5.2.5. Previous RfC Assessment 213
5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE AND INHALATION
REFERENCE CONCENTRATION 213
5.4. CANCER ASSESSMENT 222
5.4.1. Choice of Study/Data—with Rationale and Justification 225
5.4.2. Dose-Response Data 226
5.4.3. Dose Adjustments and Extrapolation Methods 229
5.4.5. Nonlinear Extrapolation Approach 247
iv DRAFT - DO NOT CITE OR QUOTE
-------
5.4.6. Uncertainties in Cancer Risk Values 248
5.4.7. Previous Cancer Assessment 255
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE 256
6.1. HUMAN HAZARD POTENTIAL 256
6.2. DOSE RESPONSE 257
6.2.1. Noncancer- Oral Exposure 257
6.2.2. Noncancer- Inhalation Exposure 259
6.2.3. Cancer 261
7. REFERENCES 268
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS
AND DISPOSITION A-l
APPENDIX B. DOSE-RESPONSE MODELING FOR DERIVING THE RfD B-l
APPENDIX C. PBPK MODELING C-l
APPENDIX D. BENCHMARK DOSE MODELING FOR DERIVING THE RfC E-1
APPENDIX E. CANCER ASSESSMENT: BMD MODELING OUTPUTS FOR LOW-DOSE
LINEAR EXTRAPOLATION APPROACH F-l
APPENDIXF. SOURCE CODE FOR PBPK MODELS G-l
v DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 2-1. Physical properties and chemical identity of carbon tetrachloride 3
Table 3-1. AUC, Cmax, and Tmax in rat tissues following administration of 179 mg/kg carbon
tetrachloride by inhalation (1000 ppm for 2 hours), oral bolus dosing, or gastric
infusion over 2 hours 10
Table 3-2. Metabolic rate constants for hepatic microsomes in vitro 14
Table 3-3. Elimination ti/2 and apparent clearance of carbon tetrachloride from rat tissues
following administration of 179 mg/kg (1,000 ppm, 2 hours) by inhalation, oral
bolus dosing, or gastric infusion over 2 hours 16
Table 3-4. Physiological parameters for the rat, monkey, and human PBPK models for
carbon tetrachloride 19
Table 3-5. Comparison of metabolism from in vitro and in vivo studies 22
Table 4-1. Mean of selected serum chemistry and hematology variables in relation to carbon
tetrachloride exposure in British chemical workers 33
Table 4-2. Urinalysis results in rats after 2-year exposure to carbon tetrachloride 55
Table 4-3. Incidence of selected nonneoplastic lesions in F344 rats exposed to carbon
tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week) 57
Table 4-4. Incidence of liver tumors in F344 rats exposed to carbon tetrachloride vapor
for 104 weeks (6 hours/day, 5 days/week) 59
Table 4-5. Incidence of selected nonneoplastic lesions in BDF1 mice exposed to carbon
tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week) 62
Table 4-6. Incidence of liver and adrenal tumors in BDF1 mice exposed to carbon
tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week) 63
Table 4-7. Hepatic toxicity in rats exposed to carbon tetrachloride by inhalation or by
equivalent oral dosing as bolus or 2-hour gastric infusion 72
Table 4-8. Genotoxicity studies of carbon tetrachloride in prokaryotic organisms 73
Table 4-9. Genotoxicity studies of carbon tetrachloride in non-mammalian eukaryotic
organisms 76
Table 4-10. Genotoxicity studies of carbon tetrachloride in mammalian cells in vitro 78
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian cells in vivo 82
Table 4-12. Challenges in evaluating carbon tetrachloride genotoxicity 93
Table 4-13. Oral toxicity studies for carbon tetrachloride 136
Table 4-14. Inhalation toxicity studies for carbon tetrachloride 141
Table 4-15. Exposure levels for necrosis/degeneration and hyperplasia/regeneration in liver
following subchronic or chronic exposure to carbon tetrachloride by gavage or
inhalation 154
Table 4-16. Dose considerations of mechanistic studies of carbon tetrachloride 158
vi DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-17. Temporal sequence and dose-response relationship for key events and liver
tumors in male and female F344 rats exposed to carbon tetrachloride vapor for
13 and 104 weeks (6 hours/day, 5 days/week) 163
Table 5-1. Serum enzyme data in male rats after 10- or 12-week exposure to carbon
tetrachloride 179
Table 5-2. Severity of liver lesions in male rats after 12-week exposure to carbon
tetrachloride 182
Table 5-3. Incidence of selected liver lesions in mice treated with carbon tetrachloride
for 90 days 183
Table 5-4. Nonneoplastic lesions (fatty change) in F344 rats exposed to carbon tetrachloride
vapor for 104 weeks (6 hours/day, 5 days/week) 196
Table 5-5. Comparisons of internal dose metrics predicted from PBPK rat models 200
Table 5-6. HEC values corresponding to BMDL values for incidence data for fatty changes of
the liver in male F344 rats 203
Table 5-7. FIEC values corresponding to BMDL values for incidence data for fatty changes of
the liver in female F344 rats (high dose dropped) 204
Table 5-8. Incidence of liver tumors in F344 rats and BDF1 mice exposed to carbon
tetrachloride vapor for 104 weeks (6 hours/day, 5 day/week) 227
Table 5-9. Incidence of adrenal tumors (pheochromocytomas) in BDF1 mice exposed to carbon
tetrachloride vapor for 104 weeks (6 hours/day, 5 day/week) 227
Table 5-10. Internal dose metrics predicted from Fisher et al. (2004) and Thrall et al. (2000)
PBPK mouse models 231
Table 5-11. BMDL values for incidence data for liver tumors (adenoma plus carcinoma) in
female F344 rats and corresponding HEC andHED values 234
Table 5-12. BMDL values for incidence data for liver tumors (adenoma plus carcinoma) in
female F344 rats (high dose dropped) and corresponding HEC and HED values ...235
Table 5-13. BMDL values for incidence data for liver tumors (adenoma plus carcinoma) in
female BDF1 mice (high dose dropped) and corresponding HEC and HED
values 236
Table 5-14. BMDL values for incidence data for liver tumors (adenoma plus carcinoma) in
female BDF1 mice (2 highest doses dropped) and corresponding HEC and HED
values 237
Table 5-15. BMDL values for incidence data for liver tumors (adenoma plus carcinoma) in male
BDF1 mice (high dose dropped) and corresponding HEC and HED values 238
Table 5-16. BMDL values for incidence data for pheochromocytomas in female BDF1 mice and
corresponding HEC and HED values 239
Table 5-17. BMDL values for incidence data for pheochromocytomas in male BDF1 mice and
corresponding HEC and HED values 240
Table 5-18. Summary of IUR estimates using linear low-dose extrapolation approach 243
vii DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-19. Summary of oral SF estimates using linear low-dose extrapolation approach and
route-to-route extrapolation 246
Table 5-20. Summary of uncertainty in the carbon tetrachloride cancer risk assessment 253
Table B-l. Serum enzyme data in male rats after 10- or 12-week exposure to carbon
tetrachloride B-l
Table B-2. Model predictions for changes in serum SDH levels (lU/mL) in male rats
exposed to carbon tetrachoride for 10 and 12 weeks B-2
Table B-3. Model predictions for changes in serum OCT levels (nmol CCh/mL) in male rats
exposed to carbon tetrachloride for 10 and 12 weeks B-4
Table B-4. Model predictions for changes in serum ALT levels (lU/mL) in male rats
exposed to carbon tetrachloride for 10 and 12 weeks B-5
Table C-l. Comparison of predicted and observed values for selected parameters from
toxicokinetic data collected from rats and mice 48 hours post exposure to a 4-hour
nose-only inhalation exposure (20 ppm carbon tetrachloride) C-3
Table C-2. Parameter values for rat and human models C-5
Table C-3. Parameter values for mouse models C-6
Table C-4. Interspecies conversion factors based on MCA dose metric (VMAXC=0.04) C-9
Table C-5. Interspecies conversion factors based on MCA dose metric (VMAXC=0.65) C-10
Table C-6. Interspecies conversion factors based on MCA dose metric (VMAXC=1.49) C-l 1
Table C-l. Interspecies conversion factors based on MCA dose metric (VMAXC=1.70) C-12
Table C-8. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=0.04) C-13
Table C-9. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=0.65) C-15
Table C-10. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=1.49) C-17
Table C-l 1. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=1.70) C-19
Table C-12. Sensitive parameters (indicated with +) in the human model C-31
LIST OF FIGURES
Figure 2-1. Carbon tetrachloride 3
Figure 3-1. Metabolic scheme for carbon tetrachloride 12
viii DRAFT - DO NOT CITE OR QUOTE
-------
Figure 3-2. Two-compartment model for simulating GI absorption of carbon tetrachloride
administered to mice as a single gavage dose in Emulphor (Fisher et al., 2004) 26
Figure 4-1. Survival curves for male and female rats 54
Figure 4-2. Survival curves for male and female mice 60
Figure 4-3. Lipid peroxidation 122
Figure 4-4. Hypothesized carcinogenic MOA 151
Figure 5-1. PODs (mg/kg-day) with corresponding derived potential oral reference values that
would result if liver toxicity was used as the critical effect 189
Figure 5-2. PODs (mg/kg-day) with corresponding derived potential oral reference values that
would result if developmental toxi city was used as the critical effect 190
Figure 5-3. PODs (mg/kg-day) with corresponding derived potential oral reference values that
would result if alternative endpoints were used as the critical effect 191
Figure 5-4. Process for analyzing animal bioassay data for deriving noncancer toxicity values
and cancer lURs and SFs using PBPK modeling 197
Figure 5-5. Internal dose metrics predicted by the PBPK rat model 201
Figure 5-6. PODs (mg/m3) with corresponding derived potential inhalation reference values that
would result if liver toxicity was used as the critical effect 210
Figure 5-7. PODs (mg/m3) with corresponding derived potential inhalation reference values that
would result if kidney toxicity was used as the critical effect 211
Figure 5-8. PODs (mg/m3) with corresponding derived potential inhalation reference values that
would result if alternative endpoints were used as the critical effect 212
Figure 5-9. Comparison of suicide inhibition profiles for liver CYP450 in microsomes prepared
from rat and human liver at substrate (carbon tetrachloride) concentrations (CVL)
similar to those predicted by the PBPK models (0.2 jiM) for exposures
corresponding to the POD for the derivation of the RfC (14.3 mg/m3, 2.27 ppm), and
at 10-fold higher concentrations (20 jiM) 219
Figure 5-10. Comparison of suicide inhibition profiles for liver CYP450 in microsomes prepared
from rat and human liver at substrate (carbon tetrachloride) concentrations (CVL)
similar to those predicted by the PBPK models (0.2 jiM) for exposures
corresponding to the POD for the derivation of the RfC (14.3 mg/m3, 2.27 ppm), and
at 10-fold higher concentrations (20 jiM) 220
Figure 5-11. Internal dose metrics predicted from the Fisher et al. (2004) and Thrall et al. (2000)
PBPK mouse models 232
Figure C-l. Comparison of observed and predicted chamber carbon tetrachloride concentrations
in closed chamber studies conducted in rats C-l
Figure C-2. Comparison of observed and predicted chamber carbon tetrachloride concentrations
in closed chamber studies conducted in mice C-Error! Bookmark not defined.
ix DRAFT - DO NOT CITE OR QUOTE
-------
Figure C-3. Comparison of the actual versus predicted concentration of carbon tetrachloride in
the expired breath of humans exposed to 10 ppm of carbon tetrachloride for 180
minutes (data from Stewart et al., 1961) C-4
Figure C-4. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent exposure concentration (EC,
left panel) and values for % delta for trend lines (right panel). VMAXC=0.40
mg/hour/kg BW0-70 C-21
Figure C-5. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent exposure concentration (EC,
left panel) and values for % delta for trend lines (right panel). VMAXC=0.65
mg/hour/kg BW0'70 C-22
Figure C-6. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent exposure concentration (EC,
left panel) and values for % delta for trend lines (right panel). VMAXC=1.49
mg/hour/kg BW0-70 C-23
Figure C-7. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent exposure concentration (EC,
left panel) and values for % delta for trend lines (right panel). VMAXC=1.70
mg/hour/kg BW0-70 C-24
Figure C-8. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent rate of uptake from GI tract to
liver (RGIL, left panel) and values for % delta for trend lines (right panel).
VMAXC=0.40 mg/hour/kg BW0-70 C-25
Figure C-9. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent rate of uptake from GI tract to
liver (RGIL, left panel) and values for % delta for trend lines (right panel).
VMAXC=0.65 mg/hour/kg BW0-70 C-26
Figure C-10. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent rate of uptake from GI tract to
liver (RGIL, left panel) and values for % delta for trend lines (right panel).
VMAXC=1.49 mg/hour/kg BW0-70 C-27
Figure C-l 1. Relationship between internal dose metric MCA (time-averaged arterial blood
concentration of carbon tetrachloride) and equivalent rate of uptake from GI tract to
liver (RGIL, left panel) and values for % delta for trend lines (right panel).
VMAXC= 1.70 mg/hour/kg BW0-70 C-28
Figure C-12. Relationship between internal dose metric MRAMKL (mean rate of carbon
tetrachloride metabolism in the liver) and equivalent exposure concentration (EC)
and values for % delta for trend lines C-29
Figure C-l3. Relationship between internal dose metric MRAMKL (mean rate of carbon
tetrachloride metabolism in the liver) and equivalent rate of uptake from GI tract to
liver (RGIL) and values for % delta for trend lines C-300
DRAFT - DO NOT CITE OR QUOTE
-------
Figure C-14. Standardized sensitivity coefficients for the MCA dose metric (average
concentration of carbon tetrachloride in blood, umol/L) simulated with the human
carbon tetrachloride PBPK model C-32
Figure C-15. Standardized sensitivity coefficients for the MRAMKL dose metric (average rate
of metabolism of carbon tetrachloride umol/hr/kg liver) simulated with the human
carbon tetrachloride PBPK model C-33
Figure E-l. Histogram of the shape parameter E-36
xi DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
ACSL Advanced Continuous Simulation Language
AIC Akaike's Information Criterion
ALP alkaline phosphatase
ALT alanine aminotransferase
AST aspartate aminotransferase
ATSDR Agency for Toxic Substances and Disease Registry
AUC area under the curve
BMD benchmark dose
BMDL benchmark dose, 95% lower bound
BMDS benchmark dose software
BMR benchmark response
BMRF benchmark response factor
BrdU 5-bromo-2'-deoxyuridine
BUN blood urea nitrogen
BW body weight
CASRN Chemical Abstracts Service Registry Number
CBZ N-benzyloxycarbonyl-valine-phenylalanine methyl ester
CC14 carbon tetrachloride
CFC chlorofluorocarbon
CHO Chinese hamster ovary
CI confidence interval
Cmax maximum tissue concentration
CPN chronic progressive nephropathy
CPK creatine phosphokinase
CYP450 cytochrome P450
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid
dpm disintegrations per minute
ECD electron capture detector
FEL frank effect level
G6Pase glucose-6-phosphatase
GCL y-glutamylcysteine ligase
GD gestational day
GDH glutamate dehydrogenase
GGT y-glutamyl transferase
GI gastrointestinal
GSH glutathione (reduced)
GST-P glutathione S-transferase placental
HA hemagglutinin
HED human equivalent dose
HEC human equivalent concentration
4-HNE 4-hydroxynonenal
IFN- y interferon- y
xn
DRAFT - DO NOT CITE OR QUOTE
-------
IgM immunoglobulin
iNOS inducible nitric oxide synthase
i.p. intraperitoneal
IRIS Integrated Risk Information System
IUR inhalation unit risk
JBRC Japan Bioassay Research Center
Km Michaelis-Menten constant
LAP leucine aminopeptidase
LH luteinizing hormone
LN lead nitrate
LDH lactate dehydrogenase
LOAEL lowest-observed-adverse-effect level
MCA mean arterial concentration
MCMC Markov Chain Monte Carlo
MCL mean liver concentration
MDA malondialdehyde
MRAMKL mean rate of metabolism in the liver
MN micronucleus
MOA mode of action
mRNA messenger RNA
MW molecular weight
NADPH nicotinamide adenine dinucleotide phosphate
NAF nafenopin
NCI National Cancer Institute
NHL non-Hodgkin's lymphoma
NK natural killer
NLM National Library of Medicine
NOAEL no-observed-adverse-effect level
NRC National Research Council
NTP National Toxicology Program
OCT ornithine carbamoyl transferase
8-OHdG 8-hydroxy-2'-deoxyguanosine
8-oxodG 8-oxo-7,8-dihydro-2'-deoxyguanosine
OR odds ratio
PBPD physiologically based pharmacodynamic
PBPK physiologically based pharmacokinetic
PFC plaque-forming cell
PH partial hepatectomy
PNMT phenylethanolamine-N-methyltransferase
PND postnatal day
POD point of departure
RfC reference concentration
RfD reference dose
RGIL rate of uptake of carbon tetrachloride from the gastrointestinal tract to liver
SAH S-adenosylhomocysteine
SAM S-adenosylmethionine
SCE sister chromatid exchange
xin
DRAFT - DO NOT CITE OR QUOTE
-------
SD standard deviation
SDH sorbitol dehydrogenase
SE standard error
SEM standard error of the mean
SF slope factor
SMR standardized mortality ratio
SOS inducible DNA repair system
ti/i half-life
TEA total bile acids
TEARS thiobarbituric acid-reactive substances
TCPOBOP l,4-bis[2-(3,5-dichloropyridyloxy)]benzene
TGF tumor growth factor
Tmax time at which the maximum occurred
TNF- a tumor necrosis factor a
TUNEL terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick
end labeling
UF uncertainty factor
UDS unscheduled DNA synthesis
Vmax maximum velocity of enzyme reaction
xiv
DRAFT - DO NOT CITE OR QUOTE
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to carbon
tetrachloride. It is not intended to be a comprehensive treatise on the chemical or toxicological
nature of carbon tetrachloride.
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 the 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
hotline.iris@epa.gov (email address).
xv DRAFT - DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER/AUTHOR
Susan Rieth
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Reeder Sams
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
AUTHORS
Mary Manibusan
Health Effects Division
Office of Pesticide Programs
U.S. Environmental Protection Agency
Washington, DC
Jennifer Jinot
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Leonid Kopylev
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Paul White
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
xvi DRAFT - DO NOT CITE OR QUOTE
-------
Paul Schlosser
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
CONTRACTOR SUPPORT
Marc Odin
Environmental Science Center
Syracuse Research Corporation
Syracuse, NY
Gary Diamond
Environmental Science Center
Syracuse Research Corporation
Syracuse, NY
Margaret Fransen
Environmental Science Center
Syracuse Research Corporation
Syracuse, NY
Julie Klotzbach
Environmental Science Center
Syracuse Research Corporation
Syracuse, NY
David Eastmond
Environmental Toxicology Graduate Program
University of California, Riverside
Riverside, CA
REVIEWERS
This document has been reviewed by EPA scientists, interagency reviewers from other
federal agencies and White House offices, and the public, and peer reviewed by independent
scientists external to EPA. A summary and EPA's disposition of the comments received from
independent external peer reviewers and from the public is included in Appendix A.
INTERNAL EPA REVIEWERS
Ted Berner, Office of Research and Development/NCEA
Glinda Cooper, Office of Research and Development/NCEA
Anthony DeAngelo, Office of Research and Development/NHEERL
xvii DRAFT - DO NOT CITE OR QUOTE
-------
Karen Hammerstrom, Office of Research and Development/NCEA
Cheryl Scott, Office of Research and Development/NCEA
Joyce Donohue, Office of Water/OST
Genetic Toxicology
Channa Keshava, Office of Research and Development/NCEA
Larry Valcovic, Office of Research and Development/NCEA
YinTak Woo, Office of Prevention, Pesticides and Toxic Substances
Immunotoxicology
Andrew Rooney, Office of Research and Development/NCEA
PBPK
Rob Dewoskin, Office of Research and Development/NCEA
Marina Evans, Office of Research and Development/NHEERL
Mode of Action
Vicki Dellarco, Office of Pollution Prevention
Danielle Devoney, Office of Research and Development/NCEA
Julie Du, Office of Water/OST
Kate Guyton, Office of Research and Development/NCEA
Rita Schoeny, Office of Water/OST
John Whalan, Office of Research and Development/NCEA
YinTak Woo, Office of Prevention, Pesticides and Toxic Substances
Jean Zodrow, U.S. EPA Region 10
EXTERNAL PEER REVIEWERS
Janusz Z. Byczkowski, DABT, D.Sc., Ph.D.
Consultant
Gary Ginsberg, Ph.D.
Connecticut Department of Public Health
Dale Hattis, Ph.D.
Clark University and George Perkins Marsh Institute
Lisa Kamendulis, Ph.D.
Department of Pharmacology and Toxicology, Indiana University
Lawrence Lash, Ph.D.
Department of Pharmacology, Wayne State University School of Medicine
xviii DRAFT - DO NOT CITE OR QUOTE
-------
Madhusudan Soni, Ph.D., FACN
Soni and Associates
xix 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 and dose-response assessment of carbon
tetrachloride. IRIS Summaries may include oral reference dose (RfD) and inhalation reference
concentration (RfC) values for chronic and other exposure durations, and a carcinogenicity
assessment.
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 specified otherwise, the RfD and RfC are derived for
chronic exposure duration.
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 ug/m3 air breathed.
Development of these hazard identification and dose-response assessments for carbon
tetrachloride has followed the general guidelines for risk assessment as set forth by the National
Research Council (NRC, 1983). 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 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
1 DRAFT - DO NOT CITE OR QUOTE
-------
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, 1996a), Guidelines for
Neurotoxicity Risk Assessment (U.S. EPA, 1998a), 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 ,4
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 February
2009.
DRAFT - DO NOT CITE OR QUOTE
-------
2. CHEMICAL AND PHYSICAL INFORMATION
Carbon tetrachloride is a colorless liquid with a sweetish odor (NLM, 2003; Lewis,
1997). Synonyms include tetrachloromethane and perchloromethane (NLM, 2003; O'Neil and
Smith, 2001). The chemical structure of carbon tetrachloride is shown in Figure 2-1. Selected
chemical and physical properties of carbon tetrachloride are listed below in Table 2-1.
Cl
Cl —C —Cl
Cl
Figure 2-1. Carbon tetrachloride.
Table 2-1. Physical properties and chemical identity of carbon tetrachloride
CASRN 56-23-5
Molecular weight
Chemical formula
Boiling point
Melting point
Vapor pressure at 25°C
Density at 20°C
•
Water solubility at 25°C
Other solubility
Partition coefficient
Flash point
Autoignition temperature
Latent heat of vaporization
Heat of fusion
Critical temperature
Critical pressure
Viscosity at 24°C
Surface tension at 20°C
Henry's law constant at 25°C
OH reaction rate constant at 25°C
Koc
153.82
CC14
76.8°C
-23°C
1.15 x 102mmHg
1.5940g/mL
5.32
5.41
7.93 x 102mg/L
Miscible with alcohol, benzene,
chloroform, ether, carbon
disulfide, petroleum ether, oils
log Kow = 2.83
Not flammable
>1,000°C
1.959x 10'J/kg
5.09 cal/g
556.35°C
4.56 x !06Pa
0.922 cp
0.027 N/m
2.76 x 10~2atmnrVmol
1.20 x 10'16 cnrVmolecule sec
71
O'Neil and Smith, 2001
O'Neil and Smith, 2001
NLM, 2003; Lide, 2000
NLM, 2003; Lide, 2000
NLM, 2003
NLM, 2003; Lide, 2000
NLM, 2003; U.S. Coast Guard,
1999
O'Neil and Smith, 2001
NLM, 2003; Horvath, 1982
NLM, 2003; O'Neil and Smith,
2001
NLM, 2003; Hansch et al, 1995
NLM, 2003; U.S. Coast Guard,
1999
Holbrook, 1993
U.S. Coast Guard, 1999
NLM, 2003; U.S. Coast Guard,
1999
Daubert and Banner, 1 995
Daubert and Banner, 1 995
U.S. Coast Guard, 1999
U.S. Coast Guard, 1999
NLM, 2003; Leighton and Calo,
1981
NLM, 2003; Atkinson, 1989
NLM, 2003
DRAFT - DO NOT CITE OR QUOTE
-------
Table 2-1. Physical properties and chemical identity of carbon tetrachloride
Bioconcentration factor
Conversion factors at 25°C
3.2-7.4
1 mg/m3 = 0. 16 ppm; 1 ppm =
6.29 mg/m3
NLM, 2003; CITI, 1992
NLM, 2003
In the United States, carbon tetrachloride is most commonly prepared by chlorinating
methane or by a chlorinating cleavage reaction with less than or equal to Cs hydrocarbons or
chlorinated hydrocarbons (Rossberg, 2002). Prior to the late 1950s, carbon tetrachloride was
produced primarily by carbon disulfide chlorination (NLM, 2003; Rossberg, 2002).
Carbon tetrachloride has been used as a dry-cleaning agent, fabric-spotting fluid, solvent,
reagent in chemical synthesis, fire extinguisher fluid, and grain fumigant (NLM, 2003; Holbrook,
1993), but its primary use was in chlorofluorocarbon (CFC) production (NLM, 2003; Rossberg,
2002). Since the mid-1970s, annual use and production has generally declined. The Consumer
Product Safety Commission banned the use of carbon tetrachloride in consumer products in the
1970s. Decline in the use of carbon tetrachloride also accompanied EPA's increased regulation
of the use of CFCs in propellants (a ban on CFCs in aerosol products went into effect in 1978),
and the adoption of the Montreal Protocol, an international agreement to reduce environmental
concentrations of ozone-depleting chemicals, which was implemented in the United States via
Title VI of the Clean Air Act Amendments of 1990 (ATSDR, 2005; Doherty, 2000; Holbrook,
1993). The ban on production and import of carbon tetrachloride in developed countries,
including the United States, took effect on January 1, 1996. Excluded from the production and
import ban is the manufacture of a controlled substance that is subsequently transformed or
destroyed and small amounts exempted for essential laboratory and analytical uses (40 CFR Part
82; 72 Fed Reg 52332, Sept 13, 2007a).
Production figures for carbon tetrachloride since the 1970s reflect the regulatory history
of the chemical. Carbon tetrachloride production peaked in the early 1970s, with annual U.S.
production exceeding one billion pounds. Production in the early 1990s had declined to
approximately 300 million pounds (Doherty, 2000). According to Agency for Toxic Substances
and Disease Registry (ATSDR), manufacture of carbon tetrachloride in the U.S. in the early
2000s was limited to one company (Vulcan Materials Company) at two plants with a combined
130 million pound capacity (ATSDR, 2005); however, these capacities were considered flexible
because other chlorinated solvents were made using the same equipment.
Historically, carbon tetrachloride was released into the environment predominantly
through direct emissions to air, with lower amounts discharged to soil and water (ATSDR,
2005). Carbon tetrachloride released to soil or water is expected to volatilize to air based on its
vapor pressure and Henry's law constant (NLM, 2003). In air, carbon tetrachloride will exist as
a vapor, as indicated by its vapor pressure (NLM, 2003). The behavior of carbon tetrachloride in
the atmosphere is the most important aspect of this chemical's environmental fate. Carbon
4 DRAFT - DO NOT CITE OR QUOTE
-------
tetrachloride does not undergo photodegradation (Holbrook, 1993) or absorb light at wavelengths
found in the troposphere and hence does not undergo direct photolysis in that region of the
atmosphere (NLM, 2003). Carbon tetrachloride that remains in the troposphere eventually rises
into the stratosphere, where it is photolyzed by the shorter wavelength light (Molina and
Rowland, 1974). When carbon tetrachloride photolyzes in the stratosphere, the chlorine radicals
responsible for the destruction of atmospheric ozone are released.
In soil, carbon tetrachloride is expected to be highly mobile based on its Koc and is
expected to leach to lower soil horizons and groundwater (NLM, 2003). BCF values indicate
that carbon tetrachloride will not bioconcentrate appreciably in aquatic or marine organisms
(NLM, 2003). Carbon tetrachloride may biodegrade in soil or water under anaerobic conditions;
however, biodegradation of carbon tetrachloride under aerobic conditions does not occur readily
(NLM, 2003; U.S. EPA, 1996b; Semprini, 1995).
DRAFT - DO NOT CITE OR QUOTE
-------
3. TOXICOKINETICS
Carbon tetrachloride is rapidly absorbed by any route of exposure in humans and animals.
Once absorbed, it is widely distributed among tissues, especially those with high lipid content,
reaching peak concentrations in
-------
the blood stream (Kim et al., 1990a).
3.1.2. Inhalation Exposure
Data from humans and animals suggest that carbon tetrachloride is rapidly absorbed
through the lungs, which is inferred from the rapid onset of symptoms of toxicity or detection of
carbon tetrachloride in blood or in exhaled air. In volunteers exposed to 10 ppm for 180
minutes, carbon tetrachloride was detectable in exhaled air within 15 minutes (Stewart et al.,
1961). Human subjects exposed to >60 mg/L (>9600 ppm) reported symptoms of toxicity within
the first minute of exposure; symptoms appeared after 3 minutes in subjects exposed to 30 mg/L
(4800 ppm) (Lehmann and Schmidt-Kehl, 1936). After male Sprague-Dawley rats were exposed
at 100 or 1000 ppm, carbon tetrachloride was detected in arterial blood in the initial 5-minute
samples (Sanzgiri et al., 1995; Bruckner et al., 1990); blood levels rose during the 2-hour
exposure period to a near steady-state level. In dogs exposed to 5000 ppm of carbon
tetrachloride, blood levels reached a near steady-state level within 2 hours (von Oettingen et al.,
1950).
Lehmann and Schmidt-Kehl (1936) estimated that approximately 63% of inhaled carbon
tetrachloride vapor was absorbed by the lungs in human subjects exposed to "a few mg per liter."
In monkeys exposed to carbon tetrachloride at 46 ppm for periods between 2 and 5 hours, an
average of 30% of the total amount inhaled was absorbed, and the rate of absorption averaged
0.022 mg/kg-minute (McCollister et al., 1951). Rats that were exposed at 4,000 ppm for 6 hours
had initial body burdens of approximately 14 mg of carbon tetrachloride and 257 ug of its
metabolite chloroform (Dambrauskas and Cornish, 1970). Initial body burdens in rats, mice, and
hamsters that were exposed to 20 ppm of carbon tetrachloride vapor for 4 hours were 7.7, 10.6,
and 4.0 mg/kg, respectively (Benson and Springer, 1999). In vitro experiments of carbon
tetrachloride indicated blood:air partition coefficients of 2.73-4.20 for human blood (Fisher et
al., 1997; Gargas et al., 1989) and 4.52 for rat blood (Gargas et al., 1986).
3.1.3. Dermal Exposure
Carbon tetrachloride is absorbed rapidly through the skin. The chemical was detected in
alveolar air within 10 minutes in human subjects who immersed their thumbs in neat liquid
(Stewart and Dodd, 1964). Animal studies have found similar results. Carbon tetrachloride was
detected in blood within 5 minutes of dermal application of neat liquid in guinea pigs (Jakobson
et al., 1982). The percutaneous absorption rate for carbon tetrachloride applied neat to the
9
abdominal skin of male ICR mice was estimated as 53.6 ± 9.3 nmoles/minute/cm (Tsuruta,
1975). Morgan et al. (1991) compared dermal absorption of carbon tetrachloride in rats when
applied neat or in aqueous solution. With neat application, maximum blood levels were reached
within 30 minutes, and approximately one quarter of the applied volume (0.54 mL) was absorbed
in a 24-hour period. With application in saturated aqueous solution, absorption was slower (peak
7 DRAFT - DO NOT CITE OR QUOTE
-------
blood levels were not attained until 10 hours after exposure), and a somewhat lower amount
(0.39 mL) was absorbed in 24 hours.
Dermal absorption of radiolabeled carbon tetrachloride vapor was low in monkeys
exposed to 485 or 1,150 ppm for about 4 hours (McCollister et al., 1951). Blood concentrations
at the end of exposure were approximately equivalent to 0.012-0.03 mg carbon tetrachloride/100
g blood but were undetectable after 48 hours; concentrations in exhaled air were equivalent to
0.0008-0.003 mg carbon tetrachloride/L but were undetectable 120 hours later. The authors
concluded that, for whole-body exposures to carbon tetrachloride vapor, the dermally absorbed
fraction would be negligible.
3.2. DISTRIBUTION
3.2.1. Oral Exposure
No data are available for the distribution of carbon tetrachloride in humans. Animal
studies indicate that the largest fraction of an absorbed oral dose of carbon tetrachloride is
initially distributed to fat. After administration of about 3,200 mg/kg to rats, peak levels of
radiolabeled carbon tetrachloride were observed after about 2 hours in blood, muscle, liver, and
brain and after 5.5 hours in fat (Marchand et al., 1970). Peak tissue levels of carbon tetrachloride
were similar in blood and muscle but were twice as high in the brain, 5 times higher in liver, and
50 times higher in fat. Similar results were obtained in rabbits treated with a low dose of carbon
tetrachloride (Fowler, 1969). Six hours after an oral dose of 1.6 mg/kg, recoveries of parent
compound totaled 787 ug/g in fat, 96 ug/g in liver, 20 ug/g in kidney, and 21 ug/g in muscle;
distributions of the carbon tetrachloride metabolites chloroform and hexachloroethane were
highest in fat and liver but were below 5 ug/g. Forty-eight hours after dosing, tissue
concentrations of the parent compound were 45 ug/g in fat, 3.8 ug/g in liver, and <1 ug/g in the
other tissues; chloroform was present at <1 ug/g in the four tissues, whereas hexachloroethane
was present at 6.8 ug/g in fat, 1 ug/g in liver, and <1 ug/g in other tissues.
3.2.2. Inhalation Exposure
A similar pattern of distribution has been found in animals exposed to carbon
tetrachloride by inhalation. Rats exposed to 4,000 ppm for 6 hours showed the largest
concentrations of carbon tetrachloride in the fat (1,674 ug/g), followed by the brain (407 ug/g),
kidney (233 ug/g), liver (136 ug/g), and blood (64 ug/g) (Dambrauskas and Cornish, 1970). The
liver also contained 10 ug/g of chloroform (as a carbon tetrachloride metabolite). Monkeys
exposed to 46 ppm of radiolabeled carbon tetrachloride vapor for 5 hours had the highest
concentration of label in fat, with decreasing amounts in the liver, bone marrow, blood, brain,
kidney, heart, spleen, muscle, lung, and bone (McCollister et al., 1951). The concentrations in
fat and liver were eight- and threefold higher, respectively, than concentrations in blood.
Bergman (1983) followed the distribution of radiolabeled carbon tetrachloride by whole-
8 DRAFT - DO NOT CITE OR QUOTE
-------
body autoradiography in mice exposed by inhalation for 10 minutes and sacrificed at time points
up to 24 hours; sections were either processed at low temperatures to retain volatile radioactivity
(primarily parent compound), evaporated to retain only nonvolatile radioactivity (metabolites), or
evaporated and then extracted to retain only protein- and nucleic acid-bound radioactivity
(metabolites covalently bound to protein and nucleic acids). Immediately after inhalation
exposure, high levels of volatile radioactivity were detectable in fat, bone marrow, and nervous
tissues (spinal cord and white matter of the brain). Nonvolatile and partly nonextractable
radioactivity was detected in the liver, kidney cortex, lung, bronchi, GI mucosa (especially in the
glandular stomach, colon, and rectum), nasal mucosa, salivary glands, vaginal and uterine
mucosa, and, interstitially, in the testis; nonvolatile radioactivity was also detected in urine and
bile. The distribution pattern of volatile carbon tetrachloride and its nonvolatile metabolites was
similar 30 minutes after exposure. Volatile radioactivity was detectable at relatively high levels
in the nervous system at 4 hours and in fat at 8 hours but not at 24 hours. The pattern of labeling
in the liver demonstrated a centrilobular concentration. Bergman (1983) reported a good
correlation between nonextractable radioactivity and published tissue concentrations of
cytochrome (CYP) P450.
Sanzgiri et al. (1997) compared the tissue distribution of carbon tetrachloride
administered by inhalation (1,000 ppm for 2 hours) and the equivalent oral dose (179 mg/kg)
given as a single bolus dose or gastric infusion over 2 hours. Table 3-1 shows area under the
curve (AUC) for the 24-hour monitoring period, the maximum tissue concentrations (Cmax), and
the times (Tmax) at which the maxima occurred. Maximal tissue concentrations were reached
quickest by gavage dosing, followed by inhalation and then gastric infusion. By all routes,
attainment of maximal levels was slower in fat than in other tissues. Maximal levels in fat were
considerably in excess of the maximal levels in other tissues, regardless of route of exposure.
Among tissues other than fat, distribution kinetics of carbon tetrachloride were generally similar
for the different tissues, except that maximal levels were higher and attained more quickly in the
liver than in other tissues following bolus oral administration.
DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-1. AUC, Cmax, and Tmax in rat tissues following administration of
179 mg/kg carbon tetrachloride by inhalation (1000 ppm for 2 hours), oral
bolus dosing, or gastric infusion over 2 hours
Tissue
Liver
Kidney
Lung
Brain
Fat
Heart
Muscle
Spleen
Inhalation
AUC
(HgXminute/
mL)
2,823
3,064
2,952
3,255
230,699
2,571
3,248
2,035
^max
G»g/g)
20
25
24
28
1506
18
18
13
T
•*• max
(min)
30
30
30
30
240
30
30
30
Oral bolus
AUC
(HgXminute/
mL)
1,023
3,029
2,908
4,223
235,471
2,747
4,117
4,096
^max
G»g/g)
58
14
10
15
246
10
7
12
T
•*• max
(min)
1
5
15
15
120
5
60
5
Gastric infusion
AUC
(HgXminute/
mL)
149
800
2,842
2,683
165,983
1,900
2,164
1,660
^max
G»g/g)
0.5
4
6
10
179
8
10
6
T
•*• max
(min)
120
120
180
150
360
120
150
150
Source: Sanzgiri et al, 1997.
Benson et al. (2001) compared the initial and delayed tissue distribution of inhaled
carbon tetrachloride in rats, mice, and hamsters exposed to 20 ppm of radiolabeled carbon
tetrachloride for 4 hours. Immediately after exposure, the percentage of the initial body burden
present in major tissues was 30% in rats and hamsters and 40% in mice; the highest proportion at
that time was in the liver of mice and hamsters and in the fat in rats. Two days later, the liver
contained the highest amount in all three species. The results in rats reflect the initial lipophilic
distribution of carbon tetrachloride and the subsequent accumulation in the liver.
3.2.3. Dermal Exposure
Few data are available regarding tissue concentrations of carbon tetrachloride following
dermal exposure. One study of guinea pigs given topical application of carbon tetrachloride
found that blood concentrations of the chemical increased during the first half hour of exposure
but then declined to about 25% of peak levels despite continued exposure over a 6-hour period
(Jakobson et al., 1982).
3.2.4. Lactational Transfer
Fisher et al. (1997) experimentally derived a human milk:blood partition coefficient of
3.26 for carbon tetrachloride, which would suggest a potential sensitive subpopulation of nursing
infants based on the possibility of lactational transfer.
10
DRAFT - DO NOT CITE OR QUOTE
-------
3.3. METABOLISM
Carbon tetrachloride is metabolized in the body, primarily by the liver, but also in the
kidney, lung, and other tissues containing CYP450. The percent of a given dose that is
metabolized varies with dose, as discussed in Section 3.4.
The metabolism of carbon tetrachloride has been extensively studied in in vivo and in
vitro mammalian systems. Based on available data, a proposed metabolic scheme for carbon
tetrachloride is illustrated in Figure 3-1. There is considerable evidence that the initial step in
biotransformation of carbon tetrachloride is reductive dehalogenation: reductive cleavage of one
carbon-chlorine bond to yield chloride ion and the trichloromethyl radical (Reinke and Janzen,
1991; Tomasi et al., 1987; McCay et al., 1984; Mico and Pohl, 1983; Slater, 1982; Poyer et al.,
1980, 1978; Lai et al., 1979).
The initial reaction step is catalyzed by an NADPH-dependent CYP450 that is inducible
by phenobarbital or ethanol (Castillo et al., 1992; Noguchi et al., 1982a; Sipes et al., 1977). In
humans and animals, CYP2E1 is the primary enzyme involved with carbon tetrachloride
bioactivation, while CYP3A may be involved under high exposure conditions (Zangar et al.,
2000; Raucy et al., 1993). As demonstrated in studies with CYP2E1 genetic knockout mice, this
enzyme is required for the development of hepatotoxicity (as measured by elevated liver
enzymes and liver histopathology) in mice exposed to carbon tetrachloride (Wong et al., 1998).
The fate of the trichloromethyl radical is dependent on the availability of oxygen and
includes several alternative pathways for anaerobic or aerobic conditions. Anaerobically, the
trichloromethyl radical may dimerize to form hexachloroethane, which has been detected in
animal tissues (Uehleke et al., 1973; Fowler, 1969). Addition of a proton and an electron to the
radical results in the formation of chloroform (CHCb), which has been detected in exposed rats
and rabbits (Reynolds et al., 1984; Ahr et al., 1980; Glende et al., 1976; Uehleke et al., 1973;
Dambrauskas and Cornish, 1970; Fowler, 1969). The trichloromethyl radical can undergo
further reductive dehalogenation catalyzed by CYP450 to form dichlorocarbene (:CC\2\ which
can bind irreversibly to tissue components or react with water to form formyl chloride (HCOC1),
which decomposes to carbon monoxide (Galelli and Castro, 1998; Pohl et al., 1984; Ahr et al.,
1980; Wolf et al., 1977). The trichloromethyl radical can bind directly to microsomal lipids and
proteins (Fanelli and Castro, 1995; Ansari et al., 1982; Villarruel et al., 1977), as well as the
heme portion of CYP450.
11 DRAFT - DO NOT CITE OR QUOTE
-------
Carbon Tetrachloride
R-CCI
adduct
•O-O-CCI
Trichloromethyl
peroxy radical
Lipid Peroxidation
R-CO-*
Cysteine
2HCI
(reduced
glutathione ,,
o
2HCI
Oxothiazolidine
carboxylic acid
GSCSG
Diglutathionyl
dithiocarbonate
CI3CCCI3
Hexachloroethane
Carbon
monoxide
2HCI
CO,
'2
Carbon
dioxide
Figure 3-1. Metabolic scheme for carbon tetrachloride.
CYP450, usually CYP2E1, but also CYP3A; R = acceptor molecule, such as protein or lipid.
Source: ACGIH, 2001.
12
DRAFT - DO NOT CITE OR QUOTE
-------
Aerobically, the trichloromethyl radical can be trapped by oxygen to form the
trichloromethyl peroxy radical, which can bind to tissue proteins (Galelli and Castro, 1998;
Packer et al., 1978) or decompose to form phosgene (COCb) (Pohl et al., 1984) and an
electrophilic form of chlorine (Pohl et al., 1984). The rate of conversion of the trichloromethyl
radical to the trichloromethyl peroxy radical (and to downstream reaction products with amino
O Q
acids and lipids) has been estimated to be approximately 10-10 L/mols (Russell et al., 1990;
Slater, 1981; Packer et al., 1978). These rates are sufficiently high to suggest that the rate of
production of the trichloromethyl peroxy radical (and, thereby, the rate of elimination of the
trichloromethyl radical) may be diffusion limited (1010-1012 L/mols; Atkins, 1998). Therefore,
limiting factors in the oxidative elimination of the trichloromethyl radical are likely to be
reactant concentrations at the site of production of the trichloromethyl radical (e.g., 62) and/or
factors that limit diffusion of the trichloromethyl radical (e.g., diffusion coefficient in cytosol).
The trichloromethyl peroxy radical is the primary initiator of lipid peroxidation that occurs from
exposure to carbon tetrachloride (Boll et al., 2001a; McCay et al., 1984; Rao and Recknagel,
1969). Carbon dioxide is generated by the hydrolytic cleavage of phosgene (Shah et al., 1979).
Phosgene may also be conjugated to reduced glutathione (GSH) to form diglutathionyl
dithiocarbonate or to cysteine to form oxothiazolidine carboxylic acid (U.S. EPA, 200la).
Continued exposure to carbon tetrachloride has been shown to temporarily reduce its
initial toxicity in rat studies (Glende, 1972). This phenomenon is related to the loss of CYP450
content (suicide inactivation), which has also been observed in treated rats (de Toranzo et al.,
1978), resulting from the formation of reactive intermediates, such as the trichloromethyl radical
(Fernandez et al., 1982; Noguchi et al., 1982b; de Groot and Haas, 1981; Glende, 1972). Under
anaerobic conditions, heme tetrapyrrolic structures of the human or rat CYP450 enzymes are
destroyed in a process that follows pseudo first-order kinetics (Manno et al., 1992, 1988).
Although the fast and slow half-lives (ti/2) for the two species are similar (3.2 and 28.9 minutes
for the rat and 4.0 and 29.8 minutes for the human), inactivation is more severe in the rat, with 1
molecule of rat CYP450 enzyme lost for every 26 molecules of substrate metabolized, compared
with a loss of 1 molecule of human enzyme for every 196 molecules of substrate processed
(Manno et al., 1992, 1988). A higher rate of inactivation of CYP450 in the rat compared to
humans has potential implications for extrapolating external and internal doses (e.g., rates of
metabolism of carbon tetrachloride) across species (see Section 5.2.2.1).
As demonstrated qualitatively by the distribution of nonvolatile radioactivity
(metabolites) in the autoradiography study by Bergman (1983) and quantitatively in other in vivo
assays (see Section 3.2), carbon tetrachloride is metabolized in many tissues throughout the body
but most significantly in the liver. The amount of carbon tetrachloride metabolized in a given
tissue is related to the CYP450 content of the tissue (Bergman, 1983; Villarruel et al., 1977). In
the liver, the greatest accumulation of carbon tetrachloride metabolites occurs in the centrilobular
region, which has high CYP450 levels (Bergman, 1983).
13 DRAFT - DO NOT CITE OR QUOTE
-------
Zangar et al. (2000) measured carbon tetrachloride metabolic rate constants for human
and animal hepatic microsomal preparations in vitro (Table 3-2). Results suggest that the
metabolic rate in humans is more similar to the rate in rats than in other rodent species.
Table 3-2. Metabolic rate constants for hepatic microsomes in vitro
Species
Human
Rat
Mouse
Hamster
Ka
m
(jiM)
56.8
59.1
29.3
30.2
V b
v max
(nmol/minute/mg protein)
2.26
3.1
2.86
4.1
aKm= Michaelis-Menten constant.
b Vmax = Maximum velocity of enzyme reaction.
Source: Zangar et al. (2000).
Metabolism of carbon tetrachloride can be induced by chemicals that increase the
expression of CYP2E1 or CYP3A (see Section 4.8.6. for further discussion).
3.4. ELIMINATION
In humans and animals exposed to carbon tetrachloride by any route, the unmetabolized
parent compound is excreted in exhaled air. Additionally, animal studies show that volatile
metabolites are released in exhaled air, whereas nonvolatile metabolites are excreted in feces and
to a lesser degree, in urine.
Six hours after an attempted suicide by ingestion of an unknown amount of carbon
tetrachloride in a mixture with methanol, the concentration of carbon tetrachloride in expired air
was -2,500 ug/L and declined to -120 ug/L after 1 day and to -1 ug/L after 20 days (Stewart et
al., 1963). In a worker acutely exposed to mixed solvent vapors, the concentration of carbon
tetrachloride in alveolar air declined from an initial value of-4,000 ppm to -0.003 ppm after 15
days (Stewart et al., 1965). Human subjects (n=6) who inhaled carbon tetrachloride vapor at 10
ppm for 3 hours had concentrations in expired air of 1 ppm 15 minutes postexposure and about
0.28 ppm 5 hours postexposure (Stewart et al., 1961). Approximately 33% of the absorbed dose
was excreted in exhaled air within 1 hour in human subjects who inhaled radiochlorine-labeled
carbon tetrachloride in a single breath (Morgan et al., 1970). Following dermal exposure to neat
carbon tetrachloride, excretion into alveolar air was detectable within 10 minutes in three human
subjects (Stewart and Dodd, 1964). Concentrations in alveolar air ranged from 0.11 to 0.83 ppm
by the end of a 30-minute exposure, peaking 30 minutes postexposure and beginning to decline 1
hour postexposure; after 5 hours, the concentrations were 0.12-0.14 ppm. Using a physiological
14 DRAFT - DO NOT CITE OR QUOTE
-------
four compartment model, Sato and Nakajima (1987) calculated that 93% of inhaled carbon
tetrachloride vapor was removed unchanged via the lungs (assuming an alveolar ventilation rate
of 336 L/hour), while 7% was cleared metabolically in humans.
Animal studies evaluated elimination of carbon tetrachloride following oral or inhalation
exposures. In rats receiving equivalent doses by inhalation or bolus gavage, terminal elimination
ti/2 values were about 4 hours (Bruckner et al., 1990).
Reynolds et al. (1984) evaluated elimination parameters during a 24-hour period in rats
exposed by gavage to [14C]-carbon tetrachloride at doses ranging from 15 to 4000 mg/kg. At the
low dose of 15 mg/kg, 19% of the administered dose was eliminated in exhaled air as the parent
compound, 28% as CC>2 (accounting for 83% of metabolites), and 0.11% as chloroform (0.3% of
metabolites); 2.9% of metabolites remained bound in the liver, while 2.7% were excreted in
urine and 11% in feces. At doses >600 mg/kg, >76% of the administered dose was exhaled as
parent compound, <2% was exhaled as CO2 (accounting for 50-60% of metabolites), and
<0.40% as chloroform (11-19% of metabolites); 2-4% of metabolites remained bound in the
liver, while 3-9% of metabolites were excreted in urine and 7-30% in feces. At 15 mg/kg, peak
exhalation rates were 11, 2.6, and 0.02 umoles/hour per kg for CC>2, parent compound, and
chloroform, respectively; the timing of the peak rates occurred in 15-45 minutes, within 2 hours,
and slightly after 2 hours for CC>2, parent compound, and chloroform, respectively. At 4,000
mg/kg, peak exhalation rates were 88, 1,550, and 3.4 umoles/hour per kg for CC>2, parent
compound, and chloroform, respectively; compared with the lower doses, peak rates were
achieved more quickly for CC>2 than for parent compound and chloroform.
In monkeys exposed by inhalation to radiolabeled carbon tetrachloride at 46 ppm for 5.75
hours, 21% of the total absorbed dose was eliminated during the initial 18 hours as carbon
dioxide and parent compound or volatile metabolite (McCollister et al., 1951). Within 75 days
following the end of exposure, 11% was eliminated as carbon dioxide and 40% was eliminated
as parent compound or volatile metabolite in exhaled breath. The majority of urinary and fecal
excretion occurred in the 5 days following exposure; a small amount of label was detectable in
feces after 12 days and in urine after 15 days.
In rats exposed to radiolabeled carbon tetrachloride vapor by inhalation at 100 or 1000
ppm for 8 hours for 1-5 days, no fecal elimination was detected (Page and Carlson, 1994); in
comparison, intravenous administration resulted in biliary and nonbiliary fecal elimination that
was <1% of the administered dose.
Sanzgiri et al. (1997) measured the elimination of carbon tetrachloride from tissues in rats
exposed to 1,000 ppm via inhalation for 2 hours or the equivalent oral dose of 179 mg/kg
administered as a single bolus dose or by intragastric infusion over 2 hours. The ti/2 of
elimination from various tissues are given in Table 3-3. Elimination ti/2 values were slowest for
fat, which is poorly perfused, but similar for the other tissues.
15 DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-3. Elimination ti/2 and apparent clearance of carbon tetrachloride
from rat tissues following administration of 179 mg/kg (1,000 ppm, 2
hours) by inhalation, oral bolus dosing, or gastric infusion over 2 hours
Tissue
Liver
Kidney
Lung
Brain
Fat
Heart
Muscle
Spleen
Inhalation
tl/2
(minutes)
249
204
226
248
665
274
218
273
Clearance
(mL/minute/kg)
63
58
61
55
0.8
70
55
88
Oral bolus
tl/2
(minutes)
323
278
442
313
780
490
649
472
Clearance
(mL/minute/kg)
175
59
62
42
0.8
65
43
44
Gastric infusion
tl/2
(minutes)
269
190
249
250
358
216
262
208
Clearance
(mL/minute/kg)
1198
224
72
67
1
94
83
108
Source: Sanzgiri et al. (1997).
Benson et al. (2001) compared elimination parameters in rats, mice, and hamsters
exposed to 20 ppm of [14C]-labeled carbon tetrachloride for 4 hours. In the 48 hours following
exposure, approximately 65-83% of the initial body burdens were eliminated as volatile organic
compounds or CC>2 in exhaled air. Elimination half-times were 7.4, 8.8, and 5.3 hours for CC>2
and 4.3, 0.8, and 3.6 hours for the volatile organic compounds for rats, mice, and hamsters,
respectively. Elimination in the urine and feces combined constituted <10% of the initial body
burden in rats and <20% in mice and hamsters.
Paustenbach et al. (1986a, b) and Veng-Pedersen et al. (1987) compared the
pharmacokinetics of carbon tetrachloride in rats exposed to 100 ppm of carbon tetrachloride
vapor in scenarios that mirror human work schedules: 8 hours/day for 5 days or 11.5 hours/day
for 4 days. Additional groups were exposed on a 2-week schedule for 5 or 3 additional days,
respectively. Following 2 weeks of exposure at 8 hours/day, 45% of the label was eliminated in
exhaled air (-97.5% as parent compound) and 48% was eliminated in feces. Exposure at 11.5
hours/day for 2 weeks resulted in elimination of 32% in exhaled air and 62% in feces. On either
schedule, <8% was excreted in urine and <2% was exhaled as CCh. The elimination profiles for
exhaled air were biphasic. For the 2-week 8 hours/day and 11.5 hours/day schedules,
elimination of the parent compound in breath had ti/2 values for the fast and slow phases of 96
and 455 minutes and 89 and 568 minutes, respectively. Similarly, ti/2 values for the fast and
slow phases of elimination of CC>2 were 305 and 829 minutes on the 8-hour schedule and 455
and 1824 minutes on the 11.5-hour schedule. The authors concluded that the longer daily
exposure placed more of the absorbed dose into the poorly-perfused fat compartment. The ti/2 of
16
DRAFT - DO NOT CITE OR QUOTE
-------
elimination in urine and feces for the 2-week exposures were 1,066 and 3,700 minutes for the 8-
hour schedule and 944 and 6,700 minutes for the 1 1.5-hour schedule.
Rats or gerbils intraperitoneally injected with carbon tetrachloride at a dose of 128-159
mg/kg eliminated 80-90% in exhaled air as carbon tetrachloride and less than 1% as CCh
(Young and Mehendale, 1989).
3.5. PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS
Physiologically based pharmacokinetic (PBPK) models are available for carbon
tetrachloride for exposures by the inhalation route (Yoon et al., 2007; Fisher et al., 2004; Thrall
et al., 2000; Benson and Springer, 1999; Evans et al., 1994; Paustenbach et al., 1988, 1987;
Gargas et al., 1986) and the oral route (Fisher et al., 2004; Semino et al., 1997; Gallo et al.,
1993). The models are based primarily on experimental data from rodents. However, Thrall et
al. (2000) derived in vivo metabolic rate constants for humans based on human in vitro metabolic
constants and in vivo/in vitro ratios for metabolic rate constants derived from animals (also
reported in Benson and Springer, 1999).
Gargas et al. (1986)
Gargas et al. (1986) used the PBPK model framework developed by Ramsey and
Andersen (1984) for styrene, together with experimentally derived tissue partition coefficients
and gas uptake data for carbon tetrachloride, to estimate in vivo metabolic rate constants for
carbon tetrachloride in rats. The model comprises a series of differential equations describing
the rate of carbon tetrachloride entry into and exit from a series of body compartments, including
liver, fat, muscle, and viscera (richly perfused organs), as well as arterial and venous blood.
Gas-uptake data were obtained in a closed recirculated exposure system. Partition coefficients
were experimentally derived in a series of in vitro studies using the tissues of interest. The
researchers found that the uptake kinetics of carbon tetrachloride were adequately described by
modeling metabolism of the compound as a single saturable process with a maximum velocity of
enzyme reaction (Vmax) of 0.92 jimol/hour (0.14 mg/hour) and a Michaelis-Menten constant (Km)
of 1.62 |imol/L (0.25 mg/L).
Paustenbach et al. (1988, 1987)
Paustenbach et al. (1988, 1987) developed a four-compartment PBPK model (similar in
structure to Gargas et al., 1986) to describe the disposition of carbon tetrachloride absorbed
during inhalation, based on the framework developed by Ramsey and Andersen (1984) and the
parameter values reported by Gargas et al. (1986). Metabolism, assumed to occur only in the
liver compartment, was modeled as a single, saturable pathway. Metabolites were apportioned
into three separate storage compartments, leading to elimination in the exhaled breath, urine, and
feces, respectively. In order to accommodate the observed biphasic elimination
1 7 DRAFT - DO NOT CITE OR QUOTE
-------
equations were included to allow for the interconversion from the urinary or fecal pools to
production of CC>2. The model also included a time delay of 23.5 hours for fecal excretion to
account for the observed delay in appearance of radioactivity in the feces. Parameter values
needed to run the model included partition coefficients (determined experimentally by vial
equilibration), biochemical constants for carbon tetrachloride metabolism (determined
experimentally by gas uptake studies), and physiological parameters (estimated from the
literature, from previous pharmacokinetic studies, and from the process of fitting the carbon
tetrachloride data during model development). Selection of the optimal parameters for fat
compartment volume, blood flow, Vmax, and Km were determined by the quality of the visual fit
of the model predictions with laboratory data; sensitivity analysis indicated that changes to other
parameters had little effect on the simulation and were thus not subject to optimization. Model
parameters are presented in Table 3-4. Calibration of the rat model was done using data for
Sprague-Dawley rats exposed to 100 ppm of carbon tetrachloride for 4, 5, 7, or 10 exposures as
reported in Paustenbach et al. (1986a, b). The model reliably predicted values for the following
experimental parameters: concentration of [14C] activity in adipose tissue, concentration of [14C]-
carbon tetrachloride in the expired breath, concentration of 14CC>2 in the expired breath, activity
of 14C in the urine, and activity of [14C] in the feces.
18 DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-4. Physiological parameters for the rat, monkey, and human
PBPK models for carbon tetrachloride
Parameter
Cardiac output (L blood/hour)3
Alveolar ventilation (L air/hour)a
Rat (0.42 kg)
8.
8
Monkey (4.6 kg)
46.4
46.4
Human (70 kg)
358
358
Tissue volumes (percent of total)
Liver
Fat
Muscle
Richly perfused organs
4
8
74
5
4
10
72
5
4
20C
62
5
Blood flow (percent of total)
Liver
Fat
Muscle
Richly perfused organs
Metabolism
Vmax (mg/hour)d
Km (mg/L)
25
4
20
51
25
4
20
51
25
6
18
51
0.35
0.25
1.91
0.25b
12.72
0.25b
a Allometrically scaled from 15 L/hr x body weight (BW)
b Assumed to be the same as in rats.
0 Tissue volume for fat in humans is shown in Table 2 of Paustenbach et al. (1988) as 10%; however, the text of
this paper states that the rat model was scaled up to humans using a fat compartment of 20% of body weight. The
20% value was determined to be correct.
d Allometrically scaled from 0.65 mg/hour x BW07.
Source: Paustenbach et al. (1988).
In order to extend the model to monkeys and humans, the rat model was scaled up,
resulting in models for monkeys and humans that were used to predict the concentration of
carbon tetrachloride in expired air. For both the monkey model and the human model, cardiac
output, alveolar ventilation, and Vmax were estimated using (body weight)0'75, and the Km was
assumed to be the same as for the rat. The rat model was scaled to monkeys, using a body
weight (BW) of 4.6 kg, a body fat estimate of 10%, and fat perfusion of 4% of cardiac output;
other parameters were assumed to be the same as in the rat. The monkey model was calibrated
by using the data of McCollister et al. (1951), which measured the concentration of expired
carbon tetrachloride after a 370-minute exposure to 50 ppm. The time course was accurately
predicted, except for long periods (>240 hours) after exposure in which the model predicted
lower concentrations than were demonstrated experimentally. The study authors suggested that
small amounts (0.4%) of carbon tetrachloride may have been converted into C2Cle, which has a
19
DRAFT - DO NOT CITE OR QUOTE
-------
longer ti/2 in adipose tissue and would account for the slow elimination of small amounts of
radiolabel. The rat model was scaled up to humans by using an experimentally measured human
blood:air partition coefficient, a body weight of 70 kg, and a fat compartment of 20% BW.
Model simulations of concentration of carbon tetrachloride in expired air over time were
compared with the data of Stewart et al. (1961), who exposed volunteers to 49 ppm carbon
tetrachloride for 70 minutes or 10 ppm carbon tetrachloride for 180 minutes; there was good
agreement between the model simulation and the measured results. The model predicted that at
concentrations up to 100 ppm, the rat, monkey, and human metabolize carbon tetrachloride in a
similar manner. Because of physiological differences, the models predicted species differences
in carbon tetrachloride accumulation in fat. The rat PBPK model accurately described carbon
tetrachloride concentrations in adipose tissue where no significant day-to-day accumulation in fat
or blood was observed following repeated exposure to 100 ppm for 8 or 11.5 hours/day, whereas
the human model predicted day-to-day increases in carbon tetrachloride in fat following
inhalation exposure to 5 ppm for 8 hours/day.
Thrall et al. (2000); Benson and Springer (1999)
Thrall et al. (2000) and Benson and Springer (1999) expanded the rat PBPK model of
Paustenbach et al. (1988) to include parameters for the mouse and the hamster. The mouse and
hamster models consist of five compartments identical to the rat model (lung, liver, fat, muscle,
and richly perfused tissues). Metabolism is still assumed to occur only in the liver and is
modeled by a single, saturable pathway that results in products that may be eliminated in the
expired air, urine, or feces. For the mouse, tissue:air partition coefficients were assumed to be
equal to those for the rat, with the exception of the blood:air coefficient, which was measured
with the vial equilibration technique. Tissue:blood partition coefficients were then calculated by
dividing the tissue:air coefficients by the blood:air coefficients. Metabolic rate constants (i.e.,
Vmax and Km) were measured in whole animals by using gas uptake studies with a closed
recirculating chamber; in comparison to the rat, the mouse has a slightly higher capacity (higher
in vivo Vmax) and lower affinity (higher in vivo Km) for metabolizing carbon tetrachloride.
Physiological parameters for the mouse model were based on published values in the literature
(Andersen et al., 1987). Model predictions for initial body burden, exhaled carbon tetrachloride,
and exhaled CC>2 were compared with data collected over a 48-hour period following a 4-hour
inhalation exposure to 20 ppm of [14C]-carbon tetrachloride (data from a personal
communication and not presented in the manuscript); ratios of predicted/observed concentrations
ranged from 1.1 to 1.4, indicating good agreement among observed and predicted values. For
the hamster, coefficients for blood:air, muscle:air, liverair, and fatair were determined by the
vial equilibration technique. Hamster tissue:air partition coefficients did not differ significantly
from those of the rat. Tissue:blood partition coefficients were then calculated by dividing the
tissue:air coefficients by the blood:air coefficients. Metabolic rate constants (i.e., Vmax and Km)
20 DRAFT - DO NOT CITE OR QUOTE
-------
were measured in whole animals by using gas uptake studies with a closed recirculating
chamber; in comparison to the rat, the hamster has a higher capacity (higher in vivo Vmax) and
lower affinity (higher in vivo Km) for metabolizing carbon tetrachloride. Physiological
parameters for the hamster model were those used in the rat model. The hamster model tended
to overpredict uptake from exposure at low concentrations and underpredict the uptake from
exposure at high concentrations (1800 ppm exposure). Model predictions for initial body
burden, exhaled carbon tetrachloride, and exhaled CC>2 were compared with data collected over a
48-hour period following a 4-hour inhalation exposure to 20 ppm of [14C]-carbon tetrachloride;
ratios of predicted/observed concentrations ranged from 0.6 to 2.1 for all three species, and from
0.6 to 1.4 for rats and mice (Thrall et al., 2000; see Appendix C for a comparison of model
predictions and experimentally-derived data).
Thrall et al. (2000) and Benson and Springer (1999) used in vitro data on metabolism of
carbon tetrachloride by human liver microsomes (Zangar et al., 2000), together with in vitro and
in vivo rodent data, to estimate the in vivo human metabolic rate constants. The calculation is
presented in Table 3-5. Briefly, in vivo Vmax/Km ratios were obtained for the rodent species after
Vmax was normalized for milligrams of liver protein. The corresponding in vitro Vmax/Km ratios
were calculated in the same manner, and the in vivo/in vitro ratios were calculated, giving values
of 1.40, 1.01, and 1.70 for the rat, mouse, and hamster, respectively. As these values were
similar, a human in vivo Vmax/Km ratio of 1.37 was estimated as the mean of the rat, mouse, and
hamster ratios. Because the human Km in vitro is similar to that of the rat, the in vivo human Km
was assumed to be the same as that of the rat, allowing for the calculation of a human in vivo
Vmax of 29.15 mg/hour. The researchers used the new value for Vmax in the human PBPK model
of Paustenbach et al. (1988), with other parameters remaining as previously described, and
compared it with the human data of Stewart et al. (1961). The model simulation of expired
carbon tetrachloride levels provided good agreement with the experimental data, particularly at
longer periods postexposure (see Appendix C for a comparison of model predictions and
experimentally-derived data).
21 DRAFT - DO NOT CITE OR QUOTE
-------
Table 3-5. Comparison of metabolism from in vitro and in vivo studies
BW (kg)
Liver weight (g)a
mg protein/g liverb
In vivo Vmax (mg/hour/kg BW)C
In vivo Vmax (mg/hour)d
In vivo Vmax (mg/hour/mg protein)
In vivo Km (mg/L)°
In vivo Vmax/Km
In vitro Vmax ((imol/hour/mg protein)6
In vitro Km (|imol/L)e
In vitro Vmax/Km (L/hour/mg protein)
Ratio (in vivo/in vitro)
Rat
0.25
10
13.8
0.4
0.15
l.lxlO~3
0.25
4.4xlO~3
0.186
59.1
3.15xlO~3
1.4
Mouse
0.025
1
21.9
0.79
5.97xlO~2
2.7 x!0~3
0.46
5.9xlO~3
0.1712
29.3
5.86xlO~3
1.01
Hamster
0.15
6
17.8
6.39
1.69
0.016f
1.14
0.014g
0.246
30.2
8.14xlO~3
1.7
Human
70
2800
12.8
1.49
29.15
S.lxlO"1
0.25h
3.2xlO~3
0.135
56.8
2.38xlO~3
1.371
a Calculated as 4% of body weight.
bFrom Reitz et al. (1996), except hamster, which was estimated as the mean of mouse and rat.
0 Rodents: experimentally measured; humans: calculated (see text).
dRodents: calculated from in vivo Vmax (mg/hour/kg BW) using BW07 (personal communication; email dated
9/5/2006, from Dr. Karla Thrall, Pacific Northwest National Laboratory, to Susan Rieth, U.S. EPA); humans:
calculated (see text).
"Data from Zangar et al. (2000).
f Corrected from value of 0.16 in Table 5 of Thrall et al. (2000) (personal communication; email dated 9/5/2006,
from Dr. Karla Thrall, Pacific Northwest National Laboratory, to Susan Rieth, U.S. EPA).
8 Corrected from value of 0.14 in Table 5 of Thrall et al. (2000) (personal communication; email dated 9/5/2006,
from Dr. Karla Thrall, Pacific Northwest National Laboratory, to Susan Rieth, U.S. EPA).
h Assumed to be equal to the rat based on in vitro Km comparisons.
'Calculated as the average of the rat, mouse, and hamster in vivo/in vitro ratios.
Source: Thrall et al. (2000).
Other Extensions of the Paustenbach et al. (1988) Model
Several other models have been developed as extensions of the Paustenbach et al. (1988)
model. Semino et al. (1997) added a GI compartment to the inhalation model of Paustenbach et
al. (1988) to describe uptake of carbon tetrachloride administered by a single gavage dose at
levels of 25 or 50 mg/kg in corn oil or at a dose of 17.25 mg/kg in 0.25% aqueous Emulphor to
male F344 rats. The GI compartment was divided into a series of sequential absorption
subcompartments, each characterized by three parameters: emptying time, absorption rate
constant (describing input to the portal circulation), and bioavailability. These parameters were
optimized against the experimental results for concentrations of parent carbon tetrachloride in
arterial blood or exhaled air. The number of subcompartments was also varied; nine
subcompartments were needed to obtain a good fit of this data set for delivery by corn oil
22
DRAFT - DO NOT CITE OR QUOTE
-------
gavage, whereas only six or seven subcompartments were needed for aqueous Emulphor. The
model simulated the higher rapid initial uptake with the aqueous vehicle and the more pulsatile
absorption profile observed from corn oil delivery following a single exposure. The
subcompartments were not intended to correspond to actual anatomic segments of the GI tract,
and the values generated for oral uptake parameters were not intended to represent true
physiological measurements.
Thrall and Kenny (1996) adapted the PBPK model of Paustenbach et al. (1988) to
simulate an intravenous route of exposure in the male F344 rat. The model added equations to
simulate the introduction of carbon tetrachloride into the mixed venous blood pool.
Physiological parameters were adjusted to account for the smaller body size of F344 rats
compared with Sprague-Dawley rats, using data from Arms and Travis (1988). The model was
used to predict the concentration of carbon tetrachloride in the expired air after a single
intravenous exposure and was compared with real-time monitoring data from rats given a single
injection of carbon tetrachloride at 0.6 or 1.5 mg/kg BW. With the exception of underestimation
of the initial peak in exhalation, the model predictions were in good agreement with the
measured data.
El-Masri et al. (1996) modified the PBPK rat model of Paustenbach et al. (1988) to
include a linked physiologically based pharmacodynamic (PBPD) model for hepatocellular
injury and animal death. First-order rate constants governed simulated cell mitosis and birth,
injury (due to carbon tetrachloride-induced vacuolation and incidental injury), repair, delay of
mitosis and repair, cell death, and phagocytosis by macrophages. Animal death was simulated to
occur when >50% of hepatocytes died. The data of Lockard et al. (1983) were used to visually
optimize the PBPD model rate constants.
Other models of carbon tetrachloride disposition were developed independent of Thrall et
al. (2000) or Paustenbach et al. (1988) and are discussed further below.
Gallo et al. (1993)
Gallo et al. (1993) developed a physiological and systems analysis hybrid
pharmacokinetic model for blood concentration-time data obtained during intravenous or oral
administration. The systems analysis procedure was based on a disposition-decomposition
method for deriving an absorption input function for each regimen. Equations were derived,
representing input into the blood, distribution to and from the blood to the peripheral tissues, and
elimination from the blood, allowing for the estimation of arterial and venous blood
concentrations but not concentrations in target tissues. Experimental data were collected for
male Sprague-Dawley rats given a single oral dose of 25 mg/kg in one of four ways (undiluted,
in corn oil, as an emulsion in 0.25% Emulphor, or in water) and from other rats receiving the
same dose in aqueous polyethylene glycol 400 as an intravenous bolus injection. A hybrid
model that combined model parameters available in the literature with the absorption input
23 DRAFT - DO NOT CITE OR QUOTE
-------
functions obtained by systems analysis adequately described the observed blood concentration-
time data. The same model using conventional first-order absorption inputs provided less
accurate fits to the data. Both the standard model and the hybrid model overestimated the initial
concentration in blood for the oral or intravenous routes.
Evans et al. (1994)
Evans et al. (1994) developed a PBPK model for carbon tetrachloride in rats based on the
Ramsey and Andersen (1984) model for styrene. Flow-limited compartments for liver, fat, and
rapidly and slowly perfused tissues were connected by arterial and venous blood. The
investigators derived partition coefficients from blood, liver, fat, and muscle samples of naive
male Fischer-344 rats. Physiological parameter values were taken from the literature.
Metabolism of carbon tetrachloride was constrained to the liver and described by Michaelis-
Menten kinetics. Vmax and Km were estimated by optimizing the model to closed-chamber gas
uptake data, generated by the study authors, for adult male Fischer-344 rats exposed to 25, 100,
250, or 1,000 ppm carbon tetrachloride for 6 hours. The resulting Vmaxc and Km values were
0.37 mg/hour/kg and 1.3 mg/L, respectively. The predicted decreases in chamber carbon
tetrachloride concentrations were similar to observations for all exposure levels and time points.
A sensitivity analysis was performed on all of the model parameters. For the low exposure (25
ppm), the blood:air partition coefficient (5.49), followed by the fatblood partition coefficient
(51.3) and fat tissue volume (8%), had the greatest effects on simulated chamber concentration.
However, the fatblood partition coefficient and fat tissue volume dominated the decrease in
chamber concentration in the 1,000-ppm exposure.
The model of Evans et al. (1994) was applied to examine the effect of methanol
pretreatment of rats (10,000 ppm for 6 hours) at 24 and 48 hours prior to 6-hour closed-chamber
carbon tetrachloride exposures of 25, 100, 250, or 1,000 ppm (Evans and Simmons, 1996).
Vmaxc was optimized against the gas uptake data from all exposure levels. A Vmaxc value of
0.48 mg/hour/kg for the 24-hour methanol pretreatment group resulted in good agreement of the
predicted and observed chamber concentrations at all exposure levels, indicating that induction
of carbon tetrachloride metabolism could be adequately simulated. Good agreement was also
achieved between predicted and observed chamber concentrations at all exposure levels for the
48-hour methanol pretreatment group. The estimated Vmaxc value of 0.18 mg/hour/kg, which
was close to the carbon tetrachloride-only value of 0.11 mg/hour/kg (from Evans et al., 1994),
indicated that the effect of methanol induction of carbon tetrachloride metabolism had practically
ceased by this time.
Yoshida et al. (1999)
Yoshida et al. (1999) used a classical compartment pharmacokinetic model to derive rates
of absorption of carbon tetrachloride in rats exposed at low concentrations in a closed chamber
24 DRAFT - DO NOT CITE OR QUOTE
-------
system. Experimentally, rats were exposed at initial concentrations between 10 and 1,000 ppb,
and the changes in chamber concentrations were measured over 6 hours. The model, like the
experimental system, had three compartments: a tank containing barium chloride to capture the
compound, the exposure chamber into which the compound was injected, and the rat. The model
consisted of three differential equations describing the apparent volumes of distribution for the
three compartments. The model included single rate constants for inhalation, exhalation, and
metabolic elimination processes in the rat. The rate constant for exhalation was determined to be
higher than that for elimination. Metabolic elimination of carbon tetrachloride was estimated as
0.53 |imol/hour/kg at 10 ppm.
Andersen et al. (1996) developed a model to describe the anaerobic in vitro metabolism
of carbon tetrachloride in a two-phase, closed-chamber headspace vial. Data were generated
from hepatic microsomal preparations from fed or fasting adult male F344 rats. Partition
coefficients were experimentally derived for phosphate buffer to air and microsomal suspension
to air. In addition to the Michaelis-Menten kinetic constants, a first-order loss-rate constant was
required for accurate fitting of the model. The model described the kinetics of anaerobic
transformation of carbon tetrachloride to chloroform.
Fisher et al. (2004)
Fisher et al. (2004) developed a PBPK model for simultaneous exposures to carbon
tetrachloride and tetrachloroethylene in mice. The model contained a four-compartment
structure (liver, fat, and richly and slowly perfused tissues) for carbon tetrachloride based on the
Ramsey and Andersen (1984) model and tetrachloroethylene based on a modified form of the
Gearhart et al. (1993) model. Absorption from the GI tract was simulated as a two-compartment,
three-parameter model (Figure 3-2). Rate coefficients were estimated by visually fitting these
parameters to blood data following single gavage doses of carbon tetrachloride (20, 50, or 100
mg/kg carbon tetrachloride alone, 10 or 100 mg/kg tetrachloroethylene alone, and 1, 5, 20, 50, or
100 mg/kg carbon tetrachloride followed 1 hour later by 10 or 100 mg/kg tetrachloroethylene; all
oral bolus doses were administered in aqueous emulsion vehicle). Metabolism for both
chemicals was represented as a saturable Michaelis-Menten pathway in the liver only. Carbon
tetrachloride-induced suicide inhibition was modeled with a second-order inhibition constant,
KD, which was used to calculate the loss of metabolic capacity (Vmaxc) for both carbon
tetrachloride and tetrachloroethylene. A submodel for trichloroacetic acid, the sole metabolite of
tetrachloroethylene oxidation, was included in which the rate of trichloroacetic acid production
in the liver was equal to the rate of tetrachloroethylene metabolism. Four compartments for
trichloroacetic acid were included: liver, kidney, and rapidly and slowly perfused tissues.
25 DRAFT - DO NOT CITE OR QUOTE
-------
Liver
K1
K3
Oral dose
C1
K2
C1
Figure 3-2. Two-compartment model for simulating GI absorption of carbon
tetrachloride administered to mice as a single gavage dose in Emulphor
(Fisher et al., 2004).
Values for rate coefficients were derived by visual fit of model predictions to observed blood carbon
tetrachloride kinetics in mice. The value for Kl was dose dependent (0.4 hour"1 for 20 mg/kg dose and 10
hours"1 for 50 and 100 mg/kg doses). Values for K2 and K3 were 2 and 0.05 hour"1, respectively.
Carbon tetrachloride partition coefficients for blood, liver, fat, and muscle (representing
slowly perfused tissue) were determined by the study authors (Fisher et al., 2004) using the vial
equilibration method of Gargas et al. (1989). Partition coefficients for tetrachloroethylene and
trichloroacetic acid were taken from Gearhart et al. (1993) and Abbas and Fisher (1997),
respectively. Physiological constants for mice were taken from the compendium of Brown et al.
(1997). Data for carbon tetrachloride gas uptake exposures of 130 ppm (Thrall et al., 2000) and
50, 450, or 1250 ppm (Fisher et al., 2004) in male B6C3F1 mice were used to optimize Vmaxc
and Km, resulting in values of 1 mg/hour/kg0'75 and 0.3 mg/L, respectively. For
tetrachloroethylene, gas uptake-derived Vmaxc and Km values of 6 mg/hour/kg0'75 and 3 mg/L,
respectively, were taken from Gearhart et al. (1993). Oral absorption rate constants for carbon
tetrachloride and tetrachloroethylene were visually fitted from the blood concentration data for
each chemical. The value for KD was estimated by optimization of the model to blood
trichloroacetic acid concentrations following co-exposures of tetrachloroethylene and carbon
tetrachloride via oral bolus dosing. See Appendix C for a summary of parameter values used in
the Fisher et al. (2004) model.
Yoonetal. (2007)
Yoon et al. (2007) explored the effect of extrahepatic carbon tetrachloride metabolism in
rats and humans on estimates of hepatic Vmax and Km. The investigators developed an eight-
26
DRAFT - DO NOT CITE OR QUOTE
-------
compartment, flow-limited PBPK model, including compartments for lung, liver, brain, kidney,
fat, rapidly and slowly perfused tissues, and the gastrointestinal tract. Physiological parameter
values were taken from the literature (Delp et al., 1991; U.S. EPA, 2000e; Brown et al., 1997).
Tissue partition coefficients for the rat were taken from Evans et al. (1994). Gas uptake data
from closed-chamber experiments (Evans et al., 1994) were used to estimate values of Vmax (0.13
mg/kr/kg0'75) and Km (1.10 mg/L) in the liver. Data for estimation of extrahepatic metabolism
were generated from in vitro CYP2E1-mediated microsomal metabolism of carbon tetrachloride
in liver, brain, skin, kidney, lung, and fat. No metabolic activity was detected in the fat, brain, or
skin. Estimates of extrahepatic in vivo metabolism in the lung and kidney were modeled as the
liver Vmax adjusted by the tissue volume-normalized ratio of Vmax, in vitro tissue / Vmax, in vitro liver-
Simulations of open-chamber inhalation exposures (ATSDR, 2005) were used to compare the
effect of the presence or absence of extrahepatic metabolism on the following dose metrics:
carbon tetrachloride blood Cmax, AUC for carbon tetrachloride in blood over a 24-hour period,
total carbon tetrachloride metabolized in the body, and carbon tetrachloride metabolized in the
liver (normalized for liver volume). The presence or absence of extrahepatic metabolism did not
affect either the estimation of hepatic Vmax and Km or the predicted dose metrics. The proportion
of liver metabolism estimated for the lung and kidney was quite small, 0.79 and 0.93%,
respectively, based on the microsomal studies. This resulted in identical values for Vmax and all
of the examined dose metrics, and similar values for Km (1.10 and 1.14 mg/L without and with
extrahepatic metabolism, respectively).
Of the PBPK models developed for carbon tetrachloride, the model by Yoon et al. (2007)
is the only one that addressed extrahepatic carbon tetrachloride metabolism. Regarding
extrahepatic metabolism, it is noted that, although rat kidney cortex and proximal tubules express
reasonable levels of CYP2E1 protein and activity for the oxidative metabolism of another
CYP2E1 substrate, trichloroethylene (Cummings et al., 2001, 2000b, 1999), the human kidney
has been reported by multiple laboratories to not express any detectable CYP2E1 protein (Amet
et al., 1997; Cummings and Lash, 2000; Cummings et al., 2000a) and to exhibit little if any
oxidative metabolism of trichloroethylene (Cummings and Lash, 2000; Cummings et al., 2000a).
27 DRAFT - DO NOT CITE OR QUOTE
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL
CONTROLS
4.1.1. Oral Exposure
4.1.1.1. Human Poisoning Incidents
Case reports reveal that individuals acutely poisoned with carbon tetrachloride can
exhibit GI toxicity (nausea, vomiting, diarrhea, and abdominal pain) and neurotoxicity
(drowsiness, coma, or seizures) (Ruprah et al., 1985; Stewart et al., 1963; New et al., 1962).
Hepatic involvement has been demonstrated by liver enlargement and significant elevations in
serum enzyme (> 100-fold increases in alanine aminotransferase [ALT] or aspartate
aminotransferase [AST]) and bilirubin levels (Ruprah et al., 1985; Stewart et al., 1963). One of
two individuals who received one 5 mL dose of carbon tetrachloride as an antihelmintic
exhibited microscopic pathology in the liver (granular degeneration); a third person who received
a second dose 2 weeks later had fatty degeneration of the liver, as well as swelling of the
proximal tubules of the kidney (Docherty andNicholls, 1923; Docherty and Burgess, 1922).
Renal effects (oliguria and increases in blood urea nitrogen [BUN]) may occur within 1-8 days
of acute exposure (New et al., 1962). Umiker and Pearce (1953) noted that, after ingestion of
fatal doses of carbon tetrachloride, the primary cause of death during the first week was hepatic
injury and afterwards was renal insufficiency. Pulmonary lesions (lung congestion, edema,
bronchopneumonia, fibrinous exudate, alveolar epithelial proliferation) appear about 8 days after
exposure and have been considered to be secondary effects of renal failure (Umiker and Pearce,
1953). Human fatalities from ingestion of carbon tetrachloride may occur with ingestion of
amounts as low as 2-3 mL (45-68 mg/kg, based on the reference adult BW of 70 kg) (Ruprah et
al., 1985;Gosselinetal., 1976).
4.1.1.2. Epidemiology Studies
Epidemiological studies have investigated possible associations between oral exposure to
carbon tetrachloride and a variety of adverse birth outcomes (Croen et al., 1997; Bove et al.,
1995, 1992a, b); however, because of multiple chemical exposures and insufficient power, these
studies are considered limited and insufficient to determine whether there is an association
between carbon tetrachloride exposure and adverse birth outcomes.
Boveetal, 1995, 1992a,b
Bove et al. (1995, 1992a,b) evaluated the relationship between contamination of public
drinking water with organic compounds (including carbon tetrachloride) and adverse birth
outcomes in a cross-sectional study of births in four counties in northern New Jersey. The study
28 DRAFT - DO NOT CITE OR QUOTE
-------
population consisted of registered live births and fetal deaths occurring from January 1, 1985, to
December 31, 1988, in 75 towns (selected from a total of 146 in the four counties), where most
residents were served by public water systems and most births occurred in the state. After
exclusion of plural births and fetal deaths from therapeutic abortions or chromosomal anomalies,
the subjects totaled 80,938 live births and 594 fetal deaths. Fetal death certificates available for
all fetal deaths with gestational age greater than 20 weeks and the New Jersey Birth Defects
Registry were used to gather data on a selection of adverse birth outcomes. A comparison group
of 52,334 births that had no adverse outcomes was included in the study to evaluate categorical
outcomes. Exposure to organic compounds was estimated from the monthly records of the 49
water companies serving the study population (water samples were collected at the tap). In
addition to carbon tetrachloride, other contaminants in the drinking water included
trihalomethanes (primarily chloroform), 1,2-dichloroethane, dichloroethylenes, 1,1,1-
trichloroethane, trichloroethylene, tetrachloroethylene, and benzene. Levels of all of these
compounds, other than benzene, were higher than carbon tetrachloride; levels of trihalomethanes
were 20- to 40-fold higher. For carbon tetrachloride, the exposed population was defined in one
of two ways: those with exposure to >1 ppb in the drinking water or those with any detectable
amount in the drinking water. In either case, the size of the comparison group with exposure to
carbon tetrachloride was small: 357 births where levels >1 ppb were detected and 1993 births
where any carbon tetrachloride was detected.
Carbon tetrachloride and the other contaminants were evaluated for effects on 13 selected
birth outcomes (birth weight among term births, term low birth weight, small for gestational age,
preterm birth, low birth weight, fetal death, central nervous system defects, neural tube defects,
oral clefts, major cardiac defects, ventricular septal defects, all cardiac defects, and all
surveillance defects). Odds ratios (ORs) for an association between each outcome and carbon
tetrachloride were calculated as the ratio of the risk of the outcome in the population with the
specified exposure (either > the detection limit or >1 ppb) to the risk in the population without
the specified exposure. ORs were adjusted for maternal age, race, education, parity, adequacy of
prenatal care, and sex of the child. Positive associations were found between exposure to carbon
tetrachloride in drinking water at concentrations above 1 ppb and certain adverse outcomes: low
birth weight (<2.5 kg) among term births (OR = 2.26, 95% confidence interval [CI]: 1.41-3.60)
and small (at or below their race-, sex- and gestation week-specific 10th percentile weight) for
gestational age (OR = 1.34, 95% CI: 1.02-1.80). These same effects, however, were also
significantly associated with exposure to trihalomethanes, which were present in higher levels
and were more prevalent in the drinking water supply (i.e., had a larger exposed population and
number of cases). While there was a statistically positive association between exposure to >1
ppb carbon tetrachloride and occurrence of neural tube defects (OR = 5.39, 95% CI: 1.31-22.2),
it was based on only two cases in the exposed population. Using a criterion of OR >1.5 without
consideration of CIs, the authors also reported positive relationships between carbon
29 DRAFT - DO NOT CITE OR QUOTE
-------
tetrachloride and several of the other adverse outcomes tested. However, the reliability of these
purported relationships is suspect without statistical support. Maternal interviews were
conducted for a sample of the study population to collect more detailed information about
potential confounders, such as maternal occupational exposures, smoking, medical histories,
height, and gestational weight gain. Adjustment for these additional risk factors had no
appreciable effect on the results for carbon tetrachloride. Interpretation of the study results is
hindered by simultaneous exposure to multiple chemicals in the drinking water, the relatively
small number of people exposed to carbon tetrachloride and the low levels to which they were
exposed, and the limited characterization of exposure to carbon tetrachloride (and the other
chemicals tested).
Croenetal, 1997
Croen et al. (1997) used data from two population-based case-control studies to
determine whether maternal residential proximity to hazardous waste sites increased the risk for
certain birth defects in California. Residential histories were obtained by interviews with
mothers of infants with specific birth defects (neural tube defects [507 cases] in one study; heart
defects [201 cases] and oral cleft defects [439 cases] in the other) and mothers of controls in the
two studies (517 for the neural tube study and 455 for the other two defects). Information was
collected on 764 inactive waste sites as well as 105 National Priority List sites. Multivariate
analysis was used to control for potential confounding effects, such as maternal race/ethnicity,
income, and education. The study found no increased risk of heart defects or oral cleft defects
among offspring of mothers living near a waste site containing carbon tetrachloride, but this
study had little power to detect effects. ORs for neural tube defects associated with carbon
tetrachloride were not provided.
4.1.2. Inhalation Exposure
4.1.2.1. Acute Exposure Incidents
The initial acute effects of carbon tetrachloride in humans exposed by inhalation are
similar to effects reported from humans exposed orally (Stewart et al., 1965; New et al., 1962;
Norwood et al., 1950); these effects include GI symptoms (nausea and vomiting, diarrhea,
abdominal pain), hepatic effects (elevated serum AST, mild jaundice, and, in fatal cases, necrosis
of the liver), and neurological effects (headache, dizziness, weakness). As with acute oral
exposure, inhalation exposure causes renal effects (oliguria, elevated BUN) that appear 1-8 days
after exposure, with an average delay of 4 days (New et al., 1962). Renal histopathological
effects in fatal cases include nephrosis, degeneration, and interstitial inflammation of the kidney
(Norwood et al., 1950). Pulmonary edema is a secondary consequence of renal insufficiency
(Umiker and Pearce, 1953; Norwood et al., 1950). Some case reports noted that a high intake of
alcohol, which can enhance carbon tetrachloride toxicity, was common among the patients
3 0 DRAFT - DO NOT CITE OR QUOTE
-------
intoxicated by inhaled carbon tetrachloride (New et al., 1962; Norwood et al., 1950).
Lehmann and Schmidt-Kehl (1936) described the neurological symptoms in humans
exposed briefly to carbon tetrachloride vapor at concentrations of >20 mg/L (>3200 ppm). No
effect was observed following exposure at 20 mg/L for 5 minutes. Exposure to 30 mg/L (4800
ppm) for 2.5 minutes resulted in slight drowsiness after 5 minutes. Exposures to 40 mg/L (6400
ppm) for 3 minutes resulted in tremor and drowsiness, followed by staggering. The highest
tested exposure, 89 mg/L (14,100 ppm) for 0.8 minutes, resulted in loss of consciousness.
Stewart et al. (1961) reported no adverse effects (such as nausea or dizziness) in male volunteers
exposed to carbon tetrachloride vapor at 49 ppm for 70 minutes or 10-11 ppm for 180 minutes.
4.1.2.2. Epidemiology Studies
Occupational exposure to unknown concentrations of carbon tetrachloride vapor for
periods between 6 weeks and 3 months resulted in GI effects (nausea, vomiting, abdominal pain,
anorexia), hepatic effects (jaundice), and neurological effects (headache, dizziness) (Norwood et
al., 1950). Kazantzis and Bomford (1960) described symptoms in 17 workers exposed to carbon
tetrachloride vapor at concentrations between 45 and 97 ppm without adequate ventilation.
Symptoms in 15/17 workers included anorexia and nausea and, in more than half of the workers,
vomiting, epigastric discomfort or distension, depression, irritability, headache, or giddiness.
Symptoms typically developed in the latter half of the workweek and cleared over the weekend.
One of the workers, who reported having symptoms for 2 years, previously had an increased
serum AST level, but levels were normal for this individual and seven others examined by the
authors for this study. Similarly, Elkins (1942) reported results of industrial hygiene evaluations
in 11 plants in which workers were exposed to carbon tetrachloride vapor. At concentrations
between 5 and <85 ppm, nausea was the most common symptom, but vomiting, headache, and
body weight loss were also observed.
Tomenson et al., 1995
Tomenson et al. (1995) conducted a cross-sectional study of hepatic function in 135
carbon tetrachloride-exposed workers in three chemical plants in northwest England and in a
control group of 276 unexposed workers. The latter came from two sites, including one of the
plants that provided workers for the exposed group and a plant nearby where carbon tetrachloride
was not used. Controls had not held jobs with potential exposure to carbon tetrachloride or other
known hepatotoxins during the previous 5 years. Subjects were administered a questionnaire
that collected information on medical history, alcohol consumption, and length of service in a job
exposed to carbon tetrachloride. Blood samples were obtained from subjects after a 12-hour fast
that included abstinence from alcohol; samples were collected for about 60 subjects over 2 weeks
in November 1986 and for the remaining subjects over 8 weeks starting in February 1987. Blood
samples were analyzed for ALT, AST, alkaline phosphatase (ALP), y -glutamyl transferase
31 DRAFT - DO NOT CITE OR QUOTE
-------
(GGT), glutamate dehydrogenase (GDH), 5'-nucleotidase, total bile acids (TEA), cholesterol,
triglycerides, and hematological variables.
The exposure assessment was based on historical personal monitoring data for various
jobs at the three plants. Subjects were placed into one of three exposure categories (low,
medium, or high), according to their current jobs. When objective monitoring data were not
available for a particular combination of job and location (as was the case for 23/40 in the low-
exposure group, 35/54 in the medium-exposure group, and 2/61 in the high-exposure group), an
industrial hygienist classified the exposure qualitatively based on comparison with similar
groups. The quantitative exposure levels nominally associated with each of these categories
were: <1 ppm for "low," 1.1-3.9 ppm for "medium," and 4 ppm-11.9 ppm for "high." Exposed
workers were also categorized according to length of time in job (<1 year, 1-5 years, and >5
years).
Study and control groups were found to be well matched for age, height, weight, work
patterns, and, generally, alcohol consumption. Almost all (97-98%) control and exposed
workers were current drinkers, and the proportions of low, medium, and high alcohol drinkers
r\
were roughly similar in the two groups (p = 0.30 for % comparison of 4 levels of alcohol use
between exposed and non-exposed). However, there was a slightly higher proportion of very
high drinkers (5-7 units every day or >8 units at least 3-4 times/week) in the exposed group
r\
(27%) than in controls (20%) (p = 0.20 for % comparison of high alcohol use between exposed
and non-exposed). Serum levels of GGT, bile acids, and triglycerides were significantly
increased in the high and/or very high alcohol consumption groups. In addition, serum levels of
GGT, cholesterol, triglycerides, AST, and 5'-nucleotidase were found to be significantly related
to age. Ages of workers in both control and exposed groups were approximately normally
distributed, with similar means and ranges.
Analysis of variance was used to investigate the relationship between carbon
tetrachloride exposure and serum chemistry and hematology variables, while controlling for age,
sampling time, and alcohol consumption. Initial analyses also included an interaction term
between carbon tetrachloride and alcohol consumption, but no evidence for any interaction was
found and the term was dropped from subsequent analyses. No analyses based on length of time
on job (i.e., duration of exposure) were presented in the published paper.
Multivariate analysis, based on simultaneous consideration of ALT, AST, ALP, and GGT
as dependent variables, revealed a statistically significant (p < 0.05) difference between exposed
and unexposed workers. There was no evidence, however, of a dose-response across the levels
of exposure. In univariate analyses, in which each dependent variable was assessed separately,
there were no significant differences between the carbon tetrachloride-exposed group and the
control group for any of the serum chemistry variables. However, there was evidence of
increased levels of ALP and GGT in the medium- and high-exposure groups, with the
differences between the medium-exposure group and controls being statistically significant (p <
32 DRAFT - DO NOT CITE OR QUOTE
-------
0.05) (see Table 4-1). GDH was significantly increased in the medium-exposure group, but
declined in the high-exposure group to the level seen in controls (see Table 4-1). There was little
difference in the mean adjusted serum ALT, AST, bile acids, and 5'-nucleotidase levels across
exposure categories.
Table 4-1. Mean of selected serum chemistry and hematology variables in
relation to carbon tetrachloride exposure in British chemical workers
Variable"
ALT (mU/mL)b
AST (mU/mL)b
ALP (mU/mL)b
GOT (mU/mL)b
GDH (mU/mL)b
TEA (umol/L)b
5'-Nucleotidase (mU/mL)
Hemoglobin (g/dL)
Packed cell volume (%)
Red blood cell count (* 1012/L)
Control
20.54(1.03)
16.48(1.02)
125.79(1.02)
26.89(1.05)
3(1.05)
1.06(1.06)
5.89(1.03)
15.97(0.08)
48.54 (0.23)
5.61 (0.03)
Exposure group
Low
20.35(1.08)
15.25(1.05)
122.2(1.05)
26.89(1.11)
3.26(1.10)
1 (1.00)
6.54(1.08)
15.6(0.19)
47.32C (0.54)
5.5 (0.08)
Medium
20.82(1.05)
15.88(1.04)
137. 10C (1.04)
33.17C(1.08)
3.57C(1.07)
1.25(1.25)
6.25(1.06)
15.39C(0.14)
47.32C(0.39)
5.47C (0.06)
High
19.39(1.06)
15.62(1.04)
135.1 (1.04)
31.5(1.08)
2.98(1.07)
1.28(1.28)
5.75(1.06)
15.71 (0.14)
48.05 (0,41)
5.5 (0.06)
"Results are presented as least square means, adjusted for age, sampling time, and alcohol consumption.
b Analyzed after logarithmic transformation; values are geometric means with standard error of the mean (SEM).
°/><0.05 (pairwise comparison).
Source: Tomensonet al. (1995).
Statistically significant changes were found for some of the hematological variables
(decreased red blood cell count, hemoglobin, and packed cell volume) in the univariate analyses
but without a dose response. Compared with the unexposed controls, there were slight (2.5-
3.5%) statistically significant decreases in all three of these variables in the medium-exposure
group and in packed cell volume in the low-exposure group (Table 4-1). Values for all three
hematological variables were similar to controls in the high-exposure group.
In an alternative analysis, a normal range was determined for each serum chemistry and
hematology variable based on the 2.5 and 97.5% quantiles in the control group. The proportion
of exposed workers exceeding the normal range was significantly elevated for ALT (8%) and
GOT (11%) but not for the other serum chemistry or hematology variables. This analysis did not
include any adjustment for alcohol intake or other potential confounders. The researchers noted
that, for the serum chemistry variables, the upper normal limits defined based on the control
group were notably higher than the upper limits of the reference ranges for these tests supplied
by the manufacturers, indicating a difference between the control group and the population used
33
DRAFT - DO NOT CITE OR QUOTE
-------
to derive the reference values, which are often hospital or university employees. This may have
been related to high alcohol consumption in the study controls, whose alcohol intake was similar
to the exposed group.
Individuals with one or more test results in excess of three standard deviations (SDs)
outside the control group mean were examined by a gastroenterologist. One exposed worker had
clinically detectable liver disease, but this could not be related to exposure to carbon
tetrachloride. The only other clinical findings were non-Hodgkin's lymphoma (NHL) in an
exposed worker and hemochromatosis in a control worker.
The observed decreases in hemoglobin, packed cell volume, and red blood cell count
were not considered to indicate a biologically significant effect of carbon tetrachloride, as the
observed changes were minimal and not clearly related to level of carbon tetrachloride exposure.
The results were generally suggestive of an effect on the liver, but were not consistent across the
liver variables or exposure levels. The overall difference seen in the multivariate analyses of the
four enzymes (ALT, AST, ALP, GOT) seemed to be driven by the increase in GOT, and to a
lesser extent in ALP, in the medium- and high-exposure groups. For GGT, the levels in the
medium and high carbon tetrachloride exposure groups were similar to the levels seen in the high
and very high alcohol use categories (geometric means of 30.04 and 32.32 mU/mL, respectively,
in these two alcohol use groups compared with 24.6 mU/mL in the low alcohol use groups).
There was little difference between the low carbon tetrachloride-exposure group (<1 ppm
estimated exposure levels) and the no-exposure group on any of the liver enzymes.
It is unclear to what extent the observed changes in serum enzyme levels reflect clinically
significant changes. The researchers suggest that their results show some enzyme leakage from
cells but without a measurable deficit in liver function (as assessed by total bile acid levels), and
they note that no effects of clinical significance were observed. Increased serum levels of ALT,
AST, ALP and GGT are indicators of liver damage (with ALP and GGT increased in exposed
workers), but none are specific for liver disease. Elevated ALP is used in the diagnosis of
hepatobiliary disease and bone disease, and elevated GGT is used in the diagnosis of liver
disease. The measurement of serum GGT levels can be used to ascertain whether observed
elevations of ALP are due to skeletal disease or reflect the presence of a hepatobiliary condition
(Tietz, 1976).
One limitation of the study is the lack of information pertaining to the reliability (e.g.,
coefficient of variation, comparison with known standards) of the enzyme measures. The
investigators noted that a follow-up study conducted at one site 3 years later revealed clear
evidence of differences in laboratory procedures between the laboratories that had performed the
testing of blood samples in the cross-sectional and follow-up studies. In addition, it was noted
that differences in the hematological variables (i.e., hemoglobin, packed cell volume, and red
blood count) were observed between the samples collected in November 1986 and those
collected in February and March of 1987.
34 DRAFT - DO NOT CITE OR QUOTE
-------
Overall, this study provides suggestive evidence of an effect from occupational carbon
tetrachloride exposure on hepatic serum enzymes, indicative of effects on the human liver.
Specifically, serum enzyme results suggested an exposure-related effect in the medium and high-
exposure categories (>l-3.9 ppm [>6.3-24.5 mg/m3] and 4-11.9 ppm [25.2-75 mg/m3]). ALP
and GGT were elevated to a similar degree in both medium- and high-exposure categories
(although the difference was statistically significant only in the medium-exposure category), and
enzyme levels in these exposure groups were comparable to the levels of ALP and GGT seen in
very high alcohol consumers. Confidence in the exposure monitoring for the medium-exposure
group is relatively low, where exposures were estimated for over half (35/54) of the workers.
Confidence in the exposure monitoring for the high-exposure group, where exposures were
measured for 59/61 workers, is higher. Because enzyme levels in these two groups were
comparable, an average concentration of the medium- and high-exposure groups (weighted by
number of subjects within specific exposure ranges) of 5.5 ppm (35 mg/m3) was considered to be
an estimate of the lowest-observed-adverse-effect level (LOAEL).b No effects on serum enzyme
levels were seen in the low-exposure category (i.e., <1 ppm [<6.3 mg/m3]). Because exposures
were estimated for more than half (23/40) of the workers in this exposure category and because
this category covers exposures <1 ppm, a no-observed-adverse-effect level (NOAEL) could not
be determined.
Seidleretal, 1999
Seidler et al. (1999) evaluated the association between maternal occupational exposure to
chemicals and the risk of infants small for gestational age in singleton births in a prospective
' An average exposure concentration for medium and high exposure categories (weighted by number of subjects
within specific exposure ranges) was calculated as follows using data in the appendix to Tomensen et al. (1995):
Exposure category
Medium
High
Sum
Average cone, for medium and
high exposure categories (ppm)
Exposure cone, (ppm)
[mid-point of range]
1.5
2.5
3.5
2.5 (estimated)3
5
7
9
11
8 (estimated)3
Number of
subjects
4
10
5
35
14
14
16
15
2
115
Product of cone, x number of subjects
(ppm-subject)
6
25
17.5
87.5
70
98
144
165
16
629
5.5b
a Estimated exposures were assumed to be the mid-point of the exposure category.
b Average calculated as the sum of the product of exposure concentration x number subjects for the individual
exposure ranges in the medium and high exposure categories divided by the total number of subjects, or 629 ppm-
subject -M15 subjects = 5.5 ppm.
35
DRAFT - DO NOT CITE OR QUOTE
-------
cohort study of 3,946 pregnant women in West Germany from 1987 to 1988. The final group of
1,865 women included those who completed a questionnaire on sociodemographic, psychosocial,
nutritional, environmental, and occupational factors, for whom pregnancy outcomes were known
and who were working at the time of the interview. Women with stillbirths, multiple births, and
incompletely recorded outcomes were excluded. A semi quantitative job-exposure matrix,
incorporating consideration of likelihood of exposure, intensity of exposure, and proportion of
time at work, was used to classify occupational exposure to eight chemicals or chemical groups,
including carbon tetrachloride. ORs were calculated, adjusting for age, smoking status, alcohol
consumption, body mass index, number of former births, and income as potential confounders.
The study found no association between occupational exposure to carbon tetrachloride and the
risk of infants small for gestational age. The power of this study was limited. Of the 1,865
births, only 64 mothers had potential exposures to carbon tetrachloride characterized as "low" or
"moderate."
Cancer studies
Several epidemiological studies have investigated potential associations between cancers
of various types and exposure to carbon tetrachloride. The subjects of all of these studies
experienced multiple chemical exposures, and the exposures were estimated qualitatively based
on historical information. These studies, therefore, can provide only suggestive evidence for
such associations.
Exposure to carbon tetrachloride was not found to be associated with cancer risk in case-
control studies for astrocytic brain cancer in white males (300 cases and 320 controls) from three
areas of the United States where a high proportion of the workforce is employed in petroleum
refining and chemical manufacture (after adjustment for several potential confounders)
(Heineman et al., 1994), for lung cancer in male employees (308 cases and 588 controls) of a
Texas chemical plant (Bond et al., 1986), for pancreatic cancer in residents (63,097 cases and
252,386 controls) from 24 U.S. states (Kernan et al., 1999), for renal cell carcinoma in
Minnesota residents (438 cases and 687 controls) (Dosemeci et al., 1999), for rectal cancer in
Montreal residents (257 cases and 533 controls) (Dumas et al., 2000), or for lymphoma in a
population (age 18-80 years) recruited from six study regions in Germany. In the general
population-based case-control studies (Seidler et al., 2007; Kernan et al., 1999; Dosemeci et al.,
1999; Dumas et al., 2000), occupation/industry information obtained from questionnaires,
interviews, or death certificates in combination with a job exposure matrix was used to
characterize chemical exposures. There was evidence for a weak association between exposure
to carbon tetrachloride and excess risk for breast cancer among white female residents of 24 U.S.
states; the OR was 1.21 (95% CI: 1.1-1.3) for those thought to have had the highest intensity of
exposure to carbon tetrachloride [based on occupation listed on death certificates] (Cantor et al.,
1995). Among white male workers at a rubber manufacturing plant in Akron, Ohio, there was a
3 6 DRAFT - DO NOT CITE OR QUOTE
-------
significant age-adjusted association between exposure to carbon tetrachloride and death from
lymphosarcoma (6 exposed out of 9 cases, OR = 4.2, p < 0.5) and lymphocytic leukemia (8
exposed out of 10 cases, OR = 15.3,/K0.001) (Wilcosky et al., 1984; Checkoway et al., 1984).
Kubale et al. (2005) reported that exposure to solvents (including carbon tetrachloride and
benzene) was significantly associated with leukemia mortality in civilian workers at the
Portsmouth Naval Shipyard in Kittery, Maine (OR = 1.03, 95% CI: 1.01-1.06). The findings
with respect to carbon tetrachloride are uncertain, however, because solvent exposures cannot be
separated, exposure misclassification was considered likely, and the phase-out of carbon
tetrachloride began in 1948, whereas the cohort considered deaths between 1952 and 1996. No
case-control studies were identified that looked for an association between carbon tetrachloride
and liver tumors or adrenal gland tumors (the tumor types found in laboratory bioassays with
carbon tetrachloride).
Spirtas et al. (1991) conducted a retrospective cohort study of 14,457 aircraft
maintenance workers at Hill Air Force Base in Utah to evaluate mortality associated with
workplace exposures, particularly trichloroethylene. Carbon tetrachloride was one of more than
20 chemicals included in the study. Increased mortality was found for NHL in white female
workers who had been exposed to carbon tetrachloride, in comparison with the Utah population
(Spirtas et al., 1991). However, in a follow-up study of the same cohort (Blair et al., 1998) that
extended the follow-up of worker mortality from 1982 to 1990, the relative risk (calculated as the
ratio of the rate of NHL mortality in the exposed and unexposed portions of the cohort, adjusted
for date of birth, calendar year of death, and sex) of NHL mortality was not significantly
increased in the female cohort (relative risk = 3.3, 95% CI: 0.9-12.7). A cohort of dry cleaners
in St. Louis, Missouri, showed slight significant excesses for deaths from all cancers
(standardized mortality ratio [SMR] = 1.2, 95% CI: 1.0-1.3), esophageal cancer (SMR = 2.1,
95% CI: 1.1-3.6), and cervical cancer (SMR = 1.7, 95% CI: 1.0-2.0) (Blair et al., 1990, 1979).
Risk of esophageal cancer was increased specifically in workers with the highest cumulative
exposure (SMR = 0.9, 0.3, and 2.8 in the low, medium, and high cumulative exposure
categories). There also appeared to be an increase in the risk of lymphatic and hematopoietic
cancers in the high-exposure group (SMR = 4.0), although this apparent increase was based on
only five cases. While some of these workers were likely to have been exposed to carbon
tetrachloride, no separate analysis was conducted for those exposed to carbon tetrachloride or
any other individual chemical. A cohort of Finnish laboratory workers exposed to carbon
tetrachloride and other chemicals showed no increased risk of cancer of any type, although the
average follow-up time of 15.7 years for the cohort may have been too short to reveal risks for
rare cancers with longer latency periods (Kauppinen et al., 2003).
An association between inhalation of carbon tetrachloride and liver cancer in humans was
suggested by two case reports (Tracey and Sherlock, 1968; Johnstone, 1948). Johnstone (1948)
reported the death of a 30-year-old female from liver cancer after 2-3 years of occupational
3 7 DRAFT - DO NOT CITE OR QUOTE
-------
exposure (assistant to a metallurgist) to carbon tetrachloride at levels that produced signs of
central nervous system toxicity, fatigue, and jaundice. Carbon tetrachloride exposure levels were
not assessed. Prior to carbon tetrachloride exposure, the woman had a history of "biliary colic"
and jaundice and had been studied for "gall bladder disease." A 66-year-old man died of
hepatocellular carcinoma 7 years after acute inhalation exposure from carpets that had been
cleaned with carbon tetrachloride (Tracey and Sherlock, 1968). The man was asymptomatic for
5 days after exposure but then developed vomiting, diarrhea, anuria, and jaundice. Although the
patient had no prior history of liver disease, he reported daily consumption of "several alcoholic
drinks"; the duration of alcohol consumption was not given. At the time of death, the liver tumor
was extensive, with little normal tissue remaining. The potential contribution of alcohol
consumption to liver disease in this patient could not be ruled out. Because of complicating
factors (e.g., alcohol consumption, previous history of liver disease), small number of individuals
involved, single exposure in one case, and relatively short time spans between exposure and
tumor appearance, a causal relationship between carbon tetrachloride and liver tumors cannot be
established from these case reports.
4.1.3. Dermal Exposure
There is evidence from one case report of health effects from exposure to carbon
tetrachloride that can at least partially be attributed to absorption across the skin (Farrell and
Senseman, 1944). The worker was exposed 8 hours/day by using a fine spray of carbon
tetrachloride to saturate a cloth wrapped around the fingers. Although some exposure is likely to
have occurred by inhalation, the authors considered absorption through the skin of the hands to
be the primary route of exposure. After an unspecified period of time at this job, the worker
developed polyneuritis. Symptoms included weakness, pain in the limbs, and loss or reduction
of certain reflexes. The patient, whose body weight was not reported, lost 8 pounds in the month
between onset of illness and hospitalization. The signs and symptoms of neurotoxicity reversed
after several months without exposure.
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS—ORAL AND INHALATION
Consistent with human data, toxicity assays in animals exposed orally or by inhalation
identify the liver to be the major target organ, with oral NOAELs between 0.71 and 0.86 mg/kg
and oral LOAELs between 7.1 and 17.8 mg/kg. Hepatic carcinogenicity has also been reported
in rats and mice exposed orally or by inhalation to carbon tetrachloride. While the liver appears
to be the primary target organ for both oral and inhalation studies, the kidney is also a sensitive
target organ for carbon tetrachloride exposure. Nephritis and nephrosis are common effects
following inhalation exposure to carbon tetrachloride.
3 8 DRAFT - DO NOT CITE OR QUOTE
-------
4.2.1. Oral Exposure
4.2.1.1. Subchronic Toxicity
Litchfield and Gartland, 1974
Litchfield and Gartland (1974) conducted a series of assays evaluating hepatic effects in
beagle dogs treated with carbon tetrachloride in gelatin capsules prior to their daily food intake.
In one experiment, groups of six male and six female young adult dogs were dosed with 797
mg/kg-day for up to 28 days. Blood samples taken before treatment and at 7-day intervals were
evaluated for serum ALT, AST, ALP, ornithine carbamoyl transferase (OCT), and creatine
kinase. At termination, livers were examined for histopathology. In a second experiment, three
female dogs were given 32 mg/kg-day for 8 weeks. Blood was sampled before treatment and at
2, 3, 5, 6, 7, and 8 weeks. Livers were examined for histopathology after sacrifice. Control
values were obtained from untreated dogs. No clinical signs of toxicity were observed. In dogs
treated at 797 mg/kg-day, increases in serum ALT levels (2- to 34-fold in 4/6 males and 6/6
females) and OCT (2- to 20-fold in 3/6 males and 6/6 females) were observed after 14-28 days.
All dogs exhibited hepatic histopathology (minimal to moderately severe centrilobular fatty
vacuolization, sometimes accompanied by single cell necrosis), the severity of which correlated
with the level of serum ALT and OCT in individual dogs. Dogs that showed no enzyme level
effect or a twofold increase only in ALT had minimal vacuolization with occasional necrosis.
Dogs that had two- to eightfold increases in ALT and two- to threefold increases in OCT had
minimal to moderate vacuolization with occasional necrosis. Dogs with 8- to 11-fold increases
in ALT and 4- to 7-fold increases in OCT had moderate vacuolation with single cell necrosis,
and those with 18- to 34-fold increases in ALT and 20-fold increases in OCT had moderately
severe vacuolation with single cell necrosis. The female dogs given 32 mg/kg-day for 8 weeks
showed no change in serum enzyme levels and no histopathology of the liver. In this study, 797
mg/kg-day was a LOAEL based on reported hepatic effects in six male and six female dogs, and
32 mg/kg-day was a NOAEL based on no hepatic effects reported in three female dogs. Given
the wide dose spacing in this study, there is considerable uncertainty about the assigned value of
the NOAEL and LOAEL.
Bruckner et al, 1986
Groups of 15-16 adult male Sprague-Dawley rats were given doses of 0, 1, 10, or 33
mg/kg of analytical-grade carbon tetrachloride by gavage in corn oil 5 days/week for 12 weeks
(time-weighted average doses of 0, 0.71, 7.1, or 23.6 mg/kg-day). Body weight was measured
twice weekly. Blood samples were taken from five rats from each group at 2-week intervals (2,
4, 6, 8, 10, and 12 weeks, and 2 weeks post-treatment; each individual animal served as a blood
donor twice, at 6-week intervals). After 12 weeks, 7-9 animals from each group were sacrificed.
The remaining animals were maintained without carbon tetrachloride treatment for an additional
2 weeks and then sacrificed. Following sacrifice, a terminal blood sample was taken by cardiac
3 9 DRAFT - DO NOT CITE OR QUOTE
-------
puncture. The liver and kidneys were removed, weighed, and processed for histopathological
examination. Blood samples were used for determination of serum ALT, OCT, and sorbitol
dehydrogenase (SDH), all of which are indicators of liver injury, and BUN, an indicator of
kidney damage. At the end of the exposure period, substantial toxicity was evident in rats
exposed to 23.6 mg/kg-day. Body weight gain in this group was significantly reduced by about
6% after 30 days and 17% after 90 days. Liver toxicity in this group was manifested by
significantly elevated ALT (up to 34 times control levels), SDH (up to 50 times control levels),
and OCT (up to 8 times control levels) from week 2 through the end of exposure, significantly
increased liverbody weight ratio, and extensive occurrence of degenerative lesions. Observed
liver lesions included lipid vacuolization, nuclear and cellular polymorphism, bile duct
hyperplasia, and periportal fibrosis. Severe degenerative changes, such as Councilman-like
bodies (single-cell necrosis), deeply eosinophilic cytoplasm, and pyknotic nuclei, were
occasionally noted as well. No evidence of nephrotoxicity was observed. Only moderate effects
were seen in animals exposed to 7.1 mg/kg-day. Body weight gain was similar to controls, and
liver toxicity was shown only by a significant (two- to threefold) elevation of SDH during the
second half of the exposure period and the presence of mild centrilobular vacuolization in the
liver. During the 2-week recovery period, serum ALT and SDH levels returned towards control
levels in both mid- and high-dose rats. Hepatic lesions were still present in both groups, but
severity was reduced for lesions other than fibrosis and bile duct hyperplasia, the severity of
which did not change. No effects were observed in rats exposed to 0.71 mg/kg-day. This study
identified aNOAEL of 0.71 mg/kg-day and a LOAEL of 7.1 mg/kg-day for carbon tetrachloride-
induced liver toxicity.
Attis et al, 1990
Allis et al. (1990) conducted a study to investigate the ability of rats to recover from
toxicity induced by subchronic exposure to carbon tetrachloride. Groups of 48 60-day-old male
F344 rats were given 0, 20, or 40 mg/kg of carbon tetrachloride 5 days/week for 12 weeks
(average daily doses of 0, 14.3, or 28.6 mg/kg-day) by gavage in corn oil. Food consumption by
cage was measured throughout the study. Rats were weighed several times during the first week
and once a week thereafter. After 12 weeks, treatment with carbon tetrachloride was stopped.
Six animals from each group were sacrificed 1, 3, 8, and 15 days after exposure termination.
Upon sacrifice, a terminal blood sample was taken for determination of total bilirubin,
triglycerides, cholesterol, ALT, AST, ALP, and lactate dehydrogenase (LDH). The liver was
weighed, and samples were taken for light microscopic examination and determination of protein
and CYP450. The remaining 24 animals were used to determine liver uptake relative to the
40 DRAFT - DO NOT CITE OR QUOTE
-------
spleen for a sulfur colloid labeled with technetium-99m and for tritiated 2-deoxyglucose0. Rats
used for this purpose were maintained as long as 22 days postexposure. The only toxicity
endpoint measured in these "remaining" animals was liver weight. Both doses of carbon
tetrachloride were hepatotoxic, although the high dose produced significantly greater toxicity
than the low dose. One day after the end of exposure, significant dose-related changes were
found for liverbody weight ratio and serum ALT, AST, and LDH (all increased) and liver
CYP450 (decreased) in both dose groups. In addition, serum ALP and cholesterol were
increased in the high-dose group. In the low-dose group, histopathological examination of the
liver revealed cirrhosis in 2/6 and vacuolar degeneration and hepatocellular necrosis in 6/6 rats;
in the high-dose group, histopathological examination revealed cirrhosis (as well as degeneration
and necrosis) in 6/6 rats. Serum enzyme levels and CYP450 returned to control levels within 8
days of the end of exposure. Severity of microscopic lesions declined during the postexposure
period, but cirrhosis persisted in the high-dose group through the end of the experiment.
Relative liver weight decreased during the postexposure period, but did not reach control levels
in the high-dose group even after 22 days. Neither of the radiolabeled tracer techniques detected
a decreased functional capacity in cirrhotic livers, a finding that could not be explained by the
investigators. The low dose of 14.3 mg/kg-day was a LOAEL for hepatic toxicity in this study.
Koporec etal, 1995
Koporec et al. (1995) evaluated the effect of different dosing vehicles on the subchronic
oral toxicity of carbon tetrachloride in the rat. Groups of 11 male Sprague-Dawley rats were
treated with carbon tetrachloride by gavage at doses of 0, 25, or 100 mg/kg, 5 days/week for 13
weeks (average daily doses of 0, 17.8, or 71.4 mg/kg-day). The compound was administered in
corn oil or as an aqueous emulsion in 1% Emulphor. An untreated control group was followed in
addition to vehicle controls. Blood samples were taken from 4 to 5 rats/group after weeks 4 and
8 for analysis of SDH and ALT. All surviving rats were sacrificed at the end of exposure at
which time additional blood samples were collected and the liver was weighed and sampled for
histopathology and biochemical studies (triglyceride, microsomal protein, CYP450, and glucose-
6-phosphatase [G6Pase]).
Mortality was found in all treated groups. The number of deaths was higher for rats
treated with the Emulphor vehicle than with corn oil and increased with dose for both vehicles.
Mortality was about 75 and 25% in the high- and low-dose Emulphor groups and about 45 and
10% in the high- and low-dose corn oil groups. No deaths occurred in any of the control groups.
Body weight decreased in a dose-related fashion throughout the study to a comparable extent in
rats treated with either vehicle. Terminal body weights were reduced about 25% (statistically
0 Relative efficiency of liver uptake of the labeled sulfur colloid is a diagnostic test for human cirrhosis and
considered by investigators to be an indirect measure of hepatocyte function. Hepatic uptake of 2-deoxyglucose is
an indicator of hepatic glucose utilization.
41 DRAFT - DO NOT CITE OR QUOTE
-------
significant) in the high-dose groups (both vehicles) and about 6% in the low-dose groups (both
vehicles). Serum chemistry analyses showed statistically significant dose-related increases in
SDH and ALT at both dose levels after 4-13 weeks of treatment with either vehicle. Increases in
SDH were as high as 10-fold in the low-dose groups and 100-fold in the high-dose groups, while
increases in ALT were about twofold in the low-dose groups and 25-fold in the high-dose
groups. The results were similar for rats treated in either vehicle. Liver microsomal enzyme
activities (CYP450 and G6Pase) were significantly reduced only in the high-dose groups, and,
again, the magnitudes of the effects were similar for rats treated in either vehicle. Absolute and
relative liver weights were slightly but significantly increased in the high-dose rats treated in
Emulphor but not in other groups. The researchers noted that the livers were perfused with
saline to facilitate collection of biochemical data and suggested that this procedure may have
influenced the liver weight results. Liver histopathology findings were similar in rats treated in
either vehicle. In the low-dose groups, lesions, seen in almost all animals, consisted primarily of
minimal-to-slight vacuolation and minimal fibrosis. In the high-dose groups, vacuolation and
fibrosis were moderate-to-moderately severe (all animals), and other lesions were also seen in all
animals, including minimal-to-slight necrosis and moderate-to-moderately severe cytomegaly,
nodular hyperplasia, oval-cell hyperplasia, and bile-duct hyperplasia. The low dose of 17.8
mg/kg-day, which produced hepatic effects in rats with either the corn oil or the Emulphor
vehicle, was considered a frank effect level (FEL) by the U.S. EPA because of the increased
mortality at this dose level. Vehicle did not influence hepatotoxicity in this study, but lethality
appeared to be enhanced by dosing in Emulphor.
Condieetal, 1986
A study comparing the effects of two different gavage vehicles on subchronic toxicity of
carbon tetrachloride was also performed in mice. CD-I mice (12/sex/group) were treated with 0,
1.2, 12, or 120 mg/kg of carbon tetrachloride (98.2% pure) by gavage in either corn oil or 1%
Tween-60 aqueous emulsion 5 days/week for 12 weeks (average daily doses of 0, 0.86, 8.6, or 86
mg/kg-day) (Condie et al., 1986). The mice were caged in groups of six and provided with food
and water ad libitum. Food and water consumption and body weights were measured twice
weekly. At terminal sacrifice, blood samples were drawn for determination of serum ALT, AST,
and LDH. The livers were examined grossly, weighed, and processed for histopathological
examination. Fifteen deaths occurred during the study, half of which were attributed to gavage
error; the others were not dose related. These early deaths were scattered over dose groups and
did not appear to influence the study outcome. Body weight was not affected by treatment in any
exposure group. Hepatotoxicity was indicated in the high-dose group (86 mg/kg-day) by
significantly elevated liver weight and liverbody weight ratio; significantly elevated ALT (77-
89 times control levels in corn oil and 10-19 times control levels in Tween-60), AST (14-15
times control levels in corn oil and 3-4 times control levels in Tween-60), and LDH (12-15
42 DRAFT - DO NOT CITE OR QUOTE
-------
times control levels in corn oil and 2-3 times control levels in Tween-60); and increased
incidence and severity of hepatic lesions, such as hepatocellular vacuolization, inflammation,
hepatocytomegaly, necrosis, and portal bridging fibrosis. At this dose, the only difference
between gavage vehicles was a greater incidence and severity of necrosis in mice given carbon
tetrachloride in corn oil. The difference between vehicles was more apparent at the middle dose
of 8.6 mg/kg-day. This dose produced significantly elevated ALT and mild-to-moderate liver
lesions in mice gavaged with corn oil but was identified as a NOAEL for mice gavaged with
Tween-60. The low dose of 0.86 mg/kg-day was identified as the NOAEL for mice gavaged
with corn oil. In general, both sexes responded similarly, with severity of histopathologic
changes in males slightly greater than females.
Hayes etal, 1986
Another study in mice was conducted at higher doses. CD-I mice (20/sex/group) were
gavaged daily with 0, 12, 120, 540, or 1,200 mg/kg-day of carbon tetrachloride (high
performance liquid chromatography grade, purity >99%) in corn oil for 90 days (Hayes et al.,
1986). An untreated control group of 20 male and 20 female mice was maintained as well. The
mice were observed for clinical signs of toxicity twice daily and weighed weekly. At
termination of exposure, the mice were sacrificed, blood was collected by cardiac puncture, and
gross necropsy was performed. Organ weights were determined for brain, liver, spleen, lungs,
thymus, kidneys, and testes, and samples were taken from the liver and kidney for
histopathological examination. The blood samples were used for comprehensive hematological
and clinical chemistry analyses. Urinalysis was also performed, although collection of urine was
not described. Determination of effect was made by comparing test groups to the vehicle
controls. Untreated controls were also compared with the vehicle controls. Observed effects
were reported in mice of both sexes at all dose levels and were generally dose-related. These
effects included increases in serum LDH, ALT, AST, ALP, and 5'-nucleotidase and a decrease in
serum glucose. Absolute and relative liver, spleen, and thymus weights were increased. A
variety of treatment-related lesions were observed in the liver, including fatty change,
hepatocytomegaly, karyomegaly, bile duct hyperplasia, necrosis, and chronic hepatitis. No
treatment-related lesions were observed in the kidney. No changes were found in urinalysis or
hematology parameters. It should be noted that, compared with untreated controls, vehicle
controls had significantly elevated serum LDH and ALT, altered organ weights, and increased
incidence of liver lesions (e.g., necrosis in 5/19 in vehicle controls versus 0/20 in untreated
controls and 20/20 in the 12 mg/kg-day group). This study failed to identify a NOAEL; the low
dose of 12 mg/kg-day was a LOAEL for hepatic effects.
43 DRAFT - DO NOT CITE OR QUOTE
-------
4.2.1.2. Chronic Toxicity and Carcinogenicity
4.2.1.2.1. Early National Cancer Institute (NCI) studies
Edwards, 1941
Researchers at the NCI performed a series of early experiments on the turnorigenicity of
orally ingested carbon tetrachloride in mice. In the first of these experiments, groups of 143
male strain C3H mice (2-3.5 months old) were treated with 0.1 mL of a 40% solution of carbon
tetrachloride in olive oil (0.04 mL or 64 mg of carbon tetrachloride) by gavage two or three
times/week for a total of 23-58 doses per mouse over a period of 8-16 weeks (Edwards, 1941).
[Because body weights were not provided, doses in mg/kg-day could not be estimated.] This
dose produced parenchymal necrosis of the liver, but no renal damage and was not lethal with
repeated administration. Necropsies performed 2-147 days after the last feeding, when the
animals were between 6 and 10 months of age, found hepatomas in 126/143 mice (88%).
Tumors were typically multiple and were similar in appearance to spontaneous hepatoma. No
metastases were found. As in spontaneous hepatoma, the tumor cells were morphologically
similar to hepatic parenchymal cells. An olive oil control group consisted of 23 male C3H mice
given 39-50 gavage doses of 0.1 mL of olive oil (two or three per week) and autopsied between
9 and 11 months of age. Only 1 of the 23 mice in this group (4%) had a hepatoma. In untreated
male C3H mice from the same stock, autopsies performed on 17 animals at 8.5-9 months of age
found no hepatic tumors, while the incidence was 10% in animals autopsied at 11 months of age
and 26% in 341 animals autopsied at 11-19 months of age.
Edwards and Dalton, 1942; Edwards et al, 1942; Edwards, 1941
Similar experiments performed by the same researchers in other strains of mice with
lower spontaneous incidence of hepatoma than C3H mice (strains A, C, Y, and L) produced
similar results (Edwards and Dalton, 1942; Edwards et al., 1942; Edwards, 1941). A lower, but
still hepatotoxic (based on histopathologically observed cirrhosis), dose was administered in one
experiment. A group of 58 strain A female mice 2.5 months of age was treated with 0.1 mL of
5% carbon tetrachloride in olive oil (0.005 mL or 8 mg of carbon tetrachloride) 3 times weekly
for 25-29 doses over a 2-month period (Edwards and Dalton, 1942). [Because body weights
were not provided, doses in mg/kg-day could not be estimated.] The mice were autopsied from 2
days to 4.5 months after the last dosing. The incidence of hepatoma was 71%. The tumors were
morphologically similar to those seen in mice treated with the higher dose. In a related
experiment by the same investigators, doses ranging from 0.005 mL (8 mg) to 0.04 mL (64 mg)
did not produce any hepatomas in 2-month-old mice treated only 1-3 times and autopsied 2-12
months later. The livers of mice in this latter experiment showed complete regeneration, with
only limited evidence of the earlier damage caused by dosing. These studies, and a subsequent
one designed specifically to investigate the possibility of a sex-related difference in susceptibility
to carbon tetrachloride tumorigenicity in C3H mice (Andervont, 1958), found no evidence of any
44 DRAFT - DO NOT CITE OR QUOTE
-------
such difference between the sexes.
Eschenbrenner and Miller, 1946
A study with multiple dose levels was conducted by Eschenbrenner and Miller (1946) in
order to investigate the relationship between necrotic damage and regenerative processes in the
liver and induction of hepatoma. Strain A mice (five/sex/group) were treated by gavage with 0,
0.125, 0.25, 0.5, or 1% of carbon tetrachloride in olive oil, receiving either 30 doses of 0.02
mL/g BW at 4-day intervals or 120 doses of 0.005 mL/g BW daily. Doses of carbon
tetrachloride, then, were 0, 10, 20, 40, or 80 mg/kg-day daily or 0, 40, 80, or 160 mg/kg-day
every 4 days for 120 days. The mice were 3 months old at the start of treatment and 7 months
old at the end of treatment. Mice were maintained for 1 month without treatment. One
additional dose was given 24 hours before sacrifice (at 8 months of age). Mice were examined
for presence of hepatomas and necrotic lesions in the liver. No necrosis or hepatomas were
found in control animals. No necrosis was observed in mice treated with either 0.005 or 0.02
mL/g of 0.125% solution (i.e., 120 doses of 10 mg/kg-day or 30 doses of 40 mg/kg-day).
Although no hepatomas were found by gross examination, two mice in the group that received
30 intermittent 40 mg/kg-day doses were found to have small tumors (hepatomas) by
microscopic examination. Necrosis was produced only with 30 intermittent doses of 80 and 160
mg/kg-day. Hepatomas were produced with 30 intermittent doses of 80 and 160 mg/kg-day as
well as 120 continuous doses of 20, 40, or 80 mg/kg-day. The investigators observed, based on
results of separate experiments involving 1 or 2 doses, that all dose levels under both dosing
regimens (except 120 daily doses of 10 mg/kg-day) were expected to have produced initial liver
necrosis, although it was not observed at terminal sacrifice.
Delia Porta etal, 1961
An oral cancer bioassay for carbon tetrachloride in hamsters was also conducted. Delia
Porta et al. (1961) treated Syrian golden hamsters (10/sex) with carbon tetrachloride by gavage
weekly for 30 weeks. For the first 7 weeks, 0.25 mL of 5% carbon tetrachloride in corn oil (12.5
uL or 20 mg of carbon tetrachloride) was administered; this dose was halved for the remainder of
the exposure period. [Because body weight was not provided, doses in mg/kg-day could not be
estimated.] Animals were observed for an additional 25 weeks prior to sacrifice. Four females
and five males died during the treatment period, and three more females died during the
observation period. The remaining three females and five males were sacrificed at the end of the
55 week. Cirrhotic changes in the liver were seen in the animals that died during treatment and
to a lesser extent in the other animals as well. Of the 10 hamsters (5 males and 5 females) that
died or were killed between weeks 43 and 55, all had liver-cell carcinomas, typically multiple,
and one had metastasized to the mesenteric and cervical lymph nodes. No liver-cell tumors were
observed in an untreated group of 109 male and 145 female hamsters from the same breeder or in
45 DRAFT - DO NOT CITE OR QUOTE
-------
another group of 50 males and 30 females given 0.5 mL of corn oil by gavage twice weekly for
45 weeks.
4.2.1.2.2. NCIbioassay. NCI (1977, 1976a, b; Weisburger, 1977) used carbon tetrachloride as a
positive control in cancer assays for chloroform, trichloroethylene, and 1,1,1-trichloroethane in
rats and mice, and findings are reported in appendices to the bioassay reports for these other
chlorinated solvents. Neoplastic and nonneoplastic incidence data were also available through
the National Toxicology Program (NTP) database search application (NTP, 2007).d Groups of
Osborne-Mendel rats (50/sex/group) were administered carbon tetrachloride by corn oil gavage
at time-weighted average doses of 47 or 94 mg/kg for males and 80 or 159 mg/kg for females, 5
days/week for 78 weeks. Rats were maintained without treatment for an additional 32 weeks.
Only 7/50 (14%) males and 14/50 (28%) females in the high-dose group and 14/50 (28%) males
and 26/50 (52%) females in the low-dose group survived to 110 weeks. In the pooled negative
control group, 26/100 (26%) males and 51/100 (51%) females survived to the end of the study.
Both doses of carbon tetrachloride resulted in marked heptotoxicity (including fatty changes),
with resultant fibrosis, cirrhosis, bile duct proliferation, and regeneration. Based on the NTP
database of neoplastic and nonneoplastic incidences (NTP, 2007), all other major organ systems
were examined for histopathological changes; however, no treatment-related effects other than
those in the liver were reported. The incidence of liver tumors was low in all groups.
Hepatocellular carcinoma was recorded in 1/99 pooled control, 2/49 low-dose, and 2/50 high-
dose males and in 0/98 pooled control, 4/49 low-dose, and 2/49 high-dose females. Neoplastic
nodules in the liver were seen in 0/99 pooled controls and 2/50 low-dose and 1/50 high-dose
males, and in 2/98 pooled controls and 2/49 low-dose and 3/49 high-dose females. The increase
in carcinomas was statistically significant in low-dose females in relation to pooled controls.
High early mortality, particularly in the high-dose group, may have affected the power of this
study to detect a carcinogenic effect.
In the same study, groups of male and female B6C3F1 mice received gavage doses of
1,250 or 2,500 mg/kg, 5 days/week for 78 weeks, and were maintained without treatment for 32
additional weeks. Mortality was markedly increased in treated mice. Survival was about 20% in
low-dose groups and <10% in high-dose groups at 78 weeks (versus 70% in control males and
90% in control females), and only one treated mouse survived to study termination at 92 weeks
(versus 50% in control males and 80% in control females). Liver toxicity (cirrhosis, bile duct
proliferation, toxic hepatitis, and fatty liver) was reported in only a few treated mice. According
to the NTP database of neoplastic and nonneoplastic incidences (NTP, 2007), the only other
nonneoplastic lesions in mice that were increased in a dose-related fashion was chronic murine
d In a few instances, the tumor incidence values differed slightly between the NCI bioassay reports where carbon
tetrachloride was included as a positive control, the Weisburger (1977) review, and the NTP database. In those
instances, the incidence value included in the Toxicological Review was taken from the NTP database.
46 DRAFT - DO NOT CITE OR QUOTE
-------
pneumonia in the lungs. Almost all treated mice, even those that died early, had hepatocellular
carcinomas (49/49 low-dose males, 47/48 high-dose males, 40/41 low-dose females, and 43/45
high-dose females). In pooled controls, incidence was only 5/77 (6%) in males and 1/80 (1%) in
females. The incidence of adrenal adenoma and pheochromocytoma was also increased in male
mice (concurrent control: 0/18, low-dose: 28/49, high-dose: 27/48) and female mice (concurrent
control: 0/18, low-dose: 15/41, high-dose: 10/45) (NTP, 2007; Weisburger, 1977).
4.2.2. Inhalation Exposure
4.2.2.1. Subchronic Toxicity
Smyth et al, 1936
Smyth et al. (1936) exposed groups of 24 guinea pigs (strain not specified) and 24
Wistar-derived rats (mixed sexes of both species) to 50, 100, 200, or 400 ppm (315, 630, 1,260,
or 2,520 mg/m3) of carbon tetrachloride vapor (>99% pure), 8 hours/day, 5 days/week for up to
10.5 months. The guinea pigs in this study received a purely vegetarian diet, but, because the
authors felt that low calcium in this diet may have affected the toxicity results, additional groups
of 16 guinea pigs fed diets supplemented with calcium were tested at concentrations of 25 ppm
(157 mg/m3), as well as 50, 100, and 200 ppm. In addition to the rats and guinea pigs, groups of
four monkeys (species and sex not specified) were exposed to 50 or 200 ppm using the same
protocol. Use of controls was not described, although the study authors state that "appropriate
controls [were] reserved." All animals were weighed weekly. Blood counts (all species) and
urinalysis (guinea pigs and monkeys) were performed monthly. The fertility of rats and guinea
pigs, which were housed in mixed-sex groups and produced litters during the study, was
monitored. All animals that survived to scheduled sacrifice (including some animals that were
sacrificed only after recovery periods of varying durations) and most of those dying during the
study were examined for gross pathology. Tissue samples for histopathological examination
were taken from the liver, kidney, adrenal gland, spleen, heart, sciatic and optic nerves, and
ocular muscle. Serum chemistry analyses were performed on some animals as well. No
statistical tests were conducted.
Guinea pigs of all exposure groups, including those that received diets supplemented with
calcium, suffered substantial mortality (>25-80% among "uninfected" guinea pigs). Mortality in
controls was not reported. In contrast, mortality among "uninfected" rats was limited to two
animals exposed to 400 ppm. No monkeys died during the study. Body weight gain was
reported to be markedly reduced among survivors in all groups of guinea pigs, compared with
that in controls. Body weight gain was also reduced by about 30% among rats exposed to 400
ppm. Too few litters were born to guinea pigs during the study to determine if exposure had any
effect, but, in rats, fertility was reduced in the 200 and 400 ppm groups. In guinea pigs, fatty
changes in the liver were seen at all dose levels, and cirrhosis developed at >50 ppm. In rats,
fatty changes were seen at >50 ppm and cirrhosis was noted at > 100 ppm. In monkeys, mild
47 DRAFT - DO NOT CITE OR QUOTE
-------
fatty degeneration of the liver was found at both 50 and 200 ppm. Other pathological changes in
animals exposed to these concentrations included renal tubular degeneration, degeneration of the
adrenal glands (with necrosis in guinea pigs), and damage to the sciatic nerve. This study did not
include concentrations low enough to identify a NOAEL for any of the three species tested. For
guinea pigs, the low concentration of 25 ppm was a frank effect level that produced substantial
mortality. For rats and monkeys, the low concentration of 50 ppm was a LOAEL that produced
fatty changes in the liver. This study provides evidence of the progression of toxic liver effects
from fatty changes in the liver at lower doses to liver cirrhosis at higher doses. Because of the
age of the study, knowledge that bacterial and viral infections were a common problem at that
time, and the confounding that pregnancy (or lack of pregnancy) could have had on body
weights, the findings from this study must be interpreted with caution.
Adams et al, 1952
Adams et al. (1952) conducted studies in which Wistar-derived rats (15-25/sex), outbred
guinea pigs (5-9/sex), outbred rabbits (1-2/sex), and Rhesus monkeys (1-2 of either sex) were
exposed to carbon tetrachloride vapor (>99% pure), 7 hours/day, 5 days/week for 6 months at
concentrations of 5, 10, 25, 50, 100, 200, or 400 ppm (31, 63, 157, 315, 630, 1260, or 2520
mg/m3). Matched control groups, both unexposed and air exposed, were included in these
experiments. Animals were observed frequently for appearance and general behavior and were
weighed twice weekly. Selected animals were used for hematological analyses periodically
throughout the study. Moribund animals and those surviving to scheduled sacrifice were
necropsied. The lungs, heart, liver, kidneys, spleen, and testes were weighed, and sections from
these and 10 other tissues were prepared for histopathological examination. In many cases,
terminal blood samples were collected and used for serum chemistry analyses, and part of the
liver was frozen and used for lipid analyses.
In this study, the primary target of carbon tetrachloride in all species was the liver. In
guinea pigs, liver effects progressed from a slight, statistical increase in relative liver weight in
females, but not males, at 5 ppm (not considered adverse by itself) to include slight-to-moderate
fatty degeneration and increases in liver total lipid, neutral fat, and esterified cholesterol at 10
ppm, and cirrhosis at 25 ppm. Liver effects became progressively more severe at higher
concentrations. Growth retardation was first observed at 25 ppm and progressed to rapid loss of
weight at 200 ppm. In the kidney, slight tubular degeneration was first observed at 200 ppm and
increased kidney weight was noted at 400 ppm. Mortality was increased at >100 ppm. A similar
progression of effects was seen in rats, with no effects at 5 ppm, mild liver changes at 10 ppm,
cirrhosis at 50 ppm, and liver necrosis, kidney effects, testicular atrophy, growth depression, and
mortality at >200 ppm. In rabbits, 10 ppm was without effect, 25 ppm produced mild liver
changes, 50 ppm produced moderate liver changes, and 100 ppm produced growth depression.
Monkeys were the most resistant species tested, with evidence of adverse effects (mild liver
48 DRAFT - DO NOT CITE OR QUOTE
-------
lesions and increased liver lipid) only at 100 ppm, the highest concentration tested. This study
identified NOAEL and LOAEL values, respectively, of 5 and 10 ppm in rats and guinea pigs, 10
and 25 ppm in rabbits, and 50 and 100 ppm in monkeys, all based on hepatotoxic effects.
Prendergast et al, 1967
Prendergast et al. (1967) exposed groups of 15 Sprague-Dawley or Long-Evans rats, 15
Hartley guinea pigs, 3 New Zealand rabbits, 2 beagle dogs, and 3 squirrel monkeys (sex not
specified) to carbon tetrachloride vapor ("highest purity available") either by continuous
exposure to 1 or 10 ppm (6.1 or 61 mg/m3) for 90 days or intermittent exposure (8 hours/day, 5
days/week) to 82 ppm (515 mg/m3) for 6 weeks. The control group consisted of 304 rats, 314
guinea pigs, 48 rabbits, 34 dogs, and 57 monkeys. In order to generate the 1 ppm concentration,
the researchers found it necessary to dilute the carbon tetrachloride in 10 ppm of n-octane.
Therefore, a vehicle control group exposed to 10 ppm of n-octane was included in this study.
Animals were observed routinely for signs of toxicity and weighed monthly. Blood samples for
hematological analysis were taken at the end of the exposure period. Following sacrifice,
animals were necropsied and sections of the heart, lung, liver, spleen, and kidney were taken for
histopathological examination. Serum chemistry and liver lipid analyses were performed on
some animals. No statistical tests were conducted.
Intermittent exposure to 82 ppm resulted in the death of 3/15 guinea pigs and 1/3
monkeys. [This compares to mortality in the control groups of 7/304 (2.3%) rats, 2/314 (0.64%)
guinea pigs, 2/48 (4.2%) rabbits, 0/34 dogs, and 1/57 (1.7%) monkeys.] Body weight gain was
reduced in all species relative to the controls, and all species except rats actually lost weight
during the study. Mottled livers were seen in all species except dogs. Histopathological
examination of the liver revealed fatty changes that decreased in severity from guinea pigs to rats
to rabbits to dogs to monkeys. Liver lipid content of guinea pigs was increased about threefold
compared with controls. The only other effect noted was interstitial inflammation in the lungs of
all species. Continuous exposure to 10 ppm resulted in the deaths of 3/15 guinea pigs. Body
weight gain was depressed in all species relative to the controls, and monkeys appeared visibly
emaciated. Gross examination showed the presence of enlarged/discolored livers in all species
except dogs. Microscopic examination revealed fatty changes in the liver that were most
prominent in rats and guinea pigs but were present in the other species as well. Lung effects
were not reported in this group. Continuous exposure to 1 ppm produced no mortality or clinical
signs of toxicity. Weight gain relative to the controls was reduced in guinea pigs, rabbits, dogs,
and monkeys but not in rats. The only histopathological findings were nonspecific inflammatory
changes in the liver, kidney, heart, and lungs. No effects were noted in the n-octane control
group. The results of this study suggest a NOAEL of 1 ppm (6.1 mg/m3) and a LOAEL of 10
ppm (61 mg/m3) for rats, guinea pigs, rabbits, dogs, and monkeys based on hepatotoxicity.
Effects on growth were reported at both exposure levels, but the data are difficult to interpret, as
49 DRAFT - DO NOT CITE OR QUOTE
-------
only starting body weights and percent change are reported, the changes did not occur in a dose-
related manner in all species, and no statistical comparisons were performed. It is unclear
whether inflammatory changes observed in the lungs of some exposed animals occurred in
controls as well.
Nagano et al, 2007a (Japan Bioassay Research Center [JBRC], 1998)
Groups of F344/DuCrj rats (10/sex/group) were exposed (whole body) to 0, 10, 30, 90,
270, or 810 ppm (0, 63, 189, 566, 1,700, or 5,094 mg/m3) of carbon tetrachloride (99.8% pure)
vapor for 6 hours/day, 5 days/week for 13 weeks (Nagano et al., 2007a). (This study was
previously available as an unpublished study by the JBRC [1998]). Rats were observed once a
day for clinical signs, behavioral changes, and mortality and were weighed weekly. Urinalysis
(pH, protein, occult blood, glucose, ketone body, bilirubin, and urobilinogen) was performed at
the end of the dosing period. Blood for hematological (erythrocytes, hemoglobin, hematocrit,
platelets, and leukocyte differential) and serum chemistry analyses (AST, ALT, LDH, ALP, total
bilirubin, creatine phosphokinase [CPK], urea nitrogen, creatinine, total protein, albumin,
albumin/globulin ratio, glucose, total cholesterol, phospholipid, sodium, potassium, chloride,
calcium, and inorganic phosphorus) was taken during euthanization at the scheduled sacrifice
after overnight fasting. All organs and tissues were examined for gross lesions, and organ
weights were recorded for the thymus, adrenal gland, ovary, testis, heart, lung, kidney, spleen,
liver and brain. Tissues (not specified) were fixed for histopathological analysis; lesions were
presented for selected tissues (liver and kidney). Additionally, livers of control and 810-ppm
male rats were sectioned for examination of hepatic altered cell foci, a preneoplastic lesion, by
immunohistochemical staining with anti-glutathione S-transferase placental (GST-P) using an
avidin-biotin-peroxidase complex method.
No deaths occurred in any group. Body weight in the 810 ppm males was lower than in
controls throughout the study. At termination, the decrease was about 20% (p<0.01). Body
weight was consistently lower than controls in the 810 ppm females as well, but the difference at
termination was slight (4%) and not statistically significant. Statistically significant, dose-related
decreases in hemoglobin and hematocrit were observed at. 90 ppm in both males and females. At
810 ppm, red blood cell count was also significantly decreased in both sexes. Serum chemistry
changes included large, statistically significant and dose-related increases in ALT, AST, LDH,
ALP, and LAP (leucine aminopeptidase) in males at. 270 ppm and females at. 90 ppm. Total
bilirubin was significantly increased in male rats at 810 ppm and female rats at >270 ppm.
Serum levels of CPK were statistically increased in females at >30 ppm, but there was little
change as exposure level increased from 90 to 810 ppm. CPK levels in males were not
statistically different from those in controls. In the urine, protein levels were increased in males
at. 270 ppm and in females at. 90 ppm. Urinary pH was decreased and the presence of occult
blood was noted in males and females at 810 ppm. Relative liver weights were significantly
50 DRAFT - DO NOT CITE OR QUOTE
-------
increased in a dose-related fashion in male rats (>10 ppm) and female rats (>30 ppm).
Significant, dose-related increases in absolute and relative weights were also recorded for the
kidneys, spleen, heart, and lungs in both males and females, primarily at >90 ppm. Females at
810 ppm also had significant reductions in absolute and relative ovary weights. Males at 270 or
810 ppm had significantly reduced absolute testes weights, but relative weights were similar to
those in controls. Dose-related increases in the incidence and severity of histopathological
lesions of the liver were observed at 10 ppm in both sexes. At the low level of 10 ppm,
treatment-related lesions included slight fatty change, cytological alteration, and granulation.
Additional lesions at higher levels included ceroid deposits, fibrosis, pleomorphism, proliferation
of bile ducts, and cirrhosis. Altered cell foci were observed in male rats at >270 ppm and in
female rats at >90 ppm (based on H&E-stained sections). The altered cell foci in 810-ppm male
rats also stained positively with the anti-GST-P antibody. Renal lesions (localized
glomerulosclerosis) were seen in the 810 ppm males and females. The low concentration of 10
ppm was a LOAEL for hepatic effects in rats (increased liver weight and histopathology). A
NOAEL was not identified.
These researchers conducted a similar study in mice. Groups of Crj:BDFl mice
(10/sex/group) were exposed (whole body) to 0, 10, 30, 90, 270, or 810 ppm (0, 63, 189, 566,
1,700, or 5,094 mg/m3) of carbon tetrachloride (99.8% pure) vapor for 6 hours/day, 5 days/week
for 13 weeks. Endpoints monitored were the same as described above for the 13-week rat study.
No treatment-related deaths occurred. Body weights were lower than in controls for most of the
study in males at 30 ppm; at termination, the decreases in these groups ranged from 8 to 15% and
were statistically significant. Body weights in treated females were similar to those in controls
throughout the study. Hematology findings included slight, significant decreases in red blood
cell count and hemoglobin at 270 ppm and hematocrit at 810 ppm in females and in hemoglobin
at 810 ppm in males. Serum chemistry changes of note included significant increases in ALT
and LAP in males and females at. 90 ppm (and ALP in males at >30 ppm), slight significant
increases in total protein and/or albumin in males and females at. 270 ppm, and a significant
increase in AST in males at 810 ppm. Urinalysis revealed no treatment-related changes in males,
but a significant decrease in the pH of urine was noted in females at 810 ppm. Organ weight
changes in treated mice included significant increases in absolute and/or relative weights of the
liver, kidney, and spleen in males and females, primarily at >90 ppm and above. Organ weight
changes in males were confounded by body weight decreases in most treated male groups.
Histopathological changes in mice were found only in the liver. In both sexes, the hepatic
lesions exhibited dose-related increases in incidence and severity. The only effect at the low
level of 10 ppm was an increase in incidence of slight cytoplasmic globular and fatty change
(large droplets) in males. Additional liver lesions noted in the higher-exposure groups were:
nuclear enlargement with atypia and altered cell foci (>270 ppm) and collapse (presumably
resulting from the necrotic loss of hepatocytes) (>30 ppm). Altered cell foci included
51 DRAFT - DO NOT CITE OR QUOTE
-------
acidophilic, basophilic, clear cell and mixed cell foci. The lowest exposure level of 10 ppm is a
minimal LOAEL for hepatic effects (slight cytological alterations) in male mice.
Benson and Springer, 1999
Groups of F344 rats, B6C3F1 mice, and Syrian hamsters (10 males/species) were
exposed by inhalation to carbon tetrachloride vapor at concentrations of 0, 5, 20, or 100 ppm
(31.5, 126, or 630 mg/m3) for 6 hours/day, 5 days/week for 12 weeks (Benson and Springer,
1999; Nikula et al., 1998). An indicator of deoxyribonucleic acid (DNA) replication, 5-bromo-
2'-deoxyuridine (BrdU), was administered to animals of all species several days prior to
sacrifice. Additional satellite groups of 5-6 animals/species were sacrificed after 1 and 4 weeks.
At sacrifice, blood was collected for ALT and SDH determinations, and liver sections were
collected for histopathological examination (quantitative evaluation of necrosis in the hepatic
parenchyma) and BrdU detection. Serum levels of ALT and SDH were significantly increased in
mice at 320 ppm and in rats and hamsters at 100 ppm. The increases in mice and hamsters were
larger than those in rats. The actual magnitude of the changes could not be assessed from the
graphical presentation of the data. The volume percent of the hepatic parenchyma that was
necrotic also was significantly increased in mice at >20 ppm and in rats and hamsters at 100
ppm. No necrosis was seen in controls or 5 ppm animals of any species. After 12 weeks, the
volume percent of necrosis in the liver of the groups showing statistically significant increases
ranged from approximately 5 to 10% in all species. More precise measures of necrosis could not
be determined from the graphical presentation of the data. BrdU labeling indices were also
significantly increased in mice at >20 ppm and hamsters at 100 ppm, but were not increased in
rats at any concentration tested (except for a small nonsignificant increase at 100 ppm). In mice,
the percent of BrdU positive hepatocytes at 12 weeks was about 20% at 20 ppm and 60% at 100
ppm. In hamsters at 100 ppm, the percent of BrdU positive hepatocytes at 12 weeks was about
40%. In controls, the percent of BrdU positive hepatocytes at 12 weeks was approximately 2%.
These results show the occurrence of hepatocellular proliferation only at doses that also
produced necrotic damage. The study identified 5 ppm as a NOAEL and 20 ppm as a LOAEL
for hepatotoxicity in mice. Hamsters and rats were less sensitive than mice, with NOAEL values
of 20 ppm and LOAEL values of 100 ppm in these species.
4.2.2.2. Chronic Toxicity and Carcinogenicity
Nagano etal, 2007b (Japan Bioassay Research Center [JBRC], 1998)
Groups of F344/DuCrj rats (50/sex/group) were exposed (whole-body) to 0, 5, 25, or 125
ppm (0, 31.5, 157, or 786 mg/m3) of carbon tetrachloride (99.8% pure) vapor for 6 hours/day, 5
days/week for 104 weeks (Nagano et al., 2007b). (This study was previously available as an
unpublished study by the JBRC [1998].) Animals were observed daily for clinical signs,
52 DRAFT - DO NOT CITE OR QUOTE
-------
behavioral changes, and mortality. Body weights were measured once a week for the first 14
weeks and every 2 weeks thereafter. Urinalysis, hematology, and clinical chemistry tests were
conducted at study termination as described above for the 13-week rat study, except that GGT
was added to the list of serum enzymes monitored. All organs and tissues were examined for
gross lesions, and organ weights were recorded for the adrenal gland, testis, ovary, heart, lung,
kidney, spleen, liver, and brain. All major tissues were examined for histopathologic changes.
Survival curves are for the male and female rat are shown in Figure 4-1. Survival was
high in all groups through week 64. After week 64, survival declined precipitously in the
125-ppm males and females. Only three males and one female from this group survived to 104
weeks. Liver tumors and chronic progressive nephropathy were the main causes of death.
Survival in the other treated groups (19-28/50 in males and 39-43/50 in females) was similar to
controls and adequate for evaluation of late developing tumors. Body weights were reduced
throughout most of the study in 125-ppm males (reduced 22% at termination) and after week 84
in 25-ppm males (reduced approximately 10% at termination). In females, body weight was
reduced during the second year of the study in both the 125-ppm (reduced 45% at termination)
and 25-ppm (reduced approximately 10% at termination) groups. The body weight decreases in
the 25-ppm males and females at termination were statistically significant. Low survival of rats
in the 125-ppm group limited statistical comparison of this group with controls.
53 DRAFT - DO NOT CITE OR QUOTE
-------
Male rat
— . • •
-
_
i
4MB
^_
•OO4X
•OMM
•o<*m
MWMW
,00?
**"
lOi.1.1
..l»iK
~~
*»
^ -
"V
7 j
T 1
^MOoa^
vi^A>
\ ' 3
n (
\
SliEVIViL HHISSL WIKBEKS
Female rat
INIJUL ' fl*T RH4
rawrnre : 11
SUK?ITAL 4KIHAL WMBEKS
Figure 4-1. Survival curves for male and female rats
Source: JBRC (1998)
54
DRAFT - DO NOT CITE OR QUOTE
-------
Hematology analyses showed trends for decreased red blood cell count, hemoglobin, and
hematocrit in males and females at 25 and 125 ppm, although only the decreases for hemoglobin
and hematocrit in 25-ppm females were statistically significant (there was no statistical
evaluation for the 125 ppm group). Serum chemistry changes included statistically significant
increases in AST (males), ALT (males and females), LDH (females), and GPT (females) at 25
ppm; the increases over control in individual serum chemistry parameters at 25 ppm ranged from
1.2- to twofold. There were also significant increases in BUN in both males and females at 25
ppm (25-63% over controls). At 125 ppm, BUN, creatinine, and inorganic phosphate were
increased by two- to threefold over the control (but were untestable statistically because of the
small number the surviving animals at 125 ppm). Consistent with the subchronic rat study, there
was a significant increase in CPK in 25-ppm females but not males. An increase was reported in
the number of male and female rats with high levels of proteinuria in the 5 and 25 ppm groups
(too few data to test in the 125 ppm group) (Table 4-2).
Table 4-2. Urinalysis results in rats after 2-year exposure to carbon
tetrachloride
Concentration
(ppm)b
Protein content of urine"
+
2+
3+
4+
Male
0
5C
25C
125
0/22 (0%)
0/31 (0%)
0/19 (0%)
0/3 (0%)
2/22 (9%)
2/31 (6%)
1/19 (5%)
0/3 (0%)
20/22 (91%)
5/31 (16%)
3/19 (16%)
3/3 (100%)
0/22 (0%)
24/31 (77%)
15/19 (79%)
0/3 (0%)
Female
0
5C
25C
125
1/39 (3%)
0/43 (0%)
0/40 (0%)
0/1 (0%)
2/39 (5%)
2/43 (5%)
0/40 (0%)
0/1 (0%)
35/39 (90%)
15/43 (35%)
3/40 (8%)
1/1 (100%)
1/39 (3%)
26/43 (60%)
37/40 (92%)
0/1 (0%)
aUrine protein concentrations were measured with a semi-quantitative dipstick test. Equivalent concentrations are:
+: 30 mg/dL; 2+: 100 mg/dL; 3+: 300 mg/dL; 4+: 1000 mg/dL (letter dated March 8, 2004, from Kasuke Nagano,
JBRC, to Mary Manibusan, U.S. EPA).
b The exposure concentrations adjusted to continuous exposure (i.e., multiplied by 5/7 x 6/24) = 0.9, 4.5, and 22.3
ppm.
0 The study report indicated that urine protein results in male and female rats in the 5- and 25-ppm groups were
statistically elevated (p < 0.01) based on a %2 test. Whether the statistical test represented a trend test or pairwise
comparison of the graded responses was unclear from the study report.
Source: JBRC (1998).
55
DRAFT - DO NOT CITE OR QUOTE
-------
Organ weight changes were generally unremarkable and limited to the 25 and 125 ppm
groups, where they were confounded by body weight decreases in both males and females. Clear
increases in the incidence and severity of nonneoplastic liver lesions (fatty change, fibrosis,
cirrhosis) were seen at 25 and 125 ppm in both males and females (Table 4-3). Liver lesions
(e.g., fatty liver, granulation) in the 5-ppm group were of similar type, incidence, and severity as
controls. In the kidney, there was a dose-related increase in the severity of chronic nephropathy
(progressive glomerulonephrosise) at 25 and 125 ppm in both males and females (Table 4-3).
Nephropathy was characterized as severe in most members of the 125-ppm group. Other dose-
related histopathological changes were increased severity of eosinophilic change (eosinophilic
globules in cytoplasm) in the nasal cavity at >25 ppm in males and >5 ppm in females and
increased incidence and severity of granulation in the lymph nodes at 125 ppm in both sexes
(Table 4-3).
e Chronic nephropathy (progressive glomerulonephrosis) is another term for the progressive renal disease in aging
rats more recently referred to as chronic progressive nephropathy (CPN) (Peter et al, 1986).
56 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-3. Incidence of selected nonneoplastic lesions in F344 rats exposed
to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)"
Lesion
Male
0 ppm
5 ppm
25 ppm
125 ppm
Female
0 ppm
5 ppm
25 ppm
125 ppm
Liver
Fatty change
+
2+
3+
3/50
1/50
7/50
30/50
9/50
27/50
22/50
5/50
1/50
3/50
4/50
18/50
27/50
4/50
17/50
29/50
Fibrosis
+
2+
43/50
2/50
34/50
11/50
Cirrhosis
+
2+
1/50
14/50
26/50
1/50
1/50
23/50
27/50
Kidney
Chronic nephropathy
+
2+
3+
16/50
26/50
7/50
8/50
32/50
9/50
9/50
23/50
18/50
8/50b
9/50b
33/50b
31/50
13/50
37/50
7/50
1/50
19/50
25/50
5/50
5/50
7/50
38/50
Nasal cavity
Eosinophilic change
+
2+
43/50
47/50
25/50
25/50
13/50
34/50
39/50
33/50
16/50
25/50
25/50
4/50
46/50
Lymph nodes
Granulation
+
2+
4/50
9/50
1/50
11/50
1/50
6/50
27/50
3/50
5/50
11/50
2/50
12/50
28/50
a A blank cell indicates that the incidence of the histopathologic finding at that severity level was zero.
The exposure concentrations were adjusted to continuous exposure (i.e., multiplied by 5/7 x 6/24) = 0.9, 4.5, and
22.3 ppm.
b The published paper of the JBRC bioassay shows an incidence (all scores combined) of 49/50 125-ppm male rats.
The study report shows a total incidence of 50/50.
Source: Nagano et al. (2007b); JBRC (1998).
The low exposure level of 5 ppm was associated with an increase in the severity of
proteinuria in male and female rats at this concentration; however, there was no effect on the
incidence of proteinuria at any exposure level. Histopathological examination revealed clear
57
DRAFT - DO NOT CITE OR QUOTE
-------
evidence of treatment-related glomerular damage (increased severity of glomerulonephrosis) in
male and female rats exposed to 25 or 125 ppm. Increases in BUN (at >25 ppm) and serum
creatinine and inorganic phosphorus (primarily at 125 ppm) show impairment of glomerular
function (i.e., decrease in glomerular filtration rate) at the same concentrations as the observed
lesions. The increased proteinuria at 5 and 25 ppm could be related to the glomerular changes
indicated by histopathology and serum chemistry results at 25 and 125 ppm. For reasons
discussed more fully in Section 4.6.2, interpretation of the observed proteinuria in the F344 rat, a
strain with a high spontaneous incidence of renal lesions, is problematic. Therefore, 5 ppm was
considered a NOAEL and 25 ppm a LOAEL for effects on the liver and kidney.
Tumor incidence data for rats are presented in Table 4-4. The incidence of hepatocellular
adenomas and carcinomas was statistically significantly increased in male and female rats at 125
ppm. The incidence of hepatocellular carcinomas in female 25-ppm rats (6%) was not
statistically elevated compared with the concurrent control, but did exceed the historical control
range for female rats from JBRC (0-2%). The increase in liver carcinoma over historical control
(2/1,797) was statistically significant (based on Fisher's exact test; two-tailed p-value = 0.0002).
No other tumors occurred with an increased incidence in treated rats. Incidences of hepatic
altered cell foci (preneoplastic lesions of the liver), including clear, acidophilic, basophilic, and
mixed cell foci, were significantly increased in the 25-ppm female rats; in males, only the
incidence of basophilic cell foci was increased at 125 ppm.
5 8 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-4. Incidence of liver tumors in F344 rats exposed to carbon
tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)a
Tumor
Hepatocellular
adenoma
Hepatocellular
carcinoma
Hepatocellular
adenoma or
carcinoma
Male
0 ppm
0/50b
l/50b
l/50b
5 ppm
1/50
0/50
1/50
25 ppm
1/50
0/50
1/50
125 ppm
21/50C
32/50c
40/50C
Female
0 ppm
0/50b
0/50b
0/50b
5 ppm
0/50
0/50
0/50
25 ppm
0/50
3/50d
3/50d
125 ppm
40/50C
15/50C
44/50c
a The exposure concentrations adjusted to continuous exposure (i.e., multiplied by 5/7 x 6/24) = 0.9, 4.5, and 22.3
ppm.
Statistically significant trend for increased tumor incidence by Peto's test (p<0.01).
0 Tumor incidence significantly elevated compared with that in controls by Fisher's exact test (p<0.01).
d Statistically significant (p < 0.001 by Fisher's exact test) in comparison to the historical control incidence
(2/1797).
Note: The historical control incidence of liver tumors in F344/DuCrj rats in JBRC studies was 1.7% (0-8%) in
males and 1.2% (0-6%) in females for hepatocellular adenoma and 0.3% (0-2%) in males and 0.1% (0-2%) in
females for hepatocellular carcinoma (based on data from 36 to 39 carcinogenicity studies carried out by JBRC;
email dated April 5, 2007, from Kasuke Nagano, JBRC, to Susan Rieth, U.S. EPA).
Sources: Nagano et al. (2007b); JBRC (1998).
These researchers also conducted a 2-year study using Crj :BDF1 mice. Groups of
Crj:BDFl mice (50/sex/group) were whole-body exposed to 0, 5, 25, or 125 ppm (0, 31.5, 157,
or 786 mg/m3) of carbon tetrachloride (99% pure) vapor for 6 hours/day, 5 days/week for 104
weeks. Endpoints monitored were the same as described above for the 2-year rat study. Survival
was high until week 64 of the study in all groups (see survival curves in Figure 4-2). Survival
decreased rapidly in 125-ppm males and females, starting at week 64, and in 25-ppm males and
females, starting at week 84. The decreases in survival were statistically significant in both
sexes at both concentrations. At 104 weeks, only one male and one female survived in the 125
ppm group and 25 males and 10 females in the 25-ppm group (versus 35 males and 26 females in
the control group). Investigators reported that liver tumors were the main cause of death at 125
ppm. At 25 ppm, deaths prior to study termination were also largely attributable to the presence
of tumors (with liver adenomas or carcinomas present in 33/39 female mice and 22/23 male mice
that died or were sacrificed prior to study termination). Body weights were markedly depressed
throughout the study in 25- and 125-ppm males and females (22-39% reduction at termination).
59
DRAFT - DO NOT CITE OR QUOTE
-------
Male
mouse
iHlUL i MXH BV1
IEK1T TlPf i II
SURVIVAL AIIIMAL N'JMBKKS
Female
mouse
^•O-CKtyOW .
4000040000 j &Ae
or~
4
t
STUB! H3 : MM
UlJUt i IKXIEE IS
JffiKJtT1 T1PE a AJ
SURVIYAL AHIHA!, H1MBE3S
r> ftt
ts pit
125 PfH
Figure 4-2. Survival curves for male and female mice
Source: JBRC(1998)
60
DRAFT - DO NOT CITE OR QUOTE
-------
The survival of only one mouse of each sex at 125 ppm prevented statistical comparisons
involving this group. Statistically significant increases in red blood cell count, hemoglobin, and
hematocrit were found in 25-ppm females. Values for these variables were also higher than in
controls (but not statistically increased) in the 25-ppm males and in the 125-ppm male and
female. This is in contrast with the significant decreases in these variables seen in the
subchronic mouse study and the rat studies.
Serum chemistry changes of interest were large, statistically significant increases in ALT,
AST, LDH, ALP, protein, total bilirubin, and BUN in males and females at 25 ppm (increases
over control ranged from 1.3- to 18-fold) and, for most of these variables, still larger increases in
the 125-ppm male and female (based on one surviving mouse/sex at terminal sacrifice).
Statistically significant decreases in ALT, AST, LDH, and CPK in 5-ppm males were not
considered to be biologically significant by the researchers (letter dated March 8, 2004, from
Kasuke Nagano, JBRC, to Mary Manibusan, U.S. EPA). The decreases were inconsistent with
the large increases seen at higher doses in males or the results in females and appeared to reflect
unusually high serum levels of these enzymes in male controls rather than reduced levels in the
5-ppm males. Levels of these enzymes in control males exceeded historical control values for
male Crj:BDFl mice in 2-year studies from the same laboratory by 1.5- to 2.5-fold; this is in
contrast to the results in females, where control values for all of these variables were within 10%
of historical control values (historical control data provided in a letter dated March 9, 2004, from
Kasuke Nagano, JBRC, to Mary Manibusan, U.S. EPA). Urinary pH was significantly decreased
in males and females at 25 ppm. The only organ weight changes of note were large significant
increases in absolute (~2.5-fold) and relative (-three- to fourfold) liver weight in 25 ppm males
and females. Liver weight data in the surviving 125 ppm male and female were consistent with
these results as well. Treatment-related nonneoplastic lesions occurred in the 25 and 125 ppm
males and females; these included increased incidence and/or severity of degeneration, cyst
formation, and deposit of ceroid in the liver, protein casts in the kidney, and extra medullary
hematopoiesis in the spleen (Table 4-5). The 25 ppm concentration was a LOAEL in this study
for effects on the liver (increased weight, serum chemistry changes indicative of damage, and
lesions), kidney (serum chemistry changes and lesions), and spleen (lesions); decreased growth;
and reduced survival. The 5 ppm level was a NOAEL.
61 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-5. Incidence of selected nonneoplastic lesions in BDF1 mice
exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5
days/week)"
Lesion
Male
0 ppm
5 ppm
25 ppm
125 ppm
Female
0 ppm
5 ppm
25 ppm
125 ppm
Liver
Degeneration
+
2+
3+
1/50
4/50
3/50
1/50
7/50
2/50
1/50
4/50
9/50
6/50
6/50
Cyst formation
+
2+
1/50
3/50
10/50
1/50
5/50
3/50
3/50
1/50
2/49
10/50
2/50
3/50
3/50
Deposition of ceroid
+
2+
Bile duct
proliferation
Centrilobular
hydropic
change
2/50
0/50
1/50
1/50
0/50
0/50
28/50
8/50
19/50
8/50
22/50
14/50
22/50
9/50
0/50
1/50
0/49
0/49
22/50
6/50
5/50
13/50
22/50
13/50
9/50
12/50
Kidney
Protein casts
+
2+
1/50
1/50
5/50
6/50
1/50
2/50
9/50
3/50
Spleen
Extramedullary hematopoiesis
+
2+
3+
15/50
12/50
1/50
15/50
8/50
2/50
14/50
25/50
5/50
5/50
26/50
12/50
8/50
7/50
3/50
11/49
4/49
5/49
11/50
18/50
7/50
4/50
30/50
9/50
a A blank cell indicates that the incidence of the histopathologic finding at that severity level was zero.
The exposure concentrations adjusted to continuous exposure (i.e., multiplied by 5/7 x 6/24) = 0.9, 4.5, and 22.3
ppm.
Sources: Nagano et al. (2007b); JBRC (1998).
Tumor incidence data in mice are presented in Table 4-6. The incidences of liver tumors
in control mice (18% in males and 4% in females for hepatocellular adenomas and 34% in males
and 4% in females for hepatocellular carcinomas) were similar to historical control data for liver
tumors in Crj :BDF1 mice in 20 studies at JBRC (see Table 4-6 for historical control liver tumor
62
DRAFT - DO NOT CITE OR QUOTE
-------
incidence). The gender differences in unexposed mice are thought to be related to inhibition of
liver tumor formation by female estrogen levels. The incidences of hepatocellular adenomas and
carcinomas were significantly elevated in both sexes at >25 ppm. At 5 ppm, the incidence of
liver adenomas in female mice (8/49 or!6%) was not statistically significantly elevated
compared to the concurrent control, but did exceed the historical control range (2-10%).
Table 4-6. Incidence of liver and adrenal tumors in BDF1 mice exposed to
carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)3
Tumor
Hepatocellular
adenoma
Hepatocellular
carcinoma
Hepatocellular
adenoma or carcinoma
Adrenal
pheochromocytomae
Male
0 ppm
9/50b
17/50b
24/50b
0/50b
5 ppm
10/50
12/50
20/50
0/50
25 ppm
27/50c
44/50c
49/50c
16/50C
125 ppm
16/50
47/50c
49/50c
32/50c
Female
0 ppm
2/50b
2/50b
4/50b
0/50b
5 ppm
8/49d
1/49
9/49
0/49
25 ppm
17/50C
33/50c
44/50c
0/50
125 ppm
5/49
48/49c
48/49c
22/49c
"The exposure concentrations adjusted to continuous exposure (i.e., multiplied by 5/7 x 6/24) = 0.9, 4.5, and 22.3
ppm.
Statistically significant trend for increased tumor incidence by Peto's test (p<0.01).
0 Tumor incidence was significantly elevated compared with controls by Fisher's exact test (p<0.01).
d Tumor incidence was significantly elevated compared with controls by Fisher's exact test (p<0.05).
e All pheochromocytomas in the mouse were benign with the exception of one malignant pheochromocytoma in
the 125-ppm male mouse group.
Note: Liver historical control data in Crj :BDF1 mice in 20 studies at JBRC: 17.1% (4-34%) in males and 5.2%
(2-10%) in females for hepatocellular adenoma and 20.1% (2-42%) in males and 2.4% (0-8%) in females for
hepatocellular carcinoma (letter dated March 8, 2004 and email dated March 9, 2004, from Kasuke Nagano,
JBRC, to Mary Manibusan, U.S. EPA).
Pheochromocytoma historical control data in Crj:BDFl mice in 32 studies at JBRC: 0.3% (range: 0-2%) in both
males and females (email dated October 15, 2005, from Kasuke Nagano, JBRC, to Mary Manibusan, U.S. EPA).
Source: Nagano et al. (2007b); JBRC (1998).
The incidence of adrenal pheochromocytoma was significantly increased in males at >25
ppm and in females at 125 ppm. This incidence exceeded the historical control incidence of
pheochromocytomas in Crj:BDFl mice in JBRC studies of 0.3% (range: 0-2%) in both males
and females (email dated October 15, 2005, from Kasuke Nagano, JBRC, to Mary Manibusan,
U.S. EPA).
63
DRAFT - DO NOT CITE OR QUOTE
-------
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND INHALATION
4.3.1. Oral Exposure
No adequate reproductive toxicity studies have been conducted in animals exposed by the
oral route. Teratogenicity has not been observed in the offspring of rats orally exposed to carbon
tetrachloride. However, total litter loss has been described at maternally toxic doses that are
higher than those associated with liver and kidney toxicity.
Alumot et al, 1976
Reproductive performance was monitored in an oral study in which rats of an unspecified
strain (18/sex/group) were fed for up to 2 years on experimental diets that had been fumigated
with carbon tetrachloride for 48 hours (Alumot et al., 1976). Doses could not reliably be
estimated. Serial matings were performed throughout the study. Rats fed fumigated food
showed no effects on reproduction (male and female fertility, litter size, and pup mortality and
body weight at birth and weaning). There was widespread occurrence of chronic respiratory
disease in animals from all groups after 14 months, but this probably did not affect the
reproductive outcomes because most reproductive activity took place during the first year of the
study (only seven successful matings occurred during the second year). Treatment-related
parental toxicity was not reported, but only parental body weight was monitored concurrently
with the reproductive part of the study. No evidence of liver toxicity was found by serum
analyses or biochemical tests at the end of the study. This study found no evidence of
reproductive or maternal effects, but doses received by the experimental animals are unknown.
Wilson, 1954
Wilson (1954) administered daily doses of 478 mg of carbon tetrachloride by gavage in
corn oil to 29 pregnant rats (strain not specified) on 1 or 2 successive days of gestation beginning
between gestational days (GDs) 7 and 11. The experiment was terminated on GD 20, at which
time surviving dams were sacrificed, uteri were examined for resorptions, and litters were
examined for external malformations. Fifty-nine percent of the dams failed to produce offspring;
this included 6 of 29 dams (21%) that died (a rate less than the 50% mortality for nonpregnant
rats given the same dose) and 11 of 29 dams (38%) that had total litter loss from early resorption.
For the 12 of 29 dams (41%) that produced offspring, the resorption rate was within normal
limits (9.1%), no fetuses were malformed, and only one litter contained fetuses with retarded
growth. Because the single dose level of carbon tetrachloride used in this study caused 21%
mortality in the dams, it is difficult to determine whether the observation of total litter loss was a
direct effect of carbon tetrachloride or was secondary to maternal toxicity.
Narotsky andKavlock, 1995; Narotsky etal., 1997a, b, 1995
Narotsky and Kavlock (1995) reported the results of a developmental toxicity screening
64 DRAFT - DO NOT CITE OR QUOTE
-------
study in rats. Groups of 16-21 timed-pregnant F344 rats were treated with 0, 112.5, or 150
mg/kg-day of carbon tetrachloride by gavage in corn oil on days 6-19 of gestation. Maternal
body weight was monitored periodically throughout gestation. The dams were allowed to litter.
Pups were examined on postnatal days (PNDs) 1, 3, and 6 and weighed on PNDs 1 and 6. Pups
found dead without gross external malformations were dissected and examined for visceral
malformations. After the final examination of their litters, dams were sacrificed and their uteri
were examined for implantation sites. Dams that did not litter by presumed day 24 of gestation
were sacrificed for uterine examination. Ammonium sulfide stain was used as needed to detect
full-litter resorption. No dams died during the study. The number of females actually pregnant
in each group was 13,9, and 14 in the control-, low-, and high-dose groups, respectively. Both
doses of carbon tetrachloride caused maternal weight loss (4-8%) early in the treatment period
and reduced extrauterine weight gain (35-45% lower than controls) over the treatment period as
a whole. The incidence of full-litter resorption was markedly increased in both dose groups: 4/9
(44%) and 10/14 (71%) in the 112.5 and 150 mg/kg-day groups, respectively (versus 0/13 in
controls). As a result, prenatal loss (reported as percent loss per litter) was significantly
increased in both dose groups. Implantation sites of the resorbed litters were not grossly visible
in most cases, requiring ammonium sulfide stain to find them. This suggested to the researchers
that the resorptions occurred early in pregnancy. Among dams that maintained their
pregnancies, resorptions were not increased nor were postnatal losses. Pup body weight was not
markedly affected by treatment. No malformations were associated with carbon tetrachloride
exposure. Reduced maternal weight gain and full-litter resorption were found at the low dose of
112.5 mg/kg-day in this study. In follow-up investigations, the researchers suggested that the
all-or-none nature of the observed resorptions points to a maternally mediated response and
produced evidence that the response is associated with reduced levels of progesterone and
luteinizing hormone (LH) in the dams (Narotsky et al., 1997a, 1995). In F344 rats administered
150 mg/kg carbon tetrachloride on GD 8, serum LH levels were significantly reduced (by 17-
69% at intervals up to 20 hours post-dosing) in animals with full-litter resorption; no adverse
developmental outcomes were observed in animals that received carbon tetrachloride and human
chorionic gonadotropin, which acted as an LH surrogate.
Narotsky et al. (1997b) compared the developmental toxicity of carbon tetrachloride
administered to rats by gavage in corn oil or an aqueous emulsion (10% Emulphor). Groups of
12-14 timed-pregnant F344 rats received carbon tetrachloride at doses of 0, 25, 50, or 75 mg/kg-
day in either vehicle on GDs 6-15. Maternal body weights were determined on GDs 5, 6, 8, 10,
13, 16, and 20. All dams were examined for clinical signs of toxicity and the day of parturition
was recorded. Pups were examined for viability and body weight on PNDs 1 and 6. Pups that
died without gross malformations were examined macroscopically for soft tissue alterations.
Dams were sacrificed on PND 6 and uterine implantation sites were counted. The uteri of
females that did not deliver were stained with 10% ammonium sulfide to detect sites of early
65 DRAFT - DO NOT CITE OR QUOTE
-------
resorption. There was no maternal mortality. Dose-related piloerection was observed in dams at
>50 mg/kg-day for both vehicles but was seen in more animals and for longer periods in the corn
oil groups. Dams exposed to 75 mg/kg-day in corn oil also exhibited kyphosis (rounded upper
back) and marked weight loss. Dams exposed to 50 and 75 mg/kg-day in water showed only
significantly reduced body weight gain. Full-litter resorption occurred with an incidence of 0/13,
0/13, 5/12 (42%), and 8/12 (67%) in the control through high-dose corn oil groups and 0/12,
0/12, 2/14 (14%), and 1/12 (8%) in the respective aqueous groups. The difference between
vehicles was statistically significant at the high dose. Among the surviving litters, there were no
effects on gestation length, prenatal or postnatal survival, or pup weight or morphology. The 25
mg/kg-day dose was a NOAEL and the 50 mg/kg-day dose a LOAEL for full-litter resorption
and maternal toxicity (piloerection) with either corn oil or aqueous vehicle, although these
effects were more pronounced with the corn oil vehicle.
Hamlin et al, 1993
Hamlin et al. (1993) treated pregnant female B6D2F1 mice with 0, 82.6, or 826 mg/kg of
carbon tetrachloride by gavage in corn oil on GDs 1-5. In this strain, GDs 1-5 are characterized
by sequential cleavage of the fertilized oocyte to generate a hatched blastocyte, with implantation
occurring on day 5 and organogenesis occurring subsequently. Therefore, dosing in this study
was limited to the preimplantation period. A total of 31 pregnant females were included in the
experiment, with a minimum of 8 in each dose group (actual group sizes were not reported).
Dams were allowed to give birth; litter size was recorded; and neonates were weighed, measured
for crown-rump length, and checked for obvious birth defects. During lactation, the pups were
weighed and measured for crown-rump length weekly. Lower incisor eruption and eye opening
were assessed in all pups on postpartum days 11 and 15, respectively. Pups were weaned on
postpartum day 22 and sacrificed. Dams were weighed weekly during pregnancy and on
postpartum day 22 just prior to sacrifice. The liver and kidneys from the dams were removed
and weighed. Liver and kidney tissue samples were collected for possible histopathological
examination at a later date but were not examined for this report. Treatment with carbon
tetrachloride had no effect on dam body weight during pregnancy or on absolute or relative liver
or kidney weight at sacrifice. Treatment also had no effect on litter size, pup size at birth, the
timing of developmental milestones (incisor eruption and eye opening), or pup growth through
weaning (a statistically significant difference in body weight between high-dose pups and
controls on day 15 postpartum was not considered to be biologically significant by the
researchers because crown-rump length was not affected and no other body weight differences
were found). No stillbirths or malformations were observed. The study report included only a
limited presentation of the results and no data were shown.
66 DRAFT - DO NOT CITE OR QUOTE
-------
4.3.2. Inhalation Exposure
The potential for reproductive toxicity of carbon tetrachloride in animals is suggested by
Bergman's (1983) finding of partly nonextractable radiolabel in the interstitial testis of mice
exposed by inhalation to [14C]-carbon tetrachloride vapor. In the subchronic inhalation study by
Adams et al. (1952), testicular atrophy was observed in rats exposed to 200 or 400 ppm (1,260 or
2,520 mg/m3) of carbon tetrachloride vapor 7 hours/day, 5 days/week for 6 months. Testicular
degeneration has also been reported in rats following repeated intraperitoneal (i.p.) doses of 1.5
mL/kg (Kalla and Bansal, 1975; Chatterjee, 1966). Smyth et al. (1936) found that fertility was
reduced in rats exposed to 200 or 400 ppm (1,260 or 2,520 mg/m3) of carbon tetrachloride vapor
8 hours/day, 5 days/week for up to 10.5 months.
The most detailed inhalation exposure study (Schwetz et al., 1974) suggests that
developmental effects of carbon tetrachloride occur at concentrations toxic to the mother and at
exposure concentrations higher than those associated with liver and kidney toxicity.
Oilman, 1971
As described in an abstract of an unpublished doctoral dissertation, Oilman (1971)
exposed groups of pregnant albino Sprague-Dawley rats to ambient air or 250 ppm (1,575
mg/m3) of carbon tetrachloride vapor for 8 hours/day on GDs 10-15. There were no adverse
effects on maternal body weight, litter size, the ratio of live to still births, or the incidence of
skeletal abnormalities.
Schwetz et al., 1974
Groups of 22-23 pregnant female Sprague-Dawley rats were exposed by inhalation to
carbon tetrachloride vapor at concentrations of 0, 334, or 1,004 ppm (0, 2,101, or 6,316 mg/m3)
for 7 hours/day on GDs 6-15 (Schwetz et al., 1974). Exposures to the two different dose levels
were not performed concurrently, so two separate control groups were used. Data from the two
control groups were combined except where they differed significantly (e.g., incidence of
delayed ossification of sternebrae). The rats were observed daily throughout pregnancy. Food
intake was monitored every other day during the experiment, and body weight was determined
on days 6, 13, and 21 of gestation. Following sacrifice on GD 21, the number and uterine
position of live, dead, and resorbed fetuses were recorded. The fetuses were weighed, measured,
and examined for external anomalies. Half of the fetuses in each litter were prepared so as to
enable detection of soft tissue anomalies upon subsequent examination, and the remainder were
prepared and examined for skeletal abnormalities. The litter was considered the unit of treatment
and observation when comparing the results from the different exposure groups. Nonpregnant
female rats were exposed simultaneously with the pregnant rats in order to monitor effects on the
liver. Serum ALT was determined in these rats throughout exposure, and some were sacrificed
for gross examination of the liver at the end of the exposure period. The remainder were
67 DRAFT - DO NOT CITE OR QUOTE
-------
sacrificed 6 days later (corresponding to the end of gestation in the pregnant rats) for ALT
analysis, gross examination of the liver, and determination of liver weight. In the 334- and 1004-
ppm groups, significant reductions in fetal body weight (7% and 14%, respectively) and crown-
rump length (3.5% and 4.5%, respectively) were found. The incidence of delayed ossification of
the sternebrae was significantly elevated in the high-dose group (13%) compared with the
concurrent control (2%) but not compared with the low-dose group or its concurrent control. No
other effects attributable to carbon tetrachloride exposure were found. No anomalies were seen
upon gross examination. A significant increase in subcutaneous edema was observed at 334 ppm
but not at 1,004 ppm. No other increases in individual soft tissue or skeletal anomalies were
reported. Maternal toxicity was also observed in both dose groups. Food consumption and body
weight were significantly reduced compared with controls, and hepatotoxicity was indicated by
significantly elevated serum ALT (fourfold increase over control), gross changes in liver
appearance, and significantly increased liver weight (26% at 334 ppm and 44% at 1,004 ppm).
This study, therefore, detected both maternal and developmental toxicity at a LOAEL of 334
ppm.
4.4. OTHER DURATION- OR ENDPOINT-SPECIFIC STUDIES
4.4.1. Acute and Short-term Toxicity Data
4.4.1.1. Oral Exposure
In animals acutely exposed to carbon tetrachloride by gavage, the liver appears to be the
primary target organ; damage to the kidney occurs at slightly higher doses (Blair et al., 1991;
Kim et al., 1990a, b; Bruckner et al., 1986; Hayes et al., 1986; Nakata et al., 1975; Litchfield and
Gartland, 1974; Korsrud et al., 1972; Gardner et al., 1925). Lung effects have also been noted
(Boyd et al., 1980; Gould and Smuckler, 1971). Hepatic toxicity is frequently measured by
significant increases in serum enzyme activities that peak between 24 and 48 hours after dosing:
ALT, AST, SDH, and OCT. The serum enzyme changes represent leakage from damaged
hepatocytes. Korsrud et al. (1972) indicated that overt hepatic necrosis was unnecessary for
detectable increases in serum enzymes. Reductions in the levels of microsomal protein,
microsomal enzymes (G6Pase), and CYP450 levels also occur after carbon tetrachloride dosing
(Kim et al., 1990a, b). Histopathological effects in the liver include centrilobular fatty
vacuolization, degeneration, necrosis, and inflammation.
Wangetal, 1997
Wang et al. (1997) monitored the time course of hepatic injury in Wistar rats treated with
3,188 mg/kg of carbon tetrachloride by gavage in corn oil. There were immediate steep declines
in the hepatic microsomal protein and CYP450 content, so that metabolic rates declined by 50%
or more, as measured in microsomal CYP content. Plasma levels of AST and ALT increased
100-fold by 24 hours. Immediate histopathological lesions of the liver included hepatocellular
68 DRAFT - DO NOT CITE OR QUOTE
-------
degeneration, necrosis, and hydropic swelling. Inflammatory cell infiltration was detectable
within 3 hours, and proliferation of mesenchymal cells began after 24 hours.
Lee et al, 1998
Lee et al. (1998) examined the time course and distribution of toxicity and repair in the
livers of male Sprague-Dawley rats 24, 36, and 48 hours after receiving 40 or 400 mg/kg carbon
tetrachloride by gavage in corn oil. Cell proliferation was monitored by pulse-labeling with
BrdU 1 hour before sacrifice. The high dose caused extensive damage in the perivenous-to-
midlobular zones. Administration of 40 mg/kg induced regenerative hepatocyte proliferation, as
indicated by a significant elevation in BrdU-positive cells in the periportal zone (the site of
necrosis) at 24 hours, increasing at 36 hours and plateauing at 48 hours. BrdU-positive cells
were close to the portal tract at 24 hours and then increasingly in the outer periportal and
midlobular zones at later times. A few hepatocytes in the perivenous zone adjacent to the area of
cell damage were labeled at all time points.
Steup et al, 1993
Steup et al. (1993) also found significantly elevated serum ALT and SDH levels in male
F344 rats 3-72 hours after they received a single dose of 80 mg/kg carbon tetrachloride by
gavage in 10% Emulphor; peak enzyme levels were at 24 hours. Hepatic GSH concentrations
were significantly elevated in treated rats at 48 hours after dosing. Six hours after treatment,
hepatocytes near terminal venules (zone 3) showed some depletion of glycogen and ballooning.
Small collections of lymphocytes were adjacent to focal necrosis of single hepatocytes. More
extensive injury involved confluent areas of necrotic cells. Hepatocellular lysis was evident by
48 hours and a mononuclear cell infiltrate concentrated around terminal hepatic venules. Mitotic
figures predominated in the cells of the surrounding tissue. By 72 hours, recovery was evident
with only a mild infiltrate of mononuclear cells at the site of injury.
Evidence of regeneration of livers in animals treated with carbon tetrachloride appears
within 48 hours of dosing. In strain A mice dosed with 2,550 mg/kg of carbon tetrachloride in
olive oil, necrosis was detectable in half the hepatocytes at 24 hours, and mitotic activity
appeared 48 hours after dosing (Eschenbrenner and Miller, 1946). Wistar rats treated with 7,970
mg/kg had peak ALT levels at 24 hours, peak AST levels at 48 hours, and significantly elevated
levels for activities of DNA-synthesizing enzymes thymidine kinase and thymidylate synthetase
at 48 and 72 hours (Nakata et al., 1975); activity levels for DNA-synthesizing enzymes were
reduced at 96 hours. Doolittle et al. (1987) found that, in male CD-I mice administered a single
gavage dose or multiple (1, 7, or 14) daily doses of carbon tetrachloride in corn oil (up to 100
mg/kg-day), dose levels high enough to elicit significant increases in serum ALT and AST also
significantly increased the number of hepatocytes in S-phase, beginning 24 hours after dosing.
69 DRAFT - DO NOT CITE OR QUOTE
-------
Multiple doses tended to lower the concentration required to induce hepatotoxicity and increased
the number of hepatocytes in S-phase (DNA-synthesizing phase of the cell-replication cycle).
The effect of dosing vehicle on carbon tetrachloride-induced hepatic toxicity has been
investigated in several studies. Kim et al. (1990a, b) reported that administration in a corn oil
vehicle resulted in lower acute hepatotoxicity (as measured by serum SDH and ALT levels over
a 72-hour period) compared with administration in an aqueous emulsion or as undiluted carbon
tetrachloride. Raymond and Plaa (1997) reported no consistent difference in serum ALT levels
measured 48 hours after dosing in male Sprague-Dawley rats given carbon tetrachloride (5.2-
25.8 mmol/kg) in corn oil, 5% aqueous Emulphor emulsion, or Tween-85 (undiluted carbon
tetrachloride was not tested).
Damage to the lung has been noted in rodents exposed to carbon tetrachloride by gavage.
After male Sprague-Dawley rats received a single dose of 4,000 mg/kg in mineral oil, pulmonary
histopathological effects included perivascular edema and mononuclear infiltration after 4 hours
and atelectasis (collapsed lung) and intraalveolar hemorrhages after 8 hours (Gould and
Smuckler, 1971). In male Swiss mice or Sprague-Dawley rats, there were significant reductions
in pulmonary CYP450 levels and the activity of the microsomal enzyme benzphetamine
demethylase 16 hours after receiving a single dose of 4,000 mg/kg of carbon tetrachloride in
50% sesame oil (Boyd et al., 1980). Clara cells showed histopathological changes (swelling and
necrosis with pyknotic nuclei), whereas the adjacent ciliated bronchiolar cells had normal
histology.
4.4.1.2. Inhalation Exposure
The central nervous system and the liver are the primary targets in acute toxicity studies
in animals exposed by inhalation. Suppression of the central nervous system occurs at relatively
high concentrations and is an immediate effect. In Wistar rats exposed for 7 hours, stupor was
observed at 4,600 ppm, incoordination at 7,300 ppm, and unconsciousness at 12,000 ppm
(Adams et al., 1952); 16-24 hours after exposure, these rats exhibited increased liver weights
and centrilobular fatty degeneration of the liver. Significant elevations in serum enzymes (ALT,
AST, SDH, and GDH) have been observed within 24 hours of acute inhalation exposures
(Paustenbach et al., 1986a, b; Siegers et al., 1985; Brondeau et al., 1983; Jaeger et al., 1975). In
addition, hepatic histopathology within 24 hours of a 4-hour exposure showed centrilobular
hydropic or necrotic parenchymal cell damage (Magos et al., 1982).
Hepatotoxicity, and to a lesser extent nephrotoxicity, appear to be the primary effects of
short-term duration inhalation exposures. Exposures of male Sprague-Dawley rats at 100 ppm, 8
or 11.5 hours/day for 5 or more days resulted in fatty changes in the liver (Paustenbach et al.,
1986a, b); nephrosis (degenerative changes in the kidney) was characterized as minor in rats
exposed for 8 hours/day but was more significant in rats exposed for 11.5 hours/day.
Plummer et al. (1990) conducted a 4-week inhalation toxicity study in male Wistar rats
70 DRAFT - DO NOT CITE OR QUOTE
-------
exposed to carbon tetrachloride vapor continuously at 16 ppm (100 mg/m3) for 24 hours/day, 7
days/week except for 1.5-hour periods on Mondays and Fridays, or discontinuously at 87 ppm
(50 mg/m3) for 6 hours/day, 5 days/week. The total time-weighted average exposures
(concentration H time) were the same: 10,507 ppm-hours for the continuous regimen and 10,458
ppm-hours for the discontinuous regimen. Liver histopathology (fibrosis and cirrhosis) was
indistinguishable between the two groups, suggesting that inhalation toxicity from carbon
tetrachloride is proportional to the product of concentration x time. In another 4-week study,
Bogers et al. (1987) exposed groups of Wistar rats to 6-hour daily exposures of carbon
tetrachloride vapor at 63 or 80 ppm, either uninterrupted or in 2-hour sessions with an
interruption of 1.5 hours; peak loads were added for some groups. At 80 ppm, serum enzyme
levels were slightly but significantly increased in the interrupted-exposure groups compared with
the uninterrupted-exposure groups (the 63 ppm groups were not compared).
4.4.1.3. Acute Studies Comparing Oral and Inhalation Exposures
The effect of route of administration on the hepatic toxicity of carbon tetrachloride has
been evaluated in rats (Sanzgiri et al., 1997; Bruckner et al., 1990). In both studies, male
Sprague-Dawley rats were exposed (nose only) to carbon tetrachloride vapor at 100 or
1,000 ppm (630 or 6,300 mg/m3) for 2 hours. The systemically absorbed doses were calculated
from measurements of minute volume and differences between concentrations in inhaled and
exhaled air over time; the doses were calculated as 18.9 and 186 mg/kg by Bruckner et al. (1990)
and as 17.5 and 179 mg/kg by Sanzgiri et al. (1997). Subsequently, groups of four to nine rats
were exposed by inhalation for 2 hours or given the same doses by gavage as a bolus delivery or
as a gastric infusion over 2 hours. Hepatotoxicity was measured by activities of SDH and ALT
in serum samples taken 24 hours after dosing, and the concentration of CYP450 and activity of
G6Pase per mg of hepatic microsomal protein. The results of the two studies are similar; those
for Sanzgiri et al. (1997) are presented in Table 4-7. SDH and ALT values were not significantly
affected by inhalation exposure at 100 ppm or gastric infusion at 17.5 mg/kg but were
significantly elevated at 1,000 ppm or 179 mg/kg. In comparison, oral bolus dosing caused more
severe elevations at both dose levels. CYP450 levels were significantly reduced in all treated
groups, with more severe effects for the gastric routes at 17.5 mg/kg and the oral bolus route at
179 mg/kg. Suppression of microsomal G6Pase activity was most severe for gastric infusion at
both doses, followed by bolus delivery at both doses. Inhalation exposure at 100 ppm slightly
decreased G6Pase activity, but exposure at 1,000 ppm was not significantly different from the
control. Overall, the results indicate more severe hepatic toxicity when carbon tetrachloride is
administered as a single bolus, compared with the same dose administered by inhalation or
gastric infusion over a longer period of time.
71 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-7. Hepatic toxicity in rats exposed to carbon tetrachloride by
inhalation or by equivalent oral dosing as bolus or 2-hour gastric
infusion
Exposure
Control3
Inhalation13
Gastric infusion
Oral bolus
Inhalation13
Gastric infusion
Oral bolus
Dose
(mg/kg)
0
17.5
17.5
17.5
179
179
179
SDH
(mU/mL)
5.2±1.0C
11.3±3.7C
6.0±1.6C
64.6±12.5d
87.6±25.7d
96.9±18.0d
269.0 ±44.7e
ALT
(mU/mL)
24.4±2.2C
19.3±1.7C
15.9±2.3C
55.5±9.9d
53.3±14.7d
81.0±8.2d
176.5 ±17.4e
P450
(nmol/mg protein)
0.81±0.02C
0.65±0.05d
0.46±0.04e
0.49±0.06e
0.61±0.04d
0.63±0.05d
0.47±0.04e
G6Pase
(umol/hour/mg protein)
14.5±0.7C
10.9±0.5d
7.3±0.7e
12.5±0.1d
14.3±0.9C
7.8±0.7e
8.9 ± 0.3d
a Controls were treated with corn oil by gavage.
b 100 or 1,000 ppm for 2 hours.
°"e Means of each parameter that are statistically equivalent share the same superscript.
Source: Sanzgiri et al. (1997).
Magos et al. (1982) compared the isotoxic oral and 4-hour inhalation concentrations of
carbon tetrachloride in Porton-Wistar or Fischer rats. For exposures by either route, Fischer rats
were twice as sensitive to hepatotoxic effects (based on SGPT and extent of liver centrilobular
damage) of carbon tetrachloride as the Porton-Wistar rats. Fischer rats required an inhalation
concentration 1.5 times lower and an oral dose 3.3 times lower than Porton-Wistar rats to
produce a 10-fold increase in serum ALT levels, measured 20 hours after exposure.
4.4.2. Genotoxicity Studies
The results of genotoxicity studies of carbon tetrachloride are summarized in Tables 4-8
to 4-11. These tables are not intended to provide an exhaustive list of genotoxicity studies for
carbon tetrachloride, but rather to represent a reasonably comprehensive summary of the
available genotoxicity literature. A review of the genotoxicity literature is also provided in
Eastmond (2008).
72
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Genotoxicity studies of carbon tetrachloride in prokaryotic organisms
Test system
S. typhimurium TA100,
TA1535
S. typhimurium his G46,
TA1950
S. typhimurium his G46,
TA1950
S. typhimurium TA98, TA100,
TA1535,TA1537, TA1538
S. typhimurium TA97, TA98,
TA100
S. typhimurium TA98, TA100,
TA1535,TA1537
S. typhimurium TA1535,
TA1538
S. typhimurium TA97, TA98,
TA100, TA1535, TA1537
S. typhimurium TA97, TA98,
TA100, TA1535
S. typhimurium TA98, TA100,
TA1535, TA1537, TA1538
S. typhimurium TA100,
TA1535
S. typhimurium TA98, TA100,
TA1535
S. typhimurium TA100,
TA1535,TA1537
Endpoint
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Test conditions
Plate incorporation assay
Spot test
Host-mediated assay in male
NMRImice
Plate incorporation assay
Plate incorporation assay
Plate incorporation assay
Preincubation assay using
capped tubes
Preincubation assay using
capped tubes
Preincubation assay using
capped tubes
Gas phase exposure in
dessicator for 7-10 hours
Gas phase exposure in
dessicator for 7-8 hours
Gas phase exposure in closed
incubation system for 48 hours
Gas phase exposure in a gas
sampling bag for 24 hours
Results3
Without
activation
-
-
NA
-(T)
-
+d
-
-
-
-
-
-
-(T)
With
activation1"
-
-
-
-(T)
-
+d
-
-
-
-
-
-
-(T)
Dosec
10,000 ug/plate
4,000 ug/plate
6,400 mg/kg
10,000 ug/plate
inDMSOd
1,000 ug/plate
inDMSOd
2,460 ug/plate
in methanol
1,230 ug/mL
3,333 ug/plate
inDMSO
3,333 ug/plate
inDMSO
ND
ND
2,830 ug/plate
50,000 ppm
Reference
McCannetal., 1975
Braun and
Schoneich, 1975
Braun and
Schoneich, 1975
De Flora, 1981
Bramsetal., 1987
Varmaetal., 1988
Uehleke et al., 1977
Zeigeretal., 1988
Zeigeretal., 1988
Simmon etal., 1977
Simmon and Tardiff,
1978
Barber etal., 1981
Araki et al., 2004
73
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Genotoxicity studies of carbon tetrachloride in prokaryotic organisms
Test system
S. typhimurium TA98
Escherichia coli
WP2ttvrA/pKM101
E. coli WP2/pKM101
E. coli WP2wvrA
S. typhimurium BA13 and
BAL13
S. typhimurium BA13 and
BAL13
S. typhimurium
TA1535/pSK1002
E. coli PQ37
E. coli WP2, WP67, CM871
E. coli WP2, WP67, CM871
E. coli WP2, WP67, CM871
E. coli K-12 343/636, K-12
343/591
Endpoint
Reverse mutation
Reverse mutation
Reverse mutation
Reverse mutation
Forward mutation
Forward mutation
DNA repair
DNA repair
Differential DNA
repair
Differential DNA
repair
Differential DNA
repair
Differential DNA
repair
Test conditions
Gas phase exposure in a gas
sampling bag for 24 hours
Gas phase exposure in a gas
sampling bag for 24 hours
Gas phase exposure in a gas
sampling bag for 24 hours
Gas phase exposure in a
desiccator
Preincubation assay for L-
arabinose resistance (AraR test)
Preincubation assay for L-
arabinose resistance (AraR test)
SOS response indicated by umu
gene expression
SOS chromotest
Liquid micromethod using
sealed plates
Preincubation assay in sealed
tubes
Spot test
Preincubation assay
Results"
Without
activation
±
±
+
ND
-
±
-
-
+
+
-
-
With
activation1"
-
±
+e
±
-
-
-
-
+
ND
ND
-
Dosec
10,000 ppm
10,000 ppm
5,000 ppm
25,000 ppm
1,230 ug/plate
inDMSOd
384 ug/plate in
DMSOd
5,300 ug/mL
1,540 ug/mL in
DMSO
12.5 ug
ND
ND
15,400 ug/mL
Reference
Araki et al., 2004
Araki et al., 2004
Araki et al., 2004
Norpothetal., 1980
Roldan-Arjona et al.,
1991
Roldan-Arjona and
Pueyo, 1993
Nakamuraetal.,
1987
Bramsetal., 1987
De Flora etal., 1984
De Flora etal., 1984
De Flora etal., 1984
Hellmer and
Bolcsfoldi, 1992
74
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Genotoxicity studies of carbon tetrachloride in prokaryotic organisms
Test system
Endpoint
Test conditions
Results"
Without
activation
With
activation1"
Dosec
Reference
a + = positive, ± = equivocal or weakly positive, - = negative, (T) = toxicity, ND = no data.
b Exogenous metabolic activation used, typically induced rat liver S9.
0 Lowest effective dose for positive results, highest dose tested for negative results, ND = no data, NA = not applicable.
d Increase in revertants not dose-related and cytotoxicity not discussed.
e Results similar with or without glutathione added to the S9 mix. Positive response is based on the magnitude of response as statistical analyses were not
performed.
DMSO = dimethyl sulfoxide; SOS = inducible DNA repair system
75
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-9. Genotoxicity studies of carbon tetrachloride in non-mammalian eukaryotic organisms
Test system
Saccharomyces cerevisiae
D7
S. cerevisiae D7
S. cerevisiae D7
S. cerevisiae RSI 12
S. cerevisiae RSI 12
S. cerevisiae RSI 12
S. cerevisiae RS112
(arrested in S phase)
S. cerevisiae RSI 12
(arrested in S phase)
S. cerevisiae RSI 12
(arrested in Gl phase)
S. cerevisiae RSI 12
(arrested in Gl phase)
S. cerevisiae AGY3
(arrested in G2 phase or
growing normally)
S. cerevisiae D61.M
Endpoint
Gene conversion
Mitotic recombination
Reverse mutation
Intrachromosomal
recombination
Intrachromosomal
recombination
Interchromosomal
recombination
Intrachromosomal
recombination
Interchromosomal
recombination
Intrachromosomal
recombination
Interchromosomal
recombination
Intrachromosomal
recombination
Aneuploidy
Test conditions
Preincubation assay in capped
tubes
Preincubation assay in capped
tubes
Preincubation assay in capped
tubes
Preincubation assay
Preincubation assay
Preincubation assay
Preincubation assay
Preincubation assay
Preincubation assay
Preincubation assay
Preincubation assay
Standard 16-hour incubation or
cold-interruption regimen
Results3
Without
activation
+ (T)
+ (T)
+ (T)
+ (T)
+ (T)
+ (T)
-
-
+ (T)
+ (T)
+ (T)
-
With
activation1"
ND
ND
ND
ND
+ (T)
+ (T)
ND
ND
ND
ND
ND
ND
Dosec
5,230 ug/mL
5,230 ug/mL
5,230 ug/mL
2,000 ug/mL
4,000 ug/mL
4,000 ug/mL
8,000 ug/mL
8,000 ug/mL
5,000 ug/mL
5,000 ug/mL
8,000 ug/mL
6,400 ug/mL
Reference
Callenetal., 1980
Callenetal., 1980
Callenetal., 1980
Brennan and
Schiestl, 1998
Schiestl et al., 1989;
Galli and Schiestl,
1998
Galli and Schiestl,
1998
Galli and Schiestl,
1998
Galli and Schiestl,
1998
Galli and Schiestl,
1996; Galli and
Schiestl, 1998
Galli and Schiestl,
1996; Galli and
Schiestl, 1998
Galli and Schiestl,
1995
Whittakeretal.,
1989
76
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-9. Genotoxicity studies of carbon tetrachloride in non-mammalian eukaryotic organisms
Test system
Aspergillus nidulans PI
A. nidulans 35
A. nidulans PI
A. nidulans PI
Drosophila melanogaster
Endpoint
Somatic segregation
due to cross over and
aneuploidy
Forward mutation
Somatic segregation
(positive for
aneuploidy; negative
for cross over)
Somatic segregation
(positive for
aneuploidy; negative
for cross over)
Mutation
Test conditions
Plate incorporation assay
Plate incorporation and
growth-mediated assays
Mitotic segregation assay
Mitotic segregation assay
Sex-linked recessive lethal
assay
Results"
Without
activation
+ (T)
±(T)
+(T)
+ (T)
With
activation1"
ND
ND
ND
ND
NA
Dosec
0.5%
0.5%
0.04%
0.0275%
25,000 ppm in
feed or 2,000
ppm injection
Reference
Gualandi, 1984
Gualandi, 1984
Crebellietal., 1988
Benigni et al., 1993
Foureman et al.,
1994
a + = positive, ± = equivocal or weakly positive, - = negative, (T) = toxicity, ND = no data.
b Exogenous metabolic activation not used for most tests because fungi have metabolic capabilities.
0 Lowest effective dose for positive results, highest dose tested for negative results, ND = no data, NA = not applicable.
77
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-10. Genotoxicity studies of carbon tetrachloride in mammalian cells in vitro
Test system
Human peripheral lymphocytes
Go
Human peripheral lymphocytes
Go
Human lymphocytes from two
donors
Human lymphocytes
Human lymphocytes
Lamb peripheral lymphocytes
Lamb peripheral lymphocytes
Lamb peripheral lymphocytes
h2El cell line (cDNA for
CYP2E1)
MCL-5 cell line (cDNA for
CYPs 1A2, 2A6, 3A4, and
2E1, and epoxide hydrolase)
AHH-1 cell line (expresses
CYP1A1)
Chinese hamster ovary cells
Endpoint
Chromosomal
aberrations
Sister chromatid
exchange
Micronucleus
formation
DNA breaks
Unscheduled DNA
synthesis
Chromosomal
aberrations
Micronucleus
formation
Sister chromatid
exchange
Micronucleus
formation
Micronucleus
formation
Micronucleus
formation
Chromosomal
aberrations
Test conditions
30 Minute incubation in sealed
tubes
30 Minute incubation in sealed
tubes
Test conducted in capped tubes
Comet assay
4-Hour culture, autoradiography
48-Hour incubation
48-Hour incubation
48-Hour incubation
Immunofluorescent labeling of
kinetochore proteins
Immunofluorescent labeling of
kinetochore proteins
Immunofluorescent labeling of
kinetochore proteins
Assay conducted in sealed flasks
Results3
Without
activation
-(T)
-(T)
(2-)d
-
-
-
+
+
+ e(T)
+ e(T)
-
-
With
activation1"
-(T)
-(T)
±
(l-)d
-
-
ND
+
±
ND
ND
ND
-
Dosec
76 ug/mL
48 ug/mL
1,540 ug/mL
3,080 ug/mL
16,000 ug/mL
16 ug/mL
8 ug/mL (w/out
activation)
16 ug/mL
(w/activation)
4 ug/mL
308 ug/mL
308 ug/mL
1,540 ug/mL
3,000 ug/mL in
DMSOf
Reference
Garry etal., 1990
Garry etal., 1990
Tafazolietal., 1998
Tafazolietal., 1998
Perocco and Prodi,
1981
Sivikova etal., 2001
Sivikova etal., 2001
Sivikova etal., 2001
Dohertyetal., 1996
Dohertyetal., 1996
Dohertyetal., 1996
Lovedayetal., 1990
78
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-10. Genotoxicity studies of carbon tetrachloride in mammalian cells in vitro
Test system
Chinese hamster ovary cells
Chinese hamster ovary cells
V79 Chinese hamster lung cell
line
V79 Chinese hamster lung cell
line
Syrian hamster embryo cells
Mouse lymphoma L5178Y
cells
Mouse lymphoma L5178Y
cells
RLi cultured cell line derived
from rat liver
RLi cultured cell line derived
from rat liver
Hepatocytes- primary cultures
from four human donors
Hepatocytes isolated from male
Sprague-Dawley rats
Hepatocytes isolated from rats
Endpoint
Sister chromatid
exchange
Lagging
chromosomes and
multipolar spindles
Aneuploidy
c-Mitosis (spindle
disturbance)
Morphological
transformation
Mutation at tk locus
DNA strand breaks
Chromosomal
aberrations
Sister chromatid
exchange
Unscheduled DNA
synthesis
Unscheduled DNA
synthesis
DNA single strand
breaks
Test conditions
Assay conducted in sealed flasks
Anaphase analysis
3 -Hour incubation
30-Minute incubation
Clonal assay
4-Hour incubation
Alkaline elution
Assay conducted in sealed flasks
Assay conducted in sealed flasks
21.5-24 24-hr incubation
periods
Autoradiography and flow
cytometric assays
Alkaline elution
Results"
Without
activation
-(T)
+
+
±(T)
±g
ND
ND
-
-
ND
-
±(T)
With
activation1"
ND
ND
ND
ND
-(T)
+(T)
ND
ND
(4-)d
ND
ND
Dosec
1,490 ug/mL
(w/out
activation) 2930
ug/mL(w/
activation) note:
both inDMSOf
8,000 ug/mL
246 ug/mL
492 ug/ml
3 ug/mL
635 ug/mL
1,007 ug/mL
0.02 ug/mL in
DMSOd
0.02 ug/mL in
DMSOd
154 ug/mL
154 ug/mL
461 ug/mL
Reference
Lovedayetal., 1990
Coutino, 1979
(Melt, 1987
(Melt, 1987
Amacher and
Zelljadt, 1983
Wangenheim and
Bolcsfoldi, 1988
Garberg et al., 1988
Dean and Hodson-
Walker, 1979
Dean and Hodson-
Walker, 1979
Butterworthetal.,
1989
Seldenetal., 1994
Sinaetal., 1983
79
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-10. Genotoxicity studies of carbon tetrachloride in mammalian cells in vitro
Test system
Hepatocytes isolated from
female Wistar rats
Hepatocytes isolated from
female Wistar rats
Hepatocytes isolated from
female Wistar rats
CalfthymusDNA
CalfthymusDNA
Mouse liver chromatin
Hepatocytes isolated from
Sprague-Dawley rats
Hepatocytes isolated from C3H
mice
Hepatocytes isolated from
Syrian golden hamsters
Endpoint
DNA single strand
breaks
DNA adduct
formation
DNA adduct
formation
DNA binding of
radiolabeled
chemical
DNA binding of
radiolabeled
chemical
DNA binding
DNA binding
DNA binding
DNA binding
Test conditions
Comet assay
MidG adducts formed secondary
to lipid peroxidation
SoxodG adducts formed
secondary to lipid peroxidation
30-min incubation with rat and
mouse microsomes
60-min incubation under a N2
atmosphere
2 and 4 hr incubation with
binding measured in DNase I-
sensitive and -resistant
chromatin DNA
Measured as radioactivity bound
to DNA after a 1-hr incubation
with microsomes
Measured as radioactivity bound
to DNA after a 1-hr incubation
with microsomes
Measured as radioactivity bound
to DNA after a 1-hr incubation
with microsomes
Results"
Without
activation
±
±
±(T)
+
ND
ND
±
±
±
With
activation1"
ND
ND
ND
+
+
+
±
±
±
Dosec
154 ug/mL
154 ug/mL
615 ug/mL
5.6 ug/mL
154 ug/ml
192 ug/mL
31 ug/mL
31 ug/mL
31 ug/mL
Reference
Beddowes etal,
2003
Beddowes etal.,
2003
Beddowes etal.,
2003
Rocchietal., 1973
DiRenzoetal., 1982
Oruambo and Van
Duuren, 1987
Castro etal., 1989
Castro etal., 1989
Castro etal., 1989
80
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-10. Genotoxicity studies of carbon tetrachloride in mammalian cells in vitro
Test system
Endpoint
Test conditions
Results"
Without
activation
With
activation1"
Dosec
Reference
a + = positive, ± = equivocal or weakly positive, - = negative, (T) = toxicity, ND = no data.
b Exogenous metabolic activation used, typically induced rat liver S9.
0 Lowest effective dose for positive results, highest dose tested for negative results, ND = no data, NA = not applicable.
d Results for the individual donors are presented.
e Increase mostly in kinetochore-positive (aneugenic) micronuclei which occurred at the lower (308 ug/mL) concentration, and some increase in kinetochore-
negative (clastogenic) micronuclei which was significantly increased at the highest (1538 ug/mL) test concentration.
f DMSO = dimethyl sulfoxide
8 Although declared positive by the authors, the induced frequency is well within the currently accepted control range.
81
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Mouse (101/H, male)
Rat (Sprague-Dawley, male)
Mouse (BDF1, male)
Mouse (BDF1, male)
Mouse (BDF1, male)
Mouse (CD-I, male)
Mouse (CD-I, male and
female)
Mouse (CD-I, male)
Rat (F344, male)
Endpoint
Chromosomal
aberrations in bone
marrow
Chromosomal
aberrations in bone
marrow
Micronucleus
formation in bone
marrow
Micronucleus
formation in bone
marrow
Micronucleus
formation in
peripheral blood
Micronucleus
formation in
peripheral blood
Micronucleus
formation in bone
marrow
DNA damage in
stomach, kidney,
bladder, lung, brain,
and bone marrow
DNA breakage
Test conditions
Metaphase analysis of samples
collected 6 to 48 hr after
dosing
Metaphase analyses from
animals sacrificed 24 hr after
dosing
Analyzed polychromatic
erythrocytes from specimens
prepared 24 hours after dosing
Analyzed polychromatic
erythrocytes from specimens
prepared 24 hours after dosing
Analyzed reticulocytes from
specimens prepared 24-72
hours after dosing
Analyzed reticulocytes from
specimens prepared 24-72
hours after dosing
Analyzed polychromatic
erythrocytes from femur bone
marrow of mice killed 24 or
48 hours after dosing
Comet assay on stomach,
kidney, bladder, lung, brain,
and bone marrow obtained 0,
3, or 24 hours after dosing
Comet assay on peripheral
blood cells
Results3
Without
activation
- (T)
"
-(T)
-(T)
"
d
-(T)
±(T)
With
activation1"
NA
NA
NA
NA
NA
NA
NA
NA
NA
Dosec
8,000 mg/kg
injected i.m.
1,600 mg/ml
by gavage
2,000 mg/kg
by gavage (2 x)
2,000 mg/kg
by gavage
3,000 mg/kg
by i.p. injection
2,000 mg/kg
by gavage in
olive oil
3,000 mg/kg
i.p. in olive oil
2,000 mg/kg
by gavage
120 mg/kg by
i.p. injection
Reference
Lil'p, 1982
Rossi etal., 1988
Moritaetal., 1997;
Suzuki etal., 1997
Moritaetal., 1997;
Suzuki etal., 1997
Suzuki etal., 1997
Moritaetal., 1997
Crebelli et al., 1999
Sasaki etal., 1998
Kadiiska et al., 2005
82
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Mouse (NMRI, male and
female)
Rat (Wistar, female, partially
hepatectomized)
Rat (F344, male)
Rat (strain and sex not
specified)
Rat (BD-VI, male)
Rat (Sprague-Dawley, male)
Mouse (CD-I, male)
Rat (Sprague-Dawley CD stain,
female)
Rat (Sprague-Dawley, male)
Endpoint
DNA strand breaks
in liver
DNA damage in
liver
DNA strand breaks
in liver
DNA breaks in liver
DNA strand breaks
in liver
DNA damage in
liver
DNA strand breaks
in liver
DNA strand breaks
in liver
DNA strand breaks
in liver
Test conditions
Alkaline elution of sample
collected 4 hr after dosing
Caffeine elution 4 or 24 hours
after dosing
Alkaline elution on primary
hepatocytes isolated from rats
sacrificed 2-48 hours after
dosing
Alkaline elution on liver
nuclei obtained 1 hr after
dosing
Alkaline elution on primary
hepatocytes isolated from rats
sacrificed 4 hours after dosing
Viscometric assay on rats
sacrificed 2 hours after dosing
Alkaline elution
Alkaline elution on primary
hepatocytes isolated from rats
dosed 21 and 4 hrs before
sacrifice
DNA strand breaks in
hepatocytes were measured by
a fluorometric assay for DNA
unwinding 1 hr after dosing
Results"
Without
activation
-
"
"
-(T)
-
+ (T)
With
activation1"
NA
NA
NA
NA
NA
NA
NA
NA
NA
Dosec
4,000 mg/kg
by gavage
800 mg/kg by
gavage in corn
oil
400 mg/kg by
corn oil gavage
4 mg/kg by i.p.
injection
4,000 mg/kg
by i.p. injection
200 mg/kg by
i.p. injection
80 mg/kg by
corn oil gavage
1,050 mg/kg
by oral gavage
in corn oil (2x)
160 mg/kg in
corn oil by i.p.
Reference
Schwarzetal., 1979
Stewart, 1981
Bermudez et al.,
1982
Kittaetal., 1982
Barbinetal., 1983
Brambilla et al.,
1983
Cans and Korson,
1984
Kitchin and Brown,
1989
Ikegwuonu and
Mehendale, 1991
83
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Rat (Wistar, male)
Rat (Wistar, male)
Mouse (CD-I, male)
Rat (Wistar, male)
Rat (Wistar, male)
Rat (Wistar, female)
Rat (Wistar, female)
Endpoint
DNA strand breaks
in liver
DNA strand breaks
in liver
DNA damage in
liver
DNA fragmentation
in liver
DNA fragmentation
in liver
Unscheduled DNA
synthesis in liver
Unscheduled DNA
synthesis in liver
Test conditions
Breaks in DNA of non-
parenchymal cells identified
by in situ nick translation 12-
96 hrs after dosing.
Breaks in DNA of non-
parenchymal cells identified
by in situ nick translation after
dosing twice a week until
week 12 with sacrifices at 3,
6, 9, 12, 15, and 18 weeks.
Comet assay on liver obtained
0, 3, or 24 hours after dosing
TUNEL f assay on rats
sacrificed 1 day after the
second dose
TUNEL f assay on rats
sacrificed at 10, 15,20,25,
and 30 hr after dosing
Animals injected with
hydroxyurea (to stop de novo
DNA synthesis) and then
[3H]-thymidine 2 hours after
dosing
Animals injected with
hydroxyurea (to stop de novo
DNA synthesis) and then
[3H]-thymidine 17 hours after
dosing
Results"
Without
activation
±(T)e
±(T)e
+ (T)
+ (T)
+ (T)
+ (T)
With
activation1"
NA
NA
NA
NA
NA
NA
NA
Dosec
1,600 mg/kg
i.p. in olive oil
2,000 mg/kg
(24x)
1,000 mg/kg
by gavage
800 mg/kg by
ip; (2x)
240 mg/kg in
corn oil by ip
4,000 mg/kg
by gavage in
liquid paraffin
4,000 mg/kg
by gavage in
liquid paraffin
Reference
Nakamura and
Hotchi, 1992
Nakamura and
Hotchi, 1992
Sasaki etal., 1998
Cabreetal., 1999
Yasuda et al., 2000
Craddock and
Henderson, 1978
Craddock and
Henderson, 1978
84
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Rat (F344, male)
Rat (F344, male)
Mouse (B6C3F1, male)
Mouse (B6C3F1, female)
Mouse (CD-I, male)
Rat (Sprague-Dawley, male)
Mouse (DC-1, male)
Endpoint
Unscheduled DNA
synthesis in liver
Unscheduled DNA
synthesis in liver
Unscheduled DNA
synthesis in liver
Unscheduled DNA
synthesis in liver
Unscheduled DNA
synthesis in liver
Unscheduled DNA
synthesis
Chromosomal
fragments and
bridges in liver
Test conditions
Rats sacrificed 2 hrs after
dosing; primary hepatocytes
isolated by liver perfusion and
cultured with [3H]-thymidine
Rats sacrificed 2-48 hours
after dosing; primary
hepatocytes isolated by liver
perfusion and cultured with
[3H]-thymidine
Rats sacrificed 12 hrs after
dosing; primary hepatocytes
isolated by liver perfusion and
cultured with [3H]-thymidine
Rats sacrificed 12 hrs after
dosing; primary hepatocytes
isolated by liver perfusion and
cultured with [3H]-thymidine
Mice sacrificed 3-48 hours
after dosing; liver cells
isolated and analyzed by
autoradiography
Unscheduled DNA synthesis
by labeling of DNA in
hydroxyurea-treated animals 1
hr after dosing
Anaphase analysis of squash
preparations prepared 72 hrs
after dosing
Results"
Without
activation
-(T)
-(T)
-(T)
-(T)
±
With
activation1"
NA
NA
NA
NA
NA
NA
NA
Dosec
lOOmg/kgby
corn oil gavage
400 mg/kg by
corn oil gavage
100 mg/kg by
oral gavage
100 mg/kg by
oral gavage
100 mg/kg by
corn oil gavage
160 mg/kg in
corn oil by i.p.
8,000 mg/kg
Reference
Mirsalis and
Butterworth, 1980
Mirsalis et al., 1982
Mirsalis, 1987;
Madleetal., 1994
Mirsalis, 1987;
Madleetal., 1994
Doolittle et al., 1987
Ikegwuonu and
Mehendale, 1991
Curtis and Tilley,
1968
85
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Rat (F344, male)
Rat (F344, male)
Rat (F344, male)
Rat (Wistar, male)
Rat (Wistar, male)
Mouse (CBAxC575BL/6,
male)
Mouse (B6C3F1, lad
transgenic; Big Blue™, male)
Endpoint
Chromosomal
aberrations in liver
Sister chromatic!
exchange in liver
Micronucleus
formation in liver
Micronucleus
formation in liver
Micronucleus
formation in liver
Micronucleus
formation and
ploidy levels in
liver
Mutations in lad
transgene in liver
Test conditions
Analyzed primary hepatocytes
cultured for 48 hrs from rats
sacrificed 0-72 hours after
dosing
Analyzed primary hepatocytes
cultured for 48 hrs from rats
sacrificed 0-72 hours after
dosing
Analyzed primary hepatocytes
cultured for 48 hrs from rats
sacrificed 0-72 hours after
dosing
Analyzed primary hepatocytes
harvested 72 hrs after dosing,
an optimal time to detect
micronuclei.
Analyzed primary hepatocytes
harvested 72 hrs after dosing,
an optimal time to detect
micronuclei.
Analyzed primary hepatocytes
from rats sacrificed 5 days
after dosing and compared
with a partially
hepatectomized control.
The target lad gene is
recovered from genomic DNA
after 5 daily doses and the
animals sacrificed 7 days after
the first dose
Results"
Without
activation
±(T)
+ (T)g
-(T)
With
activation1"
NA
NA
NA
NA
NA
NA
NA
Dosec
1,600 mg/kg
by corn oil
gavage
1,600 mg/kg
by corn oil
gavage
1,600 mg/kg
by corn oil
gavage
3,200 mg/kg
by gavage in
corn oil
3,200 mg/kg
by gavage in
corn oil
15 -Minute
inhalation at
0.05-0.1
mL/5 L
35 mg/kg-day
(5x)
Reference
Sawadaetal., 1991
Sawadaetal., 1991
Sawadaetal., 1991
Van Goethem et al.,
1993
Van Goethem et al.,
1995
Uryvaeva and
Delone, 1995
Mirsalis et al., 1994
86
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Mouse (CD2F1 lacz transgenic,
Mutamouse™, male)
Mouse (CD2F1 lacz transgenic,
Mutamouse™, male)
Rat (Wistar, male)
Mouse (Swiss, male)
Rat (Sprague-Dawley, male)
Mouse (A/J, male)
Mouse (A/J, male)
Endpoint
Mutations in the
lacz transgene in
liver
Mutations in the
lacz transgene in
liver
DNA binding in
liver
DNA binding in
liver
DNA binding in
liver
DNA binding in
liver
DNA binding in
liver
Test conditions
The target lacz gene is
recovered from genomic DNA
after a single dose with the
animals being sacrificed 14
days later
The target lacz gene is
recovered from genomic DNA
after dosing with the animals
being sacrificed 7, 14 or 28
days later
DNA extracted from liver of
rats (with or without
methylcholanthrene
pretreatment) sacrificed 12
hours after dosing
DNA extracted from liver of
mice (some pretreated with
methylcholanthrene)
sacrificed 12 hours after
dosing
DNA isolated from liver slices
of rats sacrificed 6 hours after
dosing
DNA isolated from liver slices
of mice sacrificed 6 hours
after dosing
DNA isolated from liver slices
of mice sacrificed 6 hours
after dosing
Results"
Without
activation
- (T)
-(T)
+h
±
±
+ (T)
With
activation1"
NA
NA
NA
NA
NA
NA
NA
Dosec
80 mg/kg by
gavage in corn
oil
1,400 mg/kg
by gavage
56 mg/kg i.p.
56 mg/kg i.p.
1.4 mg/kg i.p.
in olive oil
1.4 mg/kg i.p.
in olive oil
3,200 mg/kg
i.p. in olive oil
Reference
Tombolan et al.,
1999; Lambert etal.,
2005
Hachiya and
Motohashi, 2000;
Lambert etal., 2005
Rocchietal., 1973
Rocchietal., 1973
Diaz Gomez and
Castro, 1980a
Diaz Gomez and
Castro, 1980a
Diaz Gomez and
Castro, 1980a
87
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Rat (Sprague-Dawley, male)
Rat (Sprague-Dawley, male)
Mouse (C3H, male)
Hamster (Syrian golden, male)
Rat (strain and sex not
specified)
Rat (Sprague-Dawley, sex not
specified)
Rat (strain and sex not
specified)
Hamster (Syrian golden,
female)
Endpoint
DNA binding to
mitochondria and
nucleus
DNA binding in
liver
DNA binding in
liver
DNA binding in
liver
DNA adducts in
liver
DNA adducts in
liver
DNA adducts in
liver
DNA adducts in
liver and kidney
Test conditions
Mitochondrial DNA isolated
from the livers at 5 and 24 hrs
after dosing
DNA isolated from liver slices
of rats sacrificed 6 hours after
dosing
DNA isolated from liver slices
of mice sacrificed 6 hours
after dosing
DNA isolated from liver slices
of hamsters sacrificed 6 hours
after dosing
Deoxyguanosine-
malondialdehyde adducts
measured 48 hr after dosing
MidG adducts formed
secondary to lipid
peroxidation measured 4 days
after dosing
Deoxyguanosine-
malondialdehyde adducts
measured 48 hrs after dosing
13-HPO and
malondialdehyde-derived
adducts formed secondary to
lipid peroxidation detected by
[32P]-postlabelling analysis 4
hrs after treatment
Results"
Without
activation
+ (T)
±
±
±
+ (T)
+ (T)
"
±(T)
With
activation1"
NA
NA
NA
NA
NA
NA
NA
NA
Dosec
3.2 mg/kgin
corn oil
1,200 mg/kg
i.p. in olive oil
1,200 mg/kg
i.p. in olive oil
1,200 mg/kg
i.p. in olive oil
1,600 mg/kg
by gavage
0.1 mg/kg by
corn oil gavage
160 mg/kg by
oral gavage
160 mg/kg by
corn oil gavage
Reference
Levy and Brabec,
1984
Castro etal., 1989
Castro etal., 1989
Castro etal., 1989
Hadley and Draper,
1990
Chaudhary et al.,
1994
Draper etal., 1995
Wang and Liehr,
1995
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Rat (F344, male)
Rat (F344, female)
Rat (Fischer, male)
Rat (F344, male)
Rat (F344, male)
Mouse (CD-I, female)
Mouse (ICR, male)
Endpoint
DNA adducts in
liver
DNA adducts in
liver, kidney, lung,
colon, and
forestomach
DNA adducts in
liver
DNA adducts in
liver
DNA adducts in
urine
DNA binding in
liver
DNA binding in
liver
Test conditions
HNE-dG adducts formed
secondary to lipid
peroxidation
HNE-dG adducts formed
secondary to lipid
peroxidation. Samples
collected 4, 8, 16, or 24 hrs
after final dose.
8-OHdG adducts were
measured by
immunohistochemistry and
electrochemical detection at
times from 6 hrs to 7 days
8-OHdG adducts measured at
the end of week 1 after dosage
on days 1 and 4
8-OHdG adducts measured in
the urine 7 and 16 hr after a
single dose
8-oxodG measured in the
livers of 2- and 14-month
animals dosed for 3 days and
sacrificed on day 4.
[32P] -Postlabeling was used to
identify indigenous adducts
present 24 hrs after a single
injection
Results"
Without
activation
+ (T)
+ (T)
+ (T)
±(T)
+ (T)
+
+ (T)
With
activation1"
NA
NA
NA
NA
NA
NA
NA
Dosec
3,200 mg/kg
i.p. in olive oil
500 mg/kg i.p.
(Ior4x)
3,200 mg/kg
by gavage in
olive oil
400 mg/kg by
s.c. injection
(2x)
120 mg/kg by
i.p. injection
43 mg/kg i.p.
in mineral oil
1,200 mg/kg
by i.p. in corn
oil
Reference
Chung et al., 2000
Wackeretal., 2001
Takahashi et al.,
1998
Iwai et al., 2002
Kadiiska et al., 2005
Lopez-Diazguerrero
etal.,2005
Nathetal., 1990
89
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. Genotoxicity studies of carbon tetrachloride in mammalian systems in vivo
Test system
Mouse (ICR, male)
Rat (F344, male)
Rat (Wistar, male)
Endpoint
DNA binding in
liver
DNA methylation in
liver
DNA hypo-
methylation in liver
Test conditions
[32P] -Postlabeling was used to
identify indigenous and
exogenous adducts present 1,
4, and 8 weeks after two
injections given a week part.
Hydrolyzed DNA was
analyzed for aberrant
methylation as increases in 7-
methylguanine and O6-
methylguanine, 12 hrs after
dosing
The in vitro incorporation of
[3H] -methyl groups into
isolated hepatic DNA was
increased indicating that the
DNA was hypomethylated.
Results"
Without
activation
-(T)
+ (T)
+
With
activation1"
NA
NA
NA
Dosec
1,200 mg/kg
by i.p. in corn
oil
1,000 mg/kg in
corn oil
800 mg/kg by
i.p. injection
2X per week
for 3 weeks
Reference
Nathetal., 1990
Barrows and Shank,
1981
Varela-Moreiras et
al., 1995
a + = positive, ± = equivocal or weakly positive, - = negative, (T) = toxicity, ND = no data.
b Exogenous metabolic activation not applicable (NA) for these in vivo studies.
0 Lowest effective dose for positive results, highest dose tested for negative results, ND = no data, NA = not applicable.
d The small statistically significant increase detected was considered biologically insignificant by the authors (and other reviewers).
e At this dose a roughly threefold increase in micronucleus formation was seen along with a decrease in binucleated cells (about 35-50%) indicating a
cytostatic and cytotoxic effect.
f TUNEL - terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling
8 Increase was in both centromere-lacking (5.5-fold) and centromere-containing (3.6-fold) micronuclei.
h With methylcholanthrene pretreatment only.
8-OHdG = 8-hydroxy-2'-deoxyguanosine; 8-oxo-7,8-dihydro-2'-deoxyguanosine = 8-oxodG
Note: The data in the paper by Sarkar and associates (Sarkar et al., 1999) was judged to be insufficiently reliable to be included in the table.
90
DRAFT - DO NOT CITE OR QUOTE
-------
4.4.2.1. Genotoxicity Studies: Prokaryotic Organisms
As shown in Table 4-8, carbon tetrachloride was negative in most standard plate
incorporation assays for reverse mutation in Salmonella typhimurium, with or without addition of
a mammalian metabolic activation system (Brams et al., 1987; de Flora, 1981; McCann et al.,
1975). Increases in reversion frequency were reported by Varma et al. (1988), but the changes
were not dose-related. Varma et al. (1988) did not present data for the positive controls nor
discuss cytotoxicity, making it unclear how to interpret these data. Some S. typhimurium
reversion studies used modified testing techniques in order to account for the volatile nature of
carbon tetrachloride. Preincubation assays conducted in capped tubes were performed by
Uehleke et al. (1977) and Zeiger et al. (1988). Both of these research groups obtained negative
results. Gas-phase exposure studies have been conducted in various closed systems (Araki et al.,
2004; Barber et al., 1981; Simmon and Tardiff, 1978; Simmon et al., 1977). Results were
negative in most of these studies, although Araki et al. (2004) found a small increase in reversion
frequency in TA98 at concentrations of 1% (10,000 ppm) and above, when tested without
activation. It should be noted that the average control frequency of 13 revertants per plate in this
study is unusually low, and even the elevated response of 31 revertants per plate seen at the
50,000 ppm concentration is well within the range of spontaneous revertants typically seen in
TA98 controls (30-50 revertants per plate) (Maron and Ames, 1983).
In other studies using S. typhimurium, negative or equivocal results were reported for
carbon tetrachloride in a preincubation forward mutation assay using strains BA13 and BAL13
with and without metabolic activation (Roldan-Arjona and Pueyo, 1993; Roldan-Arjona et al.,
1991), and in an inducible DNA repair system (SOS) induction assay using strain
TA1535/pSK1002 (Nakamura et al., 1987). More varied results were seen in experiments using
E. coli. Carbon tetrachloride was negative in a SOS chromotest assay (Brams et al., 1987), a
spot test (De Flora et al., 1984), and a preincubation assay when evaluated for differential DNA
repair (Hellmer and Bolcsfoldi, 1992). In contrast, using E. coli strains that are more sensitive to
oxidative mutagens, increases in DNA repair were reported by De Flora et al. (1984) and
increases in reverse mutation were reported by Araki et al. (2004) and Norpoth et al. (1980). In
the DeFlora et al. (1984) study, carbon tetrachloride was more toxic to the E. coli strain CM871
(uvrA- recA- lexA-) than it was to the isogenic repair-proficient WP2 strain or WP67 (uvrA-
polA-). Although a similar pattern was seen in the presence of metabolic activation, carbon
tetrachloride was more active in the absence of activation. The differential toxicity was seen
initially using the liquid micromethod, and then confirmed using a 2-hour pre-incubation assay.
In the report of Araki et al. (2004), carbon tetrachloride produced a modest 2.5-fold increase in
mutations in the WP2uvrA/pKM101 strain of E. coli both in the presence and absence of
metabolic activation. The peak response was seen after 24 hours of exposure at a high
(20,000 ppm) concentration. The control frequencies were unusually low and the induced
response was within the control values reported by others (Martinez et al., 2000; Damment et al.,
91 DRAFT - DO NOT CITE OR QUOTE
-------
2005). Additionally, a statistically significant but well less than a twofold increase for E. coli
WP2wvrA was reported by Norpoth et al. (1980) at high levels (about 25,000 ppm) in another
gas-phase exposure study.
Carbon tetrachloride was also positive in the repair-proficient WP2/pKM101 strain of E.
coli. A doubling in mutant frequency was observed at the 5,000 ppm carbon tetrachloride
concentration and reached a fivefold increase compared to pooled controls at the 20,000 ppm
concentration. The increase was seen in experiments with and without metabolic activation as
well as with S9 plus reduced glutathione. Because the WP2 strains of E. coli have an AT base
pair at the critical mutation site within the trpE gene, they have been recommended for screening
oxidizing mutagens (Gatehouse et al., 1994; Martinez et al., 2000). This increased sensitivity to
oxidative damage may help explain both the Araki et al. (2004) and the DeFlores et al. (1984)
isolated positive results, although some aspects of the studies are still unusual. The greater
response in the repair-proficient strain seen in the Araki et al. (2004) study as compared to the
repair-deficient strain was unexpected, and led the authors to postulate that a cross-linking
metabolite might be responsible. If true, this could also be related to oxidative damage as lipid
peroxidation-derived products have been shown to form DNA and DNA-protein cross-links
(Kurtz and Lloyd, 2003; Niedernhofer et al., 2003). Again, the control frequencies reported by
Araki (2004) are lower than those reported by others (Watanabe et al., 1998), but in this case, the
induced mutant frequencies substantially exceed the control range of either group. Araki et al.
(2004) reported a 10-fold increase in mutants in the WP2/pKM101 experiments without S9.
However, approximately half of the observed increase was due to an unusually low mutant
frequency. Also, it should be noted that the results were not statistically analyzed as the
experiments were not performed in triplicate.
Some caution should be exercised in the interpretation of these and other in vitro studies
as a number of the factors listed in Table 4-12 could potentially influence the outcome of the
assays and contribute to both positive and negative results. For example, the bioactivation of
carbon tetrachloride to a mutagenic species can be affected in a variety of ways. The initial step
in the bioactivation of carbon tetrachloride is a CYP450 monooxygenase-mediated formation of
the trichloromethyl radical (Halliwell and Gutteridge, 1999; Weber et al., 2003). This radical is
highly reactive, and as a result, may not be able to cross the bacterial cell wall or membranes to
access the bacterial DNA. The trichloromethyl radical or a derived species can also react with
and inactivate the monooxygenase activation system (Weber et al., 2003), which could also
affect the outcome of the in vitro assays. In addition, many of the commonly used vehicle
solvents used for in vitro testing such as methanol, DMSO, and ethanol are also metabolized by
the cytochrome CYP450 2E1 isoform CYP2E1 (Hyland et al., 1992), the isoform primarily
involved in carbon tetrachloride metabolism, and may have interfered with the bioactivation of
carbon tetrachloride in these test systems. In addition, DMSO can act as a free radical scavenger
(Halliwell and Gutteridge, 1999).
92 DRAFT - DO NOT CITE OR QUOTE
-------
Similarly, when standard inducing procedures (Arochlor 1254 or the combination of
phenobarbitone and &eta-naphthoflavone) have been used, the levels of CYP2E1 in the rat liver
are markedly suppressed (Burke et al., 1994). This would lead to a decrease in CYP2E1 in the
S9 used for the test and could potentially contribute to the observed negative results.
Furthermore, although carbon tetrachloride has been evaluated many times in the standard
Salmonella test strains, it has not been tested in either TA102 or TA104 and only a few times in
the E. coli WP2 strains, the strains that would be the most sensitive to the oxidative DNA
damage likely to be generated during carbon tetrachloride toxicity. Because of the many
possible confounding factors, the in vitro carbon tetrachloride results should be interpreted
cautiously.
Table 4-12. Challenges in evaluating carbon tetrachloride genotoxicity
Large number of genotoxicity studies
Elevated error rates related to multiple statistical tests and comparisons
Requirement to test to high levels of toxicity to ensure a true negative response
Non-specific effects that can occur at very high chemical concentrations
Potential volatility from culture media
Requirement for metabolic activation
Downregulation of CYP2E1 synthesis shortly after carbon tetrachloride administration
Inhibition of cytochrome CYP450 monoxygenases by primary carbon tetrachloride metabolite(s)
Competitive inhibition of CYP2E1 by common solvents used as vehicles (ethanol, methanol, DMSO)
Free radical-scavenging properties of common vehicles such as DMSO
Possible inability of reactive trichloromethyl radical generated extracellularly by rat postmitochondrial
supernatant to cross the bacterial cell wall or eukaryotic cell membrane and damage the DNA of the
cell being tested
Commonly used enzyme inducers suppress CYP2E1 levels in the rat liver S9
Possible influence of dosing vehicle (corn oil, olive oil) in vivo
Concurrence of cytotoxicity and genotoxicity
Occurrence of DNA breakage during apoptotic and necrotic cell death
Occurrence of multiple reactive species and potential mechanisms of genotoxicity
Difficulties in distinguishing direct and indirect genotoxic effects
Generation of genotoxic products secondary to lipid peroxidation
Genotoxic responses occurring secondary to inflammatory responses
4.4.2.2. Genotoxicity Studies: Non-Mammalian Eukaryotic Organisms
Carbon tetrachloride has also been tested in the yeast Saccharomyces cerevisiae and the
mold Aspergillus nidulans (Table 4-9). In contrast to the bacterial results, the majority of the
studies conducted in these species have yielded positive results. However, the results obtained
from the two fungal species differ significantly, most likely due to the test strains selected and
the endpoints chosen for examination. In initial studies by Callen et al. (1980), carbon
tetrachloride induced >20-fold increases in gene conversion and mitotic conversion and a 2.5-
fold increase in reverse mutations when tested at high concentrations in the yeast D7 strain in a
preincubation assay employing capped tubes. The increases were only seen at the highest test
93 DRAFT - DO NOT CITE OR QUOTE
-------
concentration of 34 mM, one that caused extensive toxicity (90%). These initial results were
followed by a series of studies by Schiestl and co-workers using yeast strains that were designed
to detect intrachromosomal recombination (DEL assay) that results from double stranded DNA
breakage. Interchromosomal recombination can also be measured in these strains. In the initial
study using the DEL assay (Schiestl et al., 1989), carbon tetrachloride at a concentration of 8,000
|ig/mL, induced a 25-fold increase in intrachromosomal recombinants with no increase in
interchromosomal recombination. Toxicity was >99% at the highest test concentration where the
increase in recombinants was seen. Follow-up studies showed that the induced recombinants
occurred during the Gl and G2, but not S phase of the cell cycle, and in some cases an increase
in interchromosomal recombination was also seen. The dose-response curves tended to be steep
and occurred concurrently with significant toxicity (Galli and Schiestl, 1996; Galli and Schiestl,
1995). Since carbon tetrachloride did not induce recombination during S phase even though it
was toxic, the authors suggested that carbon tetrachloride acted by prematurely pushing Gl cells
into S phase and G2 cells into cell division (Galli and Schiestl, 1998). The inability to
completely repair damaged DNA prior to replication or cell division might result in DNA strand
breakage and subsequent recombination. Brennan and Schiestl (1998) showed that yeast cells
treated with carbon tetrachloride showed an increase in oxidative radical species as measured by
the intracellular oxidation of 2,7-dichlorofluorescein diacetate. N-acetylcysteine did not exhibit
a protective effect on carbon tetrachloride-induced DEL recombination, although the results are
difficult to interpret as increased toxicity was seen in cells jointly treated with carbon
tetrachloride and this sulfhydryl-containing agent.
In contrast to the recombinogenic effects seen with S. cerevisiae, the assays using A.
nidulans primarily detected an abnormal segregation of chromosomes. Following treatment with
high concentrations (0.5%) of carbon tetrachloride, Gualandi (1984) observed a significant ^20-
fold) increase in abnormal chromosome segregation and an approximately 2.5-fold increase in
forward mutations. Toxicity at the test concentration was approximately 70%. Additional
studies showed a strong correlation between toxicity and altered segregation leading to aneuploid
cells. Cysteamine (a free-radical scavenger) was also co-administered with carbon tetrachloride
and showed some protection against the induced alterations in chromosome segregation. In a
series of related studies, carbon tetrachloride was consistently shown to interfere with
chromosome segregation leading to aneuploidy. Crebelli et al. (1988) demonstrated that carbon
tetrachloride induced a 10-fold increase in chromosome segregation at the highest (0.08%)
concentration tested. Toxicity at this concentration was 72%. More modest effects
(approximately threefold) were seen beginning at lower concentrations (0.04%) that were less
toxic (18%). Notably, no increase in crossing over was seen in these experiments. Similar
results both on chromosome segregation and crossing over were observed in a follow-up study
using a narrower and somewhat lower dose range (0.01-0.03%; Benigni et al., 1993). In a
related quantitative structure-activity-relationship study of carbon tetrachloride and 23 other
94 DRAFT - DO NOT CITE OR QUOTE
-------
chlorinated aliphatic hydrocarbons, the ease at which the compounds were able to accept
electrons, as characterized by the energy of lowest unoccupied molecular orbital, was the best
predictor of their aneuploidy-inducing properties (Crebelli et al., 1992).
As indicated in Table 4-9, the genotoxic effects were seen in both Saccharomyces and
Aspergillus experiments without the use of exogenous metabolic activation. This is consistent
with studies that have shown actively growing cells of both species contain cytochrome CYP450
monooxygenase enzymes capable of bioactivating promutagens to mutagens (Bignami et al.,
1981; Callen et al., 1980). As indicated above, the studies in Saccharomyces detected primarily
recombination, whereas those in Aspergillus detected primarily alterations in chromosome
segregation. This difference in outcome appears to be due primarily to the nature of the specific
strains used and the endpoints selected for evaluation by the investigators. There was a close
association seen between cytotoxicity and the recombinogenic and aneugenic effects measured in
the two systems.
Additionally, carbon tetrachloride did not produce sex-linked recessive lethal mutations
in Drosophila melanogaster (Foureman et al., 1994).
4.4.2.3. Genotoxicity Studies: Mammalian Cells In Vitro
Numerous studies have been performed to evaluate the ability of carbon tetrachloride to
cause genotoxic effects or precursor lesions in mammalian cells in vitro (Table 4-10). These
studies have been performed using both model cell systems frequently with exogenous metabolic
activation and hepatocytes that retain their xenobiotic-metabolizing capabilities.
Studies in non-target mammalian cells. In studies using peripheral blood lymphocytes
or lymphoblastoid cells, carbon tetrachloride yielded mixed results. As part of a study of
fumigants, Garry et al. (1990) exposed G0 lymphocytes to carbon tetrachloride for 30 minutes,
then cultured the lymphocytes and measured the frequencies of chromosome aberrations and
sister chromatid exchanges (SCEs). No increases in structural aberrations or SCEs were seen.
Tafazoli et al. (1998) used the micronucleus (MN) assay to measure chromosome loss or
breakage in the peripheral lymphocytes obtained from two donors. Exposure to different
concentrations of carbon tetrachloride ranging from 1 to 40 mM did not induce a statistically
significant increase in micrenucleated cells at any concentrations except at 10 mM in one donor
with S9 mix and at 5 mM in the second donor without S9 mix. Cell division was not affected at
these mutagenic concentrations; however, the authors identified a cytotoxic concentration of 40
mM both with and without S9 mix in one donor. To measure the amount of DNA strand breaks,
Tafazoli et al. used the in vitro Comet assay with isolated lymphocytes from the donors. No
statistically significant response was found for either tail length or tail moment at concentrations
tested (5-20 mM) either with or without S9 mix. Carbon tetrachloride was also reported to be
negative when assayed for unscheduled DNA synthesis (UDS) in lymphocytes (Perocco and
95 DRAFT - DO NOT CITE OR QUOTE
-------
Prodi, 1981). Each of these studies either used high carbon tetrachloride concentrations (>1500
|ig/mL) or tested to toxic concentrations.
In contrast, when tested at relatively low concentrations, Sivikova et al. (2001) reported
that cultured ovine peripheral lymphocytes exposed to carbon tetrachloride exhibited twofold
increases in micronuclei in both the absence and presence of S9, and an approximately 25%
increase in SCEs in the absence of S9. Under similar conditions, no increase in structural
chromosome aberrations was seen, although a decrease in the mitotic index was detected.
Interestingly for both the MN and SCE experiments, the addition of vitamin E and selenium to
the cultures protected against the increases in MN and SCE, implicating a role for free radicals in
the observed genotoxic effects. In spite of the protective effects of the antioxidants, these studies
observed effects at fairly low concentrations and the greater activity in the absence of S9.
Doherty et al. (1996) reported that carbon tetrachloride induced micronuclei in two
human lymphoblastoid cell lines—one expressing CYP2E1 (h2El) and the other expressing
CYP1A2, 2A6, 3A4, and 2E1 and microsomal epoxide hydrolase (MCL-5)—but not the
CYP1 Al-expressing AHH-1 cell line. Treatment of the cells with 10 mM carbon tetrachloride
resulted in a five- and a ninefold increase in micronucleated cells in the h2El and the MCL-5
cell lines, respectively. The increases occurred mostly in kinetochore-positive micronuclei,
indicating an origin from chromosome loss. Smaller increases (-two- to fourfold) in micronuclei
originating from chromosomal breakage (kinetochore-negative) were also seen. At the 10 mM
concentration, the percentage of binucleated cells, an indicator of cell proliferation and an
indirect indicator of cytotoxicity, was 6 - 7% of the control values indicating that the increase in
micronuclei occurred primarily under conditions producing potent cytotoxic or cytostatic effects.
In other studies involving non-target cell culture systems, carbon tetrachloride was
negative for inducing structural chromosome aberrations and SCEs in Chinese hamster ovary
(CHO) cells (Loveday et al., 1990). However, in a number of other assays using CHO and V79
cells, carbon tetrachloride in the absence of exogenous activation, was reported to produce
modest increases in c-mitoses, generate multipolar spindles and lagging chromosomes during
anaphase, and interfere with chromosome segregation resulting in aneuploidy (Onfelt, 1987;
Coutino, 1979).
Carbon tetrachloride was also tested for its ability to induce morphological
transformation in Syrian hamster embryo cells (Amacher and Zelljadt, 1983). In the
transformation assay, carbon tetrachloride was tested in both RPMI 1,640 media with horse
serum and DMEM with fetal bovine serum. It was negative in the RPMI medium with 0
transformants among 2,665 colonies. In DMEM, one transformed colony was seen in 2,003
colonies scored. Although this was considered a positive result by the authors, the increase is not
statistically significant, does not meet criteria for a positive result (Kerckaert et al., 1996), and
falls within the normal control frequencies of 0-0.8% reported for this type of transformation
assay (LeBoeuf et al., 1996).
96 DRAFT - DO NOT CITE OR QUOTE
-------
In studies using mouse lymphoma (L5178Y) cells with exogenous activation, carbon
tetrachloride was inactive in inducing mutations at the tk locus when tested up to toxic
concentrations (Wangenheim and Bolcsfoldi, 1988). In a follow-up study employing similar
cells and conditions, DNA strand breaks were induced as measured by the alkaline elution assay.
The increases in strand breaks were accompanied by increases in cytotoxicity (Garberg et al.,
1988).
Studies in liver cells. Carbon tetrachloride has also exhibited mixed results when tested
in vitro using isolated hepatocytes or cell lines derived from the rat liver. In early studies by
Dean and Hodgson-Walker, carbon tetrachloride was negative for inducing structural
chromosome aberrations or SCEs when tested at a low concentration in a metabolically
competent rat liver cell line (Dean and Hodson-Walker, 1979). Similarly, no increase in UDS
was seen by Selden et al. (1994) in their studies using rat hepatocytes or by Butterworth et al.
(1989) in their UDS studies employing primary hepatocyte cultures from four human donors. In
contrast, using an alkaline elution assay on isolated rat hepatocytes, Sina and colleagues reported
a 3.1- to 5.0-fold increase in strand breaks at the highest concentration tested (3 mM), a dose that
also resulted in approximately 50-60% toxicity (Sina et al., 1983). A modest dose-related
increase in DNA strand breaks was also seen in the single cell gel electrophoresis (Comet) assay
by Beddowes et al. (2003). The increase in breaks reported by Beddowes was accompanied by
similar increases in the formation of the oxidative DNA adducts, 8-oxodeoxyguanosine and a
malondialdehyde (MDA) deoxyguanosine adduct.
The ability of bioactivated carbon tetrachloride to react directly with DNA has been
investigated by a number of investigators using isolated DNA and nuclear preparations obtained
from hepatocytes. Initial studies by Rocchi and colleagues demonstrated that when radiolabeled
carbon tetrachloride was incubated with microsomes from uninduced and 3-methylcholanthrene-
induced mice and rats, modest increases in radiolabel were recovered following extensive
washing and extraction of the DNA with several solvents (Rocchi et al., 1973). This binding was
greater in the incubations containing the 3-methylcholanthrene-induced microsomes. Similarly
DiRenzo et al. (1982) reported that significant binding of carbon tetrachloride to DNA (0.39
nmol/mg DNA) occurred following the incubation of radiolabeled carbon tetrachloride with
pronase-pretreated calf thymus DNA and microsomes from phenobarbital-induced rats. The
incubation was performed under a N2 atmosphere using conditions that, in previous studies had
resulted in maximal binding to proteins and lipids. Oruambo and Van Duuren (1987)
investigated the binding of radiolabeled carbon tetrachloride to various regions of mouse
chromatin. Following a 2-hour incubation with mouse hepatic microsomes, hepatic chromatin,
and radiolabeled carbon tetrachloride, the authors concluded that the carbon tetrachloride
metabolite(s) bound equally to both DNase I-sensitive and -resistant regions. After 4 hours of
incubation, more radiolabel was recovered associated with DNase I-resistant DNA than with
97 DRAFT - DO NOT CITE OR QUOTE
-------
DNase I-sensitive DNA. This preferential binding to transcriptionally inactive (DNase I-
resistant) sites in chromatin was seen as unique among carcinogens, and could be attributable to
changes in chromatin conformation or differential DNA repair. In addition, Castro et al. (1989)
investigated the ability of radiolabeled carbon tetrachloride to bind to the DNA of purified
nuclear preparations obtained from the livers of Sprague-Dawley rats, a strain resistant to carbon
tetrachloride carcinogen!city, and C3H mice and Syrian golden hamsters, two strains that are
sensitive to carbon tetrachloride hepatocarcinogenesis. Low levels of binding were observed,
which were increased in the mouse and hamster incubations when NADPH was included in the
microsomal incubation. The authors noted that there was no correlation between sensitivity to
carbon tetrachloride carcinogenesis (hamster > mouse » rat) and the binding of carbon
tetrachloride metabolites to DNA, either in vitro or in vivo (in vivo: hamster = mouse = rat; in
vitro with NADPH: hamster = mouse = rat; in vitro without NADPH: rat > mouse = hamster).
Overall, these data indicate that under certain conditions, carbon tetrachloride can induce
genotoxic effects in mammalian cells exposed in vitro. Although numerous negative studies
were seen, there are indications from multiple studies that at high doses, bioactivated carbon
tetrachloride is able to cause DNA breaks leading, in some cases, to chromosome breakage.
There are also multiple studies indicating that carbon tetrachloride is able to interfere with
chromosome segregation resulting in modest levels of chromosome loss and aneuploidy.
However, since exogenous bioactivation was required in some studies and not others, the
observed effects may result from both specific and non-specific mechanisms. The binding
studies using radiolabeled carbon tetrachloride (for discussion, see the following sections)
provide limited evidence that bioactivated carbon tetrachloride can bind directly to DNA. As
seen in non-mammalian assay systems, in most cases where genotoxic effects were observed,
they occurred concurrently with significant cytotoxicity.
4.4.2.4. Genotoxicity Studies: Mammalian Cells In Vivo
Carbon tetrachloride has been extensively tested for genotoxicity in mammalian systems
in vivo (Table 4-11). A number of these studies have been conducted using standard protocols
and examined genotoxicity in highly proliferating non-target organs such as the bone marrow. In
addition, a large number of studies have examined genotoxic effects or precursor lesions such as
DNA adducts occurring in the rodent liver. A summary of the important studies by target organ
and endpoint is presented below.
Chromosomal alterations and DNA breakage in non-target organs. In studies of
chromosomal alterations occurring in the bone marrow, carbon tetrachloride has shown negative
results for the induction of structural chromosome aberrations in the bone marrow of male
Sprague-Dawley rats and 101/H mice (Rossi et al., 1988; Lil'p, 1982), as well as for the
formation of micronuclei in the bone marrow and peripheral blood erythrocytes of male BDF1
98 DRAFT - DO NOT CITE OR QUOTE
-------
mice (Suzuki et al., 1997; Morita et al., 1997). Negative results were also seen for the induction
of micronucleated erythrocytes in the bone marrow and peripheral blood of both male and female
CD-I mice (Crebelli et al., 1999). In the Comet assay, no evidence of DNA breakage was seen
in the nucleated cells of the stomach, kidney, bladder, lung, brain, or bone marrow of male CD-I
mice administered 2000 mg/kg carbon tetrachloride with sampling at 0, 3, and 24 hours after
dosing (Sasaki et al., 1998). In these same animals, significant increases in DNA breakage were
seen in the liver, although this was considered by the authors to be a false positive result because
it was accompanied by evidence of necrosis in the liver. In a biomarker study, carbon
tetrachloride was also reported to induce an isolated significant increase in DNA breakage in the
Comet assay in nucleated peripheral blood cells of male F344 rats (Kadiiska et al., 2005). The
increase is of questionable relevance as it was only seen at one of the three time points tested and
only at the lower of the two doses tested.
DNA breakage in rodent liver cells. Within the rodent liver, carbon tetrachloride has
been evaluated for a range of genotoxic effects across a considerable dose range. Fourteen
studies employed the alkaline elution or similar method to determine if carbon tetrachloride is
able to induce DNA breaks in liver cells in vivo. Negative results were seen in eight of the
studies, equivocal or weak responses were seen in two studies, and positive results were seen in
four studies. When positive or equivocal responses were seen, they consistently occurred at
cytotoxic doses. A brief overview of each of the positive studies is provided below.
Nakamura and Hotchi (1992) observed an increase in DNA breakage in their studies of
DNA breakage in non-parenchymal cells. The DNA breaks were identified using an in situ nick
translation approach at time points ranging from 12 hours to 18 weeks after dosing. Although
breaks were seen, the authors argued that the breaks were most likely physiological in nature,
reflecting changes in proliferation and/or gene expression. In another series of experiments
involving the adaptation of the liver to long-term continuous carbon tetrachloride administration
to mice, Gans and Korson (1984) noted changes in the DNA synthesis of the liver nuclear DNA.
As one aspect of the study, the authors used an alkaline elution approach to study DNA damage
in the liver of CD-I mice. A maximal increase in DNA damage was seen 18 hours after
administration. The normal pattern of sedimentation was restored by 24-36 hours. The authors
stated that "these changes were observed only following doses of carbon tetrachloride which
resulted in liver necrosis. Doses of carbon tetrachloride which did not produce necrosis did not
result in a shift in the sedimentation of DNA."
Similarly, Cabre and associates detected DNA breaks in rats treated with two high doses
of carbon tetrachloride using the terminal deoxynucleotidyl transferase-mediated deoxyuridine
triphosphate nick end labeling (TUNEL) technique (Cabre et al., 1999). The TUNEL assay is
commonly used to measure DNA strand breaks occurring in apoptotic cells but also detects
breaks occurring in necrotic cells (Higami et al., 2004). Similarly, Yasuda and colleagues used
99 DRAFT - DO NOT CITE OR QUOTE
-------
the TUNEL assay to study necrotic cell death induced by carbon tetrachloride and
dimethylnitrosamine (Yasuda et al., 2000). In the Yasuda studies of carbon tetrachloride-treated
livers, TUNEL staining was closely associated with the release of lysosomal enzymes into the
cytoplasm, and an intranuclear localization of lysosomal enzymes occurred at an early stage of
subcellular damage. This pattern was notably different from that seen with the alkylating agent,
dimethylnitrosamine. Given the high doses administered and the known hepatotoxicity of
carbon tetrachloride, the observed detection of DNA strand breaks in these and the other studies
is not surprising. As mentioned earlier and for the same reason, Sasaki et al. (1998) considered
the DNA strand breaks that they observed using the Comet assay to be false positives and not
relevant to assessing genotoxic potential since evidence of necrosis was present.
VDS in the rodent liver. A number of studies have been performed to investigate the
ability of carbon tetrachloride to induce UDS in the liver of rats and mice treated in vivo. In an
initial study of de novo and repair replication of DNA in the livers of treated rats, Craddock and
Henderson (1978) reported that oral administration of 4000 mg/kg carbon tetrachloride increased
the synthesis of DNA in non-replicating hydroxyurea-treated hepatocytes 17 hours, but not 2
hours, after treatment. In the absence of the hydroxyurea treatment, extensive DNA synthesis
was seen at the 17-hour time point. Diethylnitrosamine, ethyl ethanesulfonate, aflatoxin, and
retrosine induced DNA repair replication at the earlier 2-hour sampling. The delay seen with
carbon tetrachloride was suggested by the authors as indicating that the repair was associated
with damage caused by an indirect mechanism such as deoxyribonuclease activity resulting from
lysosomal damage; however, the extensive DNA synthesis occurring at the 17-hour time point is
almost certainly due to proliferation following extensive cell death induced by carbon
tetrachloride. Under these conditions, it is not clear how efficient the hydroxyurea inhibition of
DNA synthesis would be. In a more recent study using the hydroxyurea approach, Ikegwuonu
and Mehendale (1991) saw similar results, although they saw no increase in DNA breakage
using an alkaline elution technique in a parallel study. The observations of DNA repair in the
absence of detectable DNA breaks are inconsistent and the authors concluded that the
hydroxyurea repair results were attributable to induced de novo synthesis (post replication repair)
rather than true DNA repair. It should also be noted that the use of the hydroxyurea method to
measure UDS is generally not recommended because of the complex effects of hydroxyurea in
the cell and its ability to directly induce UDS (for additional details, see Madle et al., 1994).
In six other studies conducted using the currently recommended autographic detection
method, no increase in UDS induced by carbon tetrachloride was seen even at doses exhibiting
significant toxicity. With the autographic method, DNA uptake is measured in individual cells
allowing UDS to be clearly distinguished from de novo synthesis.
To summarize the UDS results, eight in vivo studies have been performed investigating
UDS in the rodent liver following carbon tetrachloride administration. Two major methods for
100 DRAFT - DO NOT CITE OR QUOTE
-------
measuring UDS were employed, the autographic method that allows UDS in individual cells to
be measured and that is considered to be more reliable, and a less reliable method that measures
DNA synthesis in the presence of hydroxyurea, an inhibitor of global de novo DNA synthesis.
The six studies that used the autoradiographic method yielded negative results.
Chromosome aberrations and micronuclei in rodent liver cells. In cytogenetic assays of
hepatocytes isolated from treated rodents, carbon tetrachloride produced mixed, largely negative
results. In an early study by Curtis and Tiley (1968), no increase in chromosomal fragments or
bridges occurring in anaphase cells was seen in liver squash preparations of mice treated with a
high (8,000 mg/kg) dose of carbon tetrachloride. Similar negative results for structural
chromosome aberrations, SCEs and micronuclei were reported at all time points in time course
studies conducted by Sawada et al. (1991). Negative results were also reported for MN
formation and altered ploidy by Uryvaeva and Del one (1995).
In two studies conducted by Van Goethem and colleagues, however, an increase in
micronuclei was reported. In their initial study investigating the early stages of hepatic
carcinogenesis (Van Goethem et al., 1993), carbon tetrachloride was administered to male Wistar
rats at 3200 mg/kg and the frequency of micronuclei was measured in hepatocytes harvested 72
hours later. Initial studies of the mitotic index and the percent binucleated cells indicated that 72
hours was the optimal time to harvest hepatocytes for the detection of micronuclei. High intra-
animal variability was seen, but the results suggested that the hepatocytes of the carbon
tetrachloride- (and CT+NaCl-) treated mice exhibited an increase in micronuclei (1.7-7.2%) as
compared to those of control (and NaCl-treated) mice (0.2-1%). In a follow-up study, Van
Goethem and associates repeated portions of their earlier experiment (Van Goethem et al., 1995).
Three animals received carbon tetrachloride and three served as controls. The frequency of
micrenucleated hepatocytes increased from 1.5% in the controls to 7.6% in the carbon
tetrachloride-treated rats, a significant fivefold difference. Using fluorescence in situ
hybridization with a multi-centromeric rat probe, the authors attributed the increase in MN
primarily to chromosomal breakage. Based on the frequencies given in the paper, chromosome
breakage can be calculated to be 5.5-fold over the control, whereas chromosome loss can be
calculated as a 3.5-fold increase. It should be noted that the observed difference in the
proportion of centromere-containing and -lacking micronuclei in the study is attributable to a low
frequency of centromere-containing micronuclei in only one rat and is unlikely to be either
statistically or biologically significant. Based on their work and that of others (Craddock and
Henderson, 1978), the authors attributed the results to chemically-induced oxidative cellular
damage, and suggested that free radicals produced from carbon tetrachloride may disrupt
cytoplasmic organelles releasing DNase and tissue-destructive hydrolases within the cell leading
to DNA strand breaks and tissue damage. Although the sample sizes of the studies are quite
small, the two studies indicate that the MN results are reproducible and that under regenerative
101 DRAFT - DO NOT CITE OR QUOTE
-------
conditions following toxicity, an increase in chromosome breakage and possibly chromosome
loss can be detected in the regenerating cells of carbon tetrachloride-treated rats.
Sarkar et al. (1999) reported that the administration of carbon tetrachloride to mice over a
5-week period resulted in increases in structural chromosome aberrations in liver cells.
However, there are numerous and significant methodological issues with these experiments. For
example, the methods section does not adequately explain how metaphases were obtained from
either the treated or control mice that would allow structural chromosome aberrations to be
scored. Given the low number of mitotic cells in the untreated mouse liver, it would be difficult
if not impossible without mitotic stimulation to obtain 50 well-spread metaphases without the
use of colchicine or other spindle-disrupting agent. In addition, the reported frequencies of
structural aberrations including some classes of aberration, such as ring chromosomes, are
unusually high (32-48% including gaps) when compared to other studies. Because of these
concerns, this paper has not been included in Table 4-11.
Mutations in transgenic mice. The ability of carbon tetrachloride to induce mutations in
hepatocytes in vivo has been investigated in three studies using transgenic mice. The transgenic
mouse models used to evaluate carbon tetrachloride (lac I, B6C3F1; lacz CD2F1 Muta™Mice)
represent normal immunocompetent rodent strains with the addition of reporter genes for
identification of mutational events. Negative results were seen in each of the three studies. As
reported by Mirsalis and coworkers, transgenic B6C3F1 lacl mice were treated with 5 daily
doses of carbon tetrachloride at 35 mg/kg-day and the animals were sacrificed 7 days after the
first dose (Mirsalis et al., 1994; Mirsalis, 1995). Mice were implanted with an osmotic pump
that released [3H]thymidine at the beginning of the study to measure the percent of hepatocytes
in S phase (labeling index). Controls had a labeling index of 0.07% and a mutant frequency of
<6 x 10"5. Carbon tetrachloride produced a nearly 1000-fold increase in the labeling index with
no increase in the mutant frequency. The authors concluded that short bursts of cell proliferation
induced by carbon tetrachloride do not result in mutations in the liver.
As part of another study to investigate the impact of cell proliferation on liver
mutagenesis, carbon tetrachloride was administered at 80 mg/kg by i.p. injection to lacz
transgenic CD2F1 mice (Muta™Mice) and the animals were sacrificed 14 days later (Tombolan
et al., 1999; Lambert et al., 2005). The mutant frequency in the carbon tetrachlori de-treated
animals (8.6 x 10"5) was not significantly increased over that seen in the controls (5.4 x 10"5). In
non-transgenic CD2F1 mice receiving an intragastric dose of carbon tetrachloride, significant
increases in absolute and relative liver weights were seen beginning 2 days after treatment. The
percent of hepatocytes labeling with BrdU during the last 2 hours before sacrifice peaked at 59
times that of the controls at 3 days after treatment and returned to control levels by day 7.
In the third study reported by Hachiya and Motohashi (2000), the frequency of mutations
the lacZ transgene in liver of male CD2F1 lacZ transgenic mice (Muta™Mice) was determined
102 DRAFT - DO NOT CITE OR QUOTE
-------
14 days after administration of 700 mg/kg carbon tetrachloride (by oral gavage) or 7, 14, or 28
days after administration of 1,400 mg/kg. A small increase in mutant frequency, considered
biologically insignificant by the authors, was seen. The mutant frequencies for six of the nine
carbon tetrachloride-treated animals were within the control range (53 x 10"6-100.4 x 10"6). The
mutant frequencies for the other three mice exceeded the upper end of the control range by 3 to
49%. The results as analyzed by Fishers exact test were statistically significant in part because
of the large number of plaques evaluated and the fact that the Fisher's exact test does not account
for animal-to-animal variability. The authors concluded that no biologically significant increase
in the mutant frequency was seen in the carbon tetrachloride-treated mice. Other reviewers have
concurred with this conclusion (Lambert et al., 2005).
As indicated in Heddle et al. (2000), a commonly used cut-off value for a positive
response in this type of transgenic assay is at least a twofold increase over the historical negative
control mutant frequency. Although a historical control range for the Hayashi and Motohashi lab
was not presented, the range for the concurrent controls was 5.3 x 10"5 to 10 x 10"5 with a mean
of 8.2 x 10" . For comparison, a general control range suggested by Heddle et al. (2000) used for
sample size calculations is 4 x 10" to 7 x 10" . Using this as a historical control, no treatment
group exceeded twofold that of the control and only one treated animal in the study was outside
of this range. As a caveat, the numbers of animals used in the three studies were small, and the
dosing and sampling protocols did not follow those currently recommended (Heddle et al., 2000;
Lambert et al., 2005). However, the results of these three in vivo studies are consistent and
provide no evidence for the formation of carbon tetrachloride-induced mutations in the liver
following acutely toxic doses.
DNA binding by carbon tetrachloride-derived metabolites. A number of studies have
investigated the potential of carbon tetrachloride to bind covalently to DNA. Additional studies
have investigated whether DNA adducts derived from reactive oxygen species or from lipid
peroxidation-derived products are elevated following carbon tetrachloride administration. DNA
adducts from both pathways have been reported in carbon tetrachloride-treated mice, rats, and
hamsters.
In initial studies, Rocchi et al. (1973) investigated the ability of [14C]-labeled carbon
tetrachloride to bind to the DNA, RNA and proteins in the liver of male Wistar rats and male
Swiss mice. Carbon tetrachloride was injected i.p. at 56 mg/kg and the animals were sacrificed
12 hours later and the livers from the treatment groups were pooled. Half of the animals had
been previously treated with 3-methylcholanthrene to induce hepatic metabolism.
Radiochemical binding to nuclear and cytoplasmic proteins but not DNA was seen in the 3-
methylcholanthrene-pretreated and non-pretreated rats. Binding to rRNA was also seen in the 3-
methylcholanthrene-pretreated rats. In the mouse studies, DNA binding was seen in the livers of
mice pretreated with 3-methylcholanthrene, but not in mice not previously pretreated. Protein
103 DRAFT - DO NOT CITE OR QUOTE
-------
binding was seen in both groups of mice. Since the livers of the treatment groups were pooled
for analysis, no measure of variability or statistical significance could be established. In
addition, although the article mentions that the counts per minute (cpm) of the samples was at
least twice that of the background, there is no mention of controls nor information on how the
samples were corrected for radioactivity in the control samples.
Diaz Gomez and Castro (1980a) also studied the ability of [14C]-labeled carbon
tetrachloride to bind to DNA, nuclear proteins and nuclear lipids in the liver of male Sprague
Dawley rats and male Strain A/J mice. Carbon tetrachloride was injected i.p. at 1.4 mg/kg, and
the animals were sacrificed 16 hours later. Three samples, each comprised of one rat liver or the
pooled livers from 10 mice, were measured per experimental group. A small but significant
increase in radiocarbon binding was seen in both the mouse and rat samples in this experiment.
Binding to nuclear proteins and lipids was also seen in parallel experiments. In another series of
experiments, mice previously treated with phenobarbital or 3-methylcholanthrene to induce
hepatic metabolism were administered carbon tetrachloride at 1.4 mg/kg. Another group was
administered a higher (3,200 mg/kg) toxic carbon tetrachloride dose. Radiochemical binding to
mouse liver DNA was reported for the phenobarbital and 3-methylcholanthrene-pretreated mice
as well as for the mice treated with the toxic carbon tetrachloride dose. DNA binding was
slightly increased in the 3-methylcholanthrene-pretreated mice (0.84 pmol/mg) and the high-dose
mice (2.803 pmol/mg) as compared to the low-dose carbon tetrachloride-treated mice (0.72
pmol/mg). The levels of low-dose carbon tetrachloride binding to DNA were considered to be
quite low in both species with the binding in the mouse liver slightly higher than that in the rat
liver. Negative control information was not presented. In place of a true negative control, the
background radioactivity counted in the presence of DNA of 78 disintegrations per minute
(dpm). This was approximately double the background of 38 detected in the absence of DNA
and was deducted from each experimental determination.
In a follow-up study, Castro et al. (1989) investigated the relationship between the
intensities of covalent binding to liver DNA and nuclear proteins in vivo in samples obtained
from C3H mice, Syrian golden hamsters, and Sprague-Dawley rats—three species with different
susceptibilities to carbon tetrachloride-induced liver cancer—administered 1,200 mg/kg
radiolabeled carbon tetrachloride ([14C]CCl4). The authors reported that there was no correlation
between the intensity of the carcinogenic effects in these species and DNA binding, either in
vitro or in vivo. However, a good correlation was found between carcinogenicity and covalent
binding to total nuclear proteins both in vitro and in vivo. Covalent binding to liver DNA in all
three species was similar [(2.2-2.3 pmol carbon tetrachloride/mg DNA or 1.4-1.5 mol
nucleotides/mol carbon tetrachloride metabolites (x 106)]. Higher levels of covalent binding to
nuclear proteins, particularly the acidic nuclear protein fractions, were seen when expressed on a
pmol per mg basis. The authors discussed that the acidic nuclear proteins often have regulatory
functions in gene expression and that this may be important in carbon tetrachloride-induced
104 DRAFT - DO NOT CITE OR QUOTE
-------
carcinogenesis. Again, the authors indicated that they subtracted for background radioactivity
(35 dpm), but presented no data on control binding or how they corrected for control
radioactivity—a serious limitation for the use of this and other studies in assessing genotoxic
potential.
Levy and Brabec (1984) also investigated the ability of radiolabeled carbon tetrachloride
to bind to different types of DNA. After the administration of a single dose of [14C]- carbon
tetrachloride to male Sprague-Dawley rats, elevated levels of radioactivity were recovered bound
to purified mitochondrial and nuclear DNA. At both a low non-necrotizing and a high dose, 20-
to 50-fold more radioactivity was recovered bound to mitochondrial DNA than to nuclear DNA.
Binding to mitochondrial DNA also occurred when radiolabeled carbon tetrachloride was
incubated anaerobically with isolated mitochondria. Carbon tetrachloride is known to be
bioactivated in the mitochondria (Weber et al., 2003), so this report of elevated binding close to
the site of activation seems plausible. Again, there is no mention of a negative control or how
the samples were corrected for control radioactivity or counts. There is also no indication of
variability, the number of samples analyzed, or statistical significance of the results.
As described above, four studies have reported that following administration of
radiolabeled carbon tetrachloride, detectable amounts of radioactivity were recovered bound to
the extracted nuclear DNA. Significant methodological problems with each of the studies create
difficulties in interpreting the results. For one or two of the studies, basic information on sample
size, variability, and statistical significance is not provided. In addition, all studies failed to
provide data for untreated controls or indicate that the treatment samples were corrected for
control radioactivity (or dpm). For agents that bind weakly to DNA such as carbon tetrachloride,
even small increases in dpm in the controls can substantially alter the amount of binding
attributed to the chemical treatment.
Following the administration of a radiolabeled compound to an animal, the recovery of
radioactivity strongly associated with the isolated and extracted DNA is assumed to represent
covalent binding of the chemical or its metabolite to DNA. However, binding to proteins or
lipids can occur and may be recovered as contaminants within the DNA preparation (Kitta et al.,
1982). In addition, metabolic incorporation of the radiocarbon into DNA can also occur through
entry into the carbon pool of the cell with subsequent incorporation into DNA (Phillips et al.,
2000). This is a concern with carbon tetrachloride as metabolic studies have shown that
complete dechlorination of carbon tetrachloride can occur during cellular metabolism (Weber et
al., 2003; Halliwell and Gutteridge, 1999). It is therefore possible that part of the radiolabel
recovered in the in vivo 14C studies represents carbon tetrachloride-derived carbon that was
incorporated into DNA. For both of these reasons, it is important to identify the carbon
tetrachloride-derived DNA adducts to confirm that they occur in vivo. Unfortunately, this has
not yet occurred. Studies in nonaqueous model systems have shown that the trichloromethyl
radical can adduct nucleotides (Castro et al., 1994; Diaz Gomez and Castro, 1981), but it is not
105 DRAFT - DO NOT CITE OR QUOTE
-------
clear to what extent this would occur in aqueous systems or in vivo. Assuming that all of the
radiocarbon recovered represents adducts and that the levels of radioactivity in the controls are
equivalent to background, the magnitude of the DNA binding even at high toxic concentrations
is relatively low (Castro et al., 1989; Lutz, 1986; Levy and Brabec, 1984; Diaz Gomez and
Castro, 1980a; Lutz, 1979; Rocchi et al., 1973). Overall, there is limited evidence for the ability
of carbon tetrachloride metabolites to bind covalently to DNA in vivo.
Oxidative- and lipidpet•oxidation-derived'DNA adducts. Since reactive oxygen species
as well as lipid peroxidation-derived degradation products are also known to bind covalently to
DNA, numerous investigators have investigated whether oxidative adducts can be detected
following the administration of carbon tetrachloride to animals. Adducts derived from both
reactive oxygen and lipid peroxidation have been detected. Four studies employing a wide range
of doses attempted to detect DNA adducts derived from the lipid peroxidation product MDA or
similar reactive species, in the hepatic DNA of rats or hamsters. Of the four studies, two were
positive, one was equivocal, and one produced negative results. In addition, two studies detected
DNA adducts formed in the liver (as well as other tissues) from ^ram--4-hydroxy-2-nonenal (4-
HNE), another reactive species formed during lipid peroxidation. A brief description of the
individual studies follows.
In the initial study, Hadley and Draper (1990) briefly mention that the excretion of a
newly identified guanine-MDA adduct in the urine was increased 2.5-fold after the oral
administration of carbon tetrachloride to rats. No data were provided. In a later study using a
sensitive mass spectrometric method, Chaudhary et al. (1994) demonstrated that 4 days after the
administration of a 0.1 mg/kg oral dose of carbon tetrachloride to Sprague-Dawley rats, the liver
levels of the major endogenous MDA deoxyguanosine adduct increased 1.8-fold from 2.1 per
107 bases in the controls to 3.8 per 107 bases. The level of isoprostane, another product of lipid
peroxidation, was increased 16-fold in the treated animals.
In the report by Draper et al. (1995), the concentration of a deoxyguanosine-MDA
adducts in the liver was determined 48 hours after oral administration of 160 mg/kg carbon
tetrachloride to a group of five rats. A significant decrease in the level of this adduct was seen in
the carbon tetrachloride-treated rats as compared to controls. The authors suggested that in some
undetermined fashion, the liver DNA was protected from the increasing amounts of MDA
formed. They noted that under the same conditions, previous studies have shown that large
concentrations of MDA adducts with lysine, but not deoxyguanosine-MDA, are excreted in the
urine.
As part of another study to identify DNA adducts contributing to lipid hydroperoxide-
^9
mediated carcinogenesis, Wang and Liehr (1995) performed P-postlabeling to measure and
quantify the influence of carbon tetrachloride on the presence of endogenous adducts in Syrian
golden hamsters four hours after treatment with 160 mg/kg and 1600 mg/kg carbon tetrachloride.
106 DRAFT - DO NOT CITE OR QUOTE
-------
Treatment of the hamsters with the 160 mg/kg dose resulted in a doubling of renal and liver lipid
hydroperoxide levels. At the higher dose, renal lipid hydroperoxide levels were raised by 30%
but those in the liver were lowered by 50%, presumably due to lipid hydroperoxide-mediated
inactivation of metabolic enzymes required for the activation of carbon tetrachloride. The levels
of lipid hydroperoxide-derived DNA adducts in the kidney and liver varied in a comparable
manner; the measured endogenous adducts in the liver increased from -9 in the controls to -14
(expressed as relative adduct level * 108 adducts) at the low dose and decreased to ~8 at the high
carbon tetrachloride dose. Adduct levels in the kidney increased from ~11 in the controls to -25
at the low dose and -16 at the high dose. A good correlation between measured lipid
hydroperoxide levels and endogenous adducts was seen. The authors noted that the decreased
levels that were seen at the high dose were consistent with decreases in polar adducts observed
by Nath et al. (1990) in the livers of mice treated with carbon tetrachloride at 1,200 mg/kg. The
observed decrease is also similar to the decrease in the deoxyguanosine-MDA adduct seen by
Draper et al. (1995). It would appear that, at times, there can be an unusual relationship between
carbon tetrachloride dose and lipid peroxide-derived DNA adducts.
^9
Using [ P]-postlabeling combined with high-performance liquid chromatography, the
formation of trans-4-hydroxy-2-nonenal-derived cyclic adducts with deoxyguanosine was seen in
untreated rat and human tissues indicating that they are endogenous in origin (Chung et al.,
2000). Significant increases in the formation of the HNE-dG adduct were seen in the livers of
F344 rats treated with a single 3,200 mg/kg dose of carbon tetrachloride. Twenty-four hours
after treatment, the levels of the HNE-dG adducts were increased 37-fold as compared to those
of control animals (104 nmol/mol guanine versus 2.8 nmol/mol guanine). The adducts were
persistent as significant levels of the HNE-dG adducts (88 nmol/mod guanine) were present 72
hours after dosing.
The formation of l,N2-propanodeoxyguanosine adducts of trans-4-hydroxy-2-nonenal
(HNE-dGp-adducts) were measured in tissues of rats treated with carbon tetrachloride and
compared to those in control rats (Wacker et al., 2001). Carbon tetrachloride at a dosage of 500
mg/kg was administered by a single i.p. injection with sacrifices at 4, 16 and 24 hours post
injection, or by four injections at 24-hour intervals with the sacrifice occurring 8 hours after the
final dose. In the single injection studies, increases in HNE-dGp adducts were seen in the lung
and colon at various times and in the forestomach at all three time points. HNE-dGp adduct
levels also showed a nonsignificant increase in the liver and no change in the kidney. The
maximum increases seen were approximately 1.5- to 2-fold. In the multi-dose studies,
significant increases were seen in the liver (2.2-fold) and the forestomach (1.7-fold). The levels
of HNE-dGp adducts detected in the liver (2.8 per 107 normal nucleotides) in this study were of
the same order of magnitude as the adduct levels formed from MDA in the liver after treatment
with carbon tetrachloride (3.8 per 107 normal nucleotides; Chaudhary et al., 1994) and HNE
adducts found in the liver (22 per 107 normal nucleotides; Chung et al., 2000).
107 DRAFT - DO NOT CITE OR QUOTE
-------
The formation of 8-hydroxy-2'-deoxyguanosine (8-OHdG) is one of many adducts
formed between reactive oxygen species and DNA. Because of its prevalence and ease of
measurement, it is frequently used as a measure of oxidative DNA damage. Four studies have
attempted to measure 8-OHdG following the administration of carbon tetrachloride to rats or
mice. All four of the studies were positive, although the response in one was relatively weak.
In the initial study by Takahashi et al. (1998), the suitability of an antibody to detect 8-
OHdG for immunohistochemistry was determined by measuring adduct levels in hepatocyte
nuclei in a time-course study following the treatment of rats with carbon tetrachloride. Rats were
administered carbon tetrachloride at 3,200 mg/kg by gavage and sacrificed at 6 hours and
12 hours, and 1, 2, 3, and 7 days. Severe centrilobular necrosis was present by day 1. By days 2
and 3, anti-8-OHdG antibody staining was present in the mononuclear cells infiltrating the
necrotic centrilobular regions as well as in the hepatocytes in the midzonal and periportal
regions, and sinusoidal endothelial cells. At the day 2 time point, the formation of 8-OHdG in
DNA and 8-oxo-dGTPase messenger RNA (mRNA) expression were also increased by 5.1- and
1.7-fold, respectively. MDA plus 4-HNE showed peaks at 6 hours and 3 days. The findings
suggested that increased lipid peroxidation, rather than an excessive formation of 8-OHdG, was
the main contributing factor in the massive hepatic necrosis observed. The observed increase in
8-OHdG was attributed to the infiltrating mononuclear cells.
In the studies reported by Iwai et al. (2002), carbon tetrachloride was administered by
subcutaneous injection to rats twice a week at a dose of 200 mg/kg for the first 10 weeks, then at
400 mg/kg for the next 10 weeks. The rats were sacrificed at the end of week 22. At week 1, an
approximately twofold increase in 8-OHdG was seen in liver DNA of the treated rats when
compared with untreated controls. Consistent with this, the treated rats also exhibited higher
levels of 8-oxo-guanine DNA glycosylase 1 mRNA when measured using reverse-transcriptase
PCR.
Recently as part of an investigation into the susceptibility of young and old mice to
oxidative stressors, Lopez-Diazguerrero et al. (2005) administered carbon tetrachloride at a dose
of 43 mg/kg by i.p. injection on 3 consecutive days to young (2 months old) and older (14
months old) female CD-I mice. Twenty-four hours post-treatment, liver DNA in carbon
tetrachloride-treated young and old mice exhibited significant increases in 8-oxo-7,8-dihydro-2'-
deoxyguanosine (8-oxodG). The 8-oxodG levels increased from 0.5 residues/106 dG in the
young controls to 7.4 residues/106 dG in the carbon tetrachloride-treated young animals. In the
older animals, the 8-oxo-dG levels increased from 2.6 residues/106 dG in the controls to 10.1
residues/106 dG in the treated animals. The 8-oxodG levels between the treated young and old
animals did not differ significantly.
Similarly, as part of a larger study of oxidative biomarkers, Kadiiska et al. (2005)
measured the levels of 8-OHdG in the urine of male Fischer 344 rats previously administered
carbon tetrachloride at 120 mg/kg and 1,200 mg/kg by i.p. injection (urine collected 2-7 hours
108 DRAFT - DO NOT CITE OR QUOTE
-------
and 7-16 hours after carbon tetrachloride injection). Significant increases in 8-OHdG compared
to the control were seen for the low dose at 16 hours and the high dose at both sample times.
The high dose resulted in a seven- and threefold increase in the excreted adducts at the two
successive time points.
Available studies provide considerable evidence of DNA adducts derived from reactive
oxygen species or lipid peroxidation following in vivo administration. In some cases, the
relationship between dose and adduct levels appeared to be complex, without a monotonic
relationship between dose and response. In comparing the results from the various binding
studies, it should be remembered that the binding measured in radiocarbon binding studies
reflects all DNA adducts that contain the 14C label. In contrast, 8-OHdG and MDA and 4-HNE
adducts represent only a few of the many types of oxidative adducts (De Bont and van Larebeke,
2004; Halliwell and Gutteridge, 1999). When increases in these marker adducts are seen, the
total number of oxidative DNA adducts is undoubtedly larger. The overall consistency and
magnitude of the results from the oxidative adduct studies indicate that they likely represent the
major class of DNA lesion occurring in the rodent liver following carbon tetrachloride
administration.
^9
Endogenous adducts. Using the [ P]-post-labeling assay, Nath et al. (1990) investigated
the effects of carbon tetrachloride on presence of hepatic "I" spots (DNA adducts believed to be
formed from endogenous compounds) in both acute and long-term studies using 10-12 month-
old ICR mice. For the acute study, carbon tetrachloride was injected i.p. at a dose of 1,200
mg/kg. Twenty-four hours after the injection, the intensity of non-polar I-spots in the liver DNA
was increased as compared to those in corn oil-treated controls while the intensity of one polar I
spot was reduced. In contrast, in a long-term study of carbon tetrachloride, mice given two
consecutive injections of carbon tetrachloride (1,200 mg/kg) and sacrificed at 1, 4, 8, 12, and 22
weeks after the final injection, the total liver I compound levels were reduced to 17-49% of the
corresponding controls. Although there was a trend in recovery between weeks 8 and 22, the I-
compound levels remained significantly lower at week 22. The authors reported that "neither the
acute nor the chronic experiments with carbon tetrachloride produced extra spots indicative of
DNA adducts" indicating that exogenous adducts were not seen in the carbon tetrachloride-
treated mice.
Altered DNA methylation. Following carbon tetrachloride administration, a number of
studies have reported alterations in liver DNA methylation. In early studies performed by
Barrows and Shank (1981), increases in 7-methylguanine and O6-methylguanine were seen in
liver DNA 12 hours after rats were administered a single 1,000 mg/kg dose of carbon
tetrachloride. This increase was also seen in hydrazine- and ethanol-treated rats, and there was
some evidence in the hydrazine-treated rats that S-adenosylmethionine (SAM) was the methyl
109 DRAFT - DO NOT CITE OR QUOTE
-------
donor. Based on the observed results, the authors suggested that aberrant DNA methylation may
be a non-specific response to chemical injury to the liver.
More recently, Varela-Moreiras et al. (1995) investigated the effect of short-term
administration of carbon tetrachloride on hepatic DNA methylation and on SAM and S-
adenosylhomocysteine (SAH) in male Wistar rats administered 800 mg/kg carbon tetrachloride
by i.p. injection 2 times/week, for 3 weeks. Rats treated with carbon tetrachloride exhibited
hypomethylation of their hepatic DNA as measured by the extent to which the liver DNA from
the treated animals could be methylated in vitro using [3H-methyl]-SAM as a methyl donor. In
addition, decreased levels of SAM, methionine, and folate as well as increased levels of SAH
and homocysteine were seen. No changes were observed in the levels of cystathionine, reduced
glutathione, or in the activity of SAM-synthetase. The magnitude of the observed changes was
substantially reduced in animals co-administered SAM with carbon tetrachloride. The authors
proposed that "carbon tetrachloride disrupts the distribution of homocysteine between
remethylation and its degradation via the transsulphuration pathway, and that SAM, by resetting
the methylation ratio, restores this equilibrium." In eukaryotic and mammalian cells, gene
expression is influenced by the extent and patterns of DNA methylation, so the observed changes
in hepatic DNA methylation could represent an epigenetic alteration that could contribute to
carbon tetrachloride carcinogenesis.
4.4.2.5. Genotoxicity Studies: Summary of the Evidence for Genotoxic andMutagenic Effects
EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) identify a number
of criteria that should be considered in judging the adequacy of mechanistic data. These include
mechanistic relevance, number of studies of each endpoint, consistency of results in different test
systems and species, conduct of the tests according to generally accepted protocols, and degree
of consensus and general acceptance among scientists regarding the interpretation of the results.
In addition to these general considerations, evaluation of the genotoxicity data on carbon
tetrachloride poses some unique challenges. First, the genotoxicity data for carbon tetrachloride
are derived from a large number of experiments performed over a period spanning almost 40
years. Some assays were at early stages of development when performed, whereas others were
conducted under well-established protocols. As a result, the quality of the data varies widely. In
spite of this, most studies provide worthwhile information that can provide insights into the
potential of carbon tetrachloride to cause genotoxic effects. In addition, because of the large
numbers of tests performed, one would expect a number of studies to be positive due to random
chance or elevated error rates resulting from multiple comparisons. Some of the unique
challenges associated with evaluated carbon tetrachloride genotoxicity are outlined in
Table 4-12.
In accordance with the EPA mutagenicity risk assessment guidelines (U.S. EPA, 1986b),
when evaluating genotoxicity results, more weight has been given to tests performed in vivo in
110 DRAFT - DO NOT CITE OR QUOTE
-------
mammalian systems than to those performed in vitro using mammalian cells or in sub-
mammalian systems such as yeast and bacteria. Preference has also been given to results seen in
the rodent liver over those seen in other non-target tissues. This prioritization scheme is also
consistent with the current EPA carcinogen risk assessment guidelines (U.S. EPA, 2005a), which
state "Although important information can be gained from in vitro test systems, a higher level of
confidence is generally given to data that are derived from in vivo systems, particularly those
results that show a site concordance with the tumor data."
As indicated in Tables 4-8 to 4-11, well over 100 studies have been performed to assess
the genotoxic and mutagenic effects of carbon tetrachloride. A few experiments have been
conducted using human cells but none were located describing genotoxic effects in humans. A
summary evaluation by major type of genetic alteration is presented below.
Gene mutations. Intragenic or point mutations have been found in many cancer-related
genes and have been shown to play a determining role in chemical carcinogenesis (Stanley,
1995; Anderson et al., 1992; Harris, 1991). The ability of a chemical to form mutations in model
systems is an important consideration in establishing whether an agent acts through a mutagenic
MOA. There is little direct evidence that carbon tetrachloride induces intragenic or point
mutations in mammalian systems. The mutation studies that have been performed using
transgenic mice have yielded negative results, as have the vast majority of the mutagenesis
studies that have been conducted in bacterial systems. Since oxidative DNA adducts can be
converted into mutations, the inability to detect mutations in the transgenic mouse assays may be
an indication of efficient repair of oxidative lesions, a preferential formation of large
chromosomal mutations that are inefficiently detected in the transgenic models, or a reflection of
the limitations and sensitivity of the specific assays that were performed with carbon
tetrachloride. The two positive mutation/DNA damage studies conducted in E. coli were seen in
strains that are particularly sensitive to oxidative damage. Moreover, the intrachromosomal
recombination induced by carbon tetrachloride in S. cerevisiae is believed to result from double
stranded DNA breaks leading to deletion mutations. These results are consistent with DNA
breakage originating from oxidative or peroxidative stress that occurs concurrently with
cytotoxicity.
DNA strand breakage. DNA strand breakage is not a measure of mutation per se, but
can be a useful indicator of DNA damage and can contribute to an evaluation of an agent's
mutagenic potential. However, DNA breaks can also be formed during apoptotic and necrotic
cell death even by noncarcinogenic agents (Higami et al., 2004; Bergman et al., 1996; Grasl-
Kraupp et al., 1995; Elia et al., 1994), so the potential contribution of cytotoxicity to the
observed results needs to be carefully evaluated in studies reporting DNA damage. There is
some evidence that carbon tetrachloride administration results in DNA breakage and
111 DRAFT - DO NOT CITE OR QUOTE
-------
fragmentation in the liver of treated mice and rats; however, extensive hepatotoxicity was seen in
each of the studies where DNA damage has been reported. While some of the damage may be
due to reactive species formed during carbon tetrachloride metabolism and lipid peroxidation,
much of the observed damage appears to be more related to a cytotoxic response associated with
cell death than a genotoxic response leading to mutation. Indeed, the TUNEL assay used in two
of the positive carbon tetrachloride studies is commonly used as an early indicator of apoptotic
and necrotic cell death (Higami et al., 2004; Grasl-Kraupp et al., 1995).
Structural and numerical chromosome aberrations. Non-random structural and
numerical chromosomal aberrations are commonly seen in cancer cells and are believed to play
an important role in carcinogenesis (Pedersen-Bjergaard et al., 2002; Solomon et al., 1991;
Hansen and Cavenee, 1987; Oshimura and Barrett, 1986; Yunis, 1983). Furthermore, elevated
frequencies of chromosomal aberrations have been observed in humans exposed to
environmental chemicals, and recent investigations have indicated that individuals with elevated
levels of these alterations have increased risks of developing cancer (Hagmar et al., 2004;
Hagmar et al., 1998; Sorsa et al., 1992). Chromosomal alterations, measured in cell culture
systems or in animals treated in vivo, are commonly induced by carcinogenic agents, and the
evaluation of chromosomal aberrations or micronuclei is an important component of commonly
accepted genotoxicity testing schemes (Muller et al., 1999). Although less prone to problems of
cytotoxicity than the DNA breakage assays, under conditions of severe toxicity or stress,
increases in structural chromosome aberrations and micronuclei have been shown to occur
through indirect mechanisms (Galloway, 2000; Galloway et al., 1987). While aberrations
formed by noncarcinogenic agents under extreme conditions are not believed to be relevant to
mutagenic risks (Galloway, 2000), the significance of aberrations formed by carcinogens under
such conditions is less clear. For screening new chemicals, protocols have been established, at
least in vitro, to limit genotoxicity testing to concentrations that do not exhibit high toxicity
(Muller and Sofuni, 2000).
In the genotoxicity studies conducted on carbon tetrachloride, there is no evidence for
chromosomal damage when carbon tetrachloride has been tested in conventional assays for
chromosomal damage in the rat or mouse bone marrow. There is some evidence that following
high cytotoxic doses of carbon tetrachloride, increases in chromosome breakage and loss can
occur in the rat liver. The increases that have been observed have occurred exclusively at
hepatotoxic doses and have been limited in magnitude.
DNA adducts. The formation of DNA adducts within the liver following carbon
tetrachloride exposure is indicative of DNA damage occurring in the target organ. Because
adducts may be converted into mutations or DNA strand breaks, but can also be efficiently
repaired or remain unchanged in less critical non-coding sequences of DNA, these DNA adducts
112 DRAFT - DO NOT CITE OR QUOTE
-------
represent precursor lesions rather than specific mutagenic or genotoxic effects. It is generally
recognized that the types of DNA adducts formed after exposure can also provide valuable
insights into the mechanisms underlying an agent's genotoxic and mutagenic effects. There is
strong evidence of increases in DNA adducts formed from reactive oxygen species (i.e., 8-
OHdG) and lipid peroxidation products such as MDA and 4-HNE in the liver of rodents
following administration of carbon tetrachloride. Based on both in vivo and in vitro studies,
there is limited evidence for the formation of DNA adducts derived directly from carbon
tetrachloride.
UDS. The unscheduled synthesis of DNA is a measure of DNA repair and is commonly
used to assess DNA damage produced by mutagenic chemicals in the livers of treated animals.
Based on the reliable studies conducted to date, there is no evidence of UDS in the livers of
carbon tetrachloride-treated rats or mice even when tested under conditions producing significant
hepatotoxicity.
4.4.3. Initiation-promotion Studies
Tsujimura et al. (2008) examined the potential of carbon tetrachloride to induce pre-
neoplastic lesions in rat liver. Male rats (15/group) were exposed to carbon tetrachloride vapor
using nose-only inhalation exposure at concentrations of 0, 1, 5, 25, or 125 ppm for 6 hours/day,
6 days/week for 6 weeks. The numbers and area of glutathione S-transferase placental (GST-P)
positive foci were determined. Investigators also evaluated liver tissue for histopathological
changes and measured serum chemistry parameters and carbon tetrachloride concentrations in
blood.
Absolute and relative liver weights were statistically significantly increased at
concentrations >25 ppm. The areas (mm2/cm2) and numbers (number per cm2) of GST-P
positive foci were statistically significantly increased in the carbon tetrachloride-exposed rats at
25 and 125 ppm, but not at concentrations of 1 and 5 ppm. Histopathological examination of the
liver revealed centrilobular ballooning of hepatocytes, interlobular fibrosis, increased mitoses of
hepatocytes, and eosinophilic foci in all 125-ppm exposed rats. At 25 ppm, centrilobular
ballooning of hepatocytes was reported. Investigators observed microgranuloma in 14/15 rats
exposed to 25 ppm, but not in any rats exposed to 5 ppm or 125 ppm. Exposure-related changes
in liver enzymes were reported. ALP was increased at >5 ppm, and AST and ALT were
increased at >25 ppm. At 125 ppm, gamma-glutamyl transpeptidase activity and total
cholesterol were increased.
Bull et al. (2004) used an initation-promotion study design to examine how
dichloroacetate, trichloroacetate, and carbon tetrachloride, three liver carcinogens that appear to
induce liver tumors by different MO As, might interact when given as mixed exposures. Only the
carbon tetrachloride results are summarized here. B6C3F1 mice were initiated by the tumor
113 DRAFT - DO NOT CITE OR QUOTE
-------
initiator, vinyl carbamate (3 mg/kg at 2 weeks of age), and then promoted by carbon
tetrachloride for 18, 24, 30, and 36 weeks beginning at weaning (21 days of age). Initial carbon
tetrachloride doses (50, 100, and 500 mg/kg-day by gavage) were too high for study purposes
and were reduced to 5, 20, and 50 mg/kg-day. Dose-related increases in mean tumor volume
were observed with 20 and 50 mg/kg-day carbon tetrachloride, but each produced equal numbers
of tumors at 36 weeks. At doses >100 mg/kg-day, substantial increases in the number of tumors
per animal were observed, but the mean tumor size decreased. The investigators concluded that
this finding suggests that initiation occurs at carbon tetrachloride doses of >100 mg/kg-day,
perhaps as a result of a high-dose inflammatory response that is known to occur with high doses
of carbon tetrachloride. The investigators observed that trichloroacetate substantially increased
the numbers of tumors observed at early time points when combined with carbon tetrachloride
and suggested that the interaction between carbon tetrachloride and trichloroacetate may be
explained through stimulation of the growth of cells with differing phenotypes.
4.4.4. Neurotoxicity Studies
High-dose, acute toxicity studies in humans and animals reported neurotoxic effects of
carbon tetrachloride. Human case reports mention headache, drowsiness, comas, or seizures
occurring after exposure by ingestion or inhalation (Stewart et al., 1965; New et al., 1962;
Norwood et al., 1950). Lehmann and Schmidt-Kehl (1936) reported neurological symptoms
occurring after exposures of >30 mg/L (>4,800 ppm). In an acute inhalation study in rats, signs
of central nervous system depression occurred at>4,600 ppm (Adams et al., 1952).
Frantik et al. (1994) quantified the air concentrations of carbon tetrachloride and other
solvents that would produce an acute neurotoxic effect in rats and mice. Whole-body exposures
at various concentrations were undertaken for groups of four male albino Wistar rats for 4 hours
or female H mice for 2 hours; animals were then tested for the inhibition of propagation and
maintenance of an electrically evoked seizure discharge. Testing was conducted by application
of a short electrical impulse (0.2 seconds, 50 Hz, 180 volts in rats and 90 volts in mice) through
ear electrodes. The most consistent sensitive measure was the duration of tonic extension
through the hind limbs in rats and the velocity of toxic extension (reciprocal of latency) in mice.
The authors reported the "isoeffective concentration" of carbon tetrachloride in air by
interpolating to the level that would produce one-third of the maximum effect. The isoeffective
concentrations were 611 ppm (one-tailed 90% CI: 98 ppm) for rats and 1,370 ppm (one-tailed
90% CI: 465 ppm) for mice.
114 DRAFT - DO NOT CITE OR QUOTE
-------
4.4.5. Immunotoxicity Studies
Immunological effects of carbon tetrachloride have been evaluated in mice and rats
exposed by the parenteral (Kaminski et al., 1990, 1989), oral (Guo et al., 2000; Ladies et al.,
1998; Ahn and Kim, 1993; Smialowicz et al., 1991; Kaminski et al., 1989), and inhalation (Ban
et al., 2003) routes. Results of available studies indicate that carbon tetrachloride produces
adverse effects on T-cell-dependent immunity at doses that are hepatotoxic. However, it is
important to note that immunological effects were, at least in part, secondary to hepatotoxicity
and the process of hepatic repair. Information regarding the mechanism of immune system
effects and the relationship of immunotoxicity to hepatotoxicity, inflammation, and repair,
including activation of Kupffer and stellate cells, is reviewed in Section 4.5.6.
Effects of parenteral exposure of mice to carbon tetrachloride on immune function was
studied by Kaminski et al. (1990, 1989). Carbon tetrachloride was injected intraperitoneally to
female B6C3F1 mice at doses of 0, 500, 1,000, or 1,500 mg/kg-day in corn oil for 7 consecutive
days. Systemic toxicity endpoints included body weight, selected organ weights (liver, spleen,
lung, kidney, and thymus), and serum chemistry. Humoral antibody responses (the number of
antibody-forming cells) to T-cell-dependent antigen (sheep erythrocytes) and T-cell-independent
antigen (DNP-ficoll) were evaluated in vivo and in vitro. Treatment with carbon tetrachloride
had no significant effect on survival, clinical signs, body weight gain, or organ weights, except
for a decrease in thymus weight at >500 mg/kg-day. There were significant increases in serum
ALT and bilirubin at >500 mg/kg-day, albumin at >1,000 mg/kg-day, and total protein at 1,500
mg/kg-day. In vivo response to T-cell-dependent antigen was suppressed in a dose-related
manner: by 36% at 500 mg/kg-day to 53% at 1,500 mg/kg-day. The in vivo response to T-cell-
independent antigen was suppressed by 16% at the highest dose. T-cell-dependent responses
were more vulnerable to carbon tetrachloride than were T-cell-independent responses.
Kaminski et al. (1990) conducted a series of immunotoxicity experiments in female
B6C3F1 mice given carbon tetrachloride by i.p. injection or gavage in corn oil. Oral or i.p.
administration of 500-5,000 mg/kg-day for 7 consecutive days significantly reduced in vivo
T-dependent antibody response to sheep erythrocytes; the route of administration had no
significant effect. Intraperitoneal injection of 25 mg/kg-day for 30 consecutive days also
significantly reduced the in vivo T-dependent antibody response. Intraperitoneal injection at 500
or 1,000 mg/kg-day on 8 consecutive days significantly increased serum ALT (by five- and
sevenfold, respectively), but treatment at 250 mg/kg-day had no effect; no effects on body or
organ weights (spleen, liver, or thymus) were observed. Intraperitoneal injection with 5-1,000
mg/kg-day on 7 consecutive days significantly reduced the total microsomal protein content per
gram of liver. Whereas treatment at 25-100 mg/kg-day for 3 days had no effect on the T-cell-
dependent antibody response, pretreatment with 4 g/kg ethanol caused significant
immunosuppression at 50 or 100 mg/kg-day. The authors concluded that immunosuppression
following treatment with carbon tetrachloride is related to its bioactivation by microsomal
115 DRAFT - DO NOT CITE OR QUOTE
-------
enzymes.
The effects of oral exposure to carbon tetrachloride have been studied in mice (Guo et al.,
2000; Ahn and Kim, 1993) and rats (Ladies et al., 1998; Smialowicz et al., 1991). Guo et al.
(2000) administered carbon tetrachloride at doses of 0, 50, 100, 500, or 1,000 mg/kg-day by
gavage in corn oil to B6C3F1 mice on 14 consecutive days. Mice were examined for gross
pathology, at which time organ weights were recorded for thymus, lungs, liver, spleen, and
kidneys with adrenals. Blood was collected for hematology and serum chemistry analyses.
Immunological endpoints included quantification of T- and B-cells in the spleen and spleen
immunoglobulin (IgM) antibody-forming cell response and antibody liters to a T-dependent
antigen, sheep red blood cells; in addition, cellular-mediated immunity was evaluated in host
responses to infection by two bacterial strains. Treatment had no effect on mortality, the
incidence of clinical signs, body weight gain, or the weights of brain, spleen, lung, thymus, and
kidneys and no biologically significant effect on hematology parameters. Absolute liver weight
was significantly increased by 23% at 500 mg/kg-day compared with that in vehicle controls.
Significant, dose-related increases in relative liver weights were observed at 350 mg/kg-day.
Treated groups showed histopathology in the liver (cloudy swelling of hepatocytes and
centrilobular necrosis) but not in other organs. Significant dose-related changes in serum
parameters included increases in ALT (19-fold at 50 mg/kg-day), total protein (9% at 100
mg/kg-day), BUN (34% at 500 mg/kg-day), and globulin (20% at 1,000 mg/kg-day) and a
decrease in glucose (by 20% at 1,000 mg/kg-day). Exposure to carbon tetrachloride had no
effect on the mixed leukocyte response, cytotoxic T-lymphocyte activity, or natural killer (NK)
cell activity. Exposure to carbon tetrachloride reduced the humoral immune response; the IgM
antibody-forming cell response to sheep erythrocytes was suppressed at >50 mg/kg-day,
maximally by 43% at 1,000 mg/kg-day. IgM serum liters to sheep erythrocytes were
significantly reduced at > 100 mg/kg-day. Absolute numbers of CD4+CD8+ T-cells were reduced
by 40% in all dosed groups compared with vehicle controls; absolute numbers and percentages
of CD4+CD8 T-cells were reduced in the 500 mg/kg-day group. Treatment with carbon
tetrachloride reduced host resistance to both Streptococcuspneumoniae and Listeria
monocytogenes at 500 and >50 mg/kg-day, respectively. In mice, the low dose of 50 mg/kg-day
was a LOAEL for immunotoxic effects of carbon tetrachloride by oral exposure, affecting
primarily T-cell-dependent responses.
The immunotoxicity of carbon tetrachloride was investigated in male ICR mice
administered 1 mL/kg (1,590 mg/kg) carbon tetrachloride in olive oil twice weekly by gavage
(Ahn and Kim, 1993) for 4 weeks. Systemic endpoints included relative weights of liver, spleen,
and thymus. Immune response to sheep erythrocytes was assessed using hemagglutinin (HA)
liters, assays of plaque-forming cells (PFCs) and delayed-type hypersensitivity reaction, and
measurement of NK cell and phagocytic activity. Compared with control (olive oil) mice,
116 DRAFT - DO NOT CITE OR QUOTE
-------
relative liver weights were significantly increased by 12% in mice treated with carbon
tetrachloride. Relative weight of thymus and spleen were significantly decreased by 6 and 25%,
respectively, compared with that in controls. The HA titer against sheep erythrocytes and the
PFC response, both measures of T-cell-dependent antibody response, were significantly inhibited
by 56 and 40%, respectively, in mice treated with carbon tetrachloride. The delayed-type
hypersensitivity response, a measure of in vivo cell-mediated immunity, was significantly
increased by carbon tetrachloride treatment, indicating that carbon tetrachloride alters T-helper
cell function. In carbon tetrachloride-treated mice, the number of rosette-forming cells (1.90%)
was significantly decreased compared with controls (4.18%). NK cell activity, activity of
phagocytic cells, and the number of circulating leukocytes were significantly decreased by 61,
40, and 34%, respectively, in carbon tetrachloride-treated mice compared with controls. These
results demonstrate that treatment with carbon tetrachloride alters humoral and cell-mediated
immune functions.
The effect of carbon tetrachloride on humoral immunity was assessed by the IgM
response to intravenously injected sheep erythrocytes in male CD rats administered 0, 12.5, or 25
mg/kg carbon tetrachloride (eight rats per group) in corn oil by gavage 5 days/week for 30 or 90
days (Ladies et al., 1998). Carbon tetrachloride-induced hepatotoxicity was assessed by
examination of the liver by light microscopy and measurement of serum SDH activity in rats
injected with sheep erythrocytes or control vehicle. In rats treated for 30 days, administration of
12.5 and 25 mg/kg carbon tetrachloride decreased sheep erythrocyte-specific serum IgM levels
by 42 and 45%, respectively. In contrast, sheep erythrocyte-specific serum IgM levels were
unchanged compared with controls in the 12.5 mg/kg group and increased by 50% in the 25
mg/kg group in rats treated for 90 days. The authors proposed that time-dependent decreases in
metabolism of carbon tetrachloride contributed to the increased IgM response observed after 90
days of treatment with 25 mg/kg. Exposure to carbon tetrachloride did not alter the population
of splenic lymphocyte subsets (numbers of T-helper cells, T-cyt/sup cells, total T-cells, total B-
cells) or weights or morphology of lymphoid organs (spleen and thymus). Exposure to 25 mg/kg
carbon tetrachloride for 30 or 90 days and to 12.5 mg/kg for 90 days produced hepatotoxicity, as
indicated by increased relative liver weight, histopathological alterations (centrilobular fatty
changes), and increases in serum SDH activity. Results of hepatotoxicity assessments in rats
treated with sheep erythrocytes were similar to controls, indicating that exposure to sheep
erythrocytes did not interfere with the histopathological examination or measurement of serum
SDH activity.
Smialowicz et al. (1991) evaluated immunotoxicity in male F344 rats given carbon
tetrachloride by gavage at doses of 0, 5, 10, 20, or 40 mg/kg-day on 10 consecutive days.
Endpoints included body weight gain, organ weights (liver, kidney, spleen, and thymus), hepatic
microsomal protein levels, serum chemistry, and histopathology of liver and kidney.
Immunological endpoints included NK cell activity of splenocytes, cytotoxic T-lymphocyte
117 DRAFT - DO NOT CITE OR QUOTE
-------
responses, and proliferative responses of splenic lymphocytes to T-cell mitogens
(phytohemagglutinin and concanavalin A), a B-cell mitogen (S. typhimurium), and a T- and B-
cell mitogen (pokeweed mitogen). Primary antibody responses to a T-cell-dependent antigen
(sheep erythrocytes) were also tested following treatment with carbon tetrachloride at 0, 40, 80,
or 160 mg/kg-day for 10 days. Treatment at >80 mg/kg-day significantly reduced body weight
gain; separate analysis by two-way analysis of variance of 40 mg/kg-day groups and their
respective controls in three experiments indicated a significant decrease in body weight gain.
Treatment had no significant effect on the absolute or relative weights of the spleen, thymus, or
kidney or on absolute liver weight; relative liver weight was significantly increased at 40 mg/kg-
day. There were dose-related increases in AST and ALT: 47% and twofold, respectively, at 20
mg/kg-day. Whereas no hepatic histopathology was detected in control rats, there were dose-
related increases in the incidence and severity of vacuolar degeneration (minimal at 5 mg/kg-day
to mild/moderate at 40 mg/kg-day) and hepatic necrosis (none-to-minimal at 10 mg/kg-day to
minimal/mild at 40 mg/kg-day). Treatment had no significant effect on kidney histopathology or
renal serum parameters. Treatment had no effect on immunological parameters in rats at doses
that caused hepatic toxicity.
The effects of inhaled carbon tetrachloride on systemic and local immune response were
investigated in female BALB/c mice exposed to 0, 100, 200, or 300 ppm (630, 1,260, or
1,890 mg/m3) of carbon tetrachloride vapor (Ban et al., 2003). Exposure duration was not
reported; however, the maximum exposure period was most likely less than 24 hours. Immune
function was assessed for systemic (spleen) and local (lung-associated lymph nodes) effects
using the IgM response to sheep erythrocytes and interferon-y (IFN-y) production by spleen and
lung-associated lymph node cells isolated from exposed mice. Assessments of other systemic
effects of carbon tetrachloride (e.g., hepatotoxicity) were not conducted. The IgM response of
spleen cells to sheep erythrocytes, as measured by the number of PFCs, was unaffected by
carbon tetrachloride treatment. In lung-associated lymph nodes, the PFC number was
significantly increased (1.7-fold increase) in mice exposed to 300 ppm carbon tetrachloride
compared with controls, but no differences were observed in the 100 or 200 ppm carbon
tetrachloride groups. In spleen cells, carbon tetrachloride exposure had no effect on IFN-y
release, whereas IFN-y release from lung-associated lymph node cells was significantly
increased by 150 to >600% of controls in all carbon tetrachloride groups. Results of this study
indicate that inhaled carbon tetrachloride exerts immunotoxicity at the point of entry.
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION
There is considerable in vivo and in vitro evidence that may contribute to an
understanding of the MO A by which carbon tetrachloride produces toxic effects in animals
(Weber et al., 2003; Jaeschke et al., 2002; Plaa, 2000; Omura et al., 1999; Mehendale, 1990;
118 DRAFT - DO NOT CITE OR QUOTE
-------
Recknagel et al., 1989; DiRenzo et al., 1982; Slater, 1982; Gillette, 1973; Recknagel and Glende,
1973; Castro et al., 1973, 1972; Castro and Diaz Gomez, 1972). Discussion of the roles of
metabolism, lipid peroxidation, and disruption of calcium homeostasis in carbon tetrachloride
toxicity is presented below.
4.5.1. Metabolism is Required for Toxicity
Numerous studies show that metabolism of carbon tetrachloride is required for toxicity.
As discussed in Section 3.3, the initial step of carbon tetrachloride metabolism is reductive
dehalogenation by CYP450, primarily CYP2E1. Studies using CYP450 inhibitors (e.g., SKF-
525A, colchicine, silymarin, and allylisopropylacetamide) have shown that these compounds,
which inhibit activity of CYP450 enzymes and consequently prevent metabolism of carbon
tetrachloride, prevent carbon tetrachloride-induced liver damage (Martinez et al., 1995; Letteron
et al., 1990; Mourelle et al., 1988; Bechtold et al., 1982; Weddle et al., 1976).
Carbon tetrachloride itself has been shown to temporarily protect against carbon
tetrachloride toxicity by inhibiting activity of CYP450 and reducing its own metabolism. Glende
(1972) found that rats pretreated with a small, nonlethal dose of carbon tetrachloride were
protected against toxicity from a subsequent large and ordinarily lethal challenge dose of carbon
tetrachloride. Protection was not yet evident when the challenge occurred only 6 hours after the
initial dose but was complete for challenge doses administered 1-3 days after pretreatment and
was gradually less effective for subsequent challenge doses. CYP450 activity measured in this
study showed a sharp decline after the initial dose that reached a minimum at 1 day after
treatment. Gradual increases in CYP450 activity were observed at 4 days and later. The close
parallel between time course of effects on CYP450 activity and toxicity in this study is further
evidence that metabolism of carbon tetrachloride by CYP450 is required for toxicity.
Wong et al. (1998) demonstrated the specific significance of CYP2E1 to carbon
tetrachloride-induced hepatotoxicity in mice using CYP2E1 knockout mice (cyp2eT~). Twenty-
four hours after i.p. injection of 1 mL/kg (1.59 g/kg) of carbon tetrachloride to wild type mice
(cyp2eJ+/+\ there were no significant effects on survival or liver/body weight ratios, but there
was a 422-fold increase in serum ALT, a 125-fold increase in serum AST, and significant
necrosis in the centrilobular hepatocytes. In cyp2el+ + mice, serum ALT was found to be
significantly increased at 12 hours and peaked 24 hours after carbon tetrachloride dosing
(Avasarala et al., 2006). Administration of the same dose to knockout mice (cyp2el~ ") resulted
in no increase in AST, only a slight elevation in serum ALT (within normal range), and absence
of liver histopathology. Additionally, Badger et al. (1997) demonstrated that treatment of
Sprague-Dawley rats with gadolinium chloride (GdCls) decreased CYP450 levels in liver
preparations from these animals, which may explain the protective role of GdCls in carbon
tetrachloride-treated animals (see Section 4.5.6).
Carbon tetrachloride administered in vivo to guinea pigs decreased microsomal CYP450
119 DRAFT - DO NOT CITE OR QUOTE
-------
concentrations in the adrenal gland, providing evidence that the adrenal cortex is an active site of
carbon tetrachloride metabolism (Colby et al., 1981). In the adrenal gland, necrosis caused by
carbon tetrachloride is localized to the innermost region of the cortex, the zona reticularis, where
there is far greater activation of carbon tetrachloride by microsomal enzymes than other regions
of the adrenal cortex (Colby et al., 1994). The profile of CYP450 isozymes in the adrenal is
directed principally toward steroid metabolism and bears little resemblance to that in the liver.
Using an in vitro model with isolated tissue from guinea pig adrenal zona reticularis, Colby et al.
(1994) reported that carbon tetrachloride is specifically activated by a 52 kDa CY450 enzyme
associated with xenobiotic metabolism.
Chemical inducers of CYP450 that increase the activity of CYP450, and particularly
those that induce the activity of CYP2E1 specifically, potentiate carbon tetrachloride
hepatotoxicity. See Section 4.8.6 for a list of chemical CYP450 inducers, and associated
references, shown to potentiate carbon tetrachloride hepatotoxicity. In vitro, it has been shown
that hepatocyte cell lines that over-express CYP450 have increased levels of carbon
tetrachloride-induced cytotoxicity (Jaeschke et al., 2002; Takahashi et al., 2002; Dai and
Cederbaum, 1995).
4.5.2. Role of Free Radicals
The products of carbon tetrachloride metabolism by CYP2E1 include trichloromethyl and
trichloromethyl peroxy radicals (see Section 3.3). Studies with radical scavengers, such as N-
acetylcysteine, and spin-trapping agents, such as 7V-fert-butyl-a-(4-nitrophenyl)nitrone, have
shown that these agents confer a protective effect against carbon tetrachloride-induced liver
toxicity (Brennan and Schiestl, 1998; Stoyanovsky and Cederbaum, 1996; Slater, 1982),
indicating that free radicals released via metabolism of carbon tetrachloride may contribute to
carbon tetrachloride toxicity.
The trichloromethyl and trichloromethyl peroxy radicals are highly reactive species that
may produce cellular damage by covalently binding to cellular macromolecules to form nucleic
acid, protein, and lipid adducts (Recknagel and Glende, 1973). Studies using radiolabeled
carbon tetrachloride have shown irreversible binding to cellular DNA, proteins, nuclear proteins,
and lipids, following bioactivation in various in vitro and in vivo systems (Boll et al., 2001b;
Azri et al., 1991; Castro et al., 1989; DiRenzo et al., 1982; Diaz Gomez and Castro, 1980a;
Castro and Diaz Gomez, 1972; Gordis, 1969). Pulse radiolysis experiments showed that the
trichloromethyl peroxy radical is far more reactive towards cellular macromolecules than the
trichloromethyl radical (Slater, 1981; Packer et al., 1978). The trichloromethyl radical binds to
macromolecules strongly but more slowly than the more reactive trichloromethyl peroxy radical.
However, Slater (1981) concluded that most covalent binding involved the trichloromethyl
radical, because binding with the trichloromethyl peroxy radical, although faster, produces a less
stable product. This process involving the binding of the trichloromethyl radical to
120 DRAFT - DO NOT CITE OR QUOTE
-------
macromolecules is known as haloalkylation (Dianzani, 1984).
4.5.3. Lipid Peroxidation
Under oxygen-rich conditions, the trichloromethyl radical is converted to the more
reactive trichloromethyl peroxy radical. The trichloromethyl peroxy radical can attack polyenoic
(polyunsaturated) fatty acids in the cellular membrane, forming fatty acid free radicals that
initiate subsequent autocatalytic lipid peroxidation through a chain reaction (see Figure 4-3).
Although the trichloromethyl radical can also initiate lipid peroxidation, it does so at a
slow rate compared to the more reactive trichloromethyl peroxy radical (Slater, 1981). In this
process, the trichloromethyl peroxy radical abstracts a hydrogen from the methylene carbon
between two double bonds in the polyunsaturated fatty acid, generating a lipid free radical.
Rearrangement of the double bonds into a conjugated pattern shifts the location of the free
radical electron to an adjacent tetrahedral carbon, and reaction of the free radical carbon with
molecular oxygen produces a peroxylipid free radical. The peroxylipid radical can abstract a
hydrogen from a donor molecule, forming a lipid hydroperoxide, a first step in the oxidation of
the fatty acid. If the hydrogen donor is another polyunsaturated fatty acid, the process begins
again, perpetuating the lipid peroxidation (Klaassen, 1996). If the donor is a small hydrocarbon
free radical, an alkane can form.
121 DRAFT - DO NOT CITE OR QUOTE
-------
CH3-CH2-CH=CH-CH2-(CHm)n-COOR1
Any hydrogen can be removed;
hydrogens on CH2 between
two cis double bonds are
particularly accessible.
•CH3 or 'OO-CH
-3
CH4orH-OO-CH3
CH3-CH2-CH=CH-CH-(CHm)n-COORi
m = 1 or 2. Unsaturated fatty
acids will have one or more
CH=CH-CH2 combinations.
In all other cases m = 2.
R! = remainder of a membrane
phospholipid, sphingolipid, or
glycolipid.
CH3-CH2-CH-CH=CH-(CHm)n-COORi
Uo2
Alkanes
or
alkenes
CH3-CH2-CH-CH=CH-(CHm)n-COOR1
.
CH3-CH2-CH-CH=CH-(CHm)n-COOR1
•: i
'O •
Lipid ,
remnant
Aldehydes
i
t
Acids
Propagation: when R2 is another fatty
acid moiety.
Termination: when R2 is an antioxidant
or GSH.
\ ,R2-H
X
When the fatty acid is polyunsaturated,
it is likely to undergo further
peroxidation.
Hydroxylated
fatty acids
t
Malondialdehyde,
4-hydroxynonenal
Figure 4-3. Lipid peroxidation.
122
DRAFT - DO NOT CITE OR QUOTE
-------
Numerous studies have demonstrated the occurrence of lipid peroxidation following
carbon tetrachloride exposure, either by detection of conjugated dienes (a characteristic marker
of lipid peroxidation) in liver lipids (Tribble et al., 1987; Lee et al., 1982; Recknagel and Glende,
1973; Rao and Recknagel, 1969), increased exhalation of ethane or pentane (end degradation
products of peroxidized T-3 and T-6 polyunsaturated fatty acids, respectively) in treated rats
(Younes and Siegers, 1985; Gee et al., 1981), or occurrence of reactive aldehydes, such as
malonaldehyde and 4-hydroxyalkenals, frequently measured as thiobarbituric acid-reactive
substances (TEARS) (de Zwart et al., 1997; Gasso et al., 1996; Ichinose et al., 1994; Fraga et al.,
1987; Comporti, 1985; Comporti et al., 1984). TEARS form when the oxidation of the fatty acid
94- ^4-
progresses from the hydroperoxide, facilitated by the oxidation of Fe to Fe in a Fenton
reaction, leading to breaks in the fatty acid chain and the formation of aldehydes from the fatty
acid fragments (Klaassen, 1996). Among the many different aldehydes formed from lipid
peroxidation are 4-HNE and MDA.
In vitro studies have shown that 4-HNE at high concentrations (>10 uM) is a cytotoxic
product of liver microsomal lipid peroxidation because of degradation of T-6 unsaturated fatty
acids (Esterbauer et al., 1991; Van Kuijk et al., 1990). The formation of FTNE-dGp-adducts may
be relevant to the formation of cancer when these promutagenic lesions are insufficiently
repaired (Wacker et al., 2001). Wacker et al. (2001) developed a sensitive detection method for
l,N2-propanodeoxyguanosine adducts of FINE (promutagenic adducts), a specific marker for
genotoxic interaction of reactive oxygen species and lipid peroxidation products. Background
levels of adducts in various tissues in F344 rats were found in the range of 18-158 adducts/109
nucleotides. Levels of endogenous DNA adducts were higher in the liver, and lower levels were
found in kidney, lung, and colon. After induction of lipid peroxidation by a single i.p.
application of 50 uL carbon tetrachloride at a dosage of 500 mg/kg body weight, levels of HNE-
dG-adducts in the liver were elevated 1.5- to twofold compared with those in controls. The
authors concluded that these promutagenic adducts are evidence of radical-initiated lipid
peroxidation, which can lead to cancer if not repaired effectively. Other studies have also
indicated that lipid peroxidation byproducts could inhibit certain DNA repair systems and thus
indirectly increase the rate of spontaneous mutations (Curren et al., 1988; Krokan et al., 1985).
Chung et al. (2000) identified lipid peroxidation as the cause of the 37-fold increase of
HNE-dG adducts in liver tissue DNA of F344 rats after treatment with 3.2 g/kg carbon
tetrachloride via i.p. administration. Wang and Liehr (1995) found that MDA induced DNA
adducts in hamsters treated with an oral administration of 0.1 mL/kg carbon tetrachloride, and
the levels of adducts formed were directly correlated with lipid hydroperoxide concentrations.
These reactive aldehydes can form DNA adducts causing frameshift or base mispairing (G to T
and G to A mutations).
Similar to 4-HNE, MDA is a result of oxidative degradation of polyunsaturated fatty
123 DRAFT - DO NOT CITE OR QUOTE
-------
acids with more than two methylene-interrupted double bonds. In mammalian tissues, precursors
for MDA are arachidonic acid and docosahexenoic acid.
Ichinose et al. (1994) compared the in vitro production of MDA per mg microsomal
protein from hepatic microsomes in several species. The rat generated the highest amount of
MDA over 2 hours, followed by monkey, mouse, pig, cow, rabbit, sheep, horse, and dog. Using
tissue slices from male Sprague-Dawley rats incubated in 1 mM carbon tetrachloride for 2 hours,
Fraga et al. (1987) found significant increases over control values in TEARS (nmol/g tissue)
released from treated liver (-fourfold), kidney (-threefold), spleen (-twofold), and testis
(-fivefold). Abraham et al. (1999) reported significantly elevated lipid peroxide levels in the
lung (65%), testis (200%), kidney (85%), and liver (200%) of Wistar rats exposed to carbon
tetrachloride vapor over a 12-week period. The results of Fraga et al. (1987) and Abraham et al.
(1999) show that lipid peroxidation can occur in other tissues besides the liver, specifically in the
kidney, testis, spleen, and lung.
Lipid peroxidation has been proposed to disrupt cellular membranes, resulting in loss of
membrane integrity (Recknagel and Glende, 1989) and the production of reactive aldehydes that
can attack tissues and form protein and DNA adducts (Comporti, 1985; Comporti et al., 1984).
These aldehydes may diffuse from the membranes and traverse intracellularly or extracellularly
away from the point of origin to attack distant targets, acting as secondary toxicants.
Immunohistochemical procedures using antibodies directed against MDA and 4-HNE protein
adducts have been used to detect adducts in rat liver sections treated with carbon tetrachloride
(Bedossa et al., 1994). Abraham et al. (1999) reported significantly elevated protein carbonyl
content, a measure of protein adduct formation, in the liver (238%), lungs (51%), and testis
(21%) of carbon tetrachloride vapor-treated rats compared with controls.
Hartley et al. (1999) studied the temporal relationship between carbon tetrachloride-
initiated lipid peroxidation, hepatocellular damage, and formation of 4-HNE and MDA-hepatic
protein adducts, using immunohistochemical detection of aldehyde-adducted proteins in liver
sections and immunoprecipitation and immunoblotting procedures to detect and characterize
4-HNE and MDA-adducted proteins in liver homogenates from male highly alcohol-sensitive
rats treated with 1 mL/kg (1.59 g/kg) of carbon tetrachloride in mineral oil by gavage. Mineral
oil alone elicited subtle centrilobular steatosis, a slight increase in necrosis at 12 hours, and a
slight elevation of serum ALT at 24 hours. The livers of rats treated with carbon tetrachloride in
mineral oil exhibited a significant number of ballooned hepatocytes and inflammatory cells at
12 hours and progressive, massive centrilobular steatosis, inflammation, and necrosis at 18-48
hours. There was a fivefold increase in serum ALT at 6 hours after treatment, peaking at
36 hours with a 32-fold increase in ALT over control. Between 18 and 36 hours posttreatment,
TEARS values in liver homogenates of treated rats were maximal at a 2.5-fold increase over
controls. MDA-amine and 4-hydroxynonenal-sulfhydryl protein adducts were detectable at
6 hours in the midzonal region and in the centrilobular region at 12-36 hours. The
124 DRAFT - DO NOT CITE OR QUOTE
-------
correspondence in time course and location for lipid peroxidation, production of protein adducts,
and liver damage suggests that protein adducts resulting from lipid peroxidation contribute to
hepatocellular injury in carbon tetrachloride-treated rats.
Evidence of the relationship between hepatotoxicity and lipid peroxidation was also
reported by Younes and Siegers (1985). These researchers found that administration of an iron-
chelating agent, deferoxamine, suppressed both lipid peroxidation (ethane exhalation) and
hepatotoxicity (serum ALT and SDH levels) in GSH-depleted mice treated with carbon
tetrachloride. This result suggests that the observed hepatotoxic effect was secondary to lipid
peroxidation. Administration of the antioxidant vitamin E (a-tocopherol) was shown to reduce
lipid peroxidation (pentane exhalation) and metabolism (chloroform generation) in another rat
study (Gee et al., 1981).
Ciccoli et al. (1978) reported that binding of carbon tetrachloride radicals to cellular
lipids occurred in rat extrahepatic tissues, although to a lesser extent than the liver. Almost half
of the radioactivity from [14C]-labeled carbon tetrachloride incorporated into phospholipids was
found in the liver (47%); in other tissues, incorporation into phospholipids was found in
intestinal mucosa (24%), kidney (9%), adrenal gland (8%), and lung (5%), while spleen, testis,
brain, heart, and skeletal muscle lipids showed minor levels of radioactivity. Fatty acid methyl
esters prepared from the phospholipids of intestinal mucosa and kidney exhibited an electron
capture detector (ECD) response similar to the liver (indicating free radical reaction); however,
other tissues with low-level 14C incorporation showed no ECD response. In an in vitro model,
Colby (1981) and Colby et al. (1994) provided evidence that carbon tetrachloride can stimulate
lipid peroxidation in adrenal micosomes. Incubation of carbon tetrachloride plus NADPH
produced a decrease in guinea pig adrenal microsomal CYP450 content and stimulated lipid
peroxidation in adrenal zona reticularis microsomes (as indicated by rate of MDA production).
In the absence of NADPH, carbon tetrachloride did not affect lipid peroxidation and little
covalent binding was demonstrable.
Lipid peroxidation byproducts can also form promutagenic DNA adducts and modify
double-stranded DNA by formation of amino-imino propene crosslinks between the NH2 group
of the guanosine base and complementary cytosine base. In rat hepatocytes cultured with 0.25,
1, or 4 mM carbon tetrachloride, Beddowes et al. (2003) showed that carbon tetrachloride caused
a dose-dependent increase in the formation of DNA strand breaks, 8-oxodG and MDA-DNA
adducts. The increased formation of DNA strand breaks and MDA-DNA adducts was
statistically significant at 1 and 4 mM. The level of 8-oxodG was statistically elevated only at 4
mM, a concentration that caused a decrease in cellular viability. Carbon tetrachloride induced
lipid peroxidation carbonyl product formation (>twofold) at 4 mM; lower concentrations were
not studied. The formation of MDA-DNA adducts correlated with the ability of carbon
tetrachloride to induce lipid peroxidation, although failure to measure lipid peroxidation at the
two lower concentrations (0.25 and 1 mM) somewhat limits the ability to establish this
125 DRAFT - DO NOT CITE OR QUOTE
-------
correlation.
4.5.4. Depletion of Glutathione
Reduced glutathione is capable of donating a hydrogen to quench a free-radical chain
reaction and can play a key role in limiting the damage to cellular membranes caused by lipid
peroxidation. The efficacy of reduced glutathione in quenching a free radical reaction is
dependent on the activity of GSH peroxidase, the enzyme that facilitates the transfer of hydrogen
to hydrogen peroxide with the formation of glutathione disulfide and water. Cellular levels of
reduced glutathione are restored through the activity of GSH reductase using NADPH + H+ as
the hydrogen donor (Klaassen, 1996).
Cabre et al. (2000) assessed the temporal relationships between hepatic lipid
peroxidation, GSH metabolism, and development of cirrhosis in groups of 10 male Wistar rats
exposed to carbon tetrachloride. Rats were injected intraperitoneally with 0.5 mL of carbon
tetrachloride in olive oil twice weekly for 9 weeks to induce hepatic cirrhosis. By the second
week, 10/10 livers were fibrotic. Cirrhosis appeared in all 10 animals by week 9. Hepatic GSH
levels were significantly reduced, beginning at week 5, and GSH peroxidase activity was
significantly decreased at week 7 in carbon tetrachloride-treated rats; the activity of GSH
peroxidase is dependent on a sufficient level of GSH. Cytosolic GSH S-transferase activity was
also significantly inhibited in rats receiving carbon tetrachloride at week 1. TEARS (lipid
peroxides) began to be elevated by week 7. The findings of this study show that induction of
cirrhosis in rats by carbon tetrachloride produces a decrease in several components of the hepatic
GSH antioxidant system. Impairment of this hepatoprotective system was related to an increased
generation of lipid peroxides.
Gorla et al. (1983) confirmed that oral pretreatment of male Sprague-Dawley rats with 2
g/kg of GSH 30 minutes before an i.p. injection of carbon tetrachloride (1.59 mg/kg) partially
prevented the hepatic necrosis that normally occurs 24 hours after carbon tetrachloride dosing.
Treatment with cysteine, which is a precursor of GSH and, like GSH, is able to conjugate
phosgene (from chloroform) produced from carbon tetrachloride, also protected against carbon
tetrachloride hepatotoxicity when given orally 30 minutes before or 1 hour after i.p. injection of
carbon tetrachloride (de Ferreyra et al., 1974).
Gasso et al. (1996) investigated the effects of SAM availability on lipid peroxidation and
liver fibrogenesis in male Wistar rats with carbon tetrachloride-induced cirrhosis. SAM is
essential for the production of the GSH precursor homocysteine, which provides the sulfur for
the endogenous synthesis of cysteine (the source of the reactive-SH functional group in
glutathione). A SAM deficiency can also limit transmethylation reactions that function in DNA
and RNA methylation and the production of thymine for DNA repair. Gasso et al. (1996) found
that depletion of GSH triggers a feedback mechanism, leading to inactivation of SAM
synthetase, which in turn causes a further decrease in GSH. SAM synthetase is responsible for
126 DRAFT - DO NOT CITE OR QUOTE
-------
the endogenous production of SAM from the essential amino acid methionine. The deficit of
SAM could be corrected by exogenous administration of SAM but not methionine. Accordingly,
the deficit appeared to be the result of enzyme inhibition rather than methionine availability.
Carbon tetrachloride-treated rats receiving SAM for 6 weeks had significantly higher
SAM synthetase activity (156 ±5.6 pmol/minute/mg protein) than rats treated with carbon
tetrachloride alone (89.4 ±3.4 pmol/minute/mg protein) (Gasso et al., 1996). The hepatic GSH
was significantly decreased in carbon tetrachloride-treated rats (2.7 ± 13 nmol/g tissue) and
returned to normal in rats receiving SAM for 3 or 6 weeks (3.7 ± 0.13 and 3.9 ± 0.11 nmol/g
tissue). Carbon tetrachloride-treated rats receiving SAM for 6 weeks had significantly lower
liver toxicity (collagen and propyl hydroxylase activity, reduced lipid peroxidation, and less
advanced liver fibrosis). The hepatic TEARS, markers of lipid peroxidation, were also
significantly lower in rats treated with carbon tetrachloride and SAM for 6 weeks (98 ± 5 nmol/g
tissue) than rats treated with only carbon tetrachloride (134 ± 12 nmol/g tissue). In rats treated
with carbon tetrachloride and SAM for 6 weeks, serum AST (76 ± 6 U/L) and ALT (57 ± 4 U/L)
were lower than rats treated with only carbon tetrachloride (321 ± 33 and 185 ± 21 U/L,
respectively). These data provide evidence that hepatic lipid peroxidation is increased during
hepatic fibrogenesis and that exogenous SAM may lead to an increase of GSH levels, which
could prevent SAM synthetase inactivation, inhibit lipid peroxidation, and, consequently,
attenuate the development of liver fibrosis and cirrhosis.
Will et al. (1999) demonstrated in 11 untreated mammalian cell lines that the intrinsic
levels of GSH expression were inversely correlated with the background level of oxidative DNA
modifications, such as 8-hydroxyguanine. Depletion of GSH with buthionine sulphoximine, an
inhibitor of y-glutamyl-cysteine that generates the precursor to GSH (Edgren and Revesz, 1987),
increased the basal levels of oxidative DNA base modifications. Schisandrin B, a compound that
enhances the GSH antioxidant status in hepatic mitochondria, was hepatoprotective against
carbon tetrachloride exposure in Balb/c mice (Chiu et al., 2003).
4.5.5. Disruption of Calcium Homeostasis
Calcium plays an essential role in cellular physiology. Levels of calcium in the cell are
maintained far below extracellular levels by resistance of the plasma membrane to passive
diffusion of calcium across the membrane and by active transport of calcium across the cell
membrane and into the extracellular space (Klaassen, 1996). Calcium within the cell is actively
transported across the microsomal membrane into the endoplasmic reticulum and across the
mitochondrial membrane into the mitochondria. Maintenance of calcium homeostasis is vital to
cellular function, and interference with calcium homeostasis is suspected to cause cell death
(Farber, 1981).
Calcium ATPase helps maintain calcium-level homeostasis within the cell. When
cytosolic calcium levels are highly elevated, the calcium ATPase, located in the plasma
127 DRAFT - DO NOT CITE OR QUOTE
-------
membrane, is activated. Activation of calcium ATPase triggers the transport of calcium ions
from the cytosol to the endoplasmic reticulum hydrolyzing ATP in this process. This process
94-
also requires Mg to be tightly complexed to ATP. A rise in cytosolic calcium also induces the
binding of calcium ions to regulatory calcium-binding proteins, like calmodulin (a 148-residue
protein found in many cells and an essential subunit of the plasma membrane calcium ATPase).
Binding of cytosolic calcium to calmodulin triggers an allosteric activation of calcium ATPase
that accelerates the uptake of calcium ions from the cytosol by the endoplasmic reticulum to
maintain a low cytosolic concentration of <1 |iM calcium. While calmodulin complements
calcium ATPase, it also modulates the activities of a large number of calcium-dependent proteins
(Garrett and Grisham, 1999).
Studies conducted with carbon tetrachloride have reported > 100-fold increases in the
cytosolic concentration of calcium following exposure (Agarwal and Mehendale, 1986, 1984;
Long and Moore, 1986; Kroner, 1982). In a study in which hepatocytes were incubated in a
medium containing EGTA, a calcium-specific chelator, but no added calcium, treatment with
carbon tetrachloride elicited an increased calcium-dependent conversion of glycogen
phosphorylase "b" to phosphorylase "a" by phosphorylase kinase, which is stimulated by
increased intracellular calcium levels (Long and Moore, 1986). The lack of extracellular calcium
in this experimental system indicates that the carbon tetrachloride exposure released sequestered
calcium, probably from microsomes. The authors suggested that calcium could contribute to cell
death by the overstimulation of calcium-responsive cellular enzymes that initiate a cascade of
events, resulting in irreversible cell injury.
Hepatocytes treated with carbon tetrachloride had an impaired ability to maintain proper
calcium levels that was associated with inactivation of the calcium ATPase of the endoplasmic
reticulum (Lowrey et al., 1981; Moore, 1980). Administration of carbon tetrachloride caused an
85% reduction of ATP-dependent calcium uptake and calcium-sequestering capacity of the
hepatocyte endoplasmic reticulum (Moore et al., 1976). Hemmings et al. (2002) showed that
carbon tetrachloride decreased active calcium transport across the plasma and mitochondrial
membranes, as well as the endoplasmic reticulum, in rat liver. In vitro experiments confirmed
that inhibition of the plasma membrane calcium transport system by carbon tetrachloride was
rapid (within a minute) and strong (>90%) (Hemmings et al., 2002).
Carbon tetrachloride can also increase cytoplasmic calcium levels by opening certain
calcium transport channels in membranes. Liver endoplasmic reticulum contains ryanodine-
sensitive calcium-binding sites (Feng et al., 1992). Ryanodine is an alkaloid, usually found in
the skeletal and cardiac sarcoplasmic reticulum, that induces calcium release from liver
microsomes by binding to certain calcium release channels. Stoyanovsky and Cederbaum (1996)
showed that hepatic ryanodine-sensitive calcium channels may be involved in the elevation of
cytosolic calcium levels in the liver following carbon tetrachloride dosing. These researchers
observed elevated cytosolic calcium levels after treatment of hepatic microsomes with 50 uM of
128 DRAFT - DO NOT CITE OR QUOTE
-------
carbon tetrachloride. Ruthenium red, a specific inhibitor of the ryanodine receptor calcium
release channel, has been shown to block the carbon tetrachloride-induced release of calcium.
Activation of calcium-dependent cysteine proteases andphospholipases
The increase in cytosolic calcium and inhibition of the calcium pump can activate a
number of calcium-dependent cysteine proteases (e.g., calpains, known for their involvement in
proteolysis of proteins during mitosis, apoptosis, and necrosis) and phospholipases (particularly
phospholipase A2) that preferentially hydrolyze membrane lipids. Activation of these enzymes
can contribute to toxicity of carbon tetrachloride in the liver.
When calcium homeostasis has been disrupted because of the loss of microsomal
membrane integrity, increased levels of calcium leakage activate a number of cytosolic and
lysosomal degradative enzymes that are also leaked out into the extracellular space from dying
cells; these degradative enzymes can subsequently attack neighboring cells. Limaye et al. (2003)
demonstrated the involvement of calpain, a calcium-dependent cytosolic neutral cysteine
protease that leaks out from injured hepatocytes, in degrading cytoskeletal and membrane
proteins (e.g., a-fodrin, talin, filamin), and other macromolecules crucial to maintaining cellular
integrity, culminating in cell lysis and hepatocyte cell death. Calpain causes cell death by
attacking the plasma membrane, and, once the integrity of the membrane is lost, cells are
rendered highly vulnerable to destruction. Limaye et al. (2003) showed how calpain inhibition
with calpain-specific inhibitor N-benzyloxycarbonyl-valine-phenylalanine methyl ester (CBZ)
after carbon tetrachloride treatment substantially reduced the progression of injury and improved
animal survival. After 48 hours, the elevation in calpain activity was substantially in the carbon
tetrachloride + CBZ-treated rats than the carbon tetrachloride + DMSO-treated rats. (DMSO
was the vehicle used for CBZ administration.) More significantly, in rats challenged with a
normally lethal dose of carbon tetrachloride (3 mL/kg, i.p.), 75% of the male Sprague-Dawley
rats that received CBZ (60 mg/kg) 1 hour after carbon tetrachloride administration survived,
while rats treated with carbon tetrachloride alone or carbon tetrachloride and DMSO experienced
75% mortality. All control rats survived.
This study also evaluated the degradative effect of calpain on a-fodrin, a membrane
protein (Limaye et al., 2003). Calpain is known to degrade the 240-kDa fodrin to produce a 150-
kDa fragment. In rats receiving CBZ after carbon tetrachloride, the breakdown of a-fodrin was
similar to that in controls, indicating that inhibition of calpain released from dying hepatocytes
resulted in lower cellular damage. To confirm that cell death was caused by calpain, fresh
hepatocytes were incubated with calpain and 2.5 mM calcium. By the end of 240 minutes, cell
viability was decreased to 75%. Dying cells were found to develop plasma membrane blebs,
indicating cytotoxicity, which is typical of cytoskeletal damage induced by calpain. In the
presence of CBZ, hepatocytes were completely protected from calpain-mediated cell death.
Additional experiments with E64, a cell-impermeable inhibitor of calpain, also significantly
129 DRAFT - DO NOT CITE OR QUOTE
-------
reduced plasma ALT levels, suggesting that the presence of calpain in the extracellular space is
responsible for the damage to some hepatocytes.
While these results suggest that calpain is a major contributor in the progression of liver
injury, other degradative enzymes are also released into the extracellular space, such as
nucleases, acid phosphatases, and phospholipases. Loss of calcium sequestration capacity
caused by in vitro metabolism of carbon tetrachloride by isolated rat liver microsomes (e.g.,
Lowrey et al., 1981) correlates with carbon tetrachloride-dependent activation of phospholipase
A2, measured by lysophosphatide formation or release of arachidonic acid from the hydrolysis of
esterified arachidonic acid from the sn-2 position of hepatocyte phospholipids (Glende and
Pushpendran, 1986). Studies with rat hepatic microsomes demonstrated a progressive loss of
phospholipid after incubation in 5 mM CaQ2, with time-dependent losses of microsomal protein
activity (G6Pase and CYP450) that reached 80% by 3 hours (Chien et al., 1980). Quinacrine, a
phospholipase A2 inhibitor at 150 mg/kg i.p., has been shown to prevent carbon tetrachloride-
induced liver necrosis at 24 hours when administered 30 minutes before or 6 or 10 hours after
carbon tetrachloride exposure (2.5 mL/kg orally) (Gonzalez Padron et al., 1993). The authors of
this study concluded that phospholipase A2 plays a major role in carbon tetrachloride-induced
liver necrosis.
Glende and Pushpendran (1986) prelabeled hepatocytes with [3H]-arachidonic acid or
[14C]-ethanolamine and subsequently incubated the cells with carbon tetrachloride. Calcium-
activated phospholipase A2 activity was determined by measuring the release of [3H]-arachidonic
acid from cellular phospholipids labeled with arachidonate or the formation of [14C]-
lysophospholipids from cellular phospholipids labeled with ethanolamine. Treatment with 0.23-
1.3 mM of carbon tetrachloride increased the endogenous phospholipase A2 activity 1.4- to 5.3-
fold beginning within 30-60 minutes. A similar study in isolated hepatocytes revealed that
carbon tetrachloride stimulated phospholipase A2 activity (monitored by production of
lysophosphatidyl ethanolamine) within 15 minutes, succeeded within 15 minutes by
hepatotoxicity, as measured by the release of LDH from the cells into the medium (Glende and
Recknagel, 1992). This same study demonstrated that related compounds (chloroform,
bromotrichloromethane, and 1,1-dichloroethylene) similarly activate phospholipase A2 activity in
hepatocytes. The authors suggested that phospholipase A2 could contribute to hepatocyte
pathology by two different means: by increasing the hydrolysis of membrane lipids at rates
exceeding the rate of repair and/or by the phospholipase A2-dependent generation of toxic
prostanoids via initiation of the arachidonic acid cascade.
4.5.6. Immunological and Inflammatory Effects
Immunological effects of carbon tetrachloride were, at least in part, secondary to
hepatotoxicity and the process of hepatic repair. Carbon tetrachloride induces a regenerative
response in the liver similar to that observed following administration of other hepatotoxic
13 0 DRAFT - DO NOT CITE OR QUOTE
-------
chemicals (e.g., acetaminophen) or partial hepatectomy (PH) (Jeon et al., 1997; Delaney et al.,
1994). The regenerative process involves complex interactions among several cell types and cell
mediators, including the hepatic synthesis and release of serum-borne growth factors
(hepatotrophic factors) that act directly on liver cells to induce mitosis (Luster et al., 2000).
Hepatotrophic factors also appear to act on peripheral organs, most notably the spleen (Delaney
and Kaminski, 1994; Delaney et al., 1994). Results of studies on the effects of hepatotrophic
factors indicate that immune effects of carbon tetrachloride, and other hepatotoxic chemicals,
may be mediated by tumor growth factor (TGF)-pl released from the liver during the
regenerative process (Jeon et al., 1997; Delaney et al., 1994; Delaney and Kaminski, 1993).
A series of experiments conducted by Delaney and coworkers suggest that carbon
tetrachloride-induced suppression of T-cell function is mediated through serum-borne factors
(Delaney et al., 1994; Delaney and Kaminski, 1993). Serum from B6C3F1 mice treated with
250 or 500 mg/kg carbon tetrachloride in corn oil by gavage for 7 days, a dose regimen that
produced hepatotoxicity, suppressed the sheep erythrocyte-induced antibody response of carbon
tetrachloride-naive spleen cells in vitro (Delaney and Kaminski, 1993). In a subsequent study,
Delaney et al. (1994) demonstrated that carbon tetrachloride-induced suppression of the T-cell-
dependent humoral response is at least partially mediated by TGF- pi. Suppression of the sheep
erythrocyte antibody response of naive spleen cells in vitro by serum of mice exposed to carbon
tetrachloride (single oral dose of 500 or 1,000 mg/kg carbon tetrachloride in corn oil) was
abolished upon addition of TGF-p-specific antibodies to the assay. Jeon et al. (1997) reported
elevations of TGF- pi mRNA in the liver of B6C3F1 mice treated with a single hepatotoxic dose
(500 mg/kg) of carbon tetrachloride within 24 hours of exposure. Although direct effects of
carbon tetrachloride on the immune system by carbon tetrachloride have not been ruled out,
results of in vitro and in vivo studies suggest that immunotoxicity is, in part, mediated by TGF-
Pl secreted by the liver during tissue repair.
Inflammation contributes to the development of chemical-induced hepatotoxicity and
possibly to immunotoxic effects. Kupffer cells are hepatic macrophages that respond to signals
from injured hepatocytes by releasing biologically active mediators, such as prostaglandins,
reactive oxygen species, and cytokines (Luckey and Petersen, 2001). Factors released by
Kupffer cells after activation by carbon tetrachloride include nitric oxide, tumor necrosis factor-
ex (TNF-a), TGF-P, and interleukins-6, -8, and -10. The mediators produced by Kupffer cells are
involved in the regulation of the inflammatory response and fibrotic response following hepatic
injury. As discussed earlier, TGF-P 1 released from the liver plays an important role in the
immunotoxic effects of carbon tetrachloride, providing a possible link between hepatic
inflammation and Kupffer cell activation by immunotoxic events.
Stellate cells are hepatic fat-storing cells that respond to liver injury by proliferating,
migrating towards damaged areas, releasing nitric oxide and extracellular signal-regulated
kinases that perform various functions in different tissues, and increasing production of
131 DRAFT - DO NOT CITE OR QUOTE
-------
extracellular matrix, thereby promoting fibrosis (Weber et al., 2003; Marra et al., 1999). Stellate
cells are activated by TGF-a. Acute treatment with carbon tetrachloride increases the activity of
extracellular signal-regulated kinases from stellate cells (Marra et al., 1999).
Carbon tetrachloride has been shown to stimulate increases in the numbers of
immunodetectable Kupffer cells in the livers of treated rats, as well as increases in releases of
various cytokines and reactive oxidative species, corresponding to different stages of liver
histopathology (Luckey and Petersen, 2001; Alric et al., 2000). Towner et al. (1994) reported
that i.p. administration of 1,275 mg/kg of carbon tetrachloride to male Wistar rats was
characterized by hepatic edema from the accumulation of vacuoles and lipid droplets in
parenchymal cells and accumulation of phagosomes (large secondary lysosomes) and extrusion
of pseudopods in enlarged Kupffer cells. With a 1-hour intravenous pretreatment with 10 mg/kg
gadolinium trichloride (GdCb), an inhibitor of Kupffer cell activation, the parenchymal cells
were normal and Kupffer cells contained only a few secondary lysosomes. The protective effect
of GdCls was not associated with a change in detectability of carbon tetrachloride-generated
trichloromethyl radical by electron spin resonance spectroscopy.
The effects of GdCb on carbon tetrachloride-induced hepatic toxicity were evaluated in
other studies. Muriel et al. (2001) treated male Wistar rats with 4,000 mg/kg of carbon
tetrachloride by gavage in corn oil, with or without i.p. injection of 2,000 mg/kg GdCls. Twenty-
four hours later, rats treated with carbon tetrachloride showed typical hepatotoxicity (increased
serum enzymes and bilirubin, 2.5-fold increase in hepatic lipid peroxidation, and liver
histopathology: ballooning necrotic hepatocytes). Treatment with GdCb eliminated the
increases in serum biomarkers of membrane damage and hepatic lipid peroxidation and
significantly reduced the severity of hepatic necrosis. In a follow-up study of similar design,
male Wistar rats were treated with carbon tetrachloride (400 mg/kg by i.p. injection in mineral
oil 3 times/week), GdCls (20 mg/kg i.p. in saline daily), or both for 8 weeks (Muriel and
Escobar, 2003). Cotreatment with GdCb resulted in partial or complete protection against the
effects of carbon tetrachloride on serum ALT, GGT, ALP, and bilirubin; liver MDA content
(index of lipid peroxidation); liver hydroxyproline content (index of collagen content and
fibrosis); and histopathology (both necrosis and fibrosis). Depletion of liver glycogen by carbon
tetrachloride was not affected by GdCb, and GdCb itself produced a significant depletion of
glycogen.
Although multiple studies have indicated that GdCls treatment reduces or inhibits carbon
tetrachloride-induced hepatotoxicity through inactivation of Kupffer cells, GdCls may also
reduce carbon tetrachloride toxicity through other cellular mechanisms. Rose et al. (2001)
demonstrated both in vivo and in vitro that GdCb stimulated hepatocyte proliferation through a
mitogenic mechanism involving TNF-a, and promoted recovery from liver damage. GdCb has
also been shown to inhibit free radical-induced hepatocyte damage by nonselective blockage of
Na+ channels that induce necrosis in an in vitro model (Barros et al., 2001). Critical to carbon
132 DRAFT - DO NOT CITE OR QUOTE
-------
tetrachloride-induced toxicity is the generation of reactive metabolites by CYP2E1 for which
GdCls downregulates the gene expression in vivo (Okamoto, 2000; Badger et al., 1997).
Overall, multiple cellular mechanisms have been demonstrated by which GdCls reduces carbon
tetrachloride-induced toxicity and indicates that toxicity is not mediated exclusively through
inactivation of Kupffer cells.
4.5.7. Changes in Gene Expression
Changes in gene expression in response to exposure to carbon tetrachloride have been
investigated in the liver of rats and mice and in the human hepatoma cell line (lessen et al., 2003;
Fountoulakis et al., 2002; Bartosiewicz et al., 2001; Holden et al., 2000; Columbano et al., 1997;
Menegazzi et al., 1997). Many of the known upregulated genes are related to stress, DNA
damage and repair, and signal transduction, but for the most part, their specific contributions to
hepatotoxicity are not known. Fountoulakis et al. (2002) reported a fivefold increase in
expression of some genes related to stress and DNA damage repair in the livers of male Wistar
rats 6 hours after they received 400 mg/kg carbon tetrachloride. Rats receiving 3,190 mg/kg
showed 10-fold increases in expression in some genes. Some of the stress- and DNA-damage-
related genes upregulated by both doses at 24 hours included GADD45, GADD153, heat-shock
proteins, heme oxygenase, p53, c-myc, and c-jun. There were some qualitative differences in
altered gene expression at 6 and 24 hours between the two doses administered in this study,
which possibly provides a basis for the different hepatocellular responses to carbon tetrachloride-
induced injury. The hepatic expression of the Cdk inhibitor p21 in mice treated with carbon
tetrachloride occurs just prior to necrosis at 6 hours, and mice deficient in that gene do not
exhibit necrosis in response to carbon tetrachloride (Kwon et al., 2003); p21 also contributes to
the cessation of cellular proliferation that occurs later.
Intraperitoneal injection of Sprague-Dawley rats with 160 mg/kg of carbon tetrachloride
in corn oil activated c-fos and c-jun gene expression in the liver within 30 minutes (Gruebele et
al., 1996). Pretreatment of rats with diallyl sulfide, an inhibitor of CYP2E1, 3 hours before
dosing with carbon tetrachloride reduced c-jun mRNA levels by 76%. Treatment with carbon
tetrachloride also increased hepatic nuclear levels of the NF-KB transcription factor, which
regulates genes involved in responses to inflammation, apoptosis, hepatocyte proliferation, and
liver regeneration.
Columbano et al. (1997) investigated the relationship between immediate early genes and
hepatocyte proliferation through comparison of the hepatic levels of c-fos, c-jun, and LRF-1
transcripts during mouse liver cell proliferation under two conditions: (1) direct hyperplasia
induced by the primary mitogen (and hepatocarcinogen) l,4-bis[2-(3,5-
dichloropyridyloxy)]benzene (TCPOBOP), and (2) compensatory regeneration caused by a
necrogenic dose of carbon tetrachloride (single intragastric dose of 2 mg/kg in oil) or by
performing a 2/3 PH. A striking difference in the activation of early genes was observed. In
133 DRAFT - DO NOT CITE OR QUOTE
-------
spite of a rapid stimulation of S phase by the mitogen TCPOBOP, there were no changes in the
expression of c-fos, c-jun and LRF-1 or in steady state mRNA hepatic levels of IGFBP-1 (a gene
highly expressed in rat liver following PH), and only a slight increase in c-myc and PRL-1. In
contrast, a rapid, massive and transient increase in the hepatic mRNA levels of all these genes
was observed during carbon tetrachloride-induced regeneration that was comparable to those
seen following 2/3 PH. In similar research from the same laboratory, the pattern of immediate
early gene and growth factor gene expression in the rat liver induced by primary mitogens
(including lead nitrate [LN]), cyproterone acetate, or nafenopin [NAF]) was shown to differ from
that observed following compensatory liver regeneration occurring after cell loss/death and
direct hyperplasia resulting from a 2/3 PH or a necrogenic dose (2 mL/kg) of carbon
tetrachloride (Menegazzi et al., 1997). In this study, the following indicators of gene expression
were examined: modifications in the activation of two transcription factors, NF-KP and AP-1;
steady-state levels of TNF-a mRNA; and induction of the inducible nitric oxide synthase
(iNOS). Liver regeneration after treatment with carbon tetrachloride was associated with an
increase in steady-state levels of TNF-a mRNA, activation of NF-KP and AP-1, and induction of
iNOS. LN induced NF-KP, TNF-a and iNOS mRNA but not AP-1, whereas direct hyperplasia
induced by the other two primary mitogens occurred in the complete absence of modifications in
the hepatic levels of TNF-a mRNA, activation of NF-KP and AP-1, or induction of iNOS,
although the number of hepatocytes entering S phase 18-24 hours after NAF was similar to that
seen after PH. The findings from these two studies indicate that regenerative proliferation alone
does not explain the tumorigenic response associated with carbon tetrachloride in chronic
bioassays, but these data do not preclude regenerative proliferation as a biologically based
marker of such causal events.
4.5.8. Mechanisms of Kidney Toxicity
Limited data suggest that some of the same mechanisms by which carbon tetrachloride
produces damage to the liver can also operate in the kidney. Dogukan et al. (2003) observed
moderate renal histopathology (tubular necrosis, dilatation, atrophy, glomerular hypercellularity,
capillary obliteration, and interstitial fibrosis) in male Wistar rats subcutaneously injected
3 times/week with 240 mg/kg of carbon tetrachloride in olive oil for 7 weeks. The tissue damage
was associated with a significant increase in renal MDA (+34%), indicating lipid peroxidation,
and the researchers attributed the effects to oxidative stress. The tissue damage was also
accompanied by a significant decrease in renal GSH peroxidase, indicating a depletion of renal
GSH as contributing to the observed tissue damage. Studies by Fraga et al. (1987) using rat
tissue slices in vitro and Abraham et al. (1999) in rats in vivo also showed lipid peroxidation in
the kidney resulting from carbon tetrachloride exposure.
Ozturk et al. (2003) evaluated the levels of antioxidants in the kidney of Sprague-Dawley
rats subcutaneously injected with 1,594 mg/kg-day of carbon tetrachloride on 4 consecutive
134 DRAFT - DO NOT CITE OR QUOTE
-------
days. Compared with control kidneys, treated kidneys had significantly elevated activity levels
for superoxide dismutase (+30%) and catalase (+46%) but reduced activity for GSH peroxidase
(-44%) 24 hours after the last injection. The authors attributed the reduced activity of GSH
peroxidase to decreased availability of renal GSH in its reduced form. Treated kidneys showed
severe and extensive cortical histopathology: focal glomerular necrosis, tubular dilation,
epithelial vacuolization or necrosis (with detachment from the basement membrane), and protein
casts. A parallel group treated with carbon tetrachloride and betaine (a methyl group donor)
showed no differences from the control group for superoxide dismutase or GSH peroxidase,
whereas catalase was significantly elevated (+34%). Kidneys of rats treated with carbon
tetrachloride plus betaine had normal glomerular histology and only sparse tubular dilatation,
epithelial vacuolization, and few cell detachments. The authors suggested that the beneficial
effect of betaine on renal histology and GSH peroxidase activity was related to its promotion of
SAM levels, as has been demonstrated in the liver by other investigators. This study suggests
that similar toxicological mechanisms may occur in the liver and kidney of rats treated with
carbon tetrachloride.
Cytosolic phospholipase A2 levels were significantly elevated in the renal cortex and
medulla of rats with carbon tetrachloride-induced cirrhosis and ascites (Niederberger et al.,
1998). The authors attributed the increase in phospholipase A2 to the increased renal production
of prostaglandins in cirrhosis.
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS
Hepatic and renal effects are the most sensitive noncancer effects of oral or inhalation
exposure to carbon tetrachloride in humans and animals.
4.6.1. Oral
No long-term toxicity data are available for humans with quantified oral exposures to
carbon tetrachloride, but case reports identify the liver and kidney as the primary target organs
following acute exposures. Evidence of acute oral hepatotoxicity in humans comes from
observations of liver enlargement, elevated serum enzyme (AST and/or ALT), bilirubin levels, or
histopathology (hepatocyte degeneration) (Ruprah et al., 1985; Stewart et al., 1963; Docherty
and Nicholls, 1923; Docherty and Burgess, 1922). Other acute oral effects in humans include
renal toxicity, usually delayed relative to hepatic toxicity (New et al., 1962) and lung effects
secondary to renal failure (Umiker and Pearce, 1953). The prominence of hepatic injury in
acutely exposed humans suggests that hepatic toxicity observed in subchronic animal studies is
an important and relevant consideration for human health risk assessment of carbon
tetrachloride.
Studies in laboratory animals indicate that hepatic toxicity is the predominant noncancer
effect of subchronic or chronic oral exposure to carbon tetrachloride (Table 4-13). In these
13 5 DRAFT - DO NOT CITE OR QUOTE
-------
studies, evidence of hepatic damage included liver histopathology (fatty degeneration, necrosis,
fibrosis, cirrhosis, inflammation, and regenerative activity), along with increases in liver weight
and serum markers for hepatotoxicity (ALT, AST, OCT, SDH, and bilirubin) (Koporec et al.,
1995; Allis et al., 1990; Bruckner et al., 1986; Condie et al., 1986; Hayes et al., 1986; NCI, 1977,
1976a, b; Weisburger, 1977; Litchfield and Gartland, 1974; Delia Porta et al., 1961;
Eschenbrenner and Miller, 1946; Edwards and Dalton, 1942; Edwards et al., 1942; Edwards,
1941). Liver damage was produced at doses as low as 7-9 mg/kg-day in rats and mice in 90-day
corn oil gavage studies (Table 4-13). The corresponding NOAEL values were 0.7-0.9 mg/kg-
day (Bruckner et al., 1986; Condie et al., 1986). The lowest dose to produce hepatotoxicity in
90-day aqueous gavage studies was 18 mg/kg-day (Koporec et al., 1995).
Table 4-13. Oral toxicity studies for carbon tetrachloride
Species
Dose/duration
NOAEL
(mg/kg-day)
LOAEL
(mg/kg-day)
Effects at the LOAEL
Reference
Subchronic studies
Dog
(6/sex)
Dog
(3F)
Rat
(15-16 M/
group)
Rat
(6M/
group and
sacrifice
time)
Rat
(11 M/
group)
Rat
(11 M/
group)
Mouse
(12/sex/
group)
28 days in gelatin
capsule: 797 mg/kg-
day
8 weeks in gelatin
capsule: 32 mg/kg-day
5 days/week for 12
weeks by corn oil
gavage: 0, 1, 10, or 33
mg/kg-day
5 days/week for 12
weeks by corn oil
gavage: 0, 20, or 40
mg/kg-day; sacrificed
at intervals from 1 to
1 5 days post-exposure
5 days/week for 1 3
week by corn oil
gavage: 0, 25, or 100
mg/kg-day
5 days/week for 1 3
weeks by gavage in
l%Emulphor:0, 25,
or 100 mg/kg-day
5 days/week for 12
weeks by corn oil
gavage: 0, 1.2, 12, or
120 mg/kg-day
Not
determined
32
1 [0.71]a
Not
determined
Not
determined
Not
determined
1.2 [0.86] a
797
Not
determined
10[7.1]a
20 [14.3] a
25 [17.8]a
(PEL)
25 [17. 8] a
(PEL)
12 [8.6] a
Increased ALT, OCT; fatty
vacuolization with single
cell necrosis in liver
No increases in serum
enzymes; no liver
histopathology
Two- to threefold increase
in SDH; mild centrilobular
vacuolization in liver
Increased liver weight,
ALT, AST, LDH; reduced
liver CYP450; cirrhosis,
necrosis, and degeneration
in liver
10% Mortality; increased
ALT, SDH; slight hepato-
cellular vacuolization and
minimal fibrosis in liver
25% Mortality; increased
ALT, SDH; slight hepato-
cellular vacuolization and
minimal fibrosis in liver
Increased ALT; mild to
moderate hepatic lesions
(hepatocytomegaly,
necrosis, inflammation)
Litchfield and
Gartland, 1974
Litchfield and
Gartland, 1974
Bruckner et al.,
1986
Allis et al.,
1990
Koporec et al.,
1995
Koporec et al.,
1995
Condie et al.,
1986
136
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-13. Oral toxicity studies for carbon tetrachloride
Species
Mouse
(12/sex/
group)
Mouse
(20/sex/
group)
Mouse
(5/sex/
group)
Dose/duration
5 days/week for 12
weeks by gavage in
1% Tween-60: 0, 1.2,
12, or 120mg/kg-day
7 days/week for 1 3
weeks by corn oil
gavage: 0, 12, 120,
540, or 1200 mg/kg-
day
30 times in 120 days
by olive oil gavage:
0,40, 80, or 160
mg/kg-day
NOAEL
(mg/kg-day)
12 [8.6] a
Not
determined
40
LOAEL
(mg/kg-day)
120[86]a
12
80
Effects at the LOAEL
Increased liver weight,
ALT, AST, LDH; hepato-
cytomegaly, vacuolation,
inflammation, necrosis,
and fibrosis in liver
Increased liver weight,
ALT, AST, ALP, LDH, 5'-
nucleotidase; fatty change,
hepatocytomegaly,
necrosis, and hepatitis
Necrosis in liver
Reference
Condie et al,
1986
Hayes et al.,
1986
Eschenbrenner
and Miller,
1946
Chronic studies
Rat
(50/sex/
group)
Mouse
(50/sex/
group)
5 days/week for 78
weeks by corn oil
gavage: 0, 47, or 94
mg/kg-day for males;
0,80, or 159 mg/kg-
day for females
5 days/week for 78
weeks by corn oil
gavage: 0, 1250, or
2500 mg/kg-day
Not
determined
Not
determined
47
1,250
(PEL)
Increased mortality;
cirrhosis in liver
Markedly increased
mortality; cirrhosis and
other toxic lesions in liver;
adrenal
pheochromocytoma
NCI, 1977,
1976a,b
NCI, 1977,
1976a,b
Gestational exposure studies
Rat
(29 gravid
F)
Rat
(9-14
gravid F/
group)
Rat
(12-14
gravid F/
group)
Rat
(12-14
gravid F/
group)
2 days on GDs 7-1 1
by corn oil gavage:
478 mg/kg-day
GDs 6-19 by corn oil
gavage: 0, 112.5, or
1 50 mg/kg-day
GDs 6-15 by corn oil
gavage: 0, 25, 50, or
75 mg/kg-day
GDs 6-15 by gavage
in 10%Emulphor: 0,
25, 50, or 75 mg/kg-
day
Not
determined
Not
determined
25
25
478
112.5
50
50
21% Maternal mortality;
59% of dams had no
offspring, 38% because of
full-litter resorption
Reduced maternal weight
gain; markedly increased
full-litter resorption
Piloerection; markedly
increased full-litter
resorption
Piloerection; slightly
increased full-litter
resorption
Wilson, 1954
Narotsky and
Kavlock, 1995
Narotsky et al.,
1997b
Narotsky et al.,
1997b
137
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-13. Oral toxicity studies for carbon tetrachloride
Species
Mouse
(>8 gravid
F/ group)
Dose/duration
GDs 1-5 by gavage in
corn oil: 0,83, or 826
mg/kg-day
NOAEL
(mg/kg-day)
826
LOAEL
(mg/kg-day)
Not
determined
Effects at the LOAEL
No effect on dams or pups
Reference
Hamlin et al,
1993
"Duration adjusted dose provided in brackets (e.g., 1 mg/kg-day x [5 days/week/7 days/week] = 0.71 mg/kg-day).
Subchronic oral studies that also examined nonhepatic endpoints (Bruckner et al., 1986;
Hayes et al., 1986) did not observe effects in the kidneys or other organs. There was some
evidence for impairment of T-cell-dependent immunity in mice treated with 40 mg/kg-day for 14
days but not in rats at hepatotoxic doses (160 mg/kg-day for 10 days) (Guo et al., 2000;
Smialowicz et al., 1991; Kaminski et al., 1990).
There is no direct evidence for effects on reproduction or development in humans
exposed orally to carbon tetrachloride. One epidemiological study (Bove et al., 1995, 1992a, b)
suggested associations between maternal exposure to carbon tetrachloride in drinking water and
adverse birth outcomes (the strongest relationship was for low term birth weight), but subjects
were exposed to multiple chemicals and the study included only a limited characterization of
exposure. Studies in animals have found that relatively high oral doses of carbon tetrachloride
(50 mg/kg-day and above) given on GDs 6-15 produce significant prenatal loss by increasing the
incidence of full-litter resorptions (Narotsky et al.,1997a, b, 1995; Narotsky and Kavlock, 1995;
Wilson, 1954); some evidence exists that reproductive effects are a consequence of a maternally
mediated response to alterations in hormonal levels (Narotsky et al., 1995, 1997a). The doses
producing litter resorption also produced overt toxic effects in dams (piloerection, kyphosis [or
rounded upper back], and marked weight loss) and are well above the LOAELs for liver toxicity
with longer-term exposure. Although the NOAELs and LOAELs were the same, both the
clinical signs and litter resorptions were more pronounced when carbon tetrachloride was
administered in corn oil versus aqueous emulsion. Mice treated with carbon tetrachloride early
in gestation did not show these effects (Hamlin et al., 1993).
Adrenal adenoma and pheochromocytomas were observed in mice exposed to carbon
tetrachloride by gavage in an NCI bioassay in which carbon tetrachloride was used as a positive
control for liver tumors (Weisburger, 1977). These tumors may indicate a potential noncancer
health risk, as well as a cancer risk. Benign pheochromocytomas are tumors that originate in
chromaffin cells of the adrenal gland medulla and secrete excessive amounts of catecholamines,
usually epinephrine and norepinephrine. Because pheochromocytomas are not innervated,
catecholamine secretion is unregulated, producing sustained sympathetic nervous system
hyperactivity leading to hypertension, tachycardia, and cardiac arrhythmias (Hansen, 1998).
138
DRAFT - DO NOT CITE OR QUOTE
-------
Health effects related to pheochromocytoma formation in mice were not assessed in the NCI
(1977) cancer bioassay. Therefore, the potential for secondary effects of pheochromocytoma on
the cardiovascular system can only be inferred. The lowest exposure level associated with
benign pheochromocytomas in mice (LOAEL of 1,250 mg/kg-day, 5 days/week [approximately
900 mg/kg-day]) is approximately 2 orders of magnitude higher than levels at which liver effects
become apparent in experimental animals. Therefore, the available data do not identify the
adrenal gland as a sensitive target organ for carbon tetrachloride by oral administration.
Effect of Dosing Vehicle on Carbon Tetrachloride Toxicity
A number of investigators have demonstrated that the vehicle used in gavage studies to
administer carbon tetrachloride and other chlorinated solvents may affect the test chemical's
toxicity. Several investigators reported that carbon tetrachloride toxicity was enhanced if
administered in corn oil compared to an aqueous solution (Narotsky et al., 1997b; Condie et al.,
1986), whereas Kaporec et al. (1995) found that corn oil as a vehicle (compared to an aqueous
vehicle) did not significantly alter carbon tetrachloride hepatotoxicity following subchronic
exposure, and Kim et al. (1990b) observed that administration in an aqueous solution enhanced
carbon tetrachloride toxicity as compared to corn oil. Raymond and Plaa (1997) and Narotsky et
al. (1997b) found that the influence of vehicle could be dose-dependent. In their study of
developmental toxicity, Narotsky et al. (1997b) reported that maternal toxicity was slightly more
pronounced when carbon tetrachloride was administered in aqueous vehicle, but at higher doses
was more pronounced when administered in corn oil vehicle. Sanzgiri and Bruckner (1997)
found that Emulphor, a polyethoxylated vegetable oil used as an emulsifier for volatile organic
compounds and other lipophilic compounds, had no significant effect on carbon tetrachloride
acute hepatotoxicity in Sprague-Dawley rats (as measured by elevation of serum enzyme
activities of SDH and ALT) when carbon tetrachloride was administered as a single oral doses at
two dose levels (10 and 180 mg/kg) and at four concentrations of Emulphor (1, 2.5, 5 and 10%).
Blood carbon tetrachloride concentrations in these rats (measured at intervals up to 12 hours
postdosing) revealed no significant differences as a function of Emulphor concentration,
suggesting that Emulphor did not significantly affect carbon tetrachloride absorption or
distribution.
A number of explanations of the influence of vehicle on the oral toxicity of carbon
tetrachloride have been offered. Kim et al. (1990b) reported that corn oil delays carbon
tetrachloride absorption from the digestive track and thereby decreases its arterial blood
concentration. Such alterations in carbon tetrachloride pharmacokinetics could influence the
resulting toxicity. It is possible that the preservation state of corn oil might influence toxicity;
cell membranes could be altered by older oil stored under improper conditions and contaminated
with peroxides or by heated and oxygenated corn oil that could lead to the formation of reactive
oxygen radicals (Raymond and Plaa, 1997). It has been proposed that corn oil might induce
13 9 DRAFT - DO NOT CITE OR QUOTE
-------
CYP450 metabolizing enzymes that could enhance metabolism of carbon tetrachloride to
reactive, cytotoxic forms (Raymond and Plaa, 1997; Kaporec et al., 1995). High lipid intake
could possibly increase lipid levels in the liver, thereby enhancing target organ deposition of
lipophilic carbon tetrachloride. Corn oil could also directly affect the lipid composition of cell
membranes; the effects of carbon tetrachloride-derived trichloromethyl free radicals on hepatic
microsomal proteins and lipids might then be enhanced (Kim et al., 1990b).
Kaporec et al. (1995) proposed that a possible explanation for the observation of less
pronounced hepatotoxicity in mice dosed with halocarbons in aqueous media involves method
preparation. Even using methods to minimize carbon tetrachloride loss, Kaporec et al. (1995)
found that there was typically about a 20% loss of carbon tetrachloride from an aqueous
emulsion (Emulphor), but none from corn oil dosing solutions. Thus, findings of less severe
toxicity with an aqueous vehicle than corn oil vehicle may have been the result of animals
receiving a lower daily dose.
Thus, it is possible that the vehicle used in oral gavage studies to administer carbon
tetrachloride could be a potential confounding factor in toxicity assays; however, the magnitude
of the confounding and the nature of the interaction of corn oil remain uncertain.
4.6.2. Inhalation
Case reports of acute high-level exposure to carbon tetrachloride vapor or long-term
occupational exposure provide evidence of hepatotoxic and nephrotoxic effects of carbon
tetrachloride in humans. Observations indicative of an effect on the liver in these cases include
jaundice, increased serum enzyme levels, and, in fatal cases, necrosis of the liver (Stewart et al.,
1965; New et al., 1962; Kazantzis and Bomford, 1960; Norwood et al., 1950). Delayed effects
on the kidney have also been reported in acute overexposure cases. Other effects associated with
carbon tetrachloride exposure in humans are GI symptoms (nausea and vomiting, diarrhea, and
abdominal pain) and neurological effects indicative of central nervous system depression
(headache, dizziness, and weakness). Tomenson et al. (1995) conducted a cross-sectional
epidemiology study of hepatic function in workers exposed to carbon tetrachloride. They found
suggestive evidence of an effect of occupational carbon tetrachloride exposure on serum
enzymes indicative of hepatic effects at workplace concentrations in the range of 1-4 ppm.
The liver and kidney are the most prominent targets of carbon tetrachloride in subchronic
and chronic inhalation studies of laboratory animals. Hepatic toxicity in these studies was
demonstrated by histopathology (centrilobular fatty degeneration, necrosis, fibrosis, cirrhosis,
hepatitis, and regenerative activity) as well as increases in liver weight and serum markers for
liver damage (Nagano et al., 2007a, b; Benson and Springer, 1999; JBRC, 1998; Prendergast et
al., 1967; Adams et al., 1952; Smyth et al., 1936). Hepatic effects were observed in animals
exposed to carbon tetrachloride concentrations as low as 2 ppm (adjusted to continuous
exposure, see Table 4-14). Renal damage was reported less frequently in these animal studies
140 DRAFT - DO NOT CITE OR QUOTE
-------
and generally at higher concentrations than those causing liver damage. The JBRC chronic
bioassay (Nagano et al., 2007b; JBRC, 1998) found renal damage, as evidenced by
histopathology (increased severity of chronic nephropathy in the rat and protein casts in the
mouse) and changes in serum chemistry and urinalysis variables at a concentration of 4 ppm
(adjusted to continuous exposure, see Table 4-14).
Table 4-14. Inhalation toxicity studies for carbon tetrachloride
Species
Duration/
concentration
NOAEL
(ppm)
LOAEL
(ppm)
Effects at the LOAEL
Reference
Subchronic studies
Rat
(24 mixed
sex/group)
Guinea pig
(24 mixed
sex/group)
Monkey
(4/group)
Rat
(15-25/sex/
group)
Guinea pig
(5-9/sex/
group)
Rabbit
(1-2/sex/
group)
Monkey
(1-2/group)
Rat
(15/group)
8 hours/day, 5
days/week for 10.5
months: 0, 50, 100,
200, or 400 ppm
8 hours/day, 5
days/week for 10.5
months: 0, 25, 50,
100, 200, or 400 ppm
8 hours/day, 5
days/week for 10.5
months: 0, 50, or 200
ppm
7 hours/day, 5
days/week for 6
months: 0, 5, 10, 25,
50, 100, 200, or 400
ppm
7 hours/day, 5
days/week for 6
months: 0, 5, 10, 25,
50, 100, 200, or 400
ppm
7 hours/day, 5
days/week for 6
months: 0, 5, 10, 25,
50, or 100 ppm
7 hours/day, 5
days/week for 6
months: 0, 5, 10, 25,
50, or 100 ppm
24 hours/day, 7
days/week for 13
weeks: 0, 1 (in n-
octane), or 10 ppm
Not
determined
Not
determined
Not
determined
5 [If
5 [If
10 [2]a
50 [10]a
1
50 [12]a
25 [6]a
(PEL)
50 [12]a
10 [2]a
10 [2]a
25 [5]a
100 [21]a
10
Fatty change in liver
Increased mortality;
reduced body weight
gain; fatty change in liver
Mild fatty change and
degeneration in liver
Increased liver weight;
fatty degeneration in liver
Increased liver weight;
fatty degeneration in liver
Increased liver weight;
fatty degeneration and
slight cirrhosis in liver
Slight fatty degeneration
and increased lipid
content in liver
Reduced body weight
gain; enlarged liver with
fatty change
Smyth et al.,
1936
Smyth et al.,
1936
Smyth et al.,
1936
Adams et al.,
1952
Adams et al.,
1952
Adams et al.,
1952
Adams et al.,
1952
Prendergast
etal, 1967
141
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14. Inhalation toxicity studies for carbon tetrachloride
Species
Guinea pig
(15/group)
Rabbit
(3/group)
Dog
(2/group)
Monkey
(3/group)
Rat (107
sex/group)
Mouse (107
sex/group)
Rat
(10 M/
group)
Mouse
(10 M/
group)
Hamster
(10 M/
group)
Duration/
concentration
24 hours/day, 7
days/week for 13
weeks: 0, 1 (in n-
octane), or 10 ppm
24 hours/day, 7
days/week for 13
weeks: 0, 1 (in n-
octane), or 10 ppm
24 hours/day, 7
days/week for 13
weeks: 0, 1 (in n-
octane), or 10 ppm
24 hours/day, 7
days/week for 13
weeks: 0, 1 (in n-
octane), or 10 ppm
6 hours/day, 5
days/week for 13
weeks: 0, 10, 30, 90,
270, or 810 ppm
6 hours/day, 5
days/week for 13
weeks: 0, 10, 30, 90,
270, or 810 ppm
6 hours/day, 5
days/week for 12
weeks: 0, 5, 20, or 100
ppm
6 hours/day, 5
days/week for 12
weeks: 0, 5, 20, or 100
ppm
6 hours/day, 5
days/week for 12
weeks: 0, 5, 20, or 100
ppm
NOAEL
(ppm)
1
1
1
1
Not
determined
Not
determined
20 [4]a
5 [0.9]a
20 [4]a
LOAEL
(ppm)
10
10
10
10
10 [2]a
10 [2]a
100 [18]a
20 [4]a
100 [18]a
Effects at the LOAEL
Reduced body weight
gain; enlarged liver with
fatty change; 3 died,
though mortality also
reported in control group
Reduced body weight
gain; enlarged liver with
fatty change
Reduced body weight
gain; fatty change in liver
Visibly emaciated;
enlarged liver with fatty
change
Increased liver weight;
fatty change in liver
Slight cytological
alterations in the liver
Increased ALT, SDH;
necrosis in liver
Increased ALT, SDH;
necrosis and cell
proliferation in liver
Increased ALT, SDH;
necrosis and cell
proliferation in liver
Reference
Prendergast
etal, 1967
Prendergast
etal., 1967
Prendergast
etal., 1967
Prendergast
etal., 1967
Nagano et al,
2007a;
JBRC, 1998
Nagano et al.,
2007a;
JBRC, 1998
Benson and
Springer,
1999
Benson and
Springer,
1999
Benson and
Springer,
1999
142
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-14. Inhalation toxicity studies for carbon tetrachloride
Species
Duration/
concentration
NOAEL
(ppm)
LOAEL
(ppm)
Effects at the LOAEL
Reference
Chronic studies
Rat
(50/sex/
group)
Mouse
(50/sex/
Group)
6 hours/day, 5
days/week for 104
weeks: 0, 5, 25, or 125
ppm
6 hours/day, 5
days/week for 104
weeks: 0, 5, 25, or 125
ppm
5 [0.9]a
5 [0.9]a
25 [4]a
25 [4]a
Reduced body weight
gain; increased AST,
ALT, LDH, GPT, BUN,
CPK; lesions in the liver
(fatty changes, fibrosis,
cirrhosis) and kidney
(progressive
glomerulonephrosis)
Reduced survival late in
study (because of liver
tumors); reduced body
weight gain; increased
ALT, AST, LDH, ALP,
protein, total bilirubin,
and BUN; decreased
urinary pH; increased
liver weight; lesions in
the liver (degeneration),
spleen (extra medullary
hematopoiesis), and
kidney (protein casts);
benign
pheochromocytoma
(males)
Nagano et al,
2007b;
JBRC, 1998
Nagano et al.,
2007b;
JBRC, 1998
Gestational exposure study
Rat
(22-23
gravid
F/group)
7 hours/day on GDs
6-15:0, 334, or 1004
ppm
Not
determined
334 [97]a
Dam: reduced body
weight; increased liver
weight and ALT; altered
gross appearance of liver
Fetus: reduced body
weight and crown-rump
length
Schwetz et
al., 1974
aDuration adjusted concentration is provided in brackets (e.g., 10 ppm x [6 hours/day ^ 24 hours/day x 5
days/week ^ 7 days/week] = 2 ppm).
In the subchronic studies, effects on the kidneys were generally observed at
concentrations above the LOAEL for liver effects and thus are not listed in Table 4-14. With
chronic exposure, the sensitivity of the kidney and liver as target organs are comparable in the
rodent. The JBRC chronic rat study (Nagano et al., 2007b; JBRC, 1998) reported liver toxicity
(serum enzyme changes, fatty liver, fibrosis, cirrhosis) and kidney toxicity (increases in BUN,
creatinine, inorganic phosphorus, and severity of chronic progressive nephropathy [CPN]) at
exposure concentrations of 25 ppm (>4 ppm, duration-adjusted) (Table 4-14). An increase in the
143
DRAFT - DO NOT CITE OR QUOTE
-------
severity of proteinuria was reported in male and female rats at the lowest tested concentration of
5 ppm (0.9 ppm, duration adjusted). While the increased severity of proteinuria could be related
to the nephropathy observed at >25 ppm, the biological significance of the finding of proteinuria
at 5 ppm is unknown. Proteinuria (or protein in the urine) was found in essentially 100% of the
rats (both control and carbon tetrachloride-exposed), and 90% or more of all rats (again control
and carbon tetrachloride-exposed) had protein content in the urine graded as either 3+ or 4+ (see
Table 4-2). In the carbon tetrachloride-exposed animals, however, rats showed an increase in the
severity of proteinuria relative to controls (i.e., relatively more carbon tetrachloride-exposed
animals had protein content in urine graded 4+ than 3+). After 2 years of exposure to carbon
tetrachloride, proteinuria in 5-ppm rats did not progress, i.e., rats at this concentration did not
show treatment-related increases in incidence or severity of renal changes recognized as clearly
adverse (e.g., progressive glomerulonephrosis [or CPN] or measures of impaired glomerular
function, including increased levels of BUN, creatinine, and inorganic phosphorus) that were
observed at higher exposure concentrations.
Complicating interpretation of kidney effects in this study is the fact that the F344 rat is
known for its high incidence of spontaneous, age-related CPN (Hard and Seely, 2005; Chandra
and Frith, 1993/94). Chandra and Firth (1993/94) reported a background incidence of CPN of
88.8% in male and 74.5% in female F344 rats based on an examination of 491 controls from
several 2-year carcinogenicity/chronic toxicity bioassays. CPN can be seen as early as 3 months
and severity of the lesion increases with age. The presence of CPN can confound kidney lesion
diagnosis (Hard and Seely, 2005). Kidney lesions in the JBRC 13-week study of carbon
tetrachloride (Nagano et al., 2007a; JBRC, 1998) were examined with the thought that the
confounding encountered in older (2-year-old) rats would be minimized and treatment-related
lesions could be more easily distinguished from spontaneous old-age renal lesions. In the 13-
week study, the severity of proteinuria was statistically significantly increased at a concentration
of >90 ppm in females and >270 ppm in males; histopathological changes in the kidney occurred
in both sexes at >810 ppm. These effect levels are approximately 20- to 50-fold higher than the
5-ppm concentration in the chronic study at which an increase in severity of proteinuria was
observed. It is unexpected that the effect level for kidney effects would decrease by such a large
margin between subchronic and chronic exposure durations. Thus, the findings from the
subchronic study by JBRC (Nagano et al., 2007a; JBRC, 1998) are not clearly consistent with a
LOAEL for renal toxicity following chronic exposure of 5 ppm. Finally, the body of literature
for carbon tetrachloride suggests that the rat liver is a more sensitive target organ than the kidney
following exposures of subchronic duration (e.g., Nagano et al., 2007a; JBRC, 1998; Bruckner et
al., 1986; Adams et al., 1952). There are no adequate chronic studies of carbon tetrachloride
(beyond JBRC, 1998) to confirm whether the kidney may be a more sensitive target organ than
the liver following chronic exposure. The above uncertainties raise questions as to the relevance
of the finding of proteinuria in 5-ppm rats to human health assessment.
144 DRAFT - DO NOT CITE OR QUOTE
-------
In addition to adverse effects on the liver and kidney, the observation of benign
pheochromocytomas in mice exposed to carbon tetrachloride by inhalation in the JBRC chronic
study (Nagano et al., 2007b; JBRC, 1998) may indicate a potential noncancer health risk. As
noted in Section 4.6.1, benign pheochromocytomas are tumors that originate in chromaffin cells
of the adrenal gland medulla and secrete excessive amounts of catecholamines, usually
epinephrine and norepinephrine. Because pheochromocytomas are not innervated,
catecholamine secretion is unregulated, producing sustained sympathetic nervous system
hyperactivity leading to hypertension, tachycardia, and cardiac arrhythmias (Hansen, 1998).
Health effects related to pheochromocytoma formation in mice were not assessed in the JBRC
chronic inhalation exposure study. Therefore, the potential for secondary effects of
pheochromocytoma on the cardiovascular system can only be inferred. Exposure levels
associated with benign pheochromocytomas in mice (LOAELs of 4 and 22 ppm, duration-
adjusted, in male and female mice, respectively) were equal to or greater than levels associated
with hepatic and renal toxicity; thus, the adrenal gland is not the most sensitive target organ for
carbon tetrachloride following inhalation exposure.
There is no evidence for reproductive or developmental toxicity in humans exposed by
inhalation to carbon tetrachloride. One epidemiological study found no association between
maternal occupational exposure to carbon tetrachloride and infants born small for gestational age
(Seidler et al., 1999). Carbon tetrachloride has been found to produce effects in mouse testis
(Bergman, 1983), testicular atrophy, and reduced fertility in rats exposed intermittently to high
concentrations (>200 ppm) for 6 or more months (Adams et al., 1952; Smyth et al., 1936).
Testicular degeneration has also been reported in rats following repeated i.p. doses of 1.5 mL/kg
(Kalla and Bansal, 1975; Chatterjee, 1966). A definitive reproductive toxicity study has not been
performed, however. In a developmental toxicity study, Schwetz et al. (1974) found significant
reductions in fetal body weight and crown-rump length in rats exposed to carbon tetrachloride
vapor in the air during gestation but at a high concentration (334 ppm, 7 hours/day) that also
produced hepatotoxicity and reduced growth in the dams.
4.6.3. Mode of Action Information
The MO A of carbon tetrachloride-induced hepatotoxicity has been the subject of
extensive research. Mechanistic studies (described in Section 4.5) provide evidence that
metabolism of carbon tetrachloride via CYP2E1 to highly reactive free radical metabolites plays
a role in its MOA (Wong et al., 1998; Martinez et al., 1995; Letteron et al., 1990; Mourelle et al.,
1988; Bechtold et al., 1982; Weddle et al., 1976). The primary metabolites, trichloromethyl and
trichloromethyl peroxy free radicals, are highly reactive and are capable of covalently binding to
cellular macromolecules (Boll et al., 2001b; Azri et al., 1991; DiRenzo et al., 1982; Diaz Gomez
and Castro, 1980a; Castro and Diaz Gomez, 1972; Gordis, 1969). Because the toxicity of carbon
tetrachloride is secondary to its metabolism, the liver is expected to be an important target organ
145 DRAFT - DO NOT CITE OR QUOTE
-------
on the basis of its high CYP2E1 content.
The trichloromethyl peroxy and trichloromethyl radical may induce multiple cellular
effects including lipid peroxidation (de Zwart et al., 1997; Gasso et al., 1996; Ichinose et al.,
1994; Tribble et al., 1987; Lee et al., 1982; Recknagel and Glende, 1973; Rao and Recknagel,
1969) decreases in antioxidant levels (Cabre et al., 2000; Gasso et al., 1996 Gorla et al., 1983),
alterations in calcium homeostasis, and activation of calcium dependent phospholipases as
discussed in Section 4.5 (Limaye et al., 2003; Hemmings et al., 2002; Gonzalez Padron et al.,
1993; Agarwal and Mehendale, 1986, 1984; Long and Moore, 1986; Kroner, 1982; Moore et al.,
1976). Additionally, products of lipid peroxidation include reactive aldehydes that can form
protein adducts that may contribute to hepatotoxicity (Beddowes et al., 2003; Abraham et al.,
1999; Hartley et al., 1999; Bedossa et al., 1994; Comporti, 1985; Comporti et al., 1984). At this
time, the exact sequence or contribution of cellular mechanisms leading from the key event of
metabolism to carbon tetrachloride-induced hepatotoxicity (cell death) is uncertain. A
description of how carbon tetrachloride-induced noncancer effects may coincide with the
hypothesized carcinogenic MOA can be found in Figure 4-4.
Although most mechanistic studies for carbon tetrachloride have concentrated on hepatic
effects, some studies provide evidence for a similar MOA for noncancer effects in the kidney.
The distribution study of Bergman (1983) provided evidence that nonvolatile metabolites of
carbon tetrachloride accumulate in the kidney as well as the liver of mice immediately following
a 10-minute inhalation exposure (see Section 3.2). Like the liver, the kidney contains both
CYP2E1 and CYP3 A, which are able to metabolize carbon tetrachloride to the trichloromethyl
radical (Warrington et al., 2004; Koch et al., 2002; Haehner et al., 1996). Histopathological
examination in multiple studies revealed clear evidence of treatment-related glomerular damage
(increased in severity of glomerulonephrosis, BUN, proteinuria, tubular degeneration, organ
weight, and protein casts) in male and female rats exposed to carbon tetrachloride (Nagano et al.,
2007a, b; Benson and Springer, 1999; JBRC, 1998; Prendergast et al., 1967; Adams et al., 1952;
Smyth et al., 1936). Mechanistic similarities also exist between the liver and kidney regarding
increases in lipid peroxidation products (Natarajan et al., 2006; Dogukan et al., 2003; Abraham
et al., 1999; Fraga et al., 1987), reductions in GSH peroxidase activity, attributable to depleted
stores of GSH (Natarajan et al., 2006; Dogukan et al., 2003; Ozturk et al., 2003) and increased
levels of cytosolic phospholipase A2 (Niederberger et al., 1998). Based on the available data, the
kidney and liver effects associated with carbon tetrachloride appear to operate via a similar MOA
pathway.
4.7. EVALUATION OF CARCINOGENICITY
4.7.1. Summary of Overall Weight-of-Evidence
Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), carbon
tetrachloride can be classified as likely to be carcinogenic to humans by all routes of exposure.
146 DRAFT - DO NOT CITE OR QUOTE
-------
This cancer weight of evidence determination is based on (1) inadequate evidence of
carcinogenicity in humans and (2) sufficient evidence in animals (i.e., hepatic tumors in multiple
species (rat, mouse, and hamster) and by oral and inhalation routes of exposure in response to
carbon tetrachloride and evidence of pheochromocytomas in mice by oral and inhalation routes
of exposure).
Carbon tetrachloride has been shown to be a liver carcinogen in rats, mice, and hamsters
in eight bioassays of various experimental design by oral and inhalation exposure. A general
correspondence has been observed between hepatocellular cytotoxicity and regenerative
hyperplasia and the induction of liver tumors. At lower exposure levels, this correspondence is
less consistent. In particular, in the JBRC 2-year inhalation cancer bioassay in the mouse
(Nagano et al., 2007b, JBRC, 1998), the lowest exposure concentration tested (5 ppm [0.9 ppm
adjusted]; see Tables 4-5 and 4-6) was not hepatotoxic, whereas the incidence of liver adenomas
in female mice at this exposure concentration displayed a statistically significant increase
compared to concurrent and historical controls.
A hypothesized carcinogenic MOA for carbon tetrachloride-induced liver tumors has
been proposed and includes the following key events: (1) metabolism to the trichloromethyl
radical by CYP2E1 and subsequent formation of the trichloromethyl peroxy radical, (2) radical-
induced mechanisms leading to hepatocellular cytotoxicity, and (3) sustained regenerative and
proliferative changes in the liver in response to hepatotoxicity. A substantial amount of data
exists that supports these hypothesized key events in the cancer MOA for carbon tetrachloride.
Data to characterize these key events at low-exposure levels, however, are limited. This is of
particular concern for liver tumor MOA considerations in light of (1) the finding that liver
tumors in female mice occurred at non-cytotoxic doses (Nagano et al., 2007b; JBRC, 1998) and
(2) the fundamental reactivity of direct and indirect products of carbon tetrachloride metabolism.
Therefore, the MOA of carbon tetrachloride at low exposure levels can be hypothesized, but is
unknown at this time. Hypothesized MOAs are discussed further in Sections 4.7.3 and 4.7.4
below.
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence
Studies in humans are inadequate to show an association between exposure to carbon
tetrachloride and carcinogenicity. There is some evidence for certain types of cancer in
occupational populations thought to have had some exposure to carbon tetrachloride, including
NHL (Blair et al., 1998; Spirtas et al., 1991), lymphosarcoma and lymphatic leukemia
(Checkoway et al., 1984; Wilcosky et al., 1984), esophageal and cervical cancer (Blair et al.,
1990, 1979), breast cancer (Cantor et al., 1995), astrocytic brain cancer (Heineman et al., 1994),
and rectal cancer (Dumas et al., 2000). In these cases, exposure to carbon tetrachloride was
poorly characterized and confounded by simultaneous exposures to other chemicals.
Additionally, these studies were designed to evaluate tetrachloroethylene and trichloroethylene
147 DRAFT - DO NOT CITE OR QUOTE
-------
and had only limited ability to examine other chemical exposures such as carbon tetrachloride.
None of the human epidemiology studies reported associations to cancer of the liver, which is the
main site of carcinogenicity in animal studies, but this may be because of a lack of power to
detect a relatively rare human tumor.
Carbon tetrachloride has been shown to induce hepatocellular carcinomas in rodents by
oral, inhalation, and parenteral exposure. Researchers at the NCI conducted a series of gavage
studies in mice of various strains and found large increases in the incidence of liver tumors in
treated mice (Andervont, 1958; Edwards and Dalton, 1942; Edwards et al., 1942; Edwards,
1941). A similar result was obtained in hamsters (Delia Porta et al., 1961). These animal studies
were generally conducted using a single high dose of carbon tetrachloride, but one early study
was conducted with multiple dose levels in order to investigate dose-response relationships for
induction of liver tumors (Eschenbrenner and Miller, 1946). This study was conducted using
small groups of five mice of each sex per group and two dosing regimens (gavage administration
in olive oil daily or every 4 days for 4 months) that gave the same total exposure. Liver tumors
(hepatomas) were found in all strain A male and female mice that received average daily doses
as low as 20 mg/kg-day. No gross or microscopic tumors were found in mice receiving only 10
mg/kg-day.
Oral bioassays of carbon tetrachloride using groups of 50 animals/sex were conducted in
mice and rats by NCI (1977, 1976a, b) as a positive control for bioassays of chloroform,
trichloroethylene, and 1,1,1-trichloroethane. The bioassay in mice employed very high doses
(1,250 or 2,500 mg/kg, 5 day/week for 78 weeks) that produced close to 100% incidence of
hepatocellular carcinoma. The incidence of adrenal adenoma and pheochromocytoma was also
significantly increased in both dose groups in male and female mice. The bioassay in rats (47 or
94 mg/kg for males and 80 or 159 mg/kg for females, 5 days/week for 78 weeks) produced only
a low incidence of liver tumors, but high early mortality, particularly in the high-dose group,
may have affected the power of this study to detect a carcinogenic effect. Even so, the increase
in carcinomas was statistically significant in low-dose females (4/49) in relation to pooled
controls (1/99).
Carbon tetrachloride induced tumors in an inhalation bioassays in rats and mice (Nagano
et al., 2007b; JBRC, 1998). In rats, intermittent exposure (6 hours/day, 5 days/week) to 125 ppm
for 2 years produced marked significant increases in the incidence of hepatocellular carcinomas
and adenomas in both males and females. The incidence of tumors was not increased in rats
exposed to 5 or 25 ppm by the same protocol although the incidence of liver carcinoma (3/50) in
25-ppm females exceeded the range of historical control incidence from JBRC bioassays. In
mice, marked significant increases in hepatocellular carcinomas and (to a lesser extent)
adenomas occurred at both 25 and 125 ppm in both sexes. Also, a statistically significant
increase in the incidence of liver adenomas in female mice at 5 ppm (0.9 ppm adjusted) was
observed compared to the concurrent control and exceeded the historical control range for
148 DRAFT - DO NOT CITE OR QUOTE
-------
hepatocellular adenomas from JBRC 2-year bioassays. The assays in mice also found significant
increases in the incidence of benign adrenal pheochromocytomas in males at 25 or 125 ppm and
females at 125 ppm, exposure levels at or above those associated with liver hepatocellular
carcinoma and adenoma. Specifically, pheochromocytomas were identified in 32/50 high-dose
male mice, only one of which was classified as malignant (the remaining 31 pheochromocytomas
were benign) (JBRC, 1998). Benign pheochromocytomas were identified in 22/49 high-dose
female mice; none were malignant. In addition to the potential cancer risk suggested by these
tumors, benign pheochromocytomas may represent a noncancer health risk because of the
excessive secretion of catecholamines, leading to sustained and unregulated sympathetic nervous
system hyperactivity (see Section 4.6.2).
Some data from parenteral studies are also available. Subcutaneous injections of carbon
tetrachloride at an average dose of 0.29 mg/kg-day for 33-47 weeks induced hepatocellular
carcinomas in Osborne-Mendel, Japanese, and Wistar rats but not in Sprague-Dawley or black
rats (Reuber and Glover, 1970, 1967a, b). Intraperitoneal injections at an average of 86 mg/kg-
day induced hepatomas in C3H mice (Kiplinger and Kensler, 1963).
Carbon tetrachloride has been extensively studied for its genotoxic and mutagenic effects.
Overall, results are largely negative. There is little direct evidence that carbon tetrachloride
induces intragenic or point mutations in mammalian systems (Section 4.4.2). The mutagenicity
studies that have been performed using transgenic mice have yielded negative results, as have the
vast majority of the mutagenesis studies that have been conducted in bacterial systems.
However, since oxidative DNA adducts can be converted into mutations, the inability to detect
mutations in the transgenic mouse assays may be an indication of efficient repair of oxidative
lesions, a preferential formation of large chromosomal mutations that are inefficiently detected in
the transgenic models, or a reflection of the limitations and sensitivity of the specific assays that
were performed with carbon tetrachloride (see Table 4-12). The two positive mutation / DNA
damage studies conducted in E. coli were seen in strains that are particularly sensitive to
oxidative damage. Moreover, the intrachromosomal recombination induced by carbon
tetrachloride in S. cerevisiae is believed to result from double stranded DNA breaks leading to
deletion mutations. These results are consistent with DNA breakage originating from oxidative
stress or lipid peroxidation products that occur concurrently with cytotoxicity.
An evaluation based on the weight of evidence suggests that carbon tetrachloride is more
likely an indirect than a direct mutagenic agent. In general, genotoxic effects have been
observed in a consistent and close relationship with cytotoxicity, lipid peroxidation, and/or
oxidative DNA damage. Mutagenic effects, if they occur, are likely to be generated through
indirect mechanisms resulting from oxidative stress or lipid peroxidation products. Under highly
cytotoxic conditions, bioactivated carbon tetrachloride can exert genotoxic effects. These tend to
be modest in magnitude and are manifested primarily as DNA breakage and related sequelae.
Chromosome loss leading to aneuploidy may also occur to a limited extent.
149 DRAFT - DO NOT CITE OR QUOTE
-------
Challenges in evaluating the carbon tetrachloride genotoxicity database must be
acknowledged (e.g., see Table 4-12). Although the cellular effects of carbon tetrachloride are
described adequately at doses at or above those that induce cytotoxicity, there is a paucity of data
describing DNA damaging events at doses below those that are cytotoxic. Additionally, there
exists some level of uncertainty as to whether assays used to assess the genotoxicity of carbon
tetrachloride were of sufficient quality to assess genotoxicity at doses that do not induce
cytotoxicity.
4.7.3. Mode of Action Information for Liver Tumors
4.7.3.1. Hypothesized Mode of Action and Identification of Key Events
Carbon tetrachloride produced liver tumors in rats, mice, and hamsters in studies using
various experimental designs by oral and inhalation exposure (Nagano et al., 2007b; JBRC,
1998; NCI, 1977, 1976a, b; Delia Porta et al., 1961; Andervont, 1958; Edwards and Dalton,
1942; Edwards et al., 1942; Edwards, 1941). A hypothesized MOA for carbon tetrachloride-
induced liver tumors is described graphically in Figure 4-4.
Hypothesized Key Events
Hypothesized key events in the carcinogenicity of carbon tetrachloride include:
(1) metabolism to the trichloromethyl radical by CYP2E1 and subsequent formation of the
trichloromethyl peroxy radical, (2) radical-induced mechanisms leading to hepatocellular
toxicity, and (3) sustained regenerative and proliferative changes in the liver in response to
hepatotoxicity.
Metabolism of carbon tetrachloride is identified as a key event based on the following:
(1) reactive metabolites are present in the liver (Stoyanovsky and Cederbaum, 1999; Conner et
al., 1986), (2) CYP450 inhibitors prevent carbon tetrachloride-induced liver damage (Martinez et
al., 1995; Letteron et al., 1990; Mourelle et al., 1988; Bechtold et al., 1982; Weddle et al., 1976),
(3) treatment of knockout mice specific for CYP2E1 (cyp2erA) with carbon tetrachloride does not
result in hepatocellular cytotoxicity as compared to wild type (cyp2el+/+) mice, and (4) treatment
with compounds that induce CYP450s result in potentiating effects to carbon tetrachloride-
induced toxicity (Section 4.8.6).
The resulting hepatocellular toxicity has been demonstrated in numerous studies (Table
4-15) as measured by increases in liver enzymes (i.e., ALT, AST, SDH, and LDH) in plasma or
by histopathological examination. As a result of cytotoxicity in the liver of carbon tetrachloride-
treated animals, significant regenerative cellular proliferation occurs to compensate for the
necrotic or damaged tissue. As discussed in Section 4.7.2, there is a general correlation
(particularly at higher doses) between occurrence of hepatotoxicity and/or
regenerative/proliferative lesions and development of tumors. Findings from the study by JBRC
(Nagano et al., 2007b; JBRC, 1998), the only detailed study of both chronic toxicity and
150 DRAFT - DO NOT CITE OR QUOTE
-------
carcinogenicity of carbon tetrachloride available, are generally consistent with the hypothesis
that liver tumors occur at exposure levels that produced hepatotoxicity in both rats and mice.
Tumorigenesis through this hypothesized MOA resulting from carbon tetrachloride-induced
toxicity is believed to require persistent hepatocellular cytotoxicity and regenerative cellular
proliferation for tumor formation.
Key Events
Chronic exposure
to CCI4
Metabolism of
CCI4 by CYP2E1
I
Hepatocellular
cytotoxicity
Hepatocellular
regenerative
proliferation
Liver tumors
Mechanistic Events /
Areas of Uncertainty
Lipid
peroxidation
Disruption of
Ca" homeostasis
Genetic damage
or alterations
A.
B.
Demonstrated at high
doses compared to chronic
bioassays (g/kg vs mg/kg)
Level of contribution to
cytotoxicity unknown?
A.
B.
Demonstrated at lower
levels in vitro (0.5 mtvl)
Unknown cause or effect of
cytotoxicity?
D.
Background mutational
events
Reactive aldehyde-induced
mutations (HNE, MDA)
Cytotoxi city-induced
damage
Oxidative stress (8-OH-dG)
Figure 4-4. Hypothesized carcinogenic MOA
Other Mechanistic Events Hypothesized to Contribute to Liver Tumor Induction
Other biological events, including lipid peroxidation, disturbances in calcium
homeostasis, and genetic damage, are possibly involved in the induction of liver tumors by
carbon tetrachloride; however, the contribution of these events is not known. Therefore, whether
these mechanistic events represent key events in carbon tetrachloride's carcinogenicity is
unknown.
In general, mechanistic studies of carbon tetrachloride-induced lipid peroxidation have
been conducted at doses that induce significant levels of cytotoxicity (see Table 4-15).
151
DRAFT - DO NOT CITE OR QUOTE
-------
Representative studies evaluating the occurrence of lipid peroxidation are provided in
Table 4-16. These studies do not adequately characterize cellular responses that may occur at
exposures below those that induced tumors in chronic bioassays. Additionally, it is not clear at
what dose lipid peroxidation or generation of reactive aldehydes would begin to contribute to the
other effects such as cytotoxicity or genotoxicity of carbon tetrachloride. Although carbon
tetrachloride is not considered likely to be directly genotoxic, it is possible that lipid peroxidation
products generate compounds (reactive aldehydes) that may covalently bind to DNA. The low
MW aldehydes generated by lipid peroxidation have sufficiently long biological ti/2 to diffuse
from their site of formation to other parts of the cell (Slater, 1982, 1981). Nuclear DNA adducts
to these aldehydes in hepatocytes have been demonstrated in a number of studies (Beddowes et
al., 2003; Wacker et al., 2001; Chung et al., 2000; Wang and Liehr, 1995; Chaudhary et al.,
1994). One of these compounds, malonaldehyde, has been shown to be tumorigenic in Swiss
mice when applied repeatedly to the skin (Shamberger et al., 1974). In cultured rat hepatocytes,
however, the lowest concentration producing a statistically significant increase in DNA breaks
and DNA adducts generated by lipid peroxidation approached the concentration that induced
cytotoxicity (LDH leakage) (Beddowes et al., 2003). The possibility exists that reactive
aldehydes generated at low levels of carbon tetrachloride could result in increased levels of
endogenous MDA and 4-HNE DNA adducts that may contribute to the genotoxicity of carbon
tetrachloride. Additionally, based on current data sets that characterize the generation of lipid
peroxidation induced by carbon tetrachloride (Table 4-16), the doses at which this effect has
been demonstrated do not allow for a determination as to whether lipid peroxidation induces
cytotoxicity or whether cytotoxicity induces lipid peroxidation.
Disruption of calcium homeostasis as a process by which carbon tetrachloride may
induce toxicity is an area of extensive research (Hemmings et al., 2002; Long and Moore, 1986;
Kroner, 1982; Moore et al., 1976). Similar to research conducted on carbon tetrachloride-
induced lipid peroxidation, it is not established if disruption of calcium homeostasis is a cause or
an effect of cellular cytotoxicity or other cellular events hypothesized to contribute to
tumorigenicity. Some studies present evidence that disturbances in calcium homeostasis may not
be a necessary event for cell death (Albano et al., 1989; Clawson, 1989). Similarly, evaluation
of the carbon tetrachloride dose required to induce disturbances in calcium homeostasis does not
confirm this as a key event (Hemmings et al., 2002; Long and Moore, 1986; Kroner, 1982;
Moore etal., 1976).
The role of genetic damage or alteration to DNA in the cancer MOA(s) for carbon
tetrachloride has not been adequately characterized. Several cellular processes have been
proposed that may account for how genetic damage may occur, ultimately leading to genotoxic
events. The trichloromethyl and trichloromethyl peroxy free radicals are capable of covalently
binding to nucleic acids. The reactivity of these radicals, however, is such that they are not
expected to diffuse very far from their site of formation (Slater, 1982, 1981). As a result, the
152 DRAFT - DO NOT CITE OR QUOTE
-------
amount reaching the cell nucleus from microsomes would be negligible. Studies have indicated
small increases in covalent binding of trichloromethyl radical to nuclear DNA, as well as nuclear
proteins and lipids, as a result of bioactivation of carbon tetrachloride by CYP450 in the nuclear
membrane (Fanelli and Castro, 1995; Castro et al., 1989; Levy and Brabec, 1984; Diaz Gomez
and Castro, 1980a, b; Rocchi et al., 1973). There are significant methodological problems with
each of these studies, however, that confound interpretation of the results (see Section 4.4.2.4).
Additionally, the fact that carbon tetrachloride overall has not been found to be a potent mutagen
and that the few positive genotoxic results are found only at high exposure levels and generally
in concert with cytotoxic effects (Tables 4-8 to 4-11) indicates that carbon tetrachloride does not
likely induce genotoxic effects through direct binding or damage to DNA. Development of
mutations by lipid peroxidation-induced DNA damage could occur and would likely result from
the production of radicals exceeding the cell's capacity to quench and/or repair these alterations.
Genetic damage could also result from background or spontaneous mutations. In vivo
studies have estimated that background mutation frequencies may increase many fold over the
lifetime of an organism (Morley and Turner, 1999). It is generally accepted that sustained cell
proliferation in response to cell death from toxicity or other causes is a significant risk factor for
cancer (Holsapple et al., 2006). Thus, hepatic regeneration following injury from carbon
tetrachloride has the potential to result in carcinogenesis as a result of replication errors
becoming fixed mutations before DNA repair can be completed.
Many studies have characterized the formation of endogenously produced DNA adducts
(Beddowes et al., 2003; Wacker et al., 2001; Chung et al., 2000; Wang and Liehr, 1995;
Chaudhary et al., 1994), DNA strand breaks (Kadiiska et al., 2005; Yasuda et al., 2000; Gans and
Korson, 1984), chromosomal aberrations (Sawada et al., 1991), and MN formation (Uryvaeva
and Delone, 1995; Van Goethem et al., 1995). However, to date, measurement of genetic
damage to DNA has not been well characterized at or below doses at which tumors are observed
(Nagano et al., 2007b; JBRC, 1998; NCI, 1977, 1976a, b; Eschenbrenner and Miller, 1946).
Adequate dose-response studies for assays that measure genetic damaging events at or below
dose levels for which carbon tetrachloride induces tumors in chronic bioassays would help
clarify whether or not carbon tetrachloride is carcinogenic at dose levels that do not cause
cytotoxicity and cell regeneration.
4.7.3.2. Experimental Support for the Hypothesized Mode of Action
4.7.3.2.1. Strength, consistency, specificity of association. Carcinogen! city studies of carbon
tetrachloride have consistently reported carcinogenicity in the liver, independent of species,
gender, or route of administration (Nagano et al., 2007b; NCI 1977, 1976a, b; Delia Porta et al.,
1961; Andervont, 1958; Eschenbrenner and Miller, 1946; Edwards and Dalton, 1942; Edwards et
al., 1942; Edwards, 1941). Hepatic toxicity (cytotoxicity), necrosis, and regenerative
proliferation have generally been reported in animals exposed to carbon tetrachloride orally or by
153 DRAFT - DO NOT CITE OR QUOTE
-------
inhalation and are correlated with the CYP450 content. Table 4-15 shows the necrotic and
regenerative lesions observed in subchronic and chronic oral and inhalation studies of carbon
tetrachloride (only studies explicitly reporting necrotic or regenerative lesions are included). In
these studies, hepatic necrosis or degeneration was usually found in conjunction with some type
of proliferative lesion, either regenerative hepatocellular changes (Nagano et al., 2007a, b;
Benson and Springer, 1999; JBRC, 1998; Prendergast et al., 1967) or proliferation or hyperplasia
of the bile duct (Nagano et al., 2007a; JBRC, 1998; Koporec et al., 1995; Bruckner et al., 1986;
Hayes et al., 1986; NCI, 1977, 1976a, b; Prendergast et al., 1967).
Table 4-15. Exposure levels for necrosis/degeneration and hyperplasia/
regeneration in liver following subchronic or chronic exposure to carbon
tetrachloride by gavage or inhalation
Strain, species
Sprague-Dawley
rat (male)
F344 rat (male)
Sprague-Dawley
rat (male)
CD-I mouse
CD-I mouse
Strain A mouse
B6C3F1 mouse
F344 rat (male)
B6C3F1 mouse
(male)
Syrian hamster
(male)
Wistar rat
Hartley guinea
Pig
Hartley guinea
pig; Sprague-
Dawley or Long-
Evans rat
Exposure
Oral, 12 weeks
24 mg/kg-day (adjusted)
Oral, 12 weeks
14 mg/kg-day (adjusted)
Oral, 1 3 weeks
71 mg/kg-day (adjusted)
Oral, 1 3 weeks
12 mg/kg-day
Oral, 12 weeks
8.6 mg/kg-day (adjusted)
Oral, 120 days (30 doses)
80 mg/kg-dayb
Oral, 78 weeks,
892 mg/kg-dayb (adjusted)
Inhalation, 12 weeks
18 ppm (adjusted)3
Inhalation, 12 weeks
4 ppm (adjusted)3
Inhalation, 12 weeks
18 ppm (adjusted)3
Inhalation, 6 months
42 ppm
Inhalation, 1 3 weeks
10 ppm (continuous)
Inhalation, 6 weeks
20 ppm (adjusted)3
Hepatic necrosis/
degeneration
Necrosis
Necrosis
Necrosis
Necrosis
Necrosis
Necrosis
Necrosis
Necrosis
Necrosis
Necrosis
Hepatocellular
degeneration
Necrosis,
hepatocellular
degeneration
Hyperplasia/
regeneration
Bile duct hyperplasia
Nodular hepatic, bile
duct, and oval cell
hyperplasia
Bile duct hyperplasia
Bile duct proliferation
BrdU-negative
hepatocytes
BrdU-positive
hepatocytes
BrdU-positive
hepatocytes
Hepatocellular
regeneration
Hepatocellular
regeneration, bile duct
proliferation
Reference
Bruckner et al.,
1986
Allisetal, 1990
Koporec et al.,
1995
Hayes et al.,
1986
Condie et al.,
1986
Eschenbrenner
and Miller, 1946
NCI, 1977,
1976a,b
Benson and
Springer, 1999
Benson and
Springer, 1999
Benson and
Springer, 1999
Adams et al.,
1952
Prendergast et
al., 1967
Prendergast et
al., 1967
154
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-15. Exposure levels for necrosis/degeneration and hyperplasia/
regeneration in liver following subchronic or chronic exposure to carbon
tetrachloride by gavage or inhalation
Strain, species
F344 rat
BDF1 mouse
F344 rat
BDF1 mouse
Exposure
Inhalation, 1 3 weeks
2 ppm (adjusted)3
Inhalation, 1 3 weeks
5-48 ppm (adjusted)3
Inhalation, 104 weeks
5-22 ppm (adjusted)3
Inhalation, 104 weeks
5 ppmb (adjusted)3
Hepatic necrosis/
degeneration
Degeneration in
males; necrosis in
females
Hyperplasia/
regeneration
Mitosis, bile duct
proliferation, foci
Bile duct proliferation: 5
ppm, F; 16 ppm, M;
mitosis: 16 ppm, M; 48
ppm, F; foci: 48 ppm
both sexes
Foci: 5 ppm, F;
22 ppm, M3
Reference
Nagano et al,
2007a; JBRC,
1998
Nagano et al.,
2007a; JBRC,
1998
Nagano et al.,
2007b; JBRC,
1998
Nagano et al.,
2007b; JBRC,
1998
3 This concentration was adjusted to continuous exposure (e.g., a factor of 6/24 x 5/7 applied used for an inhalation
exposure administered 6 hours/day, 5 days/week).
b Hepatic tumors detected at this level.
In the 2-year inhalation studies in rats and mice by JBRC (Nagano et al., 2007b; JBRC,
1998), which are the best documented of the available chronic studies, livers of male and female
rats and male mice with adenomas or carcinomas also expressed nonneoplastic changes,
including degenerative changes, fatty liver, fibrosis, cirrhosis, and bile duct proliferation. This
association was not observed, however, in low-dose (5-ppm or 0.9-ppm duration adjusted)
female mice, where an increased incidence of liver adenomas occurred in the absence of
evidence of hepatocellular cytotoxicity.
Eschenbrenner and Miller (1946) reported development of tumors in mice at doses that
did not evidently produce necrosis, but the design of this study may have influenced this result,
as animals were sacrificed and examined 1 month after the end of the main treatment period
(animals were, however, given one last dose 24 hours prior to sacrifice). Currently, there are no
data to characterize the liver changes that may have occurred and what effect this would have on
eliciting or abating cellular cytotoxicity 24 hours prior to terminal sacrifice. The investigators
noted that all doses that induced hepatomas were likely to have caused initial necrosis based on
separate studies using one or two doses. Regenerative changes were not investigated in this part
of the study.
155
DRAFT - DO NOT CITE OR QUOTE
-------
4.7.3.2.2. Dose-response concordance. Carbon tetrachloride-induced liver tumors were seen in
rats, mice, and hamsters after oral bolus dosing in oil and in rats and mice exposed by inhalation.
Several oral studies were conducted using only a single-dose level (i.e., studies in the mouse by
Edwards, 1941; Edwards and Dalton, 1942; Edwards et al., 1942; and a study in the hamster by
Delia Porta et al., 1961) and, therefore, did not provide information on the relationship between
dose and tumor induction. The NCI (1977, 1976a, b) bioassay included two dose levels, but high
early mortality in the rat study, particularly at the high dose, limited interpretation of the results.
In the mouse study, liver carcinomas were produced at almost 100% incidence in male and
female mice of both dose groups (i.e., liver tumors were observed in 179 of 183 exposed male
and female mice). Eschenbrenner and Miller (1946) observed liver tumors in all mice treated
daily with 20 mg/kg-day or more (n = 29), but none in the 10 mice treated with 10 mg/kg-day.
The JBRC inhalation studies in rats and mice (Nagano et al., 2007b; JBRC, 1998), which used
exposure concentrations of 5, 25, or 125 ppm, showed an increase in the incidence of liver
tumors (hepatocellular adenomas or carcinomas) in rats and mice of both sexes with increasing
exposure concentration (see Tables 4-4 and 4-5).
The dose-response relationship between hepatic cytotoxicity and tumor formation is best
demonstrated by the JBRC cancer bioassay in rats and mice, which examined histopathological
changes to the liver after 13 and 104 weeks and tumor formation after 104 weeks of exposure to
carbon tetrachloride by inhalation (Nagano et al., 2007a, b; JBRC, 1998). Carbon tetrachloride
concentrations evaluated were 0, 10, 30, 90, 270, and 810 ppm in the 13-week study and 0, 5, 25,
and 125 ppm in the 104-week study. In rats exposed for 13 weeks, histopathological changes
indicative of cellular damage ("fatty change") and inflammation (granulation) were observed in
all carbon tetrachloride treatment groups. At concentrations >30 ppm, proliferative
(pleomorphism and increased mitosis) and regenerative (fibrosis, proliferative ducts, cirrhosis)
responses occurred. At concentrations >270 ppm, eosinophilic and basophilic foci, which are
associated with hyperplastic or preneoplastic changes, were observed. Similar nonneoplastic
hepatic lesions (fatty changes, granulation, cirrhosis) were observed in livers of rats exposed to
>25 ppm for 104 weeks; the incidence of nonneoplastic lesions in rats exposed to 5 ppm for 104
weeks appeared similar to that in controls. The incidence of liver tumors in rats was
significantly increased only in the 125-ppm group compared with that in concurrent controls,
although an increase in hepatocellular carcinomas in 25-ppm female rats exceeded the historical
control range. Thus, liver tumors were observed at an exposure level associated with
hepatotoxicity following subchronic and chronic exposure; tumors were not observed at an
exposure level below the level that induced cytotoxicity (<10 ppm for 13-week exposure and 5
ppm for 104-week exposure).
A similar, but less consistent, dose-response relationship for cytotoxicity and tumor
formation was observed for mice (Nagano et al., 2007b; JBRC, 1998). In mice exposed for 13
156 DRAFT - DO NOT CITE OR QUOTE
-------
weeks, dose-dependent histopathological findings indicative of cytotoxicity, damage,
proliferation, and preneoplastic changes were observed. In male mice, histopathological findings
indicative of fatty change were observed in male mice exposed to >10 ppm and in female mice
exposed to >30 ppm carbon tetrachloride. In male and female mice exposed to >30 ppm, a
significantly increased incidence of liver collapse was observed. Liver collapse was
characterized by shrunken parenchymal tissue over the centilobular area, presumably resulting
from the necrotic loss of hepatocytes, and accompanied by proliferation of the bile ducts and
oval cells. In male and female mice exposed to >270 ppm, the incidences of nuclear enlargement
of hepatocytes with atypia and altered cell foci were significantly increased. The incidence of
liver adenomas and carcinomas in male mice in the 104-week study was increased compared to
concurrent controls at >25 ppm, an exposure level that also produced cytotoxicity and similar an
exposure level (30 ppm) that produced a proliferative response in the 13-week study. In female
mice, however, the incidence of hepatocellular adenomas was statistically elevated at 5 ppm
(0.9-ppm adjusted) compared to concurrent controls although hepatocellular damage was not
observed.
Thus, the dose-response relationships between cytotoxicity and liver tumors
demonstrated by the JBRC bioassay in male and female rats and male mice generally support a
prominent role for cytotoxicity, regeneration, and proliferation in the MOA for carbon
tetrachloride-induced carcinogenesis at higher exposure levels; however, the dose-response
relationship for the female mouse suggests that these key events do not adequately describe the
MOA(s), particularly at lower exposure levels.
As summarized in Table 4-15, several subchronic inhalation and oral studies have
demonstrated that carbon tetrachloride produces hepatic toxicity and regeneration. In rodents
exposed to carbon tetrachloride vapor for 12 weeks to 6 months, LOAELs for tissue damage
were reported at concentrations ranging from 4 to 42 ppm (adjusted) and for hyperplasia/
regeneration at concentrations ranging from 4 to 20 ppm (adjusted). Thus, results of subchronic
exposure studies are consistent with results of the JBRC study in rats, showing cytotoxicity at
>10 ppm (>2 ppm adjusted) and hyperplasia/proliferation at >30 ppm (>5.4 ppm adjusted) after
13 weeks of exposure (Nagano et al., 2007a; JBRC, 1998) and cytotoxicity and
hyperplasia/regeneration at>25 ppm (>4.5 ppm adjusted) after 104 weeks of exposure (Nagano
et al., 2007b; JBRC, 1998). In rats and mice exposed orally to carbon tetrachloride for 12-17
weeks, LOAELs for tissue necrosis ranged from 8.6 to 80 mg/kg-day and for
hyperplasia/regeneration ranged from 12 to 71 mg/kg-day. Durations of the subchronic studies
were too short to evaluate tumor formation; thus, data from subchronic studies do not allow for
further definition of the dose-response relationship and time course for cytotoxicity and tumor
formation.
Significant research has been conducted on the mechanistic events that precede carbon
tetrachloride-induced hepatocellular cytotoxicity (see Section 4.5). Much of this research has
157 DRAFT - DO NOT CITE OR QUOTE
-------
focused on lipid peroxidation (de Zwart et al., 1997; Gasso et al., 1996; Ichinose et al., 1994;
Tribble et al., 1987; Lee et al., 1982; Recknagel and Glende, 1973; Rao and Recknagel, 1969),
decreases in antioxidant levels (Cabre et al., 2000; Gasso et al., 1996; Gorla et al., 1983),
alterations in calcium homeostasis, and activation of calcium-dependent phospholipases (Limaye
et al., 2003; Hemmings et al., 2002; Gonzalez Padron et al., 1993; Agarwal and Mehendale,
1986, 1984; Long and Moore, 1986; Kroner, 1982; Moore et al., 1976). Compared to doses that
result in tumor formation in chronic bioassays (5-125 ppm: Nagano et al., 2007b; JBRC, 1998;
20 mg/kg-day: Eschenbrenner and Miller, 1946), these mechanistic studies were conducted at
relatively high exposure levels (see Table 4-16). In most, if not all, mechanistic studies,
exposure levels greatly exceeded those used in chronic bioassays (e.g., on the order of grams per
kilogram (in vivo) or millimolar concentrations (>1 mM) of carbon tetrachloride). The relevance
of the mechanistic findings at these high exposure levels to lexicologically relevant exposures is
uncertain (Weber et al., 2003; Clawson, 1989; Recknagel et al., 1989; Dolak et al., 1988). The
degree to which lipid peroxidation, depletion of cellular antioxidants, alterations in calcium
homeostasis, and activation of calcium-dependent phospholipases contribute to the process of
cytotoxicity, regenerative proliferation, and tumorigenesis, and the possible reversibility of these
effects, constitutes an area of uncertainty (Weber et al., 2003; Rikans et al., 1994; Kefalas and
Stacey, 1989; Dolak et al., 1988; Sandy et al., 1988; Stacey and Klaassen, 1981).
Table 4-16. Dose considerations of mechanistic studies of carbon tetrachloride
End point
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Dose of carbon
tetrachloride
1 mL/kg
(l,590mg/kg)a
0.5 mL/kg
(800 mg/kg)a
0.5 mL (2.38
mL/kg) injected i.p.
(800 mg/kg)a
ImM
(154mg/L)
ImM
(154mg/L)
Test system
Sprague-Dawley
rats; three strains
of mice (A/J,
BALB/cJ, and
C57B1/6J
Rats and mice
Male Wistar rats
In vitro, liver
microsomes
(multiple species)
Liver slices from
Sprague-Dawley
Rats
Result
Increased conjugated dienes
in treated animals compared
to controls
Ethane production increased
in treated animals; iron
binding eliminated lipid
peroxidation (ethane) in
treated animals
TEARS significantly lower
in animals receiving SAM
Increased MDA DNA
adducts
Significant increase in
TEARS
Reference
Leeetal, 1982
Younes and Siegers,
1985
Gasso etal, 1996
Ichinose etal., 1994
Fragaetal, 1987
158
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-16. Dose considerations of mechanistic studies of carbon tetrachloride
End point
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Lipid peroxidation
Protein carbonyl
(protein adducts)
GSH modulation
GSH modulation
GSH modulation
GSH modulation
Altered Ca++
homeostasis
Altered Ca++
homeostasis
Dose of carbon
tetrachloride
500 mg/kg
3,200mg/kg
0. 1 mL/kg
(160mg/kg)a
1,590 mg/kg
2.5 mL/kg p.o. or 1
mL/kg (5 mL/kg as
a 20% solution.)
injected i.p.
(3,980 or 1,590
mg/kg)a
1,590 mg/kg
1 and 4 mM
(154 and 615 mg/L)
1 and 4 mM
(154 and 615 mg/L)
0.5 ml (2.38
mL/kg) injected i.p.
(800 mg/kg)a
Pretreated with 2
g/kg GSH 30
minutes prior to
1,590 mg/kg i.p.
carbon tetrachloride
1 ,600 mg/kg, twice
weekly for 6 weeks,
i.p.
0.1 mL/kg
(160mg/kg)a
0.3-10 mM
(46-1, 540 mg/L)
ImL/kg injected i.p.
(l,590mg/kg)a
Test system
Female F344 rats
Female F344 rats
Hamsters
Rat
Male Sprague-
Dawley rats
Rat
In vitro rat
hepatocytes
In vitro rat
hepatocytes
Wistar rats
Male Sprague-
Dawley rats
Rat
Female Balb/c
mice
In vitro
hepatocytes
Female Wistar rats
Result
Twofold induction of HNE-
dG adducts
37-Fold induction of HNE-
dG adducts
MDA DNA adducts
Significant increase in 4-
HNE and MDA adducts in
liver
Conjugated dienes or
incorporation of C14 labeled
carbon tetrachloride was not
significantly prevented by
several antioxidants
2.5-Fold increase TEARS
over controls
Significant increase in
MDA adducts
2.5-Fold increase at 4 mM
GSH decreased at 5 weeks
GSH pretreatment partially
prevented hepatic necrosis
Significant decrease in
GSH; SAM partially
prevented liver toxicity
Schisandrin B-partially
prevented hepatotoxicity
and GSH depletion
Increased activity of
phosphorylase a and
decreased activity of
endoplasmic reticulum
Ca++ pump; effects only
observed at concentrations
>1 mM
Significant decrease in
microsomal Ca++
concentration; significant
Reference
Wackeretal.,2001
Chung et al, 2000
Wang and Liehr,
1995
Hartley et al., 1999
de Ferreyra et al.,
1975
Hartley et al., 1999
Beddowes et al.,
2003
Beddowes et al.,
2003
Cabre et al., 2000
Gorlaetal, 1983
Gassoetal, 1996
Chiu et al., 2003
Long and Moore,
1986
Kroner, 1982
159
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-16. Dose considerations of mechanistic studies of carbon tetrachloride
End point
Altered Ca++
homeostasis
Altered Ca++
homeostasis
Altered Ca++
homeostasis
Phospholipase
activity
Phospholipase
activity
Phospholipase
activity
Phospholipase
activity
Dose of carbon
tetrachloride
2.5 mL/kgoral dose
by feeding tube
(3,980 mg/kg)a
0.03mL/100gto
0.125mL/100g
body weight
(1.25mL/kgby
feeding tube)
(0.48-1,990
mg/kg)a
50 uM
(7.7 mg/L)
3 mL/kg i.p.
(4,770 mg/kg)a
1 mL/kg injected
i.p.
(l,590mg/kg)a
0.23-1. 3 mM
(3 5- 200 mg/L)
1.2 mM
(185 mg/L)
Test system
Male Sprague-
Dawley rats
Male F344 rats
In vitro
hepatocytes
Male Sprague-
Dawley rats
Male Sprague-
Dawley rats
In vitro
hepatocytes
In vitro
hepatocytes
Result
increase in mitochondrial
Ca++ concentration
85% reduction in ATP-
dependent Ca++ uptake and
endoplasmic reticulum
capacity
Decreased Ca++ transport
across plasma membrane
and mitochondria
Elevated cytosolic Ca++
levels
Co-treated with CBZ
(calpain inhibitor),
decreased mortality 50%
from lethal dose of carbon
tetrachloride
Pretreated with quinacrine
(phospholipase A2 inhibitor)
Increased phospholipase A2
activity 1.4- to 5. 3 -fold
Increased phospholipase A2
activity and hepatocyte
degeneration (LDH release)
Reference
Moore et al, 1976
Hemmings et al.,
2002
Stoyanovsky and
Cederbaum, 1996
Limayeetal.,2003
Gonzalez Padron et
al., 1993
Glende and
Pushpendaran, 1986
Glende and
Recknagel, 1992
a Dose in mg/kg estimated using a density for carbon tetrachloride of 1.5940 g/mL at 20°C.
An additional area of significant uncertainty for dose-response concordance is the
possibility of genetically damaging events occurring at or below doses that induce tumors in
laboratory rodents. Because genotoxicity and mechanistic data in this portion of the dose-
response curve are limited, a low-dose mutagenic effect cannot be excluded.
4.7.3.2.3. Temporal relationship. Carbon tetrachloride is metabolized to trichloromethyl and
peroxy free radicals, which may result in radical-induced mechanisms including lipid
peroxidation and disruption of calcium homeostasis leading to hepatocellular cytotoxicity. Initial
metabolism of carbon tetrachloride to reactive radicals and subsequent events leading to
160
DRAFT - DO NOT CITE OR QUOTE
-------
cytotoxicity are ongoing processes that occur throughout exposure.
The temporal progression of nonneoplastic liver lesions following acute and subchronic
exposure is consistent with the hypothesized cytotoxic-proliferative MOA. Acute toxicity
studies on rodents treated orally with carbon tetrachloride show hepatic necrosis within 6-24
hours of dosing and evidence of compensatory hepatocellular proliferation (mitosis, BrdU-
positive labeling, or increases in DNA-synthesizing enzymes and increases in cells in S-phase) at
the same time or within 48 hours (Lee et al., 1998; Wang et al., 1997; Steup et al., 1993;
Doolittle et al., 1987; Nakata et al., 1975; Eschenbrenner and Miller, 1946). As reviewed in
Sections 4.2.1.1 and 4.2.2.1, numerous subchronic exposure studies report histopathological
findings consistent with an ongoing cycle of hepatic damage, repair, and proliferation (e.g., fatty
vacuolization and degeneration, necrosis, nuclear pleomorphism, hyperplasia, fibrosis, and
cirrhosis) (Nagano et al., 2007a; JBRC, 1998; Allis et al., 1990; Bruckner et al., 1986; Condie et
al., 1986; Litchfield and Gartland, 1974). Smyth et al. (1936), Adams et al. (1952), and Benson
and Springer (1999) clearly show a progression of liver toxicity from fatty degeneration of the
liver to liver cirrhosis and hepatocellular proliferation only at doses that produce necrotic
damage.
A temporal and dose-related progression of key events (hepatotoxicity, repair,
proliferation, and tumor development) is supported by the results of the JBRC inhalation cancer
bioassay in rats (Nagano et al., 2007b; JBRC, 1998), in which the development of hyperplastic or
preneoplastic lesions (eosinophilic and basophilic foci) following subchronic exposure to
cytotoxic levels, with subsequent development of liver tumors, is demonstrated (see Table 4-17).
Thus, in the rat, the temporal relationship of the key events is consistent with the hypothesized
MOA for carbon tetrachloride carcinogenesis. This relationship, however, is not as clearly
defined for the increased incidence of liver adenomas in female mice (Nagano et al., 2007a, b).
161 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-17. Temporal sequence and dose-response relationship for key events and
liver tumors in male and female F344 rats exposed to carbon tetrachloride vapor for
13 and 104 weeks (6 hours/day, 5 days/week)
Key event (time ->)
Exposure
level3
(ppm)
5 (0.9)
10(1.8)
25 (4.5)
30 (5.4)
90(16.1)
125 (22.3)
270 (48.2)
810(145)
Metabolism &
formation of
•O-O-CCb
(immediate and
ongoing)
+ b
+ b
+ b
+ b
+ b
+ b
+ b
+ b
13 weeks
Hepato-
toxicity c
+/— e
+
+
+
+
Regeneration
and
proliferation d
—
+
+
+
+
104 weeks
Hepato-
toxicity0
—
+
+
Regeneration
and
proliferation d
—
—
+
Liver tumors
(104 weeks)
—
+/-f
+
a The exposure concentration in parentheses is the concentration adjusted to continuous exposure (i.e., multiplied by
5/7 x 6/24)
+ b= Studies demonstrating key event were not conducted as part of the JBRC 13- and 104-week bioassays. Based
on data from acute exposure and in vitro studies (Avasarala et al, 2006; Zangar et al., 2000; Raucy et al., 1993),
metabolism of carbon tetrachloride to reactive metabolites has been demonstrated and is assumed to occur
immediately and continue throughout the duration of exposure to carbon tetrachloride at all exposure levels.
Although metabolism to reactive metabolites has been specifically demonstrated at relatively high doses, it can
reasonably be assumed that such metabolism would occur at lower exposures.
0 As indicated based on histopathological findings, including fatty change, fibrosis, cirrhosis, and/or necrosis.
d As indicated based on histopathological findings, including proliferation and hyperplasia (and in the 13-week
study, mitosis).
e An increased incidence of fatty change was observed that was not statistically significant.
f The incidence of hepatocellular carcinomas in female 25-ppm rats was not statistically elevated compared to
concurrent controls, but did exceed the historical control range for female rats from JBRC (0-2%), and increase that
was statistically significant compared to the historical control.
Note: Different exposure concentrations were used in the 13- and 104-week JBRC bioassays. Blank cells indicate
exposure concentrations not tested in either the 13- or 104-week study.
+ = Evidence demonstrating key event.
— = No evidence demonstrating key event.
+/— = equivocal
Source: Nagano et al. (2007a, b); JBRC (1998).
4.7.3.2.4. Biological plausibility and coherence. The theory that sustained cell proliferation to
replace cells killed by toxicity or viral or other insults, such as physical abrasion of tissues, can
be a significant risk factor for cancer is plausible and generally accepted (Correa, 1996). It is
logical to deduce that sustained cytotoxicity and regenerative cell proliferation may result in a
162
DRAFT - DO NOT CITE OR QUOTE
-------
greater likelihood of mutations (whether spontaneous, or directly or indirectly induced by the
chemical) being perpetuated, with the possibility of one or more of these resulting in loss of cell
cycle control and tumor development. It may also be that continuous stimulus of proliferation by
growth factors involved in inflammatory responses (e.g., TGF-a in the hepatic response to
carbon tetrachloride) increases the probability that damaged cells may slip through cell cycle
checkpoints carrying DNA alterations that would otherwise be repaired. Current views of cancer
processes support both possibilities. A high proliferation rate alone is not assumed to cause
cancer; tissues with naturally high rates of turnover do not necessarily have high rates of cancer,
and tissue toxicity in animal studies does not invariably lead to cancer. Nevertheless,
regenerative proliferation associated with persistent cytotoxicity appears to be a risk factor for
carcinogenicity.
4.7.3.3. Other Possible Modes of Action
Genotoxicity. Section 4.4.2 provides a critical review of the genotoxicity literature for
carbon tetrachloride. Various confounding factors and other challenges in evaluating
genotoxicity studies are highlighted in Table 4-12; these general features of the carbon
tetrachloride literature as well as various methodological and reporting issues in individual
studies were taken into account in the current review of the genotoxicity literature. Many of the
positive genotoxicity findings, including the following, are consistent with compounds that
induce oxidative events leading to genetic damage: (1) two positive mutation/DNA damage
studies in E. coli strains particularly sensitive to oxidative damage; (2) intrachromosomal
recombination induced by carbon tetrachloride in S. cerevisiae consistent with DNA breakage
originating from oxidative stress that occurs concurrent with cytotoxicity; (3) evidence from in
vitro and in vivo assays of DNA breakage and fragmentation in association with extensive
hepatotoxicity; and (4) DNA adducts formed from reactive oxygen species and lipid peroxidation
products (e.g., MDA and 4-HNE) in the liver of rodents following carbon tetrachloride
administration. A limited number of positive genotoxicity findings in the absence of cytotoxicity
have been reported (see Section 4.4.2 and Table 4-8 to 4-11); however, methodological or
reporting issues with many of these studies have been identified.
As a whole, the literature suggests that carbon tetrachloride is more likely an indirect than
direct mutagenic agent and that mutagenic effects, if they occur, are likely to be generated
through indirect mechanisms resulting from oxidative damage or lipid peroxidation byproducts,
which have been observed with cytotoxicity at high doses of carbon tetrachloride (Table 4-16).
Nevertheless, deficiences in this complex database must be acknowledged. To that end, if
carbon tetrachloride-associated DNA damage occurs above background levels and contributes to
low-dose mutagenic activity, some nonzero risk of carcinogenicity at doses below those
associated with cytotoxicity would be predicted.
Thus, the possible contribution of a low-dose mutagenic effect in the hypothesized MOA
163 DRAFT - DO NOT CITE OR QUOTE
-------
or alternative MO As cannot be excluded.
Epigenetic effects and changes in gene expression. As summarized in Section 4.4.2.4, a
number of studies have reported alterations in liver DNA methylation. For instance, Varela-
Moreiras et al. (1995) investigated the effect of short-term administration of carbon tetrachloride
on hepatic DNA methylation and on SAM and SAH in male Wistar rats administered 800 mg/kg
carbon tetrachloride by i.p. injection 2 times/week, for 3 weeks. Rats treated with carbon
tetrachloride exhibited hypomethylation of their hepatic DNA as measured by the extent to
which the liver DNA from the treated animals could be methylated in vitro using [3H-methyl]-
SAM as a methyl donor. In addition, decreased levels of SAM, methionine, and folate as well as
increased levels of SAH and homocysteine were seen. No changes were observed in the levels
of cystathionine or GSH, or in the activity of SAM-synthetase. The magnitude of the observed
changes was substantially reduced in animals co-administered SAM with carbon tetrachloride.
The authors proposed that "carbon tetrachloride disrupts the distribution of homocysteine
between remethylation and its degradation via the transsulphuration pathway, and that SAM, by
resetting the methylation ratio, restores this equilibrium." In eukaryotic and mammalian cells,
gene expression is influenced by the extent and patterns of DNA methylation, so the observed
changes in hepatic DNA methylation could represent an epigenetic alteration that could
contribute to carbon tetrachloride carcinogenesis.
Changes in the expression of specific genes in response to carbon tetrachloride exposure
have been investigated in the liver of rodents and in cultured human hepatoma cell line (see
Section 4.5.7) (lessen et al., 2003; Fountoulakis et al., 2002; Bartosiewicz et al., 2001; Holden et
al., 2000; Columbano et al., 1997; Menegazzi et al., 1997). Many of the known upregulated
genes are related to stress, DNA damage and repair, and signal transduction. Intraperitoneal
injection of Sprague-Dawley rats with 160 mg/kg of carbon tetrachloride in corn oil activated c-
fos and c-jun gene expression in the liver within 30 minutes (Gruebele et al., 1996).
Pretreatment of rats with diallyl sulfide, an inhibitor of CYP2E1, 3 hours before dosing with
carbon tetrachloride reduced c-jun mRNA levels by 76%. Treatment with carbon tetrachloride
also increased hepatic nuclear levels of the NF-Kp transcription factor, which regulates genes
involved in responses to inflammation, apoptosis, hepatocyte proliferation, and liver
regeneration.
Columbano et al. (1997) investigated the relationship between immediate early genes and
hepatocyte proliferation through comparison of the hepatic levels of c-fos, c-jun, and LRF-1
transcripts during mouse liver cell proliferation under two conditions: (1) direct hyperplasia
induced by the primary mitogen (and hepatocarcinogen) TCPOBOP, and (2) compensatory
regeneration caused by a necrogenic dose of carbon tetrachloride (single intragastric dose of 2
mg/kg in oil) or by performing a 2/3 PH. A striking difference in the activation of early genes
was observed. In spite of a rapid stimulation of S phase by the mitogen TCPOBOP, there were
164 DRAFT - DO NOT CITE OR QUOTE
-------
no changes in the expression of c-fos, c-jun and LRF-1 or in steady-state mRNA hepatic levels
of IGFBP-1 (a gene highly expressed in rat liver following PH), and only a slight increase in c-
myc and PRL-1. In contrast, a rapid, massive, and transient increase in the hepatic mRNA levels
of all these genes was observed during carbon tetrachloride-induced regeneration that was
comparable to those seen following 2/3 PH. In similar research from the same laboratory, the
pattern of immediate early gene and growth factor gene expression in the rat liver induced by
primary mitogens (including LN, cyproterone acetate, or NAF) was shown to differ from that
observed following compensatory liver regeneration occurring after cell loss/death and direct
hyperplasia resulting from a partial 2/3 hepatectomy or a necrogenic dose (2 mL/kg) of carbon
tetrachloride (Menegazzi et al., 1997). In this study, the following indicators of gene expression
were examined: modifications in the activation of two transcription factors, NF-KB and AP-1;
steady-state levels of TNF-a mRNA; and induction of the iNOS. Liver regeneration after
treatment with carbon tetrachloride was associated with an increase in steady-state levels of
TNF-a mRNA, activation of NF-KP and AP-1, and induction of iNOS. LN induced NF-KP,
TNF-a, and iNOS mRNA but not AP-1, whereas direct hyperplasia induced by the other two
primary mitogens occurred in the complete absence of modifications in the hepatic levels of
TNF-a mRNA, activation of NF-KP and AP-1, or induction of iNOS, although the number of
hepatocytes entering S phase 18-24 hours after NAF was similar to that seen after PH. The
findings from these two studies suggest that regenerative proliferation alone does not explain the
tumorigenic response associated with carbon tetrachloride in chronic bioassays, but these data do
not preclude regenerative proliferation as a biologically-based marker of such causal events.
4.7.3.4. Conclusions About the Hypothesized Mode of Action
The weight of evidence across different species, sexes, and routes of exposure supports
reductive dehalogenation of carbon tetrachloride by CYP2E1, sustained cytotoxicity, and
regenerative cell proliferation as key events in the MOA for carbon tetrachloride-induced tumors
of the liver. These key events generally have experimental support at high exposure levels in
terms of strength, consistency, and specificity of association; dose-response concordance;
temporal relationship; and biological plausibility and coherence. Further, these key events are
generally considered to dominate the dose-response relationship for liver tumors at high
exposures. The dose-response concordance and temporal relationship for cytotoxicity,
regenerative hyperplasia, and liver tumors in the female mouse (Nagano et al., 2007b), however,
is not consistent with this hypothesized MOA. An increased incidence of hepatocellular
adenomas occurred in the low-dose (5-ppm [0.9-ppm adjusted]) female mouse in the absence of
nonneoplastic liver toxicity, raising the possibility of another MOA operating in addition to or in
conjunction with the cytotoxic-proliferative MOA.
Considerable evidence points to the involvement of highly reactive metabolites (with the
capacity to chemically interact with DNA and other cellular macromolecules) in the processes of
165 DRAFT - DO NOT CITE OR QUOTE
-------
toxicity and carcinogenicity of carbon tetrachloride. In addition, subsequent chemical reactions
of carbon tetrachloride metabolites with cellular constituents lead to formation of reactive
oxygen species that also can damage DNA and other macromolecules. This fundamental
reactivity of both direct and indirect products of carbon tetrachloride metabolism provides a
cogent argument in favour of some degree of nonthreshold response to carbon tetrachloride
carcinogenicity.
Although the extensive genotoxicity database for carbon tetrachloride suggests that
carbon tetrachloride is not likely a direct acting mutagen, the database is complex and raises
various issues (see Table 4-12) that make it difficult to ascertain the potential genotoxicity of
carbon tetrachloride at exposures below which there is overt cytotoxicity. Positive genotoxicity
findings have generally been observed at exposures that induce cytotoxicity and regenerative cell
proliferation. Because of the difficulties in detecting genotoxic effects following treatment with
carbon tetrachloride, the many studies conducted at relatively high doses lack information
regarding dose-response and fail to characterize the role of genotoxicity at low carbon
tetrachloride exposure levels.
In summary, biological support exists for a hypothetical MOA involving metabolism of
carbon tetrachloride by CYP2E1, sustained cytotoxicity, and regenerative cell proliferation as
key events driving the steep nonlinear increase in liver tumor dose-response at relatively high
carbon tetrachloride exposures. Additionally, at high exposures, both the cytotoxicity-
regenerative proliferation-based MOA and the mutagenicity-based MOA may be operative, but it
is not possible to delineate the contribution of these potential MOA(s) to carbon tetrachloride
tumor response. Inconsistencies in the cytotoxicity-regenerative proliferation-based MOA at the
low end of the experimental exposure range suggest that MOA(s) that are independent of
cytotoxicity and regenerative cell proliferation may play a role in carbon tetrachloride liver
tumor-induction in this range. The fundamental reactivity of direct and indirect products of
carbon tetrachloride metabolism can reasonably be expected to play a role in carbon tetrachloride
carcinogenicity at all levels of exposure to carbon tetrachloride. Linear processes would likely
dominate the dose-response relationship at low exposures (i.e., exposures that are not cytotoxic).
4.7.3.5. Relevance of the Hypothesized Modes of Action to Humans
Although data are inadequate to determine the operative MOA for rodent liver tumors at
low exposures, none of the available data suggest that the hypothesized MO As are biologically
precluded in humans. Humans express ethanol-inducible CYP2E1 and phenobarbital-inducible
CYP3A in the liver, both of which are associated with the generation of trichloromethyl radical
in animals exposed to carbon tetrachloride. The antioxidant systems in animals and humans are
similar. Therefore, both the hypothesized MOA and the endogenous protective mechanisms
likely have related processes in animals and humans. Furthermore, humans exhibit the same
signs of liver toxicity that have been observed in animal studies (cirrhosis, fibrosis, steatosis,
166 DRAFT - DO NOT CITE OR QUOTE
-------
necrosis, and liver enzyme changes). Finally, the types of tumors, hepatocellular adenoma and
carcinoma, expressed consistently in several animal species exposed to carbon tetrachloride are
also found in humans.
4.7.4. Mode of Action Information for Pheochromocytomas
An increased incidence of pheochromocytomas (a neuroendocrine tumor of adrenal
chromaffm cells) associated with carbon tetrachloride administration has been observed in male
and female mice by oral (NCI, 1977, 1976a, b; Weisburger, 1977; NTP, 2007) and inhalation
exposure (Nagano et al., 2007b; JBRC, 1998), but not in rats by either route of exposure. The
MOA by which carbon tetrachloride induces pheochromocytomas in mice is not known.
Research on the mechanism(s) by which carbon tetrachloride induces toxicity in the
adrenal gland is largely limited to short-term studies of carbon tetrachloride enzyme activation in
the adrenal tissue. Colby et al. (1994, 1981) reported that carbon tetrachloride has induced
adrenocortical necrosis in humans, although reports of effects of carbon tetrachloride on the
human adrenal gland were not independently identified. In experimental animals, acute exposure
to carbon tetrachloride has produced adrenal necrosis, with effects localized to the zona
reticularis, the innermost region of the cortex (Brogan et al., 1984). This localization of toxicity
appears to be the result of greater activation of carbon tetrachloride by microsomal enzymes in
the zona reticularis (Colby et al., 1994; Brogan and Colby, 1983). In vitro studies showed that
preincubation of adrenal microsomes with 1-aminobenzotriazole (ABT), a CYP450 suicide
inhibitor, prevented the effects of carbon tetrachloride on lipid peroxidation and covalent binding
(Colby et al., 1994). It would appear that carbon tetrachloride metabolism plays a role in the
induction of toxicity in the adrenal gland as it does in the liver.
Malendowicz and Colby (1982) administered 0.2 mL (-1,750 mg/kg) carbon
tetrachloride to Wistar rats by gavage once a week for 20 weeks or 0.1 mL (-880 mg/kg) per day
for 7 or 14 days. The investigators reported that their findings suggest that carbon tetrachloride
may influence plasma corticosteroid levels through effects on adrenal steroid production as well
as hepatic reductive steroid metabolism, resulting in overall decreases in circulating
corticosteroid concentrations. No evidence is available, however, that suggests any association
between effects on corticosteroid balance and induction of pheochromocytomas.
Unlike liver carcinogenicity, none of the key events (subsequent, perhaps, to activation to
a biologically active metabolite) in the development of carbon tetrachloride-induced
pheochromocytomas has been elucidated. In general, few chemicals have been reported to cause
pheochromocytomas in mice. Of 514 technical reports published by NTP, exposure to only
seven chemicals has been associated with pheochromocytomas in mice with no apparent
common denominator (Tischler et al., 2004; Hill et al., 2003).
167 DRAFT - DO NOT CITE OR QUOTE
-------
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
Age (e.g., childhood, senescence), gender, nutritional status, disease status, and exposure
to other chemicals are all factors that might influence susceptibility to carbon tetrachloride.
These factors are described further below.
Hypothesized key events leading to carbon tetrachloride-induced liver toxicity and
carcinogenicity (e.g., metabolism to trichloromethyl radical by CYP2E1 and subsequent
formation of trichloromethyl peroxy radical; cytotoxicity; sustained regenerative and
proliferative changes in the liver in response to hepatotoxicity; epigenetic alterations; gene
expression changes; and DNA damage and fixation leading to mutagenic activity) involve
metabolic and cellular processes common to cells at all life stages. Because metabolism is a
hypothesized key event in carbon tetrachloride toxicity, heterogeneity in the human population in
microsomal enzymes responsible for carbon tetrachloride metabolism could also influence
susceptibility to carbon tetrachloride. Quantitative information on variation in human hepatic
levels of CYP2E1 and other CYP450 enzymes demonstrates considerable intrahuman variability.
For example, Lipscomb et al. (1997) reported a sevenfold range in activity of CYP2E1 among
hepatic microsomal samples from 23 subjects. Snawder and Lipscomb (2000) demonstrated 36-,
13-, 11-, 2-, 12-, and 22-fold differences in CYP1A, CYP2B, CYP2C6, CYP2C11, CYP2E1, and
CYP3 A protein content, respectively, between the highest and lowest samples from 40 samples
of microsomes from adult human liver organ donors.
4.8.1. Possible Childhood Susceptibility
Limited data on CYP450 enzymes are available to evaluate the relative susceptibility of
children to carbon tetrachloride. As observed in adult animals, the initiating event for liver
toxicity and carcinogenicity is metabolism of carbon tetrachloride by CYP2E1 to reactive
metabolites. Assuming that this is the initiating key event in the MOA for all age groups,
susceptibility to carbon tetrachloride at all life stages is related to the presence of functional
microsomal enzymes (particularly CYP2E1 but also CYP3 A). Hepatic concentrations of
CYP2E1 do not achieve adult levels until sometime between 1 and 10 years, although large
increases in hepatic CYP2E1 protein occur postnatally between 1 and 3 months in humans
(Vieira et al., 1996). Thus, age-related differences in CYP450, as described below, could
potentially affect susceptibility. To the extent that hepatic CYP2E1 levels are lower, infants and
children would be less susceptible to free radical-induced liver injury from carbon tetrachloride
than adults. There is some evidence from the therapeutic drug literature that CYP3 A levels also
change with age, but in a pattern different from CYP2E1. Based on ti/2 results for several
therapeutic drugs metabolized by the CYP3A family (Ginsberg et al., 2002), enzyme levels were
lower than the adult up to 2 months of age, but from 6 months to 2 years of age were
significantly higher than the adult. To the extent that CYP3 A levels are relatively higher than
the adult and CYP3 A plays a significant role in carbon tetrachloride metabolism, infants and
168 DRAFT - DO NOT CITE OR QUOTE
-------
young children could be relatively more susceptible to liver injury from carbon tetrachloride.
Work conducted by Zangar et al. (2000), however, suggests that CYP2E1 is the major human
enzyme in the adult responsible for carbon tetrachloride bioactivation at lower, environmentally
relevant levels (i.e., levels that are not hepatotoxic). Only at higher carbon tetrachloride levels,
CYP3 A and possibly other CYP450 forms may contribute to carbon tetrachloride metabolism.
Therefore, assuming that CYP2E1 is the more effective metabolizing enzyme in children as it is
in adults at environmentally-relevant exposure levels, infants and children would likely be less
susceptible to liver injury from carbon tetrachloride than adults to the extent that hepatic
CYP2E1 levels are lower. Carbon tetrachloride-specific enzyme data for younger populations
are not available, however, to confirm these assumptions.
Low levels of CYP2E1 mRNA begin to be elevated in human fetal brains after week 7
and increase thereafter to week 16 (the oldest stage examined) (Brzezinski et al., 1999).
CYP2E1 function was analyzed in prenatal human brain and liver tissues (7-17 weeks of
gestation) using three assays (Boutelet-Bochan et al., 1997). Low levels of CYP2E1 expression
were detected in fetal brain tissue, with some evidence for increasing expression at later stages of
gestation; weaker levels were identified in fetal liver. In fetal brain, CYP2E1 was not detected
with the less sensitive assay (Northern blot), and expression measured with the two more
sensitive assays (RT-PCR and RNase protection assays) were considerably weaker than those
measured in adult human or rat liver samples. The results suggested that, during gestation weeks
8-17, the fetal brain might be more vulnerable than the liver to toxic effects from exposure to
carbon tetrachloride. Carpenter et al. (1996) detected functional CYP2E1 in human fetal livers at
19 weeks of gestation. However, when related to weight unit of microsomal protein, the
CYP2E1 content of fetal livers was considerably lower than in adults. In an in vitro experiment,
exposure to ethanol or clofibrate induced expression of CYP2E1 in hepatocytes from a 20-week
fetus, which suggests that maternal alcohol intake might enhance CYP2E1 in the human fetus.
Given that the maternal liver mass and hepatocellular CYP2E1 content are so much higher than
the fetal values, it would seem that fetuses would have only a slight vulnerability from maternal
exposure to carbon tetrachloride at low levels. For inhalation exposures, the arterial blood flow
does not perfuse the liver before reaching the fetus; therefore, this observation may apply more
to oral exposures than to inhalation exposures.
An unknown factor in fetal vulnerability is the expression of CYP450 in the placenta.
Two different laboratories have detected CYP2E1 in human placentas. Hakkola et al. (1996)
detected several different enzymes in human placentas, including CYP2D6, CYP4B1, and
several forms of CYP3A and CYP2E1; there was considerable variation in expression among the
different individuals. Rasheed et al. (1997) compared the levels of CYP2E1 protein in western
immunoblots of microsomes taken at delivery from placentas of 12 African-American women.
None of the women who abstained from ethanol had detectable levels of placental CYP2E1,
whereas the protein was detectable in blots for 6/8 drinkers. The median head circumference at
169 DRAFT - DO NOT CITE OR QUOTE
-------
birth was significantly smaller (33.2 cm) in children with detectable CYP2E1 compared with
those without detectable enzyme (37 cm, p = 0.04). The study provides suggestive evidence that
placental CYP2E1 is inducible by alcohol consumption, although there are individual variations.
Theoretically, fetuses of mothers who drink ethanol would be potentially more susceptible to
injury from carbon tetrachloride exposure.
Carpenter et al. (1996) measured the amount and activity of CYP2E1 in fetal (GD 20)
and maternal rat liver and brain, following maternal exposure to a 5% ethanol diet. Rates of
metabolism for chlorzoxazone and N-nitrosodimethylamine were used to evaluate functional
activity of CYP2E1. In untreated or pair-fed rats, the amount of CYP2E1 in maternal or fetal
brain was several hundred-fold lower than in the respective livers. Ethanol exposure increased
the level of CYP2E1 protein by 1.4-fold in the maternal liver and 2.4-fold in the fetal liver
compared with the untreated or pair-fed groups but had no effect on CYP2E1 levels in maternal
or fetal brain. Hepatic CYP2E1 function, as exemplified by chlorzoxazone 6-hydroxylation, was
elevated 2.1-fold in ethanol-exposed maternal liver but not significantly in fetal liver.
Demethylation of N-nitrosodimethylamine was elevated about 1.5-fold in maternal and fetal
livers after ethanol exposure. Cambon-Gros et al. (1986) demonstrated the formation of
trichloromethyl radicals in maternal and fetal rat liver exposed to carbon tetrachloride on GD 20.
The results of these studies suggest that maternal ethanol ingestion might increase the
susceptibility of fetuses to hepatotoxicity from exposure to carbon tetrachloride.
Developmental studies in rats demonstrated that total litter loss was the primary effect of
maternal exposure between GDs 6 and 15 (Narotsky et al., 1997b; Narotsky and Kavlock, 1995).
The MOA for developmental effects has not been explored, so it is unknown whether placental
expression of CYP2E1 may contribute to the litter loss, as CYP2E1 contributes to liver
cytotoxicity.
While some information is available on the activity of enzymes involved in the
metabolism of carbon tetrachloride in children, little lifestage-specific information on the levels
of antioxidants (e.g., GSH) was identified.
In summary, there is no direct evidence for increased or decreased susceptibility to
carbon tetrachloride in children. The relatively lower activity of CYP2E1 (the major human
enzyme responsible for carbon tetrachloride bioactivation at environmentally-relevant exposure
levels) in infants and children compared to adults suggests the possibility of lower susceptibility
to carbon tetrachloride-induced liver injury for younger life stages. Too little is known,
however, about changes in activity of other enzyme levels with age to support a conclusion that
children are at decreased risk. CYP3 A levels are higher in children 6 months to 2 years than in
adults (although CYP3 A is less likely to contribute to carbon tetrachloride metabolism at
environmentally-relevant exposure levels than CYP2E1). Further, little lifestage-specific
information on the levels of antioxidants (e.g., GSH), another factor likely to contribute to
susceptibility to carbon tetrachloride toxicity, is available. No information is available to support
170 DRAFT - DO NOT CITE OR QUOTE
-------
an evaluation of differences in childhood susceptibility to possible effects of carbon tetrachloride
on the adrenal medulla (as suggested by the increased incidence of pheochromocytomas in
mice).
4.8.2. Possible Effects of Aging
The overall vulnerability to carbon tetrachloride is affected directly by the rate of
generation of reactive intermediates, a function of microsomal CYP activity, and inversely by the
antioxidant content. Compared with young/mature adults, older organisms exhibit changes,
usually decreases, in these parameters that vary independently in different tissues.
Studies evaluating the capacity for drug metabolism in the human liver during different
life stages reported a reduction in activity for CYP3A3/4 and CYP2E1 in the elderly (i.e.,
individuals older than 65 years) (reviewed in Tanaka, 1998). Total immunoreactive CYP3A
protein (the sum of CYP3A4 and CYP3A5) per mg hepatic microsomal protein was significantly
reduced by 90% in samples from men aged 61-72 years (n = 5) compared with those from men
aged 21-40 years (n = 5) (Patki et al., 2004). McLean and Le Couteur (2004) suggested that the
reduction in phase I enzyme activity may be related not to deficits intrinsic to the liver
microsomal monooxygenase systems, but rather to structural changes in the liver with age (e.g.,
thickening and defenestration of the sinusoidal endothelium of the liver) that may reduce oxygen
availability for phase I enzymes that are directly dependent on oxygen supply as a substrate.
Studies in experimental animals also provide evidence of age-related changes in CYP
activity. Although no significant age-related variations in hepatic CYP2E1 mRNA content were
noted in adult (18-month-old) male Wistar rats compared with 8-month-old rats, CYP2E1
activity (assayed as chlorzoxazone oxidation) was significantly reduced by 46% in the older
group (Wauthier et al., 2004); this study found no age-related changes for hepatic CYP3A1, 3A2,
3A9, or 3A23 mRNA or protein levels in rats. Wauthier et al. (2004) attributed age-related
reductions in hepatic CYP2E1 activity to posttranslational modifications, possibly from the
reactive oxygen species commonly generated by this CYP. A photoperiodicity study reported
that increases in hepatic CYP3 A-dependent erythromycin N-demethylase activity, which is
elevated after Wistar rats are exposed to a dark cycle, were twofold lower in the livers of 22-
month-old rats compared with 10-week-old rats (Martin et al., 2003). Total immunoreactive
CYP3A content was reduced in the hepatic microsomes of 2-year-old compared with 1-year-old
male CD-I mice and was associated with a reduced clearance of the substrate midazolam
(Warrington et al., 2000). These results suggest that the metabolism of carbon tetrachloride
would be slower in the liver of old compared with younger organisms.
Warrington et al. (2004) compared age-related changes in microsomal CYP3A and
NADPH-reductase in the liver and kidney in male F344 rats at 2-4 months (young), 13-14
months (intermediate), and 25-26 months (old). Expression of CYP3A protein in the kidney was
only 1% of that in the liver. The net CYP3A content of the liver was significantly reduced in old
171 DRAFT - DO NOT CITE OR QUOTE
-------
rats compared with young or intermediate rats and involved both immunodetectable bands in
western blots. Conversely, a 50% increase in one isoform of CYP3A was detected in the kidneys
of old rats compared with the intermediate group; an 11% net increase in renal CYP3A was not
statistically significant. Age-related decreases (by 23-36%) in the expression of NADPH-
reductase occurred in the liver and kidney of male F344 rats, but compared with that in young
rats, the decline was statistically significant only in the liver of old rats (Warrington et al., 2004).
The results of this study suggest that the capacity to initiate the metabolism of carbon
tetrachloride is reduced in the liver but possibly increased in the kidney of older organisms
compared with younger animals.
Antioxidant content is also reduced in aging animals compared with younger life stages.
Hepatic GSH content was 35% lower in 24-28-month-old male F344 rats compared with 2-5-
month-old rats (Suh et al., 2004); the decline was related to significant decreases in the level and
activity of y -glutamylcysteine ligase (GCL), the rate-controlling enzyme in the synthesis of
GSH. The ultimate reduction in enzyme activity in old rats was related to an age-related
decrease in a transcription factor, nuclear factor erythroid-related factor 2, that governs the
expression of GCL (Suh et al., 2004). In the liver of 18-month-old male Wistar rats, the GSH
content was significantly reduced by 34% compared with 8-month-old rats, and the level of
TEARS was increased by 287% compared with that in 3-month-old rats (Wauthier et al., 2004).
One study reported a significant age-related reduction in GSH peroxidase activity in the kidney,
but not the liver, of 24-month-old male F344 rats compared with 6-month-old rats (Tian et al.,
1998). Significant decreases in GSH (-20 and -15%), GSH peroxidase activity (-59 and -37%),
and increases in TEARS (+54 and +23%) were noted, respectively, in the liver and kidney of 22-
month-old Wistar rats compared with those of 10-week-old animals (Martin et al., 2003). These
studies suggest that older animals are at greater risk than younger animals of oxidative damage
following exposure to carbon tetrachloride. Studies vary as to whether the age-related changes
are more significant in the kidney or liver, possibly because of strain differences.
In general, aging is associated with constriction of the kidney arterioles and reduced renal
blood flow as well as with reductions in kidney mass and the number of functioning nephrons
(U.S. EPA, 2001b). The result of these changes is a decrease in glomerular filtration rate.
Because of their reduced glomerular function, aged adults are likely to be more sensitive than
younger adults to a chemical, such as carbon tetrachloride, that targets the glomerulus. The
manifestations of renal disease in 2-year-old rats that had been exposed to high concentrations of
carbon tetrachloride in air for most of their lifetimes were increased severity of glomerular
lesions associated with aging (progressive glomerulonephrosis) and impaired glomerular
function (decreased glomerular filtration rate, as indicated by increases in serum levels of BUN,
creatinine, and inorganic phosphorous) in comparison with untreated concurrent controls.
Whether older populations would likely be more susceptible to carbon tetrachloride
toxicity is difficult to determine. Evidence for a reduction in CYP3 A and CYP2E1 activity in
172 DRAFT - DO NOT CITE OR QUOTE
-------
the liver with age would suggest an age-related reduction in the generation of reactive
metabolites from carbon tetrachloride and possibly a corresponding reduction in susceptibility;
however, evidence for reduction in antioxidant content in aging animals would result in a relative
increased risk of oxidative damage in older animals. Functional changes in the kidney with age
and increases in kidney CYP3 A activity (as suggested by experimental animal studies) indicate
that older populations may be at greater risk of carbon tetrachloride-associated kidney damage.
4.8.3. Possible Gender Differences
The extent to which men and women differ in susceptibility to carbon tetrachloride
toxicity is not known. No human data are available to suggest there are gender differences in the
toxicity or carcinogenicity of carbon tetrachloride.
Animal subchronic and chronic toxicity studies by the oral or inhalation route did not
report any significant gender differences in susceptibility to cancer or noncancer effects from
carbon tetrachloride. One study in rats exposed by i.p. injection measured a 2.5-fold increase in
the serum level of hepatic enzymes, a longer period of hepatic injury, and more evidence of
hepatic regeneration in females compared with males (Moghaddam et al., 1998); male livers had
20% more CYP2E1 activity than female livers. The significance of this observation is uncertain,
given the modest difference and the absence of other corroborating data. There was no basis for
assuming gender differences in susceptibility.
4.8.4. Nutritional Status
Fasting or food deprivation has been shown to increase the toxicity of carbon
tetrachloride, as demonstrated by histopathology of the liver, increased serum enzyme levels, or
increased generation of chloroform (Qin et al., 2007; Seki et al., 2000; Shertzer et al., 1988; Sato
andNakajima, 1985; Pentz and Strubelt, 1983; Yoshimine and Takagi, 1982). Decreasing levels
of GSH have been detected in food-restricted animals (Gonzalez-Reimers et al., 2003; Harris and
Anders, 1980; Nakajima and Sato, 1979). The basis for the increased toxicity caused by fasting
is the increase in lipolysis, which generates acetone, an inducer of CYP2E1 (Bruckner et al.,
2002). Bruckner et al. (2002) established that a circadian rhythmicity of vulnerability to carbon
tetrachloride in rats was based on the increased levels of acetone that occur during overnight
fasting. Peak levels of serum SDH, ALT, and isocitrate dehydrogenase were significantly higher
in fasted rats than in fed rats (for example, peak SDH levels were 7 times higher with fasting).
Fasted rats also showed significantly more covalent binding of radiolabeled carbon tetrachloride
to microsomal protein and significantly higher CYP2E1 activities.
Carbon tetrachloride toxicity is also affected by the level of antioxidants in the diet. Rats
fed a diet low in vitamin E, methionine, and selenium (a cofactor for GSH reductase) showed an
increase in lipid peroxidation and liver damage that was reversed by supplementing the diet with
one or more of the antioxidants (Parola et al., 1992; Sagai and Tappel, 1978; Hafeman and
173 DRAFT - DO NOT CITE OR QUOTE
-------
Hoekstra, 1977; Taylor and Tappel, 1976). Addition of vitamin A (retinoic acid or retinol) to
basal diet reduced the hepatic effects of carbon tetrachloride in mice (Rosengren et al., 1995;
Kohno et al., 1992), although it had the opposite effect in rats (Badger et al., 1996; El Sisi et al.,
1993a, b).
Dietary mineral content can also be important. Rats fed a diet deficient in zinc showed
an increase in hepatotoxicity from carbon tetrachloride (DiSilvestro and Carlson, 1994). Cabre
et al. (2000) assessed the time course of hepatic lipid peroxidation and GSH metabolism in
Wistar rats injected with 0.5 mL of carbon tetrachloride to induce hepatic cirrhosis. Inclusion of
zinc in the diet delayed the appearance of cirrhosis and prevented the rise in lipid peroxides. The
protective effect of zinc was independent of GSH levels, which were reduced by carbon
tetrachloride.
4.8.5. Disease Status
Based on experimental findings from rodent studies, there is some reason to suspect that
people with diabetes may have altered susceptibility to hepatotoxic effects from carbon
tetrachloride. Studies in rats have found that rats made diabetic by pretreatment with the
diabetogenic agents alloxan or streptozotocin display markedly enhanced hepatotoxicity in
comparison with nondiabetic rats (Sawant et al., 2007, 2004; Watkins et al., 1988; Hanasono et
al., 1975). The relevance of this finding to humans is uncertain, although it has been reported
that diabetics have nearly twofold higher risk of acute liver failure due to drug-induced toxicities
and chronic liver disease (Sawant et al., 2007). Streptozotocin-induced diabetes not only failed
to enhance the hepatotoxicity of carbon tetrachloride but actually protected against lethality of
the compound in mice (Shankar et al., 2003; Gaynes and Watkins, 1989).
There has been some investigation of the mechanism by which diabetes potentiates
carbon tetrachloride hepatotoxicity in rats. Diabetic rats do not gain weight as normal rats do,
raising the possibility that the enhanced toxicity in diabetic rats is a result of associated
starvation (see Section 4.8.4). However, data for a pair-fed control group in the Hanasono et al.
(1975) study showed that the restriction in food intake could account for only a small portion of
the observed hepatotoxicity in diabetic Sprague-Dawley rats. (Diabetes was induced by
treatment with alloxam monohydrate or streptozotocin.) Treatment of diabetic rats with insulin
controlled the diabetic state and prevented any enhancement of carbon tetrachloride
hepatotoxicity in these rats (Watkins et al., 1988; Hanasono et al., 1975), suggesting that the
diabetic state and not the presence of inducer chemicals potentiates carbon tetrachloride
hepatotoxicity. Serum glucose levels in the diabetic rats were not sensitive predictors of the
extent of hepatotoxicity in the Hanasono et al. (1975) study (e.g., 40 mg alloxan and 65 mg
streptozotocin produced similar plasma glucose levels, but the increase in serum ALT associated
with carbon tetrachloride treatment was twofold higher in the latter experiment), suggesting that
other metabolic effects of diabetes are more important to the effect on carbon tetrachloride
174 DRAFT - DO NOT CITE OR QUOTE
-------
toxicity.
Because ketones and compounds metabolized to ketones have been found to potentiate
the toxicity of carbon tetrachloride and other haloalkanes (see Section 4.8.6), presumably by
enhancing expression of CYP2E1 leading to increased activation of the hepatotoxicant, it has
been suggested that ketosis associated with diabetes might be responsible for the observed effect
(Hewitt et al., 1980). However, there are several lines of evidence suggesting that ketonemia and
increased bioactivation may not be the critical features of diabetes leading to enhanced toxicity
of carbon tetrachloride. In the study by Hanasono et al. (1975), alloxan and streptozotocin both
potentiated carbon tetrachloride-induced hepatotoxicity, even though alloxan-induced diabetes in
rats is characterized by a marked persistent increase in ketone bodies and streptozotocin-induced
diabetes is not. Both alloxan and streptozotocin have been reported to decrease CYP450 activity
(Watkins et al., 1988; Hanasono et al., 1975). Sawant et al. (2004) found no effect on hepatic
microsomal CYP2E1 levels or activity, lipid peroxidation, GSH, or covalent binding of carbon
tetrachloride in the liver in rats with streptozotocin-induced diabetes. Time course studies
performed by Sawant et al. (2004) found that the initial liver injury produced by carbon
tetrachloride in diabetic rats was similar to that in nondiabetic rats but that the effect progressed
only in the diabetic rats. Sawant et al. (2007) reported that liver injury initiated by non-lethal
doses of carbon tetrachloride progressed to hepatic failure and death of diabetic Sprague-Dawley
rats because liver cells failed to advance from Go/Gi to S-phase, thereby unabling S-phase DNA
synthesis (a critical step in cell division) and inhibiting tissue repair. A more detailed
understanding of the mechanism would be needed to predict how diabetes might affect carbon
tetrachloride toxicity in humans.
4.8.6. Exposure to Other Chemicals
Factors that increase the expression of CYP2E1 or CYP3A are likely to increase
susceptibility to carbon tetrachloride exposure (all other things being the same) because the
relatively higher rate of metabolism on a per cell basis would significantly increase the rate of
generation of trichloromethyl radicals in the liver and kidney. Heavy consumers of ethanol,
which induces CYP2E1, are therefore more vulnerable to carbon tetrachloride (Manno et al.,
1996). Manno et al. (1996) described case reports of two workers who consumed 120 or 250
grams of ethanol per day and were the only individuals to develop severe hepatotoxicity and
nephrotoxicity following a 2-hour exposure to carbon tetrachloride vapors used in a fire
extinguisher (Manno et al., 1996); their nonsymptomatic colleagues, who also were exposed,
consumed <50 g ethanol/day. Cases of acute carbon tetrachloride poisoning often involved
individuals who were alcohol consumers (New et al., 1962). Enhanced toxicity from
concomitant or preceding ethanol consumption and exposure to carbon tetrachloride has been
verified in animal studies (Wang et al., 1997; Plummer et al., 1994; Hall et al., 1991; Ikatsu et
al., 1991; Kniepert et al., 1990; Reinke et al., 1988; Sato andNakajima, 1985; Strubelt, 1984;
175 DRAFT - DO NOT CITE OR QUOTE
-------
Teschke et al., 1984; Harris and Anders, 1980; Sato et al., 1980).
Potentiation of carbon tetrachloride hepatotoxicity has also been observed following
exposure to other chemical inducers of CYP450, including isopropanol which converts to
acetone (Rao et al., 1996; Folland et al., 1976; Traiger and Plaa, 1971), methanol (Allis et al.,
1996; Harris and Anders, 1980), 2-butanol (Traiger and Bruckner, 1976), tert-butanol (Ray and
Mehendale, 1990; Harris and Anders, 1980), and other aliphatic alcohols (Ray and Mehendale,
1990); acetone, methyl ethyl ketone, methyl isobutyl ketone, 2-butanone, and other ketones
(Raymond and Plaa, 1995; Charbonneau et al., 1986; Pilon et al., 1986; Plaa and Traiger, 1972);
phenobarbital (Abraham et al., 1999; Sundari et al., 1997; Hocher et al., 1996; Cornish et al.,
1973; Garner and McLean, 1969); DDT (McLean and McLean, 1966); polychlorinated and
polybrominated biphenyls (Kluwe et al., 1979); and mirex and chlordecone (Soni and
Mehendale, 1993; Kodavanti etal., 1992; Mehendale, 1992, 1991, 1990; Bell and Mehendale,
1987, 1985; Curtis et al., 1979). Coexposure to nicotine in drinking water also increased hepatic
effects of carbon tetrachloride, although this was thought to be because of a synergistic effect on
lipid peroxidation produced by both chemicals rather than induction of CYP450 (Yuen et al.,
1995).
There is also limited evidence for a reduction in carbon tetrachloride hepatotoxicity
associated with reduced bioactivation of the chemical. Coexposure to carbon tetrachloride and
carbon disulfide both in rats and human workers resulted in hepatic and neurological effects
associated with carbon disulfide but no effects characteristic of carbon tetrachloride (Peters et al.,
1987; Seawright et al., 1980). The researchers attributed this result to destruction of CYP450 by
carbon disulfide and reduced bioactivation of carbon tetrachloride. Pretreatment with LN
reduced the hepatotoxicity of carbon tetrachloride, apparently because of the ability of lead to
inhibit CYP450 (Calabrese et al., 1995).
176 DRAFT - DO NOT CITE OR QUOTE
-------
5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)
5.1.1. Choice of Principal Study and Critical Effect—with Rationale and Justification
Epidemiological studies of long-term exposure to carbon tetrachloride are inadequate to
establish whether an association exists between oral exposure and adverse birth outcomes (the
only health outcome evaluated in these studies). Case reports of human poisoning identify the
liver and kidney as primary target organs of acute carbon tetrachloride exposure, but do not
provide data useful for dose-response analysis.
Several subchronic oral toxicity studies, including Bruckner et al. (1986), Condie et al.
(1986), Hayes et al. (1986) and Allis et al. (1990), provide liver toxicity data that were
considered for dose-response analysis. Hayes et al. (1986) and Allis et al. (1990) reported liver
toxicity at the lowest dose tested (i.e., a NOAEL was not identified) and are thus less suitable for
defining a point of departure (POD) for the RfD. Further, in the Hayes et al. (1986) study, which
included both a vehicle (corn oil) and untreated control group, the vehicle controls themselves
had significantly elevated serum enzyme levels, altered organ weights, and increased incidence
of liver necrosis. This type of corn oil vehicle response was not seen in other studies. The Allis
et al. (1990) protocol also provided data less amenable to dose-response analysis. Male rats were
sacrificed in groups of six at various time points after exposure was terminated (1, 3, 8, and 15
days), and results at these various time points could not be combined.
Subchronic gavage studies by Bruckner et al. (1986) in male rats and Condie et al. (1986)
in male and female mice provided the best available characterizations of the dose response for
ingested carbon tetrachloride at low doses. Bruckner et al. (1986) identified a NOAEL of
1 mg/kg and a LOAEL of 10 mg/kg in rats administered carbon tetrachloride 5 days/week by
gavage in corn oil (0.71 and 7.1 mg/kg-day, respectively, adjusted to daily exposure). Condie et
al. (1986) identified a NOAEL of 1.2 mg/kg and a LOAEL of 12 mg/kg in similarly treated mice
(0.86 and 8.6 mg/kg-day, respectively, adjusted to daily exposure). In both studies, the LOAEL
of 10-12 mg/kg (average daily dose of 7-9 mg/kg-day) produced hepatotoxicity, indicated by
increased serum activity of enzyme markers of liver damage and direct histopathological
determination of liver lesions. More marked effects on the liver were found at higher doses in
both studies. Liver effects were also observed in numerous other studies in animals. The
LOAELs from Bruckner et al. (1986) and Condie et al. (1986) are consistent with the LOAELs
from Hayes et al. (1986) [12 mg/kg-day] and Allis et al. (1990) [14.3 mg/kg-day].
5.1.2. Methods of Analysis—Including Models
The most sensitive endpoints identified for effects of carbon tetrachloride by oral
exposure relate to liver toxicity (including serum enzyme changes and liver histopathology) in
177 DRAFT - DO NOT CITE OR QUOTE
-------
the subchronic corn oil gavage studies of Bruckner et al. (1986) in male rats and Condie et al.
(1986) in male and female mice. Sensitive endpoints in both studies were evaluated for
suitability for benchmark dose (BMD) modeling. For suitable data sets, BMD modeling
methodology (U.S. EPA, 2000c, 1995) was used to analyze the data.
Serum Chemistry Data
Condie et al. (1986) reported liver enzyme changes in carbon tetrachloride-exposed mice;
however, the median of 8-12 determinations was reported without a standard error (SE) or
standard deviation (SD) (only the minimum and maximum of the range were reported). Without
a mean and SE or SD, BMD analysis cannot be performed. Therefore, the NOAEL of
1.2 mg/kg-day, 5 days/week, was identified as a possible POD for this data set.
Serum chemistry data in male rats from Bruckner et al. (1986) are presented in Table 5-1.
Of the enzymes monitored, only SDH showed a clear statistically and biologically significant
increase in the 10 mg/kg dose group. The data for the 10- and 12-week blood draws were
similar. Therefore, both 10- and 12-week data were used for dose-response modeling using
BMD methods.
Serum activity of SDH is widely used in toxicity studies as an indicator of hepatocellular
injury. It is a specific and sensitive biomarker of liver damage. SDH is located in the cytosol
and mitochondria of liver cells. It is found at low levels in normal serum and erythrocytes.
Presence of increased activity in serum indicates leakage from hepatocytes secondary to cell
damage. In acute studies with carbon tetrachloride, serum SDH activity was a particularly
sensitive indicator of liver toxicity, with increases found at doses similar to, or even lower than,
those producing cellular damage visible by light microscopy (Paustenbach et al., 1986b; Korsrud
et al., 1972). In the Bruckner et al. (1986) study, the lowest administered dose at which serum
SDH activity was increased was also the lowest dose at which liver lesions were observed.
Use of elevated serum SDH activity as a critical effect for derivation of the RfD is
supported by results of a study examining the use of serum liver enzymes as predictors of
hepatotoxicity (Travlos et al., 1996). The relationship between the activity of serum liver
enzymes (ALT, SDH, ALP, and TEA) and liver histopathology was examined for 50 chemicals
and three chemical mixtures using 1-, 2-, 3-, and 13-week clinical chemistry measurements and
13-week histopathology assessments in male and female F344 rats, although carbon tetrachloride
was not tested. Treatment-related changes in serum liver enzymes were determined using the
Jonksheere-Terpstra trend test at the 0.05 level or Dunn's test at the 0.01 level; serum liver
enzyme activities were not reported. An association was observed between treatment-related
increases in SDH and ALT activities and the development of histopathological changes to the
liver. SDH was a more sensitive predictor of histopathological changes than ALT, with SDH
activity predicting 13-week histopathological changes in rats of both sexes with 76-92%
accuracy, compared with 56-83% accuracy for ALT. If both SDH and ALT were elevated,
178 DRAFT - DO NOT CITE OR QUOTE
-------
positive terminal histopathological changes were predicted with 100% accuracy from the 2-, 3-,
and 13-week clinical chemistry measurements. TEA and ALP were predictive of histopathology
results with 20-85 accuracy and 29-82% accuracy, respectively. Based on these findings,
statistically significant elevations in serum SDH and ALT activity are sensitive markers for liver
toxicity, with SDH predicting histopathological changes to the liver with higher accuracy than
ALT. As shown in Table 5-1, serum liver enzyme activity for SDH in the Bruckner et al. (1986)
study was significantly elevated after 10 and 12 weeks of exposure in the mid- and high-dose
groups and ALT was significantly elevated in the mid- and high-dose groups after 12-weeks of
exposure. In addition, treatment-related histopathologic findings were observed in the mid-dose
group (lipid vacuolization), with more extensive findings in the high-dose group (lipid
vacuolization, nuclear and cellular pleomorphism, bile duct hyperplasia, and periportal fibrosis)
after 12-weeks of exposure (see Liver Histopathologic Changes below). Thus, carbon
tetrachloride-induced elevations in SDH and ALT are valid markers of histopathological changes
to the liver in the Bruckner et al. (1986) study.
Table 5-1. Serum enzyme data in male rats after 10- or 12-week exposure
to carbon tetrachloride
Daily dose
(mg/kg-day)
0
1
10
33
SDH (IU/mL)a
10 weeks
3. 5 ±0.4
2.3 ±0.6
7.6±2.5b
134.8 ±15.0b
12 weeks
3.2 ±0.4
1.9±0.1
8.7±2.0b
145.7 ± 57. 9b
OCT (nmol CO2/mL)a
10 weeks
28 ±8
23 ±3
55 ±10
148±48b
12 weeks
45 ±4
61 ±12
69 ±16
247±31b
ALT (IU/mL)a
10 weeks
18±1
20 ±1
23 ±1
617 ±334
12 weeks
20 ±0.3
19±1
27 ± 2b
502±135b
aValues presented are mean ± standard error for groups of five rats at 10 weeks and seven to nine rats at 12 weeks.
V<0.05.
SDH = sorbitol dehydrogenase
OCT = ornithine carbamoyl transferase
ALT = alanine aminotransferase
Source: Bruckner et al. (1986).
All of the models for continuous data in U.S. EPA's benchmark dose software (BMDS)
(version 1.4.1) (U.S. EPA, 2007b) were fit to the 10- and 12-week SDH data. An increase in
SDH activity two times the control mean, representing an increase in serum enzyme level
considered to be biologically significant, was used as the benchmark response (BMR). Several
expert organizations, particularly those concerned with early signs of drug-induced
hepatotoxicity, have identified an increase in liver enzymes compared with concurrent controls
of two to fivefold as an indicator of concern for hepatic injury (EMEA, 2006; Boone et al., 2005;
179
DRAFT - DO NOT CITE OR QUOTE
-------
FDA Working Group, 2000). Dr. James Bruckner, University of Georgia and principal
investigator of the study used to derive the RfD, considered a twofold increase in SDH to be an
indication of a lexicologically significant response (personal communication, November 7, 2006,
with Susan Rieth, U.S. EPA). Because ALT is the liver enzyme that is generally measured
clinically, most expert organizations similarly focus on ALT as an indicator of liver injury in
preclinical (animal) studies. Because SDH, like ALT, is one of the more specific indicators of
hepatocellular damage in most animal species and generally parallels changes in ALT in toxicity
studies where liver injury occurs, a similar twofold increase in SDH is considered indicative of
liver injury in experimental animals.
BMD modeling results for the 10- and 12-week SDH data are presented in Appendix B.
None of the models for continuous data in BMDS provided an adequate fit of the 12-week SDH
data. The 3rd degree polynomial and power models provided adequate fits of the 10-week SDH
data (based on a goodness-of-fit p-value of >0.1). The power model provided the better fit of the
data (based on the lower Akaike's Information Criterion [AIC] value) and was therefore selected
as the basis for a candidate POD; this model estimated a BMD2X of 7.32 mg/kg-day and the 95%
lower confidence limit on the BMD (BMDL2x) of 5.46 mg/kg-day.
BMD modeling was also performed using the 10- and 12-week OCT and ALT data from
Bruckner et al. (1986) (see Appendix B for a more detailed summary of model results). OCT
data could not adequately be fit by the models available in BMDS. The power model provided
an adequate fit of the 10-week ALT data, yielding a BMD2x and BMDL2X of 14.7 and 13.21
mg/kg-day, respectively; however, as shown in Table 5-1, the standard error of the mean (SEM)
ALT for the high-dose (33 mg/kg-day) male rats was extremely large (617 ± 334). Bruckner et
al. (1986) noted: "There was a pronounced rise in GPT [ALT] at 10 and 12 weeks. Scrutiny of
values of individual animals revealed that dramatic increases in two rats at each time point were
largely responsible for the late increase in GPT [ALT] activity." In light of the large variation in
response at 33 mg/kg-day, relatively high uncertainty is associated with this quantitative analysis
using the 10-week ALT data set. The polynomial and power models provided adequate fits of
the 12-week ALT data (based on a goodness-of-fit p-value >0.1). The polynomial model, which
provided a better fit (based on lower AIC values) of the data using both n = 7 and 9, estimated a
BMD2x and BMDL2x of 13.0 and 11.8 mg/kg-day, respectively. The values of the BMD and
BMDL were not sensitive to the value of n.
Overall, a dose-response analysis of 10- and 12-week liver enzyme data from Bruckner et
al. (1986) reveals that the 10-week SDH data provide the most sensitive estimates of the BMD2x
and BMDL2x, or 7.32 and 5.46 mg/kg-day, respectively. For purposes of comparison across
chemicals, the BMD and BMDL corresponding to a change in the mean response equal to one
control SD from the control mean were also calculated for the 10-week SDH data, consistent
with BMD guidance (U.S. EPA, 2000c). The BMDiso and BMDLiso were 5.5 and 3.8 mg/kg-
day, respectively.
180 DRAFT - DO NOT CITE OR QUOTE
-------
Liver Histopathologic Changes
Liver lesion incidence data from the Bruckner et al. (1986) study in male rats and the
Condie et al. (1986) study in male and female mice exhibit inductions of hepatic lesions due to
carbon tetrachloride at 10-12 mg/kg. Table 5-2 presents liver pathology data from the Bruckner
et al. (1986) study. Data were displayed as mean severity scores. Incidence data were not
presented directly, although it can be inferred that incidence was 0% where severity is 0. In
addition, a statement in Bruckner et al. (1986) implied that incidence was 100% for lipid
vacuolation in the 10 mg/kg dose group.
Table 5-2. Severity of liver lesions in male rats after 12-week exposure to
carbon tetrachloride
Daily dose
(mg/kg-day)
0
1
10
33
Lipid vacuolationa
Ob
0
3.7C
4
Nuclear and cellular
pleomorphism"
0
0
0
5.7
Bile duct
hyperplasia"
0
0
0
4
Periportal fibrosis"
0
0
0
3.7
""Severity graded from 0 (absent) to 8 (severe); values presented are means for groups of 6-7 rats.
b Severity score of 0 implies incidence of 0%.
0 Text reports that "each animal" in this group showed the lesion, implying incidence of 100%.
Source: Bruckner et al. (1986).
It can be seen that lipid vacuolation was the only lesion to occur in the 10 mg/kg group,
making this the most sensitive pathology endpoint in the study, and that the incidence (not
reported but assumed from the text of the paper) of this lesion increased from 0% at 1 mg/kg to
100% at 10 mg/kg.
In the Condie et al. (1986) study, exposure to carbon tetrachloride by gavage in corn oil
or Tween-60 aqueous emulsion produced a variety of liver lesions (hepatocellular vacuolization,
inflammation, hepatocytomegaly, necrosis, portal bridging fibrosis) in male and female mice at
the high dose of 120 mg/kg. However, only necrosis (minimal to mild) in males and
hepatocytomegaly (severity unranked) in males and females treated using a corn oil vehicle
occurred with statistically elevated incidence in the 12 mg/kg dose group. Incidence data for
these lesions, which represent the most sensitive effects of carbon tetrachloride in mice, are
shown in Table 5-3. For all three of these histopathologic lesions, incidence increased from 0%
in the 1.2 mg/kg group to 60-90% in the 12 mg/kg group.
181
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-3. Incidence of selected liver lesions in mice treated with carbon
tetrachloride for 90 days
Sex
M
M
F
Vehicle
Corn oil
Corn oil
Corn oil
Lesion
Necrosis
Hepatocytomegaly
Hepatocytomegaly
Incidence at daily dose
0 mg/kg-day
0/10
0/10
0/10
1.2 mg/kg-day
0/9
0/9
0/9
12 mg/kg-day
9/1 Oa
8/1 Oa
6/1 Oa
120 mg/kg-day
9/1 Oa
10/10a
9/9a
a p < 0.05 by Fisher's exact test conducted for EPA.
M=males, F=females
Source: Condie et al. (1986).
The histopathology data from Bruckner et al. (1986) and Condie et al. (1986) are,
therefore, consistent with a POD between 1 and 10 mg/kg in male rats and 1.2 and 12 mg/kg in
mice, but do not provide sufficient information on response in the vicinity of the BMR (typically
10% for quantal data) (U.S. EPA, 2000c) to objectively inform the shape of the dose-response
curve in the region of interest. At the LOAELs (approximately 10-12 mg/kg-day) in these
studies, the response rate was 60-100%, whereas the response at the dose below the LOAEL was
0%. The incidence data do, however, support the BMD2X of 7.32 mg/kg and BMDL2X of
5.46 mg/kg estimated from the increase in 10-week serum SDH observed in the Bruckner et al.
(1986) study.
The NOAEL of 1.2 mg/kg-day, 5 days/week for liver enzyme changes from Condie et al.
(1986) was considered as a POD for the carbon tetrachloride RfD; however, the data provided in
the study report were insufficient to allow BMD modeling of these data. Because the NOAELs
and LOAELs in the Bruckner et al. (1986) study were similar to those in Condie et al. (1986),
and because the data reported in Bruckner et al. (1986) supported BMD modeling and thus
provided better resolution of the dose-response relationship in the low-dose region, the BMDL2x
based on Bruckner et al. (1986) was selected as the POD for the carbon tetrachloride RfD.
Consideration of PBPK Models for Interspecies Extrapolation
Three PBPK models of oral exposures have been reported; two rat models (Semino et al.,
1997; Gallo et al., 1993) and a mouse model (Fisher et al., 2004). These models implement
different approaches to simulate the complex kinetics of absorption of carbon tetrachloride that
follows an oral gavage dose of carbon tetrachloride in corn oil or emulsifiers (e.g., Emulphor).
Oral absorption of carbon tetrachloride in corn oil (and Emulphor) exhibits a pulsatile behavior,
evident from multiple peaks of carbon tetrachloride concentrations in blood that occur during the
first 12-20 hours following a gavage dose (Fisher et al., 2004; Semino et al., 1997; Gallo et al.,
1993). Semino et al. (1997) successfully modeled this pulsatile behavior in the rat with a multi-
182
DRAFT - DO NOT CITE OR QUOTE
-------
compartment model in which first-order absorption from 6 to 9 compartments was scheduled at
different times following the dose (i.e., absorption was zero until the scheduled activation of
each compartment). The scheduling was accomplished using the SCHEDULE command in
Advanced Continuous Simulation Language (ACSL), which cannot be implemented repeatedly;
therefore, the implementation is not directly amenable to continuous simulation of multiple
exposures. The approach also required calibration of the model against blood concentration
kinetics for a specific dose of carbon tetrachloride (e.g., 25 mg/kg). The dose-dependence of the
resulting parameter values was not evaluated and, therefore, extrapolation to other dose levels
would be highly uncertain. Gallo et al. (1993) successfully simulated the oral absorption of
carbon tetrachloride in corn oil with multiple zero-order absorption rates (e.g., jig/hour) that
were estimated by fitting to observed blood carbon tetrachloride kinetics. Although this
approach successfully reproduced the blood carbon tetrachloride absorption kinetics following a
25 mg/kg dose to the rat, implementation of this approach would require calibration of the zero-
order absorption rates to each data set (i.e., blood kinetics following the dose levels of interest).
Fisher et al. (2004) simulated oral absorption of carbon tetrachloride in an aqueous emulsion
vehicle (similar to Emulphor) in the mouse with a two-compartment, three-parameter model (see
Figure 3-2). Rate coefficients were estimated by visually fitting these parameters to blood
kinetics following single oral gavage doses of carbon tetrachloride. One of the parameters in the
absorption model was varied with dose in order to simulate dose-dependent absorption kinetics;
as a result, similar to the Gallo et al. (1993) approach, implementation of the two-compartment,
three-parameter model would require calibration to blood kinetics for the dose levels of interest.
The above approaches to simulating oral absorption kinetics of carbon tetrachloride were
not implemented in the dosimetry analysis of oral bioassay data for two major reasons:
(1) predictions of oral absorption kinetics of carbon tetrachloride would be highly uncertain for
doses other than those to which the above models had been specifically calibrated; and
(2) extrapolation of these absorption models to humans also would be highly uncertain. An
alternative approach that simulates a time-averaged daily absorption rate and bioavailability
might suffice for simulating long-term average blood (arterial) concentrations of carbon
tetrachloride that would result from repeated oral exposures to carbon tetrachloride. Estimates of
liver metabolism rates would be less certain, however, since carbon tetrachloride is simulated in
the PBPK models as a nonlinear function of carbon tetrachloride delivery to the liver (i.e., from
absorption and from arterial blood). As a result, large fluctuations in absorption rate could result
in similarly large fluctuations in metabolism rates that may not be accurately represented by
simulations of time-averaged rates of absorption. Therefore, EPA does not consider currently
available PBPK models to be adequate for interspecies dosimetry extrapolations of carbon
tetrachloride administered to animals by gavage (e.g., in corn oil) to continuous exposures in
humans. As described in Section 5.4.3.4, however, the human PBPK model has been used to
extrapolate dosimetry in the human across routes (i.e., inhalation to oral).
183 DRAFT - DO NOT CITE OR QUOTE
-------
As noted above, the BMDL2X of 5.46 mg/kg-day estimated from the increase in serum
SDH activity in male rats in the Bruckner et al. (1986) study was selected as the POD for
derivation of the RfD. The BMDL2X of 5.46 mg/kg-day was derived from a study with an
intermittent dosing schedule. In the absence of a suitable PBPK model, the BMDL is adjusted to
an average daily dose according to the following equation:
BMDL 2X-ADJ = BMDL2X x 5 days/7 days
= 5.46 mg/kg-day x 5 days/7 days
= 3.9 mg/kg-day
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs)
An RfD of 0.004 mg/kg-day for carbon tetrachloride is derived by applying a composite
UF of 1,000 to the BMDL2X-ADJ of 3.9 mg/kg-day, as follows:
RfD = BMDL2X-ADJ/UF
3.9 mg/kg-day/1000
0.0039 mg/kg-day or
0.004 mg/kg-day (rounded to one significant figure)
The composite UF of 1,000 includes a factor of 3 (10°5) to extrapolate from a subchronic
to chronic duration of exposure, a factor of 10 to protect susceptible individuals, a factor of 10 to
extrapolate from rats to humans, and a factor of 3 to account for database deficiencies, lacking an
adequate multigeneration study of reproductive function.
• A default 10-fold UF for intraspecies differences (UFH) was selected to account for
variability in susceptibility among members of the human population in the absence of
quantitative information on the variability of human response to carbon tetrachloride.
Factors that could contribute to a range of human response to carbon tetrachloride were
discussed in Section 4.8. Intrahuman variability in CYP450 levels that are responsible
for metabolism of carbon tetrachloride to reactive metabolites has been documented (see
Section 4.8). This variation in CYP450, which is likely influenced by age-related
differences or other factors (e.g., exposure to other chemicals that induce or inhibit
microsomal enzymes), could alter susceptibility to carbon tetrachloride toxicity.
Individual variability in nutritional status, alcohol consumption, or the presence of
underlying disease could also alter metabolism of carbon tetrachloride or antioxidant
protection systems. To account for these uncertainties, a factor of 10 was included for
individual variability.
184 DRAFT - DO NOT CITE OR QUOTE
-------
A default 10-fold UF for interspecies extrapolation (UFA) was selected to account for
potential pharmacokinetic and pharmacodynamic differences between rats and humans.
Metabolism of carbon tetrachloride to reactive species is the initial key event in the
development of carbon tetrachloride toxicity. Also critical to carbon tetrachloride
toxicity are cellular antioxidant systems that function to quench the lipid peroxidation
reaction, thereby preventing damage to cellular membranes. PBPK models available for
carbon tetrachloride were found unsuitable for repeat-dose oral scenarios, and could not
be used for interspecies extrapolation. In the absence of data to quantify specific
interspecies differences or a suitable PBPK model, a UF of 10 is included.
A UF of 3 (10°5) for subchronic to chronic extrapolation (UFs) was selected based on the
following: (1) Qualitative information demonstrating that the target of toxicity following
chronic oral exposure is the liver. The NCI oral cancer bioassay in rats and mice (NTP,
2007; NCI, 1977, 1976a, b; Weisburger, 1977) did not include an adequate evaluation of
low-dose exposures; in rats, there was marked hepatotoxicity at the lowest dose tested,
and in mice, survival was low in dosed animals because of the high incidence of liver
tumors. For these reasons, the bioassay was not suitable for dose-response analysis.
Nevertheless, complete nonneoplastic incidence data available through an NTP (2007)
database of neoplastic and nonneoplastic data did not identify carbon tetrachloride-related
histopathological changes in any organ systems or tissues other than the liver. Therefore,
the NCI bioassay clearly identified the liver as a target organ following chronic
exposures, consistent with the findings from subchronic oral studies and subchronic and
chronic inhalation studies.
(2) Knowledge of the relationship between effect levels in subchronic and chronic
inhalation studies. The JBRC inhalation bioassay, which included 13-week and 2-year
inhalation studies in rats and mice (Nagano et al., 2007a, b; JBRC, 1998), provides
information on the relationship between NOAELs and LOAELs from subchronic and
chronic exposure durations. In the 13-week study, liver toxicity (increased liver weight
and fatty liver) was observed in rats and mice at the lowest exposure concentration tested
(LOAEL = 2 ppm, duration adjusted). Following chronic exposure, the LOAEL based on
liver and kidney effects was 4 ppm (duration adjusted) and the NOAEL was 0.9 ppm
(duration adjusted); the LOAEL concentration in the chronic study was, in fact, twofold
higher than the LOAEL from the subchronic study. Other subchronic inhalation studies
in rats and mice support a NOAEL in the range of 0.9-4 ppm (see Table 4-14), which is
similar to or within fourfold of the NOAEL from the JBRC chronic inhalation bioassay.
(3) Early onset of liver toxicity. Cytotoxicity occurs early in the sequence of events. For
185 DRAFT - DO NOT CITE OR QUOTE
-------
example, Bruckner et al. (1986) observed increases in liver enzymes and liver cell
vacuolization after 4 days of exposure in an 11-day oral toxicity study, and increases in
liver enzymes at week 2 in a 12-week oral toxicity study.
Thus the data suggest that an increase in the duration of the exposure may not increase
the incidence and/or severity of the liver toxicity.
A UF to account for extrapolation from a LOAEL to a NOAEL (UFL) 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 an increase
in SDH activity 2 times the control mean was selected under an assumption that it
represents a minimal biologically significant change.
A UF to account for deficiencies in the database (UFo) of 3 (10°5) was selected. The oral
database for this chemical includes extensive testing for subchronic toxicity in animals, a
number of tests of immunotoxic potential, limited chronic oral bioassays in both rats and
mice, and limited human data. Developmental toxicity testing by the oral route has been
conducted. Testing for developmental toxicity by two groups of investigators (Narotsky
and Kavlock, 1995; Wilson, 1954) found full-litter resorption at doses accompanied by
some degree of maternal toxicity, ranging from piloerection to mortality. Because both
studies used relatively high doses, neither study identified a NOAEL. The low dose of
carbon tetrachloride (25 mg/kg-day) used in Narotsky et al. (1997b) caused neither
maternal nor developmental effects when administered in either aqueous or corn oil
vehicles, albeit the group sizes (12-14 dams/dose level) were smaller than the group size
used in the typical developmental toxicity study. Nevertheless, the NOAEL in this
developmental study (25 mg/kg-day) exceeds the POD for the RfD based on liver effects
by over 6-fold and the LOAEL (50 mg/kg-day) by 13-fold, and is consistent with
developmental toxicity endpoints as less sensitive than measures of hepatotoxicity. Also,
as noted in Section 4.8.1 (Possible Childhood Susceptibility), the available life stage
information on microsomal enzyme activity, and in particular CYP2E1, suggests that the
developing organism would be no more susceptible to free radical-induced liver injury
from carbon tetrachloride than adults. The carbon tetrachloride database lacks an
adequate multigeneration study of reproductive function by any route of exposure. A
database UFo of 3 was applied to account for the lack of a multigeneration reproductive
toxicity study.
186 DRAFT - DO NOT CITE OR QUOTE
-------
5.1.4. RfD Comparison Information
PODs and potential oral RfDs based on selected studies included in Table 4-13 are
arrayed in Figures 5-1 to 5-3, and provide perspective on the RfD supported by Bruckner et al.
(1986). These figures should be interpreted with caution because the PODs across studies are
not necessarily comparable, nor is the confidence in the data sets from which the PODs were
derived the same. PODs in these figures may be based on a NOAEL, LOAEL, or BMDL (in the
case of the principal study), and the nature, severity, and incidence of effects occurring at a
LOAEL are likely to vary. To some extent, the confidence associated with the resulting potential
RfD is reflected in the magnitude of the total UF applied to the POD (i.e., the size of the bar);
however, the text of Sections 5.1.1 and 5.1.2 should be consulted for a more complete
understanding of the issues associated with each data set and the rationale for the selection of the
critical effect and principal study used to derive the RfD.
The predominant noncancer effect of subchronic and chronic oral exposure to carbon
tetrachloride is hepatic toxicity. Figure 5-1 provides a graphical display of dose-response
information from five studies that reported liver toxicity in experimental animals following
subchronic oral exposure to carbon tetrachloride, including the PODs that could be considered in
deriving the oral RfD. As discussed in Sections 5.1.1 and 5.1.2, among those studies that
demonstrated liver toxicity, the study by Bruckner et al. (1986) provided the data set most
appropriate for deriving the RfD. Possible RfDs that might be derived from each of these studies
are also presented. Although the RfD based on Bruckner et al. (1986) is not the lowest among
candidate studies, it is considered the most scientifically rigorous. The POD is based on BMD
methods, which has an inherent advantage over use of a NOAEL or LOAEL by making greater
use of all the data from the study. Because the studies by Hayes et al. (1986) and Allis et al.
(1990) identified only a LOAEL for liver effects, the RfD associated with these studies is driven
lower by use of a larger composite UF.
Studies in experimental animals have also found that relatively high doses of carbon
tetrachloride during gestation can produce prenatal loss; these doses also produced overt toxic
effects in the dams. A graphical display of dose-response information from three developmental
studies is provided in Figure 5-2.
Figure 5-3 displays PODs for the major targets of toxicity associated with oral exposure
to carbon tetrachloride. For the reasons discussed in Section 5.1.2, liver effects in the rat
observed in the study by Bruckner et al. (1986) are considered the most appropriate basis for the
carbon tetrachloride RfD. The POD is lower than that for developmental toxicity, and the
resulting RfD should adequately protect against developmental effects of carbon tetrachloride.
187 DRAFT - DO NOT CITE OR QUOTE
-------
100
10
t
"3)
£
0.1
0.01
0.001
0.0001
¥
The RfD in the dashed
circle was selected as
the final RfD for carbon
tetrachloride.
• POD
[JTJAnimal-to-human
r~| Human variation
g| LOAEL to NOAEL
Q Subchr to Chronic
| Database deficiencies
O RfD
Condieetal, 1986; 12-
wk mouse study (corn
oil gavage); liver
enzyme activity,
histopathology; NOAEL
Hayes etal, 1986; 13-
wk mouse study; liver
wt, enzyme activity,
Allis etal, 1990; 12-wk
rat study; liver wt,
enzyme activity,
Bruckner etal, 1986;
12-wk rat study; SDH
activity; BMDL-2x
histopathology; LOAEL histopathology; LOAEL
(1)
(2)
Condieetal, 1986; 12-
wk mouse study
(Tween-60 gavage);
liverwt, enzyme activity,
histopathology; NOAEL
(1) Magnitude of effect at the LOAEL: liver weight (f 15-19%); enzyme activity (f <6-fold); 100% necrosis.
(2) Magnitude of effect at the LOAEL: liver weight (j 30%); enzyme activity (f 3-5X); 100% necrosis.
Figure 5-1. PODs (mg/kg-day) with corresponding derived oral reference values that would result if liver toxicity
was used as the critical effect.
188
DRAFT - DO NOT CITE OR QUOTE
-------
1000.00
0.01
• POD
IHTH Animal-to-human
| | Human variation
^ LOAEL to NOAEL
I I Subchr to Chronic
^| Database deficiencies
O RfD
Narotskyand Kavlock, 1995; rat;
GD6-19; resorptions; LOAEL (1)
Narotskyetal, 1997b; rat; GD6-15
(corn oil gavage); resorptions;
NOAEL
Narotskyetal, 1997b; rat; GD6-15
(Emulphor gavage); resorptions;
NOAEL
(1) Magnitude of effect at the LOAEL: 44% resorptions
Figure 5-2. PODs (mg/kg-day) with corresponding derived oral reference values that would result if
developmental toxicity was used as the critical effect.
189
DRAFT - DO NOT CITE OR QUOTE
-------
100
10
O)
j*:
"01
E.
0)
M
Q
0.1
0.01
0.001
/
/
/
/ ,
/
/
/
/
/
/
1
1
\
\
\ 1
\
\
\
\
/
irffl
s.
-**. ^ .
nnf
X
\
\
\
\
tin ^
\
\
\
\
\
\
\
I
•
i
• /
1 /
• /
' !
i t
t
Ifffl
I
i
r
Fhe RfD in the dashed
;ircle was selected as
he final RfD for carbon
etrachloride.
• POD
[TJ] Animal-to-human
| | Human variation
^LOAELtoNOAEL
| | Subchr to Chronic
| Database deficiencies
ORfD
POD based on:
Liver: BMDLfor
elevated serum SDH
activity in the rat; 12-
week oral (gavage)
study (Bruckner et al.,
1986)
Developmental:
NOAELfor
resorptions in the rat;
exposure on GD 6-15
(Narotsky et al,
1997b)
Liver
Developmental
Figure 5-3. PODs (mg/kg-day) with corresponding derived oral reference values that would result if alternative
endpoints were used as the critical effect.
190
DRAFT - DO NOT CITE OR QUOTE
-------
5.1.5. Previous RfD Assessment
The previous oral RfD for carbon tetrachloride (verified on 05/20/85 and posted on the
IRIS database in 1987) was 0.0007 mg/kg-day, based on the NOAEL of 1 mg/kg (daily dose of
0.7 mg/kg-day) and the LOAEL of 10 mg/kg (daily dose of 7 mg/kg-day) for liver lesions
(evidenced by mild centrilobular vacuolation and significantly increased serum SDH activity) in
rats treated for 12 weeks (5 days/week) with carbon tetrachloride by gavage in corn oil by
Bruckner et al. (1986). (A 1983 draft of the Bruckner et al. (1986) study was used as the basis
for the RfD by the RfD Work Group. The published version of the study did not necessitate a
change to the RfD.) The RfD of 0.0007 mg/kg-day was calculated by applying a UF of 1,000
(three factors of 10 to account for interspecies and interhuman variability and extrapolation from
subchronic to chronic exposure) to the NOAEL of 0.7 mg/kg-day.
The current RfD relies on the same principal study as the previous RfD, but applies
benchmark dose analysis to derive the POD (3.9 mg/kg-day), whereas the previous RfD used the
NOAEL (0.7 mg/kg-day) as the POD. Both RfDs were derived using a total UF of 1000,
although some of the individual UFs differed. The previous RfD incorporated a UF of 10 to
account for extrapolation from subchronic to chronic exposure, whereas the current RfD includes
a subchronic to chronic UF of 3 based on a more thorough analysis of the available oral and
inhalation literature. The current RfD also includes a database UF of 3; the previous RfD
(posted in 1987) predated the institution of the database UF.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
5.2.1. Choice of Principal Study and Critical Effect—with Rationale and Justification
As noted in Section 4.6.2, the predominant targets of toxicity of carbon tetrachloride in
humans (based on case reports of acute, high-level exposure, or long-term occupational
exposure) and experimental animals following inhalation exposure are the liver and kidney.
Only one cross-sectional epidemiological study of hepatic function in workers (Tomenson et al.,
1995) provides data that can be considered for use in dose-response analysis.
Tomenson et al. (1995) conducted a cross-sectional study of hepatic function in 135
carbon tetrachloride-exposed workers in three chemical plants in northwest England and in a
control group of 276 unexposed workers. The exposure assessment was based on historical
personal monitoring data for various jobs at the three plants. Subjects were placed into one of
three exposure categories—low (<1 ppm), medium (1.1-3.9 ppm), or high (>4 ppm)—according
to their current jobs. Multivariate analysis, based on simultaneous consideration of ALT, AST,
ALP, and GGT as dependent variables, revealed a statistically significant (p < 0.05) difference
between exposed and unexposed workers. Univariate analyses (in which each dependent
variable was assessed separately) showed evidence of increased levels of ALP and GGT in the
medium- and high-exposure groups, with the differences between the medium-exposure group
and controls being statistically significant (p < 0.05). In an alternative analysis, the proportion of
191 DRAFT - DO NOT CITE OR QUOTE
-------
exposed workers exceeding the normal range (i.e., the 2.5 and 97.5% quantiles of the control
group) was significantly elevated for ALT (8%) and GGT (11%) but not for the other serum
chemistry variables. There was little difference between the low carbon tetrachloride-exposure
group (<1 ppm estimated exposure levels) and the control group on any of the liver enzymes.
Overall, this study suggests an effect of occupational carbon tetrachloride exposure on the liver
at exposures in the range of >l-3.9 ppm (6.3-24.5 mg/m3); this exposure range is considered a
LOAEL. The low exposure category in this study (<1 ppm or <6.3 mg/m3) is a NOAEL.
Because of study uncertainties described in Section 4.1.2.2, these values of the NOAEL and
LOAEL must be considered similarly uncertain.
A number of experimental animal studies that identified the liver and kidney as targets of
carbon tetrachloride toxicity were considered as the basis for RfC derivation. The most robust
study was the 2-year inhalation bioassay by JBRC (Nagano et al., 2007b; JBRC, 1998), which
used 50 animals/sex/group and examined an extensive set of endpoints of toxicity. The exposure
concentration of 25 ppm, 6 hours/day, 5 days/week in this study (corresponding to a continuous
exposure level of 4.5 ppm) produced evidence of liver and renal toxicity in both male and
female F344/DuCrj rats. The lowest exposure concentration in this study, 5 ppm (0.9 ppm,
adjusted to continuous exposure), was considered a NOAEL. As described in Section 4.2.2.2,
carbon tetrachloride-induced liver toxicity at >25 ppm was evidenced by serum chemistry
changes (including significant increases in ALT, AST, LDH, LAP, and GGT) and
histopathologic changes (fatty change, fibrosis, and cirrhosis) (see Table 4-3). In the kidney,
there was a dose-related increase in the severity of chronic nephropathy (progressive
glomerulonephrosis or CPN) (see Table 4-3) and a significant increase in BUN in rats exposed to
>25 ppm. Because of the high spontaneous rate of chronic nephropathy in F344 rats, the
incidence of chronic nephropathy was close to 100% in all dose groups, including the control,
and a dose-related increase in incidence could not be demonstrated. As discussed in Section
4.6.2, the severity (but not incidence) of proteinuria was increased in all carbon tetrachloride-
exposed rats. Because this observation was difficult to interpret and its biological significance
was uncertain, it was not used to define the NOAEL and LOAEL for kidney effects. For these
reasons, hepatic effects in this study were considered the more appropriate and sensitive measure
of carbon tetrachlori de-related toxicity.
Hepatic effects observed in the chronic rat inhalation study are consistent with the overall
carbon tetrachloride database. Epidemiological literature, in particular a cross-sectional study of
hepatic function in carbon tetrachloride-exposed workers (Tomenson et al., 1995) reported some
evidence of carbon tetrachlori de-associated effects on hepatic serum enzymes. Subchronic
studies in a number of experimental species (Adams et al., 1952; Prendergast et al., 1967;
f The exposure of 25 ppm for 6 hours/day, 5 days/week was adjusted to continuous exposure as follows: 25 ppm x 6
hours/24 hours x 5 days/7 days = 4.5 ppm
192 DRAFT - DO NOT CITE OR QUOTE
-------
Benson and Springer, 1999) identified aNOAEL for liver effects in the range of 0.9-4 ppm
(adjusted to continuous exposure). These subchronic studies used exposure durations of 12-26
weeks (versus 104 weeks in the JBRC bioassay) and experimental protocols that were less
rigorous than the JBRC bioassay. Therefore, these studies were considered less appropriate as
the basis for the RfC. In the chronic mouse study by JBRC (Nagano et al., 2007b; JBRC, 1998),
the NOAEL for liver toxicity was 0.9 ppm (adjusted to continuous exposure). This NOAEL is
the same as that for rats in the JBRC bioassay; however, the incidences of specific liver lesions
in the mouse were lower than those in the rat. Renal effects were observed in the JBRC chronic
mouse study (Nagano et al., 2007b; JBRC, 1998) and in subchronic animal studies, but generally
at concentrations higher than those that produced liver effects or occurred at a lower incidence
than liver effects.
At the lowest tested concentration of 5 ppm in the JBRC study (corresponding to a
continuous exposure level of 0.9 ppm), an increase in severity of proteinuria in male and female
rats was reported. As discussed in Section 4.6.2, the adversity of the proteinuria findings at this
exposure concentration is uncertain, and the evidence as a whole supports liver toxicity as the
endpoint of concern.
In addition to proteinuria, the only other effect reported at 5 ppm in the chronic rat study
was an increase in severity of eosinophilic change in the nasal cavity of the female rats (Nagano
et al., 2007b; JBRC, 1998). A similar effect in males was seen only at >25 ppm. This change,
by itself, is not considered to represent an adverse effect. Even in the high-exposure group that
experienced severe renal and hepatic effects, the nasal lesion was graded at only moderate
severity and was not accompanied by any other, more clearly adverse effects in the nasal cavity.
Nonvolatile and partly nonextractable radioactivity was detected in the nasal mucosa after
inhalation of radiolabeled carbon tetrachloride in mice (Bergman, 1983), suggesting that some
inhaled carbon tetrachloride is metabolized in the nasal cavity. However, there are no other
reports of lesions or irritant effects produced by carbon tetrachloride vapor in either humans or
animals.
By inhalation, benign pheochromocytomas, that could represent a potential noncancer
health hazard, were reported in mice in the JBRC inhalation bioassay (Nagano et al., 2007b;
JBRC, 1998). This benign tumor was observed only in mice (i.e., no increase in
pheochromocytomas was observed in rats in either NCI, 1977 or Nagano et al., 2007b) and thus
may represent a strain-specific finding. No data are available, however, to establish whether this
response is species specific. Developmental toxicity (reduced fetal body weight and crown-rump
length) was reported in a single inhalation study (Schwetz et al., 1974) at a concentration that
also produced toxicity in the dam. Because neither benign pheochromocytomas nor
developmental toxicity occurred at a concentration below those associated with liver toxicity and
because level of response was less robust than for endpoints of liver toxicity, these endpoints
were not considered appropriate as the basis for the RfC.
193 DRAFT - DO NOT CITE OR QUOTE
-------
The hepatic effects observed in the JBRC chronic inhalation bioassay (Nagano et al.,
2007b; JBRC, 1998) were considered the most appropriate basis for RfC derivation. Fatty
change in the liver of rats was selected as the specific endpoint for dose-response analysis
because this histopathologic lesion is indicative of cellular damage and appears to be a more
sensitive endpoint than other histopathologic changes (i.e., fibrosis and cirrhosis) that were also
present in 25-ppm rats in the JBRC study. General information on liver toxicants reveals that a
sufficient intracellular concentration of fatty acids can lead to injury of cell membranes, thereby
contributing to necrosis, inflammation, and progression to fibrosis and cirrhosis (Lieber, 2004;
Brunt and Tiniakos, 2002). Liver serum enzyme activities were also increased in male and
female rats and mice exposed to 25 ppm; however, serum enzyme levels were considered a less
consistent and reliable indicator of liver damage in this study than histopathologic changes. In
the mouse, the overall increase in liver enzyme levels was not monotonic (i.e., levels at 5 ppm
were lower than control levels). In the rat, liver enzyme level increases at 25 ppm were
considered modest (i.e., increases over control of only 40-90%). Further, reliable liver enzyme
data were not available for 125-ppm rats or mice because of the high mortality at this exposure
concentration (1-3 surviving animals/group at study termination) and because blood
biochemistry was not performed on animals that died before study termination. Therefore, liver
enzyme data were considered a less appropriate endpoint, compared with fatty change, for dose-
response analysis.
The occupational study by Tomensen et al. (1995) was also considered as the basis for
RfC derivation, using the estimated LOAEL of 5.5 ppm (35 mg/m3) as the POD. As discussed
more fully in Section 4.1.2.2, exposures for almost two-thirds of the workers were estimated, so
that there is some uncertainty in the value of the LOAEL. Although the data from the Tomensen
et al. (1995) study was not used to derive the RfC, the study was considered in an examination of
potential RfC values that would be obtained using alternative PODs (see Section 5.2.4).
5.2.2. Methods of Analysis—Including Models
Candidate RfCs for carbon tetrachloride were derived from data on fatty changes to the
liver in male and female rats; incidence data are summarized in Table 5-4.
194 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-4. Nonneoplastic lesions (fatty change) in F344 rats exposed to carbon
tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Species
Rat
Rat
Sex
Mb
Fc
Lesion type
fatty change
fatty change
Lesion severity"
l+and2+
1+, 2+, and 3+
Number of rats with lesions
Dose
0 ppm
4
6
5 ppm
7
7
25 ppm
39
49
125 ppm
49
46
a Severity rating: 1+, slight; 2+, moderate; 3+ marked.
b Number of male rats examined: 50/group; number of male rats surviving to study termination: 0 ppm, 22/50; 5
ppm, 29/50; 25 ppm, 19/50; 125 ppm, 3/50.
0 Number of female rats examined: 50/group; number of female rats surviving to study termination: 0 ppm, 39/50;
5 ppm, 43/50; 25 ppm, 39/50; 125 ppm, 1/50.
Sources: Nagano et al. (2007b); JBRC (1998).
The general procedure for analysis of the animal bioassay data for PBPK analysis is
depicted in Figure 5-4. Exposure levels studied in the 2-year rat bioassay (Nagano et al., 2007b;
JBRC, 1998) were converted to estimates of internal doses by application of a PBPK model.
BMD modeling methodology (U.S. EPA, 2000c, 1995) was used to analyze the relationship
between the estimated internal doses and response (i.e., fatty change of the liver). The resulting
BMDL values were converted to estimates of equivalent human exposure concentrations (HECs)
by applying a human PBPK model.
195
DRAFT - DO NOT CITE OR QUOTE
-------
Animal External Doses or
Exposure Concentrations
i
Animal PBPK
r
Animal Internal Doses
(MCA, MRAMKL)
1
Dose-respons
r
Benchmark Doses (BMDLs)
Expressed in Units of
Internal Dose (MCA,
MRAMKL)
i
Human PBPK
r
Human External Doses or
Exposure Concentrations
Corresponding to BMDLs
i
r
Point of Departure
MCA, time-averaged arterial blood concentration of carbon tetrachloride (umol/L);
MRAMKL, time-averaged rate of metabolism of carbon tetrachloride (umol/hour/kg
liver); PBPK, physiologically-based pharmacokinetics model
Figure 5-4. Process for analyzing animal bioassay data for deriving
noncancer toxicity values and cancer lURs and SFs using PBPK modeling.
5.2.2.1. PBPK Modeling for Internal Dose Metrics
Estimation of internal doses corresponding to the exposure concentrations studied in the
2-year rat bioassay (Nagano et al., 2007b; JBRC, 1998) was accomplished using a PBPK model
for the rat (Thrall et al., 2000; Benson and Springer, 1999; Paustenbach et al., 1988) (see
Sections 3.5 for description of the model). The review, selection, and application of the chosen
PBPK models was informed by an EPA report (U.S. EPA, 2006c) that addresses the application
and evaluation of PBPK models. The PBPK model was used to simulate internal dose metrics
corresponding to intermittent exposure (6 hours/day, 5 days/week) to concentrations of 5, 25, and
125 ppm, as studied in the 2-year bioassay (Nagano et al., 2007b; JBRC, 1998). Internal dose
metrics were selected that were considered to be most relevant to the toxicity endpoints of
interest (e.g., liver toxicity), based on consideration of evidence for MOA of carbon
196
DRAFT - DO NOT CITE OR QUOTE
-------
tetrachloride. Two dose metrics were selected based on available information on the
mechanisms of carbon tetrachloride liver toxicity: (1) time-averaged arterial blood concentration
of carbon tetrachloride (MCA, jimol/L); and (2) time-averaged rate of metabolism of carbon
tetrachloride (MRAMKL, |imol/hour/kg liver).
Liver metabolism rate was selected as the primary dose metric for liver effects, based on
evidence that metabolism of carbon tetrachloride via CYP2E1 to highly reactive free radical
metabolites plays a crucial role in its MOA in producing liver toxicity (described in Section 4.5).
The primary reactive metabolites that are thought to participate in carbon tetrachloride toxicity
are the trichloromethyl radical (-CCls) and the trichloromethyl peroxy radical (-O-OCCb),
although other reactive species may also contribute to a lesser extent (e.g., dichlorocarbene,
:CC\i). The role of these species in oxi dative injury is discussed further in Sections 4.5.2 and
4.5.3. The trichloromethyl radical is a product of carbon tetrachloride metabolism by CYP450.
It rapidly reacts with oxygen to produce the corresponding peroxy radical, which is more highly
reactive than the trichloromethyl radical (Russell et al., 1990; Slater, 1981; Packer et al., 1978).
The trichloromethyl peroxy radical is thought to be the dominant intermediate in the initiation of
lipid peroxidation associated with carbon tetrachloride hepatotoxicity (Slater, 1981). Elimination
of trichloromethyl radical, by reaction with oxygen to form the trichloromethyl peroxy radical
and downstream reaction products with amino acids, protein, and lipid, is extremely rapid (e.g.,
near the limits of diffusion) relative to the production of the trichloromethyl radical by CYP450.
(See Section 3.3 for a discussion of the rates of conversion of carbon tetrachloride to the
trichloromethyl radical and the trichloromethyl radical to the trichloromethyl peroxy radical.)
The large difference in rates of production and elimination of the trichloromethyl radical (i.e.,
1 &
10 -10 fold difference) has several implications. (1) Limiting factors in the elimination of the
trichloromethyl radical are likely to be reactant concentrations at the site of production of the
trichloromethyl radical (e.g., 62,) and/or factors that limit diffusion of the trichloromethyl radical
(e.g., diffusion coefficient in cytosol). (2) Similarly, limiting factors in the elimination of the
trichloromethyl peroxy radical are likely to be reactant concentrations (e.g., intracellular amino
acids, lipid, protein, non-protein sulfhydryls) and/or diffusion, all of which are expected to be
similar in rodents and humans. (3) These elimination reactions are likely to occur within a
relatively short diffusion distance from the site of production of the trichloromethyl radical
(Slater, 1981), resulting in most (if not all) of the production of reaction products of the
trichloromethyl and trichloromethyl peroxy radicals occurring within the tissue where
trichloromethyl radical is produced (e.g., within liver and other tissues having appreciable
CYP450 activity). Suicide inhibition and destruction of CYP450 by carbon tetrachloride is
consistent with the reactivity of these species on a very local histological scale. It follows from
the above considerations that equal rates of hepatic metabolism of carbon tetrachloride by
197 DRAFT - DO NOT CITE OR QUOTE
-------
CYP450 in rodents and humans would be expected to yield equal rates of elimination of
trichloromethyl and trichloromethyl peroxy radicals.
Finally, carbon tetrachloride metabolism is known to lead to lipid peroxidation. Once
initiated, lipid peroxidation is a self-perpetuating process that continues as a chain reaction
(MacNee and Rahman, 2004). As such, the generation of lipid peroxides is not expected to be
enzymatically driven. Accordingly, the rate of hepatic metabolism of carbon tetrachloride
should be a reasonable internal dose surrogate for these radical species in liver.
Uncertainty regarding the accuracy of available PBPK models to simulate carbon
tetrachloride is recognized. These uncertainties include the following: (1) estimates of the Km
and Vmax for the CYP2E1 pathway in the rat and human and potential dose-dependence of these
parameters (e.g., suicide inhibition and induction); (2) relative contributions of extra-hepatic
tissues to carbon tetrachloride metabolism (all of which is assigned to the liver in PBPK models
used in this analysis); and (3) magnitude of direct contribution of carbon tetrachloride (i.e.,
parent compound) to liver toxicity. Given the above uncertainties, arterial blood concentration
of carbon tetrachloride was also included in the analysis as a more proximal dose metric to liver
metabolism.
The two dose metrics, MCA and MRAMKL, were simulated in the rat PBPK model as
time-averaged values, with the averaging time being the chronic exposure period (e.g., 2 years).
The time-averaged dose metrics were calculated as follows (Equations 5-1 and 5-2):
t Eq. (5-1)
MRAMKL = AUClUMKL = AMKL
t t Eq. (5-2)
where:
MCA = time-averaged arterial blood concentration of carbon tetrachloride (|imol/L)
AUCcA = area under the arterial concentration (CA) - time profile (|imol-hour/L)
MRAMKL = time-averaged rate of metabolism of carbon tetrachloride (|imol/hour/kg
liver weight)
AUCRAMKL = area under the rate of metabolism (RAMKL) - time profile (|imol/kg liver
weight)
AMKL = cumulative amount of carbon tetrachloride metabolized (|imol/kg liver)
t = time (hours)
Internal dose metrics corresponding to the exposure concentrations studied in the 2-year
rat inhalation bioassay (Nagano et al., 2007b; JBRC, 1998) are presented in Table 5-5. Two
values for Vmaxc (maximum rate of hepatic metabolism of carbon tetrachloride) have been
reported for the rat; both estimates are represented in the data presented in Table 5-5. Gargas et
al. (1986) derived a value for Vmaxc of 0.4 mg/hour/kg BW°'70, based on the results of gas uptake
198 DRAFT - DO NOT CITE OR QUOTE
-------
studies in rats. Paustenbach et al. (1988) derived a value of 0.65 mg/hour/kg BW°'70, based on a
reanalysis of data for a subset of the rats used in the Gargas et al. (1986) study. Increasing Vmaxc
from 0.4 to 0.65 mg/hour/kg BW°'70 resulted in lower values for the MCA dose metric and higher
values for the MRAMKL dose metric (Table 5-5). Comparisons of internal doses predicted for
various exposure concentrations are shown in Figure 5-5. The effect of varying Vmaxc on
MRAMKL becomes more pronounced as exposure concentration increases. This pattern reflects
the increasing influence of Vmax on rate of metabolism at higher exposures concentrations that
result in liver carbon tetrachloride concentrations that exceed the Km.
Table 5-5. Comparisons of internal dose metrics predicted from PBPK rat models"
Exposure
(ppm)
5
25
125
MCA
(umol/L)
VmaxC=0.40
0.128
0.708
3.892
Vmax=0.65
0.116
0.653
3.775
MRAMKL
(umol/hr/kg liver)
VmaxC=0.40
3.813
12.092
24.320
Vmax=0.65
4.991
17.626
36.266
a Values are for 0.452 kg rat.
Sources: Thrall et al. (2000); Paustenbach et al. (1988); Gargas et al. (1986).
199
DRAFT - DO NOT CITE OR QUOTE
-------
VmaxC=0.40
- - - •VmaxC=0.65
0
50 100 150
AIR(ppm)
200
250
VmaxC=0.40
- - -VmaxC=0.65
0
50 100 150
AIR (ppm)
200
250
Dose metrics shown are time-averaged arterial concentration of carbon tetrachloride
(MCA, umol/L, upper panel), and time-averaged rate of metabolism of carbon
tetrachloride (MRAMKL, umol/hour/kg liver, lower panel). The dose metrics are plotted
against exposure concentration (6 hours/day, 5 days/week, 2 years) for a 0.452 kg rat.
Sources: Thrall et al. (2000); Paustenbach et al. (1988).
Figure 5-5. Internal dose metrics predicted by the PBPK rat model.
200
DRAFT - DO NOT CITE OR QUOTE
-------
5.2.2.2. BMD Modeling
BMD modeling methodology (U.S. EPA, 2000c, 1995) was used to analyze data on
estimated internal doses (i.e., MCA, MRAMKL) and incidence data (i.e., fatty changes of the
liver) from the 2-year rat bioassay (Nagano et al., 2007b; JBRC, 1998). All of the models for
dichotomous data in U.S. EPA's BMDS (version 1.4.1) (U.S. EPA, 2007b) were fit to the
incidence data for rats.
Internal doses associated with a BMR of 10% extra risk were calculated. A BMR of 10%
extra risk of fatty changes in the liver was selected because the POD associated with this BMR
fell near the low end of the range of experimental data points (see plots in Appendix D). As
noted in U.S. EPA (2000c), "[t]he major aim of benchmark dose modeling is to model the dose-
response data for an adverse effect in the observable range (i.e., across the range of doses for
which toxicity studies have reasonable power to detect effects) and then select a 'benchmark
dose' at the low end of the observable range to use as a 'point of departure'."
In the male rat, the best fit of the data was provided by the log-logistic model using MCA
9
as the dose metric and the logistic model using MRAMKL as the dose metric (based on % p>0.1
and lowest AIC value). For female rats, no models provided an adequate fit to the data when all
r\
dose groups were included, as assessed by the % goodness-of-fit test (i.e., application of the
r\
models in BMDS yielded % ^-values in all cases <0.1). After dropping the highest dose, the
r\
multistage model provided the best fit of the female incidence data (based on % p>0.l and
lowest AIC value) using either dose metric. Summaries of the resulting BMDio and BMDLio
values for male and female rats are shown in Tables 5-6 and 5-7 (columns 3 and 4). Details of
the BMD modeling are provided in Appendix D.
201 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-6. HEC values corresponding to BMDL values for incidence data for fatty
changes of the liver in male F344 rats
BMR
(l)b
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Metric
(2)
MCA
MRAMKL
BMD modeling"
VMAXCR=0.4
(3)
BMD10:0.14
(34.26)
BMDL10: 0.079
(19.68)
BMD10: 3.26
(26.27)
BMDL10: 2.59
(20.38)
VMAXCR=0.65
(4)
BMD10:0.12
(32.36)
BMDL10: 0.071
(19.42)
BMD10: 4.60
(28.72)
BMDL10: 3.65
(22.42)
VMAXCH
(5)
0.40
0.65
1.49
1.70
0.40
0.65
1.49
1.70
HEC
VMAXCR=0.4
(6)
5.396
5.712
6.338
6.436
23.793
17.160
11.826
11.343
VMAXCR=0.65
(7)
4.830
5.113
5.671
5.760
35.243
24.773
16.794
16.093
Rats were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, 25, 125 ppm.
BMR, benchmark response; HEC, human equivalent concentration, mg/m3; MCA, time-averaged arterial blood
concentration, umol/L; MRAMKL, time-averaged rate of metabolism per kg liver, umol/hr/kg liver; VMAXC,
maximum rate of metabolism in humans (H) or rat (R), mg/hr/kg B W°70
aMCA, log-logistic model provided the best fit; MRAMKL, logistic model provided the best fit. Values in
parentheses are animal exposure concentrations (mg/m3) corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
202
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-7. HEC values corresponding to BMDL values for incidence data for fatty
changes of the liver in female F344 rats (high dose dropped)
BMR
(l)b
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Metric
(2)
MCA
MRAMKL
BMD modeling"
VMAXCR=0.4
(3)
BMD10:0.12
(29.53)
BMDL10:
0.085(21.13)
BMD10: 3.77
(31.97)
BMDL10: 2.82
(22.37)
VMAXCR=0.65
(4)
BMD10:0.11
(29.75)
BMDL10: 0.078
(21.29)
BMD10: 5.42
(34.35)
BMDL10: 3.75
(23.08)
VMAXCH
(5)
0.40
0.65
1.49
1.70
0.40
0.65
1.49
1.70
HEC
VMAXCR=0.4
(6)
5.815
6.156
6.831
6.937
26.259
18.838
12.935
12.405
VMAXCR=0.65
(7)
5.298
5.608
6.222
6.319
36.337
25.478
17.246
16.524
Rats were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, 25 ppm (125 ppm dose dropped).
BMR, benchmark response; HEC, human equivalent concentration, mg/m3; MCA, time-averaged arterial blood
concentration, umol/L; MRAMKL, time-averaged rate of metabolism per kg liver, umol/hr/kg liver; VMAXC,
maximum rate of metabolism in humans (H) or rat (R), mg/hr/kg B W°70
a MCA, multistage (2); MRAMKL, multistage (3). Values in parentheses are animal exposure concentrations
(mg/m3) corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
5.2.2.3. PBPK Modeling of Human Equivalent Exposure Concentrations
Interspecies extrapolation (i.e., rat-to-human) of carbon tetrachloride inhalation
dosimetry was accomplished using a human PBPK model described in Paustenbach et al. (1988),
Thrall et al. (2000), and Benson and Springer (1999). The human PBPK model was used to
estimate continuous human equivalent concentrations (HECs, in mg/m3) that would result in
values for the internal dose metrics, MCA or MRAMKL, equal to the BMDLio values for fatty
changes of the liver.
The approach used to derive the HECs for each dose metric was as follows:
(1) The human PBPK model was used to calculate internal doses corresponding to a series of
exposure concentrations (EC, continuous exposure, mg/m3). For the dose metric MCA, the
human PBPK model was run at intervals over the range from 0.1 to 100 ppm (0.63-629 mg/m3);
for MRAMKL, the human PBPK model was run at intervals from 1 to 300 ppm (6.3-
1,887 mg/m3).
(2) For each internal dose, conversion factors were calculated as the following corresponding
ratios:
• EC/MCA (to relate a continuous chronic human inhalation exposure in mg/m3 [EC] to
an internal dose using MCA as the dose metric);
203
DRAFT - DO NOT CITE OR QUOTE
-------
• EC/MRAMKL (to relate a continuous chronic human inhalation exposure in mg/m3
[EC] to an internal dose using MRAMKL as the dose metric); and
(3) Conversion factors were calculated for each of four assumed values of Vmaxc in the human
PBPK model: 0.40, 0.65, 1.49, or 1.70 mg/hour/kg BW0'70. These conversion factors are
provided in Appendix C. Trend equations were also developed to permit the calculation of EC
for any value of MCA or MRAMKL (see Appendix C).
Estimates of the dose metrics, MCA and MRAMKL, were sensitive to the value assigned
to the Vmaxc parameter (see Figure 5-5). Several values for Vmaxc in animals and humans have
been reported (Thrall et al., 2000; Benson and Springer, 1999; Paustenbach et al., 1988; Gargas
et al., 1986); therefore, evaluation of uncertainty in this parameter was introduced into the
analysis by assuming various reported values for Vmaxc in the estimation of HECs. Thrall et al.
(2000) and Benson and Springer (1999) derived a value of 1.49 mg/hour/kg BW0'70 for humans,
based on an analysis of data on in vivo (gas uptake) studies in rodents and in vitro studies of
metabolism of carbon tetrachloride in rodent and human liver samples. Thrall et al. (2000) also
derived a value of 1.7 mg/hour/kg BW0'70 for hamsters, based on the results of closed chamber
gas uptake studies. The value of 1.49 mg/hour/kg BW0'70 for humans (Thrall et al., 2000; Benson
and Springer, 1999), the value of 1.70 mg/hour/kg BW0'70 for the hamster (Thrall et al., 2000),
and the two values estimated for the rat (0.4, 0.65 mg/hour/kg BW0'70; Paustenbach et al., 1988;
Gargas et al., 1986) were used in the estimation of HECs. Estimated values for HECs
corresponding to BMDLio values for fatty changes of the liver as reported in the 2-year rat
inhalation bioassay (Nagano et al., 2007b; JBRC, 1998) for alternative values of Vmaxc in the rat
and human are presented in Tables 5-6 and 5-7 (columns 6 and 7).
A human Vmaxc estimated from in vitro human data can reasonably be presumed to be
more relevant than a human Vmaxc based entirely on rodent data. In addition, because the MOA
for carbon tetrachloride-induced hepatotoxicity involves metabolism to reactive metabolites in
the liver, HECs based on the MRAMKL dose metric is the most proximate to the critical effect.
Therefore, the human Vmaxc estimated from in vitro human data (1.49 mg/hour/kg BW0'70) and
the dose metric MRAMKL are considered to yield the most appropriate estimate of the HEC.
No information is available to establish a rat Vmaxc of either 0.4 or 0.65 mg/hour/kg BW0'70 as
the more scientifically defensible value for this parameter. Therefore, HECs derived using these
two rat Vmnxc values were averaged to derive the POD for the carbon tetrachloride RfC.
Accordingly, the POD based on male rat data was calculated as (11.826 + 16.794) - 2 = 14.3
mg/m3. In the female rat, the HEC was similarly calculated as (12.935 + 17.246) -2 = 15.1
mg/m3. The HEC based on data for the male rat (14.3 mg/m3) is the lower of the two values, and
was selected as the POD for RfC derivation.
204 DRAFT - DO NOT CITE OR QUOTE
-------
5.2.3. RfC Derivation—Including Application of Uncertainty Factors
An RfC of 0.1 mg/m3 for carbon tetrachloride is derived by applying a composite UF of
100 to the HEC of 14.3 mg/m3, as follows:
RfC = HEC/UF (5-3)
14.3 mg/m3/100
= 0.143 mg/m3 or 0.1 mg/m3
The composite UF of 100 includes a factor of 10 to protect susceptible individuals, a
factor of 3 (10°5) to adjust for pharmacodynamic differences in the extrapolation from rats to
humans, and a factor of 3 (10°5) to account for an incomplete database lacking an adequate
multigeneration study of reproductive function.
• A default 10-fold UF for intraspecies differences (UFn) was selected to account for
variability in susceptibility among members of the human population in the absence of
quantitative information on the variability of human response to carbon tetrachloride.
Factors that could contribute to a range of human response to carbon tetrachloride were
discussed in Section 4.8. Intrahuman variability in CYP450 levels that are responsible
for metabolism of carbon tetrachloride to reactive metabolites has been documented (see
Section 4.8). This variation in CYP450, which is likely influenced by age-related
differences or other factors (e.g., exposure to other chemicals that induce or inhibit
microsomal enzymes), could alter susceptibility to carbon tetrachloride toxicity.
Individual variability in nutritional status, alcohol consumption, or the presence of
underlying disease could also alter metabolism of carbon tetrachloride or antioxidant
protection systems. To account for these uncertainties, a factor of 10 was applied for
individual variability.
• A UF of 3 (10°5) was selected for interspecies extrapolation (UFA) to account for
potential pharmacodynamic differences between rats and humans. As pharmacokinetic
and pharmacodynamic components are assumed to contribute equally to the uncertainty
in interspecies extrapolation and the product of the two components is assumed by
default to be 10, a numeric value of 10°5 (3.2, expressed as the numeral 3 after rounding)
is assigned to each component. Cellular antioxidant systems function to quench the lipid
peroxidation reaction and prevent damage to cellular membranes. A pharmacokinetic
model was used to adjust for pharmacokinetic differences across species; therefore, an
additional UF was not included for pharmacokinetic differences between species. In the
absence of data to quantify specific interspecies differences for cellular protective
mechanisms, a UF of 3 is applied to account for species differences in
205 DRAFT - DO NOT CITE OR QUOTE
-------
pharmacodynamics.
A UF to account for extrapolation from a LOAEL to a NOAEL (UFi) 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 of a 10% change in fatty
changes of the liver was selected under an assumption that it represents a minimal
biologically significant change.
A UF to extrapolate from a subchronic to a chronic exposure duration (UFS) was not
necessary because the RfC was derived from a study using a chronic exposure protocol.
• A UF to account for database deficiencies (UFo) of 3 (10°5) was selected. The
inhalation database for this chemical includes extensive testing for subchronic
toxicity in animals, 2-year chronic inhalation bioassays in rats and mice, one study of
immunotoxic potential, and human epidemiology data. Testing for developmental
toxicity was limited to one inhalation study in the rat that found effects only at high,
maternally toxic exposure concentrations. This study did not use an exposure
concentration low enough to identify a NOAEL for either maternal or fetal toxicity.
Nevertheless, the developmental effects at the LOAEL were modest, and were limited
to decreased fetal body weight (7%) and decreased crown-rump length (3.5%). The
LOAEL for developmental effects (in the presence of maternal toxicity) in this study
(334 ppm) was 66-fold higher than the NOAEL from the principal study (5 ppm).
Developmental toxicity has been tested more extensively by the oral route, although
all adequate studies were conducted in the same species (rat); the oral NOAEL for
developmental toxicity exceeded both the oral NOAEL and LOAEL for liver toxicity.
As noted in Section 4.8.1 (Possible Childhood Susceptibility), microsomal enzymes
that are responsible for metabolizing carbon tetrachloride, particularly CYP2E1, are
lower in the developing organism than the adult, and do not achieve adult levels in
humans until sometime between 1 and 10 years. Thus, lifestage information on
microsomal enzyme activity suggests that the developing organism would be no more
susceptible to free radical-induced liver injury from carbon tetrachloride than adults.
The available information suggests that further developmental toxicity testing would
not likely result in a POD lower than that based on liver toxicity. The database lacks
an adequate multigeneration study of reproductive function by any route of exposure;
therefore, a threefold UF was applied.
206 DRAFT - DO NOT CITE OR QUOTE
-------
5.2.4. RfC Comparison Information
PODs and potential inhalation RfCs based on selected studies included in Table 4-14 are
arrayed in Figures 5-6 to 5-8, and provide perspective on the RfC supported by Nagano et al.
(2007b; JBRC, 1998). These figures should be interpreted with caution because the PODs across
studies are not necessarily comparable, nor is the confidence in the data sets from which the
PODs were derived the same. PODs in these figures may be based on a NOAEL, LOAEL, or
BMDL (in the case of the principal study), and the nature, severity, and incidence of effects
occurring at a LOAEL are likely to vary. In addition, PBPK modeling for animal to human
extrapolation was applied to data from the principal study, whereas the default approach (i.e.,
application of an UF of 10) was used for other animal data sets. To some extent, the confidence
associated with the resulting potential RfC is reflected in the magnitude of the total UF applied to
the POD (i.e., the size of the bar); however, the text of Sections 5.2.1 and 5.2.2 should be
consulted for a more complete understanding of the issues associated with each data set and the
rationale for the selection of the critical effect and principal study used to derive the RfC.
As discussed in Section 4.6.2, the liver and kidney are the predominant targets of carbon
tetrachloride toxicity in laboratory animals in subchronic and chronic inhalation studies (Nagano
et al., 2007a,b; Benson and Springer, 1999; JBRC, 1998; Prendergast et al., 1967; Adams et al.,
1952; Smyth et al., 1936) and in humans based on case reports and studies in exposed workers.
Benign pheochromocytomas from the adrenal gland medulla, that could represent a potential
noncancer health hazard, were observed by inhalation only in mice in the JBRC chronic bioassay
(Nagano et al., 2007b; JBRC, 1998). A single study of developmental toxicity (Schwetz et al.,
1974) found significant reductions in fetal body weight and crown-rump length in rats at a
carbon tetrachloride concentration that also produced hepatotoxicity and reduced growth in the
dams. This set of literature was evaluated in selecting the most appropriate study and endpoint
to use as the basis for the RfC, with particular consideration given to the overall strength of the
evidence for a given measure of toxicity, consistency of the finding across studies, relevance to
humans, sensitivity of the endpoint, and rigor of a given study.
Figure 5-6 provides a graphical display of dose-response information from one
occupational cross sectional study and five experimental animal data sets that reported liver
toxicity; all animal studies identified a NOAEL for liver toxicity of approximately 6 mg/m3 or
0.9 ppm (adjusted to continuous exposure) and the study of exposed workers (Tomensen et al.,
1995) identified a LOAEL of approximately to 12.5 mg/m3 or 2 ppm (adjusted to continuous
exposure).8 As discussed in Section 5.2.1, the JBRC study in the rat (Nagano et al., 2007b;
JBRC, 1998), which identified a NOAEL for liver toxicity of 5.7 mg/m3 or 0.9 ppm (adjusted to
continuous exposure), was a sensitive and well-conducted study of carbon tetrachloride toxicity,
8 The workplace exposure concentration of 35 mg/m3 was adjusted to continous exposure by multiplying by
(10 m3/day ^ 20 m3/day) x (5 days/week ^ 7 days/week), where 10 mVday is an estimate of an 8-hour tinier
average occupational respiratory rate and 20 m3/day an estimate of an average daily respiratory rate.
207 DRAFT - DO NOT CITE OR QUOTE
-------
and was selected as the basis for the RfC. Dose-response analysis of the data from this study,
which included BMD and PBPK modeling, yielded a POD of 14.3 mg/m3. Possible RfCs that
might be derived from other studies demonstrating liver toxicity are also presented in Figure 5-6.
Although the RfC based on the JBRC rat data is not the lowest among candidate studies, it is
considered to be the most scientifically rigorous and associated with a lower degree of
uncertainty than other experimental animal studies. The POD is based on a study of chronic
toxicity data (rather than the subchronic exposures used in Benson and Springer, 1999, and
Adams et al., 1952), the application of BMD methods, which has an inherent advantage over the
use of a NO AEL or LOAEL by making greater use of all the data from the study, and the use of
PBPK modeling for interspecies extrapolation. As shown in Figure 5-6, the use of PBPK
modeling also resulted in the application of a smaller composite UF to the POD, (i.e., smaller
degree of uncertainty than with other data sets to which the default UF of 10 for interspecies
extrapolation was applied). The RfC derived using data from the JBRC rat study is consistent
with the potential RfC derived from the Tomensen et al. (1995) study. Tomensen et al. (1995)
reported a statistically significant increase in two of four serum enzymes indicative of liver
function in workers exposed to approximately 35 mg/m3 (5.5 ppm) carbon tetrachloride (adjusted
to continuous exposure: 12.5 mg/m3). Using 12.5 mg/m3 as the POD and applying a composite
UF of 300 (10 for variation in sensitivity in the human population, 10 for extrapolation from a
LOAEL to a NO AEL, and 3 for database deficiencies), the potential RfC is estimated to be 0.04
mg/m3. Because the Tomensen et al. (1995) noted that "there was no evidence of effects of clear
clinical significance on the liver function of workers exposed to carbon tetrachloride at the levels
indicated," it could be argued that a UF for LOAEL to NO AEL extrapolation of 3 (rather than a
full UF of 10) might be appropriate. In this case, the potential RfC estimated from Tomensen et
al. (1995) serum enzyme data would be 0.1 mg/m3. Thus, the potential RfCs of 0.04-0.1 mg/m3
estimated from Tomensen et al. (1995) are consistent with the RfC of 0.1 mg/m3 derived from
the JBRC rat bioassay (Nagano et al., 2007b; JBRC, 1998), and supports this RfC for liver
effects derived from animal data.
The most sensitive study of kidney toxicity was the JBRC bioassay in the rat and mouse
(Figure 5-7) (Nagano et al., 2007b; JBRC, 1998). As discussed in Section 5.2.1., kidney effects
occurred at a concentration similar to liver effects, but at lower incidence.
Figure 5-8 displays PODs for all major targets of carbon tetrachloride toxicity by the
inhalation route, including liver, kidney, adrenal gland, and developmental toxicity. For the
reasons discussed in Section 5.2.1, liver effects in the rat observed in the JBRC study are
considered the most appropriate basis for the carbon tetrachloride RfC. The POD based on liver
effects is similar to the PODs associated with kidney effects and effects on the adrenal gland
(benign pheochromocytomas); however, a smaller composite UF was applied to the POD for
liver effects because PBPK modeling was used for interspecies extrapolation. The greatest
degree of uncertainty is associated with the potential RfC for developmental toxicity. While this
208 DRAFT - DO NOT CITE OR QUOTE
-------
relatively large UF drives down the value of the potential RfC for developmental toxicity, the
RfC based on liver effects should be adequately protective.
209 DRAFT - DO NOT CITE OR QUOTE
-------
100
10
O)
o
§
0.001
The RfC in the
dashed circle was
selected as the
final RfC for carbon
tetrachloride.
0.01
Tomensen et al, 1995; Benson & Springer
occupational epid 1999; 12-w k mouse
study; liver enzymes; study; liver enzymes,
LOAEL(1) histopath; NOAEL
JBRC1998;2-yr
mouse study; liver
weight, enzyme
activity, histopath;
NOAEL
Adams etal, 1952;
6-month rat study;
liver weight, fatty
liver; NOAEL
Adams etal, 1952;
6-month guinea pig
study; liver weight,
fatty liver; NOAEL
Nagano etal., 2007b;
2-yr rat study; liver
enzymes, histopath,
including fatty liver;
BMDL10[HEC]
• POD
]J| Animal-to-human
^| Human variation
2 LOAEL to NOAEL
^ Subchr to Chronic
• Database deficiencies
ORfC
Note: The RfC based on
Nagano et al. (2007b)
used BMD methods and
PBPK modeling, and
therefore is not directly
comparable to the other
RfCs for liver endpoints.
(1) Magnitude of effect at the LOAEL: liver enzyme levels (f <23%)
Figure 5-6. PODs (mg/m3) with corresponding derived potential inhalation reference values that would result if liver
toxicity was used as the critical effect.
210
DRAFT - DO NOT CITE OR QUOTE
-------
10.00
1.00
O)
o
0)
u
0.10
0.01
• POD
nT|Animal-to-human
^| Human variation
^LOAELtoNOAEL
^| Subchr to Chronic
| Database deficiencies
0 RfC
Nagano et al., 2007b; 2-yr rat study; serum chemistry,
histopathology; NOAEL
Nagano et al., 2007b; 2-yr mouse study; serum chemisty,
histopathology; NOAEL
Figure 5-7. PODs (mg/m ) with corresponding derived potential inhalation reference values that would result if
kidney toxicity was used as the critical effect.
211
DRAFT - DO NOT CITE OR QUOTE
-------
1000
100
O)
o
10
0.1
0.01
• POD
III! Animal-to-human
QJ Human variation
0LOAELtoNOAEL
I I Subchr to Chronic
| Database deficiencies
O RfC
1
\ /
Kidney
Adrenal gland (benign
pheochromocytoma)
Liver
The RfC in the
dashed circle was
selected as the final
RfC for carbon
tetrachloride.
Developmental
POD based on:
Kidney: NOAEL for serum
chemistry and histopathology
changes in the rat and mouse;
2-year inhalation study
(Nagano etal., 2007b)
Benign pheochromocvtoma:
NOAEL in the mouse; 2-year
inhalation study (JBRC, 1998)
Liver: BMDL[HEq for fatty liver
in the rat; 2-year inhalation
study (JBRC, 1998)
Developmental: LOAEL for
decreased fetal body weight
and crown-rump length;
exposure on GD 6-15 (Schwetz
etal., 1974)
Note: The RfC based on liver
toxicity used BMD methods
and PBPK modeling, and
therefore is not directly
comparable to the other organ-
specific RfCs.
Figure 5-8. PODs (mg/m ) with corresponding derived potential inhalation reference values that would result if
alternative endpoints were used as the critical effect.
212
DRAFT - DO NOT CITE OR QUOTE
-------
5.2.5. Previous RfC Assessment
An inhalation assessment for carbon tetrachloride was not previously available on IRIS.
5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE AND INHALATION
REFERENCE CONCENTRATION
Risk assessments need to describe associated uncertainty. The following discussion
identifies uncertainties associated with the RfD and RfC for carbon tetrachloride. As presented
earlier in this section (see Sections 5.1.2 and 5.1.3 for the RfD and Sections 5.2.2 and 5.2.3 for
the RfC), the UF approach (U.S. EPA, 2002, 1994b) was used to derive the RfD and RfC for
carbon tetrachloride. Using this approach, the POD was divided by a set of factors to account for
uncertainties associated with a number of steps in the analysis, including extrapolation from
responses observed in animal bioassays to humans and from data from subchronic exposure to
chronic exposure, a diverse population of varying susceptibilities, and to account for database
deficiencies. Because information specific to carbon tetrachloride was unavailable to fully
inform many of these extrapolations, default factors were generally applied.
A broad range of animal toxicity data and more limited range of human study data are
available to assess carbon tetrachloride hazard (see Section 4). Human studies include case
reports of acute human exposure (both oral and inhalation) and occupational epidemiology
studies. The animal toxicology literature includes subchronic and chronic animal studies by the
oral and inhalation routes, developmental toxicity studies by the oral and inhalation routes,
studies of immunotoxic potential, extensive literature on genotoxicity, and numerous mechanistic
toxicity studies. In addition, carbon tetrachloride has been used in hundreds of studies as a
classic inducer of liver toxicity. Nevertheless, gaps in the carbon tetrachloride database have
been identified; uncertainties associated with these data deficiencies are discussed more fully
below.
Selection of the critical effect for reference value determination. Liver toxicity was
selected as the critical effect for both the RfD and RfC (specifically, elevated liver enzymes
[Bruckner et al., 1986] in the case of the RfD and fatty change of the liver [Nagano et al., 2007b;
JBRC, 1998] in the case of the RfC). The liver has been established as a sensitive target of
toxicity across animal species and routes of exposure. Case reports of human poisonings identify
the liver as a target organ of acute carbon tetrachloride exposure, and an occupational
epidemiology study of workers exposed to carbon tetrachloride (Tomenson et al., 1995) provides
evidence of impaired liver function in humans following prolonged exposure. Thus, there is
little uncertainty related to the relevance of the critical effect to human health assessment.
Kidney toxicity associated with carbon tetrachloride inhalation exposure has been seen
less consistently in experimental animal studies. Nagano et al. (2007b; also reported as JBRC,
1998) reported an increase in the severity of proteinuria in rats at the lowest concentration tested
213 DRAFT - DO NOT CITE OR QUOTE
-------
in a 2-year bioassay. This kidney finding occurred at an exposure level fivefold lower than the
concentration associated with fatty changes of the liver; however, given the uncertainties in this
endpoint discussed in Section 4.6.2, proteinuria was not used as the critical effect for the RfC.
Use of proteinuria data as the basis for the RfC would have yielded a lower POD than liver data.
Dose-response modeling. BMD modeling was used to estimate the POD for both the
RfD and RfC. BMD modeling has advantages over a POD based on a NOAEL or LOAEL
because, in part, the latter are a reflection of the particular exposure concentration or dose at
which a study was conducted. A NOAEL or LOAEL lacks characterization of the dose-response
curve, and for this reason, is less informative than a POD obtained from BMD modeling. The
selected models—the power model in the case of the RfD and the logistic model in the case of
the RfC—provided the best mathematical fits to the experimental data sets (as determined by the
lowest AIC), but do not necessarily have greater biological support over the various models
included in BMDS. Other models in BMDS yield estimates of the POD both higher and lower
than the PODs used to derive the RfD and RfC.
Animal to human extrapolation. Extrapolating dose-response data from animals to
humans is another source of uncertainty. The effect and the magnitude of the effect at the POD
in rodents are extrapolated to human response. Uncertainty in interspecies extrapolation can be
separated into two general areas—toxicokinetic and toxicodynamic. A UF of 3 was used to
account for toxicodynamic differences between animals and humans. A PBPK model was
available for the inhalation pathway and was used in deriving the RfC to address the
toxicokinetic portion of interspecies extrapolation. Availability of an inhalation PBPK model
generally reduces the toxicokinetic component of uncertainty associated with animal to human
extrapolation by moving away from default assumptions about kinetic differences between
animals and humans. Any PBPK model, however, has its own associated uncertainties related to
model structure and parameters (e.g., inclusion of appropriate parameters and interrelationships
between parameters), and to values assigned to parameters. A sensitivity analysis was performed
for the human PBPK model (see Section C.4 in Appendix C). The maximum rate of metabolism
(Vmaxc) was a sensitive parameter for both dose metrics utilized. Other sensitive chemical-
specific parameters included the blood:air partition coefficient and Michaelis-Menten coefficient
for metabolism (Km) using MCA as the internal dose metric, and liverblood, slowly-
perfused:blood, and readily-perfused:blood partition coefficients for MRAMKL as the dose
metric.
In general, relatively high confidence is assumed for values of physiological parameters
(e.g., tissue volumes and blood flows), since these are amenable to direct (and corroborated)
observation in animals and humans. Similarly, relatively high confidence is also assumed for
values of partition coefficients that have been directly measured in rodent and human tissues,
214 DRAFT - DO NOT CITE OR QUOTE
-------
especially if independent estimates yield values within expected intra- and inter-laboratory
variability. Although different values for the blood:air partition coefficient were used in the
human (2.64; Paustenbach et al.,1988) and rat model (4.52; Gargas et al., 1986), these
differences are within a range of expected variability for these parameter values, within and
across species. Studies in which identical methods have been applied to estimation of blood:air
partition coefficients have obtained variation in values across species. For example, estimates of
partition coefficients for carbon tetrachloride in humans (H) and rats (R) from a single laboratory
were 2.73 ± 0.23 (SE) and 4.52 ± 0.35, respectively (H/R = 0.60; Gargas et al., 1989). The
above value for humans is similar to the value reported by Paustenbach et al. (1988), based on
similar methods, 2.64 ± 0.07 (SE), and that was used in PBPK modeling in the current
assessment. Estimates for 59 chemicals in human and rat blood yielded H/R ratios that ranged
from 0.33 to 1.08 (mean = 0.64; Gargas et al., 1989). Estimates for seven chemicals in human
and mouse (M) blood yielded H/M ratios of 0.32-1.54 (mean = 0.68; Gargas et al., 1989).
Independent estimates of partition coefficients for carbon tetrachloride in the same species (i.e.,
estimates from different laboratories) have also shown variability to different degrees (e.g., for
mouse: 7.83 ±2.18 [SD, Thrall et al., 2000] and 3.8 [SD not reported, Fisher et al., 2004]).
Independent estimates for rats are: 4.52 ± 0.35 (SE, Gargas et al., 1986), 5.49 ± 0.95 (SD, Evans
et al., 1994), and 4.11 ± 0.25 (Uemitsu, 1986). Mechanisms for apparent differences across
species may include interspecies differences in blood composition, including lipid, protein, and
water (Beliveau et al., 2005). Inter-laboratory variability in methods may also contribute to
variability in values reported from different laboratories (e.g., preservatives and metabolic
inhibitors added to blood). Given the above observations and the absence of conclusive evidence
to argue that the interspecies differences in reported values for the blood:air partition coefficient
are not real differences, EPA has used the reported measured values for blood:air partition
coefficients in each species in PBPK modeling to support derivation of toxicity values. These
are the values used in the reported calibration of each model (e.g., human, mouse, rat) and would
be most consistent with other estimated parameters (e.g., Vmax, Km) that also relied on the
measured values for partition coefficients (see below).
Metabolism parameters in PBPK models used in this analysis were fit to gas uptake data
in rodents. In this procedure, elimination kinetics from the chamber atmosphere (after
accounting for leaks and adsorption) are attributed to metabolism, and metabolism parameter
values (Vmax, Km) are adjusted to achieve the best fit to observations. A range of chamber
concentrations is studied, presumed to span Km, in order to provide sensitivity of the
data fitting procedure to values of Km. Although this procedure yields a measurement of whole-
body metabolism kinetics, parameters other than Vmax and Km may influence chamber
elimination kinetics (e.g., parameters that influence carbon tetrachloride concentration in the
liver such as tissue:blood partition coefficients). Uncertainties in estimates for these parameters
will contribute to uncertainties in the estimates of Vmax and Km. For these reasons, greater
215 DRAFT - DO NOT CITE OR QUOTE
-------
uncertainty is assumed for estimates of metabolism parameters than for physiological parameters
and directly measured tissue partition coefficients. Additional uncertainty enters the dosimetry
calculations from extrapolation of values for Vmax and Km from rodents to humans, when, as in
this case, no validating estimates of these parameters in humans are available. For all of the
above reasons, various values for Vmaxc were considered in modeling rodents and humans in
order to capture reported uncertainty in carbon tetrachloride metabolism kinetics in the estimates
of rodent internal dose metrics and corresponding HECs.
In the carbon tetrachloride RfC analysis, uncertainty was examined by using two dose
metrics and alternative values of Vmaxc- MRAMKL was considered the more scientifically
appropriate dose metric for liver toxicity; MCA was included given uncertainties in modeling
carbon tetrachloride metabolism and uncertainties regarding the magnitude of direct contribution
of carbon tetrachloride (as parent compound) to liver toxicity. MRAMKL provided HEC (and
thus RfC) values that were two- to sevenfold higher than those derived using MCA (depending
on the value of Vmaxc used).
Estimates of the dose metrics, MCA and MRAMKL, were sensitive to the value assigned
to the Vmaxc parameter (see Figure 5-5 and Tables 5-6 and 5-7). Several values for Vmaxc in
animals and humans have been reported (Thrall et al., 2000, Benson and Springer, 1999;
Paustenbach et al., 1988; Gargas et al., 1986); therefore, evaluation of uncertainty in this
parameter was introduced into the analysis by assuming various reported values for Vmaxc in the
estimation of HECs. Thrall et al. (2000) and Benson and Springer (1999) derived a value of
1.49 mg/hour/kg BW°'70 for humans, based on an analysis of data on in vivo (gas uptake) studies
in rodents and in vitro studies of metabolism of carbon tetrachloride in rodent and human liver
samples. Thrall et al. (2000) also derived a value of 1.7 mg/hour/kg BW°'70 for hamsters, based
on the results of closed chamber gas uptake studies. The value of 1.49 mg/hour/kg BW°'70 for
humans (Thrall et al., 2000; Benson and Springer, 1999), the value of 1.70 mg/hour/kg BW0'70
for the hamster (Thrall et al., 2000), and the two values estimated for the rat (0.4 and 0.65
mg/hour/kg BW0'70; Paustenbach et al., 1988; Gargas et al., 1986) were used in the estimation of
HECs. In general, increasing Vmaxc from 0.4 to 1.7 mg/hour/kg BW0'70 resulted in higher values
for HECs based on the MCA dose metric and lower values for HECs based on the MRAMKL
dose metric. This pattern reflects the effect of higher rates of metabolism and blood clearance at
any given exposure concentration that result from higher values for Vmax. Higher rates of
metabolism decrease the corresponding exposure concentration required to achieve a given value
of MRAMKL and increase the corresponding exposure concentration required to achieve a given
value of MCA. The effect of increasing Vmaxc was more pronounced on HECs based on the
MRAMKL dose metric. This pattern reflects the increasing influence of Vmax on metabolism
rate at higher exposure concentrations that result in liver carbon tetrachloride concentrations that
exceed the Km. The Vmaxc upon which the RfC was based, i.e., a Vmaxc based on in vitro human
data, was considered most scientifically defensible; other values of Vmaxc yielded HEC (and thus
216 DRAFT - DO NOT CITE OR QUOTE
-------
RfC) values that ranged from 4% smaller to twofold higher.
Suicide inhibition of CYP450 was not explicitly simulated in PBPK models used to
predict internal doses of carbon tetrachloride or to extrapolate external doses across species.
However, since estimates for Vmax and Km were based on in vivo gas uptake studies, the
parameter estimates reflect time-averaged estimates of these parameters during the 6-hour
duration (1-1,000 ppm) gas uptake measurements (Gargas et al. 1986; Thrall et al. 2000), during
which suicide inhibition of CYP450 probably occurred (Uemitsu, 1986). Therefore, average
rates of metabolism simulated in the animal PBPK models for exposure concentrations (e.g., 5-
125 ppm) and exposure durations (6 hours) would be expected to reflect the average Michaelis-
Menten parameter values estimated in gas uptake studies for similar exposure concentrations and
durations. What may not be accurately captured in the simulations are the effects of repeated
daily exposures on CYP450 activity and metabolism rates (i.e., cumulative effects of suicide
inhibition and induction). The rat PBPK model was able to simulate fat [14C] levels observed
during repeated exposures to [14C]-labeled carbon tetrachloride (8 hours/day, 5 days/7 days over
14 days) and excretion rates of [14C] following cessation of repeated exposures to [14C]-labeled
carbon tetrachloride (Paustenbach et al., 1988), suggesting that metabolism rates over this
duration of exposure were not substantially over- or under-predicted by the model.
Nevertheless, suicide inhibition, in the absence of induction of CYP450, would be
expected to decrease the rate of metabolism of carbon tetrachloride; therefore, models that do not
simulate suicide inhibition may over-predict the dose metric MRAMKL (rate of metabolism of
carbon tetrachloride per kg liver). The dose metric MCA (concentration of carbon tetrachloride
in blood) is relatively insensitive to Vmax (as a result of the competing respiratory elimination
pathways for carbon tetrachloride) and, therefore, quasi-steady-state values for the MCA metric
would not be expected to be appreciably affected by suicide inhibition of CYP450.
Over-prediction of MRAMKL could have several potential effects on the interspecies
extrapolation of carbon tetrachloride dosimetry. Over-prediction of MRAMKL in the animal
models may result in over- or under-prediction of BMD and BMDL values, depending on the
dose range and form of the dose-response model (e.g., linear versus nonlinear). Over-prediction
of the BMDL selected as the POD may result in over-prediction of the corresponding HEC. On
the other hand, over-prediction of MRAMKL in the human PBPK model would result in under-
prediction of the corresponding HEC.
The direction and magnitude of the effect of suicide inhibition of CYP450 on the RfC
(which was based on the MRAMKL metric) cannot be determined with any certainty without
validated animal and human PBPK models that simulate CYP450 suicide inhibition and
induction. However, it is possible to approximate the relative magnitudes of effect that
interspecies differences in suicide inhibition rates might have on CYP450 metabolism rates, by
making some simplifying assumptions. The magnitude of the effect of the inactivation of
CYP450 on the rate of carbon tetrachloride metabolism will depend on the inactivation
217 DRAFT - DO NOT CITE OR QUOTE
-------
coefficient (i.e., mole CYP450 inactivated/mole carbon tetrachloride metabolized), levels of
CYP450 in liver (nmol CYP450/g liver), Vmax (nmol/minute/g liver), Km (|iM in liver), and
carbon tetrachloride concentration in liver relative to Vmax (e.g., «Km versus >Km). Given
similar values for Vmax and Km in rat and human liver (Zanger et al., 2000), the dominant
variables affecting interspecies differences in suicide inhibition would be the inactivation
coefficient, which is substantially greater in the rat (-0.04) compared to humans (-0.005; Manno
et al., 1992, 1988), and higher basal liver CYP450 levels in rats (1.5 nmol CYP450/mg
microsomal protein) compared to humans (0.2 nmol CYP450 mg protein; Manno et al., 1992).
Figures 5-9 and 5-10 compare predicted effects of suicide inhibition on CYP450 activity
in rat and human liver microsomes, assuming (1) a constant substrate (carbon tetrachloride)
concentration; (2) effects of suicide inhibition are restricted to removal of CYP450 activity and,
thereby, reduction of Vmax (i.e., no effects on Km); and (3) the removal rate of CYP450 is given
by the suicide inhibition coefficient (moles CYP450 inactivated/mole carbon tetrachloride
metabolized). As shown in Figures 5-9 and 5-10, at carbon tetrachloride concentrations in liver
(PBPK model parameter CVL) similar to those predicted by the PBPK models (0.1-0.2 |iM) for
exposures corresponding to the POD for the derivation of the RfC (14.3 mg/m3, 2.27 ppm),
inhibition has a relatively minor effect on metabolism rate, and rates of decline of metabolism
are similar in rat and human microsomes. At higher concentrations (approaching the Km),
inhibition is more pronounced. Although this simple model of suicide inhibition in isolated liver
microsome preparations cannot accurately reflect all events that occur in vivo (e.g., effects of
CYP450 inhibition and carbon tetrachloride concentrations in liver, CYP450 induction), the
model supports the conclusion that suicide inhibition would have relatively minor effects on the
extrapolation of carbon tetrachloride external doses across species, in the low-dose range
relevant to the derivation of the RfC.
218 DRAFT - DO NOT CITE OR QUOTE
-------
*c*
a
0
Q.
M
C
E"
u
"5
E
^c_
ns
ce
Om ">
Onn/i
Onn?
n nnn
c
CVL=0.2nM Human Rat
™ ^ ^ ^ ^ ^ ^B JHL ^ ^
D 500 1000 1500 2000 25
Time(min)
00
CVL=20^M
• Human
Rat
1000 1500
Time(min)
2000
2500
Rate is expressed in units of nmol carbon tetrachloride metabolized/min/mg protein.
Time is in units of minutes of reaction time and assumes instantaneous inactivation,
whereas observed kinetics of inactivation appears to be first-order with multiple phases
(half-lives of-3-4 and 29 minutes).
Sources: Mannoetal. (1998, 1992).
Figure 5-9. Comparison of suicide inhibition profiles for liver CYP450 in
microsomes prepared from rat and human liver at substrate (carbon
tetrachloride) concentrations (CVL) similar to those predicted by the PBPK
models (0.2 uM) for exposures corresponding to the POD for the derivation
of the RfC (14.3 mg/m3, 2.27 ppm), and at 10-fold higher concentrations
(20 uM).
219
DRAFT - DO NOT CITE OR QUOTE
-------
100
_> 80
o
ii
4-J
"5 60
S?
¥ 40
> 20
0
CVL=0.2
• Human — — Rat
500 1000 1500
Time(min)
2000
2500
100
CVL=20^M
• Human — — Rat
500 1000 1500
Time(min)
2000
2500
The change in VMAX (percent of value at time=0) is shown in the vertical axis.
Time is in units of minutes of reaction time and assumes instantaneous
inactivation, whereas observed kinetics of inactivation appears to be first-order
with multiple phases (half-lives of-3-4 and 29 minutes).
Sources: Mannoetal. (1998, 1992).
Figure 5-10. Comparison of suicide inhibition profiles for liver CYP450 in
microsomes prepared from rat and human liver at substrate (carbon
tetrachloride) concentrations (CVL) similar to those predicted by the PBPK
models (0.2 uM) for exposures corresponding to the POD for the derivation
of the RfC (14.3 mg/m3, 2.27 ppm), and at 10-fold higher concentrations
(20 uM).
220
DRAFT - DO NOT CITE OR QUOTE
-------
An adequate PBPK model for the oral pathway was not available and, thus, PBPK
modeling could not be used for interspecies extrapolation in developing the RfD. In the absence
of information to quantitatively assess oral toxicokinetic or toxicodynamic differences between
animals and humans, a 10-fold UF was used to account for uncertainty in extrapolating from
laboratory animals to humans in the derivation of the RfD associated with this 10-fold UF.
The magnitude of possible over- or underestimation of interspecies differences
introduced by the use of default factors cannot be determined.
Intrahuman variability. Heterogeneity among humans is another source of uncertainty
(Lipscomb and Kedderis, 2002). Carbon tetrachloride-specific data on human variation is not
available. Quantitative information on variation in human hepatic levels of CYP2E1 and other
CYP450 enzymes is available, however, and demonstrates considerable intrahuman variability
(see Section 4.8 for additional information). Accordingly, a default UF of 10 was used to
account for uncertainty associated with human variation in the derivation of the RfD and RfC.
Human variation may be larger or smaller; however, carbon tetrachloride-specific data to
examine the potential magnitude of over- or underestimation are unavailable.
Subchronic to chronic exposure extrapolation. Because the available chronic oral
toxicity studies for carbon tetrachloride were not considered adequate for derivation of the oral
RfD, subchronic toxicity studies were used, and a UF of 3 was applied to extrapolate those data
obtained from a study of subchronic exposure to chronic exposure. This UF is based on the
assumption that an effect seen at a shorter duration will also be seen after a lifetime of exposure,
but at a lower exposure level or with greater severity. In the absence of information to inform
this extrapolation, a subchronic to chronic UF of 10 is typically applied. Inhalation data for
carbon tetrachloride and other chemical-specific information (see Section 5.1.3) indicate that a
full default UF of 10 would overestimate the difference in response following subchronic and
chronic oral exposures. The availability of carbon tetrachloride-specific information reduces the
uncertainty in extrapolating from subchronic to chronic exposure data.
Data gaps. Considering the database for carbon tetrachloride, it is possible that certain
endpoints of toxicity or certain sensitive lifestages have not been evaluated that could result in
PODs lower than those for which study data are available. The carbon tetrachloride database
lacks an adequate multigeneration study of reproductive toxicity by any route of exposure. The
absence of these types of studies introduces uncertainty in the RfD and RfC. The magnitude of
this uncertainty cannot be quantified.
221 DRAFT - DO NOT CITE OR QUOTE
-------
Vehicle effects. The vehicle used in oral gavage studies to administer carbon
tetrachloride could be a potential confounding factor in the toxicity assays. Investigators have
variably reported that (compared to an aqueous vehicle) corn oil either enhanced carbon
tetrachloride toxicity (Narotsky et al., 1997b; Condie et al., 1986), did not significantly affect
toxicity (Kaporec et al., 1995), or reduced toxicity (Kim et al., 1990b), or that influences of
vehicle could be dose-dependent (Raymond and Plaa, 1997; Narotsky et al., 1997b). The
polyethoxylated vegetable oil Emulphor has been shown to not influence carbon tetrachloride
acute hepatotoxicity, absorption, or distribution (Sanzgiri and Bruckner, 1997). Thus, it is
possible that the vehicle used in gavage studies to administer carbon tetrachloride could
influence the observed toxicity; however, given the variable effects of corn oil (versus an
aqueous vehicle), the magnitude of the confounding and the nature of the interaction of corn oil
remain uncertain.
5.4. CANCER ASSESSMENT
Several epidemiological studies (including several case-control studies and one
retrospective cohort study) have investigated potential associations between cancers of various
types and exposure to carbon tetrachloride. In all the available studies, subjects experienced
multiple chemical exposures, and the exposures were estimated qualitatively based on historical
information. These studies, therefore, can provide only limited evidence for an association
between carbon tetrachloride exposure and cancer, and are not useful for dose-response analysis.
Studies in experimental animals suggest that the primary cancer risk associated with
exposure to carbon tetrachloride is development of liver cancer. Carbon tetrachloride produced
hepatocellular carcinomas in rats, mice, and hamsters in oral studies and in rats and mice by
inhalation exposure. In addition to liver tumors, adrenal pheochromocytomas were observed in
male and female mice by oral (NTP, 2007; Weisburger, 1977) and inhalation (Nagano et al.,
2007b; JBRC, 1998) exposure. No increase in pheochromocytomas was observed in rats.
Selection of a low-dose extrapolation approach to assess cancer risk for carbon
tetrachloride was guided by EPA's (2005 a) Guidelines for Carcinogen Risk Assessment.
According to these guidelines, a linear extrapolation approach is used as the default approach:
[w]hen the weight of evidence evaluation of all available data are insufficient to establish
the mode of action for a tumor site and when scientifically plausible based on the
available data,... because linear extrapolation generally is considered to be a health-
protective approach.
A nonlinear extrapolation approach should be selected for assessing cancer risk:
when there are sufficient data to ascertain the mode of action and conclude that it is not
222 DRAFT - DO NOT CITE OR QUOTE
-------
linear at low doses and the agent does not demonstrate mutagenic or other activity
consistent with linearity at low doses. Special attention is important when the data
support a nonlinear mode of action but there is also a suggestion of mutagenicity.
Depending on the strength of the suggestion of mutagenicity, the assessment may justify
a conclusion that mutagenicity is not operative at low doses and focus on a nonlinear
approach, or alternatively, the assessment may use both linear and nonlinear approaches.
Both linear and nonlinear approaches may be presented "[w]here alternative approaches
with significant biological support are available for the same tumor response and no scientific
consensus favors a single approach" or "when there are multiple modes of action." The
Guidelines for Carcinogen Risk Assessment also suggest that "[i]f there are multiple modes of
action at a single tumor site, one linear and another nonlinear, that both approaches are used to
decouple and consider the respective contributions of each mode of action in different dose
ranges."
As discussed in Section 4.7.3.4, the mechanisms underlying the induction of liver toxicity
by carbon tetrachloride has been extensively investigated. Biological support exists for a
hypothetical MOA involving metabolism of carbon tetrachloride by CYP2E1, sustained
cytotoxicity, and regenerative cell proliferation as key events driving the steep nonlinear increase
in liver tumor dose-response at relatively high carbon tetrachloride exposures. Several pieces of
evidence suggest, however, that carbon tetrachloride carcinogenicity is not explained by a
cytotoxic-proliferative MOA. This evidence, described further in the paragraphs that follow,
include: an increased incidence of liver tumors in the low-dose female mouse (Nagano et al.,
2007b) in the absence of nonneoplastic liver toxicity; induction of pheochromocytomas in mice,
a tumor for which the MOA(s) is/are unknown; fundamental reactivity of the chemical; and
absence of data on low-dose genotoxicity.
At lower exposure levels, the correspondence between hepatocellular cytotoxicity and
regenerative hyperplasia and the induction of liver tumors is inconsistent. In particular, liver
findings from the JBRC bioassay (Nagano et al., 2007b; JBRC, 1998) suggest that mouse
hepatocarcinogenicity cannot be explained in terms of the cytotoxic-proliferative MOA. An
increased incidence of hepatocellular adenomas occurred in the low-dose (0.9-ppm adjusted)
female mouse in the absence of nonneoplastic liver toxicity, raising the possibility of another
MOA operating in addition to or in conjunction with the cytotoxic-proliferative MOA.
Cytotoxicity and cellular regeneration have been documented at somewhat higher doses (5 ppm
adjusted; see Table 4-15), even though they have not been directly observed at the dose level
(0.9 ppm adjusted) inducing a significant increase in liver adenomas in the female mouse model.
Carbon tetrachloride also induced pheochromocytomas in male and female mice by oral
(NTP, 2007; Weisburger, 1977) and inhalation (Nagano et al., 2007b; JBRC, 1998) exposure.
The MOA(s) for the induction of pheochromocytomas in the mouse is unknown.
223 DRAFT - DO NOT CITE OR QUOTE
-------
Other considerations suggest that the carbon tetrachloride database is insufficient for
ruling out other MO As at low exposure levels, in particular considerations related to the
compound's genotoxicity and general reactivity. Carbon tetrachloride is metabolized to reactive
species (trichloromethyl and trichloromethyl peroxy radical), and subsequent chemical reactions
of carbon tetrachloride metabolites with cellular constituents lead to formation of reactive
oxygen species that also can damage DNA and other macromolecules. The potential exists for
biologically-active carbon tetrachloride metabolites to react with macromolecules at low
exposures (i.e., exposure levels below doses that are cytotoxic); however, data to characterize
this low-exposure activity are limited.
Results of extensive testing for genotoxic and mutagenic potential are largely negative,
and a weight of evidence analysis of the genotoxicity literature suggests that carbon tetrachloride
is more likely an indirect than direct mutagenic agent; however, the nature of the database does
not characterize the role of genotoxicity at low doses of carbon tetrachloride. There is little
direct evidence that carbon tetrachloride induces intragenic or point mutations in mammalian
systems. The mutagenicity studies that have been performed using transgenic mice have yielded
negative results, as have the vast majority of the mutagenicity studies that have been conducted
in bacterial systems. Under highly cytotoxic conditions, bioactivated carbon tetrachloride can
exert genotoxic effects. These tend to be modest in magnitude and are manifested primarily as
DNA breakage and related sequelae. Chromosome loss leading to aneuploidy may also occur to
a limited extent. The fact that carbon tetrachloride overall has not been found to be a potent
mutagen and that positive genotoxic results are found only at high exposure levels and generally
in concert with cytotoxic effects (see Tables 4-8 to 4-11) indicates that carbon tetrachloride does
not likely induce genotoxic effects through direct binding or damage to DNA. The majority of
genotoxicity studies, however, have been conducted at relatively high exposure levels such that
the potential for genotoxic activity at low doses cannot be determined.
Thus, as summarized above and in Section 4.7.3.4, biological support exists for a
hypothesized cytotoxicity-regenerative MO A as a major MO A driving the steep nonlinear
increase in liver tumor dose-response at relatively high carbon tetrachloride exposures.
Additionally, at high exposures, both the cytotoxicity-regenerative proliferation-based MOA and
a mutagenicity-based MOA may be operative, but it is not possible to delineate the contribution
of these possible MOA(s) to carbon tetrachloride tumor response. Inconsistencies at the low end
of the experimental exposure range (including evidence from the JBRC bioassay that indicates
that female mouse liver tumors cannot simply be explained in terms of the cytotoxic-proliferative
MOA, the findings of pheochromocytomas in mouse by oral [NCI bioassay] and inhalation
[JBRC bioassay] exposure for which the MOA is unknown, and insufficient data at low doses to
rule out the possibility of low-dose genotoxicity or other biological responses to a reactive
chemical) suggest that other (or another) MO As independent of cytotoxicity and regenerative
cell proliferation may be operative in this range. Therefore, consistent with the 2005 Guidelines
224 DRAFT - DO NOT CITE OR QUOTE
-------
for Carcinogen Risk Assessment (U.S. EPA, 2005a), linear low-dose extrapolation as a default
approach is applied to data for liver tumors and pheochromocytomas.
Broader science considerations based on scientific literature not specific to carbon
tetrachloride also support inferences about potential risks of carbon tetrachloride at lower doses.
EPA guidance and reports from expert advisory bodies (including NRC, 2009) have provided
broad and long-standing scientific arguments in favor of low-dose linear approaches to cancer
risk assessment based on the following principles:
• A chemical's carcinogenic effects may act additively to ongoing biological processes,
given that diverse human populations already have substantial background incidence of
various tumors (e.g., Crump et al., 1976);
• A broadening of the dose-response curve in the human population (i.e., less rapid fall-off
with dose) and, accordingly, a greater potential for risks from low-dose exposures (see
Zeise et al., 1987; Lutz et al., 2005) would result for two reasons. First, even if there is a
threshold concentration at the cellular level, that threshold is likely to be different among
different individuals. Secondly, greater variability in response to exposures in the
heterogeneous human population would be anticipated than in controlled laboratory
species and conditions (due to, for example, genetic variability, disease states, nutrition,
age).
• The general use of linear extrapolation provides plausible upper-bound risk estimates and
also provides consistency across assessments.
An alternative nonlinear extrapolation approach for carbon tetrachloride liver tumors is
presented in Section 5.4.5. This approach is consistent with empirical evidence that supports the
hypothesis that liver carcinogenicity occurs at carbon tetrachloride exposures that also induce
hepatocellular toxicity and a sustained regenerative and proliferative response, and that
exposures that do not cause sustained cytotoxicity and regenerative cell proliferation would be
protective of liver tumors if this is the primary MOA. Because evidence indicates that carbon
tetrachloride-induced liver tumors are not explained only by a cytotoxic-proliferative MOA and
the unknown MOA for carbon tetrachloride-induced pheochromocytomas, EPA does not
consider a nonlinear extrapolation approach for carbon tetrachloride to be supported.
5.4.1. Choice of Study/Data—with Rationale and Justification
5.4.1.1. Inhalation Data
As noted previously, epidemiological studies of populations exposed to carbon
225 DRAFT - DO NOT CITE OR QUOTE
-------
tetrachloride provide only limited evidence for an association between carbon tetrachloride
exposure and human cancer and are not adequate for dose-response analysis.
The only chronic bioassay of carbon tetrachloride by the inhalation route is the 104-week
inhalation bioassay in rats and mice conducted by JBRC (Nagano et al., 2007b; JBRC, 1998), a
bioassay that provides data adequate for dose-response modeling. In this bioassay, F344 rats and
BDF1 mice were exposed to 0, 5, 25, or 125 ppm carbon tetrachloride, 6 hours/day, 5 days/week,
for 2 years. Carbon tetrachloride produced a statistically significant increase in hepatocellular
adenomas and carcinomas in rats and mice of both sexes, and adrenal pheochromocytomas in
mice of both sexes.
5.4.1.2. Oral Data
Studies of carbon tetrachloride carcinogen!city by the oral exposure route are not
sufficient to derive a quantitative estimate of cancer risk using low-dose linear approaches. No
epidemiological investigations of the possible carcinogenicity of carbon tetrachloride associated
with oral exposure have been performed. The cancer studies by Edwards et al. (1942) in the
mouse and Delia Porta et al. (1961) in the hamster included a control and only one dose group,
and animals were dosed for less than a lifetime (2 months and 30 weeks, respectively). Neither
study provided body weight information, so doses could not be estimated with certainty. Despite
the relatively short dosing periods and the fact that animals were kept on study for less than a
lifetime (approximately 6.5 months in the case of Edwards et al., 1942, and approximately 1 year
in the case of Delia Porta et al., 1961), liver tumor incidence was high (71% in the case of
Edwards et al., 1942, and 100% of the hamsters that died or were sacrificed between weeks 43
and 55 in the case of Delia Porta et al., 1961). In the NCI bioassays (1977, 1976a, b), liver tumor
incidence in the mouse was virtually 100% in both dose groups. In the rat, liver tumor incidence
was low and failed to show a dose-response relationship (in the female rat, tumor incidence was
higher in the low-dose group [4/46] than in the high-dose group [1/30], presumably because
early mortality in the high-dose group precluded tumor formation). Thus, none of the available
oral studies of carbon tetrachloride carcinogenicity provided data sets amenable to dose-response
modeling.
5.4.2. Dose-Response Data
5.4.2.1. Inhalation Data
Dose-response modeling was performed for five tumor responses from the JBRC
bioassay: adenoma and carcinoma of the liver in female rats, adenoma and carcinoma of the liver
in male and female mice, and pheochromocytomas in male and female mice. Incidence data for
liver tumors and pheochromocytomas are summarized in Tables 5-8 and 5-9 below.
226 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-8. Incidence of liver tumors in F344 rats and BDF1 mice exposed
to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Tumor
Male
0 ppm
5 ppm
25 ppm
125 ppm
Female
0 ppm
5 ppm
25 ppm
125 ppm
RAT
Hepatocellular
adenoma or
carcinoma
l/50a
1/50
1/50
40/50b
0/50a
0/50
3/50
44/50b
MOUSE
Hepatocellular
adenoma or
carcinoma
24/503
20/50
49/50b
49/50b
4/50a
9/49
44/50b
48/49b
a Statistically significant trend for increased tumor incidence by Peto's test (p< 0.01).
b Tumor incidence significantly elevated compared with that in controls by Fisher's exact test (p< 0.01).
Sources: Nagano et al. (2007b); JBRC (1998).
Table 5-9. Incidence of adrenal tumors (pheochromocytomas) in BDF1
mice exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day,
5 days/week)
Tumor
Adrenal
pheochromocytoma0
Male
0 ppm
0/50a
5 ppm
0/50
25 ppm
16/50b
125 ppm
32/50b
Female
0 ppm
0/50a
5 ppm
0/49
25 ppm
0/50
125 ppm
22/49b
a Statistically significant trend for increased tumor incidence by Peto's test (p < 0.01).
b Tumor incidence significantly elevated compared with controls by Fisher's exact test (p < 0.01).
0 All pheochromocytomas in the mouse were benign with the exception of one malignant pheochromocytoma in
the 125-ppm male mouse group.
Sources: Nagano et al. (2007b); JBRC (1998).
The male rat data for liver adenomas and carcinomas were not modeled because this data
set lacked the resolution desired for dose-response modeling. Tumor frequency increased from
control levels to close to maximal response without any intervening dose levels having
submaximal responses. In the female rat, lower but biologically significant levels of response
were seen at intermediate dose levels. Further, the incidence of liver tumors was higher in the
female rat compared with the male rat, such that the female rat data would provide the higher
estimate of risk of the two data sets.
227
DRAFT - DO NOT CITE OR QUOTE
-------
For the female mouse, the bioassay data set contained two exposure concentrations (mid-
and high-exposure concentrations) at which close to maximal responses were seen. Preliminary
r\
fitting of a multistage model revealed that: (1) a fit with an adequate % based p-value was not
obtained, and (2) the fit and parameter estimates were highly sensitive to the precise finding of
48/49 tumors at the highest concentration. (A hypothetical shift of the data to 49/49 tumors led
to a good model fit with different powers of the multistage model involved in the fit.) As these
distinctions were not judged biologically based, multistage model fits below were conducted
without use of the highest exposure concentration data, an approach commonly used in BMD
modeling when high dose data are not compatible with model fits.)
Dose-response modeling was also conducted for pheochromocytomas observed in the
JBRC mouse bioassay. These tumors, with one exception, were characterized as benign rather
than malignant. Unlike liver tumors associated with carbon tetrachloride exposure, which have
been observed in numerous bioassays in multiple species, pheochromocytomas have been
observed in only one species (mouse). Thus, the finding of pheochromocytomas in the mouse
may present a less certain human health hazard than does the finding of liver tumors in
experimental animals. The decision to develop dose-response models for pheochromocytomas
was based on guidance provided in the 2005 Guidelines for Carcinogen Risk Assessment (U.S.
EPA, 2005a), which states that "benign tumors that are not observed to progress to malignancy
are assessed on a case-by-case basis." The benign tumor type seen in the mouse has a human
equivalent that is damaging to human health and can lead to fatal sequelae. In humans,
pheochromocytomas are rare and usually benign neuroendocrine tumors, but may also present as
or develop into a malignancy (Eisenhofer et al., 2004). Salmenkivi et al. (2004) noted that
approximately 10% of pheochromocytomas in humans metastasize. The presence of one
observed malignant tumor in the mouse study also suggests the potential for these benign tumors
to progress to malignancy. The oral NCI bioassay characterized adrenal gland tumors simply as
"pheochromocytoma" (incidence data are provided in Weisburger, 1977, and NTP, 2007). This
characterization suggests that the tumors were not malignant, although the status as benign or
malignant was not clearly established. Finally, Salmenkivi et al. (2004) observed that while
most pheochromocytomas are benign, differentiating a benign tumor from a malignant tumor
only by histological criteria is difficult. Thus, it was considered appropriate to conduct dose-
response modeling for pheochromocytomas.
In the analyses of the mouse and rat carbon tetrachloride inhalation data that follow,
incidence reflects that of benign or malignant tumors combined. Data are not available to
indicate whether malignant tumors developed specifically from progression of the benign
tumors; however, etiologically similar tumor types (i.e., benign and malignant tumors of the
same cell type) were combined for these analyses because of the possibility that the benign
tumors could progress to the malignant form (U.S. EPA, 2005a).
228 DRAFT - DO NOT CITE OR QUOTE
-------
5.4.2.2. Oral Data
As noted above, oral cancer bioassay data for carbon tetrachloride are not adequate for
dose-response analysis. Therefore, PBPK modeling was applied to extrapolate inhalation tumor
data to the oral route. Because liver tumors and pheochromocytomas have been observed in
experimental animals following both inhalation and oral exposures, the data sets evaluated as the
basis for the IUR were considered appropriate for estimation of an oral SF. The route-to-route
extrapolation method is described further below.
5.4.3. Dose Adjustments and Extrapolation Methods
5.4.3.1. General Approach to Modeling and Extrapolation of Animal Data to Humans
Cancer risk estimates were obtained by straight line extrapolation from the POD to zero
as described in the EPA's Guidelines for Carcinogenic Risk Assessment (U.S. EPA, 2005a). As
stated in the guidelines, "The linear approach is to draw a straight line between a point of
departure from observed data, generally as a default, an LED [lower bound of effective dose]
chosen to be representative of the lower end of the observed range, and the origin (zero
incremental dose, zero incremental response)." Linear extrapolation is used as the approach in
the absence of data supporting a biologically based model for extrapolation outside of the
observed range (U.S. EPA, 2005a).
The general procedure for deriving the POD from animal bioassay data is the same as
that used to derive the POD for RfC derivation and is depicted in Figure 5-4. Exposure levels
studied in the 2-year rat and mouse bioassays (Nagano et al., 2007b; JBRC, 1998) were
converted to estimates of internal doses by application of the rat and mouse PBPK models.
BMD modeling methodology (U.S. EPA, 2000c, 1995) was used to analyze the relationship
between the estimated internal doses and response (i.e., liver tumors and pheochromocytomas).
The resulting BMDL values were converted to estimates of HECs and human equivalent doses
(HEDs) by applying the human PBPK model.
5.4.3.2. PBPK Modeling for Internal Dose Metrics
Estimation of internal doses corresponding to the exposure concentrations studied in the
2-year rat and mouse bioassays (Nagano et al., 2007b; JBRC, 1998) was accomplished using
PBPK models of the rat (Thrall et al., 2000; Benson and Springer, 1999; Paustenbach et al.,
1988) and mouse (Fisher et al., 2004; Thrall et al., 2000), respectively (see Sections 3.5 and
Appendix C for description of the models). The review, selection and application of the chosen
PBPK models was informed by an EPA report (U.S. EPA, 2006c), which addresses the
application and evaluation of PBPK models. The PBPK models were used to simulate internal
dose metrics corresponding to exposure concentrations studied in the 2-year bioassays: 5, 25, and
125 ppm, 6 hours/day, 5 days/week (Nagano et al., 2007b; JBRC, 1998). Internal dose metrics
229 DRAFT - DO NOT CITE OR QUOTE
-------
were selected that were considered to be most relevant to the toxicity endpoints of interest (i.e.,
liver tumors and pheochromocytomas), based on consideration of evidence for MOA of carbon
tetrachloride. Two dose metrics were selected based on available information on the
mechanisms of carbon tetrachloride liver toxicity: (1) time-averaged arterial blood concentration
of carbon tetrachloride (MCA, jimol/L); and (2) time-averaged rate of metabolism of carbon
tetrachloride (MRAMKL, jimol/hour/kg liver). Liver metabolism rate was selected as the
primary dose metric for liver effects based on evidence that metabolism of carbon tetrachloride
via CYP2E1 to highly reactive free radical metabolites plays a crucial role in its MOA in
producing liver toxicity (described in Section 4.5). Further support for rate of hepatic
metabolism as a dose metric is provided in Section 5.2.2.1. Because of acknowledged
uncertainties regarding the accuracy of available PBPK models to simulate carbon tetrachloride
(see Section 5.2.2.1), arterial blood concentration of carbon tetrachloride was also included in the
analysis of liver tumor data as a more proximal dose metric to liver metabolism.
Data on incidence of adrenal pheochromocytomas in mice were also analyzed. The
MRAMKL dose metric was excluded from consideration in the analysis of pheochromocytomas
on the basis that reactive metabolites of carbon tetrachloride formed in the liver are unlikely to
be sufficiently stable to contribute to toxicity or transformations of cells in the adrenal gland.
Although it is possible that local generation of reactive metabolites may contribute to the
production of pheochromocytomas, PBPK models available for this analysis do not simulate
uptake and metabolism of carbon tetrachloride in the adrenal gland. The model of Yoon et al.
(2007) is the only one available that includes extra-hepatic metabolism, specifically in lung and
kidney. Based on the model and estimates of Michaelis-Menten parameters for carbon
tetrachloride metabolism in lung and kidney, metabolism in each of these tissues was estimated
to be <1% of that in the liver, and had a negligible effect on MCA and MRAMKL. It would be
expected, however, that rates of metabolism in all tissues, including the adrenal gland, would be
dependent on delivery of carbon tetrachloride to these tissues and, thereby, would be correlated
with blood concentrations of carbon tetrachloride. Therefore, the MCA dose metric was used to
represent the internal dose in BMD modeling of pheochromocytoma incidence in mice.
The two dose metrics, MCA and MRAMKL, were simulated in the rat and mouse PBPK
models as time-averaged values, with the averaging time being the chronic exposure period (e.g.,
2 years). See Equations 5-1 and 5-2 (Section 5.2.2.1) for the calculation of the time-averaged
dose metrics.
Internal dose metrics corresponding to the exposure concentrations studied in the 2-year
rat inhalation bioassay (Nagano et al., 2007b; JBRC, 1998) for two values of Vmaxc were
provided previously in Table 5-5 (see Section 5.2.2.1). Internal dose metrics corresponding to
the exposure concentrations studied in the 2-year mouse inhalation bioassay (Nagano et al.,
2007b; JBRC, 1998) as derived from the Fisher et al. (2004) and Thrall et al. (2000) PBPK
models are presented in Table 5-10. The Fisher et al. (2004) model predicts lower values for
230 DRAFT - DO NOT CITE OR QUOTE
-------
MCA than the Thrall et al. (2000) model. This is at least partly explained by the higher values
for tissue:blood partition coefficients in the Fisher et al. (2004) model, which results in a larger
fraction of the body burden outside of the vascular compartment. The Fisher et al. (2004) model
predicts higher values for MRAMKL at exposure concentrations above approximately 40 ppm.
At least two factors contribute to this pattern: (1) the higher liverblood partition coefficient in
the Fisher et al. (2004) model results in higher concentrations of carbon tetrachloride in the liver;
and (2) the higher Vmaxc in the Fisher et al. (2004) model results in increases in liver metabolism
rate at any given liver concentration of carbon tetrachloride, with the more pronounced
enhancement of metabolism at liver concentrations above the Km. The exposure concentration-
dependence of the dose metrics estimated from both models is shown in Figure 5-11.
Table 5-10. Internal dose metrics predicted from Fisher et al. (2004) and Thrall et
al. (2000) PBPK mouse models
Exposure
(ppm)
5
25
125
MCA
(umol/L)
Fisher et al.
0.111
0.603
3.315
Thrall et al.
0.213
1.226
6.856
MRAMKL
(umol/hr/kg liver)
Fisher et al.
12.666
41.675
71.589
Thrall et al.
15.456
43.599
63.596
Values are for 0.036 kg mouse.
MCA, time-averaged arterial concentration of carbon tetrachloride; MRAMKL, time-averaged rate of metabolism of
carbon tetrachloride per kg liver.
231
DRAFT - DO NOT CITE OR QUOTE
-------
12.0
10.0
8.0
<
O 6.0
4.0
2.0
0.0
0
0
50 100 150
AIR (ppm)
200
250
50 100 150
AIR (ppm)
200
250
Dose metrics shown are time-averaged arterial concentration of carbon tetrachloride
(MCA, umol/L, upper panel), and time-averaged rate of metabolism of carbon
tetrachloride (MRAMKL, umol/hr/kg liver, lower panel). The dose metrics are plotted
against exposure concentration (6 hours/day, 5 days/week, 2 years) for a 0.036 kg mouse.
Figure 5-11. Internal dose metrics predicted from the Fisher et al. (2004)
and Thrall et al. (2000) PBPK mouse models.
232
DRAFT - DO NOT CITE OR QUOTE
-------
5.4.3.3. BMD Modeling of Response Data from Animal Bioassays
BMD modeling methodology (U.S. EPA, 2000c, 1995) was used to analyze data on
estimated internal doses (i.e., MCA, MRAMKL) and incidence data (i.e., liver tumors in rats,
and liver tumors and adrenal pheochromocytomas in mice) from the 2-year rat and mouse
inhalation bioassays (Nagano et al., 2007b; JBRC, 1998). The multistage model in U.S. EPA's
BMDS (version 1.4.1) (U.S. EPA, 2007b) was fit to the tumor incidence data for rats and mice.
When adequate fit could not be achieved with the multistage model, other models from the
BMDS suite of models were fit. The results of the BMD modeling are summarized below;
detailed model outputs are provided in Appendix E.
Female F344 rat — hepatocellular adenomas + carcinomas
Internal doses associated with a BMR of 5% extra risk of liver tumors were calculated. A
BMR of 5% excess risk was in the low range of experimental data for the rat (see Appendix E)
In addition, a BMR of 5% excess risk was preferred over a BMR of 10% (a 10% BMR is
commonly used in BMD modeling as a means of facilitating comparison across assessments and
endpoints) in the interest of moving the POD further from the range where hepatocellular
toxicity and a proliferative/regenerative response was observed and where tumor induction is
more likely influenced by the hypothesized cytotoxic-proliferative MOA.
BMD modeling using the multistage model in BMDS was performed using the female rat
liver tumor incidence data shown in Table 5-8 and internal doses shown in Table 5-5. A
summary of the resulting BMDs and BMDLs values is presented in Table 5-11 (columns 2
and 3).
233 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-11. BMD values for incidence data for liver tumors (adenoma and
carcinoma) in female F344 rats and corresponding HEC and HED values
Metric
(l)b
MCA
MRAMKL
BMD modeling"
VMAXCR=
0.4
(2)
BMD5: 0.61
(136.8)
BMDL5: 0.39
(90.29)
BMD5: 9.82
(109.0)
BMDL5: 8.40
(85.71)
VMAXCR
= 0.65
(3)
BMD5: 0.59
(143.2)
BMDL5: 0.35
(88.94)
BMD5: 14.6
(116.4)
BMDL5: 12.3
(91.37)
VMAXCH
(4)
0.40
0.65
1.49
1.70
0.40
0.65
1.49
1.70
HEC
VMAXCR
= 0.4
(5)
26.083
27.605
27.605
31.273
107.759
63.915
39.635
37.771
VMAXCR
= 0.65
(6)
23 922
25.318
28.203
28.667
236.171
105.882
59.326
56.236
HED
VMAXCR
= 0.4
(7)
3.65
4.27
6.35
6.87
5.10
3.03
1.88
1.79
VMAXCR
= 0.65
(8)
3.37
3.96
5.95
6.44
11.19
5.01
2.81
2.66
Rats were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Internal doses modeled
correspond to exposure concentrations: 0, 5, 25, or 125 ppm.
HEC, human equivalent concentration, mg/m3; HED, human equivalent dose, mg/kg-day; MCA, time-averaged
arterial blood concentration, (imol/L; MRAMKL, time-averaged rate of metabolism per kg liver, (imol/hr/kg liver;
VMAXC, maximum rate of metabolism in humans (H) or rat (R), mg/hr/kg B W
a MCA, multistage (2-stage); MRAMKL, multistage (4-stage)
BMR (benchmark response) = 5%. Values in parentheses are animal exposure concentrations (mg/m3)
corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
A second analysis was performed to examine the effect on the cancer risk estimate of
using only carbon tetrachloride cancer response data at exposure levels below those associated
with evidence of cell replication. In the female F344 rat, the 3/50 hepatocellular carcinoma
response at 25 ppm (an exposure concentration at which cytotoxicity occurred but below which
regenerative proliferation was reported; see Table 4-17) is statistically significant (two-tailed p-
value of 0.0002) when compared to the historical control incidence of 2/1,797 for female rats for
the same strain and research center (email data April 5, 2007, from Kasuke Nagano, JBRC, to
Susan Rieth, U.S. EPA). A comparison to concurrent controls in the JBRC study did not yield a
statistically significant difference in response; however, because the observed carcinomas in
female rats at 25 ppm are part of a trend of increasing carcinoma incidences with increasing
exposure, it is reasonable to consider the tumors to be biologically significant.
As noted above, cytotoxicity was reported in female rats at 25 ppm in the 104-week
study, but regeneration and proliferation were not reported at this exposure level; additionally,
regeneration and proliferation were not observed in 13-week studies at 30 ppm and below
(Table 4-17). Thus, the tumor response at 25 ppm can be considered as potentially independent
of, or at most minimally influenced by, regenerative proliferation.
234
DRAFT - DO NOT CITE OR QUOTE
-------
A multistage POD model of the control, 5 ppm, and 25 ppm exposure groups
(Table 5-12, columns 2 and 3) is provided for comparison with the results above for the full data
set (see Table 5-11, columns 2 and 3).
Table 5-12. BMD values for incidence data for liver tumors (adenoma and
carcinoma) in female F344 rats (high dose dropped) and corresponding HEC and
HED values
Metric
(l)b
MCA
MRAMKL
BMD modeling"
VMAXCR=
0.4
(2)
BMD5: 0.65
(145.1)
BMDL5: 0.35
(81.62)
BMD5: 11.6
(145.1)
BMDL5: 6.92
(65.27)
VMAXCR
= 0.65
(3)
BMD5: 0.60
(145.4)
BMDL5: 0.32
(81.88)
BMD5: 16.7
(143.3)
BMDL5: 9.76
(67.82)
VMAXCH
(4)
0.40
0.65
1.49
1.70
0.40
0.65
1.49
1.70
HEC
VMAXCR
= 0.4
(5)
23.339
24.701
27.512
27.965
79.943
50.626
32.384
30.918
VMAXCR
= 0.65
(6)
21.459
22.713
25.288
25.701
140.519
77.275
46.414
44.157
HED
VMAXCR
= 0.4
(7)
3.29
3.88
5.84
6.33
3.79
3.05
1.53
1.46
VMAXCR
= 0.65
(8)
2.40
3.61
5.48
5.95
6.66
3.66
2.20
2.09
Rats were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Internal doses modeled
correspond to exposure concentrations: 0, 5, or 25 (125 ppm exposure was dropped).
HEC, human equivalent concentration, mg/m3; HED, human equivalent dose, mg/kg-day; MCA, time-averaged
arterial blood concentration, (imol/L; MRAMKL, time-averaged rate of metabolism per kg liver, ^mol/hr/kg liver;
VMAXC, maximum rate of metabolism in humans (H) or rat (R), mg/hr/kg B W°70
a MCA, multistage (2-stage); MRAMKL, multistage (2-stage)
BMR (benchmark response) = 5%. Values in parentheses are animal exposure concentrations (mg/m3)
corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
See Appendix E for the BMDS model outputs and graphs of the modeled data.
Female BDF1 mouse - hepatocellular adenomas + carcinomas
As with the female rat liver tumor data, EPA considered a BMR of 5% excess risk in the
interest of moving the POD further from the range where hepatocellular toxicity and a
proliferative/regenerative response was observed and where tumor induction may more likely be
influenced by a cytotoxic-proliferative MOA. In the case of the female mouse liver tumor data,
however, a BMR of 5% fell well below the experimental range; therefore, a BMR of 10% was
used in the BMD modeling of female mouse liver tumor data.
BMD modeling using the multistage model in BMDS was performed using the female
mouse liver tumor incidence data shown in Table 5-8 and internal doses shown in Table 5-10.
As noted in Section 5.4.2.1, the multistage model fits below were conducted without use of the
235
DRAFT - DO NOT CITE OR QUOTE
-------
highest exposure concentration data, an approach commonly used in BMD modeling when high
dose data are not compatible with model fits. A summary of the resulting BMDio and BMDLio
values is presented in Table 5-13 (columns 2 and 3).
Table 5-13. BMD values for incidence data for liver tumors (adenoma and
carcinoma) in female BDF1 mice (high dose dropped) and corresponding HEC and
HED values
Metric
(l)b
MCA
MRAMKL
BMD modeling"
Fisher
(2)
BMD10:0.10
(28.41)
BMDL10: 0.047
(13.54)
BMD10: 9.71
(23.45)
BMDL10: 6.32
(14.82)
Thrall
(3)
BMD10:0.19
(28.20)
BMDL10: 0.088
(13.42)
BMD10: 10.4
(19.98)
BMDL10: 7.59
(14.16)
VMAXCH
(4)
0.40
0.65
1.49
1.70
0.40
0.65
1.49
1.70
HEC
Fisher
(5)
3.197
3.385
3.753
3.811
70.278
45.526
29.466
28.152
Thrall
(6)
6.042
6.396
7.097
7.208
91.709
56.492
35.646
34.005
HED
Fisher
(7)
0.50
0.61
0.99
1.08
3.33
2.16
1.40
1.33
Thrall
(8)
0.94
1.14
1.82
2.00
4.34
2.68
1.69
1.61
Mice were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, or 25 ppm (125 ppm exposure dropped)
HEC, human equivalent concentration, mg/m3; HED, human equivalent dose, mg/kg-day; Fisher, Fisher et al.
(2004) model; MCA, time-averaged arterial blood concentration, (imol/L; MRAMKL, time-averaged rate of
metabolism per kg liver, ^mol/hr/kg liver; Thrall, Thrall et al. (2000) model; VMAXCH, maximum rate of
metabolism in humans, mg/hr/kg BW°70
a MCA, multistage (2-stage); MRAMKL, multistage (2-stage)
BMR (benchmark response) = 10%. Values in parentheses are animal exposure concentrations (mg/m3)
corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
As with the rat, a second analysis was performed with female mouse liver tumor data to
examine the effect on the cancer risk estimate of using only carbon tetrachloride cancer response
data at exposure levels below those associated with evidence of cell replication. A multistage
model POD calculation using only the control and 5-ppm exposure group (Table 5-14, columns 2
and 3) is provided for comparison with the results above for the full data set (Table 5-13,
columns 2 and 3). As discussed further in Section 5.4.4.2, the analysis based on the control and
5-ppm dose groups provides a less informative characterization of the dose-response curve than
does the analysis based on the control, 5-ppm and 25-ppm dose groups. The latter analysis does,
however, provide some information on the dose-response relationship for liver tumors in the
female mouse at concentrations below levels documented to cause cell replication.
236
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-14. BMD values for incidence data for liver tumors (adenoma and
carcinoma) in female BDF1 mice (2 highest doses dropped) and corresponding HEC
and HED values
Metric
(l)b
MCA
MRAMKL
BMD modeling"
Fisher
(2)
BMD10:0.10
(28.41)
BMDL10: 0.044
(12.68)
BMD10: 11.6
(28.51)
BMDL10: 5.05
(11.72)
Thrall
(3)
BMD10: 0.20
(29.61)
BMDL10: 0.085
(12.97)
BMD10: 14.2
(28.49)
BMDL10:6.16
(11.33)
VMAXCH
(4)
0.40
0.65
1.49
1.70
0.40
0.65
1.49
1.70
HEC
Fisher
(5)
3.025
3.202
3.550
3.605
52.187
35.277
23.367
22.358
Thrall
(6)
5.792
6.132
6.804
6.910
67.796
44.180
28.683
27.410
HED
Fisher
(7)
0.48
0.58
0.94
1.03
2.47
1.67
1.11
1.06
Thrall
(8)
0.91
1.09
1.75
1.92
3.21
2.09
1.36
1.30
Mice were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0 or 5 (25 ppm and 125 ppm exposures were dropped)
HEC, human equivalent concentration, mg/m3; HED, human equivalent dose, mg/kg-day; Fisher, Fisher et al.
(2004) model; MCA, time-averaged arterial blood concentration, (imol/L; MRAMKL, time-averaged rate of
metabolism per kg liver, (imol/hr/kg liver; Thrall, Thrall et al. (2000) model; VMAXCH, maximum rate of
metabolism in humans, mg/hr/kg BW°70
a MCA, multistage (2-stage); MRAMKL, multistage (2-stage)
BMR (benchmark response) = 10%. Values in parentheses are animal exposure concentrations (mg/m3)
corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
See Appendix E for the BMDS model outputs and graphs of the modeled data.
Male BDF1 mouse - hepatocellular adenomas + carcinomas
Internal doses associated with a BMR of 10% extra risk of liver tumors were calculated
for the male mouse. As with the female mouse liver tumor data, a BMR of 10% was used in the
BMD modeling.
Similar to the male rat data for liver adenomas and carcinomas, the male mouse data
provided poor resolution of the dose-response relationship for liver tumors. Tumor incidence in
5-ppm male mice was below the control level, and was close to maximal response (49/50) at the
mid- and high-exposure groups, without any intervening dose levels having submaximal
responses. BMD modeling of this data set (shown in Table 5-8) and internal doses (shown in
Table 5-10) revealed that none of the dichotomous models in BMDS provided an adequate fit of
the liver tumor data. Therefore, multistage model fits were conducted without use of the highest-
exposure group (125-ppm) data. A marginal fit of the data was obtained when the multistage
model was applied to this reduced data set. A summary of the resulting BMDio and BMDLio
values is presented in Table 5-15 (columns 2 and 3).
237
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-15. BMD values for incidence data for liver tumors (adenoma and
carcinoma) in male BDF1 mice (high dose dropped) and corresponding HEC and
HED values
Metric
(l)b
MCA
MRAMKL
BMD modeling"
Fisher
(2)
BMD10:0.19
(52.89)
BMDL10: 0.064
(18.35)
BMD10: 13.4
(33.52)
BMDL10:7.31
(17.29)
Thrall
(3)
BMD10: 0.39
(55.37)
BMDL10:0.12
(18.14)
BMD10: 14.2
(28.49)
BMDL10: 8.82
(16.66)
VMAXCH
(4)
0.40
0.65
1.49
1.70
0.40
0.65
1.49
1.70
HEC
Fisher
(5)
4.33
4.59
5.09
5.17
86.55
53.95
34.25
32.68
Thrall
(6)
8.26
8.74
9.72
9.88
116.95
67.89
41.70
39.72
HED
Fisher
(7)
0.68
0.83
1.33
1.46
4.10
2.56
1.62
1.55
Thrall
(8)
1.28
1.56
2.48
2.71
5.54
3.22
1.98
1.88
Mice were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, or 25 ppm (125 ppm exposure dropped)
HEC, human equivalent concentration, mg/m3; HED, human equivalent dose, mg/kg-day; Fisher, Fisher et al.
(2004) model; MCA, time-averaged arterial blood concentration, (imol/L; MRAMKL, time-averaged rate of
metabolism per kg liver, (imol/hr/kg liver; Thrall, Thrall et al. (2000) model; VMAXCH, maximum rate of
metabolism in humans, mg/hr/kg BW° 70
a MCA, multistage (3-stage); MRAMKL, multistage (3-stage)
BMR (benchmark response) = 10%. Values in parentheses are animal exposure concentrations (mg/m3)
corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
See Appendix E for the BMDS model outputs and graphs of the modeled data.
Female and male BDF1 mouse - pheochromocytomas
Internal doses associated with a BMR of 10% extra risk of pheochromocytomas were
calculated. BMD modeling in BMDS was performed using the female and male mouse
pheochromocytoma incidence data shown in Table 5-9 and internal doses shown in Table 5-10.
The multistage model was used to fit female mouse pheochromocytoma data. The multistage
model did not provide an adequate fit of the male mouse data for this tumor type; therefore, for
this data set, other models for dichotomous data in BMDS were run. The log-probit model
without restriction on the slope parameter provided the best fit of the male mouse
pheochromocytoma data (based on %2 p>0.\ and lowest AIC value). Bayesian analysis (see
Appendix E) confirmed BMDS results and provided an explanation as to why the slope
parameter of the log-probit model should not be constrained. Summaries of the resulting
and BMDLio values for the female and male mouse are presented in Table 5-16 (columns 2
and 3) and Table 5-17 (columns 2 and 3), respectively.
238
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-16. BMD values for incidence data for pheochromocytomas in female
BDF1 mice and corresponding HEC and HED values
Metric
(l)b
MCA
BMD modeling"
Fisher
(2)
BMD10: 1.43
(3529)
BMDL10: 1.14
(285.2)
Thrall
(3)
BMD10: 2.95
(353.3)
BMDL10: 2.34
(284.8)
VMAXCH
(4)
0.4
0.65
1.49
1.7
HEC
Fisher
(5)
74.551
78.636
88.173
89.826
Thrall
(6)
149.096
156.027
174.686
178.325
HED
Fisher
(7)
9.66
10.73
14.20
15.05
Thrall
(8)
18.54
19.90
24.34
25.44
Mice were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, 25, or 125 ppm
HEC, human equivalent concentration, mg/m3; HED, human equivalent dose, mg/kg-day; Fisher, Fisher et al.
(2004) model; MCA, time-averaged arterial blood concentration, (imol/L; Thrall, Thrall et al. (2000) model;
VMAXCH, maximum rate of metabolism in humans, mg/hr/kg B W°70
a Multistage (2-stage) model
BMR (benchmark response) = 10%. Values in parentheses are animal exposure concentrations (mg/m )
corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
Table 5-17. BMD values for incidence data for pheochromocytomas in male BDF1
mice and corresponding HEC and HED values
Metric
(l)b
MCA
BMD modeling"
Fisher
(2)
BMD10: 0.26
(71.36)
BMDL10:0.15
(42.13)
Thrall
(3)
BMD10: 0.53
(73.43)
BMDL10: 0.30
(43.38)
VMAXCH
(4)
0.40
0.65
1.49
1.70
HEC
Fisher
(5)
10.19
10.79
12.00
12.20
Thrall
(6)
19.96
21.13
23.56
23.95
HED
Fisher
(7)
1.56
1.91
3.04
3.33
Thrall
(8)
2.87
3.41
5.21
5.67
Mice were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, 25, or 125 ppm)
HEC, human equivalent concentration, mg/m3; HED, human equivalent dose, mg/kg-day; Fisher, Fisher et al.
(2004) model; MCA, time-averaged arterial blood concentration, (imol/L; Thrall, Thrall et al. (2000) model;
VMAXCH, maximum rate of metabolism in humans, mg/hr/kg B W°70
a log-probit model
BMR (benchmark response) = 10%. Values in parentheses are animal exposure concentrations (mg/m3)
corresponding to BMD and BMDL values.
b Number in parentheses indicates the column number.
See Appendix E for the BMDS model outputs and graphs of the modeled data.
239
DRAFT - DO NOT CITE OR QUOTE
-------
5.4.3.4. PBPKModeling of Human Equivalent Exposure Concentrations and Doses
Interspecies extrapolation (i.e., rat-to-human, mouse-to-human) and route-to-route
extrapolation of carbon tetrachloride inhalation dosimetry was accomplished using the human
PBPK model described in Paustenbach et al. (1988), Thrall et al. (2000), and Benson and
Springer (1999). The human PBPK model was used to estimate HECs (in mg/m3) or HEDs (i.e.,
daily ingested doses, in mg/kg-day) that would result in values for the internal dose metrics,
MCA or MRAMKL, equal to the respective BMDLs for each toxicity endpoint (i.e., liver tumors
in rats, liver tumors and adrenal pheochromocytomas in mice).
The approach used to derive the HECs and HEDs for each dose metric was as follows:
(1) The human PBPK model was used to calculate internal doses corresponding to a series of
exposure concentrations (EC, continuous exposure, mg/m3). For the dose metric MCA, the
human PBPK model was run at intervals over the range from 0.1 to 100 ppm (0.63-629 mg/m3);
for MRAMKL, the human PBPK model was run at intervals from 1 to 300 ppm (6.3-
1,887 mg/m3).
(2) For each of these internal doses, the human PBPK model was also used to calculate
equivalent rates of uptake of carbon tetrachloride from the GI tract to liver (RGIL, mg/kg-day)
that yielded the same internal doses. Values for uptake (RGIL) were used as estimates of HEDs
(mg/kg-day). This simple method of approximating the HED from the RGIL assumes that a
given ingestion dose of carbon tetrachloride (mg/kg-day) would result in the same dose delivered
from the GI tract to the liver and that the liver dose would be delivered at a constant rate during
the day (i.e., conceptually equivalent to, and simulated in the PBPK model as, a constant rate of
infusion of carbon tetrachloride into the liver). HED values derived from RGIL values are
approximations because they do not account for the possibility that bioavailability of ingested
carbon tetrachloride may be <100% or that the rate of absorption of ingested carbon tetrachloride
may not be constant throughout the day (i.e., bolus effects). Thus, this approximation method
assumes complete absorption from the GI tract. Information on bioavailability and absorption
kinetics of carbon tetrachloride in humans is not available. However, as discussed in Section
5.1.2, under certain dosing conditions (e.g., gavage in corn oil), absorption of carbon
tetrachloride from the GI tract may actually exhibit pulsatile behavior (Fisher et al., 2004;
Semino et al., 1997; Gallo et al., 1993).
(3) For each internal dose, conversion factors were calculated as the following corresponding
ratios:
• EC/MCA (to relate a continuous chronic human inhalation exposure in mg/m3 [EC] to
an internal dose using MCA as the dose metric);
• RGIL/MCA (to relate the rate of uptake of carbon tetrachloride from the
gastrointestinal tract to the liver (i.e., chronic daily ingested dose in mg/kg-day
[RGIL] to an internal dose using MCA as the dose metric);
240 DRAFT - DO NOT CITE OR QUOTE
-------
• EC/MRAMKL (to relate a continuous chronic human inhalation exposure in mg/m3
[EC] to an internal dose using MRAMKL as the dose metric); and
• RGIL/MRAMKL (to relate the rate of uptake of carbon tetrachloride from the
gastrointestinal tract to the liver in mg/kg-day [RGIL] to an internal dose using
MRAMKL as the dose metric).
(4) Conversion factors were calculated for each of four assumed values of Vmaxc in the human
PBPK model: 0.40, 0.65, 1.49, or 1.70 mg/hour/kg BW0'70. These conversion factors are
provided in Appendix C. Trend equations were also developed to permit the calculation of EC
or RGIL for any value of MCA or MRAMKL (see Appendix C).
Estimated values for inhalation HECs corresponding to BMDLs for the 2-year rat and
mouse inhalation bioassays (Nagano et al., 2007b; JBRC, 1998) for different tumor types and
alternative values of Vmaxc are presented in Tables 5-11 to 5-17, columns 5 and 6. Estimated
values for oral HEDs are presented in Tables 5-11 to 5-17, columns 7 and 8. As noted in the
discussion of the RfC derivation, estimates of the dose metrics, MCA and MRAMKL, were
sensitive to the value assigned to the Vmaxc parameter (see Figures 5-5 and 5-9), and the
inclusion of these alternative Vmaxc values provides some indication of the uncertainty in the
modeling. As in the derivation of the RfC, the human Vmaxc estimated from in vitro human data
(1.49 mg/hour/kg BW070) was considered to yield the most appropriate estimate of the HEC and
HED, and was used as the basis for cancer risk estimates. As discussed in Section 5.4.3.2, the
dose metric MRAMKL was considered to be the most appropriate dose metric to represent
internal doses in modeling liver tumors in rats and mice, and MCA was considered to be the
appropriate dose metric to represent internal doses in modeling pheochromocytoma incidence in
mice; these dose metrics were used as the basis for cancer risk estimates.
For the rat model, no information is available to establish whether a rat Vmaxc of 0.4 or
0.65 mg/hour/kg BW0'70 is the more scientifically defensible value for this parameter. Therefore,
the cancer risk values derived using these two rat Vmaxc values were averaged to derive the final
cancer risk values for carbon tetrachloride. Similarly, for the mouse, it cannot be established
whether the Fisher et al. (2004) or Thrall et al. (2000) model provides the more accurate
prediction of the internal dose for the mouse. Therefore, the cancer risk values derived using
these two mouse models were averaged to derive the final cancer risk values for carbon
tetrachloride (see Section 5.4.4 below).
5.4.4. Inhalation Unit Risk and Oral Slope Factor
5.4.4.1. Inhalation Unit Risk
IUR estimates based on the five tumor data sets analyzed in Section 5.4.3.3 were
calculated as follows:
241 DRAFT - DO NOT CITE OR QUOTE
-------
IUR = BMR / HEC
The lURs are provided in Table 5-18. The highest IUR was associated with
pheochromocytomas in the male mouse [5.6 x 10"6 (jig/m3)"1, or rounded to one significant
figure, 6 x 10"6 (jig/m3)"1]. Incidence of liver tumors was also increased in male mice. Because
different internal dose metrics were used in the dose-response analysis of liver tumors
(MRAMKL) and pheochromocytomas (MCA), the addition of individual tumor risks to obtain a
composite risk for the male mouse could not be performed. Uncertainty in the estimate of the
IUR associated with male mouse liver tumors also argues against risk addition. As noted in
Section 5.4.3.3, data from the male mouse provided a poor resolution of the dose-response
relationship for liver tumors. A marginal fit of this data set with the multistage model in BMDS
was obtained only when the highest dose group was dropped.
242 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-18. Summary of IUR estimates using linear low-dose extrapolation approach
Tumor
Female rat
hepatocellular
adenoma + carcinoma
Female mouse
hepatocellular
adenoma + carcinoma
Male mouse
hepatocellular
adenoma + carcinoma
Female mouse
pheochromocytoma
Male mouse
pheochromocytoma
Dose groups
modeled
0, 5, 25,
125 ppm
0, 5, 25 ppm
0, 5, 25 ppm
0, 5 ppm
0, 5, 25 ppm
0, 5, 25,
125 ppm
0, 5, 25,
125 ppm
Model
parameters
MRAMKL; VmaxR = 0.4
BMR = 5%
MRAMKL; VmaxR =
0.65
BMR = 5%
MRAMKL; VmaxR = 0.4
BMR = 5%
MRAMKL; VmaxR =
0.65
BMR = 5%
MRAMKL; Fisher
model
BMR =10%
MRAMKL; Thrall
model
BMR =10%
MRAMKL; Fisher
model
BMR =10%
MRAMKL; Thrall
model
BMR =10%
MRAMKL; Fisher
model
BMR =10%
MRAMKL; Thrall
model
BMR =10%
MCA; Fisher model
BMR =10%
MCA; Thrall model
BMR =10%
MCA; Fisher model
BMR =10%
MCA; Thrall model
BMR =10%
HEC
(mg/m3)
39.63
59.32
32.33
46.41
29.46
35.64
23.37
28.68
34.25
41.70
88.17
174.69
12.00
23.56
Average HEC
(mg/m3)3
49.48
39.37
32.55
26.03
37.98
131.4
17.78
IUR estimate11
(us/m3)-1
l.Ox 10'6
1.3 x lO'6
3.1 x 10"6
3.8 x lO'6
2.6 x lO'6
7.6 x 10'7
5.6 x 10'6
a For the rat, the average represents an arithmetic mean of the two HEC values based on VmaxR values of 0.65 and 0.4
mg/hr/kg B W°70; for the mouse, the average represents an arithmetic mean of the two HEC values based the Fisher
and Thall models.
b The IUR was calculated as the BMR - HEC.
Carbon tetrachloride also induced both liver tumors and pheochromocytomas in the
female mouse. For the same reason as the male mouse (i.e., different internal dose metrics were
used in the dose-response analysis), the risks associated with female liver tumors and
243
DRAFT - DO NOT CITE OR QUOTE
-------
pheochromocytomas could not be summed. To ensure that the composite tumor risk in the
female mouse did not exceed that associated with pheochromocytomas in the male mouse, a
bounding exercise was performed by summing the lURs for female mouse liver tumors and
pheochromocytomas [i.e., 3 x 10"6 + 8 x 10"7 (ug/m3)"1 = 4 x 10"6 (ug/m3)"1], a procedure that
results in an overestimation of composite risk. This bounding exercise confirms that the highest
value of the IUR is derived from male mouse pheochromocytoma data.
The IUR for carbon tetrachloride via the inhalation pathway is estimated as 6 x 10"6
(ug/m3)"1 based on pheochromocytomas in the male mouse. This data set was judged to be
applicable, scientifically sound, and yielded the highest estimate of risk. The slope of the linear
extrapolation from the central estimate based on pheochromocytomas in the male mouse is
calculated as 0.1 ^ (3.13 x 104 ug/m3) = 3.2 x 10"6 (ug/m3)"1, or rounded to one significant figure,
3 x 10"6 (ug/m3)-1.8
5.4.4.2. OralSF
Oral SF estimates based on the five inhalation tumor data sets analyzed in Section 5.4.3.3
and use of the human PBPK model of Paustenbach et al. (1988) and Thrall et al. (2000) to
r\
perform route-to-route extrapolation are provided in Table 5-19. The highest oral SF [6.5 x 10"
1 91
(mg/kg-day)- , or rounded to one significant figure, 7x10" (mg/kg-day)" ] was associated with
female mouse hepatocellular adenomas or carcinomas (using tumor data from the 0, 5, and 25-
ppm-exposure groups). An analysis of liver tumor data using only the 0- and 5-ppm groups
yielded a higher SF, but because it is based on only two data points and thus provides a less
informative characterization of the dose-response curve for female mouse liver tumors, the SF
based on analysis of data from the 0, 5, and 25-ppm groups is considered more appropriate. The
analysis based on tumor response data using only the 0- and 5-ppm groups was performed to
The slope of the linear extrapolation from the central estimate POD was calculated based on incidence data for
pheochromocytomas in male BDF1 mice. The central estimate POD (expressed as HEC) is:
Metric
MCA
BMCV
Fisher
0.26
Thrall
0.53
VMAXCH
1.49
HEC (central estimate)13
Fisher
20.7
Thrall
41.9
Average0
31.3
Mice were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, 25, or 125 ppm
HEC, human equivalent concentration, mg/m3; Fisher, Fisher et al. (2004) model; MCA, time-averaged arterial
blood concentration, (imol/L; Thrall, Thrall et al. (2000) model; VMAXCH, maximum rate of metabolism in
humans, mg/hr/kg BW°70
a Log-probit model
BMR (benchmark response) = 10%.
b The HEC values corresponding to BMCio values (as central estimates) were calculated using the values in Table
C-6 in Appendix C.
0 The average represents an arithmetic mean of the two HEC values based the Fisher and Thrall models.
244
DRAFT - DO NOT CITE OR QUOTE
-------
examine the effect on the liver cancer risk estimate of using only carbon tetrachloride response
data at exposure levels below those associated with evidence of cell replication. This analysis
reveals that dropping the 25-ppm group data had a relatively small impact on the SF (i.e.,
9 91
7 x 10" versus 8 x 10" [mg/kg-day]" ). A similar analysis of female rat liver tumor data
revealed a similarly negligible impact of performing a dose-response analysis on data points
99 1
below those associated with evidence of cell replication (i.e., 2 x 10" vs 3 x 10" [mg/kg-day]" ;
see Table 5-19).
245 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-19. Summary of oral SF estimates using linear low-dose extrapolation
approach and route-to-route extrapolation
Tumor
Female rat
hepatocellular
adenoma + carcinoma
Female mouse
hepatocellular
adenoma + carcinoma
Male mouse
hepatocellular
adenoma + carcinoma
Female mouse
pheochromocytoma
Male mouse
pheochromocytoma
Dose Groups
Modeled
0, 5, 25,
125 ppm
0, 5, 25 ppm
0, 5, 25 ppm
0, 5 ppm
0, 5, 25 ppm
0,5,25,
125 ppm
0, 5, 25,
125 ppm
Model
Parameters
MRAMKL; VmaxR = 0.4
BMR = 5%
MRAMKL; VmaxR = 0.65
BMR = 5%
MRAMKL; VmaxR = 0.4
BMR = 5%
MRAMKL; VmaxR = 0.65
BMR = 5%
MRAMKL; Fisher model
BMR =10%
MRAMKL; Thrall model
BMR =10%
MRAMKL; Fisher model
BMR =10%
MRAMKL; Thrall model
BMR =10%
MRAMKL; Fisher model
BMR =10%
MRAMKL; Thrall model
BMR =10%
MCA; Fisher model
BMR =10%
MCA; Thrall model
BMR =10%
MCA; Fisher model
BMR =10%
MCA; Thrall model
BMR =10%
HED
(mg/kg-d)
1.88
2.81
1.53
2.20
1.40
1.69
1.11
1.36
1.62
1.98
14.2
24.34
3.04
5.21
Average
HED
(mg/kg-d)a
2.34
1.86
1.54
1.24
1.8
19.27
4.12
Oral SF
Estimate
(mg/kg-day)1
2.1 x lO'2
2.7 x lO'2
6.5 x lO'2
8.1 x lO'2
5.6 x lO'2
5.2 x lO'3
2.4 x lO'2
a For the rat, the average represents an arithmetic mean of the two HEC values based on VmaxR values of 0.65 and
0.4 mg/hr/kg BW070; for the mouse, the average represents an arithmetic mean of the two HEC values based the
Fisher and Thall models.
b The oral SF was calculated as the BMR - HED.
Carbon tetrachloride also induced pheochromocytomas in the female mouse. For the
same reason provided for the male mouse tumor data used to derive the IUR, the estimated risks
from the individual tumors could not be summed because different internal dose metrics were
used in the dose-response/PBPK analysis. Because the SF associated with pheochromocytomas
is an order of magnitude smaller than the SF associated with liver tumors in the female mouse,
the pheochromocytoma data would be expected to contribute negligibly to the total cancer risk
estimate.
246
DRAFT - DO NOT CITE OR QUOTE
-------
The oral SF for carbon tetrachloride is estimated as 7 x 10"2 (mg/kg-day)"1 based on
female mouse liver tumors. This data set was judged to be applicable, scientifically sound, and
yielded the highest estimate of risk. The slope of the linear extrapolation from the central
estimate based on female mouse liver tumors is calculated as 0.1 ^ 2.27 mg/kg-day = 4.4 x 10"2
(mg/kg-day)"1, or rounded to one significant figure, 4 x 10"2 (mg/kg/day)"1.9
It is noted that whereas the male mouse pheochromocytoma data set yielded the highest
estimate of the IUR for carbon tetrachloride based on analysis of tumor data from the JBRC
bioassay (Nagano et al., 2007b) and application of PBPK modeling for interspecies
extrapolation, female mouse liver tumor data yielded the highest estimate of the oral SF based on
the data from the same bioassay and route-to-route extrapolation using PBPK modeling. While
it may appear counterintuitive that the use of data from a single inhalation bioassay (Nagano et
al., 2007b) could result in the use of different data sets for estimating cancer potency by the oral
and inhalation routes, the situation arises because of the use in PBPK modeling of different dose
metrics for the liver and adrenal gland that could result in different relationships between
environmental exposure and internal dose within a species (i.e., rat in the current bioassay) and
across species (i.e., rats and humans).
5.4.5. Nonlinear Extrapolation Approach
As noted above, empirical evidence for carbon tetrachloride, particularly from studies
using relatively high exposure levels, supports a MOA for liver tumors that includes the
following hypothesized key events: (1) metabolism to the trichloromethyl radical by CYP2E1
and subsequent formation of the trichloromethyl peroxy radical, (2) radical-induced mechanisms
The slope of the linear extrapolation from the central estimate POD was calculated based on incidence data for
liver tumors in female BDF1 mice. The central estimate POD (expressed as HED) is:
Metric
MRAMKL
BMD10a
Fisher
9.71
Thrall
10.4
VMAXCH
1.49
HED (central estimate)13
Fisher
2.19
Thrall
2.35
Average0
2.27
Mice were exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week). Doses modeled
correspond to exposure concentrations: 0, 5, or 25 ppm (125 ppm exposure dropped)
HED, human equivalent dose, mg/kg-day; Fisher, Fisher et al. (2004) model; MRAMKL, time-averaged rate of
metabolism per kg liver, (imol/hr/kg liver; Thrall, Thrall et al. (2000) model; VMAXCH, maximum rate of
metabolism in humans, mg/hr/kg BW°70
a multistage model
BMR (benchmark response) = 10%.
b The HED values corresponding to BMDio values (as central estimates) were calculated using the values in Table
C-10 in Appendix C.
0 The average represents an arithmetic mean of the two HED values based the Fisher and Thall models.
247
DRAFT - DO NOT CITE OR QUOTE
-------
leading to hepatocellular toxicity, and (3) sustained regenerative and proliferative changes in the
liver in response to hepatotoxicity. These postulated key events are consistent with a hypothesis
that liver carcinogenicity occurs at exposures that also induce hepatocellular toxicity and a
sustained regenerative and proliferative response, and that exposures that do not cause
hepatotoxicity are not expected to result in liver cancer. For this hypothesized MOA for carbon
tetrachloride liver carcinogenicity, a nonlinear approach to low-dose extrapolation may be
considered appropriate.
The RfD of 0.004 mg/kg-day and RfC of 0.1 mg/m3 derived in Sections 5.1 and 5.2
represent the outcome of nonlinear assessments based on hepatotoxicity associated with oral
exposures (RfD) and inhalation exposures (RfC) to carbon tetrachloride. Doses (or
concentrations) of carbon tetrachloride below the RfD (or RfC) that do not cause sustained
hepatocellular cytotoxicity and regenerative cell proliferation would be expected to be protective
of liver tumors if this is the primary MOA for liver tumors. This harmonized approach between
noncancer and cancer endpoints transparently utilizes a key event (cytotoxicity or hepatotoxicity)
in the hypothesized nonlinear MOA to derive the RfD and RfC. Based on the MOA consistent
with nonlinearity, the RfD of 0.004 mg/kg-day and RfC of 0.1 mg/m3 can be used to assess the
potential risk of liver cancer from carbon tetrachloride exposure.
The application of a nonlinear approach for liver tumors is based on MOA information
specific to that tumor type and does not apply to the occurrence of liver tumors at non-cytotoxic
doses (as was seen in female mice in the JBRC study), nor does it apply to pheochromocytomas.
As noted above, pheochromocytomas were reported in mice in the JBRC 104-week bioassay
(Nagano et al., 2007b; JBRC, 1998). Unlike liver tumors associated with carbon tetrachloride
exposure, which have been observed in numerous bioassays in multiple species and by multiple
routes of exposure, pheochromocytomas have been observed only in the mouse. Nevertheless,
the RfD and RfC based on liver toxicity cannot be assumed to be protective for the potential
cancer risk associated with carbon tetrachloride-induced pheochromocytomas.
5.4.6. Uncertainties in Cancer Risk Values
As in most risk assessments, extrapolation of the available experimental data for carbon
tetrachloride to estimate potential cancer risk in human populations introduces uncertainty in the
risk estimation. Several types of uncertainty may be considered quantitatively, whereas others
can only be addressed qualitatively. Thus, an overall integrated quantitative uncertainty analysis
cannot be developed. Major sources of uncertainty in the cancer assessment for carbon
tetrachloride are summarized in this section and in Table 5-20 at the end of this section.
Relevance to humans. As noted in EPA's 2005 Guidelines for Carcinogen Risk
Assessment, "... agents observed to produce tumors in both humans and animals have produced
248 DRAFT - DO NOT CITE OR QUOTE
-------
tumors either at the same site (e.g., vinyl chloride) or different sites (e.g., benzene) (NRC, 1994).
Hence, site concordance is not always assumed between animals and humans." Thus, it is not
clear whether the tumors observed in rodent bioassays would be predictive of human tumors of
the same or different sites.
The MOA for liver tumor induction has not been established, but the hypothesized MO As
that have been investigated are assumed to be relevant to humans (Section 4.7.3.5). There is no
available evidence in humans for hepatic cancer associated with carbon tetrachloride exposure.
The experimental animal literature, however, shows carbon tetrachloride to consistently induce
liver tumors across species and routes of exposure. Further, there are similarities between
experimental animals and humans in terms of carbon tetrachloride metabolism, antioxidant
systems, and evidence for the liver as a sensitive target organ. Together, this supports a
conclusion that experimental evidence for liver cancer is relevant to humans.
Pheochromocytomas, on the other hand, were observed in only one species (the mouse).
In humans, pheochromocytomas are rare catecholamine-producing neuroendocrine tumors that
are usually benign, but may also present as or develop into a malignancy (Eisenhofer et al., 2004;
Salmenkivi et al., 2004; Tischler et al., 1996). In humans, hereditary factors have been identified
as important in the development of pheochromocytomas (Eisenhofer et al., 2004). In the mouse,
few chemicals have been reported to cause adrenal medullary tumors (Hill et al., 2003), and the
MOA for this tumor in mice is unknown. The relevance of mouse pheochromocytomas to
humans is similarly unknown, although parallels between this tumor in the mouse and human led
investigators to concluded that the mouse might be an appropriate model for human adrenal
medullary tumors (Tischler et al., 1996). Like the human, pheochromocytomas in the mouse are
relatively rare, as are metastases. Both the morphological variability of the mouse
pheochromocytomas and the morphology of the predominant cells are comparable to those of
human pheochromocytomas. An important characteristic of mouse pheochromocytomas is
expression of immunoreactive phenylethanolamine-N-methyltransferase (PNMT); human
pheochromocytomas are also usually PNMT-positive (Tischler et al., 1996). Overall, this
supports a conclusion that experimental evidence for pheochromocytomas is potentially relevant
to humans.
Choice of low-dose extrapolation approach. The MOA is a key determinant of which
approach to apply for estimating low-dose cancer risk. The MOA of carbon tetrachloride liver
carcinogenicity has been investigated extensively; however, much of this research has been
conducted at relatively high exposure levels. The MOA(s) at low exposure levels is not known.
For liver tumors, a nonlinear extrapolation approach was explored in Section 5.4.5 as an
alternative to the linear low-dose extrapolation approach for cancer risk estimation. Such an
249 DRAFT - DO NOT CITE OR QUOTE
-------
approach would be supported by a conclusion that the hypothesized cytotoxicity-proliferative
MOA is operative at all doses; however, bioassay results inconsistent with the nonlinear
approach include evidence of female mouse hepatocarcinongenicity at non-cytotoxic doses,
potential for genotoxicity at low doses, and the absence of MOA information regarding the
observed pheochromocytomas in mice.
The linear extrapolation approach assumes that some cancer risk exists at all non-zero
exposures, and that this risk increases linearly with exposure. While consistent with the
recognized biological reactivity of carbon tetrachloride, uncertainties in this low-dose
extrapolation approach are associated with the lack of MOA information at low exposures.
Additional MOA information in the low-dose region to establish whether a linear or nonlinear
approach applies to carbon tetrachloride liver tumors would significantly reduce the uncertainty
associated with estimating the magnitude of liver tumor risk.
The effect on risk estimates derived using a linear extrapolation approach of using only
data on carbon tetrachloride liver tumor response at levels below those associated with increased
cell replication was examined. The risk calculations did not prove particularly sensitive to the
limitation of data points to those below which increased cell replication was reported (see Tables
5-18 and 5-19). This consistency in cancer risk estimates provided some confidence that the IUR
and SF estimates based on liver tumor data are not driven by high doses associated with
significant hepatotoxicity.
In data sets where early mortality is observed, methods that can reflect the influence of
competing risks and intercurrent mortality on site-specific tumor incidence rates are preferred.
Survival curves for female rats and mice from the JBRC bioassay (see Figures 4-1 and 4-2) show
early mortality in some treated groups. Because liver tumors were the primary cause of early
deaths in these groups, failure to apply a time to tumor analysis is not likely to significantly
influence the IUR for liver tumors. The impact on the IUR for pheochromocytomas is unknown.
Under the linear low-dose extrapolation approach, cancer risk estimates were calculated
by straight line extrapolation from the POD to zero, with the multistage model used to derive the
POD. (The one exception is the male mouse pheochromocytoma data set, where the log-probit
model was used.) It is unknown how well this extrapolation procedure predicts low-dose risks
for carbon tetrachloride. The multistage model does not represent all possible models one might
fit, and other models could conceivably be selected to yield different results consistent with the
observed data, both higher and lower than those included in this assessment.
For pheochromocytomas, only a linear low-dose extrapolation approach was used to
estimate human carcinogenic risk in the absence of any information on the MOA for this tumor.
MOA information to inform the approach to low-dose extrapolation for carbon tetrachloride-
250 DRAFT - DO NOT CITE OR QUOTE
-------
induced pheochromocytomas would significantly reduce the uncertainty associated with the
magnitude of risk from exposure to this tumor type.
Cancer risk estimates for liver tumors and pheochromocytomas developed using a linear
low-dose extrapolation approach were not combined because different dose metrics were used in
the dose-response/PBPK analysis of these two tumor types. Deriving the IUR or oral SF for data
on one tumor site, however, may underestimate the carcinogenic potential of carbon
tetrachloride. For the IUR based on male mouse pheochromocytomas, because of the poor
resolution of the dose-response relationship for male rodent liver tumors, the magnitude of the
potential risk underestimation cannot be characterized. Because the SF based on female mouse
liver tumors was an order of magnitude greater than that for female mouse pheochromocytomas,
any underestimation of the SF is expected to be small.
Interspecies extrapolation. Extrapolating dose-response data from animals to humans
was accomplished using PBPK models in the rat, mouse, and human. Availability of a PBPK
model generally reduces the pharmacokinetic component of uncertainty associated with animal
to human extrapolation; however, any PBPK model has its own associated uncertainties.
Specific uncertainties in the PBPK modeling for carbon tetrachloride were discussed previously
in Section 5.3.
Route-to-route extrapolation for the oral SF. Studies of carbon tetrachloride
carcinogenicity by the oral route were determined to be insufficient to derive a quantitative
estimate of cancer risk. Therefore, a human PBPK model was used to extrapolate inhalation data
to the oral route. A simple approximation method was used that assumed continuous infusion of
carbon tetrachloride from the human GI tract to the liver and that absorption of carbon
tetrachloride from the GI tract is essentially complete. Doses extrapolated from inhalation to
oral exposures in this analysis were approximations because they did not account for oral
bioavailability or absorption kinetics, information that is not available for carbon tetrachloride.
To the extent that GI absorption is less than 100%, the current estimation method for route-to-
route extrapolation would tend to overestimate the SF.
Statistical uncertainty at the POD. Parameter uncertainty can be assessed through
confidence intervals. Each description of parameter uncertainty assumes that the underlying
model and associated assumptions are valid. For the log-probit model applied to the male mouse
pheochromocytoma data, there is a reasonably small degree of uncertainty at the 10% excess
incidence level (the POD for linear low-dose extrapolation); the lower bound on the BMD (i.e.,
the BMDLio) is 1.8-fold lower than the BMD. For the multistage model applied to the female
mouse liver tumor data, there is similarly a reasonably small degree of uncertainty at the 10%
251 DRAFT - DO NOT CITE OR QUOTE
-------
excess incidence level; the lower bound on the BMD (i.e., the BMDLio) is approximately 1.5-
fold lower than the BMD.
Bioassay selection. The study by Nagano et al. (2007b; also reported as JBRC, 1998)
was used for development of the IUR. A full report of the bioassay findings was published in
2007, although the study itself was conducted in the mid-1980s. Although not a recently
conducted study, this bioassay was well-designed, using two species (rats and mice), four dose
groups, including an appropriate untreated control, and 50 animals/sex/group. Examination of
toxicological endpoints in both sexes of rats and mice was appropriate. No issues were identified
with this bioassay that might have contributed to uncertainty in the cancer assessment.
Alternative bioassays for developing an IUR were unavailable.
Choice of species/gender. For liver tumors, modeling was performed using the JBRC
inhalation bioassay data for the female mouse and female rat. The male rat liver tumor data were
not modeled because these data sets lacked the resolution desired for dose-response modeling;
the male mouse liver data were modeled, but provided similarly poor dose-response curve
resolution. Tumor frequencies increased from control levels to close to maximal responses
without any intervening dose levels having submaximal responses. In the female mice and rats,
lower but biologically significant levels of tumor response were seen at intermediate dose levels.
Also, notably, increased levels of hepatocellular proliferation were not reported for rodents at
these intermediate levels, so that dose-response modeling based on these data may be more
applicable to an evaluation of cancer risk at noncytotoxic exposures. There is no indication that
male rodents are more sensitive to carbon tetrachloride liver tumor induction and that use of
female data only underestimated potential risk. For pheochromocytomas, JBRC inhalation data
sets for both male and female mice were amenable to modeling, and the data set yielding the
highest estimate of cancer risk could be selected.
Human population variability. Neither the extent of interindividual variability in carbon
tetrachloride metabolism nor human variability in response to carbon tetrachloride has been fully
characterized. Factors that could contribute to a range of human response to carbon tetrachloride
include variations in CYP450 levels because of age-related differences or other factors (e.g.,
exposure to other chemicals that induce or inhibit microsomal enzymes), genetic polymorphisms
in drug metabolism enzymes, transporters, and receptors (all of which can markedly affect
susceptibility to a toxic chemical), nutritional status, alcohol consumption, or the presence of
underlying disease that could alter metabolism of carbon tetrachloride or antioxidant protection
systems. Incomplete understanding of the potential differences in metabolism and susceptibility
252 DRAFT - DO NOT CITE OR QUOTE
-------
across exposed human populations represents a source of uncertainty.
Table 5-20. Summary of uncertainty in the carbon tetrachloride cancer risk
assessment
Consideration/
approach
Impact on cancer
risk estimate
Decision
Justification
Human relevance
of rodent tumor
data
If rodent tumors
proved not to be
relevant to humans,
unit risk would not
apply, i.e., human
risk would J,
Liver tumors in rats
and mice and
pheochromocytomas
in mice are relevant
to human exposure
As noted in EPA's 2005 Guidelines for Carcinogen
Risk Assessment, "... site concordance is not always
assumed between animals and humans." Thus, it is
not clear whether the tumors observed in rodent
bioassays would be predictive of human tumors of
the same or different sites.
Liver: The experimental animal literature shows
carbon tetrachloride to consistently induce liver
tumors across species and routes of exposure.
Although there is no evidence in humans for
hepatic cancer associated with carbon tetrachloride
exposure, the hypothesized MOAs are considered
relevant to humans. Experimental animals and
humans are similar in terms of carbon tetrachloride
metabolism, antioxidant systems, and evidence for
the liver as a sensitive target organ. Together, this
evidence supports a conclusion that experimental
evidence for liver cancer is relevant to humans.
Pheochromocytomas: Pheochromocytomas were
observed in the mouse only. In humans,
pheochromocytomas are rare catecholamine-
producing neuroendocrine tumors that are usually
benign, but may also present as or develop into a
malignancy. Hereditary factors have been
identified as important in pheochromocytoma
development. The mouse has been characterized as
possibly an appropriate model for human adrenal
medullary tumors.
Low-dose
extrapolation
approach
Departure from
EPA's Guidelines
for Carcinogen Risk
Assessment POD
paradigm, if
justified, could J, or
| unit risk an
unknown extent
Liver: Nonlinear
approach and linear
approach presented.
Under the linear
extrapolation
approach, a POD-
based straight-line
extrapolation was
applied
Pheochromocytoma:
Linear approach,
using a POD-based
straight-line
extrapolation
Liver: Biological support is available for a
cytotoxic-proliferative MOA that is consistent with
a nonlinear extrapolation approach; however, other
evidence suggests that hepatocarcinogenicity may
not be explained only in terms of this hypothesized
MOA. Where data are not strong enough to
ascertain the MOA, EPA's 2005 Guidelines for
Carcinogen Risk Assessment recommend
application of a linear low-dose extrapolation
approach in addition to a nonlinear approach.
Pheochromocytoma: Application of a linear
approach where the MOA has not been established
is consistent with EPA's 2005 Guidelines for
Carcinogen Risk Assessment.
253
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-20. Summary of uncertainty in the carbon tetrachloride cancer risk
assessment
Consideration/
approach
Interspecies
extrapolation
using PBPK
model
Route-to-route
extrapolation
using PBPK
model
Statistical
uncertainty at
POD
Bioassay
Species/gender
combination
Human
population
variability in
metabolism and
response/
sensitive
subpopulations
Impact on cancer
risk estimate
|IUR
The magnitude of
uncertainty cannot
be quantified;
however,
assumption of
complete GI
absorption may
overestimate the SF.
| IUR and SF by
1.5-1. 8-fold if
BMD used as the
POD rather than
lower bound on
POD
Alternative
bioassay, if
available, could | or
J, SF by an unknown
extent
Human risk could |
or J,, depending on
relative sensitivity
Low-dose risk could
t or I to an
unknown extent
Decision
PBPK modeling used
to extrapolate rodent
tumor data to humans
A human PBPK
model was used to
extrapolate inhalation
data to the oral route
BMDL (preferred
approach for
calculating
reasonable upper
bound SF)
JBRC bioassay
Female mouse and
rat liver tumors
Male and female
mouse pheo-
chromocytomas
Considered
qualitatively
Justification
PBPK modeling is considered to reduce the
uncertainty in extrapolating rodent tumor data to
humans.
Studies of carbon tetrachloride carcinogenicity by
the oral route were determined insufficient to derive
a quantitative estimate of cancer risk. A simple
approximation method was used that assumed
continuous infusion of carbon tetrachloride from
the human GI tract to the liver.
Size of bioassay results in sampling variability;
lower bound is 95% confidence interval on
administered exposure.
Alternative bioassays were unavailable.
It was assumed that humans are as sensitive as the
most sensitive rodent gender/species tested; true
correspondence is unknown.
For liver tumors, female mouse and female rat data
from the JBRC bioassay were considered more
amenable for modeling and demonstrating a
response that may be more relevant to lower dose
conditions than males. For pheochromocytomas,
JBRC inhalation data sets for both male and female
mice were amenable to modeling, and the data set
yielding the highest estimate of cancer risk could be
selected.
No data to support range of human
variability /sensitivity. Factors that could contribute
to a range of human response to carbon
tetrachloride include variations in CYP450 levels,
genetic polymorphisms, nutritional status, alcohol
consumption, or the presence of underlying disease
that could alter metabolism of carbon tetrachloride
or antioxidant protection systems. On balance,
available data do not indicate that children would
necessarily be more sensitive.
254
DRAFT - DO NOT CITE OR QUOTE
-------
5.4.7. Previous Cancer Assessment
The previous cancer assessment for carbon tetrachloride was posted on the IRIS database
in 1987. At that time, carbon tetrachloride was classified as a B2 carcinogen (probable human
carcinogen), based on the finding of treatment-related hepatocellular carcinomas in rats, mice
and hamsters. In the previous assessment, an oral SF of 1.3 x 10"1 (mg/kg-day)"1 was derived
using linear extrapolation procedures and liver tumor data sets from the hamster (Delia Porta et
al., 1961), mouse (Edwards et al., 1942; NCI, 1977, 1976a, b), and rat (NCI, 1977, 1976a, b). In
the current assessment, the available oral bioassay data were not considered adequate for dose-
response analysis, and a SF was derived instead by application of a PBPK model to extrapolate
inhalation bioassay data to the oral route. The resulting SF [7 x 10"2 (mg/kg-day)"1] is
approximately twofold smaller than the previous SF.
An IUR of 1.5 x 10"5 (jig/m3)"1 was derived previously from the oral SF by route-to-route
extrapolation (assuming an air intake of 20 m3/day, body weight of 70 kg, and 40% absorption
rate by humans). The current IUR [6 x 10"6 (jig/m3)"1] was derived using a chronic inhalation
bioassay (Nagano et al., 2007b) that was not available at the time of the previous assessment and
PBPK modeling for interspecies extrapolation.
255 DRAFT - DO NOT CITE OR QUOTE
-------
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE
6.1. HUMAN HAZARD POTENTIAL
Carbon tetrachloride is rapidly absorbed by any route of exposure. Once absorbed, it is
widely distributed among tissues, especially those with high lipid content, reaching peak
concentrations in
-------
hepatocellular cytotoxicity and regenerative hyperplasia and the induction of liver tumors is
inconsistent. In particular, an increased incidence of hepatocellular adenomas in the low-dose
(5-ppm) female mouse in the absence of nonneoplastic liver toxicity (Nagano et al., 2007b;
JBRC, 1998) suggests that mouse hepatocarcinogenicity cannot simply be explained in terms of
the hypothesized cytotoxic-proliferative MOA. Studies of genotoxic and mutagenic potential are
largely negative. There is little direct evidence that carbon tetrachloride induces intragenic or
point mutations in mammalian systems. Mutagenicity studies performed using transgenic mice
have yielded negative results, as have the vast majority of the mutagenesis studies that have been
conducted in bacterial systems. Under highly cytotoxic conditions, bioactivated carbon
tetrachloride can exert genotoxic effects. These tend to be modest in magnitude and are
manifested primarily as DNA breakage and related sequelae. Chromosome loss leading to
aneuploidy may also occur to a limited extent. The fact that carbon tetrachloride overall has not
been found to be a potent mutagen and that positive genotoxic results are found at high exposure
levels and generally in concert with cytotoxic effects indicates that carbon tetrachloride does not
likely induce genotoxic effects through direct binding or damage to DNA. The nature of the
genotoxicity database, however, poses distinct challenges to the evaluation of carbon
tetrachloride genotoxicity, particularly at low exposure levels. Information on the biological
activity of carbon tetrachloride at low exposures is far less complete than at higher (cytotoxic)
exposure levels. Considerable evidence points to the involvement of reactive metabolites and
reaction products of carbon tetrachloride with cellular constituents in the induction of liver
toxicity and carcinogenicity by carbon tetrachloride. In light of the fundamental reactivity of
products of carbon tetrachloride metabolism, uncertainties about genotoxic activity at low
exposures, and empirical data from rodent bioassays that suggest that mouse
hepatocarcinogenicity cannot be explained in terms of a cytotoxic-proliferative MOA alone, the
MOA(s) for carbon tetrachloride-induced liver tumors at low exposure levels is/are unknown.
The MOA for pheochromocytomas induced by carbon tetrachloride is unknown. Under the
Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), carbon tetrachloride is
characterized as likely to be carcinogenic to humans by all routes of exposure.
6.2. DOSE RESPONSE
6.2.1. Noncancer - Oral Exposure
The most sensitive endpoints identified for effects of carbon tetrachloride by oral
exposure relate to liver toxicity in the subchronic corn oil gavage studies of Bruckner et al.
(1986) in rats and Condie et al. (1986) in mice. The Bruckner et al. (1986) study identified
serum enzyme changes and liver histopathology as the most sensitive endpoints for carbon
tetrachloride. Serum SDH was the most sensitive serum chemistry endpoint and was considered
a marker of histopathologic changes. Another target of carbon tetrachloride toxicity following
oral exposure considered in the selection of the critical effect was the developing organism.
257 DRAFT - DO NOT CITE OR QUOTE
-------
Studies in experimental animals found that relatively high doses of carbon tetrachloride during
gestation can produce prenatal loss; these doses also produced overt toxic effects in the dams.
Carbon tetrachloride doses associated with liver toxicity were lower than those associated with
developmental toxicity.
BMD modeling methods were used to calculate the POD for deriving the RfD by
estimating the effective dose at a specified level of response (BMDX) and its 95% lower bound
(BMDLX) for liver enzyme changes. An increase in SDH activity 2 times the control mean was
used as the BMR. All of the models for continuous data in U.S. EPA's BMDS (version 1.4.1)
(U.S. EPA, 2007b) were fit to the 10- and 12-week SDH data. None of the models for
continuous data in BMDS provided an adequate fit of the 12-week SDH data. The power model,
which provided the best fit to the data, estimated a BMD2X of 7.32 mg/kg-day and a BMDL2X of
5.46 mg/kg-day.
Liver lesion incidence data from the Bruckner et al. (1986) study in rats and the Condie et
al. (1986) study in mice do not provide adequate information in the response region of concern
(i.e., 10% increase in extra risk over controls) to allow for BMD modeling of these endpoints
(U.S. EPA, 2000c). The NOAEL of 1 and LOAEL of 10-12 mg/kg-day in these studies do,
however, support the BMD2X of 7.32 mg/kg-day and the BMDL2X of 5.46 mg/kg-day estimated
from the increase in serum SDH observed in the Bruckner et al. (1986) study.
The BMDL2X of 5.46 mg/kg estimated from the increase in serum SDH activity in rats in
the Bruckner et al. (1986) subchronic toxicity study was used as the POD for derivation of the
RfD. Use of the modeled BMDL provides an inherent advantage over use of a NOAEL or
LOAEL by making greater use of the available data. Because of the absence of a suitable PBPK
model for oral exposure to carbon tetrachloride, one was not used for this assessment. Because
the BMDL2X of 5.46 mg/kg was derived from a study (Bruckner et al., 1986) with an intermittent
dosing schedule, it was adjusted to an average daily dose prior to application of UFs (BMDL2x-
ADJ = 3.9 mg/kg-day). Applying a composite UF of 1,000 to the BMDLADJ of 3.9 mg/kg-day
yields an RfD of 0.004 mg/kg-day for carbon tetrachloride. The composite UF of 1,000 includes
a factor of 10 to protect susceptible individuals, a factor of 10 to extrapolate from rats to humans,
a factor of 3 (10°5) to extrapolate from a subchronic to a chronic duration of exposure, and a
factor of 3 (10°5) to account for database deficiencies. Information was unavailable to
quantitatively assess toxicokinetic or toxicodynamic differences between animals and humans
and the potential variability in human susceptibility (factors that could contribute to a range of
human response include variations in CYP450 levels, nutritional status, alcohol consumption, or
the presence of underlying disease); thus, the UF selected for uncertainties related to both
interspecies and intraspecies was the default of 10. A UF of 3 for subchronic to chronic
extrapolation was selected based on: (1) qualitative information demonstrating that the target of
toxicity following chronic oral exposure as the liver; (2) knowledge of the relationship between
effect levels in subchronic and chronic inhalation studies; and (3) early onset of liver toxicity. A
258 DRAFT - DO NOT CITE OR QUOTE
-------
database UF of 3 was selected to account for a lack of an adequate multigeneration study of
reproductive function.
To provide perspective on the RfD supported by Bruckner et al. (1986), PODs and oral
RfDs based on other selected studies of carbon tetrachloride oral toxicity are arrayed in Figures
5-1 to 5-3. The predominant noncancer effect of subchronic and chronic oral exposure to carbon
tetrachloride is hepatic toxicity. Figure 5-1 provides a graphical display of five studies that
reported liver toxicity in experimental animals following subchronic oral exposure, including the
PODs, applied UFs, and potential RfDs for comparison to the RfD derived from the Bruckner et
al. study. Studies in experimental animals have also reported developmental toxicity (prenatal
loss) at relatively high doses of carbon tetrachloride during gestation. A graphical display of
information from three developmental studies is provided in Figure 5-2. Figure 5-3 displays
PODs for the major targets of toxicity associated with oral exposure to carbon tetrachloride. For
the reasons discussed in Section 5.1.2, liver effects in the rat observed in the study by Bruckner
et al. (1986) are considered the most appropriate basis for the carbon tetrachloride RfD. The text
of Sections 5.1.1 and 5.1.2 should be consulted for a more complete understanding of the issues
associated with each data set and the rationale for the selection of the critical effect and principal
study used to derive the RfD.
Confidence in the principal study, Bruckner et al. (1986), is medium. The 12-week
gavage study is a well-conducted, peer-reviewed study that used three dose groups plus a control
and collected interim data at 2-week intervals. The study is limited by relatively small group
sizes (5 to 9 rats/group) and investigation of only two target organs (liver and kidney).
Confidence in the oral database is medium. Two chronic oral animal studies were designed as
cancer bioassays, and one of the two included only limited investigation of noncancer endpoints.
The second chronic bioassay by NCI provided complete nonneoplastic incidence data; however,
because of the marked hepatotoxicity in dosed rats at the lowest dose tested and the low survival
in dosed mice as a result of the high incidence of liver tumors, the bioassay was not suitable for
dose-response analysis. The toxicity of carbon tetrachloride has been more thoroughly
investigated in a number of oral toxicity studies of subchronic duration, and a number of tests of
immunotoxic potential are available. The oral database lacks an adequate multigeneration study
of reproductive function. Testing for developmental toxicity has been performed in only one
species. Overall confidence in the RfD is medium.
6.2.2. Noncancer - Inhalation Exposure
The most sensitive endpoint identified for effects of carbon tetrachloride by inhalation
exposure was liver toxicity in the chronic rat study by JBRC (Nagano et al., 2007b; JBRC,
1998), manifested at an exposure concentration of 25 ppm by elevated serum enzymes, fatty
change, fibrosis, and cirrhosis. Other targets of carbon tetrachloride toxicity considered in the
selection of the critical effect included the kidney, the adrenal gland, and the developing
259 DRAFT - DO NOT CITE OR QUOTE
-------
organism.
PBPK and BMD modeling methods were used to calculate the POD for deriving the RfC.
Exposure levels studied in the 2-year JBRC rat bioassay were converted to estimates of internal
dose metrics by application of PBPK models (Thrall et al., 2000; Benson and Springer, 1999;
Paustenbach et al., 1988); rate of carbon tetrachloride metabolism in the liver was considered the
most appropriate dose metric for liver toxicity. BMD modeling methodology (U.S. EPA, 2000c,
1995) was used to analyze the relationship between the estimated internal doses and response
(i.e., fatty change of the liver) by estimating the effective dose at a specified level of response
(BMDX) and its BMDLX. A 10% extra risk of fatty changes of the liver was used as the BMR.
All of the models for dichotomous data in U.S. EPA's BMDS (version 1.4.1) (U.S. EPA, 2007b)
were fit to the incidence data for fatty liver in male and female rats. In the male rat, the logistic
model provided the best fit of the data. For female rats, no models provided an adequate fit to
the data when all dose groups were included, as assessed by the j^ goodness-of-fit test. After
dropping the highest dose, the multistage model provided the best fit of the data. The resulting
BMDLio values (expressed as internal doses) were converted to estimates of equivalent human
exposure concentrations (HECs) by applying a human PBPK model and assuming a value for the
human Vmaxc estimated from in vitro human data. An HEC of 14.3 mg/m3 is used as the POD
for RfC derivation. An RfC of 0.1 mg/m3 for carbon tetrachloride is derived by applying a
composite UF of 100 to the FIEC of 14.3 mg/m3. The composite UF of 100 includes a factor of
10 to protect susceptible individuals, a factor of 3 (or 10°5) to extrapolate from rats to humans,
and a factor of 3 (or 10°5) to account for an incomplete database. Information was unavailable to
quantitatively assess the potential variability in human susceptibility (factors that could
contribute to a range of human response include variations in CYP450 levels, nutritional status,
alcohol consumption, or the presence of underlying disease); thus, a default UF of 10 was
selected to account for the uncertainty in intraspecies variability. A pharmacokinetic model was
used to adjust for pharmacokinetic differences across species. A UF of 3 was selected for
interspecies extrapolation to account for potential pharmacodynamic differences between rats
and humans. A database UF of 3 was selected to account for a lack of a multigeneration
reproductive toxicity.
To provide perspective on the RfC derived using data from the JBRC inhalation bioassay
in the rat, PODs and potential inhalation RfCs based on other selected studies of carbon
tetrachloride inhalation toxicity are arrayed in Figures 5-6 to 5-8. The liver and kidney are the
predominant targets of carbon tetrachloride toxicity in subchronic and chronic inhalation studies
in laboratory animals and in humans based on case reports and studies in exposed workers.
Figures 5-6 and 5-7 provide graphical displays of information from studies that reported liver or
kidney toxicity in experimental animals following subchronic oral exposure, including the PODs,
applied UFs, and potential RfDs for comparison to the RfD derived from JBRC liver data.
Benign pheochromocytomas from the adrenal gland medulla, which could represent a potential
260 DRAFT - DO NOT CITE OR QUOTE
-------
noncancer health hazard, were observed following inhalation exposure only in mice in the JBRC
chronic bioassay. A single study of developmental toxicity found significant reductions in fetal
body weight and crown-rump length in rats at a carbon tetrachloride concentration that was also
toxic to the dams. Figure 5-8 displays PODs for all major targets of carbon tetrachloride toxicity
by the inhalation route. For the reasons discussed in Section 5.2.2, liver effects in the rat
observed in the study by JBRC are considered the most appropriate basis for the carbon
tetrachloride RfC. The text of Sections 5.2.1 and 5.2.2 should be consulted for a more complete
understanding of the issues associated with each data set and the rationale for the selection of the
critical effect and principal study used to derive the RfC.
Confidence in the principal study, the JBRC bioassay, is high. This chronic study was
well conducted, using two species (rats and mice), 3 dose groups, and 50 animals/sex/group.
The JBRC chronic study was preceded by a 13-week subchronic study, and an extensive set of
endpoints was examined in both studies. Confidence in the database, which includes the JBRC
2-year chronic inhalation bioassays in rats and mice, subchronic toxicity studies, and one study
of immunotoxic potential, is medium. Testing for developmental toxicity by inhalation exposure
found effects only at high, maternally toxic exposure concentrations but was limited to a single
inhalation study in a single species that did not test an exposure concentration low enough to
identify a NOAEL for maternal or fetal toxicity. The database lacks an adequate inhalation
multigeneration study of reproductive function. Overall confidence in the RfC is medium.
6.2.3. Cancer
The MOA of carbon tetrachloride-induced liver tumors and pheochromocytomas has not
been established. Therefore, consistent with the Guidelines for Carcinogen Risk Assessment
(U.S. EPA, 2005a), a low-dose linear extrapolation approach has been applied to the quantitative
evaluation of carbon tetrachloride carcinogen!city.
The 104-week inhalation bioassay in rats and mice conducted by JBRC (Nagano et al.,
2007b; JBRC, 1998) provided data adequate for dose-response modeling of the inhalation
pathway and was used as the basis for the IUR. Exposure levels studied in the 2-year JBRC rat
and mouse bioassay were converted to estimates of internal dose metrics by application of a
PBPK model. BMD modeling methodology (U.S. EPA, 2000c, 1995) was used to analyze the
relationship between the estimated internal doses and response (i.e., liver tumors in rats and mice
and pheochromocytomas in mice). The resulting BMDL values were converted to estimates of
equivalent FfECs by applying a human PBPK model. Data for male mouse pheochromocytomas
yielded the highest estimate of the IUR of those data sets modeled [i.e., 6 x 10"6 (jig/m3)"1].
Studies of carbon tetrachloride carcinogenicity in humans and experimental animals by
the oral exposure route are not sufficient to derive a quantitative estimate of cancer risk using
low-dose linear approaches. Therefore, PBPK modeling was applied to extrapolate inhalation
tumor data to the oral route. Because liver tumors and pheochromocytomas have been observed
261 DRAFT - DO NOT CITE OR QUOTE
-------
in experimental animals following both inhalation and oral exposures, the data sets evaluated as
the basis for the IUR were considered appropriate for estimation of an oral SF. Data for female
mouse liver tumors yielded the highest estimate of the SF of those data sets modeled [i.e.,
7 x 10'2 (mg/kg-day)'1].
An alternative nonlinear approach was also presented for quantitative dose-response
analysis of liver tumor data consistent with the evidence that supports a hypothesized MOA for
carbon tetrachloride-induced liver tumors that includes the following key events: (1) metabolism
to the trichloromethyl radical by CYP2E1 and subsequent formation of the trichloromethyl
peroxy radical, (2) radical-induced mechanisms leading to hepatocellular cytotoxicity, and
(3) sustained regenerative and proliferative changes in the liver in response to hepatotoxicity. As
a whole, empirical evidence provides significant biological support for the hypothesis that liver
carcinogenicity by this MOA occurs at carbon tetrachloride exposures that also induce
hepatocellular toxicity and a sustained regenerative and proliferative response, and that
exposures that do not cause hepatotoxicity are not expected to result in liver cancer. The RfD
and RfC were quantitatively derived based upon hepatotoxicity (cytotoxicity), a key event for the
hypothesized nonlinear MOA. Therefore, under an assumption of nonlinearity, doses (or
concentrations) of carbon tetrachloride below the RfD of 0.004 mg/kg-day or RfC of 0.1 mg/m3
that do not cause sustained cytotoxicity and regenerative cell prolifereation would be expected to
be protective of liver tumors if this is the primary MOA for liver tumors. The application of a
nonlinear approach for liver tumors is based on MOA information specific to that tumor type and
does not apply to the occurrence of liver tumors at non-cytotoxic doses nor to the occurrence of
pheochromocytomas for which the MOA is unknown. Therefore, the RfD and RfC based on
liver toxicity cannot be assumed to be protective for the potential cancer risk associated with
carbon tetrachloride-induced pheochromocytomas.
Uncertainties in the Cancer Dose-Response Assessment. Major uncertainties in the
cancer assessment are described below:
Relevance to humans. As noted in EPA's 2005 Guidelines for Carcinogen Risk
Assessment, "... agents observed to produce tumors in both humans and animals have
produced tumors either at the same site (e.g., vinyl chloride) or different sites (e.g.,
benzene) (NRC, 1994). Hence, site concordance is not always assumed between animals
and humans." Thus, it is not clear whether the tumors observed in rodent bioassays
would be predictive of human tumors of the same or different sites.
The MOA for liver tumor induction has not been established, but the
hypothesized MO As that have been investigated are assumed to be relevant to humans
(Section 4.7.3.5). There is no available evidence in humans for hepatic cancer associated
262 DRAFT - DO NOT CITE OR QUOTE
-------
with carbon tetrachloride exposure. The experimental animal literature, however, shows
carbon tetrachloride to consistently induce liver tumors across species and routes of
exposure. Further, there are similarities between experimental animals and humans in
terms of carbon tetrachloride metabolism, antioxidant systems, and evidence for the liver
as a sensitive target organ. Together, this supports a conclusion that experimental
evidence for liver cancer is relevant to humans.
Pheochromocytomas, on the other hand, were observed in only one species (the
mouse). In humans, pheochromocytomas are rare catecholamine-producing
neuroendocrine tumors that are usually benign, but may also present as or develop into a
malignancy (Eisenhofer et al., 2004; Salmenkivi et al., 2004; Tischler et al., 1996). In
humans, hereditary factors have been identified as important in the development of
pheochromocytomas (Eisenhofer et al., 2004). In the mouse, few chemicals have been
reported to cause adrenal medullary tumors (Hill et al., 2003), and the MOA for this
tumor in mice is unknown. The relevance of mouse pheochromocytomas to humans is
similarly unknown, although parallels between this tumor in the mouse and human led
investigators to concluded that the mouse might be an appropriate model for human
adrenal medullary tumors (Tischler et al., 1996). Like the human, pheochromocytomas
in the mouse are relatively rare, as are metastases. Both the morphological variability of
the mouse pheochromocytomas and the morphology of the predominant cells are
comparable to those of human pheochromocytomas. An important characteristic of
mouse pheochromocytomas is expression of immunoreactive PNMT; human
pheochromocytomas are also usually PNMT-positive (Tischler et al., 1996). Overall, this
supports a conclusion that experimental evidence for pheochromocytomas is potentially
relevant to humans.
Choice of low-dose extrapolation approach. The MOA is a key determinant of which
approach to apply for estimating low-dose cancer risk. The MOA of carbon tetrachloride
liver carcinogenicity has been investigated extensively; however, much of this research
has been conducted at relatively high exposure levels. The MOA(s) at low exposure
levels is not known.
For liver tumors, a nonlinear extrapolation approach was explored in Section 5.4.5
as an alternative to the linear low-dose extrapolation approach for cancer risk estimation.
Such an approach would be supported by a conclusion that the hypothesized cytotoxicity-
proliferative MOA is operative at all doses; however, bioassay results inconsistent with
the nonlinear approach include evidence of female mouse hepatocarcinongenicity at non-
263 DRAFT - DO NOT CITE OR QUOTE
-------
cytotoxic doses, potential for genotoxicity at low doses, and the absence of MO A
information regarding the observed pheochromocytomas in mice.
The linear extrapolation approach assumes that some cancer risk exists at all non-
zero exposures, and that this risk increases linearly with exposure. While consistent with
the recognized biological reactivity of carbon tetrachloride, uncertainties in this low-dose
extrapolation approach are associated with the lack of MO A information at low
exposures. Additional MOA information in the low-dose region to establish whether a
linear or nonlinear approach applies to carbon tetrachloride liver tumors would
significantly reduce the uncertainty associated with estimating the magnitude of liver
tumor risk.
The effect on risk estimates derived using a linear extrapolation approach of using
only data on carbon tetrachloride liver cancer response at levels below those associated
with increased cell replication was examined. The risk calculations did not prove
particularly sensitive to the limitation of data points to below which increased cell
replication was reported. This consistency in cancer risk estimates provided some
confidence that the IUR and SF estimates based on liver tumor data are not driven by
high doses associated with significant hepatotoxicity.
In data sets where early mortality is observed, methods that can reflect the
influence of competing risks and intercurrent mortality on site-specific tumor incidence
rates are preferred. Survival curves for female rats and mice from the JBRC bioassay
(see Figures 4-1 and 4-2) show early mortality in some treated groups. Because liver
tumors were the primary cause of early deaths in these groups, failure to apply a time to
tumor analysis is not likely to significantly influence the IUR for liver tumors. The
impact on the IUR from pheochromocytomas is unknown.
Under the linear low-dose extrapolation approach, cancer risk estimates were
calculated by straight line extrapolation from the POD to zero, with the multistage model
used to derive the POD. (The one exception is the male mouse pheochromocytoma data
set, where the log-probit model was used.) It is unknown how well this extrapolation
procedure predicts low-dose risks for carbon tetrachloride. The multistage 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.
For pheochromocytomas, only a linear low-dose extrapolation approach was used
to estimate human carcinogenic risk in the absence of any information on the MOA for
this tumor. MOA information to inform the approach to low-dose extrapolation for
264 DRAFT - DO NOT CITE OR QUOTE
-------
carbon tetrachloride-induced pheochromocytomas would significantly reduce the
uncertainty associated with the magnitude of risk from exposure to this tumor type.
Cancer risk estimates for liver tumors and pheochromocytomas developed using a
linear low-dose extrapolation approach were not combined because different internal
dose metrics were used in the dose-response/PBPK analysis of these two tumor types.
Deriving the IUR or oral SF for data on one tumor site, however, may underestimate the
carcinogenic potential of carbon tetrachloride. For the IUR based on male mouse
pheochromocytomas, because of the poor resolution of the dose-response relationship for
male mouse liver tumors, the magnitude of the potential risk underestimation cannot be
characterized. Because the SF based on female mouse liver tumors was an order of
magnitude greater than that for female mouse pheochromocytomas, any underestimation
of the SF is expected to be small.
Interspecies extrapolation. Extrapolating dose-response data from animals to humans
was accomplished using PBPK models in the rat, mouse, and human. Availability of a
PBPK model generally reduces the pharmacokinetic component of uncertainty associated
with animal to human extrapolation; however, any PBPK model has its own associated
uncertainties. Specific uncertainties in the PBPK modeling for carbon tetrachloride are
discussed in Section 5.3.
Route-to-route extrapolation for the oral SF. Studies of carbon tetrachloride
carcinogenicity by the oral route were determined to be insufficient to derive a
quantitative estimate of cancer risk. Therefore, a human PBPK model was used to
extrapolate inhalation data to the oral route. A simple approximation method was used
that assumed continuous infusion of carbon tetrachloride from the human GI tract to the
liver and that absorption of carbon tetrachloride from the GI tract is essentially complete.
Doses extrapolated from inhalation to oral exposures in this analysis were
approximations because they did not account for oral bioavailability or absorption
kinetics, information that is not available for carbon tetrachloride. To the extent that GI
absorption is less than 100%, the current estimation method for route-to-route
extrapolation would tend to overestimate the SF.
Statistical uncertainty at the POD. Parameter uncertainty can be assessed through
confidence intervals. Each description of parameter uncertainty assumes that the
underlying model and associated assumptions are valid. For the log-probit model applied
to the male mouse pheochromocytoma data, there is a reasonably small degree of
265 DRAFT - DO NOT CITE OR QUOTE
-------
uncertainty at the 10% excess incidence level (the POD for linear low-dose
extrapolation); the lower bound on the BMD (i.e., the BMDLio) is 1.8-fold lower than the
BMD. For the multistage model applied to the female mouse liver tumor data, there is
similarly a reasonably small degree of uncertainty at the 10% excess incidence level; the
lower bound on the BMD (i.e., the BMDLio) is approximately 1.5-fold lower than the
BMD.
Bioassay selection. The study by Nagano et al. (2007b; also reported as JBRC, 1998)
was used for development of the IUR. A full report of the bioassay findings was
published in 2007, although the study itself was conducted in the mid-1980s. Although
not a recently conducted study, this bioassay was well-designed, using two species (rats
and mice), four dose groups, including an appropriate untreated control, and 50
animals/sex/group. Examination of toxicological endpoints in both sexes of rats and
mice was appropriate. No issues were identified with this bioassay that might have
contributed to uncertainty in the cancer assessment. Alternative bioassays for developing
an IUR were unavailable.
Choice of species/gender. For liver tumors, modeling was performed using the JBRC
inhalation bioassay from the female mouse and female rat. The male rat liver tumor data
were not modeled because these data sets lacked the resolution desired for dose-response
modeling. The male mouse liver data were modeled, but provided similarly poor dose-
response curve resolution. Tumor frequencies increased from control levels to close to
maximal responses without any intervening dose levels having submaximal responses. In
the female mice and rats, lower but biologically significant levels of tumor response were
seen at intermediate dose levels. Also, notably, increased levels of hepatocellular
proliferation were not reported for rodents at these intermediate levels, so that dose-
response modeling based on these data may be more applicable to an evaluation of cancer
risk at noncytotoxic exposures. There is no indication that male rodents are more
sensitive to carbon tetrachloride liver tumor induction and that use of female data only
underestimated potential risk. For pheochromocytomas, JBRC inhalation data sets for
both male and female mice were amenable to modeling, and the data set yielding the
highest estimate of cancer risk could be selected.
Human population variability. Neither the extent of interindividual variability in carbon
tetrachloride metabolism nor human variability in response to carbon tetrachloride has
been fully characterized. Factors that could contribute to a range of human response to
carbon tetrachloride include variations in CYP450 levels because of age-related
266 DRAFT - DO NOT CITE OR QUOTE
-------
differences or other factors (e.g., exposure to other chemicals that induce or inhibit
microsomal enzymes), genetic polymorphisms in drug metabolism enzymes, transporters,
and receptors (all of which can markedly affect susceptibility to a toxic chemical),
nutritional status, alcohol consumption, or the presence of underlying disease that could
alter metabolism of carbon tetrachloride or antioxidant protection systems. Incomplete
understanding of the potential differences in metabolism and susceptibility across
exposed human populations represents a source of uncertainty.
267 DRAFT - DO NOT CITE OR QUOTE
-------
7. REFERENCES
Abbas, R; Fisher, JW. (1997) A physiologically based pharmacokinetic model for trichloroethylene and its
metabolites, chloral hydrate, trichloroacetate, dichloroacetate, trichloroethanol, and trichloroethanol glucuronide in
B6C3F1 mice. Toxicol Appl Pharmacol 147:15-30.
Abraham, P; Wilfred, G; Catherine, SP; et al. (1999) Oxidative damage to the lipids and proteins of the lungs, testis
and kidney of rats during carbon tetrachloride intoxication. Clin Chim Acta 289:177-179.
ACGIH (American Conference of Governmental Industrial Hygienists). (2001) Documentation of threshold limit
values and biological exposure indices for chemical substances in the workroom air. 7 edition. Supplement.
Cincinnati, OH: American Conference of Governmental Industrial Hygienists.
Adams, EM; Spencer, HC; Rowe, VK; et al. (1952) Vapor toxicity of carbon tetrachloride determined by
experiments on laboratory animals. Arch Ind Hyg Occup Med 6:50-66.
Adelman, R; Saul, RL; Ames, BN. (1988) Oxidative damage to DNA: relation to species metabolic rate and life
span. Proc Natl Acad Sci USA 85:2706-2708.
Agarwal, AK; Mehendale, HM. (1984) Excessive hepatic accumulation of intracellular Ca2+ in chlordecone
potentiated CC14 toxicity. Toxicology 30:17-24.
Agarwal, AK; Mehendale, HM. (1986) Effect of chlordecone on carbon tetrachloride-induced increase in calcium
uptake in isolated perfused rat liver. Toxicol Appl Pharmacol 83:342-348.
Ahn, YK; Kim, JH. (1993) Preventive effects of diphenyl dimethyl dicarboxylate on the immunotoxicity of carbon
tetrachloride in ICR mice. J Toxicol Sci 18(3): 185-195.
Ahr, HJ; King, LJ; Nastainczyk, W; et al. (1980) The mechanism of chloroform and carbon monoxide formation
from carbon tetrachloride by microsomal cytochrome P-450. Biochem Pharmacol 29:2855-2861.
Albano, E; Carini, R; Parola, M; et al. (1989). Effects of carbon tetrachloride on calcium homeostasis. A critical
reconsideration. Biochem Pharmacol 38:2719-25.
Allis, JW; Ward, TR; Seely, JC; et al. (1990) Assessment of hepatic indicators of subchronic carbon tetrachloride
injury and recovery in rats. Fundam Appl Toxicol 15:558-570.
Allis, JW; Brown, BL; Simmons, JE; et al. (1996) Methanol potentiation of carbon tetrachloride hepatotoxicity: the
central role of cytochrome P450. Toxicology 112:131-140.
Alric, L; Orfila, C; Carrere, N; et al. (2000) Reactive oxygen intermediates and eicosanoid production by Kupffer
cells and infiltrated macrophages in acute and chronic liver injury induced in rats by CC14. Inflamm Res 49:700-
707.
Alumot, E; Nachtomi, E; Mandel, E; et al. (1976) Tolerance and acceptable daily intake of chlorinated fumigants in
the rat diet. Food Cosmet Toxicol 14:105-111.
Amacher, DE; Zelljadt, I. (1983) The morphological transformation of Syrian hamster embryo cells by chemicals
reportedly nonmutagenic to Salmonella typhimurium. Carcinogenesis 4(3):291-296.
Amet, Y; Berthou, F; Founder, G; et al. (1997) Cytochrome P450 4A and 2E1 expression in human kidney
microsomes. Biochem Pharmacol 53:765-771.
Andersen, ME; Clewell, HJ, III; Gargas, ML. (1987) Physiologically based pharmacokinetics and the risk
assessment process for methylene chloride. Toxicol Appl Pharmacol 87:185-205.
268 DRAFT - DO NOT CITE OR QUOTE
-------
Andersen, NJ; Waller, CL; Adamovic, JB; et al. (1996) A pharmacokinetic model of anaerobic in vitro carbon
tetrachloride metabolism. ChemBiol Interact 101:13-31.
Anderson, MW; Reynolds, SH; You, M; et al. (1992) Role of proto-oncogene activation in carcinogenesis. Environ
Health Perspect 98:13-24.
Andervont, HB. (1958) Induction of hepatomas in strain C3II mice with 4-o-tolylazo-o-toluidine and carbon
tetrachloride. J Natl Cancer Inst 20:431-438.
Ansari, GA; Moslen, MT; Reynolds, ES. (1982) Evidence for in vivo covalent binding of CC13 derived from CC14
to cholesterol of rat liver. Biochem Pharmacol 31:3509-3510.
Araki, A; Kamigaitao, N; Sasaki, T; et al. (2004) Mutagenicity of carbon tetrachloride and chloroform in Salmonella
typhimurium TA98, TA100, TA1535, and TA1537, and Escherichia coli WP2uvrA/pKM101 and WP2/pKM101,
using a gas exposure method. Environ Mol Mutagen 43:128-133.
Arms, AD; Travis, CC (1988) Reference physiological parameters in pharmacokinetic modeling. Oak Ridge, TN:
Oak Ridge National Laboratory. EPA/600/6-88/004..PB88-196019.
Atkins, P. (1998) Diffusion controlled reactions. In: Physical chemistry (6th ed.), New York, NY: Freeman, pp. 825-
828.
Atkinson, R. (1989) Kinetics and mechanisms of the gas-phase reactions of the hydroxyl radical with organic
compounds. Monograph No. 1. New York, NY: American Chemical Society and the American Institute of Physics;
p. 66.
ATSDR (Agency for Toxic Substances and Disease Registry). (2005) Toxicological profile for carbon tetrachloride.
Public Health Service, U.S. Department of Health and Human Services, Atlanta, GA. Available online at
http://www.atsdr.cdc.gov/toxprofiles (accessed June 30, 2009). PB2006-100002.
Avasarala, S; Yang, L; Sun, Y; et al. (2006) A temporal study on the histopathological, biochemical and molecular
responses of CCl(4)-induced hepatotoxicity in Cyp2el-null mice. Toxicology 228(2-3):310-322.
Azri, S; Mata, HP; Gandolfi, AJ; et al. (1991) CC14-induced cytochrome P-450 loss and lipid peroxidation in rat
liver slices. In: Witmer, CM; ed. Biological reactive intermediates IV: molecular and cellular effects and their
impact on human health. New York, NY: Plenum Press; pp. 669-674.
Badger, DA; Sauer, J-M; Hoglen, NC; et al. (1996) The role of inflammatory cells and cytochrome P450 in the
potentiation of CC14-induced liver injury by a single dose of retinol. Toxicol Appl Pharmacol 141:507-519.
Badger, DA; Kuester, RK; Sauer, JM; et al. (1997) Gadolinium chloride reduces cytochrome P450: Relevance to
chemical-induced hepatotoxicity. Toxicology 121:143-153.
Ban, M; Hettich, D; Bonnet, P. (2003) Effect of inhaled industrial chemicals on systemic and local immune
response. Toxicology 184(1):41-50.
Barber, ED; Donish, WH; Mueller, KR. (1981) A procedure for the quantitative measurement of the mutagenicity of
volatile liquids in the Ames Salmonella/microsome assay. Mutat Res 90:31-48.
Barbin, A; Beresiat, JC; Bartsch, H. (1983) Evaluation of DNA damage by the alkaline elution technique in liver,
kidneys and lungs of rats and hamsters treated with N-nitrosodialkylamines. Carcinogenesis 4:541-545.
Barros, L; Stutzin, A; Calixto, A; et al. (2001) Nonselective cation channels as effectors of free radical-induced rat
liver cell necrosis. Hepatology 33:114-122.
269 DRAFT - DO NOT CITE OR QUOTE
-------
Barrows, LR; Shank, RC (1981)Aberrant methylation of liver DNA in rats during hepatotoxicity. Toxicol
Appl Pharmacol 60:334-345.
Bartosiewicz, MJ; Jenkins, D; Penn, S; et al. (2001) Unique gene expression patterns in liver and kidney associated
with exposure to chemical toxicants. Pharmacol Exp Ther 297:895-905.
Bechtold, MM; Gee, DL; Bruenner, U; et al. (1982) Carbon tetrachloride-mediated expiration of pentane and
chloroform by the intact rat: the effects of pretreatment with diethyl maleate, SKF-525A and phenobarbital. Toxicol
Lett 11:165-171.
Beddowes, EJ; Fau, SP; Chipman, JK. (2003) Chloroform, carbon tetrachloride and glutathione depletion induce
secondary genotoxicity in liver cells via oxidative stress. Toxicology 187:101-115.
Bedossa, P; Houglum, K; Trautwein, C; et al. (1994) Stimulation of collagen alpha 1(1) gene expression is
associated with lipid peroxidation in hepatocellular injury: a link to tissue fibrosis? Hepatology 19:1262-1271.
Beliveau, M; Lipscomb, J; Tardif, R; et al. (2005) Quantiative structure-property relationships for interspecies
extrapolation of the inhalation pharmacokinetics of organic chemicals. Chem Res Toxicol 18:475-485.
Bell, AN; Mehendale, HM. (1985) The effect of dietary exposure to a mirex plus chlordecone combination on CC14
hepatotoxicity. Fundam Appl Toxicol 5:679-687.
Bell, AN; Mehendale, HM. (1987) Comparative changes in hepatic DNA, RNA, protein, lipid, and glycogen
induced by a subtoxic dose of CC14 in chlordecone, mirex, and phenobarbital pretreated rats. Toxicol Lett 35:191-
200.
Benigni, R; Andreoli, C; Conti, L; et al. (1993) Quantitative structure-activity relationship models correctly predict
the toxic and aneuploidizing properties of six halogenated methanes in Aspergillus nidulans. Mutagenesis 8:301-
305.
Benson, JM; Springer DL. (1999) Improved risk estimates for carbon tetrachloride. Final report. U.S. Department of
Energy, Albuquerque, New Mexico; Report DE-FC04-96AL76406. Project No. 54940.
Benson, JM; Tibbetts, BM; Thrall, KD; et al. (2001) Uptake, tissue distribution, and fate of inhaled carbon
tetrachloride: comparison of rat, mouse, and hamster. Inhal Toxicol 13:207-217.
Bergman, K. (1983) Application and results of whole-body autoradiography in distribution studies of organic
solvents. Cnt Rev Toxicol 12:59-118.
Bergman, K; Muller, L; and Teigen, SW. (1996) Series: current issues in mutagenesis and carcinogenesis, No.
65. The genotoxicity and carcinogenicity of paracetamol: a regulatory (re)view. Mutat Res 349:63-288.
Bermudez, E; Mirsalis, JC; Bales, HC. (1982) Detection of DNA damage in primary cultures of rat hepatocytes
following in vivo and in vitro exposure to genotoxic agents. Environ Mutagen 4:667-679.
Bignami, M; Conti, G; Crebelli, R; et al. (1981) Growth-mediated metabolic activation of promutagens in
Aspergillus nidulans. Mutat Res 80:265-272.
Blair, A; Decoufle, P; Grauman, D. (1979) Causes of death among laundry and dry cleaning workers. Am J Public
Health 69:508-511.
Blair, A; Stewart, PA; Tolbert, TE. (1990) Cancer and other causes of death among a cohort of dry cleaners. Br J Ind
Med 47:162-168.
Blair, PC; Thompson, MB; Wilson, RE; et al. (1991) Correlation of changes in serum analytes and hepatic
histopathology in rats exposed to carbon tetrachloride. Toxicol Lett 55:149-159.
270 DRAFT - DO NOT CITE OR QUOTE
-------
Blair, A; Hartge, P; Stewart, PA; et al. (1998) Mortality and cancer incidence of aircraft maintenance workers
exposed to trichloroethylene and other organic solvents and chemicals: extended follow-up. Occup Environ Med
55:161-171.
Rogers, M; Appelman, LM; Feron, VJ; et al. (1987) Effects of the exposure profile on the inhalation toxicity of
carbon tetrachloride in male rats. J Appl Toxicol 7:185-191.
Boll, M; Weber, LWD; Becker, E; et al. (2001a) Pathogenesis of carbon tetrachloride-induced hepatocyte injury:
bioactivation of CC14 by cytochrome P450 and effects on lipid homeostasis. Z Naturforsch [C] 56:111-121.
Boll, M; Weber, LWD; Becker, E; et al. (2001b) Hepatocyte damage induced by carbon tetrachloride: inhibited
lipoprotein secretion and changed lipoprotein composition. Z Naturforsch [C] 56:283-290.
Bond, GG; Flores, GH; Shellenberger, RJ; et al. (1986) Nested case-control study of lung cancer among chemical
workers. Am JEpidemiol 124:53-66.
Boone, L; Meyer, D; Cusick, P; et al. (2005) Selection and interpretation of clinical pathology indicators of hepatic
injury in preclinical studies. Vet Clin Pathol 34(3): 182-188.
Boutelet-Bochan, H; Huang, Y; Juchau, MR. (1997) Expression of CYP2E1 during embryogenesis and fetogenesis
in human cephalic tissues: implications for the fetal alcohol syndrome. Biochem Biophys Res Commun 238:443-
447.
Bove, FJ; Fulcomer, MC; Klotz, JB; et al. (1992a) Population-based surveillance and etiological research of adverse
reproductive outcomes and toxic wastes. Report on phase IV-A: public drinking water contamination and birth
weight, fetal deaths, and birth defects. A cross-sectional study. New Jersey Department of Health, Trenton, New
Jersey.
Bove, FJ; Fulcomer, MC; Klotz, JB; et al. (1992b) Population-based surveillance and etiologic research of adverse
reproductive outcomes and toxic wastes. Report on phase IV-B: public drinking water contamination and birth
weight, fetal deaths, and birth defects. A case-control study. New Jersey Department of Health, Trenton, New
Jersey.
Bove, FJ; Fulcomer, MC; Klotz, JB; et al. (1995) Public drinking water contamination and birth outcomes. Am J
Epidemiol 141:850-862.
Boyd, MR; Statham, CN; Longo, NS. (1980) The pulmonary Clara cell as a target for toxic chemicals requiring
metabolic activation: studies with carbon tetrachloride. J Pharmacol Exp Ther 212:109-114.
Brambilla, G; Carlo, P; Finollo, R; et al. (1983) Viscometric detection of liver DNA fragmentation in rats treated
with minimal doses of chemical carcinogens. Cancer Res 43:202-209.
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 38:123-133.
Braun, R; Schoneich, J. (1975) The influence of ethanol and carbon tetrachloride on the mutagenic effectivity of
cyclophosphamide in the host-mediated assay with Salmonella typhimurium. Mutat Res 31:191-194.
Brennan, RJ; Schiestl, RH. (1998) Chloroform and carbon tetrachloride induce intrachromosomal recombination and
oxidative free radicals in Saccharomyces cerevisiae. Mutat Res 397:271-278.
Brogan, WC; Colby, HD. (1983) Carbon tetrachloride (CC14) toxicity in the guinea pig adrenal cortex. West VA
MedJ79(12):274.
271 DRAFT - DO NOT CITE OR QUOTE
-------
Brogan, WC; Eacho, PI; Hinton, DE; et al. (1984) Effects of carbon tetrachloride on adrenocortical structure and
function in guinea pigs. Toxicol Appl Pharmacol 75:118-127.
Brondeau, MT; Bonnet, P; Guenier, JP; et al. (1983) Short-term inhalation test for evaluating industrial
hepatotoxicants in rats. Toxicol Lett 19:139-146.
Brooks, S. (1998) Markov chain Monte Carlo method and its application. The Statistician 47(1):69-100.
Brown, RP; Delp, MD; Lindstedt, SL; et al. (1997) Physiological parameter values for physiologically based
pharmacokinetic models. Toxicol Ind Health 13:407-484.
Bruckner, JV; MacKenzie, WF; Muralidhara, S; et al. (1986) Oral toxicity of carbon tetrachloride: acute, subacute
and subchronic studies in rats. Fundam Appl Toxicol 6:16-34.
Bruckner, JV; Kim, HJ; Muralidhara, S; et al. (1990) Influence of route and pattern exposure on the
pharmacokinetics and hepatotoxicity of carbon tetrachloride. In: Gerrity, TR; Henry, CJ, eds. Principle of route to
route extrapolation for risk assessment. New York, NY: Elsevier Science Publishing Co., Inc.; pp. 271-284.
Bruckner, JV; Ramanathan, R; Lee, KM; et al. (2002) Mechanisms of circadian rhythmicity of carbon tetrachloride
hepatotoxicity. J Pharmacol Exp Ther 300:273-281.
Brunt, EM; Tiniakos, DG. (2002) Pathology of steatohepatitis. Best Practices Res Clin Gastroenterol 16(5):691-707.
Brzezinski, MR; Boutelet-Bochan, H; Person, RE; et al. (1999) Catalytic activity and quantitation cytochrome P-450
2E1 in prenatal human brain. J Pharmacol Exp Ther 289:1648-1653.
Bull, RJ; Sasser, LB; Lei, XC. (2004) Interactions in the tumor-promoting activity of carbon tetrachloride,
trichloroacetate, and dichloroacetate in the liver of male B6C3F1 mice. Toxicology 199:169-183.
Burke, DA; Wedd, DJ; Herriott, D; et al. (1994) Evaluation of pyrazole and ethanol induced S9 fraction in
bacterial mutagenicity testing. Mutagenesis 9:3-29.
Butterworth, BE; Smith-Oliver, T; Loury, EL; et al. (1989) Use of primary cultures of human hepatocytes
in toxicology studies. Cancer Res 49:1075-1084.
Cabre, M; Ferre, N, Folch, J; et al. (1999) Inhibition of hepatic cell nuclear DNA fragmentation by zinc in carbon
Tetrachloride-treated rats. J Hepatol 31:228-234.
Cabre, M; Camps, J; Paternain, JL; et al. (2000) Time-course of changes in hepatic lipid peroxidation and
glutathione metabolism in rats with carbon tetrachloride-induced cirrhosis. Clin Exp Pharmacol Physiol 27:694-699.
Calabrese, EJ; Baldwin, LA; Leonard, DA; et al. (1995) Decrease in hepatotoxicity by lead exposure is not
explained by its mitogenic response. J Appl Toxicol 15:129-132.
Callen, DF; Wolf, CR; Philpot, RM. (1980) Cytochrome P-450 mediated genetic activity and cytotoxicity of seven
halogenated aliphatic hydrocarbons in Saccharomyces cerevisiae. Mutat Res 77:55-63.
Cambon-Gros, C; Deltour, P; Boigegrain, RA; et al. (1986) Short communications: radical activation of carbon
tetrachloride in foetal and maternal rat liver microsomes. Biochem Pharmacol 35:2041-2044.
Cantor, KP; Stewart, PA; Brinton, LA; et al. (1995) Occupational exposures and female breast cancer mortality in
the United States. J Occup Environ Med 37:336-348.
Carpenter, SP; Lasker, JM; Raucy, JL. (1996) Expression, induction, and catalytic activity of the ethanol-inducible
cytochrome P450 (CYP2E1) in human fetal liver and hepatocytes. Mol Pharmacol 49:260-268.
272 DRAFT - DO NOT CITE OR QUOTE
-------
Casella, G; George, E. (1992) Explaining the Gibbs sampler. Am Stat 46(3):167-174.
Castillo, T; Koop, DR; Kamimura, S; et al. (1992) Role of cytochrome P-450 2E1 in ethanol-, carbon tetrachloride-
and iron-dependent microsomal lipid peroxidation. Hepatology 16:992-996.
Castro, JA; Diaz Gomez, MI. (1972) Studies on the irreversible binding of 14C-CC14 to microsomal lipids in rats
under varying experimental conditions. Toxicol Appl Pharmacol 23:541-552.
Castro, JA; de Castro, CR; de Fenos, OM; et al. (1972) Effect of cystamine on the mixed function oxygenase system
from rat liver microsomes and its preventive effect on the destruction of cytochrome P450 by carbon tetrachloride.
Pharmacol Res Commun 4:185-190.
Castro, JA; de Ferreyra, EC; de Castro, CR; et al. (1973) Studies on the mechanism of cystamine prevention of
several liver structural and biochemical alterations caused by carbon tetrachloride. Toxicol Appl Pharmacol 24:1-19.
Castro, GD; Diaz Gomez, MI; Castro, JA. (1989) Species differences in the interaction between CC14 reactive
metabolites and liver DNA or nuclear protein fractions. Carcinogenesis 10:289-294.
Castro, GD; Simpson, JT; Castro, JA. (1994) Interaction of trichloromethyl free radicals with thymine in a
model system: a mass spectrometric study. Chem Biol Interact 90:3-22.
Chandra, M; Frith, CH. (1993/94) Non-neoplastic renal lesions in Sprague-Dawley and Fischer-344 rats. Exp Toxic
Pathol 45:439-447.
Charbonneau, M; Oleskevich, S; Brodeur, J; et al. (1986) Acetone potentiation of rat liver injury induced by
trichloroethylene-carbon tetrachloride mixtures. Fundam Appl Toxicol 6:654-661.
Chatterjee, A. (1966) Testicular degeneration in rats by carbon tetrachloride intoxication. Experientia 22:394-396.
Chaudhary, AK; Nokubo, M; Reddy, GR; et al. (1994) Detection of endogenous malondialdehyde-deoxyguanosine
adducts in human liver. Science 265:1580-1582.
Checkoway, H; Wilcosky, T; Wolf, P; et al. (1984) An evaluation of the associations of leukemia and rubber
industry solvent exposures. Am J Ind Med 5:239-249.
Chib, S; Greenberg, E. (1995) Understanding the Metropolis-Hastings algorithm. Am Statistician 49(4):327-335.
Chien, KG; Sherman, SC; Mittnacht, S, Jr; et al. (1980) Membrane structure and function subsequent to calcium
activation of an endogenous phospholipase. Arch Biochem Biophys 205:614-622.
Chiu, PY; Tang, MH; Mak, DH; et al. (2003) Hepatoprotective mechanism of schisandrin B: role of mitochondrial
glutathione antioxidant status and heat shock proteins. Free Radic Biol Med 35:368-380.
Chung, F-L; Nath, RG; Ocando, J; et al. (2000) Deoxyguanosine adducts of t-4-hydroxy-2-nonal are endogenous
DNA lesions in rodents and humans: detection and potential sources. Cancer Res 60:1507-1511.
Ciccoli, L; Casini, AF; Beneditti, A. (1978) Free radical damage produced by carbon tetrachloride in the lipids of
various rat tissues. Agents Actions 8(3):303-310.
CITI (Chemicals Inspection and Testing Institute). (1992) Biodegradation and bioaccumulation data of existing
chemicals based on the CSCL Japan. Chemicals Inspection and Testing Institute, Chemical Industry Ecology-
Toxicology and Information Center, Tokyo, Japan; pp. 2-9.
Clawson, GA. (1989) Mechanisms of tetrachloride hepatotoxicity. Pathol Immunopathol Res 8:104-112.
Colby, HD. (1981) Chemical suppression of steriodogenesis. Environ Health Persp 38:119-127.
273 DRAFT - DO NOT CITE OR QUOTE
-------
Colby, HD; Brogan, WC; Miles, PR. (1981) Carbon tetrachloride-induced changes in adrenal microsomal mixed-
function oxidases and lipid peroxidation. Toxicol Appl Pharmacol 60:492-499.
Colby, HD; Purcell, H; Kominami, S; et al. (1994) Adrenal activation of carbon tetrachloride: role of microsomal
P450 isozymes. Toxicology 94:31-40.
Columbano, A; Ledda-Columbano, GM; Pibiri, M; et al. (1997) Increased expression of c-fos, c-jun and LRF-1 is
not required for in vivo priming of hepatocytes by the mitogen TCPOBOP. Oncogene 14:857-63.
Comporti, M. (1985) Biology of disease: lipid peroxidation and cellular damage in toxic liver injury. Lab Invest
53:599-623.
Comporti, M; Benedetti, A; Ferrali, M; et al. (1984) Reactive aldehydes (4-hydroxyalkenals) originating from the
peroxidation of liver microsomal lipids: biological effects and evidence for their binding to microsomal protein in
CC14 or BrCC13 intoxication. Front Gastrointest Res 8:46-62.
Condie, LW; Laurie, RD; Mills, T; et al. (1986) Effect of gavage vehicle on hepatotoxicity of carbon tetrachloride in
CD-I mice: corn oil versus Tween-60 aqueous emulsion. Fundam Appl Toxicol 7:199-206.
Cornish, HH; Ling, BP; Earth, ML. (1973) Phenobarbital and organic solvent toxicity. Am Ind Hyg Assoc J 34:487-
492.
Conner, HD; Thurman, RG; Galizi, MD; et al. (1986) The formation of a novel free radical metabolite from CC14 in
the perfused rat liver and in vivo. J Biol Chem 261:4542-4548.
Correa, P. (1996) Morphology and natural history of cancer precursors. In: Schottenfield, D; Fraumeni, JF; eds.
Cancer epidemiology and prevention. New York:, NY Oxford University Press.
Coutino, RR. (1979) Analysis of anaphase in cell culture: an adequate test system for the distinction between
compounds which selectively alter the chromosome structure or the mitotic apparatus. Environ Health Perspect
31:131-136.
Craddock, VM; Henderson, AR. (1978) De novo and repair replication of DNA in liver of carcinogen-treated
animals. Cancer Res 38:2135-2143.
Crebelli, R; Andreoli, C; Carere, A; et al. (1992) The induction of mitotic chromosome malsegregation in
Aspergillus nidulans. Quantitative structure activity relationship (QSAR) analysis with chlorinated aliphatic
hydrocarbons. Mutat Res 266:117-134.
Crebelli, R; Benigni, R; Franekic, J; et al. (1988) Induction of chromosome malsegregation by halogenated organic
solvents in Aspergillus nidulans: unspecific or specific mechanism? Mutat Res 201:401-411.
Crebelli, R; Carere, A; Leopardi, P. (1999) Evaluation of 10 aliphatic halogenated hydrocarbons in the mouse bone
marrow micronucleus test. Mutagenesis 14:207-215.
Croen, LA; Shaw, GM; Sanbonmatsu, L; et al. (1997) Maternal residential proximity to hazardous waste sites and
risk for selected congenital malformations. Epidemiology 8:347-354.
Crump, KS; Hoel, DG; Langley, CH; et al. (1976) Fundamental carcinogenic processes and their implications for
low dose4 risk assessment. Cancer Res 36:2973-2979.
Cummings, BS; Lash, LH. (2000) Metabolism and toxicity of trichloroethylene and S-(l,2-dichlorovinyl)-L-cysteine
in freshly isolated human proximal tubular cells. Toxicol Sci 53:458-466.
274 DRAFT - DO NOT CITE OR QUOTE
-------
Cummings, BS; Zangar, RC; Novak, RF; et al. (1999) Cellular distribution of cytochromes P-450 in the rat kidney.
Drug Metab Dispos 27:542-548.
Cummings, BS; Lasker, JM; Lash, LH. (2000a) Expression of glutathione-dependent enzymes and cytochrome
P450s in freshly isolated and primary cultures of proximal tubular cells from human kidney. J Pharmacol Exp Ther
293:677-685.
Cummings, BS; Parker, JC; Lash, LH. (2000b) Role of cytochrome P450 and glutathione S-transferase in
metabolism and cytotoxicity of trichloroethylene in rat kidney. Biochem Pharmacol 59:531-543.
Cummings, BS; Parker, JC; Lash, LH. (2001) Cytochrome P-450-dependent metabolism of trichloroethylene in rat
kidney. Toxicol Sci 60:11-19.
Curren, RD; Yang, LL; Conklin, PM; et al. (1988) Mutagenesis of xeroderma pigmentosum fibroblasts by acrolein.
Mutat Res 209:17-22.
Curtis, HJ; Tilley, J. (1968) Chromosome aberrations in liver forced to regenerate by chemical or surgical methods.
JGerontol 23:140-141.
Curtis, LR; Williams, WL; Mehendale, HM. (1979) Potentiation of the hepatotoxicity of carbon tetrachloride
following preexposure to chlordecone (kepone) in the male rat. Toxicol Appl Pharmacol 51:283-293.
Dai, Y; Cederbaum, AL (1995) Inactivation and degradation of human cytochrome P4502D1 by CC14 in a
transfected HepG2 cell line. J Pharmacol Exp Ther 275:1614-1622.
Dambrauskas, T; Cornish, HH. (1970) Effect of pretreatment of rats with carbon tetrachloride on tolerance
development. Toxicol Appl Pharmacol 17:83-97.
Damment, SJ; Beevers, C; Gatehouse, DG. (2005) Evaluation of the potential genotoxicity of the phosphate binder
lanthanum carbonate. Mutagenesis 20:29-37.
Daubert, TE; Danner, RP. (1995) Physical and thermodynamic properties of pure chemicals: data compilation.
Supplement 5. Washington, DC: Taylor and Francis.
Davies, B; Morris, T (1993) Physiological parameters in laboratory animals and humans. Pharm Res 10:1093-1095.
Dean, BJ; Hodson-Walker, G. (1979) An in vitro chromosome assay using cultured rat-liver cells. Mutat Res
64:329-337.
De Bont, R; van Larebeke, N. (2004) Endogenous DNA damage in humans: a review of quantitative data.
Mutagenesis 19:169-185.
Dedrick, RL; Bishoff, KB. (1980) Species similarities in pharmacokinetics. Fed Proc 39:54-59.
de Ferreyra, EC; Castro, JA; Diaz Gomez, MI; et al. (1974) Prevention and treatment of carbon tetrachloride
hepatotoxicity by cysteine: studies about its mechanism. Toxicol Appl Pharmacol 27:558-568.
de Flora, S. (1981) Study of 106 organic and inorganic compounds in the salmonella/microsome test.
Carcinogenesis 2:283-298.
de Flora, S; Zanacchi, P; Camoirano, A; et al. (1984) Genotoxicity activity and potency of 135 compounds in the
Ames reversion test and in a bacterial DNA-repair test. Mutat Res 133:161-198.
de Groot, H; Haas, W. (1981) Self-catalysed, O2-independent inactivation of NADPH- or dithionite-reduced
microsomal cytochrome P-450 by carbon tetrachloride. Biochem Pharmacol 30:2343-2347.
275 DRAFT - DO NOT CITE OR QUOTE
-------
Delaney, B; Kaminski, NE. (1993) Induction of serum-borne immunomodulatory factors in B6C3F1 mice by carbon
tetrachloride. I. Carbon tetrachloride-induced suppression of helper T-lymphocyte function is mediated by a serum
borne factor. Toxicology 85(l):67-84.
Delaney, B; Kaminski, NE. (1994) Induction of serum borne immunomodulatory factors in B6C3F1 mice by carbon
tetrachloride. Exposure to carbon tetrachloride produces an increase in B-cell number and function. Toxicology
88(1-3):201-212.
Delaney, B; Strom, SC; Collins, S; et al. (1994) Carbon tetrachloride suppresses T-cell-dependent immune
responses by induction of transforming growth factor-beta 1. Toxicol Appl Pharmacol 126(1 ):98-107.
Delia Porta, GD; Terracini, B; Shubik, P. (1961) Induction with carbon tetrachloride of liver cell carcinomas in
hamsters. J Natl Cancer Inst 26:855-863.
Delp, MD; Manning, RO; Bruckner, JV; Armstrong, RB. (1991) Distribution of cardiac output during diurnal
changes of activity in rats. Am J Physiol 261 :H1487-H1493.
de Toranzo, EG; Diaz Gomez, MI; Castro, JA. (1978) Carbon tetrachloride activation, lipid peroxidation and liver
necrosis in different strains of mice. Res Commun Chem Pathol Pharmacol 19:347-352.
de Zwart, LL; Venhorst, J; Groot, M; et al. (1997) Simultaneous determination of eight lipid peroxidation
degradation products in urine of rats treated with carbon tetrachloride using gas chromatography with electron-
capture detection. J ChromatogrB Biomed Sci Appl 694:277-287.
Dianzani, MU. (1984) Lipid peroxidation and haloalkylation: two distinct mechanisms for CC14-induced liver
damage. In: Calandra, S; Carulli, N; Salvioli, G.. eds. Liver and lipid metabolism, Amsterdam: Elsevier Exerpta
Medica, pp. 39-50
Diaz Gomez, MI; and Castro, JA. (1981) Reaction of trichloromethyl free radicals with deoxyribonucleic acid bases.
Res Commun Chem Pathol Pharmacol 32:147-153.
Diaz Gomez, MI; Castro, JA. (1980a) Covalent binding of carbon tetrachloride metabolites to liver nuclear DNA,
proteins and lipids. Toxicol Appl Pharmacol 56:199-206.
Diaz Gomez, MI; Castro, JA. (1980b) Nuclear activation of carbon tetrachloride and chloroform. Res Commun
Chem Pathol Pharmacol 27:191 -194.
DiRenzo, AB; Gandolfi, AJ; Sipes, IG. (1982) Microsomal bioactivation and covalent binding of aliphatic halides to
DNA. Toxicol Lett 11:243-252.
DiSilvestro, RA; Carlson, GP. (1994) Effects of mild zinc deficiency, plus or minus acute phase response, on CC14
hepatotoxicity. Free Radic Biol Med 16:57-61.
Docherty, JF; Burgess, E. (1922) The action of carbon tetrachloride on the liver. Br Med J 2:907-908.
Docherty, JF; Nicholls, L. (1923) Report of three autopsies following carbon tetrachloride treatment. Br Med J
2:753.
Dogukan, A; Akpolat, N; Celiker, H; et al. (2003) Protective effect of interferon-alpha on carbon tetrachloride -
induced nephrotoxicity. JNephrol 16:81-84.
Doherty, AT; Ellard, S; Parry, EM; et al. (1996) An investigation into the activation and deactivation of chlorinated
hydrocarbons to genotoxins in metabolically competent human cells. Mutagenesis 11:247-274.
276 DRAFT - DO NOT CITE OR QUOTE
-------
Doherty, RE. (2000) A history of the production and use of carbon tetrachloride, tetrachloroethylene,
trichloroethylene and 1,1,1-trichloroethane in the United State: Part 1—historical background; carbon tetrachloride
and tetrachloroethylene. J Environ Forensics 1:69-81.
Dolak, JA; Waller, RL; Glende, EA; et al. (1988) Liver cell calcium homeostasis in carbon tetrachloride liver cell
injury: new data with Fura2. J Biochem Toxicol 3:329-42.
Doolittle, DJ; Muller, G; Scribner, HE. (1987) Relationship between hepatotoxicity and induction of replicative
DNA synthesis following single or multiple doses of carbon tetrachloride. J Toxicol Environ Health 22:63-78.
Dosemeci, M; Cocco, P; Chow, W-H. (1999) Gender differences in risk of renal cell carcinoma and occupational
exposures to chlorinated aliphatic hydrocarbons. Am J Ind Med 36:54-59.
Draper, HH; Agarwal, S; Nelson, DE; et al. (1995) Effects of peroxidative stress and age on the concentration of a
deoxyguanosine-malondialdehyde adduct in rat DNA. Lipids 30:959-961.
Dumas, S; Parent, ME; Siemiatycki, J; et al. (2000) Rectal cancer and occupational risk factors: a hypothesis-
generating, exposure-based case-control study. Int J Cancer 87:874-879.
Eastmond, DA. (2008) Evaluating genotoxicity data to identify a mode of action and its application in estimating
cancer risk at low doses: a case study involving carbon tetrachloride. Environ Mol Mut 49:132-141.
Edgren, M; Revesz, L. (1987) Compartmentalized depletion of glutathione in cells treated with buthionine
sulphoximine. Br J Radiol 60:723-724.
Edwards, JE. (1941) Hepatomas in mice induced with carbon tetrachloride. J Natl Cancer Inst 2:197-199.
Edwards, JE; Dalton, AJ. (1942) Induction of cirrhosis of the liver and of hepatomas in mice with carbon
tetrachloride. J Natl Cancer Inst 3:19-41.
Edwards, J; Heston, WE; Dalton, AJ. (1942) Induction of the carbon tetrachloride hepatoma in strain L mice. J Natl
Cancer Inst 3:297-301.
Eisenhofer, G; Bornstein, SR; Brouwers, FM; et al. (2004) Malignant pheochromocytoma: current status and
initiatives for future progress. Endocrine-Related Cancer 11:423-436.
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 24:181-191.
Elkins, HB. (1942) Maximal allowable concentrations. II. Carbon tetrachloride. J Ind Hyg Toxicol 24:233-235.
El Masri, HA; Thomas, RS; Sabados, GR; et al. (1996) Physiologically based pharmacokinetic/pharmacodynamic
modeling of the toxicologic interaction between carbon tetrachloride and kepone. Arch Toxicol 70:704-713.
El Sisi, AD; Earnest, DL; Sipes, IG. (1993a) Vitamin A potentiation of carbon tetrachloride hepatotoxicity: role of
liver macrophages and active oxygen species. Toxicol Appl Pharmacol 119:295-301.
El Sisi, AD; Hall, P; Sim, W-LW; et al. (1993b) Characterization of vitamin A potentiation of carbon tetrachloride -
induced liver injury. Toxicol Appl Pharmacol 119:280-288.
Eschenbrenner, AB; Miller, E. (1946) Liver necrosis and the induction of carbon tetrachloride hepatomas in strain A
mice. J Natl Cancer Inst 6:325-341.
Esterbauer, H; Schaur, RJ; Zollner, H. (1991) Chemistry and biochemistry of 4-hydroxynonenal, malondialdehyde
and related aldehydes. Free Radic Biol Med 11:81-128.
277 DRAFT - DO NOT CITE OR QUOTE
-------
EMEA (European Medicines Agency). (2006) Draft guidelines on detection of early signals of drug-induced
hepatotoxicity in non-clinical studies. Committee for Medicinal Products for Human Use (CHMP), London.
Adoption by CHMP for release for consultation 28 June 2006. EMEA/CHMP/SWP/150115/2006. Available online
at http://www.emea.europa.eu/index/indexhl.htm (accessed June 23, 2009).
Evans, MV; Simmons, JE. (1996) Physiologically based pharmacokinetic estimated metabolic constants and
hepatotoxicity of carbon tetrachloride after methanol pretreatment in rats. Toxicol Appl Pharmacol 140:245-253.
Evans, MV; Crank, WD; Yang, HM; et al. (1994) Application of sensitivity analysis to a physiologically based
pharmacokinetic model for carbon tetrachloride in rats. Toxicol Appl Pharmacol 128:36-44.
Fanelli, SL; Castro, JA. (1995) Covalent binding of carbon tetrachloride reactive metabolites to liver microsomal
and nuclear lipid and phospholipid classes from Sprague Dawley and Osborne Mendel male rats. Teratog Carcinog
Mutagen 15:155-166.
Farber, JL. (1981) The role of calcium in cell death. Life Sci 29:1289-1295.
Farrell, CL; Senseman, LA. (1944) Carbon tetrachloride polyneuritis. A case report. RI Med J 27:334-346.
FDA Working Group. (2000) Nonclincial assessment of potential hepatotoxicity in man (a concept paper meant to
provide a framework for discussion at a February 12-13, 2001 public workshop on drug-induced hepatotoxicity).
Available online at http://www.fda.gov/cder/livertox/preclinical.pdf (accessed June 24, 2009).
Feng, L; Pereira, B; Kraus-Friedmann, N. (1992) Different localization of inositol 1,4,5-trisphosphate and ryanodine
binding sites in rat liver. Cell Calcium 13:79-87.
Fernandez, G; Villarruel, MC; de Toranzo, EGD; et al. (1982) Covalent binding of carbon tetrachloride metabolites
to the heme moiety of cytochrome P-450 and its degradation products. Res Commun Chem Pathol Pharmacol
35:283-290.
Fisher, J; Mahle, D; Bankston, L; et al. (1997) Lactational transfer of volatile chemicals in breast milk. Am Ind Hyg
AssocJ 58:425-431.
Fisher, J; Lumpkin, M; Boyd, J; et al. (2004) PBPK modeling of the metabolic interactions of carbon tetrachloride
and tetrachloroethylene in B6C3F1 mice. Environ Toxicol Pharmacol 16:93-105.
Folland, DS; Schaffner, W; Ginn, HE; et al. (1976) Carbon tetrachloride toxicity potentiated by isopropyl alcohol.
Investigation of an industrial outbreak. J Am Med Assoc 236:1853-1856.
Fountoulakis, M; de Vera, MC; Crameri, F; et al. (2002) Modulation of gene and protein expression by carbon
tetrachloride in the rat liver. Toxicol Appl Pharmacol 183:71-80.
Foureman, P; Mason, JM; Valencia, R; et al. (1994) Chemical mutagenesis testing in drosophila. X. Results of 70
coded chemicals tested for the National Toxicology Program. Environ Mol Mutagen 23:208-277.
Fowler, JSL. (1969) Carbon tetrachloride metabolism in the rabbit. Br J Pharmacol 37:733-737.
Fraga, CG; Zamora, R; Tappel, AL. (1987) Damage to protein syntheses concurrent with lipid peroxidation in rat
liver slices: effect of halogenated compounds, peroxides, and vitamin E. Arch Biochem Biophys 270:84-91.
Frantik, R; Hornychova, M; Horvath, M. (1994) Relative acute neurotoxicity of solvents: isoeffective air
concentrations of 48 compounds evaluated in rats and mice. Environ Res 66:173-185.
Galelli, ME; Castro, JA. (1998) Effect of trichloromethyl and trichloromethyl peroxyl free radicals on protein
sulfhydryl content studies in model and enzymatic carbon tetrachloride activation systems. Res Commun Mol Pathol
Pharmacol 100:227-238.
278 DRAFT - DO NOT CITE OR QUOTE
-------
Galli, A; Schiestl, RH. (1995) Salmonella test positive and negative carcinogens show different effects on
intrachromosomal recombination in G2 cell cycle arrested yeast cells. Carcinogenesis 16:659-663.
Galli, A; Schiestl, RH. (1996) Effects of Salmonella assay negative and positive carcinogens on intrachromosomal
recombination in Gj-arrested yeast cells. Mutat Res 370:209-221.
Galli, A; Schiestl, RH. (1998) Effect of salmonella assay negative and positive carcinogens on intrachromosomal
recombination in S-phase arrested yeast cells. Mutat Res 419:53-68.
Gallo, JM; Cheung, LL; Kim, JJ; et al. (1993) A physiological and system analysis hybrid pharmacokinetic model to
characterize carbon tetrachloride blood concentrations following administration in different oral vehicles. J
Pharmacokinet Biopharm 21:551-567.
Galloway, SM. (2000) Cytotoxicity and chromosome aberrations in vitro: experience in industry and the case for an
upper limit on toxicity in the aberration assay. Environ Mol Mutagen 35:191-201.
Galloway, SM; Deasy, DA; Bean, CL; et al. (1987) Effects of high osmotic strength on chromosome aberrations,
sister-chromatid exchanges and DNA strand breaks, and the relation to toxicity. Mutat Res 189:15-25.
Gans, JH; Korson, R. (1984) Liver nuclear DNA synthesis in mice following carbon tetrachloride administration or
partial hepatectomy. Proc Soc ExperBiol Med 175:237-242.
Garberg, P; Akerblom, EL; Bolcsfoldi, G. (1988) Evaluation of a genotoxicity test measuring DNA-strand breaks in
mouse lymphoma cells by alkaline unwinding and hydroxyapatite elution. Mutat Res 203:155-176.
Gardner, GH; Gove, RC; Gustafson, RK; et al. (1925) Studies on the pathological histology of experimental carbon
tetrachloride poisoning. Bull Johns Hopkins Hosp 36:107-133.
Gargas, ML; Andersen, ME; Clewell, HJ, III. (1986) A physiologically based simulation approach for determining
metabolic constants from gas uptake data. Toxicol Appl Pharmacol 86:341-352.
Gargas, ML; Burgess, RJ; Voisard, DE; et al. (1989) Partition coefficients of low-molecular-weight volatile
chemicals in various liquids and tissues. Toxicol Appl Pharmacol 98:87-99.
Garner, RC; McLean, AEM. (1969) Increased susceptibility to carbon tetrachloride poisoning in the rat after
pretreatment with oral phenobarbitone. Biochem Pharmacol 18:645-650.
Garrett, RH; Grisham, CM. (1999) Biochemistry. 2nd edition. New York, NY: Saunders College Publishing.
Garry, VF; Nelson, RL; Griffith, J; et al. (1990) Preparation for human study of pesticide applicators: sister
chromatid exchanges and chromosome aberrations in cultured human lymphocytes exposed to selected fumigants.
Teratog Carcinog Mutagen 10:21-29.
Gasso, M; Rubio, M; Varela, G; et al. (1996) Effect of S-adenosylmethionine on lipid peroxidation and liver
fibrogenesis in carbon tetrachloride-induced cirrhosis. J Hepatol 25:200-205.
Gatehouse, D; Haworth, S; Cebula, T; et al. (1994) Recommendations for the performance of bacterial mutation
assays. Mutat Res 312:217-233.
Gaynes, BI; Watkins, JB, III. (1989) Carbon tetrachloride and the sorbitol pathway in the diabetic mouse. Comp
Biochem PhysiolB 94:213-217.
Gearhart, JM; Mahle, DA; Greene, RJ; et al. (1993) Variability of physiologically based pharmacokinetic (PBPK)
model parameters and their effects on PBPK model predictions in a risk assessment for perchloroethylene (PCE).
Toxicol Lett 68:131-144.
279 DRAFT - DO NOT CITE OR QUOTE
-------
Gee, DL; Bechtold, MM; Tappel, AL. (1981) Carbon tetrachloride-induced lipid peroxidation: simultaneous in vivo
measurements of pentane and chloroform exhaled by the rat. Toxicol Lett 8:299-306.
Gilks, W; Richardson, S; Spiegelhalter, D. (1998) Markov Chain Monte Carlo in practice. New York, NY: Chapman
& Hall/CRC.
Gillespie, WR; Cheung, LL; Kim, HJ; et al. (1990) Application of system analysis to toxicology: characterization of
carbon tetrachloride oral absorption kinetics. In: Gentry, TR; Henry, CJ, eds. Principles of route-to-route
extrapolations for risk assessment. New York, NY: Elsevier Science Publishing Company, pp. 285-295.
Gillette, JR. (1973) Factors that affect the covalent binding and toxicity of drugs. In: Pharmacology and the future of
man: proceedings of the 5th international congress on pharmacology; July 1972; San Francisco, CA. Vol. 2. Basel,
Switzerland: S. Karger AG, pp. 187-202.
Oilman, MR. (1971) A preliminary study of the teratogenic effects of inhaled carbon tetrachloride and ethyl alcohol
consumption in the rat [dissertation]. Drexel University, Philadelphia, PA.
Ginsberg, G; Hattis, D; Sonawane, B; et al. (2002) Evaluation of child/adult pharmacokinetic differences from a
database derived from the therapeutic drug literature. Tox Sci 66:185-200.
Glende, EA. (1972) Carbon tetrachloride-induced protection against carbon tetrachloride toxicity. The role of the
liver microsomal drug-metabolizing system. Biochem Pharmacol 21:1697-1702.
Glende, EA; Pushpendran CK. (1986) Activation of phospholipase A2 by carbon tetrachloride in isolated
hepatocytes. Biochem Pharmacol 35:3301-3307.
Glende, EA; Recknagel, RO. (1992) Phospholipase A2 activation and cell injury in isolated rat hepatocytes exposed
to bromotrichloromethane, chloroform, and 1,1-dichloroethylene as compared to effects of carbon tetrachloride.
Toxicol Appl Pharmacol 113:159-162.
Glende, EA, Jr; Hruszkewycz, AM; Recknagel, RO. (1976) Critical role of lipid peroxidation in carbon
tetrachloride-induced loss of aminopyrine demethylase, cytochrome P-450 and glucose 6-phosphatase. Biochem
Pharmacol 25:2163-2170.
Gonzalez Padron, A; de Toranzo, EG; Castro, JA. (1993) Late preventive effects of quinacrine on carbon
tetrachloride induced liver necrosis. Arch Toxicol 67:386-391.
Gonzalez-Reimers, E; Lopez-Lirola, A; Olivera, RM; et al. (2003) Effects of protein deficiency on liver trace
elements and antioxidant activity in carbon tetrachloride-induced liver cirrhosis. Biol Trace Elem Res 93:127-139.
Gordis, E. (1969) Lipid metabolites of carbon tetrachloride. J Clin Invest 48:203-209.
Gorla, N; de Ferreyra, EC; Villarruel, MC; et al. (1983) Studies on the mechanism of glutathione prevention of
carbon tetrachloride-induced liver injury. Br J Exp Pathol 64:388-395.
Gosselin, RE; Hodge, HC; Smith, RP; et al. (1976) Clinical toxicology of commercial products. In: Acute poisoning.
4th edition. Baltimore, MD: The Williams and Wilkins Co., pp. 13, 92-97, 110.
Gould, VE; Smuckler, EA. (1971) Alveolar injury in acute carbon tetrachloride intoxication. Arch Intern Med
128:109-117.
Grasl-Kraupp, B; Ruttkay-Nedecky, B; Koudelka, H; et al. (1995) In situ detection of fragmented DNA (TUNEL
assay) fails to discriminate among apoptosis, necrosis, and autolytic cell death: a cautionary note. Hepatology
21:1465-1468.
280 DRAFT - DO NOT CITE OR QUOTE
-------
Grube, K; Br|rkle, A. (1992) PolycADP-ribose polymerase activity in mononuclear leukocytes of 12 mammailian
species correlates with species-specific life span. Proc Natl Acad Sci USA 89:11759-11763.
Gruebele, A; Zawaski, K; Kaplan, D. (1996) Effects on signal transduction as demonstrated by altered immediate-
early (c-Fos and c-Jun) gene expression and nuclear AP-1 and NK-kB transcription factor levels. Drug Metab
Dispos 24:15-22.
Gualandi, G. (1984) Genotoxicity of the free-radical producers CC14 and lipoperoxide in Aspergillus nidulans. Mutat
Res 136:109-114.
Gubskii, lul; Kurskii, MD; Zadorina, OV; et al. (1990) Calcium transport in endoplasmic reticulum of the rat liver
during lipid peroxidation. Biokhimiia 55:12-22.
Guo, TL; McCay, JA; Brown, RD; et al. (2000) Carbon tetrachloride is immunosuppressive and decreases host
resistance to Listeria monocytogenes and Streptococcus pneumoniae in female B6C3F1 mice. Toxicology 154:85-
101.
Hachiya, N; Motohashi, Y. (2000) Examination of lacZ mutant induction in the liver and testis of Muta Mouse
following injection of halogenated aliphatic hydrocarbons classified as human carcinogens. Ind Health 38:213-220.
Hadley, M; Draper, HH. (1990) Isolation of a guanine-malondialdehyde adduct from rat and human urine. Lipids
25:82-85.
Haehner, BD; Gorski, JC; Vandenbranden, M; et al. (1996) Bimodal distribution of renal cytochrome P450 3A
activity in humans. Mol Pharmacol 50:52-59.
Hafeman, DG; Hoekstra, WG. (1977) Protection against carbon tetrachloride-induced lipid peroxidation in the rat by
dietary vitamin E, selenium and methionine as measured by ethane evolution. J Nutr 107:656-665.
Hagmar, L; Bonassi, S; Stromberg, U; et al. (1998) Chromosomal aberrations in lymphocytes predict human cancer:
a report from the European Study Group on Cytogenetic Biomarkers and Health (ESCH). Cancer Res 58:4117-4121.
Hagmar, L; Stromberg, U; Bonassi, S; et al. (2004) Impact of types of lymphocyte chromosomal aberrations on
human cancer risk: results from Nordic and Italian cohorts. Cancer Res 64:2258-2263.
Hakkola, J; Raunio, H; Purkunen, R; et al. (1996) Detection of cytochrome P450 gene expression in human placenta
in first trimester of pregnancy. Biochem Pharmacol 52:379-383.
Hall, PM; Plummer, JL; Ilsley, AH; et al. (1991) Hepatic fibrosis and cirrhosis after chronic administration of
alcohol and "low-dose" carbon tetrachloride vapor in the rat. Hepatology 13:815-819.
Halliwell, B; and Gutteridge, J. (1999) Free radicals in biology and medicine. New York, NY: Oxford University
Press.
Hamlin, G.P., Kholkute, SD; Dukelow, WR. (1993) Toxicology of maternally ingested carbon tetrachloride (CC14)
on embryonal and fetal development and in vitro fertilization in mice. Zool Sci 10:111-116.
Hanasono, GK; Cote, MG; Plaa, GL. (1975) Potentiation of carbon tetrachloride-induced hepatotoxicity in alloxan-
or streptozotocin-diabetic rats. J Pharmacol Exp Ther 192:592-604.
Hansch, C; Leo, A; Hoekman, D. (1995) Exploring QSAR: Hydrophobic, electronic, and steric constants. ACS
professional reference book. Vol. 2. In: Heller, SR; ed. Washington, DC: American Chemistry Society; p. 3.
Hansen, M. (1998) Disorders of somatic and motor autonomic function. In: Pathophysiology: foundations of disease
and clinical intervention. Philadelphia, PA: W.B. Saunders Company, pp. 644-645.
281 DRAFT - DO NOT CITE OR QUOTE
-------
Hansen, MF; Cavenee, WK. (1987) Genetics of cancer predisposition. Cancer Res 47:5518-5527.
Hard, GC; Seely, JC. (2005) Recommendations for the interpretation of renal tubule proliferative lesions occurring
in rat kidneys with advanced chronic progressive nephropathy (CPN). Toxicol Pathol 33:641-649.
Harris, CC. (1991) Chemical and physical carcinogenesis: advances and perspectives for the 1990s. Cancer Res
51:5023s-5044s.
Harris, RN; Anders, MW. (1980) Effect of fasting, diethyl maleate and alcohols on carbon tetrachloride-induced
hepatotoxicity. Toxicol Appl Pharmacol 56:191-198.
Hart, RN; Setlow, RB. (1974) Correlation between deoxyribonucleic acid excision-repair and lifespan in a number
of mammalian species. Proc Natl Acad Sci USA 71:2169-2173.
Hartley, DP; Kolaja, KL; Reichard, J; et al. (1999) 4-Hydroxynonenal and malondialdehyde hepatic protein adducts
in rats treated with carbon tetrachloride: immunochemical detection and lobular localization. Toxicol Appl
Pharmacol 161:23-33.
Hattis, D; Rahmioglu, N; Verma, P; et al. (2009) Preliminary operational classification system for non-mutagenic
modes of action for carcinogenesis. Grit Rev Toxicol 39(2):97—98.
Hayes, JR; Condie, LW; Borzelleca, IF. (1986) Acute, 14-day repeated dosing, and 90-day subchronic toxicity
studies of carbon tetrachloride in CD-I mice. Fundam Appl Toxicol 7:454-463.
Heddle, JA; Dean, S; Nohmi, T; et al. (2000) In vivo transgenic mutation assays. Environ Mol Mutagen 35:253-259.
Heineman, EF; Cocco, P; Gomez, MR; et al. (1994) Occupational exposure to chlorinated aliphatic hydrocarbons
and risk of astrocytic brain cancer. Am J Ind Med 26:155-169.
Hellmer, L; Bolcsfoldi, G. (1992) An evaluation of the E. coli K-12 uvrB/recA DNA repair host-mediated assay. I.
In vitro sensitivity of the bacteria to 61 compounds. Mutat Res 272:145-160.
Hemmings, SJ; Pulga, VB; Tran ST; et al. (2002) Differential inhibitory effects of carbon tetrachloride on the
hepatic plasma membrane, mitochondrial and endoplasmic reticular calcium transport systems: implications to
hepatotoxicity. Cell Biochem Funct 20:47-59.
Hewitt, WR; Miyajima, H; Cote, MG; et al. (1980) Modification of haloalkane-induced hepatotoxicity by exogenous
ketones and metabolic ketosis. Fed Proc 39:3118-3123.
Higami, Y; Tsuchiya, T; To, K; et al. (2004) Expression of DNase gamma during Fas-independent apoptotic DNA
fragmentation in rodent hepatocytes. Cell Tissue Res 316:403-407.
Hill, GD; Pace, V; Persohn, E; et al. (2003) A comparative immunohistochemical study of spontaneous and
chemically induced pheochromocytomas in B6C3F1 mice. Endocrine Path 14:81-91.
Hocher, B; Zart, R; Diekmann, F; et al. (1996) Paracrine renal endothelin system in rats with liver cirrhosis. Br J
Pharmacol 118:220-227.
Holbrook, MT. (1993) Carbon tetrachloride. In: Kroschwitz, JI; Howe-Grant, M; eds. Kirk-Othmer encyclopedia of
chemical technology. 4th edition. Vol. 5. New York, NY: John Wiley and Sons, pp. 1062-1072.
Holden, PR; James, NH; Brooks, AN; et al. (2000) Identification of a possible association between carbon
tetrachloride-induced hepatotoxicity and interleukin-8 expression. J Biochem Mol Toxicol 14:283-290.
Holsapple, MP; Pitot, HC; Cohen, SM; et al. (2006) Mode of action relevance of rodent liver tumors to human risk.
Toxicol Sci89(l):51-56.
282 DRAFT - DO NOT CITE OR QUOTE
-------
Horvath, AL. (1982) Halogenated hydrocarbons: solubility-miscibility with water. New York, NY: Marcel Dekker,
Inc., p. 889.
Hyland, R; Gescher, A; Thummel, K; et al. (1992) Metabolic oxidation and toxification of N-methylformamide
catalyzed by the cytochrome P450 isoenzyme CYP2E1. Mol Pharmacol 41:259-266.
Ichinose, T; Miller, MG; Shibamoto, T. (1994) Determination of free malonaldehyde formed in liver microsomes
upon CC14 oxidation. J Appl Toxicol 14:453-455.
Ikatsu, H; Okino, T; Nakajima, T. (1991) Ethanol and food deprivation induced enhancement of hepatotoxicity in
rats given carbon tetrachloride at low concentration. Br J Ind Med 48:636-642.
Ikegwuonu, FI; Mehendale, HM. (1991) Biochemical assessment of the genotoxicity of the in vitro interaction
between chlordecone and carbon tetrachloride in rat hepatocytes. J Appl Toxicol 11:303-310.
ILSI/EPA (International Life Sciences Institute/U.S. Environmental Protection Agency). 1994. Physiological
Parameter Values for PBPK Models. International Life Sciences Institute Risk Science Institute, Office of Health
and Environmental Assessment, U.S. Environmental Protection Agency.
Iwai, S; Karim, R; Kitano, M; et al. (2002) Role of oxidative DNA damage caused by carbon tetrachloride-induced
liver injury — enhancement of MelQ-induced glutathione S-transferase placental form-positive foci in rats. Cancer
Lett 179:15-24.
Jaeger, RJ; Conolly, RB; Murphy, SD. (1975) Short-term inhalation toxicity of halogenated hydrocarbons. Effects
on fasting rats. Arch Environ Health 30:26-31.
Jaeschke, H; Gores, GJ; Cederbaum, Al; et al. (2002) Mechanisms in hepatotoxicity. Toxicol Sci 65:166-176.
Jakobson, I; Wahlberg, JE; Holmberg, B; et al. (1982) Uptake via the blood and elimination of 10 organic solvents
following epicutaneous exposure of anesthetized guinea pigs. Toxicol Appl Pharmacol 63:181-187.
JBRC (Japan Bioassay Research Center). (1998) Subchronic inhalation toxicity and carcinogenicity studies of
carbon tetrachloride in F344 rats and BDF1 mice (Studies Nos. 0020, 0021, 0043, and 0044). Kanagawa, Japan
Industrial Safety and Health Association, Japan Bioassay Research Center, Kanagawa, Japan. Unpublished report to
the Ministry of Labor. Hirasawa Hadano Kanagawa, 257 Japan.
Jegga, A; Inga, A, Menendaez, D; et al. (2008) Functional evolution of the p53 regulatory network through its target
response elements. Proc Natl Acad Sci USA 105:944-949.
Jeon, YJ; Han, SH; Yang, KH; et al. (1997) Induction of liver-associated transforming growth factor beta 1 (TGF-
beta 1) mRNA expression by carbon tetrachloride leads to the inhibition of T helper 2 cell-associated lymphokines.
Toxicol Appl Pharmacol 144(l):27-35.
Jessen, BA; Mullins, JS; De Peyster, A; et al. (2003) Assessment of hepatocytes and liver slices as in vitro test
systems to predict in vivo gene expression. Toxicol Sci 75:208-222.
Johnstone, RT. 1948. Occupational medicine and industrial hygiene. St. Louis, MO: CV Mosby Co., pp. 148-158.
Kadiiska, MB; Gladen, BC; Baird, DD; et al. (2005) Biomarkers of oxidative stress study II: are oxidation products
of lipids, proteins, and DNA markers of CC14 poisoning? Free Radic Biol Med 38:698-710.
Kalla, NR; Bansal, MP. (1975) Effect of carbon tetrachloride on gonadal physiology in male rats. Acta Anat 91:380-
385.
283 DRAFT - DO NOT CITE OR QUOTE
-------
Kaminski, NE; Jordan, SD; Holsapple, MP. (1989) Suppression of humoral and cell-mediated immune responses by
carbon tetrachloride. Fundam Appl Toxicol 12:117-128.
Kaminski, NE; Barnes, DW; Jordan, SD; et al. (1990) The role of metabolism in carbon tetrachloride-mediated
immunosuppression: in vivo studies. Toxicol Appl Pharmacol 102:9-20.
Kaporec, KP; Kim, HJ; MacKenzie, WF; et al. (1995) Effect of oral dosing vehicles on the subchronic
hepatotoxicity of carbon tetrachloride in the rat. J Toxicol Environ Health 44:13—27.
Kauppinen, T; Pukkala, E; Saalo, A; et al. (2003) Exposure to chemical carcinogens and risk of cancer among
Finnish laboratory workers. Am J Ind Med 44:343-350.
Kazantzis, G; Bomford, RR. (1960) Dyspepsia due to inhalation of carbon tetrachloride vapor. Lancet 1:360-362.
Kefalus, V; Stacey, NH. (1989) Potentiation of carbon tetrachloride-induced lipid peroxidation by trichloroethylene
in isolated rat hepatocytes: no role in enhanced toxicity. Toxicol Appl Pharmacol 101(1):158-169.
Kerckaert, GA; Isfort, RJ; Carr, GJ; et al. (1996) A comprehensive protocol for conducting the Syrian hamster
embryo cell transformation assay at pH 6.70. Mutat Res 356:65-84.
Kernan, GJ; Ji, B-T; Dosemeci, M; et al. (1999) Occupational risk factors for pancreatic cancer: a case-control study
based on death certificates from 24 U.S. states. Am J Ind Med 36:260-270.
Kim, HJ; Bruckner, JV; Dallas, CE; et al. (1990a) Effect of dosing vehicles on the pharmacokinetics of orally
administered carbon tetrachloride in rats. Toxicol Appl Pharmacol 102:50-60.
Kim, HJ; Odendhal, S; Bruckner, JV. (1990b) Effect of oral dosing vehicles on the acute hepatotoxicity of carbon
tetrachloride in rats. Toxicol Appl Pharmacol 102:34-49.
King-Herbert, A; Thayer, K. (2006) NTP workshop: Animal models for the NTP rodent cancer bioassay: Stocks and
strains—Should we switch? Toxicol Pathol 34:802-805.
Kiplinger, GF; Kensler, CJ. (1963) Failure of phenoxybenzamine to prevent formation of hepatomas after chronic
carbon tetrachloride administration. J Natl Cancer Inst 30:837-843.
Kitchin, KT; Brown, JL. (1989) Biochemical effects of three carcinogenic chlorinated methanes in rat liver. Teratog
Carcinog Mutagen 9:61-69.
Kitta, D; Schwarz, M; Tennekes, HA; et al. (1982) Covalent binding of CCl4-intermediates to reduced pyridine
nucleotides in mouse liver. In Snyder, D; Parke, V; Kocsis, JJ; et al., eds. Biological reactive intermediates-II:
chemical mechanisms and biological effects. New York, NY: Plenum Press, pp. 769-777'.
Klaassen, CD; ed.. (1996) Casarett and Doull's toxicology: the basic science of poisons. 5th edition. New York, NY:
McGraw-Hill, Health Professions Division.
Kluwe, WM; Herrmann, CL; Hook JB. (1979) Effects of dietary polychlorinated biphenyls and polybrominated
biphenyls on the renal and hepatic toxicities of several chlorinated hydrocarbon solvents in mice. J Toxicol Environ
Health 5:605-615.
Kniepert, E; Siegemund, A; Gorisch, V. (1990) Influence of ethanol pretreatment of differing duration on toxic
effects of carbon tetrachloride in rats. Biomed Biochim Acta 49:1097-1102.
Koch, I; Weil, R; Wolbold, R; et al. (2002) Interindividual variability and tissue-specificity in the expression of
cytochrome P450 3A mRNA. Drug Metab Dispos 30:1108-1114.
284 DRAFT - DO NOT CITE OR QUOTE
-------
Kodavanti, PR; Kodavanti, UP; Faroon, OM; et al. (1992) Pivotal role of hepatocellular regeneration in the ultimate
hepatotoxicity of CC14 in chlordecone-, mirex-, or phenobarbital-pretreated rats. Toxicol Pathol 20:556-569.
Kohno, H; Hoshino, Y; Katoh, S; et al. (1992) Effect of retinoic acid on liver transglutaminase activity and carbon
tetrachloride-induced liver damage in mice. Experientia 48:386-388.
Koporec, KP; Kim, HJ; MacKenzie, WF; et al. (1995) Effect of oral dosing vehicles on the subchronic
hepatotoxicity of carbon tetrachloride in the rat. J Toxicol Environ Health 44:13-27.
Kopylev, L; Chen, C; White, P. (2007). Towards quantitative uncertainty assessment for cancer risks: central
estimates and probability distributions of risk in dose-response modeling. Regul Toxicol Pharmacol 49:203-207.
Korsrud, GO; Grice, HC; McLaughlan, JM. (1972) Sensitivity of several serum enzymes in detecting carbon
tetrachloride-induced liver damage in rats. Toxicol Appl Pharmacol 22:474-483.
Krokan, H; Grafstrom, RC; Sundqvist, K; et al. (1985) Cytotoxicity, thiol depletion and inhibition of O6-
methylguanine-DNA methyltransferase by various aldehydes in cultured human bronchial fibroblasts.
Carcinogenesis 6:1755-1759.
Kroner, H. (1982) The intracellular distribution of liver cell calcium in normal rats and one hour after administration
of CC14. Biochem Pharmacol 31:1069-1073.
Kubale, TL; Daniels, RD; Yiin, JH; et al. (2005) A nested case-control study of leukemia mortality and ionizing
radiation at the Portsmouth Naval Shipyard. Radiat Res 164:810-819.
Kurtz, AJ; Lloyd, RS. (2003) 1 ,N2-deoxyguanosine adducts of acrolein, crotonaldehyde, and trans-4-
hydroxynonenal cross-link to peptides via Schiff base linkage. J Biol Chem 278:5970-5976.
Kwon, YH; Jovanovic, A; Serfas, MS; et al. (2003) The Cdk inhibitor p21 is required for necrosis, but it inhibits
apoptosis following toxin-induced liver injury. J Biol Chem 278:30348-30355.
Ladies, GS; Smith, C; Elliott, GS; et al. (1998) Further evaluation of the incorporation of an immunotoxicological
functional assay for assessing humoral immunity for hazard identification purposes in rats in a standard toxicology
study. Toxicology 126(2):137-152.
Lai, EK; McCay, PB; Noguchi, T; et al. (1979) In vivo spin-trapping of trichloromethyl radicals formed from carbon
tetrachloride. Biochem Pharmacol 28:2231-2235.
Lambert, IB; Singer, TM; Boucher, SE; et al. (2005) Detailed review of transgenic rodent mutation assays. Mutat
Res 590:1-280.
LeBoeuf, RA; Kerckaert, GA; Aardema, MJ; et al. (1996) The pH 6.7 Syrian hamster embryo cell transformation
assay for assessing the carcinogenic potential of chemicals. Mutat Res 356:85-127.
Lee, PY; McCay, PB; Hornbrook, KR. (1982) Evidence for carbon tetrachloride-induced lipid peroxidation in
mouse liver. Biochem Pharmacol 31:405-409.
Lee, VM; Cameron, RG; Archer, MC. (1998) Zonal location of compensatory hepatocyte proliferation following
chemically induced hepatotoxicity in rats and humans. Toxicol Pathol 26:621-627.
Lehmann, KB; Schmidt-Kehl, L. (1936) The thirteen most important chlorinated aliphatic hydrocarbons from the
standpoint of industrial hygiene. Arch Hygiene 116:132-200.
Leighton, DT, Jr; Calo, JM. (1981) Distribution coefficients of chlorinated hydrocarbons in dilute air-water systems
for groundwater contamination applications. J Chem Eng 26:382-585.
285 DRAFT - DO NOT CITE OR QUOTE
-------
Letteron, P; Labbe, G; Degott, C; et al. (1990) Mechanism for the protective effects of silymarin against carbon
tetrachloride-induced lipid peroxidation and hepatotoxicity in mice. Evidence that silymarin acts both as an inhibitor
of metabolic activation and as a chain-breaking antioxidant. Biochem Pharmacol 39:2027-2034.
Levy, GN; Brabec, MJ. (1984) Binding of carbon tetrachloride metabolites to rat hepatic mitochondrial DNA.
Toxicol Lett 22:229-234.
Lewis, RJ, Sr; ed. (1997) Hawley's condensed chemical dictionary. 13th edition. New York, NY: John Wiley and
Sons, Inc., p. 213.
Lide, DR; ed. (2000) CRC handbook of chemistry and physics. 81st Edition. Boca Raton, FL: CRC Press LLC, pp.
3-207.
Lieber, CS. (2004) Alcoholic fatty liver: its pathogenesis and mechanism of progression to inflammation and
fibrosis. Alcohol 34:9-19.
Lil'p, IG. (1982) [Chromosome instability in 101/H and C57BL/6 strain mice during aging]. Genetika 18:1976-1982.
Limaye, PB; Apte, UM; Shankar, K; et al. (2003) Calpain released from dying hepatocytes mediates progression of
acute liver injury induced by model hepatotoxicants. Toxicol Appl Pharmacol 191:211-226.
Lindstedt, SL; Calder, WA. (1981) Body size, physiological time, and longevity of homeothermic animals. Quart
RevBiol56:l-16.
Lipscomb, JC; Kedderis, GL. (2002) Incorporating human interindividual biotransformation variance in health risk
assessment. Sci Total Environ 288:13-21.
Lipscomb, JC; Garrett, CM; Snawder, JE. (1997) Cytochrome P450-dependent metabolism of trichloroethylene:
interindividual differences in humans. Toxicol Appl Pharmacol 142:311-318.
Litchfield, MH; Gartland, CJ. (1974) Plasma enzyme activity and hepatocellular changes in the beagle dog after
single or repeated administration of carbon tetrachloride. Toxicol Appl Pharmacol 30:117-128.
Lockard, VG; Mehendale, HM; O'Neal, RM. (1983) Chlordecone-induced potentiation of carbon tetrachloride
hepatotoxicity: a morphometric and biochemical study. Exp Mol Pathol 39:246-255.
Long, RM; Moore, L. (1986) Elevated cytosolic calcium in rat hepatocytes exposed to carbon tetrachloride. J
Pharmacol Exp Ther 238:186-191.
Lopez-Diazguerrero, NE; Luna-Lopez, A; Gutierrez-Ruiz, MC; et al. (2005) Susceptibility of DNA to oxidative
stressors in young and aging mice. Life Sci 77:2840-2854.
Loveday, KS; Anderson, BE; Resnick, MA; et al. (1990) Chromosome aberration and sister chromatid exchange
tests in Chinese hamster ovary cells in vitro. V: results with 46 chemicals. Environ Mol Mutagen 16:272-303.
Lowrey, KE; Glende, A, Jr; Recknagel, RO. (1981) Destruction of liver microsomal calcium pump activity by
carbon tetrachloride and bromotrichloromethane. Biochem Pharmacol 30:135-140.
Luckey, SW; Petersen, DR. (2001) Activation of Kupffer cells during the course of carbon tetrachloride-induced
liver injury and fibrosis in rats. Exp Mol Pathol 71:226-240.
Luster, MI; Simeonova, PP; Gallucci, RM; et al. (2000) Immunotoxicology: role of inflammation in chemical-
induced hepatotoxicity. Int J Immunopharmacol 22(12): 1143-1147.
Lutz, WK. (1979) In vivo covalent binding of organic chemicals to DNA as a quantitative indicator in the process of
chemical carcinogenesis. Mutat Res 65:289-356.
286 DRAFT - DO NOT CITE OR QUOTE
-------
Lutz, WK. (1986) Quantitative evaluation of DNA binding data for risk estimation and for classification of direct
and indirect carcinogens. J Cancer Res Clin Oncol 112:85-91.
Lutz, WK; Gay lor, DW; Conolly, RB; et al. (2005) Nonlinearity and thresholds in dose-response relationships for
carcinogenicity due to sampling variation, logarithmic dose scaling, or small differences in individual susceptibility.
Toxicol Applied Pharmacol 207:8565-8569.
MacNee, W; Rahman, I. (2004) Oxidative stress in chronic obstructive pulmonary disease. In: Oxygen/nitrogen
radicals: lung injury and disease. Vallyathan, V; Shi, X; Castranova, V, eds. New York, NY: Marcel Dekker, Inc.
Madle, S; Dean, SW; Andrae, U; et al. (1994) Recommendation for the performance of UDS tests in vitro and in
vivo. MutatRes 312(3):263-285.
Magos, L; Snowden, R; White, INH; et al. (1982) Isotoxic oral and inhalation exposure of carbon tetrachloride in
Porton-Wistar and Fischer rats. J Appl Toxicol 2:238-240.
Malendowicz, LK; Colby, HD. (1982) Effects of carbon tetrachloride on adrenocortical function in rats. Toxicol
Appl Pharmacol 65:32-37.
Manno, M; deMatteis, F; King, LJ. (1988) The mechanism of the suicidal, reductive inactivation of microsomal
cytochrome P-450 by carbon tetrachloride. Biochem Pharmacol 37:1981-1990.
Manno, M; Ferrara, R; Cazzaro, S; et al. (1992) Suicidal inactivation of human cytochrome P-450 by carbon
tetrachloride and halothane in vitro. Pharmacol Toxicol 70:13-18.
Manno, M; Rezzadore, M; Grossi, M; et al. (1996) Potentiation of occupational carbon tetrachloride toxicity by
ethanol abuse. Hum Exp Toxicol 15:294-300.
Marchand, C; McLean, S; Plaa, GL. (1970) The effect of SKF 525A on the distribution of carbon tetrachloride in
rats. J Pharmacol Exp Ther 174:232-238.
Maron, DM; Ames, BN. (1983) Revised methods for Salmonella mutagenicity testing. Mutat Res 113(3-4): 173-215.
Marra, F; Arrighi, MC; Fazi, M; et al. (1999) Extracellular signal-regulated kinase activation differentially regulates
platelet-derived growth factor's actions in hepatic stellate cells, and is induced by in vivo liver injury in the rat.
Hepatology30:951-958.
Martin, C; Dutertre-Catella, H; Radionoff, M; et al. (2003) Effect of age and photoperiodic conditions on
metabolism and oxidative stress related markers at different circadian stages in rat liver and kidney. Life Sci 73:327-
335.
Martinez, A; Urios, A; Blanco, M. (2000) Mutagenicity of 80 chemicals in Escherichia coli tester strains IC203,
deficient in OxyR, and its oxyR(+) parent WP2 uvrA/pKMlOl: detection of 31 oxidative mutagens. Mutat Res
467:41-53.
Martinez, M.; Mourelle, M.; Muriel, P. (1995) Protective effect of colchicine on acute liver damage induced by
CC14. Role of cytochrome P-450. J Appl Toxicol 15:49-52.
McCann, J; Choi, E; Yamasaki, E; et al. (1975) Detection of carcinogens as mutagens in the salmonella/microsome
test: assay of 300 chemicals. ProcNat Acad Sci 72:5135-5139.
McCay, PB; Lai, EK; Poyer, JL; et al. (1984) Oxygen- and carbon-centered free radical formation during carbon
tetrachloride metabolism. Observations of lipid radicals in vivo and in vitro. J Biol Chem 259:2135-2143.
287 DRAFT - DO NOT CITE OR QUOTE
-------
McCollister, DD; Beamer, WH; Atchison, GJ; et al. (1951) The absorption, distribution and elimination of
radioactive carbon tetrachloride by monkeys upon exposure to low vapor concentrations. J Pharmacol Exp Ther
102:112-124.
McLean, AEM; McLean, EK. (1966) The effect of diet and l,l,l-trichloro-2,2-bis-(p-chlorophenyl)ethane (DDT) on
microsomal hydroxylating enzymes and on sensitivity of rats to carbon tetrachloride poisoning. Biochem J 100:564-
571.
McLean, AJ; Le Couteur, DG. (2004) Aging biology and geriatric clinical pharmacology. Pharmacol Rev 56:163-
184.
Mehendale, HM. (1990) Potentiation of halomethane hepatotoxicity by chlordecone: a hypothesis for the
mechanism. Med Hypotheses 33:289-299.
Mehendale, HM. (1991) Commentary: role of hepatocellular regeneration and hepatolobular healing in the final
outcome of liver injury. A two-stage model of toxicity. Biochem Pharmacol 42:1155-1162.
Mehendale, HM. (1992) Biochemical mechanisms of biphasic dose-response relationships: role of hormesis. In:
Calabrese, EJ; ed. Biological effects of low level exposures to chemicals and radiation. Chelsea, MI: Lewis
Publishers, pp. 59-94.
Menegazzi, M; Carcereri-De Prati, A; Suzuki, H; et al. (1997) Liver cell proliferation induced by nafenopin and
cyproterone acetate is not associated with increases in activation of transcription factors NF-kappaB and AP-1 or
with expression of tumor necrosis factor alpha. Hepatology 25:585-592.
Mico, BA; Pohl, LR. (1983) Reductive oxygenation of carbon tetrachloride. Trichloromethylperoxyl radical as a
possible intermediate in the conversion of carbon tetrachloride to electrophilic chlorine. Arch Biochem Biophys
225:596-609.
Mirsalis, JC. (1987) In vivo measurement of unscheduled DNA synthesis and S-phase synthesis as an indicator of
hepatocarcinogenesis in rodents. Cell Biol Toxicol 3:165-173.
Mirsalis, JC. (1995) Transgenic models for detection of mutations in tumors and normal tissues of rodents. Toxicol
Lett 82-83:131-134.
Mirsalis, JC; Butterworth, BE. (1980) Detection of unscheduled DNA synthesis in hepatocytes isolated from rats
treated with genotoxic agents: an in vivo-in vitro assay for potential carcinogens and mutagens. Carcinogenesis
1:621-625.
Mirsalis, JC; Tyson, CK; Butterworth, BE. (1982) Detection of genotoxic carcinogens in the in vivo-in vitro
hepatocyte DNA repair assay. Environ Mutagen 4:553-562.
Mirsalis, JC; Monforte, JA; Winegar, RA. (1994) Transgenic animal models for measuring mutations in vivo. Grit
Rev Toxicol 24:255-280.
Moghaddam, AP; Eggers, JS; Calabrese, EJ. (1998) Evaluation of sex difference in tissue repair following acute
carbon tetrachloride toxicity in male and female Sprague-Dawley rats. Toxicology 130:95-105.
Molina, MD; Rowland, FS. (1974) Predicted present stratospheric abundances of chlorine species from
photodissociation of carbon tetrachloride. Geophys Res Lett 1:309-312.
Moore, L. (1980) Inhibition of liver-micro some calcium pump by in vivo administration of CC14, CHC13, and 1,1-
dichloroethylene (vinylidene chloride). Biochem Pharmacol 29:2505-2511.
Moore, L; Rodman-Daveport, G; Landon, EJ. (1976) Calcium uptake of a liver microsomal subcellular fraction in
response to in vivo administration of carbon tetrachloride. JBiol Chem 251:1197-1201.
288 DRAFT - DO NOT CITE OR QUOTE
-------
Morgan, A; Black, A; Belcher, DR. (1970) The excretion in breath of some aliphatic halogenated hydrocarbons
following administration by inhalation. Ann Occup Hyg 13:219-233.
Morgan, DL; Copper, SW; Carlock, DL; et al. (1991) Dermal absorption of neat and aqueous volatile organic
chemicals in the Fischer 344 rat. Environ Res 55:51-63.
Morita, T; Asano, N; Awogi, T; et al. (1997) Evaluation of the rodent micronucleus assay in the screening of IARC
carcinogens (groups 1, 2A and 2B) the summary report of the 6th collaborative study by CSGMT/JEMS MMS.
Collaborative Study of the Micronucleus Group Test. Mammalian Mutagenicity Study Group. Mutat Res 389:3-122.
Morley, AA; Turner, DR. (1999) The contribution of exogenous and endogenous mutagens to in vivo mutations.
Mutat Res 428(1-2):! 1-15.
Mourelle, M; Villalon, C; Amezcua, JL. (1988) Protective effect of colchicine on acute liver damage induced by
carbon tetrachloride. Hepatology 6:337-342.
Muller, L; Sofuni, T. (2000) Appropriate levels of cytotoxicity for genotoxicity tests using mammalian cells in vitro.
Environ Mol Mutagen 35:202-205.
Muller, L; Kikuchi, Y; Probst, G; et al. (1999) ICH-harmonised guidances on genotoxicity testing of
Pharmaceuticals: evolution, reasoning and impact. Mutat Res 436:195-225.
Muriel, P; Escobar, Y. (2003) Kupffer cells are responsible for liver cirrhosis induced by carbon tetrachloride. J
ApplToxicol 23:103-108.
Muriel, P; Alba, N; Perez-Alvarez, VM; et al. (2001) Kupffer cells inhibition prevents hepatic lipid peroxidation and
damage induced by carbon tetrachloride. Comp Biochem Physiol C Toxicol Pharmacol Endrocrinol 130:219-226.
Nagano, K. (2004a). Letter dated March 8, 2004. From Kasuke Nagano, JBRC, to Mary Manibusan, U.S. EPA.
Nagano, K. (2004b). Letter dated March 9, 2004. From Kasuke Nagano, JBRC, to Mary Manibusan, U.S. EPA.
Nagano, K. (2004c). Email dated March 9, 2004. Subject: Carbon tetrachloride 2-year chronic bioassay. From
Kasuke Nagano, JBRC, to Mary Manibusan, U.S. EPA.
Nagano, K. (2005). Email dated October 15, 2005. Subject: Carbon tetrachloride 1998 inhalation study. From
Kasuke Nagano, JBRC, to Mary Manibusan, U.S. EPA.
Nagano, K. (2007). Email dated April 5, 2007. Subject: Historical control data. From Kasuke Nagano, JBRC, to
Susan Rieth, U.S. EPA.
Nagano, K; Umeda, Y; Saito, M; et al. (2007a) Thirteen-week inhalation toxicity of carbon tetrachloride in rats and
mice. J Occup Health 49:249-259.
Nagano, K; Sasaki, T; Umeda, Y; et al. (2007b) Inhalation carcinogenicity and chronic toxicity of carbon
tetrachloride in rats and mice. Inhal Toxicol 19:1089-1103.
Nakajima, T; Sato, A. (1979) Enhanced activity of liver drug-metabolizing enzymes for aromatic and chlorinated
hydrocarbons following food deprivation. Toxicol Appl Pharmacol 50:549-556.
Nakamura, T; Hotchi, M. (1992) Changes in DNA strand breaks in non-parenchymal cells following hepatocyte
regeneration in CC14-induced rat liver injury. Virchows Arch B Cell Pathol Incl Mol Pathol 63:11-16.
Nakamura, S; 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 192:239-246.
289 DRAFT - DO NOT CITE OR QUOTE
-------
Nakata, R; Tsukamoto, I; Miyoshi, M; et al. (1975) Liver regeneration after carbon tetrachloride intoxication in the
rat. Biochem Pharmacol 34:586-588.
Narotsky, MG; Kavlock, RJ. (1995) A multidisciplinary approach to toxicological screening: II. Developmental
toxicity. J Toxicol Environ Health 45:145-171.
Narotsky, MG; Hamby, BT; Best, DS; et al. (1995) Carbon tetrachloride (CC14)-induced pregnancy loss in F-344
rats: luteinizing hormone (LH) levels and rescue by human chorionic gonadotropin (hCG). Biol Reprod 52(Suppl
1):172.
Narotsky, MG; Brownie, CF; Kavlock, RJ; et al. (1997a) Critical period of carbon tetrachloride-induced pregnancy
loss in Fischer 344 rats, with insights into the detection of resorption sites by ammonium sulfide staining. Teratology
56:252-261.
Narotsky, MG; Pegram, RA; Kavlock, RJ. (1997b) Effect of dosing vehicle on the developmental toxicity of
bromodichloromethane and carbon tetrachloride in rats. Fundam Appl Toxicol 40:30-36.
Natarajan, SK; Basivireddy, J; Ramachandran, A; et al. (2006) Renal damage in experimentally-induced cirrhosis in
rats: role of oxygen free radicals. Hepatology 43:1248-1456.
Nath, RG; Li, DH; Randerath, K. (1990) Acute and long-term effects of carbon tetrachloride on DNA modifications
(I-compounds) in male mouse liver. Chem Biol Interact 76:343-357.
NCI (National Cancer Institute). (1976a) Report on the carcinogenesis bioassay of chloroform. National Institutes of
Health, Bethesda, MD.
NCI (National Cancer Institute). (1976b) Carcinogenesis bioassay of trichloroethylene. National Cancer Institute
carcinogenesis technical report series, No. 2. National Institutes of Health, Bethesda, MD; NCI-CG-TR-2.
NCI (National Cancer Institute). (1977) Bioassay of 1,1,1-trichloroethane for possible carcinogenicity. National
Cancer Institute carcinogenesis technical report series, No. 3. National Institutes of Health, Bethesda, MD; NCI-
CG-TR-3.
New, PS; Lubash, GD; Scherr, L; et al. (1962) Acute renal failure associated with carbon tetrachloride intoxication.
J Am Med Assoc 181:903 -906.
Niederberger, M; Gines, P; Martin PY; et al. (1998) Increased renal and vascular cytosolic phospholipase A2
activity in rats with cirrhosis and ascites. Hepatology 27:42-47.
Niedemhofer, LJ; Daniels, JS; Rouzer, CA; et al. (2003) Malondialdehyde, a product of lipid peroxidation, is
mutagenic in human cells. J Biol Chem 278:31426-31433.
Nikula, K; Benson, J; Barr, E; et al. (1998) Comparative interspecies metabolism of carbon tetrachloride,
cytolethality and regenerative cell proliferation. Toxicol Sci 42:368.
NLM (National Library of Medicine). (2003) Carbon tetrachloride. Hazardous Substances Data Bank (HSDB).
National Institutes of Health, U.S. Department of Health and Human Services, Bethesda, MD. Available online at
http://toxnet.nlm.nih.gov (accessed June 24, 2009).
Noguchi, T; Fong, K-L; Lai, EK; et al. (1982a) Selective early loss of polypeptides in liver microsomes of CC14-
treated rats. Relationship to cytochrome P-450 content. Biochem Pharmacol 31:609-614.
Noguchi, T; Fong, K-L; Lai, EK; et al. (1982b) Specificity of a phenobarbital-induced cytochrome P-450 for
metabolism of carbon tetrachloride to the trichloromethyl radical. Biochem Pharmacol 31:615-624.
290 DRAFT - DO NOT CITE OR QUOTE
-------
Norpoth, K; Reisch, A; Heinecke, A. (1980) Biostatistics of Ames-test data. In: Norpoth, K; Garner, RC, eds. Short-
term test systems for detecting carcinogens. New York, NY: Springer-Verlag, pp. 312-322.
Norwood, WD; Fuqua, PA; Scudder, BC. (1950) Carbon tetrachloride poisoning. Arch Ind Hyg Occup Med 1:90-
100.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press.
NRC (National Research Council). (1994) Science and judgment in risk assessment. Washington, DC: National
Academy Press.
NRC (National Research Council). (2009) Science and decisions: advancing risk assessment. Washington, DC:
National Academy Press.
NTP (National Toxicology Program). (2007) National Toxicology database search application. Standard Toxicology
& Carcinogenesis Studies. Search results for carbon tetrachloride (CAS No. 56-23-5). Available at: .
Okamoto, T. (2000) Suppression of cytochrome P450 gene expression in the livers of mice with concanacalin A-
induced hepatitis. Eur J Pharmacol 394:157-161.
Omura, M; Katsumata, T; Misawa, H; et al. (1999) Decrease in protein kinase and phosphatase activities in the liver
nuclei of rats exposed to carbon tetrachloride. Toxicol Appl Pharmacol 160:192-197.
O'Neil, MJ; Smith, A; eds. (2001) The Merck index: an encyclopedia of chemicals, drugs, and biologicals. 13th
edition. Whitehouse Station, NJ: Merck and Co., Inc.; pp. 305-306.
Onfelt, A. (1987) Spindle disturbances in mammalian cells. III. Toxicity, c-mitosis and aneuploidy with 22 different
compounds. Specific and unspecific mechanisms. Mutat Res 182:135-154.
Oruambo, IF; Van Duuren, BL. (1987) Distribution of carbon tetrachloride-metabolite(s) to DNase I-sensitive and -
resistant chromatin. Cancer Lett 37:311-316.
Oshimura, M; Barrett, JC. (1986) Chemically induced aneuploidy in mammalian cells: mechanisms and biological
significance in cancer. Environ Mutagen 8:129-159.
Ozturk, F; Ucar, M; Ozturk, 1C; et al. (2003) Carbon tetrachloride-induced nephrotoxicity and protective effect of
betaine in Sprague-Dawley rats. Urology 62:353-356.
Packer, JE; Slater, TF; Willson, RL. (1978) Reactions of the carbon tetrachloride-related peroxy free radical
(CC13O.2) with amino acids: Pulse radiolysis evidence. Life Sci 23:2617-2620.
Page, DA; Carlson, GP. (1994) The role of the intestinal tract in the elimination of carbon tetrachloride. Toxicol
Appl Pharmacol 124:268-274.
Parola, M; Leonarduzzi, G; Biasi, F; et al. (1992) Vitamin E dietary supplementation protects against carbon
tetrachloride-induced chronic liver damage and cirrhosis. Hepatology 16:1014-1021.
Patki, KC; von Moltke, LL; Harmatz, JS; et al. (2004) Effect of age on in vitro triazolam biotransformation in male
human liver microsomes. J Pharmacol Exp Ther 308:874-879.
Paul, BB; Rubinstein, D. (1963) Metabolism of carbon tetrachloride and chloroform by the rat. J Pharmacol Exp
Ther 141:141-148.
291 DRAFT - DO NOT CITE OR QUOTE
-------
Paustenbach, DJ; Carlson, GP; Christian, JE; et al. (1986a) A comparative study of the pharmacokinetics of carbon
tetrachloride in the rat following repeated inhalation exposures of 8 and 11.5 hr/day. Fundam Appl Toxicol 6:484-
497.
Paustenbach, DJ; Christian, JE; Carlson, GP; et al. (1986b) The effect of an 11.5-hr/day exposure schedule on the
distribution and toxicity of inhaled carbon tetrachloride in the rat. Fundam Appl Toxicol 6:472-483.
Paustenbach, DJ; Clewell, HJ, III; Gargas, ML; et al. (1987) Development of a physiologically based
pharmacokinetic model for multiday inhalation of carbon tetrachloride. In: Pharmacokinetics in risk assessment.
Vol. 8. Washington, DC: National Academies Press, pp. 312-326.
Paustenbach, DJ; Clewell, HJ, III; Gargas, ML; et al. (1988) A physiologically based pharmacokinetic model for
inhaled carbon tetrachloride. Toxicol Appl Pharmacol 96:191-211.
Pedersen-Bjergaard, J; Andersen, MK; Christiansen, DH; et al. (2002) Genetic pathways in therapy-related
myelodysplasia and acute myeloid leukemia. Blood 99:1909-1912.
Pentz, R; Strubelt, O. (1983) Fasting increases the concentrations of carbon tetrachloride and of its metabolite
chloroform in the liver of mice. Toxicol Lett 16:231-234.
Perocco, P; Prodi, G. (1981) DNA damage by haloalkanes in human lymphocytes cultured in vitro. Cancer Lett
13:213-218.
Peter, CP; Burek, JD; van Zwieten, MJ. (1986) Spontaneous nephropathies in rats. Toxicol Pathol 14:91-100,
Peters, HA; Levine, RL; Matthews, CG; et al. (1987) Synergistic neurotoxicity of carbon tetrachloride/carbon
disulfide (80/20 fumigants) and other pesticides in grain storage workers. Acta Pharmacol Toxicol (Copenhagen)
59:535-546.
Phillips, DH; Farmer, PB; Beland, FA; et al. (2000) Methods of DNA adduct determination and their application to
testing compounds for genotoxicity. Environ Mol Mutagen 35:222-233.
Pilon, D; Brodeur, J; Plaa, GL. (1986) 1,3-Butanediol-induced increases in ketone bodies and potentiation of CC14
hepatotoxicity. Toxicology 40:165-180.
Plaa, GL. (2000) Chlorinated methanes and liver injury: highlights of the past 50 years. Ann Rev Pharmacol Toxicol
40:42-65.
Plaa, GL; Traiger, GJ. (1972) Mechanism of potentiation of CC14-induced hepatotoxicity. In: Loomis, TA; ed.
Pharmacology and the future of man. Vol. 3: Toxicological problems. Basel, Switzerland: Larger, pp 100-113.
Plummer, JL; Hall, PM; Ilsley, AH; et al. (1990) Influence of enzyme induction and exposure profile on liver injury
due to chlorinated hydrocarbon inhalation. Pharmacol Toxicol 67:329-335.
Plummer, JL; Hall, PM; Ilsley, AH; et al. (1994) Dose-response relationships in hepatic injury produced by alcohol
and carbon tetrachloride. Alcohol Clin Exper Res 18:1523-1526.
Pohl, LR; Schulick, RD; Highet, RJ; et al. (1984) Reductive-oxygenation mechanism of metabolism of carbon
tetrachloride to phosgene by cytochrome P-450. Mol Pharmacol 25:318-321.
Poyer, JL; Floyd, RA; McCay, PB; et al. (1978) Spin-trapping of the trichloromethyl radical produced during
enzymic NADPH oxidation in the presence of carbon tetrachloride or bromotrichloromethane. Biochim Biophys
Acta 539:402-409.
292 DRAFT - DO NOT CITE OR QUOTE
-------
Poyer, JL; McCay, PB; Lai, EK; et al. (1980) Confirmation of assignment of the trichloromethyl radical spin adduct
detected by spin trapping during 13C-carbon tetrachloride metabolism in vitro and in vivo. Biochem Biophys Res
Commun 94:1154-1160.
Prendergast, JA; Jones, RA; Jenkins, LJ, Jr; et al. (1967) Effects on experimental animals of long-term inhalation of
trichloroethylene, carbon tetrachloride, 1,1,1-trichloroethane dichlorodifluoromethane, and 1,1-dichloroethylene.
Toxicol Appl Pharmacol 10:270-289.
Qin, L-Q; Wang, Y; Xu, J-Y; et al. (2007) One-day dietary restriction changes hepatic metabolism and potentiates
the hepatotoxicity of carbon tetrachloride and chloroform in rats. Tohoku J Experimental Med 212:379-387.
Ramsey, JC; Andersen, ME. (1984) A physiologically based description of the inhalation pharmacokinetics of
styrene in rats and humans. Toxicol Appl Pharmacol 73:159-175.
Rao, KS; Recknagel, RO. (1969) Early incorporation of carbon-labeled carbon tetrachloride into rat liver particulate
lipids and proteins. Exp Mol Pathol 10:219-228.
Rao, PS; Dalu, A; Kulkarni, SG; et al. (1996) Stimulated tissue repair prevents lethality in isopropanol-induced
potentiation of carbon tetrachloride hepatotoxicity. Toxicol Appl Pharmacol 140:235-244.
Rasheed, A; Hines, RN; McCarver-May, DG. (1997) Variation in induction of human placental CYP2E1: possible
role in susceptibility to fetal alcohol syndrome? Toxicol Appl Pharmacol 144:396-400.
Raucy, JL; Kraner, JC; Lasker, JM. (1993) Bioactivation of halogenated hydrocarbons by cytochrome P4502E1. Grit
Rev Toxicol 23:1-20.
Ray, SD; Mehendale, HM. (1990) Potentiation of CC14 and CHC13 hepatotoxicity and lethality by various alcohols.
Fundam Appl Toxicol 15:429-440.
Raymond, P; Plaa, GL. (1995) Ketone potentiation of haloalkane-induced hepato- and nephrotoxicity. I. Dose-
response relationships. J Toxicol Environ Health 45:465-480.
Raymond, P; Plaa, GL. (1997) Effect of dosing vehicle on the hepatotoxicity of CC14 and hepatotoxicity of CHC13
in rats. J Toxicol Environ Health 51:463-476.
Recknagel, RO; Glende, EA. (1973) Carbon tetrachloride hepatotoxicity: an example of lethal cleavage. Crit Rev
Toxicol 2:263-297.
Recknagel, RO; Glende, EA; Dolak, JA; et al. (1989) Mechanisms of carbon tetrachloride toxicity. Pharmac Ther
43:139-154.
Recknagel, RO; Glende, EA; Dolak, JA; et al. (1989) Mechanisms of carbon tetrachloride toxicity. Pharmacol Ther
43:139-154.
Reinke, LA; Janzen, EG. (1991) Detection of spin adducts in blood after administration of carbon tetrachloride to
rats. ChemBiol Interact 78:155-165.
Reinke, LA; Lai, EK; McCay, PB. (1988) Ethanol feeding stimulates trichloromethyl radical formation from carbon
tetrachloride in liver. Xenobiotica 18:1311-1318.
Reitz, RH; Gargas, ML; Mendrala, AL; et al. (1996) In vivo and in vitro studies of perchloroethylene metabolism
for physiologically based pharmacokinetic modeling in rats, mice, and humans. Toxicol Appl Pharmacol 136:289-
306.
Reuber, MD; Glover EL. (1967a) Cholangiofibrosis in the liver of buffalo strain rats injected with carbon
tetrachloride. Br J Exp Pathol 48:319-322.
293 DRAFT - DO NOT CITE OR QUOTE
-------
Reuber, MD; Glover EL. (1967b) Hyperplastic and early neoplastic lesions of the liver in buffalo strain rats of
various ages given subcutaneous carbon tetrachloride. J Natl Cancer Inst 38:891-899.
Reuber, MD; Glover EL. (1970) Cirrhosis and carcinoma of the liver in male rats given subcutaneous carbon
tetrachloride. J Natl Cancer Inst 44:419-427.
Reynolds, ES; Treinen, RJ; Fairish, HH. (1984) Metabolism of [14C]carbon tetrachloride to exhaled, excreted and
bound metabolites. Biochem Pharmacol 33:3363-3374.
Rikans, LE; Hornbrook, KR; Cai, Y. (1994) Carbon tetrachloride hepatotoxicity as a function of age in female
Fischer 344 rats. Mech Ageing Dev 76:89-99.
Rocchi, P; Prodi, G; Grilli, S; et al. (1973) In vivo and in vitro binding of carbon tetrachloride with nucleic acids and
proteins in rats and mouse liver. Int J Cancer 11:419-425.
Roldan-Arjona, T; Pueyo, C. (1993) Mutagenic and lethal effects of halogenated methanes in the Ara test of
Salmonella typhimurium: quantitative relationship with chemical reactivity. Mutagenesis 8:127-131.
Roldan-Arjona, T; Garcia-Pedrajas, MD; Luque-Romero, FL; et al. (1991) An association between mutagenicity of
the Ara test of Salmonella typhimurium and carcinogenicity in rodents for 16 halogenated aliphatic hydrocarbons.
Mutagenesis 6:199-205.
Rose, ML; Bradford, BU; Germolec, DR; et al. (2001) Gadolinium chloride-induced hepatocyte proliferation is
prevented by antibodies to tumor necrosis factor a. Toxicol and Appl Pharmacol 170:39-45.
Rosengren, RJ; Sauer, J-M; Hooser, SB; et al. (1995) The interactions between retinol and five different
hepatotoxicants in the Swiss Webster mouse. Fundam Appl Toxicol 25:281-292.
Rossberg, M. (2002) Chlorinated hydrocarbons. In: Gerhartz, W; Yamamoto, YS; Campbell, FT; eds. Ullmann's
encyclopedia of industrial chemistry. 5th edition. Vol. A6. New York, NY: VCH Publishers, pp. 370-371.
Rossi, AM; Zaccaro, L; Rosselli, F; et al. (1988) Clastogenic effects induced in mice and rats by l,4-bis[2-(3,5-
dichloropyridyloxy)]-benzene, a phenobarbital-like enzyme inducer and liver tumour promoter. Carcinogenesis
9:1147-1151.
Ruprah, M; Mant, TGK; Flanagan, RJ. (1985) Acute carbon tetrachloride poisoning in 19 patients: implications for
diagnosis and treatment. Lancet 1:1027-1029.
Russell, JJ; Seetula, JA; Gutman, D; et al. (1990) Kinetics and thermochemistry of the equilibrium CC13 + O2 ~
CC1302. J Phys Chem 94:3277-3283.
Sagai, M; Tappel, AL. (1978) Effect of vitamin E on carbon tetrachloride-induced lipid peroxidation as
demonstrated by in vivo pentane production. Toxicol Lett 2:149-155.
Salmenkivi, K; Heikkila, P; Haglund, C; et al. (2004) Malignancy in pheochromocytomas. APMIS 112:551-559.
Sandy, MS; Di Monte, D; Smith, MT. (1988) Relationships between intracellular vitamin E, lipid peroxidation, and
chemical toxicity in hepatocytes. Toxicol Appl Pharmacol 93(2):288-97.
Sanzgiri, UY; Bruckner, JV. (1997) Effect of Emulphor, an emulsifier, on the pharmacokinetic and hepatotoxicity of
oral carbon tetrachloride in the rat. Fundam Appl Toxicol 36:54-61.
Sanzgiri, UY; Kim, HJ; Muralidhara, S; et al. (1995) Effect of route and pattern of exposure on the
pharmacokinetics and acute hepatotoxicity of carbon tetrachloride. Toxicol Appl Pharmacol 134:148-154.
294 DRAFT - DO NOT CITE OR QUOTE
-------
Sanzgiri, UY; Srivattsan, V; Muralidhara, S; et al. (1997) Uptake, distribution, and elimination of carbon
tetrachloride in rat tissue following inhalation and ingestion exposures. Toxicol Appl Pharmacol 143:120-129.
Sarkar, A; Pradhan, S; Mukhopadhyay, I; et al. (1999) Inhibition of early DNA-damage and chromosomal
aberrations by Trianthema portulacastruml in carbon tetrachloride-induced mouse liver damage. Cell Biol Int
23:703-708.
Sasaki, YF; Saga, A; Akasaka, M; et al. (1998) Detection of in vivo genotoxicity of haloalkanes and haloalkenes
carcinogenic to rodents by the alkaline single cell gel electrophoresis (comet) assay in multiple mouse organs. Mutat
Res 419:13-20.
Sato, A; Nakajima, T. (1985) Enhanced metabolism of volatile hydrocarbons in rat liver following food deprivation,
restricted carbohydrate intake, and administration of ethanol, phenobarbital, poly chlorinated biphenyl and 3-
methylcholanthrene: a comparative study. Xenobiotica 15:67-75.
Sato, A; Nakajima, T. (1987) Pharmacokinetics of organic solvent vapors in relation to their toxicity. Scand J Work
Environ Health 13:81-93.
Sato, A; Nakajima, T; Koyama, Y. (1980) Effects of chronic ethanol consumption on hepatic metabolism of
aromatic and chlorinated hydrocarbons in rats. Br J Ind Med 37:382-286.
Sawada, S; Yamanaka, T; Yamatsu, K; et al. (1991) Chromosome aberrations, micronuclei and sister-chromatid
exchanges (SCEs) in rat liver induced in vivo by hepatocarcinogens including heterocyclic amines. Mutat Res
251:59-69.
Sawant, SP; Dnyanmote, AV; Shankar, K; et al. (2004) Potentiation of carbon tetrachloride hepatotoxicity and
lethality in Type 2 diabetic rats. J Pharmacol Exp Ther 308:694-704.
Sawant, SP; Dnyanmote, AV; Mehendale, HM. (2007) Mechanisms of inhibited liver tissue repair in toxicant
challenged type 2 diabetic rats. Toxicology 232:200-215.
Schiestl, RH; Gietz, RD; Mehta, RD; et al. (1989) Carcinogens induce intrachromosomal recombination in yeast.
Carcinogenesis 10:1445-1455.
Schwarz, M; Hummel, J; Appel, KE; et al. (1979) DNA damage induced in vivo evaluated with a non-radioactive
alkaline elution technique. Cancer Lett 6:221-226.
Schwetz, BA; Leong, BKJ; Gehring, PJ. (1974) Embryo- and fetotoxicity of inhaled carbon tetrachloride, 1,1-
dichloroethane and methyl ethyl ketone in rats. Toxicol Appl Pharmacol 28:452-464.
Seawright, AA; Wilkie, IW; Costigan, P; et al. (1980) The effect of an equimolar mixture of carbon tetrachloride
and carbon disulphide on the liver of the rat. Biochem Pharmacol 29:1007-1014.
Seidler, A; Raum, E; Arabin, B; et al. (1999) Maternal occupational exposure to chemical substances and the risk of
infants small-for-gestational-age. Am J Ind Med 36:213-222.
Seidler, A; Mohner, M; Berger, J; et al. (2007) Solvent exposure and malignant lymphoma: a population-based case-
control study in Germany. J Occup Med Toxicol 2:2.
Seki, M; Kasama, K; Imai, K. (2000) Effect of food restriction on hepatotoxicity of carbon tetrachloride in rats. J
Toxicol Sci 25:33-40.
Selden, JR; Dolbeare, F; Miller, JE; et al. (1994) Validation of a flow cytometric in vitro DNA repair (UDS) assay
in rat hepatocytes. Mutat Res 315:147-167.
295 DRAFT - DO NOT CITE OR QUOTE
-------
Semino, G; Lilly, P; Andersen, ME. (1997) A pharmacokinetic model describing pulsatile uptake of orally-
administered carbon tetrachloride. Toxicology 117:25-33.
Semprini, L. (1995) In situ bioremediation of chlorinated solvents. Environ Health Perspect 103:101-105.
Shah, H; Hartman, SP; Weinhouse, S. (1979) Formation of carbonyl chloride in carbon tetrachloride metabolism by
rat liver in vitro. Cancer Res 39:3942-3947.
Shamberger, RJ; Andreone, TL; Willis, CE. (1974) Antioxidants and cancer. IV. Initiating activity of
malonaldehyde as a carcinogen. J Nat Cancer Inst 53:1771-1773.
Shankar, K; Vaidya, VS; Apte, UM; et al. (2003) Type 1 diabetic mice are protected from acetaminophen
hepatotoxicity. Toxicol Sci 73:220-234.
Shertzer, HG; Reitman, FA; Tabor, MW. (1988) Influence of diet on the expression of hepatotoxicity from carbon
tetrachloride in ICR mice. Drug Nutr Interact 5:275-282.
Siegers, CP; Horn, W; Younes, M. (1985) Effect of hypoxia on the metabolism and hepatotoxicity of carbon
tetrachloride and vinylidene chloride in rats. Acta Pharmacol Toxicol 56:81-86.
Simmon, VF; Tardiff, RG. (1978) The mutagenic activity of halogenated compounds found in chlorinated drinking
water. Water Chlorination Impact Health Effects 2:417-431.
Simmon, VF; Kauhanen, K; Tardiff, R. (1977) Mutagenic activity of chemicals identified in drinking water. In:
Scott, D; Bridges, BA; Sobels, FH; eds. Progress in genetic toxicology: proceedings of the 2nd international
conference on environmental mutagens; July 1977; Edinburgh, Scotland. Amsterdam, Holland: Elsevier/North-
Holland Biomedical Press, pp. 249-258.
Sina, JF; Bean, CL; Dysart, GR; et al. (1983) Evaluation of the alkaline elution/rate hepatocyte assay as predictor of
carcinogenic/mutagenic potential. MutatRes 113:357-391.
Sipes, IG; Krishna, G; Gillette, JR. (1977) Bioactivation of carbon tetrachloride, chloroform and
bromotrichloromethane: role of cytochrome P-450. Life Sci 20:1541-1548.
Sivikova, K; Piesova, E; Dianovsky, J. (2001) The protection of vitamin E and selenium against carbon
tetrachloride-induced genotoxicity in ovine peripheral blood lymphocytes. Mutat Res 494:135-142.
Slater, TF. (1981) Free radicals as reactive intermediates in tissue injury. Adv Exp Med Biol 136:575-589.
Slater, TF. (1982) Activation of carbon tetrachloride: chemical principles and biological significance. In: McBrien,
DCH; Slater, TF; eds. Free radicals, lipid peroxidation and cancer. New York, NY: Academic Press, pp. 243-274.
Smialowicz, RJ; Simmons, JE; Luebke, RW; et al. (1991) Immunotoxicologic assessment of subacute exposure of
rats to carbon tetrachloride with comparison to hepatotoxicity and nephrotoxicity. Fundam Appl Toxicol 17:186-
196.
Smith, A; Gelfand, A. (1992) Bayesian statistics without tears: A sampling-resampling perspective. Am Stat
46(2):84-89.
Smyth, HF; Smyth, HF, Jr; Carpenter, CP. (1936) The chronic toxicity of carbon tetrachloride; animal exposure and
field studies. J Ind Hyg Toxicol 18:277-298.
Snawder, JE; Lipscomb, JC. (2000) Interindividual variance of cytochrome P450 forms in human hepatic
microsomes: correlation of individual forms with xenobiotic metabolism and implications in risk assessment. Regul
Toxicol Pharmacol 32:200-209.
296 DRAFT - DO NOT CITE OR QUOTE
-------
Solomon, E; Borrow, J; Goddard, AD. (1991) Chromosome aberrations and cancer. Science 254:1153-1160.
Soni, MG; Mehendale, HM. (1993) Hepatic failure leads to lethality of chlordecone-amplified hepatotoxicity of
carbon tetrachloride. Fundam Appl Toxicol 21:442-450.
Sorsa, M; Wilbourn, J; Vainio, H. (1992) Human cytogenetic damage as a predictor of cancer risk. IARC Scientific
Publications:543-554.
Spiegelhalter, D; Thomas, A; Best, N. (2003) WinBugs Version 1.4 user manual. Available online at
http://www.mrc-bsu.cam.ac.uk/bugs/winbugs/manuall4.pdf (accessed June 24, 2009).
Spirtas, R; Stewart, PA; Lee, JS; et al. (1991) Retrospective cohort mortality study of workers at an aircraft
maintenance facility. I. Epidemiological results. Br J Ind Med 48:515-530.
Stacey, NH; Klaassen, CD. (1981) Inhibition of lipid peroxidation without prevention of cellular injury in isolated
rat hepatocytes. Toxicol Appl Pharmacol 58(1) 8-18.
Stanley, LA. (1995) Molecular aspects of chemical carcinogenesis: the roles of oncogenes and tumour suppressor
genes. Toxicology 96:173-194.
Steup, DR; Hall, P; McMillan, DA; et al. (1993) Time course of hepatic injury and recovery following
coadministration of carbon tetrachloride and trichloroethylene in Fischer 344 rats. Toxicol Pathol 21:327-334.
Stewart, BW. (1981) Generation and persistence of carcinogen-induced repair intermediates in rat liver DNA in
vivo. Cancer Res 41:3228-3243.
Stewart, RD; Dodd, HC. (1964) Absorption of carbon tetrachloride, trichloroethylene, tetrachloroethylene,
methylene chloride and 1,1,1-trichloroethane through the human skin. Am Ind Hyg Assoc J 25:439-446.
Stewart, RD; Gay, HH; Erley, DS; et al. (1961) Human exposure to carbon tetrachloride vapor. J Occup Expos
3:586-590.
Stewart, RD; Boettner, EA; Southworth, RR; et al. (1963) Acute carbon tetrachloride intoxication. J Am Med Assoc
183:94-97.
Stewart, RD; Dodd, HC; Erley, DS; et al. (1965) Diagnosis of solvent poisoning. J Am Med Assoc 193:115-118.
Stoyanovsky, DA; Cederbaum, AL. (1999) Metabolism of carbon tetrachloride to trichloromethyl radical: An ESR
and HPLC-EC study. Chem Res Toxicol 12(8):730-736.
Stoyanovsky, DA; Cederbaum, AL (1996) Thiol oxidation and cytochrome P450-dependent metabolism of CC14
triggers Ca2+ release from liver microsomes. Biochemistry 35:15839-15845.
Strubelt, O. (1984) Alcohol potentiation of liver injury. Fundam Appl Toxicol 4:144-151.
Suh; JH; Shenvi, SV; Dixon, BM; et al. (2004) Decline in transcriptional activity of Nrf2 causes age-related loss of
glutathione synthesis, which is reversible with lipoic acid. Proc Natl Acad Sci USA 101:3381-3386.
Sundari, PN; Wilfred, G; Ramakrishna, B. (1997) Does oxidative protein damage play a role in the pathogenesis of
carbon tetrachloride-induced liver injury in the rat? Biochim Biophys Acta 1362(2-3):169-176.
Suzuki, H; Hirano, N; Watanabe, C; et al. (1997) Carbon tetrachloride does not induce micronucleus in either mouse
bone marrow or peripheral blood. Mutat Res 394:77-80.
Tafazoli, M; Baeten, A; Geerlings, P; et al. (1998) In vitro mutagenicity and genotoxicity study of a number of
short-chain chlorinated hydrocarbons using the micronucleus test and the alkaline single cell gel electrophoresis
297 DRAFT - DO NOT CITE OR QUOTE
-------
technique (Comet assay) in human lymphocytes: a structure-activity relationship (QSAR) analysis of the genotoxic
and cytotoxic potential. Mutagenesis 13:115-126.
Takahashi, S; Hirose, M; Tamano, S; et al. (1998) Immunohistochemical detection of 8-hydroxy-2'-deoxyguanosine
in paraffin-embedded sections of rat liver after carbon tetrachloride treatment. Toxicol Pathol 26:247-252.
Takahashi, S; Takahashi, T; Mizobuchi, S; et al. (2002) Increased cytotoxicity of carbon tetrachloride in a human
hepatoma cell line overexpressing cytochrome P450 2E1. J Int Med Res 30:400-405.
Tanaka, E. (1998) In vivo age-related changes in hepatic drug-oxidizing capacity in humans. J Clin Pharmacol
Therapeut 23:247-255.
Taylor, SL; Tappel, AL. (1976) Effect of dietary antioxidants and phenobarbital pretreatment on microsomal lipid
peroxidation and activation by carbon tetrachloride. J Life Sci 19:1151-1160.
Teitz, NW, ed. (1976) Fundamentals of clinical chemistry. Philadelphia, PA: W.B. Saunders Company.
Teschke, R; Vierke, W; Gellert, J. (1984) Effect of ethanol on carbon tetrachloride levels and hepatotoxicity after
acute carbon tetrachloride poisoning. Arch Toxicol 56:78-82.
Thrall, K. (2006). Email dated 9/5/2006, RE: a final follow-up to your 2000 paper on carbon tetrachloride
comparative metabolism, to Susan Rieth, U.S. EPA.
Thrall, KD; Kenny, DV. (1996) Evaluation of a carbon tetrachloride physiologically based pharmacokinetic model
using real-time breath-analysis monitoring of the rat. Inhal Toxicol 8:251-261.
Thrall, KD; Vucelick, ME; Gies, RA; et al. (2000) Comparative metabolism of carbon tetrachloride in rats, mice,
and hamsters using gas uptake and PBPK modeling. J Toxicol Environ Health A 60:531-548.
Tian, L; Cai, Q; Wei, H. (1998) Alterations of antioxidant enzymes and oxidative damage to macromolecules in
different organs of rats during aging. Free Radic Biol Med 29:1477-1484.
Tischler, AS; Sheldon, W; Gray, R. (1996) Immunohistochemical and morphological characterization of
spontaneously occurring pheochromocytomas in the aging mouse. Vet Pathol 33:512-520.
Tischler, AS; Powers, JF; Alroy, J. (2004) Animal models of pheochromocytoma. Histol Histopathol 19:883-895.
Tomasi, A; Albano, E; Banni, S; et al. (1987) Free-radical metabolism of carbon tetrachloride in rat liver
mitochondria. A study of the mechanism of action. Biochem J 246:313-317.
Tombolan, F; Renault, D; Brault, D; et al. (1999) Effect of mitogenic or regenerative cell proliferation on lacz
mutant frequency in the liver of MutaTMMice treated with 5, 9-dimethyldibenzo[c,g]carbazole. Carcino gene sis
20:1357-1362.
Tomenson, JA; Baron, CE; O'Sullivan, JJ; et al. (1995) Hepatic function in workers occupationally exposed to
carbon tetrachloride. Occup Environ Med 52:508-514.
Towner, RA; Reinke, LA; Janzen, EG; et al. (1994) In vivo magnetic resonance imaging study of Kupffer cell
involvement in CC14-induced hepatotoxicity in rats. Can J Physiol Pharmacol 72(5):441-446.
Tracey JP, Sherlock P. (1968) Hepatoma following carbon tetrachloride poisoning. NY J Med 68:2202-2204.
Traiger, GJ; Bruckner, JV. (1976) The participation of 2-butanone in 2-butanol-induced potentiation of carbon
tetrachloride hepatotoxicity. J Pharmacol Exp Therap 196:493-500.
298 DRAFT - DO NOT CITE OR QUOTE
-------
Traiger, GJ; Plaa, GL. (1971) Differences in the potentiation of carbon tetrachloride in rats by ethanol and
isopropanol pretreatment. Toxicol Appl Pharmacol 20:105-112.
Travis, CC. (1990) Tissue dosimetry for reactive metabolites. Risk Anal 10:317-321.
Travlos GS; Minis RW; Elwell MR; et al. (1996) Frequency and relationships of clinical chemistry and liver and
kidney histopathology findings in 13-week toxicity studies in rats. Toxicology 107:17-29.
Tribble, DL; Aw, TY; Jones, DP. (1987) The pathophysiological significance of lipid peroxidation in oxidative cell
injury. Hepatology 7:377-386.
Tsuda, H; Matsumoto, K; Ogino, H, et al. (1993) Demonstration of initiation potential of carcinogens by induction
of preneoplastic glutathione S-transferase P-form-positive liver cell foci: possible in vivo assay system for
environmental carcinogens. Jpn J Cancer Res 84:230-236.
Tsujimura, K; Ichinose, F; Kara, T; et al. (2008) The inhalation exposure of carbon tetrachloride promote rat liver
carcinogenesis in a medium-term liver bioassay. Toxicology Letters 176:207-214.
Tsuruta, H. (1975) Percutaneous absorption of organic solvents. Comparative study of the in vivo percutaneous
absorption of chlorinated solvents in mice. Ind Health 13:227-236.
Uehleke, H; Hellmer, KH; Tabarelli, S. (1973) Binding of 14C-carbon tetrachloride to microsomal proteins in vitro
and formation of CHC13 by reduced liver microsomes. Xenobiotica 3:1-11.
Uehleke, H; Werner, T; Greim, H; et al. (1977) Metabolic activation of haloalkanes and tests in vitro for
mutagenicity. Xenobiotica 7:393-400.
Uemitsu, N. (1986) Inhalation pharmacokinetics of carbon tetrachloride in rats based on arterial blood:inhaled air
concentration ratios. Toxicol Appl Pharmacol 83:20-29.
Umiker, W; Pearce J. (1953) Nature and genesis of pulmonary alterations in carbon tetrachloride poisoning. Arch
Pathol 55:203-217.
Uryvaeva, IV; Delone, GV. (1995) An improved method of mouse liver micronucleus analysis: an application to
age-related genetic alteration and polyploidy study. Mutat Res 334:71-80.
U.S. Coast Guard. (1999) Carbon tetrachloride. Chemical Hazards Response Information System (CHRIS)
Hazardous Chemical Data. Department of Transportation, Washington, DC; Available online at
http://www.chrismanual.com/findform.htm (accessed June 23, 2009).
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment of chemical
mixtures. Fed Registr 51(185):34014-34025. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed
January 15,2009).
U.S. EPA (Environmental Protection Agency). (1986b) Guidelines for mutagenicity risk assessment. Fed Regist
51(185):34006-34012. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (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/backgr-d.htm (accessed January
15,2009)..
U.S. EPA (Environmental Protection Agency). (1991) Guidelines for developmental toxicity risk assessment. Fed
Regist 56(234):63798-63826. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15,
2009).
299 DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA (Environmental Protection Agency). (1994a) Interim policy for particle size and limit concentration
issues in inhalation toxicity studies. Fed Regist 59(206):53799 (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (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/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (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 January 15, 2009).
U.S. EPA (Environmental Protection Agency). (1996a) Guidelines for reproductive toxicity risk assessment. Fed
Regist 61(212):56274-56322. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15,
2009).
U.S. EPA (Environmental Protection Agency). (1996b) Symposium on natural attenuation of chlorinated organics in
groundwater. Office of Research and Development, Washington, DC; EPA-540/R-96/509. National Technical
Information Service, Springfield, VA; AD-A319114/5.
U.S. EPA (Environmental Protection Agency). (1998a) Guidelines for neurotoxicity risk assessment. Fed Regist
63(93):26926-26954.
U.S. EPA (Environmental Protection Agency). (2000a) Science policy council handbook: peer review. 2nd edition.
Prepared by the Office of Science Policy, Office of Research and Development, Washington, DC; EPA 100-B-OO-
001. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed June 23, 2009).
U.S. EPA (Environmental Protection Agency). (2000b) Science policy council handbook: risk characterization.
Prepared by the Office of Science Policy, Office of Research and Development, Washington, DC; EPA 100-B-OO-
002. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2000c) 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/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2000e) Toxicological review of vinyl chloride (CAS No. 75-01-4):
In support of summary information on the integrated risk information system (IRIS), Washington, DC; EPA/635R-
00/004. Available online at http://www.epa.gov/iris (accessed June 24, 2009).
U.S. EPA (Environmental Protection Agency). (200la) Toxicological review of chloroform. Integrated Risk
Information System (IRIS), National Center for Environmental Assessment, Washington, DC; EPA/635/R-01/001.
Available online at http://www.epa.gov/iris (accessed June 24, 2009).
U.S. EPA (Environmental Protection Agency). (200Ib) Exploration of aging and toxic response issues [final report].
Prepared by Versar, Inc. for U.S. Environmental Protection Agency. Risk Assessment Forum, Washington, DC;
EPA/630/R-01/003.
U.S. EPA (Environmental Protection Agency). (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/backgr-d.htm (accessed June 24, 2009).
U.S. EPA (Environmental Protection Agency). (2005a) Guidelines for carcinogen risk assessment. Risk Assessment
Forum, Washington, DC; EPA/630/P-03/001b. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed
January 15,2009).
3 00 DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA (Environmental Protection Agency). (2005b) Supplemental guidance for assessing susceptibility from
early-life exposure to carcinogens. Risk Assessment Forum, Washington, DC; EPA/630/P-03/003F. Available
online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2006a) Science policy council handbook: Peer review. 3rd edition.
Science Policy Council, Washington, DC. Available online at
http://www.epa.gov/ncea/iris/Peer_Review_Handbook_2006_3rd_edition.pdf (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (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 January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2006c) Approaches for the application of physiologically based
pharmacokinetic (PBPK) models and supporting data in risk assessment (final report). U.S. Environmental
Protection Agency, Washington, D.C., EPA/600/R-05/043F.
U.S. EPA (Environmental Protection Agency). (2007a) Protection of stratospheric ozone: extension of global
laboratory and analytical use exemption for essential class I ozone-depleting substances. Fed Regist
72(177):52332-52337. Available online at http://frwebgate4.access.gpo.gov/cgi-
bm/PDFgate.cgi?WAISdocID=467979422640+16+2+0&WAISaction=retneve (accessed July 1, 2009.)
U.S. EPA (Environmental Protection Agency). (2007b) Benchmark dose software (HMDS) version 1.4.1. last
modified February 2007. Available online at http://www.epa.gov/ncea/bmds.htm (accessed June 23, 2009).
Van Goethem, F; Ghahroudi, MA; Castelain, P; et al. (1993) Frequency and DNA content of micronuclei in rat
parenchymal liver cells during experimental hepatocarcinogenesis. Carcinogenesis 14:2397-2406.
Van Goethem, F; de Stoppelaar, J; Hoebee, B; Kirsch-Volders, M. (1995) Identification of clastogenie and/or
aneugenic events during the preneoplastic stages of experimental rat hepatocarcinogenicity by fluorescence in site
hybridization. Carcinogenesis 16(8):1825-34.
Van Kuijk, FJ; Holte, LL; Dratz, EA. (1990) 4-Hydroxyhexenal: a lipid peroxidation product derived from oxidized
docosahexaenoic acid. Biochim Biophys Acta 1043:116-118.
Varela-Moreiras, G; Alonso-Aperte, E; Rubio, M; et al. (1995) Carbon tetrachloride-induced hepatic injury is
associated with global DNA hypomethylation and homocysteinemia: effect of S-adenosylmethionine treatment.
Hepatology22:1310-1315.
Varma, MM; Ampy, FR; Verma, K; et al. (1988) In vitro mutagenicity of water contaminants in complex mixtures. J
Appl Toxicol 8:243-248.
Veng-Pedersen, P; Paustenbach, DJ; Carlson, GP; et al. (1987) A linear systems approach to analyzing the
pharmacokinetics of carbon tetrachloride in the rat following repeated exposures of 8 and 11.5 h/day. Arch Toxicol
60:355-364.
Vieira, I; Sonnier, M; Cresteil, T. (1996) Developmental expression of CYP2E1 in the human liver.
Hypermethylation control of gene expression during the neonatal period. Eur J Biochem 238:476-483.
Vijg, J; Mullaart, E; van der Schans, GP; et al. (1984) Kinetics of ultraviolet induced DNA excision repair in rat and
human fibroblasts. MutatRes 132:129-138.
Villarruel, M; de Toranzo, EGD; Castro, JA. (1977) Carbon tetrachloride activation, lipid peroxidation and the
mixed function oxygenase activity of various rat tissues. Toxicol Appl Pharmacol 41:337-344.
von Oettingen, WF. (1964) The halogenated hydrocarbons of industrial and toxicological importance. In: Browning,
E; ed. Elsevier monographs on toxic agents. New York, NY: Elsevier Publishing Co.
3 01 DRAFT - DO NOT CITE OR QUOTE
-------
von Oettingen, WF; Powell, CC; Sharpless, NE; et al. (1950) Comparative studies of the toxicity and
pharmacodynamic action of chlorinated methanes with special reference to their physical and chemical
characteristics. Arch Int Pharmacodyn 81:17-34.
Wacker, M; Wanek, P; Eder, E. (2001) Detection of 1, N2-propanodeoxyguanosine adducts of trans-4-hydroxy-2-
nonenal after gavage of trans-4-hydroxy-2-nonenal or induction of lipid peroxidation with carbon tetrachloride in
F344rats. Chem Biol Interact 137:269-283.
Wang, MY; Liehr, JG. (1995) Lipid hydroperoxide-induced endogenous DNA adducts in hamsters: possible
mechanism of lipid hydroperoxide-mediated carcinogenesis. Arch Biochem Biophys 316:38-46.
Wang, P-Y; Kaneko, T; Tsukada, H; et al. (1997) Time courses of hepatic injuries induced by chloroform and by
carbon tetrachloride: comparison of biochemical and histopathological changes. Arch Toxicol 71:638-645.
Wangenheim, J; Bolcsfoldi, G. (1988) Mouse lymphoma L5178Y thymidine kinase locus assay of 50 compounds.
Mutagenesis 3:193-205.
Warrington, JS; Poku, JW, von Moltke, LL; et al. (2000) Effects of age on in vitro midazolam biotransformation in
male CD-I mouse liver microsomes. J Pharmacol Exp Ther 292:1024-1031.
Warrington, JS; Von Moltke, LL; Greenblatt, DJ. (2004) Age-related differences in CYP3A expression and activity
in the rat liver, intestine and kidney. J Pharmacol Exp Ther 309:720-729.
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/pKMlOl:
collaborative study II. Mutat Res 412:17-31.
Watkins, JB, III; Sanders, RA; Beck, LV. (1988) The effect of long-term streptozotocin-induced diabetes on the
hepatotoxicity of bromobenzene and carbon tetrachloride and hepatic biotransformation in rats. Toxicol Appl
Pharmacol 93:329-338.
Wauthier, V; Verbeeck, RK; Caulderon, PB. (2004) Age-related changes in the protein and mRNA levels of
CYP2E1 and CYP3A isoforms as well as in their hepatic activities in Wistar rats. What role for oxidative stress?
Arch Toxicol 78:131-138.
Weber, LW; Boll, M; Stampfl, A. (2003) Hepatotoxicity and mechanism of action of haloalkanes: carbon
tetrachloride as a toxicological model. Crit Rev Toxicol 33:105-136.
Weddle, CE; Hornbrook, KR; McCay, PB. (1976) Lipid peroxidation and alteration of membrane lipids in isolated
hepatocytes exposed to carbon tetrachloride. J Biol Chem 251:4973-4978.
Weisburger, EK. (1977) Carcinogenicity studies on halogenated hydrocarbons. Environ Health Perspect 21:7-16.
West, GB; Woodruff, WH; Brown, JH. (2002) Allometric scaling of metabolic rate from molecules and
mitochondria to cells and mammals. Proc Natl Acad Sci USA 99:2473-2478
Whittaker, SG; Zimmerman, FK; Dicus, B; et al. (1989) Detection of induced mitotic chromosome loss in
Saccharomyces cerevisiae—an interlaboratory study. Mutat Res 224:31-78.
Wilcosky, TC; Checkoway, H; Marshall, EG; et al. (1984) Cancer mortality and solvent exposures in the rubber
industry. Am Ind Hyg Assoc J 45:809-811.
Will, O; Mahler, HC; Arrigo, AP; et al. (1999) Influence of glutathione levels and heat shock on the steady state
levels of oxidative DNA base modifications in mammalian cells. Carcinogenesis 20:333-337.
3 02 DRAFT - DO NOT CITE OR QUOTE
-------
Wilson, JG. (1954) Influence of the offspring of altered physiologic states during pregnancy in the rat. Ann NY
Acad Sci 57:517-525.
Wolf, CR; Mansuy, D; Nastainczyk, W; et al. (1977) The reduction of polyhalogenated methanes by liver
microsomal cytochrome P-450. Mol Pharmacol 13:698-705.
Wong, FW; Chan, W; Lee, SS. (1998) Resistance to carbon tetrachloride-induced hepatotoxicity in mice which lack
CYP2E1 expression. Toxicol Appl Pharmacol 153:109-118.
Yasuda, M; Okabe, T; Itoh, J; et al. (2000) Differentiation of necrotic cell death with or without lysosomal
activation: application of acute liver injury models induced by carbon tetrachloride (CC14) and dimethylnitrosamine
(DMN). J Histochem Cytochem 48:1331-1339.
Yoon, M; Madden, MC; Barton, HA. (2007) Extrahepatic metabolism by CYP2E1 in PBPK modeling of lipophilic
volatile organic chemicals: Impacts on metabolic parameter estimation and prediction of dose metrics. J Toxicol
Environ Health, Part A 70:1527-1541.
Yoshida, T; Andoh, K; Fukuhara, M. (1999) Estimation of absorption of trihalomethanes and carbon tetrachloride in
low-level exposure by inhalation pharmacokinetic analysis in rats. Arch Environ Contam Toxicol 36:347-354.
Yoshimine, K; Takagi, M. (1982) Effects of starvation and protein deficiency on the acute carbon tetrachloride -
induced hepatotoxicity. Bull Tokyo Med Dent Univ 29:37-46.
Younes, M; Siegers, CP. (1985) The role of iron in the paracetamol- and CC14-induced lipid peroxidation and
hepatotoxicity. Chem Biol Interact 55:327-334.
Young, RA; Mehendale, HM. (1989) Carbon tetrachloride metabolism in partially hepatectomized and sham-
operated rats pre-exposed to chlordecone (kepone). J Biochem Toxicol 4:211-219.
Yuen, ST; Gogo, AR; Luk, ISC; et al. (1995) The effect of nicotine and its interaction with carbon tetrachloride in
the rat liver. Pharmacol Toxicol 77:225-230.
Yunis, J.J. (1983) The chromosomal basis of human neoplasia. Science 221:227-236.
Zangar, RC; Benson, JM; Burnett, VL; et al. (2000) Cytochrome P450 2E1 is the primary enzyme responsible for
low-dose carbon tetrachloride metabolism in human liver microsomes. Chem Biol Interact 125:233-243.
Zeiger, E; Anderson, B; Haworth, S; et al. (1988) Salmonella mutagenicity tests: IV. Results from the testing of 300
chemicals. Environ Mol Mutagen ll(Suppl. 12): 1-158.
Zeise, L; Wilson, R; Crouch, EAC. (1987) Dose-response relationships for carcinogens: a review. Environ Health
Perspect 73:259-308.
3 03 DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION
The Toxicological Review of Carbon Tetrachloride has undergone a formal external peer
review performed by scientists in accordance with EPA guidance on peer review (U.S. EPA,
2006a, 2000a). The external peer reviewers were tasked with providing written answers to
general questions on the overall assessment and on chemical-specific questions in areas of
scientific controversy or uncertainty. A summary of significant comments made by the external
reviewers and EPA's responses to these comments arranged by charge question follow. In many
cases the comments of the individual reviewers have been synthesized and paraphrased in
development of Appendix A. EPA received no scientific comments on this assessment from the
public.
EXTERNAL PEER REVIEW PANEL COMMENTS
The reviewers made several editorial suggestions to clarify specific portions of the text.
These changes were incorporated in the document as appropriate and are not discussed further.
I. General Comments
1. Is the Toxicological Review logical, clear and concise? Has EPA accurately, clearly and
objectively represented and synthesized the scientific evidence for noncancer and cancer
hazards?
Comments: All six peer reviewers agreed or generally agreed that the Toxicological Review was
logical and clear. One of these reviewers noted that certain aspects of the pharmacokinetic
modeling were not sufficiently described. One reviewer considered the Toxicological Review to
be concise, whereas three reviewers did not. Two of these reviewers pointed to redundancy in
the document and suggested that text be synthesized in a tabular format, or that discussions of
the MOA be shortened with reference to the initial text location. One reviewer suggested
changes for improved accuracy or clarity throughout the Toxicological Review and identified
some relevant references that were not cited. This reviewer considered these errors or lack of
analysis to be relatively minor and ones that might not significantly influence the overall
evaluation of noncancer and cancer hazards (although may modestly influence the specific
uncertainty factors used).
A-1 DRAFT - DO NOT CITE OR QUOTE
-------
Response: A more detailed discussion of the PBPK modeling was provided in the Toxicological
Review (see response to RfC Charge Question #5 for more detailed response). Sections of the
Toxicological Review, and in particular those sections dealing with MO A, were revised to
reduce redundancy. Errors or omissions of references identified by the peer reviewers were
addressed.
Comments: One reviewer identified two main problems with the utilization of the available
evidence: (1) the use of the rate of metabolism per unit liver tissue dose metric for PBPK
modeling with no additional pharmacokinetic correction between species, and (2) the selection of
a doubling of a particular enzyme level as the BMR, to be identified as the functional
replacement for a NOAEL. Another reviewer questioned the large range in the value of the
blood/air partition coefficient in humans and rats.
Response: These comments are addressed in response to comments on RfD Charge Question #3
and RfC Charge Question #5.
Comment: One reviewer noted that CYP enzyme inactivation is more severe in the rat (1
molecule of enzyme lost for every 26 molecules of substrate metabolized in the rat versus 1
molecule of enzyme lost for every 196 molecules of substrate metabolized in the human) and
that a 7.5-fold difference in metabolism-dependent inactivation would be expected to have a
large influence on the extent of carbon tetrachloride bioactivation and potential for increased risk
in humans as compared to rats. This reviewer further observed that a 27% lower Vmax in humans
versus rats (based on Table 3-5, in vitro and in vivo metabolism data for four species) may
mitigate some of the effect of omitting consideration of interspecies differences in rates of CYP
inactivation.
Response: Suicide inhibition is identified in the Toxicological Review as a contributor to
uncertainty in the application of PBPK models to interspecies extrapolation of carbon
tetrachloride toxicokinetics (Section 5.2.2.1); however, the uncertainty was not addressed
quantitatively because information and models to support such an assessment are not currently
available. A discussion of major issues associated with the fact that suicide inhibition of
CYP450 was not explicitly simulated in PBPK models (to predict internal doses of carbon
tetrachloride or to extrapolate external doses across species) was added to Section 3.3,
Metabolism, and Section 5.3, Uncertainties in the Oral Reference Dose and Inhalation Reference
Concentration, under the subheading "Animal to human extrapolation." Based on the analyses
presented in Section 5.3, the model supports the conclusion that suicide inhibition would have
A-2 DRAFT - DO NOT CITE OR QUOTE
-------
relatively minor effects on the extrapolation of carbon tetrachloride external exposures across
species in the low-dose range relevant to the derivation of the RfC.
Comment: One reviewer pointed to the discussion of Yoon et al. (2007) regarding the
extrahepatic metabolism of carbon tetrachloride. This reviewer noted that while rat kidney
cortex and proximal tubules express reasonable levels of CYP2E1 protein and activity for the
oxidative metabolism of the CYP2E1 substrate trichloroethylene, human kidney has been
reported by multiple laboratories to not express any detectable CYP2E1 protein and to exhibit
little if any oxidative metabolism of trichloroethylene. This reviewer acknowledged that because
extrahepatic metabolism is calculated to contribute only a minor proportion to total metabolism
(<1%), this interspecies difference has no significant influence on the conclusions. For the sake
of correctness, however, this reviewer recommended that these interspecies (rodent versus
human) and interorgan (kidney versus liver) differences in CYP2E1 expression and activity be
properly noted. Six references (Cummings and Lash, 2000; Cummings et al., 2001, 2000a,b,
1999; Amet et al., 1997) were provided by this reviewer for consideration.
Response: Pertinent findings from the literature cited by the reviewer were incorporated in
Sections 3.5 and 5.4.3.2.
Comments: One reviewer offered several comments regarding the interpretation of genotoxicity
studies and the strength of the conclusions that were synthesized from those studies. Although
the reviewer did not necessarily disagree with the qualitative conclusions provided in Section
4.4.2, the reviewer suggested stronger statements may be achievable in this section related to the
results of the genotoxicity studies for carbon tetrachloride. Lastly this reviewer suggested brief
descriptions of in vivo mouse strains be added to Section 4.4.2.4.
Response: Section 4.4.2.5 is intended to provide a summary of the genotoxicity literature for
carbon tetrachloride and observations about interpretation of positive and negative findings in
particular bioassays. Conclusions related to carbon tetrachloride's genotoxic potential as it
relates to MOA are presented in Section 4.7.3.4. A brief description of the transgenic mouse
strains was added to Section 4.4.2.4 (Mutations in transgenic mice).
2. Please identify any additional studies that should be considered in the assessment of the
noncancer and cancer health effects of carbon tetrachloride.
Comments: Peer reviewers identified the following additional studies for consideration:
A-3 DRAFT - DO NOT CITE OR QUOTE
-------
Colby et al. (1994) Adrenal activation of carbon tetrachloride: role of microsomal P450 isozymes.
Toxicology 94:31-40.
Eastmond (2008) Environ Mol Mutagen 49:132-141.
Two reviewers identified additional initiation-promotion studies (see below). One of
these reviewers noted that these types of studies should be part of the evidence that carbon
tetrachloride is a well known promoter at high dose, and suggested that this literature be
examined to see if data are available for the evaluation of initiating potential of carbon
tetrachloride in such two-stage designs.
Tsuda, et al. (1993) Jap J Cane Res 84:230-236.
Tsujimura, K; Ichinose, F; Kara, T; Yamasaki, K; Otsuka, M; Fukushima, S. (2008) Toxicol Lett
176(3):207-14.
Bull, RJ; Sasser, LB; Lei, XC. (2004) Interactions in the tumor-promoting activity of carbon tetrachloride,
trichloroacetate, and dichloroacetate in the liver of male B6C3F1 mice. Toxicology 199(2-3):169-183.
One reviewer stated that the following references on renal vs. hepatic CYP2E1 in rats
versus humans and on the human inter-individual variability in CYP expression should be
considered:
Amet, Y; Berthou, F; Fournier, G; Dreano, Y; Bardou, L; Cledes, J; Menez, J-F. (1997) Cytochrome P450
4A and 2E1 expression in human kidney microsomes. Biochem Pharmacol 53:765-771.
Cummings, BS; Lash, LH. (2000) Metabolism and toxicity of trichloroethylene and S-(l,2-dichlorovinyl)-
L-cysteine in freshly isolated human proximal tubular cells. Toxicol Sci 53:458-466.
Cummings, BS; Lasker, JM; Lash, LH. (2000a) Expression of glutathione-dependent enzymes and
cytochrome P450s in freshly isolated and primary cultures of proximal tubular cells from human kidney. J
Pharmacol Exp Ther 293:677-685.
Cummings, BS; Parker, JC; Lash, LH. (2000b) Role of cytochrome P450 and glutathione S-transferase a in
metabolism and cytotoxicity of trichloroethylene in rat kidney. Biochem Pharmacol 59:531-543.
Cummings, BS; Parker, JC; Lash, LH. (2001) Cytochrome P-450-dependent metabolism of
trichloroethylene in rat kidney. Toxicol Sci 60:11-19.
Cummings, BS; Zangar, RC; Novak, RF; Lash, LH. (1999) Cellular distribution of cytochromes P-450 in
the rat kidney. Drug Metab Dispos 27:542-548.
Two reviewers were not aware of any additional studies that should be included in the
assessment.
A-4 DRAFT - DO NOT CITE OR QUOTE
-------
Response: A discussion of the promotion study by Tsujimura et al. (2008) and initiation-
promotion study by Bull et al. (2004) was added to the Toxicological Review (Section 4.4.3). A
summary of Tsuda et al. (1993) was not added. In this study, carbon tetrachloride was used as a
promoter, but the lack of an appropriate control limited the utility of this study for evaluating
carbon tetrachloride promotion properties.
A summary of the findings of Colby et al. (1994) on carbon tetrachloride-induction of
effects on the adrenal gland was added to Sections 4.5 (mechanistic data) and 4.7.4 (MOA for
pheochromocytomas). Citation to the paper by Eastmond was added.
A discussion of renal versus hepatic CYP2E1 in rats and humans and human
interindividual variability in CYP expression based on Amet et al. (1997), Cummings and Lash
(2000), and Cummings et al. (2001, 2000a, b, 1999) was added to Section 3.5.
3. Please discuss research that you think would be likely to increase confidence in the
database for future assessments of carbon tetrachloride.
Comments: The peer reviewers identified the following areas of research to increase confidence
in the database.
Carcinogen!city/chronic toxicity:
• Studies that characterize carcinogenic activity at lower dose levels (i.e., a bioassay with
lower doses, and/or studies evaluating preneoplastic lesion development at lower dose
levels).
• A new oral cancer bioassay with administration of a wide range of doses, including those
below which hepatotoxicity occurs, to eliminate the need for route-to-route extrapolation
as well as to provide better data for RfD estimation.
• Studies on cancer endpoints in CYP2E1 knockout mice (either cancer bioassay or studies
of preneoplastic lesions).
• Repeat of studies where control animals exhibit higher rates of liver cancer than historical
controls.
• Classical initiation-promotion liver studies in which low doses of carbon tetrachloride are
given in conjunction with promoters (partial hepatectomy, phorbol esters, etc.) to
determine whether carbon tetrachloride has initiating potential in rodent liver. One
reviewer suggested such studies using a system described in Tsujimura et al. (2008),
where various amounts/durations of carbon tetrachloride are administered either before
known promoters of liver tumors or after known initiators to improve our information on
dose response for different kinds of cancer-enhancing activities for carbon tetrachloride.
A-5 DRAFT - DO NOT CITE OR QUOTE
-------
• Studies of the mechanism of adrenal tumor induction to understand if parent compound
or a metabolite is the key dose metrics for PK modeling and whether there is a potential
nonlinearity in the dose response. Colby et al. (1994) was identified as a possible useful
start.
Genetic toxicology:
• Assessments of genotoxicity and mutagenicity at lower (non-cytolethal) dose levels to
establish whether DNA damage can really occur at doses relevant to environmental or
occupational exposures.
Kinetic information:
• Measurements in comparable rat and human liver metabolism systems of the rates of
destruction of the reactive metabolites of carbon tetrachloride or steady state
concentrations of those metabolites as indexed by rates of formation of metabolite-
specific adducts. If metabolite elimination rates are in fact slower in people than in
rodents, then steady state concentrations of metabolites should be greater in human than
in the rodent systems for a given rate of metabolite formation. To be fully credible, such
comparisons should be done with fresh liver systems (e.g., slices, isolated hepatocytes)
that preserve as much of the in vivo concentrations of enzymes and cofactors as possible.
• More complete human metabolism data in both liver and extrahepatic tissues.
• More complete analysis of human variation, including genetic polymorphisms, in
enzymes that metabolize carbon tetrachloride, including CYP2E1 and CYP3A4.
Noncancer toxicity:
• Studies of developmental toxicity and reproductive toxicity (including a
multigenerational toxicity study).
• Studies to explore the potential for carbon tetrachloride to be endocrine disruptive
(hormonal mimic or impairment of hormonal systems).
• Studies to elucidate dose-response relationships for more sensitive tests of liver effects,
including cell replication, lipid peroxidation, and s-adenosylmethionine depletion.
• In general, enhanced assessment of toxicity at lower dose levels.
A-6 DRAFT - DO NOT CITE OR QUOTE
-------
Epidemiology:
• Epidemiology studies of exposed workers to follow up on the suggestive evidence of
lymphocytic cancer and further explore the potential for adrenal, liver, and other tumors.
The epidemiology studies may be enhanced by phenotyping individuals for CYP2E1
level and by genotyping individuals for glutathione transferase polymorphisms and for
other factors that may modify anti-oxidant and cellular defense status.
Response: No response needed.
4. Please comment on the identification and characterization of sources of uncertainty in
Sections 5 and 6 of the Toxicological Review. Please comment on whether the key sources
of uncertainty have been adequately discussed. Have the choices and assumptions made in
the discussion of uncertainty been transparently and objectively described? Has the
impact of the uncertainty on the assessment been transparently and objectively described?
Comments: Three peer reviewers believed that the key sources of uncertainty were adequately
discussed. Two reviewers offered specific comments on individual uncertainty factors; these
comments are summarized and addressed in response to RfD Charge Question #4.
One reviewer stated that the characterization of uncertainty in Sections 5 and 6 could be
more complete and more descriptive, and suggested that thought be given to weighting these
uncertainties in terms of how much they affect the confidence in the overall assessment (low,
medium, or high importance). This reviewer identified the following uncertainties not
specifically elaborated in text or tables: 1) dose metric for adrenal tumors; 2) interaction with
other chemicals that may induce or inhibit CYP2E1 or detoxification pathways; 3) disease
processes (specifically diabetes as a condition that could elevate CYP2E1 levels, leading to
additional uncertainty over population variability) and genetic polymorphisms; and
4) uncertainty regarding effect of time-weight averaging exposure in the Japanese inhalation
bioassay and whether hepatotoxicity may be occurring and then repaired at the 5-ppm exposure
level so that the net result is no evidence of toxicity at this dose. This reviewer stated that an
attempt at expressing this on page 238 needed to be made more coherent and further developed
in the uncertainty section.
One reviewer recommended that the implicit assumption of passive destruction of the
reactive metabolites at identical rates in humans and rodents be articulated and raised questions
as to whether the appropriate causal dose metric (gross metabolism rate vs AUC of the active
metabolites) was used and interspecies projections correctly performed.
A-7 DRAFT - DO NOT CITE OR QUOTE
-------
Response: Discussion of uncertainties associated with the dose metrics (both for effects on the
liver and adrenal gland) was expanded in Section 5.3, Uncertainties in the Oral Reference Dose
and Inhalation Reference Concentration, under the subheading "Animal to human extrapolation."
EPA judged that the discussions of interactions of carbon tetrachloride with other chemicals and
susceptibility associated with disease processes and genetic polymorphisms are most
appropriately addressed in Section 4.8, Susceptible Populations and Life Stages, and were not
expanded on in Sections 5 or 6. The exposure regimen used in the JBRC inhalation bioassay
(Nagano et al., 2007b) and the adjustment of the intermittent exposure (6 hours/day) to
continuous 24-hour exposure was taken into account in the PBPK modeling of the experimental
exposure. EPA does not consider this a significant source of uncertainty to be included in
Sections 5 or 6. The comment regarding differences in rates of passive destruction of the
reactive metabolites across species and implications for selection of dose metric is addressed in
response to comments on RfC Charge Question #5.
Chemical-Specific Charge Questions:
(A) Oral reference dose (RfD) for carbon tetrachloride
1. A 12-week oral gavage study in the rat by Bruckner et al. (1986) was selected as the
basis for the RfD. Please comment on whether the selection of this study as the principal
study is scientifically justified. Has this study been transparently and objectively described
in the Toxicological Review? Are the criteria and rationale for this selection transparently
and objectively described in the document? Please identify and provide the rationale for
any other studies that should be selected as the principal study.
Comments: Four reviewers considered the selection of the Bruckner et al. (1986) study to be
scientifically justified and the rationale for its selection clearly explained. One of these
reviewers suggested that the discussion be consolidated to make reasoning for this selection even
more transparent. A fifth reviewer stated that "The choice of the Bruckner et al. (1986) ... study
is not clearly incorrect," but would have considered preferable some integrative calculation
across different data sets rather than the determination of the RfD based on a single study and
single data set within that study. The sixth reviewer stated that the choice of Bruckner et al.
(1986) may in fact be the best choice for RfD derivation, although this reviewer observed that
leakage of SDH may not be the most sensitive indicator of hepatotoxicity and suggested that
consideration be given to low-dose biochemical perturbations (as an additional source of
database uncertainty in RfD derivation). This reviewer noted that evidence suggests that other
A-8 DRAFT - DO NOT CITE OR QUOTE
-------
carbon tetrachloride effects may be detectable at lower doses, although this is not known because
careful dose-response studies for these effects have not been reported down to low doses.
One reviewer's comment concerning the use of 10-week versus 12-week data from the
Bruckner et al. (1986) study is summarized in response to RfD Charge Question #2.
Response: In response to the reviewer that suggested that some integrative calculation across
different data sets would have been preferable, EPA notes that BMD analysis was performed
using data for SDH, OCT and ALT. As detailed in Appendix B, none of the models in BMDS
provided an adequate fit to the OCT data. ALT data provided higher BMD and BMDL values
than did SDH data. In light of the analysis by Travlos et al. (1996) of serum liver enzymes as
predictors of hepatotoxicity that showed SDH to be a more sensitive predictor of
histopathological changes than ALT, EPA considers the BMDL based on SDH data alone to be a
sensitive and appropriate basis for the carbon tetrachloride RfD. EPA performed an integrative
analysis by considering serum enzyme changes in the context of levels that also induced
histopathological changes (see Section 5.1.2). Indicators of hepatotoxicity possibly more
sensitive than increases in SDH could not be considered because no such experimental data have
been collected for carbon tetrachloride.
2. An increase in serum sorbitol dehydrogenase (SDH) activity was selected as the most
appropriate critical effect for the RfD because it is considered by EPA to be an indicator of
hepatocellular injury and a biomarker of an adverse effect. Please comment on whether
the rationale for the selection of this critical effect is scientifically justified. Are the criteria
and rationale for this selection transparently and objectively described in the Toxicological
Review? Please provide a detailed explanation. Please identify and provide the rationale
for any other endpoints that should be considered in the selection of the critical effect.
Comments: Five reviewers agreed that the selection of SDH activity as the critical effect was
scientifically justified, and four of these reviewers specifically considered the selection of the
critical effect to have been adequately explained in the Toxicological Review. A sixth reviewer
stated that within the limits of the testing done and parameters measured, SDH activity may be
the most useful and reasonably sensitive endpoint to date, but noted the uncertainty due to data
gaps in reproductive testing, the potential for low-dose biochemical effects that are part of the
hepatotoxic MO A, and lack of a long-term oral study with adequate sensitivity and
histopathology to test whether liver histopathology could be more sensitive than liver enzyme
leakage for POD selection.
A-9 DRAFT - DO NOT CITE OR QUOTE
-------
One of the reviewers who concurred with selection of SDH activity as the critical effect
questioned why SDH rather than some of the more commonly measured parameters, such as
AST (aspartate aminotransferase) or ALT (alanine aminotransferase), was chosen, and suggested
that the Toxicological Review include a statement as to how typically SDH is used as a metric of
hepatic function.
Two reviewers questioned the exclusion of the data for the 12-week time point (on the
basis that group sizes for the 12-week data were provided as a range of 7 to 9 rats whereas the
group size for the 10-week data was 5 rats). One of these reviewers noted that a range of group
sizes does not make the data unusable and suggested that the BMD analysis be conducted by
assuming 8 animals/group at all doses. A statistical analysis performed by the second reviewer
revealed that the uncertainty in the standard deviations estimated for the 12-week exposure
groups was less than the uncertainty in the estimates of standard deviation for the 10-week
exposures that were selected for BMD analysis (a coefficient of variation [CV] of 27.5% for the
12-week exposure groups versus a CV of 36.7% for the 10-week exposure groups, assuming the
same mean and standard deviation). Therefore, this reviewer recommended that the 12-week
results be used in preference to the 10-week results, or preferably that the BMD calculations be
performed for both periods of exposure and that the results be combined in some reasonable way.
Response: EPA recognizes the possibility that if studies were conducted that involved longer
exposure durations or examined other endpoints of toxicity (e.g., reproductive toxicity endpoints
or low-dose biochemical effects), a more sensitive endpoint for carbon tetrachloride could be
identified. This uncertainty is addressed by the application of uncertainty factors for database
deficiencies (UF = 3) and subchronic to chronic extrapolation (UF = 3).
In response to the question of how typically SDH is used as a metric of hepatic function,
EPA notes that Travlos et al. (1996) reviewed serum enzyme data for 61 13-week toxicity studies
in male and female F344 rats conducted for the NTP by eight contract laboratories following a
standard protocol established by the NTP. Of these 61 studies, SDH was measured in male rats
in 58 of the studies and in females in 57 studies. ALT was measured in male rats in 61 studies
and female rats in 60 studies. This review, while limited to NTP protocols, suggests that SDH is
a commonly measured metric of hepatic function. SDH was selected as the critical effect for the
carbon tetrachloride RfD because it was the most sensitive of the three serum enzymes (SDH,
OCT, and ALT) measured by Bruckner et al. (1986).
An analysis of serum enzyme data collected after 12-weeks of exposure in the Bruckner
et al. (1986) study was added to Appendix B and integrated in Section 5.1.2. As discussed in
Section 5.1.2, the BMDL based on 10-week SDH data provided the lowest POD and was
retained as the basis for the RfD.
A-10 DRAFT - DO NOT CITE OR QUOTE
-------
3. Benchmark dose (BMD) modeling methods were applied to SDH data to derive the point
of departure (POD) for the RfD. Please comment on whether BMD modeling is the best
approach for determining the POD. Has the BMD modeling been appropriately conducted
and objectively and transparently described? Is the benchmark response (BMR) selected
for use in deriving the POD (i.e., an increase in SDH activity two times the control mean)
scientifically justified? Has it been transparently and objectively described? Please
identify and provide rationales for any alternative approaches (including the selection of
the BMR, model, etc.) for the determination of the POD and discuss whether such
approaches are preferred to EPA's approach.
Comments: All peer reviewers considered BMD modeling for use in deriving the POD to be
appropriate.
Three peer reviewers considered the selection of an increase in SDH activity two times
the control mean as the BMR to be scientifically justified. One reviewer considered a twofold
SDH increase to be too large a change to be considered the functional equivalent of a NOAEL.
This reviewer recommended a shift in the mean of 1 standard deviation (SD) (based on
observations in the control group) as the BMR. This reviewer observed that in the current
assessment the SD for the control group is 0.4 x 5°5 = 0.9; 1 SD above the mean would be about
4.4 lU/ml rather than the doubling to 7 lU/ml that was used for the BMR. This reviewer further
observed that a doubling of the group mean enzyme level represents a movement for the average
animal of about 3.5/0.9 = 3.9 SDs. Two reviewers did not provide comments on the selection of
the BMR.
One reviewer's comment concerning the use of 10-week versus 12-week data from the
Bruckner et al. (1986) study is summarized in response to RfD Charge Question #2.
Response: EPA notes that a BMD in not equivalent to a NOAEL. Rather, a BMD (or BMC) is defined
as "a dose or concentration that produces a predetermined change in response rate of an adverse
effect (called the benchmark response or BMR) compared to background" (IRIS glossary at
http://www.epa.gov/ncea/iris/help_gloss.htm#b). The BMR of a twofold SDH increase was
selected consistent with EPA''s Benchmark Dose Technical Guidance Document (U.S. EPA,
2000c), that states that "[i]f there is a minimal level of change in the endpoint that is generally
considered to be biologically significant (for example, a change in average adult body weight of
10%, or the doubling of average level for some liver enzyme), then that amount of change can be
used to define the BMR." As discussed in Section 5.1.2, the scientific literature supports a
twofold increase in liver enzyme levels as a minimally biologically significant change. EPA's
A-11 DRAFT - DO NOT CITE OR QUOTE
-------
BMD guidance further suggests that "a change in the mean equal to one control standard
deviation from the control mean" be used "in the absence of any other idea of what level of
response to consider adverse." For purposes of comparison across chemicals, the BMD and
BMDL corresponding to a change in the mean response equal to one control SD from the control
mean were also calculated for the 10-week SDH data and are presented in Section 5.1.2 and
Appendix B.
4. Please comment on the selection of the uncertainty factors applied to the POD for the
derivation of the RfD. For instance, are they scientifically justified and transparently and
objectively described in the document? If changes to the selected uncertainty factors are
proposed, please identify and provide a rationale(s). Please comment specifically on the
following uncertainty factors:
• An intraspecies (human variability) uncertainty factor of 10 was applied in deriving
the RfD because the available quantitative information on the variability in human
response to carbon tetrachloride is considered insufficient to move away from the
default uncertainty factor of 10.
• A subchronic to chronic uncertainty factor of 3, rather than a default of 10, was
used in light of limited chronic oral study data and more extensive inhalation study
data that informed the progression of toxicity from subchronic to chronic exposure
durations.
• A database uncertainty factor of 3 was used to account for lack of adequate
reproductive toxicity data for carbon tetrachloride, and in particular absence of a
multigeneration reproductive toxicity study.
Are the criteria and rationale for the selection of these uncertainty factors transparently
and objectively described in the document? Please comment on whether the application of
these uncertainty factors has been scientifically justified?
Comments: Five peer reviewers considered the uncertainty factor for intraspecies extrapolation
of 10 and for subchronic to chronic extrapolation of 3 to be scientifically justified; four reviewers
considered the database uncertainty factor of 3 to be scientifically justified. Although the charge
did not include a specific question about the interspecies (animal to human) uncertainty factor,
two reviewers agreed that the uncertainty factor of 10 was appropriate. Four reviewers
considered the criteria and rationale for the selection of the uncertainty factors to be transparent
and objective.
Of the reviewers who considered the intra- and interspecies uncertainty factors to be
appropriate, one reviewer observed that the lack of a PBPK model for refining the RfD
A-12 DRAFT - DO NOT CITE OR QUOTE
-------
derivation requires further explanation given the use of this technique to extrapolate kinetics
across species for inhalation exposure. A second reviewer observed (response to General Charge
Question #4) that the clarity of the text could be improved by listing the applied uncertainty
factors in a separate table with the abbreviations UFA, UFn, UF^ UFs, and UFo.
One reviewer did not consider the rationale for the intraspecies (human variability)
uncertainty factor to be scientifically justified. This reviewer observed that the Toxicological
Review did not seem to account for much of the known information on variation and genetic
polymorphisms in CYP2E1 and CYP3A4 or for the stated differences in rates of enzyme
inactivation in rat and human liver microsomes.
One reviewer considered the justification for the subchronic to chronic uncertainty factor
of 3 provided in the Toxicological Review (i.e., that inhalation studies failed to show a difference
between subchronic and chronic dose response) to be weak. This reviewer observed that
inhalation exposure may not be as sensitive as oral exposure to the buildup of toxicity from
carbon tetrachloride dosing (related to first pass delivery from oral but not inhalation exposure
and to gavage dosing that delivers a higher acute dose compared to inhalation - factors that can
combine to cause the peak exposure at the target site to be greater after oral exposure).
Two reviewers raised questions about the database uncertainty factor, but did not offer an
alternative value for this factor. One of these reviewers noted that it could be argued that the
database uncertainty factor of 3 is too low in light of possible upstream effects in the form of
lipid peroxidation, GSH depletion, macromolecular binding, and derangement in calcium
homeostasis. This reviewer further acknowledged because low-dose mechanistic studies are
unavailable and because the point at which any perturbations might be considered adverse would
be difficult to establish, a database uncertainty factor of 3 can be acceptable under the current
circumstances. The second reviewer pointed to the following language in the discussion of data
gaps - "the absence of these types of studies (i.e., an adequate multigeneration study of
reproductive toxicity) introduces uncertainty... the magnitude of this uncertainty cannot be
quantified" - and asked, if the magnitude of uncertainty due to missing data is unknown, why
would not the default uncertainty factor of 10 be used rather than an uncertainty factor of 3.
Response: Further discussion of the rationale for not applying a PBPK model for RfD derivation
was added to Section 5.1.2.
The abbreviations UFA, UFn, UFL, UFs, and UFo were added to the Toxicological
Review in the discussion of the uncertainty factors.
Additional discussion of CYP450 variation in the human population was added to Section
4.8, Susceptible Populations and Life Stages. Reference to this section was added to the
justification for the intraspecies UF for both the RfD and RfC.
A-13 DRAFT - DO NOT CITE OR QUOTE
-------
With regard to the comment on the subchronic to chronic UF of 3, EPA notes that
inhalation study information that revealed no difference between subchronic and chronic dose
response was only one of the factors that contributed to EPA's determination that a full 10-fold
UF for subchronic to chronic extrapolation was not warranted. Other considerations included
available chronic oral toxicity data and the observation of early onset of toxicity following oral
exposure.
Consistent with input from several peer reviewers, a database UF of 3 was retained. A
database UF of 3 to account for the lack of a multigeneration reproductive toxicity study in the
presence of developmental toxicity information is consistent with EPA practice (U.S. EPA,
2002).
(B) Inhalation reference concentration (RfC) for carbon tetrachloride
1. The JBRC et al. (1998) 2-year inhalation bioassay in the rat was selected as the basis for
the RfC. Please comment on whether the selection of this study as the principal study is
scientifically justified. Has the rationale for this selection been transparently and
objectively described in the Toxicological Review? Are the criteria and rationale for this
selection transparently and objectively described in the document? Please identify and
provide the rationale for any other studies that should be selected as the principal study.
Comments: All six peer reviewers considered the selection of JBRC et al. (1998) as the principal
study to be appropriate.
Response: No response needed.
2. Fatty changes in the liver was selected as the critical effect for the RfC because it is
considered by EPA to be an adverse effect. Please comment on whether the selection of this
critical effect is scientifically justified. Are the criteria and rationale for this selection
transparently and objectively described in the Toxicological Review? Please comment on
whether EPA's rationale about the adversity of the critical effect has been adequately and
transparently described and is supported by the available data. Please provide a detailed
explanation. Please identify and provide the rationales for any other endpoints that should
be considered in the selection of the critical effect.
Comments: Five peer reviewers considered the selection of fatty changes in the liver as the
critical effect to be scientifically justified.
A-14 DRAFT - DO NOT CITE OR QUOTE
-------
A sixth reviewer considered liver effects to be the most sensitive endpoint for deriving an
RfC, and the criteria and rationale for the selection of fatty changes as the critical effect to be
transparently and objectively described in the Toxicological Review. This reviewer suggested
that additional discussion and literature citations be included to firm the association between
fatty liver (seen in this study) and assumed cell damage. This reviewer observed that fibrotic
changes in the liver may be more representative of sustained cellular damage and therefore the
more biologically relevant endpoint, but further that since the NOAEL and LOAEL for fatty
liver changes and fibrosis were the same, selecting fibrosis as the critical effect would not change
the NOAEL and LOAEL values used to derive the RfC.
Response: Fatty liver was retained as the critical effect for the RfC and discussion of the
association between fatty liver and subsequent liver fibrosis and cirrhosis was added to Section
5.2.1.
3. An increase in the severity (but not incidence) of proteinuria in low-dose male and
female rats was reported in the 2-year JBRC (1998) bioassay. Because the biological
significance of this finding in F344/DuCrj rats was considered unclear (see Section 4.6.2 of
the Toxicological Review), proteinuria was not used as the critical effect for the RfC.
Please comment on whether the decision not to use proteinuria as the critical effect is
scientifically sound and has been transparently and objectively described in the
Toxicological Review.
Comments: Five peer reviewers agreed that the decision not to use proteinuria as the critical
effect was scientifically sound and transparently and objectively described. One of these
reviewers further proposed the inclusion of an analysis of the implications for the RfC of using
proteinuria as the basis for calculating an RfC, and recommended the inclusion of alternative
RfC calculations using this endpoint to make the consequence of the choice of liver effects as the
primary focus for the RfC more transparent.
One reviewer considered this decision to be "a questionable call by EPA." This reviewer
stated that arguments against using the chronic proteinuria data are not compelling because
relying upon the subchronic study to dictate the dose response for chronic nephrotoxicity may
underestimate the potential for the kidney to accumulate damage related to carbon tetrachloride.
This reviewer considered proteinuria to be a logical early signal of renal pathology, with the high
frequency in aged animals making interpretation more complex. This reviewer suggested that
the description in Section 4.6.2 show incidence and severity data for this endpoint and related
renal toxicity endpoints to better document the relevance (or lack thereof) of proteinuria to
A-15 DRAFT - DO NOT CITE OR QUOTE
-------
carbon tetrachloride risk. This reviewer observed that the proteinuria data add to the uncertainty
regarding proper selection of the key endpoint.
Response: The implications of not using proteinuria as the basis for the RfC are discussed in
Section 5.3, Uncertainties in the Oral Reference Dose and Inhalation Reference Concentration,
Selection of the critical effect for reference value determination. The text was revised to indicate
that proteinuria occurred at an exposure level fivefold lower than the concentration associated
with fatty liver. Because the dose-response analysis of data for incidence of fatty liver
incorporated BMD and PBPK modeling, the consequence of an alternative analysis using
proteinuria data (without the application of BMD and PBPK modeling) cannot be directly
established. Given the uncertainties in the proteinuria findings in the rat, EPA determined that
an analysis of kidney data using BMD and PBPK modeling is not warranted. EPA notes that a
statement is included in Section 5.3 acknowledging that use of proteinuria data as the critical
effect would have yielded a lower POD than the liver data.
In response to the reviewer who did not find compelling EPA's argument for not using
proteinuria as the basis for the RfD because of subchronic study data considerations, EPA notes
that conclusions about the biological significance of proteinuria were based on a number of
considerations in addition to analysis of the 13-week study findings, including: (1) 100%
incidence of proteinuria in all rats, including the control, (2) >90% incidence of 3+ or 4+
proteinuria in all rats, including the control, (3) lack of progression of proteinuria in the 5-ppm
rats after two years of exposures, i.e., lack of treatment-related increases in incidence or severity
of other renal changes, (4) the occurrence of proteinuria in an animal model (F344 rat) known for
its high incidence of spontaneous, age-related chronic progressive nephropathy that complicates
interpretation of kidney findings, and (5) the body of carbon tetrachloride literature that suggests
that the liver is a more sensitive target organ that the kidney. EPA notes that Table 4-2 in
Section 4.2.2.2 presents all available information on proteinuria from the JBRC study. Thus,
documentation of incidence and severity data for proteinuria was not repeated in the synthesis
section (Section 4.6.2) as recommended by the reviewer; however, reference to Table 4-2 is
provided in this section.
4. BMD methods were applied to incidence data for fatty changes in the liver to derive the
POD for the RfC. Please provide comments on whether BMD modeling is the best
approach for determining the POD. Has the BMD modeling been appropriately conducted
and objectively and transparently described? Has the BMR selected for use in deriving the
POD (i.e., 10% extra risk of fatty liver) been scientifically justified? Has it been
transparently and objectively described? Please identify and provide rationales for any
A-16 DRAFT - DO NOT CITE OR QUOTE
-------
alternative approaches (including BMR, model, etc.) for the determination of the POD and
discuss whether such approaches are preferred to EPA's approach.
Comments: All six peer reviewers considered BMD modeling and the choice of the BMR for use
in deriving the POD to be appropriate.
Response: No response needed.
5. PBPK modeling was used to extrapolate the POD from rats to humans and from
inhalation to oral dose estimates. Please comment on whether the PBPK modeling for
interspecies and route-to-route extrapolation is scientifically justified. Has the modeling
been transparently and objectively described in the Toxicological Review? Does the model
properly represent the toxicokinetics of the species under consideration? Was the model
applied properly? Are the model assumptions, parameter values, and selection of dose
metrics clearly presented and scientifically supported? Has the sensitivity analysis been
clearly presented, and appropriately characterized and considered? Has the uncertainty
been accurately captured and considered?
Comments: Three reviewers considered the application of PBPK models for interspecies
extrapolation to be scientifically appropriate and transparently described.
Response: No response needed.
Comment: One reviewer observed that the description of PBPK modeling applied to extrapolate
animal to human carbon tetrachloride dosimetry (Appendix C) is a potential source of confusion
because of the overwhelming amount of detailed (and sometime redundant) information and
inconsistencies in Section 3.5 (e.g., different values of QCC and QPC).
Response: Inconsistencies in the body of the Toxicological Review and Appendix C were
corrected. The objective of Section 3.5 is to describe models that have been reported in the
literature, whereas Appendix C describes the models and parameter values used in the
implementation of these models in deriving toxicity values. Exact concordance between
parameter values in the two sections is not expected, since some parameter values were selected
based on consideration of multiple factors (e.g., multiple independent estimates of the values),
and units of parameters reported in the literature were not always the same as the units of
parameters implemented in the models described in Appendix C. For example, the reviewer
A-17 DRAFT - DO NOT CITE OR QUOTE
-------
noted a discrepancy between the values for QPC and QCC reported in Section 3.5 (Table 3-4)
based on Paustenbach et al. (1988) and those reported in Table C-2 of Appendix C. Values in
Table 3-4 are for QC and QP (L/hr) for a 0.42-kg rat and 70-kg human (as reported in
Paustenbach et al., 1988), whereas in Table C-2, these values were converted to values for QCC
and QPC (L/hr/kg BW°'74), as described in Paustenbach et al. (1988). The same applies to values
for VMAX reported in Table 3-4; corresponding values for Vmaxc (mg/hr/kg BW°'7) are reported in
Table C-2. The data presentation in Sections 3.5 and Appendix C was made clearer by including
the unsealed values for these parameters in the revised footnote to Table 3-4 (for comparison to
Table C-2).
Data presented in Table C-l were incorrectly cited as a personal communication on page
C-2 (and in Section 3.5); these data were reported in Thrall et al. (2000). The text was revised
accordingly.
Comment: One reviewer pointed to inconsistencies in reporting PBPK model parameterization
between the text (Table 3-4) and Appendix C (Table C-2), in particular with respect to human
cardiac output (QC) and alveolar ventilation (QP).
Response: Values of 256 L/hr for QC and 254 L/hr for QP were derived from Table 2 of
Paustenbach et al. (1988); however, in the text on page 196 of the same publication, the values
are given for QC and QP as 348 L/hr for a 70-kg human, as derived from QCC (or QPC)=15
L/hr x BW°'74. The reason for the discrepancy between the text and Table 2 of Paustenbach et al.
(1988) is not apparent. To improve clarity and comparability between Table 3-4 and Table C-2,
Table 3-4 was revised to present the scaled values from the allometric scaling functions for
cardiac output and alveolar ventilation (i.e., 15 L/hr x BW°'74) and VMAX (i.e., 0.65 mg/hr x
BW07) reported by Paustenbach et al. (1988).
Comment: One reviewer stated that, in light of nonlinear pharmacokinetics and toxicodynamics,
it would be more appropriate to apply the uncertainty factors relevant to animal and human
variability to internal dosimetrics rather than to the predicted human external exposure
concentration of carbon tetrachloride.
Response: The reviewer is suggesting that the uncertainty factor used to account for possible
interspecies differences in pharmacokinetics and/or pharmacodynamics (e.g., 10°5 used in
derivation of the RfC) be applied to the animal internal dosimetry (e.g., MCA, MRAMKL) and
not to the human exposure concentration (HEC). This approach might yield a lower reference
value than if the same uncertainty factor is applied to the human external dose, if the unadjusted
A-18 DRAFT - DO NOT CITE OR QUOTE
-------
human external dose was in the nonlinear range of the external dose-internal dose relationship.
However, as a general principle, the EPA applies uncertainty factors to estimates of HECs for the
following reasons. The HEC is intended to be an estimate of the most likely value for the HEC
equivalent to the POD from the internal dose-response relationship. Uncertainty factors are then
applied to the most likely estimate to account for various categories of uncertainty that might
result in an overestimation of the HEC. Applying all uncertainty factors to the HEC achieves
greater transparency in the quantitative treatment of uncertainty than distributing uncertainty
factors across different points in the derivation of the HEC.
Comment: One reviewer noted that the two rates (MCA and MRAMKL) selected as internal
dosimetry for derivation of the RfC and cancer SF are both time-averaged values. The reviewer
further claimed that the dynamics of the PBPK model prediction were lost because the animal
exposure dosage was also adjusted from 6/24 hours and 5/7 days to the average continuous
exposure of 24 hours/day, 7 days/week.
Response: Inhalation exposures to animals were simulated in PBPK models as 6 hour/day, 5
day/week exposures. Simulations of equivalent human exposures assumed continuous (24
hour/day, 7 day/week) exposures. The text was revised to increase transparency (Section
5.2.2.1).
Comment: One reviewer suggested that the PBPK modeling be made more transparent by listing
the inhalation concentration in the rodent corresponding to the BMD and BMDL.
Response: Tables that present HECs (Tables 5-6, 5-7, and 5-11 though 5-17) were revised to
include external exposure concentrations that correspond to reported BMD and BMDL values.
Comment: One reviewer recommended that the text explain the major rodent-human differences
that yield greater dosimetry in rodents and the confidence one has that these physiologic and
metabolic differences are accurate (e.g., the percentage of body fat and metabolic rate appears to
be backfits).
Response: Values for tissue volumes (i.e., fraction of body weight) and metabolism parameters
were taken from the documentation on the models (i.e., Thrall et al., 2000; Paustenbach et al.,
1988). The metabolism parameters were derived in the above studies from fitting data on closed
chamber elimination kinetics. Paustenbach et al. (1988) also adjusted values for the fat fraction
(VFC) and blood flow (QFC) of rats to improve fit to the gas uptake data. The values used in the
A-19 DRAFT - DO NOT CITE OR QUOTE
-------
Paustenbach et al. (1988) model and in the PBPK models used to derive toxicity values in the
Toxicological Review (VFC, 8%; QFC, 4%) are within the range of reported observations for
these parameters: VFC, 7-15% for adult rats weighing 250-500 g; QFC, 4-7% (Davies and
Morris, 1983; ILSI, 1994).
Results of sensitivity analyses of the two internal dose metrics used in deriving toxicity
values (MCA, MRAMKL) are provided in Appendix C.4. The relative volume of fat (i.e., fat
volume as a fraction of body weight, VFC) was not a sensitive parameter for either dose metric
(sensitivity coefficient <0.01). The metabolism parameter VMAXC was a sensitive parameter
for both dose metrics (sensitivity coefficient >0.1). Increasing VMAXC decreases MCA and
increases MRAMKL.
Section 5.3 (animal to human extrapolation) was revised to discuss the relative
confidence in PBPK model parameter values, including physiological parameters, partition
coefficients, and metabolism parameters.
Comment: One reviewer observed that the blood:air partition coefficients was measured as being
lower in humans than rodents and suggested that the confidence in these data be described as it is
pivotal in creating cross species dosimetry differences. A second reviewer considered it unusual
that there should be a large range of values for the blood:air partition coefficient (2.73 to 4.20 in
humans, Fisher et al., 1997 and Gargas et al., 1989; 4.52 for rats, Gargas et al., 1986).
Response: Confidence in parameter values that are measured in the species being simulated (e.g.,
blood:air partition coefficient) are, in general, considered to be more certain than those that are
extrapolated across species by applying generic allometric scaling factors (e.g., VMAX, KM). The
importance of uncertainty in the estimate of the blood:air partition coefficient depends on the
internal dose metric used in the internal dosimetry modeling. The MRAMKL metric (used as the
basis for the RfC and oral cancer SF) is relatively insensitive to uncertainties in the blood:air
partition coefficient, whereas the MCA metric (used as the basis for the cancer IUR) is highly
sensitive to this partition coefficient. The sensitivity analysis for this and other model
parameters is presented in Appendix C.4, Figures C-14 and C-15.
Although different values for the blood:air partition coefficient were used in the human
and rat models for carbon tetrachloride, these differences were within a range of expected
variability for these parameter values, within and across species. Section 5.3 was revised to
provide additional discussion of literature values for the blood:air partition coefficient.
Comment: One reviewer disagreed with the choice of dose metric in the PBPK model and its
interspecies projection. This reviewer noted that the implicit conclusion made in the assessment,
A-20 DRAFT - DO NOT CITE OR QUOTE
-------
that there will be equal toxic and carcinogenic effects across species for an equal rate of
production of reactive metabolites per unit liver tissue, would be correct if the rates of
destruction of the reactive metabolites across species are the same. The reviewer further noted,
however, that evidence to support this assumption was not provided. This reviewer
recommended that the implicit assumption of passive destruction of the reactive metabolites at
identical rates in humans versus rodents be articulated, together with the mechanistic reasoning
and prior experience with other chemicals that could lead to different assumptions as to the
appropriate causal dose metric (gross metabolism rate versus AUC of the active metabolites) and
interspecies projection rules for causally-relevant delivered dose.
This reviewer further stated that unless both the production and loss of the reactive
metabolites can be included in pharmacokinetic models based on reasonable empirical data, EPA
should apply the BW"° 25 correction to account for likely slower elimination of the active
metabolites in humans relative to rats.
Response: This comment applies to the carbon tetrachloride RfC and IUR, both of which were
derived using a PBPK model for interspecies extrapolation and rate of metabolism of carbon
tetrachloride in the liver as the dose metric. EPA acknowledges uncertainty in the assumption of
equal toxic and carcinogenic effects in liver across species for an equal rate of production of
metabolites per unit liver. Species differences could arise from various mechanisms, including
quantitative differences in clearance of reactive metabolites of carbon tetrachloride and
quantitative differences in mechanisms that participate in quenching lipid peroxide cascades
and/or repairing lipid peroxides (e.g., glutathione peroxidase), that scavenge or reduce oxygen
radicals (e.g., superoxide dismutase, GSH), or that repair DNA damage (e.g., glycolases, ligases,
polymerases).
Nevertheless, empirical data specific to carbon tetrachloride metabolism indicate that
equal rates of metabolism of carbon tetrachloride by CYP450 in rodents and humans would be
expected to yield equal rates of elimination of trichloromethyl and trichloromethyl peroxy
radicals. Sections 3.3 and 5.2.2.1 were revised to include a discussion of the generation and
elimination of reactive metabolites of carbon tetrachloride. Based on the considerations
presented in Section 5.2.2.1, EPA determined that a reasonable modeling approximation was to
simulate the elimination of the trichloromethyl radical, in both rodents and humans, as occurring
with the same, high rate relative to the much slower production of the radical. This is analogous
to a flow-limited system, in which the amounts of reaction products of the trichloromethyl
radical produced over time (i.e., AUC) are limited by the rate of production of the
trichloromethyl radical (i.e., via CYP450) and the availability of reactants for the trichloromethyl
and trichloromethyl peroxy radicals (e.g., intracellular amino acids, lipid, protein in the liver). In
A-21 DRAFT - DO NOT CITE OR QUOTE
-------
carbon tetrachloride PBPK models applied in the current assessment, interspecies scaling of the
production of the trichloromethyl radical is modeled with species-specific values for Michaelis-
Menten rate coefficients for CYP450-mediated metabolism of carbon tetrachloride (i.e., VMAX
scaled to BW° 7). Tissue concentrations of reactants for the trichloromethyl and trichloromethyl
peroxy radicals (e.g., amino acids, lipid, protein) are assumed to be the same in rodent and
human liver. Therefore, the AUC for the concentration of trichloromethyl and trichloromethyl
peroxy radicals in liver would be expected to scale with the rates of production of metabolism of
carbon tetrachloride to the trichloromethyl radical in liver, which are simulated in both species,
and with liver volumes, which scale directly with body weight (i.e., liver volume is assumed to
be 0.04 of body weight). Given the highly reactive nature of carbon tetrachloride and the
available rate constant information for carbon tetrachloride metabolites, the additional scaling
factor of the elimination rate proposed by the reviewer (i.e., BW"° 25) is not necessary. Scaling
dosimetry of reactive metabolites that are eliminated by spontaneous processes (i.e., not
metabolism) directly with body weight (i.e., BW1) has been discussed elsewhere (e.g., Travis,
1990). It is emphasized that the determination not to apply an additional scaling factor of
BW"°25 was based on the strength of the available carbon tetrachloride data and information on
the biochemical reaction mechanism and should not be construed as precedent for other
compounds where such data and information are not available.
Comment: One reviewer observed that the HECs obtained from VMAXC values of 0.4 and 0.65
were averaged. This reviewer noted that using the lower value might be considered more
conservative, but that use of either the average of the two HECs or the lower value yielded the
same RfC when rounded to one significant figure.
Response: The rationale for averaging, i.e., that there is no empirical basis for selecting either
end of the range as the more likely estimate for the RfC, is provided in Section 5.2.2.3.
Comment: One reviewer pointed to the absence of explicit source codes of PBPK model(s) used
in POD extrapolations for derivation of the RfC (both, CSL and CMD files).
Response: ACSL csl file, and m files for the human, rat, and mouse models were added as a new
Appendix F.
6. Please comment on the selection of the uncertainty factors applied to the POD for the
derivation of the RfC. If changes to the selected uncertainty factors are proposed, please
A-22 DRAFT - DO NOT CITE OR QUOTE
-------
identify and provide a rationale(s). Please comment specifically on the following
uncertainty factors:
• An intraspecies (human variability) uncertainty factor of 10 was applied in deriving
the RfC because the available quantitative information on the variability in human
response to carbon tetrachloride is considered insufficient to move away from the
default uncertainty factor of 10.
• An interspecies uncertainty factor of 3 was used to address pharmacodynamic
uncertainty only, because PBPK modeling was used to address pharmacokinetic
extrapolation from rodents to humans. This contrasts with using the full default
interspecies uncertainty factor of 10 for the RfD where an oral PBPK model to
support interspecies extrapolation is not available.
• A database uncertainty factor of 3 was used to account for lack of adequate
reproductive toxicity data for carbon tetrachloride, and in particular absence of a
multigeneration reproductive toxicity study.
Are the criteria and rationale for the selection of these uncertainty factors transparently
and objectively described in the document? Please comment on whether the application of
these uncertainty factors has been scientifically justified?
Comments: All six peer reviewers agreed with the application of an intraspecies uncertainty
factor of 10. One of these reviewers noted while information is available regarding CYP
expression that was not considered and that could mitigate some of the variability, the choice of
the default UF of 10 is probably reasonable based on the desire to err on the side of
conservatism.
Five reviewers agreed with the application of an interspecies uncertainty factor of 3. One
of these reviewers suggested that a discussion of whether the use of an uncertainty factor of 3
was adequate for the interspecies extrapolation from rat to hamster be included to provide
support for the use of an interspecies uncertainty factor. A sixth reviewer reiterated that a
BW"°25 correction should be added to account for likely slower elimination of the active
metabolites in humans relative to rats, which would lower the RfC by a factor of about fourfold
[(70/0.25)-0'25].
Five reviewers agreed with the application of a database uncertainty factor of 3. A sixth
reviewer observed that there may be sufficient uncertainty with regards to proteinuria being the
driving endpoint instead of fatty liver to increase the database uncertainty factor to 10. In lieu of
this, this reviewer noted that EPA could model the proteinuria data to study the implications of
this apparent lowest LOAEL and either use that determination directly for RfC derivation, or use
it to further inform the magnitude of the database UF.
A-23 DRAFT - DO NOT CITE OR QUOTE
-------
Two reviewers specifically offered the opinion that the criteria and rationale for the
selection of the uncertainty factors were transparent and objective. A third offered the same
opinion, with the exception of the failure in the document to correct for the likely difference in
detoxification rate of the active metabolites in humans versus rodents.
One reviewer reiterated the suggestions on improvement of the clarity of presentation of
uncertainty factors (see RfD Charge Question #5) and further recommended that uncertainty
factors be applied to the internal dosimetric rather than to the predicted human external exposure
concentration of carbon tetrachloride.
Response: With respect to the interspecies uncertainty factor, EPA does not consider a discussion
of interspecies extrapolation from rat to hamster would provide relevant support for the
interspecies uncertainty factor applied to account for uncertainty in the extrapolation from data in
rats to humans. The comments related to application of body weight scaling for interspecies
extrapolation and application of uncertainty factors to the internal dosimetric are addressed in
response to comments on RfC Charge Question #5. The comment related to uncertainty
associated with data for proteinuria is addressed in response to comments on RfC Charge
Question #3. The comment related to correction for likely differences in detoxification rate of
the active metabolites in humans versus rodents is addressed in response to comments on
General Charge Question #1. The comment regarding the clarity of the uncertainty factor
presentation and application is addressed in response to comments on RfD Charge Question #5.
The abbreviations UFA, UFn, UFL, UFs, and UFo were added in the discussion of the uncertainty
factors applied in deriving the RfC.
(C) Carcinogenicity of carbon tetrachloride
1. Under EPA's 2005 Guidelines for Carcinogen Risk Assessment
(www.epa.gov/iris/backgr-d.htm), the Agency concluded that carbon tetrachloride is likely
to be carcinogenic to humans by all routes of exposure. Please comment on the cancer
weight of evidence characterization. Has the scientific justification for the weight of
evidence descriptor been sufficiently, transparently and objectively described? Do the
available data for both liver tumors in rats and mice and pheochromocytomas in mice
support the conclusion that carbon tetrachloride is a likely human carcinogen? Has the
scientific justification for deriving a quantitative cancer assessment been transparently and
objectively described?
A-24 DRAFT - DO NOT CITE OR QUOTE
-------
Comments: Four peer reviewers agreed with the conclusion that carbon tetrachloride is likely to
be carcinogenic to humans by all routes of exposure.
One reviewer expressed concerns about the overall conclusion that carbon tetrachloride
should be considered "likely to be carcinogenic to humans by all routes of exposure," as it was
not clear how this related to the previous assessment from 1991 that assigned a weight-of-
evidence descriptor of "probably a human carcinogen." This reviewer stated that the previous
conclusion was based on sufficient evidence in animals whereas the newly proposed designation
of "likely to be carcinogenic to humans" would be based on sufficient evidence in animals and
humans. This reviewer further observed that considering that liver effects are considered to be
primary, it was unclear how the absence of liver tumors in humans could be reconciled with the
designation of "likely to be carcinogenic in humans."
Another reviewer did not specifically offer a comment on the cancer weight-of-evidence
descriptor. Rather, this reviewer offered the opinion that it is logical to postulate that
hepatocarcinogenicity could be mechanistically relevant to humans, but believed there is no such
a parallelism with mouse pheochromocytomas, and that the conclusion that "... experimental
evidence for pheochromocytomas is potentially relevant to humans..." bears a great degree of
uncertainty. This reviewer suggested that the uncertainty regarding pheochromocytoma in
humans should be better emphasized in Section 6.2.3, Relevance to humans.
Response: As noted in EPA's 2005 Guidelines for Carcinogen Risk Assessment, evidence
consistent with the descriptor of "likely to be carcinogenic to humans" covers a broad range of
data combinations, including, for example, "an agent that has tested positive in animal
experiments in more than one species, sex, strain, site, or exposure route, with or without
evidence of carcinogenicity in humans" (U.S. EPA, 2005a, p. 2-55). The cancer findings for
carbon tetrachloride, which include tumors in three species (rat, mouse and hamster), two sites
(liver and adrenal gland), and two routes of exposure (oral and inhalation), are consistent with
this cancer weight-of-evidence descriptor.
In response to the reviewer who suggested that the uncertainty regarding
pheochromocytomas in humans be better emphasized in Section 6.2.3, Relevance to humans,
EPA notes that the fact that pheochromocytomas were observed only in mice does not
necessarily lead to the conclusion of uncertain relevance to humans. As noted in the Guidelines
for Carcinogen Risk Assessment (U.S. EPA, 2005a), "agents observed to produce tumors in both
humans and animals have produced tumors either at the same site... or different sites," and
therefore "site concordance is not always assumed between animals and humans," particularly
where MOA information does not lead to an anticipation of site concordance. EPA considers
the discussion of uncertainties related to the relevance of mouse pheochromocytomas to humans
A-25 DRAFT - DO NOT CITE OR QUOTE
-------
to be appropriately captured in Section 6.2.3, Relevance to humans, where it is acknowledged
that the relevance is unknown, but that the mouse has been identified as a potentially appropriate
model for human adrenal medullary tumors.
2. In the Toxicological Review, EPA discussed a MOA for liver cancer involving
metabolism, cytotoxicity, and regenerative proliferation leading to tumor induction as key
events occurring at relatively high exposure levels. EPA also discussed that carbon
tetrachloride carcinogenicity may not be explained by a cytotoxic-proliferative MOA only
and that a MOA involving genetic damage may also be operative at high exposure levels
and may predominate at noncytotoxic (low) exposures. Please provide detailed comments
on whether this analysis regarding carbon tetrachloride's MOA(s) is scientifically justified.
In particular, please provide comments on EPA's evaluation of the carbon tetrachloride
genotoxicity database and EPA's judgments about potential low-dose genotoxicity given
the limited information at low doses. Has the MOA for liver cancer been transparently and
objectively described in the document? Considerations should include the scientific
support regarding the plausibility for each of the hypothesized MO As, and the
characterization of uncertainty regarding these MOAs.
Comments: The peer reviewers offered a range of opinions on EPA's presentation of
hypothesized cancer MOAs for carbon tetrachloride. Three reviewers generally agreed that the
inclusion of hypothesized MOAs at high and low doses is appropriate. One of these reviewers
observed that the various MOA discussions in the document tended to emphasize the high-dose
phenomena, with the low-dose MOA discussion brought in secondarily mostly to explain one
data point rather than as a primary mechanism with sufficient footing to drive low-dose
extrapolation. A second reviewer emphasized that reactive metabolites are expected to be
formed at low and high doses. A third reviewer noted that available evidence supporting the
cancer MOA involving hepatic cytotoxicity, necrosis and cellular regeneration is well presented
and more convincing that a MOA involving genetic damage, but both MOAs appear to
contribute.
Three reviewers questioned whether a second MOA involving low-dose genetic damage
was adequately scientifically supported. One of these reviewers considered that the discussion of
MOAs involved in hepatocarcinogenicity of carbon tetrachloride considered only the two
extreme alternatives in a somewhat simplistic manner, e.g., either cytotoxicity/regeneration or
genetic damage, and avoided discussing the epigenetic mechanisms that this reviewer believed
were most probably involved in both cancer and noncancer effects caused by environmentally-
relevant, low concentrations of carbon tetrachloride, a pro-oxidant chemical. This reviewer
A-26 DRAFT - DO NOT CITE OR QUOTE
-------
believed that an increased proliferation rate appears earlier and at significantly lower
concentrations than those needed for noticeable cytotoxicity, and further that the biomarkers of
cellular proliferation relate to the dose of pro-oxidant nonlinearly. This reviewer concluded that
it is likely that the epigenetic mechanisms (e.g., oncogene depression or activation) rather than
genotoxicity or necrosis/regeneration may be responsible for carcinogenicity observed in carbon
tetrachloride-treated animals and that, therefore, genotoxicity at low exposures is not a plausible
MO A for carbon tetrachloride.
The second of these three reviewers concluded that the preponderance of data for carbon
tetrachloride supports a MO A for liver tumors that includes the following key events:
(1) metabolism to reactive intermediates, (2) radical-induced mechanisms leading to
hepatocellular toxicity, and (3) sustained regenerative and proliferative changes, and that these
key events are consistent with a hypothesis that exposures that do not cause hepatotoxicity are
not expected to result in liver cancer. This reviewer stated that the scientific basis for this MOA
and the characterization of uncertainties for this MOA were adequately addressed and described
in the Toxicological Review.
The third reviewer believed that there are ample data to support a MOA involving
cytotoxicity-proliferation. With the exception of the unexplained hepatocellular adenomas in
female mice at low doses, this reviewer knew of no data that support any other mechanism of
action. While appropriate to suggest an additional mechanism to be consistent with unexplained
data, this reviewer was not sure that the mouse data provide a strong rationale for an alternate
MOA.
Response: Section 4.7 was rewritten to more clearly articulate the hypothesized liver tumor
MOA at high and low exposure levels. One characteristic of the mechanistic database that was
evaluated in the MOA analysis is that the majority of available studies were conducted at
relatively high doses (Table 4-16).
Evidence for an epigenetic component to the cancer MOA was added to Section 4.7.3.3.
Support for MO As other than a cytotoxi city/regenerative proliferation MOA is not
limited to the incidence of female mouse liver tumors (Nagano et al., 2007b). Other
considerations that suggest that the carbon tetrachloride database is insufficient to rule out other
MO As at low exposure levels include: (1) carbon tetrachloride's general reactivity (i.e., carbon
tetrachloride is metabolized to the reactive species trichloromethyl and trichloromethyl peroxy
radical that can react with cellular constituents and lead to formation of reactive oxygen species
that also can damage DNA and other macromolecules) and (2) insufficient data to ascertain
whether or not carbon tetrachloride is genotoxic at low exposures.
A-27 DRAFT - DO NOT CITE OR QUOTE
-------
Comments: Two of the six reviewers provided specific opinions related to the weight of evidence
for genotoxicity. One of these reviewers stated that the genetic toxicology database is overall not
supportive of mutagenesis as being a primary mechanism; however, given important
uncertainties in the genotoxicity studies, the fact that in a few studies carbon tetrachloride has
showed a genotoxic effect in the absence of S-9 mix, and the likely formation of radicals at sub-
toxic doses, this reviewer thought some consideration should be given to genotoxicity as the
explanation for the tumor response in the JBRC study at a relatively low, non-toxic levels. The
second reviewer believed that using a weight-of-evidence approach, the scientific data show that
carbon tetrachloride is not genotoxic or mutagenic and therefore did not agree with the
conclusions drawn in the Toxicological Review concerning potential for low-dose genotoxicity
of carbon tetrachloride. This reviewer also suggested carbon tetrachloride could induce a
hormetic response in that moderate "priming doses" of liver toxicants such as carbon
tetrachloride can induce detoxifying and/or DNA repair enzymes and reduce or prevent cellular
damage caused by carbon tetrachloride.
Response: EPA maintains that, given the highly reactive biological activity of carbon
tetrachloride and demonstration of a genotoxic response at high-exposure levels, the contribution
of genotoxicity to the cancer MOA for carbon tetrachloride cannot be excluded. Significant
literature that suggests that carbon tetrachloride induces a hormetic response is unavailable. The
three citations on hormesis provided by the peer reviewer are papers on chloroform or a general
review of hormesis and are not specific to carbon tetrachloride.
3. Regarding liver cancer, two approaches to dose-response assessment for the inhalation
exposure route are presented in the Toxicological Review—a nonlinear low-dose approach
and a linear low-dose extrapolation approach. Do you agree with EPA regarding the
support for a nonlinear extrapolation approach consistent with a MOA involving
hepatocellular cytotoxicity and regenerative hyperplasia? Do you agree with EPA
regarding the support for applying the default linear extrapolation approach due to
uncertainty in understanding the cancer MOA at low doses? Please provide detailed
comments on whether the inclusion of both approaches to dose-response assessment is
scientifically sound and transparently and objectively described in the document.
Comments: Three reviewers generally agreed with the presentation of both linear and nonlinear
approaches in the assessment. One of these reviewers concluded that given the suggestive
evidence of low-dose carcinogenesis below toxicity thresholds and uncertainties with respect to
genotoxicity, the recommendation in the Toxicological Review of a linear low-dose modeling
A-28 DRAFT - DO NOT CITE OR QUOTE
-------
approach is a prudent way to deal with the uncertainties in a reasonably health protective
manner. This reviewer suggested that the Toxicological Review attempt to bring these different
approaches together into a unified synthesis and provide perspective on the difference between
the approaches (i.e., that if one chose the nonlinear approach, one would be out of bounds for
protecting public health if in fact the low-dose linear model is correct). The second of these
reviewers thought it appropriate to present the low-dose linear approach, but considered it to be a
default approach with little scientific support, whereas the nonlinear extrapolation approach has a
good deal of scientific support from the literature. This reviewer believed that the document fell
short in not making some judgment as to the relative strength of the two proposed approaches.
The third reviewer believed that both nonlinear and linear approaches were well described in the
document and that a well-balanced explanation of the support and deficiencies for both methods
was clearly presented. This reviewer further noted that while the nonlinear extrapolation
approach appears more consistent with the MOA involving hepatocellular cytotoxicity and
regenerative hyperplasia, the default linear may also be considered given the uncertainty in
understanding the cancer MOA at low doses.
One reviewer did not agree with the application of the default linear extrapolation
approach due to uncertainty in understanding the cancer MOA at low doses. Rather, this
reviewer believed that the available data supported key events involving hepatocellular
cytotoxicity and regenerative hyperplasia consistent with a nonlinear MOA. Further, this
reviewer questioned the biological significance of female mouse liver tumors (18%) at 5 ppm in
the JBRC study that was statistically significantly elevated relative to historical but not study
controls, and that was lower than the incidence produced by 25 ppm (88%) and 125 ppm (98%)
carbon tetrachloride in male mice or by 125 ppm in either male or female rats (80-88%). This
reviewer also noted that epidemiological studies have not identified an association between
human exposures to carbon tetrachloride and increased liver cancer incidence.
In contrast, another reviewer disagreed with a presentation of the nonlinear threshold-
implying calculations. This reviewer suggested that where linear and upward-turning nonlinear
MO As are present in the same system, the dose response in the low-dose region will tend toward
linearity. In this case, EPA should therefore do the best it can to estimate the low-dose slope.
The reviewer appended an extended excerpt from a 2007 white paper prepared for EPA
discussing relevant issues. In addition, this reviewer proposed the depletion of S-adenosyl
methionine (SAM) as one of the likely components of the carbon tetrachloride MOA mentioned
in the Toxicological Review. This reviewer noted as relevant the discussion of the dose-
response relationship for dichloroacetate (DCA), which is also thought to act via this process
(and cited: Hattis, D, Rahmioglu, N, Verma, P, Hartman, K, Kozlak, M, and Goble, R. A
preliminary operational classification system for non-mutagenic MO As for carcinogenesis.
A-29 DRAFT - DO NOT CITE OR QUOTE
-------
Critical Reviews in Toxicology, 2008, in press.) According to this reviewer, the proposed MOA
for DCA is decreased methylation of the promoter regions of the proto-oncogenes c-jun and c-
myc and increased expression of the corresponding m-RNAs. Data from a 2-year carcinogenesis
dose-response study for DCA did not indicate appreciable nonlinearity over the fairly wide dose
range studied. A sixth reviewer did not provide a specific opinion as to whether the inclusion of
both approaches to dose-response assessment is supported. This reviewer raised questions as to
the support for a linear approach, noting that under the assumption that the short-lived free
radical metabolites of carbon tetrachloride and the peroxidative products are responsible for
hepatocarcinogenicity, it may be unrealistic to expect a linear proliferative response versus time-
averaged integrated carbon tetrachloride dosimetrics. Given this and consistent with both
epigenetic cancer and noncancer MOA (e.g., represented by the RfD/RfC), this reviewer
considered a nonlinear approach to dose-response to be more appropriate and more relevant to
potential hepatocarcinogenesis than the linear extrapolations or even a simplified MOA
involving hepatocellular cytotoxicity and regenerative hyperplasia.
Response: Consistent with the overall input received from the peer reviewers, a default linear
approach and a nonlinear approach (as an alternative low-dose extrapolation approach) were
retained.
With respect to the recommendation suggested by two reviewers that the assessment
provide either a unified synthesis of the linear and nonlinear approaches or judgment as to the
relative strength of the two approaches, EPA notes that providing the risk at an exposure
equivalent to the RfD or RfC under the assumption of the linear low-dose extrapolation approach
is essentially conducting a risk assessment for an exposure scenario of lifetime exposure at that
exposure level, and thus falls outside the scope of an IRIS Toxicological Review. Further, EPA
does not believe that providing judgments as to the relative strength of the two approaches is
scientifically supported. In response to the reviewer who disagreed with the application of the
default linear extrapolation on the basis that data support key events consistent with a nonlinear
MOA, EPA notes that the incidence of hepatocellular adenomas in 5-ppm female mice in the
JBRC study was, in fact, statistically elevated relative to both historical and study controls and
that the combined incidence of hepatocellular adenomas and carcinomas showed a positive trend.
As this reviewer observed, epidemiological studies have not identified an association between
human exposures to carbon tetrachloride and increased liver cancer incidence; however, no case-
control studies were identified that specifically looked for this association. Further, EPA notes
that site concordance is not necessarily assumed between animals and humans (U.S. EPA,
2005a).
A-3 0 DRAFT - DO NOT CITE OR QUOTE
-------
In response to the reviewer who provided a white paper entitled, "Uncertainties in Risk
Assessment for Carcinogenesis: A Road Map Toward Practical Improvements," EPA notes that
much of the discussion in this white paper extends beyond the scope of the carbon tetrachloride
assessment, although some of the concepts that support the application of a low-dose linear
extrapolation are presented in Section 5.4, Cancer Assessment.
Additional discussion of the role of SAM depletion in the carbon tetrachloride cancer
MOA was added to Section 4.7.3.3.
Additional responses to comments regarding EPA's decision to recommend the linear
low-dose extrapolation approach for assessing cancer risk are provided in response to
Carcinogenicity Charge Question #8.
4. Is EPA's characterization of mouse pheochromocytomas, including their relevance to
human cancer risk, transparently and objectively described in the Toxicological Review?
EPA applied a linear extrapolation approach to pheochromocytoma data from the JBRC
inhalation bioassay in mice in the absence of MOA information. Please comment on the
scientific justification for quantification of cancer risk for this tumor type, considering
relevance to humans. Has the dose-response modeling been appropriately and objectively
conducted? Are the results objectively and transparently described?
Comments: Two peer reviewers agreed with the characterization of pheochromocytomas as
relevant to humans, the dose-response assessment, and with the characterization of the
uncertainty in dose response. One of these reviewers expressed the opinion that the fact that the
tumors are benign does not materially detract from their relevance as indicators of a carbon
tetrachloride-induced carcinogenic process. This reviewer cited Colby et al. (1994) as additional
evidence for carbon tetrachloride activation in the rodent and suggested that additional literature
search and evaluation be conducted regarding the potential MOA for adrenal tumors. This
reviewer further suggested that greater emphasis be placed on the uncertainties in PBPK
modeling and cancer risk estimation for this endpoint given the lack of MOA information and
uncertainty with regards to the key dose metric for estimating internal dose and risk. The second
reviewer believed it would be informative to include an alternative linear low-dose model
estimate based on liver tumors only and addition of a statement about the fraction of animals
with pheochromocytomas that also had liver tumors. Finally, this reviewer thought it would be
useful to put the result in perspective by showing where the carbon tetrachloride slope factor fits
among the slope factors calculated for other small molecular weight chlorinated hydrocarbons
(e.g., vinyl chloride, methylene chloride).
A-31 DRAFT - DO NOT CITE OR QUOTE
-------
Three peer reviewers did not agree with EPA's approach to pheochromocytomas. Two
reviewers observed that pheochromocytomas may represent a strain-specific finding and that no
increases in pheochromocytomas have been observed in epidemiological studies. For these
reasons, one of these two reviewers did not consider a linear extrapolation based on
pheochromocytomas in mice to be justified. A second reviewer did not agree that data for
pheochromocytomas should override the conclusions based on the use of liver tumors as the
primary response because: (1) pheochromocytomas have been observed at higher doses than
those that cause liver tumors; (2) the relevance to humans is questionable as this tumor has not
been previously observed in carbon tetrachloride-exposed individuals; and (3) these tumors are
almost always benign.
One reviewer did not directly address the charge question. This reviewer observed that
the issue of mouse pheochromocytomas was adequately described qualitatively and
quantitatively and characterized sufficiently in this Toxicological Review; however, their
relevance to human cancer risk was considered highly uncertain.
Response: A search of the literature for additional information on the effects of carbon
tetrachloride on the adrenal gland was performed. Relevant findings, including those of Colby et
al. (1994), were added to Sections 4.5 (mechanistic data) and 4.7.4 (MOA for
pheochromocytomas).
Uncertainties in the PBPK modeling for pheochromocytomas, and in particular selection
of the dose metric, given the lack of MOA information for this tumor were addressed in Section
5.4.3.2.
Estimates of cancer risk using a linear low-dose extrapolation approach are presented for
all individual tumor types in Tables 5-18 and 5-19. A statement about the fraction of animals
with pheochromocytomas that also had liver tumors has no bearing on the estimate of cancer risk
and is not necessary.
The presentation of cancer slope factors or unit risks for other small molecular weight
chlorinated hydrocarbons is outside the scope of the Toxicological Review for carbon
tetrachloride. EPA disagrees with reviewers who considered the use of pheochromocytoma data
for cancer dose-response modeling to be unsupported on the grounds that the tumor may be
specific to the mouse, that epidemiological evidence for an association between carbon
tetrachloride exposure and pheochromocytomas is absent, and that the tumors are almost always
benign. As discussed in Section 5.4.2.1, pheochromocytomas seen in the mouse have a human
equivalent. In humans, pheochromocytomas are rare catecholamine-producing neuroendocrine
tumors that are usually benign, but may also present as or develop into a malignancy (Eisenhofer
et al., 2004; Salmenkivi et al., 2004; Tischler et al., 1996). Hereditary factors in humans have
A-32 DRAFT - DO NOT CITE OR QUOTE
-------
been identified as important in the development of pheochromocytomas (Eisenhofer et al., 2004).
Although epidemiological studies have not been conducted to investigate whether an association
exists between carbon tetrachloride exposure and pheochromocytomas, some evidence supports
a conclusion that experimental evidence for pheochromocytomas is potentially relevant to
humans. Parallels between this tumor in the mouse and humans led investigators to concluded
that the mouse might be an appropriate model for human adrenal medullary tumors (Tischler et
al., 1996). Like the human, pheochromocytomas in the mouse are relatively rare, as are
metastases. Both the morphological variability of the mouse pheochromocytomas and the
morphology of the predominant cells are comparable to those of human pheochromocytomas.
Finally, an important characteristic of mouse pheochromocytomas is expression of
immunoreactive phenylethanolamine-N-methyltransferase (PNMT); human pheochromocytomas
are also usually PNMT-positive (Tischler et al., 1996). Considering what is known about
pheochromocytomas in rodents and humans and in the absence of MO A information, there is no
basis upon which to conclude that these tumors are not relevant to humans.
Finally, pheochromocytomas in the mouse associated with carbon tetrachloride exposure
cannot necessarily be characterized as benign only. One pheochromocytoma in the JBRC
bioassay (Nagano et al., 2007b) was identified as malignant, and the oral NCI bioassay
(Weisburger, 1977; NTP, 2007) did not indicate whether pheochromocytomas were benign or
malignant. Salmenkivi et al. (2004) observed that while most pheochromocytomas are benign,
differentiating a benign from a malignant tumor only by histological criteria is difficult.
Therefore, EPA does not consider it appropriate to exclude pheochromocytomas from analysis of
cancer risk because they are often benign.
5. Nonlinear approach: The Toxicological Review finds that the RfD of 0.004 mg/kg-day
and the RfC of 0.1 mg/m3 be used to assess liver cancer risk for carbon tetrachloride under
the assumption of a MOA consistent with low-dose nonlinearity. Please provide detailed
comments on whether this nonlinear approach is scientifically justified. Has this approach
been transparently and objectively described in the document? Are there other nonlinear
approaches to evaluating liver cancer risk for carbon tetrachloride that should be
presented in the Toxicological Review? Please comment on the utility of including these
alternative nonlinear approaches. Please comment on the confidence that EPA should have
that there is not a cancer risk for exposures below the level of the RfD/RfC.
Comments: Four peer reviewers generally considered the nonlinear approach to be appropriately
presented in the Toxicological Review and to be the preferred and more scientifically supported
approach for carbon tetrachloride cancer assessment. One of these reviewers observed that while
A-3 3 DRAFT - DO NOT CITE OR QUOTE
-------
the linear approach is health-protective, it can result in exaggerated risk estimates in comparison
to the alternative approach (i.e., an epigenetic approach discussed by this reviewer in response to
Cancer Charge Question #2), which this reviewer considered insufficiently explored in this
Toxicological Review. The other three reviewers did not identify alternative nonlinear
approaches that should be applied to characterize liver cancer risk from carbon tetrachloride.
It was the opinion of two reviewers that the choice of the low-dose linear approach was
the most clear, prudent and scientifically defensible approach. One of these reviewers noted that
while the nonlinear approach is reasonable to consider, the disparity between the nonlinear and
linear approaches is so large that they cannot easily be used in tandem when making risk
judgments. This reviewer further noted that an advocate of the nonlinear approach might use an
additional uncertainty factor for potential carcinogenicity (e.g., lOx) to get the target dose down
in a range that has a better chance of protecting against both cancer and noncancer endpoints -
an approach that has been used for "Group C" carcinogens in certain regulatory settings (e.g.,
U.S. EPA's Office of Drinking Water). The second reviewer did not consider a low-dose
nonlinear assumption to be compatible with the expected linear production of DNA reactive
metabolites at low doses.
Response: Section 5.4, Cancer Assessment, was restructured to make it clear that EPA has
applied a linear extrapolation approach to the carbon tetrachloride tumor data consistent with the
2005 Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a). These guidelines
recommend the application of a linear extrapolation approach as the default approach "[w]hen
the weight of evidence evaluation of all available data are insufficient to establish the mode of
action for a tumor site and when scientifically plausible based on the available data,... because
linear extrapolation generally is considered to be a health-protective approach." EPA considers
the current understanding of carbon tetrachloride-induced liver tumors and pheochromocytomas
to be consistent with the application of a linear extrapolation approach. Discussion of a
nonlinear approach was moved to Section 5.4.5 and is presented as an alternative approach
supported by empirical (bioassay) evidence for liver cancer at relatively high exposures of
carbon tetrachloride. Additional response to the suggestion that an epigenetic approach be
considered is addressed in response to Cancer Charge Question #2.
6. Linear extrapolation: The Toxicological Review describes the alternative approaches for
incorporating low-dose linearity that were applied to four tumor datasets from JBRC
(1998) (female rat and mouse liver tumors and male and female mouse
pheochromocytomas). These included (1) POD-based straight line risk calculations and
(2) similar risk calculations (for liver tumor data sets only) that examined the effect on risk
A-34 DRAFT - DO NOT CITE OR QUOTE
-------
estimates of using only data on carbon tetrachloride cancer response at exposure levels
below those for which increased cell replication was reported. In addition, a Bayesian
approach was applied to male mouse pheochromocytoma data to investigate the
distribution of the slope parameter in the log-probit model. Please comment on whether
the linear extrapolation approaches are scientifically plausible given potential for a
cytotoxic MOA at higher doses and other MOAs at lower doses. Please comment on EPA's
choice of using data for pheochromocytomas in the male mouse as the basis for the
inhalation unit risk and data for female mouse liver tumors as the basis for the oral slope
factor. Has the rationale for including a low-dose linear extrapolation been transparently
and objectively described in the document? In the above analyses, a BMR of 5% was used
for the female rat liver tumor data set, and a BMR of 10% was used for the other tumor
data sets. Please comment on the scientific justification for the selection of these BMRs. Is
the rationale transparently and objectively described in the document?
Comments: Two reviewers generally concurred with EPA's linear low-dose analysis. Two
reviewers considered it appropriate to present a linear low-dose extrapolation approach as an
alternative approach, but that based on available evidence the nonlinear method seems more
appropriate. One of these two reviewers recommended the addition of an evaluative statement
regarding the likelihood that a linear approach is correct as compared with the nonlinear
extrapolation approach. A fifth reviewer believed that the linear extrapolation of cancer data to
low-dose exposures to carbon tetrachloride is difficult to defend and is not a preferable approach.
A sixth reviewer did not agree that a linear assessment is justified for carbon tetrachloride.
Response: EPA notes that a linear extrapolation approach was selected as a default approach in
large part because of the absence of an understanding of carbon tetrachloride tumor induction in
the low-dose region. According to the 2005 Guidelines for Carcinogen Risk Assessment (U.S.
EPA, 2005a), a linear extrapolation approach is selected because it is generally considered to be
a health-protective approach. As such, it is not possible to provide a statement of the likelihood
that either the linear or nonlinear approach is correct. See also the response to Cancer Charge
Question #5 to which similar comments regarding the linear and nonlinear approaches were
provided.
Comments: Two reviewers agreed that the Toxicological Review clearly describes the
procedures (i.e., assumptions and modeling) for low-dose linear extrapolation and a third
generally agreed with the EPA analyses and choices. A fourth reviewer considered the cancer
modeling approaches to be "reasonable explorations of the dose response at low dose." This
A-3 5 DRAFT - DO NOT CITE OR QUOTE
-------
reviewer further noted, however, that little importance should be given to the run of the
inhalation data in which the top doses were discarded as that is a dose response involving only a
low dose and a control and that this exercise should just be seen as a screening-level cross check.
It was suggested that other techniques to test whether low-dose response is compatible with the
remainder of the dose response might be more helpful in determining whether the dose response
might be different if more sub-toxic doses were available.
Response: Text was added to Section 5.4.3.3, under Female BDF1 mouse - hepatocellular
adenomas + carcinomas, noting the limitations of the dose-response analysis based on control
and 5-ppm liver tumor data only. Section 5.4.4.2 describes this analysis as a less informative
characterization of the dose-response curve than the analysis based on data for the control, 5-
ppm, and 25-ppm dose groups. While limited, this analysis revealed that the elimination of data
points with evidence of cell replication had small impact on the estimate of the inhalation unit
risk (see Table 5-18). EPA is not aware of other techniques that could be applied to this tumor
data set to explore the effects of cytotoxicity on the shape of the dose response curve.
Comments: One reviewer stated that the use of data for pheochromocytomas in the male mouse
as the basis for the inhalation unit risk appears sound and provides the highest risk estimate.
Two reviewers questioned the relevance of pheochromocytomas to humans, and one of these two
further indicated that these tumors should not be used in derivation of a cancer risk value.
One reviewer expressed some reservations about the exclusive use of a probit model for
the pheochromocytoma Bayesian analysis because it implies an individual threshold-type dose
response for which there is no specific justification.
One reviewer questioned the switching of tumor endpoint when going from the inhalation
unit risk to the oral slope factor given that both the oral and inhalation slope factors are based
upon the same (inhalation) bioassay. This reviewer further observed that a systemic target site
like the adrenal gland would logically be the risk driver for inhalation exposure and the liver
would be the driver for oral exposure because oral exposure leads to first pass metabolism in the
liver whereas inhalation exposure leads to greater systemic doses of parent compound and more
opportunity for extra-hepatic targeting of the tumor response. Given the uncertainties in the
MOA and PBPK modeling of the adrenal tumor dose response, this reviewer believed that it may
be more appropriate to use female liver tumors as the basis for both the IUR and oral SF. If
retained, the reviewer recommended a straightforward explanation of the reasons for switching
endpoint with dose route.
A-3 6 DRAFT - DO NOT CITE OR QUOTE
-------
Response: Comments regarding the relevance of pheochromocytomas and use of
pheochromocytoma data for derivation of cancer risk estimates are addressed in response to
Carcinogenicity Charge Question #4.
The log-probit model (without restriction on the slope parameter) was the only model in
the BMDS suite of models that provided an adequate fit (a p-value for fit >0.1 as recommended
by the BMDS guidance [U.S. EPA, 2000c]) and therefore was the model used to estimate the
POD. Bayesian analysis was used to provide more detail on why restricting the slope parameter
is inappropriate.
An explanation for the use of different tumor types as the basis for the IUR and oral SF
where tumor incidence data from the same study (i.e., Nagano et al., 2007b) were used as the
basis for both values was added to Section 5.4.4.2.
Comment: Two reviewers considered the choice of a BMR of 5% for female rat liver tumor data
and a BMR of 10% for the other tumor data sets to be scientifically justified. One reviewer
stated that the objective of using a BMR of 5% for all tumor sites (i.e., a POD as far removed
from the hepatotoxic portion of dose response as possible) made sense, but this reviewer
believed that backing off to the 10% BMR for all tumor endpoints other than female rat liver was
not well justified. This reviewer suggested a graphic depiction of where on the dose-response
curve a 5 and 10% BMR lies and description of potential risk implications (i.e., would you tend
to get higher slope factor with a 5 vs 10% BMR?). Other reviewers did not offer an opinion
regarding selection of the BMR.
Response: The Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend that
for each tumor response, a POD from the observed data be estimated to mark the beginning of
extrapolation to lower doses and that this POD be an estimated dose near the lower end of the
observed range without significant extrapolation to lower doses. Appendix E provides the dose-
response curves from BMDS for the models that provided the best fit of the tumor data from the
JBRC bioassay (Nagano et al., 2007b) (see list of tumor data sets modeled in Table 5-18). As
the plots in Appendix E show, a BMR of 10% for male and female mouse liver tumors and male
mouse pheochromocytomas results in a BMDL at the low end of the observed range (i.e., on
either side of the lowest exposure concentration). Use of a BMR of 5% would move the BMDL
lower on the dose-response curve and would result in more significant extrapolation below the
observed range. For these tumor data sets, therefore, a BMR of 5% is not supported.
In the case of female mouse pheochromocytomas, the current BMR of 10% yields a
BMDL somewhat above the mid-dose group. Because there were no pheochromocytomas in the
control, low-, and mid-dose groups in the female mouse, a BMR below the mid-dose group
A-3 7 DRAFT - DO NOT CITE OR QUOTE
-------
would, for this data set, be outside the observed range (i.e., the range that produced a tumor
response). Therefore, for this pheochromocytoma data set, a BMR of 5% is similarly not
supported.
7. The conclusion was reached that studies of carbon tetrachloride carcinogenicity by the
oral exposure route are not sufficient to derive a quantitative estimate of cancer risk using
oral cancer response data and low-dose linear approaches. Please provide detailed
comments on whether this judgment is scientifically justified. Has EPA's judgment been
transparently and objectively described in the Toxicological Review? EPA used a PBPK
model to extrapolate inhalation data to derive an oral cancer risk estimate. Please
comment on EPA's application of a PBPK model for route-to-route extrapolation to derive
an oral cancer risk estimate from the inhalation data. Please provide detailed comments on
whether this approach is scientifically justified. Has EPA's judgment been transparently
and objectively described in the document?
Comments: Three reviewers supported the conclusion that studies on carbon tetrachloride
carcinogenicity by the oral route were insufficient to derive a quantitative estimate of cancer risk.
One reviewer considered EPA's judgment that oral studies provide inadequate data for dose-
response assessment to be "basically correct," with the one possible exception of the adrenal
tumor response seen in the oral NCI mouse study in which high doses yielded approximately a
50% response in male mice and 20% response in female mice. This reviewer observed that
analysis of adrenal tumor data could provide interesting comparison to the inhalation dose
response for this endpoint after correction for internal dose differences across dose routes. A
fifth reviewer agreed that EPA had sound reasons for concluding that the available
carcinogenesis studies by the oral route are considerably less than ideal, but added that it is not
impossible to use these data.
Response: In response to the reviewer who suggested that a comparison be made of inhalation
exposure concentrations corresponding to the BMD for 20% and 50% adrenal tumor responses in
the mouse inhalation study with equivalent oral doses (i.e., in terms of internal dose metrics,
MCA and/or MRAMKL), EPA noted that the available PBPK models are not considered
adequate for simulating internal doses from gavage studies (e.g., NCI, 1977). Challenges in
simulating absorption kinetics in carbon tetrachloride gavage studies include pulsatile absorption
kinetics, which are vehicle-dependent (e.g. corn oil, Emulphor) and may be dose-dependent
(Fisher et al., 2004; Semino et al., 1997; Gallo et al., 1993). Available models that have been
developed to simulate the relatively complex kinetics of carbon tetrachloride absorption in rodent
A-3 8 DRAFT - DO NOT CITE OR QUOTE
-------
gavage studies have required calibration of the absorption parameters to the specific observations
being simulated and have not been successfully validated to extrapolation to other dosing
regimens (Fisher et al., 2004; Semino et al., 1997; Gallo et al., 1993). The uncertainties in
applying existing PBPK models for this purpose are described in Section 5.1.2.
Comments: Three reviewers considered the use of a PBPK model for the inhalation-to-oral
exposure extrapolation to be supported. One of these three reviewers qualified this comment
with the observation that the analysis used a low-dose linear extrapolation, an approach with
which this reviewer was not in agreement. One reviewer reiterated a previous comment on
PBPK model use in derivation of the RfC; i.e., that the use of time-weighted carbon tetrachloride
dosimetry has a questionable relevance to mechanisms of carcinogenesis.
Response: With regard to the comments on low-dose linear extrapolation, see response to
comments under Cancer Charge Questions 3 and 6. The comment related to the use of time-
weighted carbon tetrachloride dosimetry is addressed in response to RfC Charge Question #5
where this reviewer offered the same comment.
Comment: One reviewer did not consider the inhalation to oral extrapolation to be adequately
explained, including the basis for the rate of uptake of carbon tetrachloride from the GI tract
(RGIL) and the assumptions of whether human oral exposure is relatively constant or bolus in
nature.
Response: The RGIL parameter is an estimate of the rate of transfer of carbon tetrachloride from
the gastrointestinal tract to liver (mg/kg-day) that is equivalent, in terms of internal dose (i.e.,
MCL or MRAMKL), to a continuous inhalation exposure (HEC, mg/m3). In extrapolating
human inhalation exposures to equivalent human ingestion doses (HED), the RGIL value was
used to estimate the HED. Section 5.4.3.4 was revised to clarify that HED values were
calculated from the predicted relationship (i.e., from the human PBPK model) between the HEC
and RGIL for the purpose of making the route-to-route extrapolation.
Values for RGIL in humans cannot be derived from available studies for two reasons:
(1) no ingestion studies have been reported that allow estimates to be made of carbon
tetrachloride bioavailability or absorption kinetics in humans; and (2) studies conducted in
rodents in which animals received gavage doses of carbon tetrachloride have shown that
absorption kinetics can be complex (e.g., pulsatile), dependent on vehicle (e.g., corn oil,
Emulphor, water), and may also be dose-dependent (Fisher et al., 2004; Semino et al., 1997;
Gallo et al., 1993). Based on the absence of data on absorption kinetics and bioavailability in
A-3 9 DRAFT - DO NOT CITE OR QUOTE
-------
humans, the simplest conceptual model was adopted for the purpose of making the inhalation-to-
oral extrapolation in humans, where the primary interest is a continuous exposure scenario (e.g.,
drinking water); it was assumed that bioavailability was 100% and absorption rate during the day
for any given oral dosage (mg/kg-day) was constant. The Toxicological Review (Section
5.4.3.4) was revised to clarify this point.
Note that the RGIL was used for route-to-route extrapolation in humans in deriving the
oral cancer slope factor; the absence of an adequate model for simulating bioavailability and
absorption kinetics in gavage studies conducted in rodents precluded using PBPK models for
animal-to-human extrapolations of internal dosimetry in deriving the RfD (see Section 5.1.2).
Comment: One reviewer found some contradiction in the document because PBPK modeling
was not used in connection with the interspecies projection of the RfD but was used to derive an
oral cancer SF.
Response: In the derivation of the oral SF, the human PBPK model was used to extrapolate from
inhalation exposures to oral dosages that would result in the same values for internal dose
metrics. In applying the PBPK model, however, it is acknowledged that this approach would
only approximate oral dosimetry because it does not account for oral bioavailability or
absorption kinetics, information that is not available for carbon tetrachloride (see Section
5.4.3.4). Had route-to-route extrapolation not been performed, an oral SF could not have been
derived because oral cancer bioassay data for carbon tetrachloride are not adequate for dose-
response analysis (see Section 5.4.1.2). EPA considers the uncertainties associated with use of
the PBPK model for route-to-route extrapolation to be acceptable and preferable to not having
any quantitative estimate of oral cancer potency for carbon tetrachloride on IRIS.
On the other hand, a route-to-route extrapolation was not needed for deriving the RfD,
since adequate animal oral studies were available. The determination made in the derivation of
the RfD was that PBPK models were not sufficiently developed to extrapolate internal dose
estimates across species, in particular for extrapolation of internal doses resulting from oral
gavage doses (e.g., in corn oil or other vehicles) to continuous exposures of carbon tetrachloride
exposures (e.g., in water or food; see Section 5.1.2).
EPA does not view these two determinations (i.e., to use a PBPK model for dosimetry
extrapolations across routes for continuous exposures in humans, but not use PBPK models for
dosimetry extrapolations across species and from gavage dosing in one species to continuous
dosing in humans) to be in conflict. The Toxicological Review has been revised to include these
considerations (see Section 5.1.2).
A-40 DRAFT - DO NOT CITE OR QUOTE
-------
8. EPA's 2005 Guidelines for Carcinogen Risk Assessment provides guidance on choosing
an approach for dose-response extrapolation below the observed data. Relevant language
related to choosing an extrapolation approach is provided in Section 5.4.3 of the
Toxicological Review. In this section of the Toxicological Review, a linear low-dose
extrapolation approach is recommended for assessing carbon tetrachloride cancer risk
over a nonlinear approach due to uncertainty in understanding the cancer MOA as well as
some bioassay evidence inconsistent with a nonlinear MOA at low exposure levels. Please
comment on the scientific justification for this recommendation. Has this recommendation
been transparently and objectively described in the document?
Comments: Two reviewers generally agreed with the recommendation to use a low-dose linear
model.
One reviewer believed it appropriate to present a possible, alternative risk assessment
approach; however, this reviewer identified concerns with both the validity of some of the data
and their relevance to humans that makes the linear approach much less likely to yield accurate
estimates of risk. This reviewer also observed that the document lacks an evaluation of the
likelihood of one approach over the other providing an accurate assessment.
One reviewer concurred that an alternative MOA may be operative in the carbon
tetrachloride carcinogenesis at low exposure levels and considered the linear low-dose
extrapolation approach to be adequately described, but stated that "the data lacks for support of a
linear approach."
One reviewer stated that the recommendation to apply a linear low-dose extrapolation
approach was not convincing, and suggested that an alternative nonlinear approach be used. The
reviewer also noted that the oral cancer slope factor at a 10"6 risk level would require enforcing a
concentration in drinking water below practical quantifiable limits for carbon tetrachloride.
Another reviewer reiterated their disagreement with the recommendation based on the following:
(1) "the slight increase in tumor response (5 ppm, female mice) was limited to female mice, and
not in male mice or male and female rats", and "[t]he tumor response at 5 ppm in female mice
(18%) was considerably lower than the incidence produced by 25 (88%) and 125 ppm (98%)
carbon tetrachloride in male mice (98%) and by 125 ppm in either male or female rats (80-
88%)." Further, epidemiological studies have not identified an association between human
exposures to carbon tetrachloride and increased liver cancer incidence; (2) pheochromocytomas
in mice were classified as benign, were only observed in mice by two separate routes of
administration, and epidemiological studies did not reveal increases in pheochromocytomas in
exposed humans; and (3) a weight-of-evidence approach of the scientific data supports that
carbon tetrachloride is not genotoxic or mutagenic.
A-41 DRAFT - DO NOT CITE OR QUOTE
-------
Response: The recommendation to apply a linear extrapolation approach for carbon tetrachloride
cancer dose-response assessment is consistent with the 2005 Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005a) that state that a linear approach should be used with agents whose
MOA is considered to be linear in the region below the POD as well as when the available data
are insufficient to establish the MOA for a tumor site. EPA notes that for carbon tetrachloride a
linear extrapolation approach was selected as a default approach in large part because of the
absence of an understanding of carbon tetrachloride tumor induction in the low-dose region. See
also responses to Cancer Charge Questions #3 and 6 that address comments from the peer
reviewers similar to those offered in response to the above charge question.
In response to the reviewer who noted that the oral cancer slope factor would result in
drinking water limits below practical quantifiable limits, EPA notes that drinking water standards
(maximum contaminant levels or MCLs) are set as close to the health based level as feasible, but
also take into consideration best available technology, treatment techniques, and cost.
A-42 DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX B. DOSE-RESPONSE MODELING FOR DERIVING THE RfD
Serum enzyme data (indicators of liver toxicity) from Bruckner et al. (1986) are
summarized in Table B-l.
Table B-l. Serum enzyme data in male rats after 10- or 12-week
exposure to carbon tetrachloride
Daily dose
(mg/kg-day)
0
1
10
o o
JJ
SDH (IU/mL)a
10 weeks
3. 5 ±0.4
2.3 ±0.6
7.6±2.5b
134.8 ± 15. Ob
12 weeks
3. 2 ±0.4
1.9±0.1
8.7±2.0b
145.7 ±57.9b
OCT (nmol CO2/mL)a
10 weeks
28 ±8
23 ±3
55 ±10
148 ± 48b
12 weeks
45 ±4
61 ±12
69 ±16
247±31b
ALT (IU/mL)a
10 weeks
18±1
20 ±1
23 ±1
617 ±334
12 weeks
20 ±0.3
19±1
27±2b
502±135b
aValues presented are mean ± SE for groups of five rats at 10 weeks and seven to nine rats at 12 weeks.
V<0.05
Source: Bruckner et al. (1986).
B.I. BMD Modeling of SDH
SDH data for the 10- and 12-week time points were used for BMD analysis. Because the
precise group sizes were not known for the 12-week data (a range of 7-9 rats per group was
reported), BMD modeling for 12-week data were run using an n of both 7 and 9 rats/group to
bracket the values of the BMD and BMDL.
All of the models for continuous data in U.S. EPA's BMDS (version 1.4.1) (U.S. EPA,
2007) were fit to the 10- and 12-week serum SDH data from Bruckner et al. (1986) (see Table B-
1). Because of the nonhomogeneous variances in the SDH data, a nonhomogeneous variance
model was used in running each of the models in BMDS. A twofold increase in mean control
SDH was used as the BMR (see Section 5.1.2. for the rationale for using this BMR), with
"relative deviation" selected as the BMR type. As stated in U.S. EPA's BMD technical guidance
(U.S. EPA, 2000c), relative deviation means the BMR will be the background estimate (PO) plus
(or minus) the product of the background estimate times the BMR Factor (BRMF) entered by the
user, or
BMR = PO ± (BMRF*PO)
To achieve a doubling of the control mean, a benchmark response factor (BMRF) of one was
B-l
DRAFT - DO NOT CITE OR QUOTE
-------
used. Thus, the BMR was calculated as PO + (1 x PO) or 2 x PO. It should be noted that HMDS
uses the fitted, or estimated, value for the mean and SD to calculate the BMR and BMD. For
example, for the 10-week SDH data, the value estimated by BMDS for the control SDH mean
was 2.71 lU/mL (see detailed model run; a box appears around the estimated mean). Thus, for
this data set, the BMR using relative deviation (as the BMR type) and a BMRF of 1 was
calculated as BMR = 2.71 + (1 x 2.71) = 5.42.
Modeling results are summarized in Table B-2. The 3rd degree polynomial and power
models provided adequate fits of the 10-week SDH data (based on a goodness-of-fit p-value
>0.1); with both models, the modeling of the variance (test 3 in BMDS output) was marginally
adequate (p-value = 0.07515). The power model, which provided the better fit of the data (based
on the lower AIC value), gave an estimated BMD2X of 7.32 mg/kg-day and BMDL2X of
5.46 mg/kg-day (see the detailed model run at the end of this appendix). None of the models for
continuous data in BMDS provided an adequate fit of the 12-week SDH data (i.e., the linear,
polynomial, and power models yielded a p-value <0.0001, and there were insufficient degrees of
freedom to run the Hill model).
Table B-2. Model predictions for changes in serum SDH levels (lU/mL) in
male rats exposed to carbon tetrachloride for 10 and 12 weeks
Model
/7-value3
AIC for fitted
model
BMD2X
(mg/kg-day)
BMDL2X
(mg/kg-day)
10-WEEK DATA
Linear3
Polynomial (3rd degree)b'c
Powerd
Hilld
O.0001
0.253
0.264
NAe
--
85.95
85.88
87.84
--
7.15
7.32
8.88
--
4.29
5.46
5.49
12-WEEK DATA: all continuous models provided a significant lack of fit.
ap-value for Test 4: Does the model fit? Values <0.10 fail to meet conventional goodness-of-fit criteria.
bBetas restricted to. 0.
0 Insufficient degrees of freedom to fit higher degree polynomials.
d Power restricted to. 1.
e Insufficient degrees of freedom.
For purposes of comparison across chemicals, the BMD and BMDL corresponding to a
change in the mean response equal to one control SD from the control mean were also calculated
for the 10-week SDH data (using the power model, which provided the best fit of the data as
described above), consistent with BMD guidance (U.S. EPA, 2000c):
BMDiso: 5.5 mg/kg-day
BMDLiso: 3.8 mg/kg-day
B-2
DRAFT - DO NOT CITE OR QUOTE
-------
B.2. BMD Modeling of OCT and ALT
BMD modeling was also conducted for OCT and ALT. Available continuous variable
models in EPA's BMDS (linear, polynomial, power, and Hill models; BMDS version 1.4.1; U.S.
EPA, 2007b) were fit to the 10- and 12-week data shown in Table B-l for changes in serum OCT
and ALT in male rats exposed to carbon tetrachloride (Bruckner et al., 1986). For each of these
endpoints, a twofold increase in mean enzyme level was used as the BMR (see Section 5.1.2.),
with relative deviation as the BMR type and a BMRF of one (see Section B.I). A
nonhomogeneous variance model was used in running each of the models in BMDS.
Modeling results for OCT data are summarized in Table B-3. None of the models for
continuous data provided an adequate fit to the 10- or 12-week OCT data (based on a goodness-
of-fitp-value>0.1).
Modeling results for ALT data are summarized in Table B-4. The power model provided
an adequate fit of the 10-week ALT data; however, as shown in Table B-l, the SEM of the mean
ALT for the high-dose (33 mg/kg-day) rats was extremely large (617 ± 334). Bruckner et al.
(1986) noted: "There was a pronounced rise in GPT [ALT] at 10 and 12 weeks. Scrutiny of
values of individual animals revealed that dramatic increases in two rats at each time point were
largely responsible for the late increase in GPT [ALT] activity." In light of the large variation in
response at 33 mg/kg-day, relatively high uncertainty is associated with quantitative analysis
using the 10-week ALT data set. The polynomial and power models provided adequate fits of
the 12-week ALT data (based on a goodness-of-fit p-value >0.1). The polynomial model, which
provided a better fit of the data using both an n = 7 and 9 (based on lower AIC values), gave an
estimated BMD2X of 13.0 mg/kg-day and a BMDL2X of 11.8 mg/kg-day. The values of the
BMD and BMDL were not sensitive to the value of n. Model outputs for the ALT data sets are
provided at the end of this appendix.
B-3 DRAFT - DO NOT CITE OR QUOTE
-------
Table B-3. Model predictions for changes in serum OCT levels (nmol COi/mL) in
male rats exposed to carbon tetrachloride for 10 and 12 weeks
Model
p value"
AIC for fitted
model
BMD2X
(mg/kg-day)
BMDL2X
(mg/kg-day)
10-WEEK DATA
Linear3
Polynomial (2nd degree)b'c
Powerd
Hilld
0.0449
0.0427
0.0553
NAe
157.57
157.47
157.04
158.60
8.04
11.4
11.04
10.12
4.44
5.86
6.19
6.52
12-WEEK DATA
Linearb
n=7
n=9
0.04507
0.03479
239.05
313.74
9.08
9.00
5.78
5.79
Polynomial13'
n=7 (2nd degree)
n=9 (2nd degree)
0.0499
0.02905
238.70
313.79
14.8
14.1
7.62
7.14
Powerd
n=7
n=9
0.04347
0.02376
238.93
314.14
15.4
14.8
7.37
6.80
Hilld
n=7
n=9
NAe
NAe
241.04
316.31
11.7
11.7
10.0
6.60
a Values <0.10 fail to meet conventional goodness-of-fit criteria.
b Betas restricted to >0.
0 Insufficient degrees of freedom to fit higher degree polynomials.
d Power restricted to >1.
eNA = not available; insufficient degrees of freedom.
Source: Bruckner et al. (1986).
B-4
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-4. Model predictions for changes in serum ALT levels (lU/mL) in male
rats exposed to carbon tetrachloride for 10 and 12 weeks
Model
p value"
AIC for fitted
model
BMD2X
(mg/kg-day)
BMDL2X
(mg/kg-day)
10-WEEK DATA
Linear3
Polynomial (3rd degree)b'c
Powerd
Hilld
O.0001
0.01022
0.1145
NAe
291.27
123.31
118.70
120.70
33.05
13.66
14.66
NAf
0.0071
12.7
13.2
NAf
12-WEEK DATA
Linearb
n=7
n=9
0.0001
O.0001
353.58
454.34
Failed
Failed
0.66
0.53
Polynomial13'
n=7 (3rd degree)
n=9 (3rd degree)
0.5311
0.631
159.75
212.94
13.0
13.0
11.8
11.8
Powerd
n=7
n=9
0.6561
0.8388
160.72
214.13
12.9
12.9
11.9
11.8
Hilld
n=7
n=9
NA
NA
162.72
216.13
10.88
11.8
Failed
Failed
a Values <0.10 fail to meet conventional goodness-of-fit criteria.
b Betas restricted to >0.
0 Insufficient degrees of freedom to fit higher degree polynomials.
d Power restricted to >1.
feNA = not available; insufficient degrees of freedom (BMD software could not generate a model output).
Source: Bruckner et al. (1986).
B-5
DRAFT - DO NOT CITE OR QUOTE
-------
BMDS model output for SDH levels (10 weeks)
Power Model. (Version: 2.6; Date: 12/06/2005)
Input Data File: G:\CARBON TET\BMD\RFD\SDH.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\RFD\SDH.plt
BMDS MODEL RUN - Power Model
The form of the response function is:
Y[dose] = control + slope * dose^power
Dependent variable = MEAN
Independent variable = Dose(mg/kg-d)
The power is restricted to be greater than or equal to 1
The variance is to be modeled as Var(i) = alpha*mean(i)^
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =
rho =
control =
slope =
power =
289.698
0
2.3
0.0106715
2.69605
Asymptotic Correlation Matrix of Parameter Estimates
alpha rho control slope
alpha
rho
control
slope
power
1
-0.87
-0.45
-0.17
0.19
-0.87
1
0.32
0.14
-0.18
-0.45
0.32
1
-0.12
0.1
-0.17
0.14
-0.12
1
-0.99
power
0.19
-0.18
0.1
-0.99
1
Parameter Estimates
Variable
alpha
rho
control
slope
power
Estimate
0.393849
1.64633
2.70501
0.0161484
2.57243
Std. Err.
0.284596
0.261152
0.432245
0.0130984
0.243917
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.163949
1.13449
1.85783
-0.00952409
2.09436
0.951647
2.15818
3.5522
0.0418208
3.0505
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
1
10
33
3.5
2 .3
7.6
135
[TTTTI
2 .72
8.74
133
0.9
1.3
5.6
33.5
1.42
1.43
3.74
35.1
1.25
-0.658
-0.681
0.125
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} =
B-6
DRAFT - DO NOT CITE OR QUOTE
-------
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = alpha*(Mu(i))
Model R: Yi = Mu + e(i)
Var{e(i)} =
Likelihoods of Interest
Model Log(likelihood)
Al -64.456951
A2 -34.731110
A3 -37.319331
fitted -37.942951
R -91.888765
# Param ' s
5
8
6
5
2
AIC
138.913902
85.462220
86.638662
85.885902
187.777530
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels? (A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Test
Test 1
Test 2
Test 3
Test 4
Tests of Interest
-2*log(Likelihood Ratio) Test df
114.315
59.4517
5.17644
1.24724
p-value
<.0001
<.0001
0.07515
0.2641
The p-value for Test 1 is less than .05. There appears to be a difference between response
and/or variances among the dose levels. It seems appropriate to model the data.
The p-value for Test 2 is less than .1. A non-homogeneous variance model appears to be
appropriate.
The p-value for Test 3 is less than .1. You may want to consider a different variance model.
The p-value for Test 4 is greater than .1. The model chosen seems to adequately describe the
data.
Benchmark Dose Computation
Specified effect = 1
Risk Type = Relative risk
Confidence level = 0.95
HMD = 7.32096
BMDL = 5.46287
B-7
DRAFT - DO NOT CITE OR QUOTE
-------
o>
CO
C
o
Q.
CO
CD
o:
CO
CD
Power Model with 0.95 Confidence Level
150
100
50
Power
10
15
20
25
30
dose
13:2712/282006
DRAFT - DO NOT CITE OR QUOTE
-------
BMDS model output for ALT levels (10 weeks)
Power Model. (Version: 2.6; Date: 12/06/2005)
Input Data File: G:\CARBON TET\BMD\RFD\ALT.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\RFD\ALT.plt
BMDS MODEL RUN
The form of the response function is:
Y[dose] = control + slope * dose^power
Dependent variable = MEAN
Independent variable = Dose(mg/kg-d)
The power is restricted to be greater than or equal to 1
The variance is to be modeled as Var(i) = alpha*mean(i)^
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 139449
rho = 0
control = 18
slope = 3.57427
power = 1.46475
Asymptotic Correlation Matrix of Parameter Estimates
alpha
rho
control
slope
power
alpha
1
-0.97
-0.11
-0.51
0.64
-
0
rho
0.97
1
.068
0.56
-0.7
control
-0.11
0.068
1
-0.34
0.27
slope
-0.51
0.56
-0.34
1
-0.98
power
0.64
-0.7
0.27
-0.98
1
Parameter Estimates
Variable
alpha
rho
control
slope
power
Estimate
0.000240402
3 .29767
19.0745
0.000186379
4.29657
Std. Err.
0.000362454
0.461545
0.631297
0.00024718
0.473086
95.t
Wald Confidence Interval
Lower Conf. Limit
-0.000469995
2.39305
17.8372
-0.000298084
3.36933
Upper Conf. Limit
0.0009508
4.20228
20.3119
0.000670842
5.2238
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
B-9
DRAFT - DO NOT CITE OR QUOTE
-------
0
1
10
33
18
20
23
617
19.1
19.1
22 .8
643
2.2
2 .2
2 .2
747
2
2
2 .68
661
-1.2
1.03
0.197
-0.0863
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = alpha*(Mu(i))
Model R: Yi = Mu + e(i)
Var{e(i)} = Si
Likelihoods of Interest
Model Log(likelihood)
Al -126.223088
A2 -52.674727
A3 -53.102306
fitted -54.347748
R -131.425960
# Param ' s
5
8
6
5
2
AIC
262 .446176
121.349453
118.204612
118.695497
266.851919
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Test
Test 1
Test 2
Test 3
Test 4
Tests of Interest
-2*log(Likelihood Ratio) Test df
157.502
147.097
0.855159
2.49088
p-value
<.0001
<.0001
0.6521
0.1145
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
model appears to be appropriate
A non-homogeneous variance
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Relative risk
B-10
DRAFT - DO NOT CITE OR QUOTE
-------
Confidence level = 0.95
HMD = 14.6575
BMDL = 13.205
-------
BMDS model output for ALT levels (12 weeks)
Polynomial Model. (Version: 2.7; Date: 12/06/2005)
Input Data File: G:\CARBON TET\BMD\RFD\BRUCKNER 12-WK DATA\ALT-12N7.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\RFD\BRUCKNER 12-WK DATA\ALT-12N7.pit
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*dose^2 + ...
Dependent variable = MEAN
Independent variable = Dose(mg/kg-d)
The polynomial coefficients are restricted to be positive
The variance is to be modeled as Var(i) = alpha*mean(i)^r
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1
rho = 0
beta_0 = 20
beta_l = 0
beta_2 = 0
beta 3 = 0
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
alpha 8.5932e-005 8.29597e-005 -7.66661e-005 0.00024853
rho 3.65721 0.244569 3.17787 4.13656
beta_0 19.3388 0.558043 18.2451 20.4325
beta_l 0 NA
beta_2 0 NA
beta_3 0.00876918 0.00160884 0.00561592 0.0119224
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -beta_l -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
alpha rho beta_0 beta_3
alpha 1000
rho 0100
beta 00010
B-12 DRAFT - DO NOT CITE OR QUOTE
-------
beta 3
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
1
10
33
7
7
7
7
20
19
27
502
19.3
19.3
28.1
334
0.8
2 .6
5.3
357
2 .09
2 .09
4 .13
383
0.838
-0.44
-0.709
1.16
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma^2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = alpha*(Mu(i))
Model R: Yi = Mu + e(i)
Var{e(i)} = Si
Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log(likelihood)
-157.028066
-69.790014
-75.241703
-75.874569
-170.807125
# Param ' s
5
8
6
4
2
AIC
324 .056132
155.580028
162 .483407
159.749137
345.614250
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test
Test 1
Test 2
Test 3
Test 4
-2*log(Likelihood Ratio) Test df
202.034
174.476
10.9034
1.26573
p-value
<.0001
<.0001
0.004289
0.5311
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is less than .1. You may want to consider a
different variance model
B-13
DRAFT - DO NOT CITE OR QUOTE
-------
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Relative risk
Confidence level = 0.95
HMD = 13.0164
BMDL =
11.791
Polynomial Model with 0.95 Confidence Level
900
800
700
g 600
° 500
CD
K 400
cc
1 300
200
100
0
Pi*^l\/ni*^fn i^ I
: polynomial
r
r
\ //
- ^^^
r ^^^___^-^^
! BMDL BMQ
'-_
\
> H
H
H
H
0 5 10 15 20 25 30
dose
17:1812/282006
B-14
DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX C. PBPK MODELING
C.I. Paustenbach et al. (1988) and Thrall et al. (2000) PBPK Models (rat, mouse, human)
Detailed summaries of the Paustenbach et al. (1988) and Thrall et al. (2000) PBPK
models appear in Section 3.5. Source code for the rat, mouse, and hamster models (reported in
Thrall et al., 2000) in Advanced Continuous Simulation Language (ACSL) was graciously
provided to Syracuse Research Corporation (SRC) by Dr. Karla Thrall. Included with the code
were data collected from gas uptake studies conducted in these species (also reported in Thrall et
al., 2000). Accuracy of the implementation of the rat and mouse models in ACSL (version
11.8.4) was verified by comparing model predictions to observations from the closed chamber
studies. These simulations are shown in Figures C-l and C-2. The comparisons of observed and
predicted closed chamber CCU concentrations as a function of exposure times match those
reported in Figure 2 of Thrall et al. (2000).
10000
1000 :
100 :
o
o
Data points are observations (provided by Thrall) for exposures for three rats per
chamber (body weight, 0.24 kg); lines are simulations. The non-specific loss rate of
carbon tetrachloride from the chamber was assumed to be 0.05 hour"1 (from Thrall).
Partition coefficients were from Thrall source code.
Figure C-l. Comparison of observed and predicted chamber carbon
tetrachloride concentrations in closed chamber studies conducted in rats.
C-l
DRAFT - DO NOT CITE OR QUOTE
-------
10000
1000 -
Q.
CL
C
g
to
o
O
(5
.Q
E
ra
6
100 -
Gas Chamber Simulations - Mouse
Data points are observations (provided by Thrall) for exposures for seven mice
per chamber (body weight, 0.024 kg); lines are simulations. The non-specific loss
rate of CCU from the chamber was assumed to be 0.05 hour"1 (from Thrall source
code). Partition coefficients were from Thrall source code.
Figure C-2. Comparison of observed and predicted chamber carbon
tetrachloride concentrations in closed chamber studies conducted in mice.
As noted above, Thrall et al. (2000) compared model predictions for the rat and mouse
with experimental data collected over a 48-hour period following a 4-hour nose-only inhalation
exposure to 20 ppm of [14C]-carbon tetrachloride. This comparison of PBPK model-predicted
and experimentally-observed values for selected parameters is provided in Table C-l. Thrall et
al. (2000) also compared the model simulation for humans with human data of Stewart et al.
(1961); (see Figure C-3). As this figure shows, the model simulation of expired carbon
tetrachloride levels provided good agreement with the experimental data, particularly at longer
periods postexposure.
C-2
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-l. Comparison of predicted and observed values for selected parameters
from toxicokinetic data collected from rats and mice 48 hours post exposure to a 4-
hour nose-only inhalation exposure (20 ppm carbon tetrachloride)
Species
Rat
Mouse
Parameter
Initial body burden
Total amount trapped by KOH
Total amount trapped on charcoal0
Total amount metabolized
Initial body burden
Total amount trapped by KOHb
Total amount trapped on charcoal0
Total amount metabolizedd
Model
(umol)
7.8
2.8
4.1
3.7
2.2
0.95
0.94
1.3
Data
(umol equivalents of
CC14± SD)a
11.7±0.54
2.7 ±0.25
7.4 ±0.44
3.7 ±0.22
2.0 ±0.48
0.69±0.11
0.76 ±0.37
1.2±0.11
Ratio
(predicted/observed)
0.7
1.0
0.6
1.0
1.1
1.4
1.2
1.1
a n = 3^- animals.
b 14CO2 measured using a KOH trap.
0 Parent compound (14CC14) measured using a charcoal trap.
d Represents the sum of radioactivity (in (imol equivalents) in urine, feces, and trapped on KOH (CO2).
Sources: Thrall et al. (2000); Benson and Springer (1999).
c-:
DRAFT - DO NOT CITE OR QUOTE
-------
100
Time, hr
Figure C-3. Comparison of the actual versus predicted concentration of
carbon tetrachloride in the expired breath of humans exposed to 10 ppm of
carbon tetrachloride for 180 minutes (data from Stewart et al., 1961).
Sources: Thrall et al. (2000); Benson and Springer (1999)
Parameter values for the rat and human models used in the Paustenbach et al. (1988) and
Thrall et al. (2000) models are summarized in Table C-2. Parameter values for the mouse are
shown in Table C-3.
C-4
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-2. Parameter values for rat and human models"
Parameter
BW
VLC
we
VSC
VRC
QCC
QPC
QLC
QFC
QSC
QRC
PB
PL
PF
PS
PR
VmaxC
Kmx
Definition
Body weight (kg)
Liver volume (fraction of body)
Fat volume (fraction of body)
Slowly -perfused tissue volume (fraction of body)
Rapidly -perfused tissue volume (fraction of body)
Cardiac output (L/hour-kg BW074)
Alveolar ventilation rate (L/hour-kg BW° 74)
Liver blood flow (fraction of cardiac output)
Fat blood flow (fraction of cardiac output)
Slowly -perfused blood flow (fraction of cardiac output)
Rapidly -perfused blood flow (fraction of cardiac output)
Blood:air partition coefficient
Liverblood partition coefficient
Fatblood partition coefficient
Slowly -perfused partition coefficient
Readily -perfused partition coefficient
Maximum rate of metabolism (mg/hour-kg B W° 7)
Michaelis-Menten coefficient for metabolism (mg/L)
Rat model
0.452b
0.04c'd
0.08c'd
0.74c'd
0.05c'd
15c'd
15c'd
0.25c'd
0.04c'd
0.2c'd
0.51c'd
4.52e
3.14e
79.42e
le
3.14f
0.4e,0.65c
0.25c'd
Human model
70
0.04C
0.2s
0.62C
0.05C
15C
15C
0.25C
0.06C
0.1 8C
0.51C
2.64C
3.14e
79.42e
le
3.14f
0.4e,0.65c, 1.49d, 1.7d
0.25c'd
a See summary of the Paustenbach et al. (1988) and Thrall et al. (2000) models in Section 3.5 for discussion the
source of parameter values.
b Time-weighted mean body weight for the exposure group of interest (0.452 kg, male rats) and an exposure of 3
ppm, 6 hours/day, 5 days/week (based on Nagano et al., 2007b; JBRC, 1998).
'Paustenbach et al. (1988).
d Thrall et al. (2000).
eGargasetal. (1986).
f Partition coefficient for readily-perfused is assumed to be equal to that of liver.
8 Adjusted from reported value of 0.1 in Paustenbach et al. (1988).
C-5
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-3. Parameter values for mouse models"
Parameter
BW
VLC
we
VSC
VRC
QCC
QPC
QLC
QFC
QSC
QRC
PB
PL
PF
PS
PR
VmaxC
Kmx
Kl
K2
K2
Definition
Body weight (kg)
Liver volume (fraction of body)
Fat volume (fraction of body)
Slowly -perfused tissue volume (fraction of body)
Richly -perfused tissue volume (fraction of body)
Cardiac output (L/hour-kg BWSF)d'h
Alveolar ventilation rate (L/hour-kg BWSF)d'h
Liver blood flow (fraction of cardiac output)
Fat blood flow (fraction of cardiac output)
Slowly -perfused blood flow (fraction of cardiac output)
Richly -perfused blood flow (fraction of cardiac output)
Blood:air partition coefficient
Liverblood partition coefficient
Fatblood partition coefficient
Slowly -perfused partition coefficient
Richly -perfused partition coefficient
Maximum rate of metabolism (mg/hour-kg B WSF)fj
Michaelis-Menten coefficient for metabolism (mg/L)
GI absorption rate coefficient Cl -liver (hour"1)
GI absorption rate coefficient C1-C2 (hour"1)
GI absorption rate coefficient C2-liver (hour"1)
Thrall et al.
(2000)
0.036b
0.04C
0.04C
0.78C
0.05C
28c'd
28c'd
0.24C
0.05C
0.19C
0.52C
7.83e
2.08e
23. Oe
0.61e
2.08e
0.79e'f
0.46e
-
-
-
Fisher et al.
(2004)
-
0.04g
0.04g
0.69g
0.1 4g
30g"h
30&h
0.24g
0.05g
0.1 7g
0.54g
3.81
4.81
91. 41
2.51
4.81
11J
0.31
0.4, 10k
2k
0.051
a See Paustenbach et al. (1988) and Thrall et al. (2000) for discussion the source of parameter values.
bReference value for mouse body weight in a chronic study (0.036 kg; U.S. EPA, 1988)
c Andersen etal, 1987
d SF, scaling factor; QC (L/hour)=QCC*BW074; QP (L/hour)=QPC*BW074
e Thrall source code (CARBON TETRACHLORIDE PBPK MODEL KD THRALL 3/98 ITRICCL4.ACSL).
Thrall et al. (2000) reported the tissue:blood partition coefficients for the mouse were based on values for blood:air
for the mouse (7.83) from Thrall et al. (2000) and tissue:air values (liver:air=14.2; muscle:air=4.54; fat:air=359)
from Gargas et al. (1986). The corresponding tissue:blood values would be: PL=1.81; PF=45.85; PS=0.58;
PR=1.81.
f SF, scaling factor; VMAX=VBMAXC*BW070
8 Brown etal., 1997
h SF, scaling factor; QC (L/hour)=QCC*BW075; QP (L/hour)=QPC*BW075
1 Fisher et al. (2004) vial equilibrium measurements
1 VMAX=VBMAXC*BW075
kFisher et al. (2004) fit to closed chamber data.
'Fisher et al. (2004) fit to gavage blood data. Kl values are 0.4 hr"1 for 20 mg/kg dose and 10 hr"1 for 50 and 100
mg/kg dose.
C-6
DRAFT - DO NOT CITE OR QUOTE
-------
C.2. Fisher PBPK Model (mouse)
A detailed summary of the mouse PBPK model developed by Fisher et al. (2004) is
provided in Section 3.5. This model was reconstructed from the information provided in their
paper.
Fisher et al. (2004) performed gas uptake experiments with mice at four concentrations of
carbon tetrachloride to estimate metabolic constants. Metabolic constants provided a good fit
between model predictions and observations for the gas uptake study.
Parameter values for the mouse used in the Fisher et al. (2004) model are summarized in
Table C-3 and are compared with the mouse parameter values from the Thrall et al. (2000)
model. Values for Km and Vmaxc used in the two models are similar: 0.3 mg/L, 1 mg/hour/kg0'75
(Fisher et al., 2004) compared to 0.46 mg/L, 0.79 mg/hour/kg0-70 (Thrall et al., 2000); although
different allometric scaling factors were used to scale Vmax to BW. The corresponding Vmax
values for a 0.036-kg mouse are 0.077 mg/hour (Thrall et al., 2000) and 0.082 mg/hour (Fisher et
al., 2004). Tissue partition coefficients used in the Fisher et al. (2004) model were 2-4 times
higher than in the Thrall et al. (2000) model.
C.3. PBPK Modeling of Human Equivalent Concentrations and Doses
Interspecies extrapolation (i.e., rat-to-human, mouse-to-human) and route-to-route
extrapolation of carbon tetrachloride inhalation dosimetry was accomplished using a human
PBPK model described in Paustenbach et al. (1988), Thrall et al. (2000), and Benson and
Springer (1999). The human PBPK model was used to estimate the continuous chronic human
inhalation exposure in mg/m3 (abbreviated as EC in the following tables) or the rate of uptake of
carbon tetrachloride from the GI tract to the liver (i.e., chronic daily ingested dose) in mg/kg-day
(abbreviated RGIL in the following tables) that would result in values for the internal dose
metrics, MCA or MRAMKL, equal to the respective BMDLs for each toxicity endpoint (i.e.,
RfC: fatty liver degeneration; cancer: liver tumors in rats, liver tumors and adrenal
pheochromocytomas in mice). This procedure is described in Section 5.4.3.4.
Conversion factors that relate EC or RGIL to the two dose metrics (MCA and
MRAMKL) for each of the assumed values of human Vmaxc (0.40, 0.65, 1.49, or 1.70
mg/hour/kg BW°'70) are provided in Tables C-4 to C-l 1. Figures C-4 to C-l 1 display plots of
MCA and corresponding values of EC or RGIL predicted from the human PBPK model, with
trend equations developed to permit the calculation of EC or RGIL for any value of MCA.
Trend equations shown on the plots are power functions fit to each data set using the method of
least squares (Microsoft Excel). The corresponding fit to the PBPK model predictions were
evaluated by R2 (shown on the trend plots) and the magnitude of the difference between PBPK
model predictions and the trend function predictions (i.e., shown in the plots of % delta, where %
C-7 DRAFT - DO NOT CITE OR QUOTE
-------
delta = 100*[Trend-PBPK]/PBPK). If values for % delta uniformly <5% could not be achieved
with single trend functions applied to the full ranges of internal dose metric values presented in
Tables C-4 to C-l 1, trend functions were developed for subsets of the full MCA range that
yielded achieved % delta values <5%. Similar plots were developed for the dose metric
MRAMKL (see Figures C-12 and C-13).
C-8 DRAFT - DO NOT CITE OR QUOTE
-------
Table C-4. Interspecies conversion factors based on MCA dose metric
(VMAXC=0.40)
EC
(ppm)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
o
J
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
EC
(mg/m3)
0.6290
1.258
1.887
2.516
3.145
3.774
4.403
5.032
5.661
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
MCA
(jimol/L)
0.009182
0.01837
0.02757
0.03678
0.04599
0.05522
0.06445
0.07369
0.08293
0.09219
0.1852
0.2790
0.3735
0.4687
0.5646
0.6611
0.7583
0.8560
0.9543
1.961
2.995
4.045
5.103
6.167
7.234
8.304
9.375
10.447
RGIL
(mg/kg-day)
0.1016
0.2021
0.3019
0.4007
0.4987
0.5959
0.6923
0.7880
0.8829
0.9772
1.887
2.752
3.584
4.392
5.183
5.961
6.729
7.490
8.245
15.67
23.06
30.47
37.91
45.36
52.82
60.29
67.77
75.25
RGIL/EC
(mg/kg-day/
mg/m3)
0.1614
0.1607
0.1600
0.1592
0.1586
0.1579
0.1572
0.1566
0.1560
0.1554
0.1500
0.1458
0.1424
0.1396
0.1373
0.1354
0.1337
0.1323
0.1311
0.1245
0.1222
0.1211
0.1205
0.1202
0.1200
0.1198
0.1197
0.1196
EC/MCA
(mg/m3/
jimol/L)
68.51
68.48
68.45
68.42
68.38
68.35
68.32
68.29
68.26
68.23
67.94
67.65
67.37
67.11
66.85
66.60
66.36
66.14
65.92
64.17
63.01
62.21
61.63
61.20
60.87
60.60
60.39
60.21
RGIL/MCA
(mg/kg-day/
jimol/L)
11.06
11.00
10.95
10.89
10.84
10.79
10.74
10.69
10.65
10.60
10.19
9.864
9.595
9.370
9.180
9.016
8.874
8.749
8.640
7.992
7.699
7.534
7.428
7.355
7.302
7.261
7.229
7.203
EC, air exposure concentration; MCA, time-averaged arterial concentration of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW°70).
C-9
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-5. Interspecies conversion factors based on MCA dose metric
(VMAXC=0.65)
EC
(ppm)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
o
J
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
EC
(mg/m3)
0.6290
1.258
1.887
2.516
3.145
3.774
4.403
5.032
5.661
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
MCA
(jimol/L)
0.008674
0.01735
0.02604
0.03474
0.04344
0.05215
0.06087
0.06959
0.07832
0.08706
0.1748
0.2633
0.3525
0.4424
0.5330
0.6243
0.7162
0.8087
0.9019
1.864
2.866
3.893
4.936
5.988
7.047
8.110
9.176
10.244
RGIL
(mg/kg-day)
0.1182
0.2350
0.3504
0.4645
0.5774
0.6890
0.7995
0.9088
1.0171
1.1243
2.147
3.097
3.994
4.853
5.683
6.489
7.279
8.055
8.821
16.21
23.51
30.85
38.22
45.63
53.05
60.50
67.95
75.41
RGIL/EC
(mg/kg-day/
mg/m3)
0.1879
0.1868
0.1857
0.1846
0.1836
0.1826
0.1816
0.1806
0.1797
0.1787
0.1706
0.1641
0.1588
0.1543
0.1506
0.1474
0.1447
0.1423
0.1402
0.1289
0.1246
0.1226
0.1215
0.1209
0.1205
0.1202
0.1200
0.1199
EC/MCA
(mg/m3/
jimol/L)
72.52
72.49
72.46
72.43
72.40
72.37
72.34
72.31
72.28
72.25
71.96
71.66
71.37
71.09
70.81
70.54
70.27
70.00
69.75
67.51
65.85
64.63
63.72
63.03
62.48
62.05
61.70
61.41
RGIL/MCA
(mg/kg-day/
jimol/L)
13.63
13.54
13.45
13.37
13.29
13.21
13.14
13.06
12.99
12.91
12.28
11.760
11.331
10.969
10.661
10.395
10.164
9.961
9.780
8.699
8.203
7.923
7.743
7.619
7.529
7.460
7.406
7.362
EC, air exposure concentration; MCA, time-averaged arterial concentration of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW°70).
C-10
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-6. Interspecies conversion factors based on MCA dose metric
(VMAXC=1.49)
EC
(ppm)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
o
J
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
EC
(mg/m3)
0.6290
1.258
1.887
2.516
3.145
3.774
4.403
5.032
5.661
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
MCA
(jimol/L)
0.007827
0.01566
0.02349
0.03133
0.03917
0.04702
0.05487
0.06272
0.07058
0.07844
0.1573
0.2365
0.3161
0.3962
0.4766
0.5575
0.6388
0.7205
0.8027
1.650
2.545
3.482
4.454
5.453
6.470
7.501
8.542
9.590
RGIL
(mg/kg-day)
0.1742
0.3457
0.5146
0.6808
0.8447
1.0060
1.1651
1.3219
1.4766
1.6291
3.053
4.326
5.487
6.559
7.564
8.514
9.419
10.288
11.130
18.67
25.69
32.67
39.74
46.90
54.13
61.42
68.76
76.13
RGIL/EC
(mg/kg-day/
mg/m3)
0.2770
0.2748
0.2727
0.2706
0.2686
0.2665
0.2646
0.2627
0.2608
0.2590
0.2427
0.2293
0.2181
0.2085
0.2004
0.1934
0.1872
0.1817
0.1769
0.1484
0.1361
0.1299
0.1263
0.1243
0.1229
0.1221
0.1215
0.1210
EC/MCA
(mg/m3/
jimol/L)
80.37
80.35
80.33
80.31
80.29
80.27
80.25
80.23
80.21
80.19
79.99
79.80
79.60
79.39
79.19
78.98
78.78
78.57
78.36
76.24
74.16
72.26
70.61
69.22
68.06
67.09
66.28
65.59
RGIL/MCA
(mg/kg-day/
jimol/L)
22.26
22.08
21.90
21.73
21.56
21.40
21.23
21.07
20.92
20.77
19.41
18.294
17.358
16.557
15.871
15.272
14.744
14.278
13.864
11.316
10.095
9.384
8.922
8.601
8.367
8.188
8.049
7.938
EC, air exposure concentration; MCA, time-averaged arterial concentration of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW°70).
C-ll
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-7. Interspecies conversion factors based on MCA dose metric
(VMAXC=1.70)
EC
(ppm)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
o
J
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
EC
(mg/m3)
0.6290
1.258
1.887
2.516
3.145
3.774
4.403
5.032
5.661
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
MCA
(jimol/L)
0.007709
0.01542
0.02314
0.03086
0.03858
0.04630
0.05403
0.06177
0.06950
0.07724
0.1548
0.2327
0.3110
0.3896
0.4686
0.5480
0.6277
0.7078
0.7883
1.616
2.488
3.403
4.355
5.336
6.340
7.361
8.394
9.435
RGIL
(mg/kg-day)
0.1882
0.3735
0.5557
0.7351
0.9118
1.0857
1.2571
1.4259
1.5924
1.7565
3.284
4.642
5.873
7.005
8.060
9.051
9.993
10.893
11.758
19.40
26.38
33.28
40.25
47.32
54.49
61.73
69.03
76.36
RGIL/EC
(mg/kg-day/
mg/m3)
0.2993
0.2969
0.2945
0.2922
0.2899
0.2877
0.2855
0.2834
0.2813
0.2792
0.2610
0.2460
0.2334
0.2227
0.2135
0.2055
0.1986
0.1924
0.1869
0.1542
0.1398
0.1323
0.1280
0.1254
0.1238
0.1227
0.1219
0.1214
EC/MCA
(mg/m3/
jimol/L)
81.60
81.58
81.56
81.54
81.53
81.51
81.49
81.47
81.46
81.44
81.26
81.09
80.91
80.73
80.54
80.36
80.17
79.98
79.79
77.83
75.84
73.94
72.22
70.73
69.45
68.36
67.45
66.67
RGIL/MCA
(mg/kg-day/
jimol/L)
24.42
24.22
24.02
23.82
23.63
23.45
23.26
23.09
22.91
22.74
21.21
19.948
18.885
17.978
17.200
16.517
15.920
15.390
14.915
12.003
10.602
9.780
9.242
8.868
8.595
8.386
8.224
8.093
EC, air exposure concentration; MCA, time-averaged arterial concentration of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW°70).
C-12
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-8. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=0.40)
EC
(ppm)
1
2
o
J
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
EC
(mg/m3)
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
691.9
754.8
817.8
880.7
943.6
1006
1069
1132
1195
1258
1321
1384
1447
1510
1573
1636
MRAMKL
(nmol/hr/kg
liver)
0.7352
1.433
2.093
2.719
3.311
3.872
4.402
4.903
5.377
5.826
9.196
11.24
12.57
13.48
14.15
14.65
15.05
15.36
15.62
15.84
16.02
16.18
16.31
16.43
16.53
16.63
16.71
16.78
16.85
16.91
16.97
17.02
17.06
17.11
17.15
RGIL
(mg/kg-day)
0.2980
0.5960
0.8940
1.192
1.490
1.788
2.086
2.384
2.682
2.980
5.959
8.938
11.92
14.89
17.87
20.85
23.83
26.80
29.78
32.75
35.73
38.70
41.68
44.65
47.63
50.60
53.57
56.54
59.52
62.49
65.46
68.43
71.40
74.37
77.34
RGIL/EC
(mg/kg-day/
mg/m3)
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04736
0.04736
0.04735
0.04735
0.04735
0.04734
0.04734
0.04734
0.04733
0.04733
0.04733
0.04732
0.04732
0.04732
0.04732
0.04731
0.04731
0.04731
0.04730
0.04730
0.04730
0.04730
0.04729
0.04729
EC/MRAMKL
(mg/m3/
jimol/hr/kg liver)
8.556
8.782
9.015
9.254
9.498
9.749
10.004
10.264
10.529
10.798
13.681
16.792
20.025
23.329
26.675
30.049
33.442
36.849
40.265
43.689
47.119
50.553
53.990
57.430
60.873
64.318
67.765
71.213
74.662
78.112
81.563
85.015
88.468
91.921
95.375
RGIL/MRAMKL
(mg/kg-day/
jimol/hr/kg liver)
0.4053
0.4161
0.4271
0.4384
0.4500
0.4618
0.4739
0.4862
0.4987
0.5115
0.6480
0.7953
0.9483
1.105
1.263
1.423
1.583
1.744
1.906
2.068
2.230
2.393
2.555
2.718
2.880
3.043
3.206
3.369
3.532
3.695
3.858
4.021
4.184
4.347
4.510
C-13
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-8. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=0.40)
EC
270
280
290
300
EC
1698
1761
1824
1887
MRAMKL
17.19
17.22
17.25
17.28
RGIL
80.31
83.28
86.24
89.21
RGIL/EC
0.04728
0.04728
0.04728
0.04728
EC/MRAMKL
98.830
102.284
105.740
109.195
RGIL/MRAMKL
4.673
4.836
4.999
5.162
EC, air exposure concentration; MRAMKL, time-averaged rate of metabolism of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW°70).
C-14
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-9. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=0.65)
EC
(ppm)
1
2
3
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
EC
(mg/m3)
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
691.9
754.8
817.8
880.7
943.6
1006
1069
1132
1195
1258
1321
1384
1447
1510
1573
1636
MRAMKL
(nmol/hr/kg
liver)
0.9770
1.920
2.830
3.706
4.550
5.361
6.140
6.888
7.607
8.296
13.772
17.33
19.71
21.36
22.57
23.47
24.18
24.74
25.19
25.57
25.89
26.16
26.40
26.60
26.78
26.94
27.08
27.21
27.33
27.43
27.52
27.61
27.69
27.76
27.83
RGIL
(mg/kg-day)
0.2980
0.5960
0.8940
1.192
1.490
1.788
2.086
2.384
2.682
2.980
5.959
8.938
11.92
14.89
17.87
20.85
23.83
26.80
29.78
32.75
35.73
38.71
41.68
44.66
47.63
50.60
53.57
56.55
59.52
62.49
65.46
68.44
71.40
74.38
77.34
RGIL/EC
(mg/kg-day/
mg/m3)
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04736
0.04736
0.04736
0.04735
0.04735
0.04735
0.04734
0.04734
0.04734
0.04733
0.04733
0.04733
0.04733
0.04732
0.04732
0.04731
0.04731
0.04731
0.04730
0.04730
0.04730
0.04730
0.04730
0.04729
EC/MRAMKL
(mg/m3/
jimol/hr/kg liver)
6.438
6.552
6.669
6.789
6.913
7.041
7.171
7.305
7.443
7.583
9.135
10.889
12.768
14.723
16.726
18.760
20.815
22.886
24.967
27.057
29.153
31.254
33.359
35.467
37.578
39.691
41.805
43.922
46.039
48.158
50.278
52.398
54.519
56.641
58.764
RGIL/MRAMKL
(mg/kg-day/
jimol/hr/kg liver)
0.3050
0.3104
0.3159
0.3216
0.3275
0.3335
0.3397
0.3461
0.3525
0.3592
0.4327
0.5157
0.6047
0.697
0.792
0.888
0.986
1.084
1.182
1.281
1.380
1.479
1.579
1.679
1.778
1.878
1.978
2.078
2.178
2.278
2.378
2.479
2.579
2.679
2.779
C-15
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-9. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=0.65)
EC
270
280
290
300
EC
1698
1761
1824
1887
MRAMKL
27.89
27.95
28.01
28.06
RGIL
80.32
83.28
86.25
89.22
RGIL/EC
0.04729
0.04728
0.04728
0.04728
EC/MRAMKL
60.887
63.010
65.134
67.259
RGIL/MRAMKL
2.879
2.979
3.080
3.180
EC, air exposure concentration; MRAMKL, time-averaged rate of metabolism of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW070).
C-16
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-10. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=1.49)
EC
(ppm)
1
2
3
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
EC
(mg/m3)
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
691.9
754.8
817.8
880.7
943.6
1006
1069
1132
1195
1258
1321
1384
1447
1510
1573
1636
MRAMKL
(nmol/hr/kg
liver)
1.3834
2.749
4.095
5.423
6.731
8.020
9.289
10.537
11.764
12.971
23.832
32.48
39.11
44.09
47.83
50.68
52.88
54.62
56.01
57.15
58.10
58.90
59.57
60.16
60.67
61.11
61.50
61.85
62.17
62.45
62.70
62.93
63.14
63.34
63.52
RGIL
(mg/kg-day)
0.2980
0.5960
0.8940
1.192
1.490
1.788
2.086
2.384
2.682
2.980
5.960
8.940
11.92
14.90
17.87
20.85
23.83
26.81
29.79
32.76
35.74
38.71
41.69
44.66
47.64
50.61
53.59
56.56
59.53
62.51
65.47
68.45
71.42
74.39
77.36
RGIL/EC
(mg/kg-day/
mg/m3)
0.04737
0.04738
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04736
0.04736
0.04736
0.04736
0.04735
0.04735
0.04735
0.04734
0.04734
0.04734
0.04733
0.04733
0.04733
0.04733
0.04732
0.04732
0.04732
0.04731
0.04731
0.04730
0.04730
0.04730
EC/MRAMKL
(mg/m3/
jimol/hr/kg liver)
4.547
4.577
4.608
4.640
4.672
4.706
4.740
4.776
4.812
4.850
5.279
5.810
6.434
7.134
7.891
8.689
9.516
10.365
11.230
12.107
12.992
13.885
14.783
15.685
16.590
17.499
18.410
19.323
20.238
21.154
22.071
22.989
23.909
24.829
25.750
RGIL/MRAMKL
(mg/kg-day/
jimol/hr/kg liver)
0.2154
0.2168
0.2183
0.2198
0.2213
0.2229
0.2246
0.2263
0.2280
0.2297
0.2501
0.2752
0.3048
0.338
0.374
0.411
0.451
0.491
0.532
0.573
0.615
0.657
0.700
0.742
0.785
0.828
0.871
0.914
0.958
1.001
1.044
1.088
1.131
1.175
1.218
C-17
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-10. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=1.49)
EC
270
280
290
300
EC
1698
1761
1824
1887
MRAMKL
63.68
63.83
63.97
64.10
RGIL
80.33
83.30
86.27
89.24
RGIL/EC
0.04730
0.04729
0.04729
0.04729
EC/MRAMKL
26.671
27.593
28.516
29.439
RGIL/MRAMKL
1.261
1.305
1.349
1.392
EC, air exposure concentration; MRAMKL, time-averaged rate of metabolism of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW070).
C-18
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-ll. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=1.70)
EC
(ppm)
1
2
3
4
5
6
7
8
9
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
EC
(mg/m3)
6.290
12.58
18.87
25.16
31.45
37.74
44.03
50.32
56.61
62.90
125.8
188.7
251.6
314.5
377.4
440.3
503.2
566.1
629.0
691.9
754.8
817.8
880.7
943.6
1006
1069
1132
1195
1258
1321
1384
1447
1510
1573
1636
MRAMKL
(nmol/hr/kg
liver)
1.4401
2.865
4.273
5.665
7.040
8.398
9.738
11.060
12.365
13.650
25.429
35.14
42.84
48.77
53.31
56.79
59.49
61.62
63.33
64.72
65.88
66.84
67.66
68.37
68.98
69.51
69.99
70.40
70.78
71.11
71.42
71.69
71.94
72.17
72.38
RGIL
(mg/kg-day)
0.2980
0.5960
0.8940
1.192
1.490
1.788
2.086
2.384
2.682
2.980
5.960
8.939
11.92
14.90
17.88
20.85
23.83
26.81
29.79
32.76
35.74
38.71
41.69
44.66
47.64
50.61
53.59
56.56
59.53
62.51
65.48
68.45
71.42
74.39
77.36
RGIL/EC
(mg/kg-day/
mg/m3)
0.04737
0.04738
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04737
0.04736
0.04736
0.04736
0.04735
0.04735
0.04735
0.04735
0.04734
0.04734
0.04734
0.04733
0.04733
0.04733
0.04733
0.04732
0.04732
0.04731
0.04731
0.04731
0.04731
0.04730
EC/MRAMKL
(mg/m3/
jimol/hr/kg liver)
4.368
4.392
4.417
4.442
4.468
4.494
4.522
4.550
4.579
4.608
4.947
5.370
5.874
6.448
7.080
7.754
8.459
9.187
9.933
10.691
11.459
12.234
13.015
13.801
14.591
15.383
16.179
16.976
17.775
18.576
19.378
20.181
20.985
21.790
22.596
RGIL/MRAMKL
(mg/kg-day/
jimol/hr/kg liver)
0.2069
0.2081
0.2092
0.2104
0.2117
0.2129
0.2142
0.2155
0.2169
0.2183
0.2344
0.2544
0.2782
0.305
0.335
0.367
0.401
0.435
0.470
0.506
0.543
0.579
0.616
0.653
0.691
0.728
0.766
0.803
0.841
0.879
0.917
0.955
0.993
1.031
1.069
C-19
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-ll. Interspecies conversion factors based on MRAMKL dose metric
(VMAXC=1.70)
EC
270
280
290
300
EC
1698
1761
1824
1887
MRAMKL
72.57
72.75
72.92
73.07
RGIL
80.34
83.30
86.28
89.24
RGIL/EC
0.04730
0.04730
0.04730
0.04729
EC/MRAMKL
23.403
24.210
25.017
25.825
RGIL/MRAMKL
1.107
1.145
1.183
1.221
EC, air exposure concentration; MRAMKL, time-averaged rate of metabolism of carbon tetrachloride; RGIL, rate of
uptake of carbon tetrachloride from GI-tract to liver; VMAXC, maximum rate of metabolism of carbon tetrachloride
(mg/hr/kg BW070).
C-20
DRAFT - DO NOT CITE OR QUOTE
-------
y=B4.6658x
R2-=O:999& -
0123456
400
350
300
§250
1200
0150
100
50
0
MCA (1 -5), VMAXC=0.40
0123456
MCA (umol/L)
80
70
60
£50
I 40
O 30
LU
20
10
MCA (0.01 - 1), VMAXC=0.40
0.0
0.2
0.4 0.6
MCA (umol/L)
0.8
1.0
?
0
-D
£
o
LU
c _,
n
-1 ^
c
MCA (0.01 - 5), VMAXC=0.40
i «*•••••••"
i ;-"--
}- 4 i.«u**2 --3-- --4-- --5-- --(>
MCA (umol/L)
8 -I
0 n "
"° n
<£ u
0 -1 ^
^ 2
MCA(1 -5), VMAXC=0.40
«.
}- -A-- --2-- -3- --4-- --&-- -fi
MCA (umol/L)
C
J3 1
0 n "
~° n
^ u
0 -1 4
LU
c
MCA (0.01 - 1), VMAXC=0.40
•
^ m
\± ' ' . ' •
}- ---Q--'-- 0-- -4- --1-- --\- -A
MCA (umol/L)
Figure C-4. Relationship between internal dose metric MCA (time-averaged
arterial blood concentration of carbon tetrachloride) and equivalent
exposure concentration (EC, left panel) and values for % delta for trend lines
(right panel). VMAXC=0.40 mg/hour/kg BW "
0.70
C-21
DRAFT - DO NOT CITE OR QUOTE
-------
234
MCA (umol/L)
400
350
300
|250
I 200
O 150
LU
100
50
0
MCA(1 - 5), VMAXC=0.65
23456
MCA (umol/L)
0.4 0.6
MCA (umol/L)
S3
0
•Q
£
o
III
m -,
,
o
*<
m
MCA (0.01 - 5), VMAXC=0.65
I
1 .*•••••'
I .-..•-• ""
• ««w
J'-.i^i.fc-*"" 2- --3-- --4-- --5-- --(>
MCA (umol/L)
MCA (umol/L)
i1
0
T3
#
O
LLI
5
4 -
3
2
1 -
-10
-2
-3
-4 -
MCA (0.01 - 1), VMAXC=0.65
•
_ _____ f ^
Q 0:2- - - -0-.4 0.6 -0:8 -HO- 1-2
MCA (umol/L)
Figure C-5. Relationship between internal dose metric MCA (time-averaged
arterial blood concentration of carbon tetrachloride) and equivalent
exposure concentration (EC, left panel) and values for % delta for trend lines
(right panel). VMAXC=0.65 mg/hour/kg BW "
0.70
C-22
DRAFT - DO NOT CITE OR QUOTE
-------
2 3
MCA (umol/L)
400
350
300
f 250
I 200
O 150
LU
100
50
0
MCA (1 -5), VMAXC=1.49
23456
MCA (umol/L)
80
70
60
fso
'40-
O 30
LU
20
10 --
0.0 0.2 0.4 0.6 0.8
MCA (umol/L)
1.0
1.2
10 -p
8 --
6 --
1 2
3 0
O -2
MCA(1 -5), VMAXC=1.49
MCA (umol/L)
s
0
-o
^
O
LU
0 -
-1 4
c, _
MCA (1 -5), VMAXC=1.49
•
.••;••••
}-- -4-- --2r- -3-- --4-- --',)
MCA (umol/L)
0
O
4 -
3
2 -
1 -
n
-10
-3
-4 -
c
MCA (0.01 - 1), VMAXC=1.49
v T . •
0 *OT2* - * -0.4 €.6 -O8 1-.& 42
MCA (umol/L)
Figure C-6. Relationship between internal dose metric MCA (time-averaged
arterial blood concentration of carbon tetrachloride) and equivalent
exposure concentration (EC, left panel) and values for % delta for trend lines
(right panel). VMAXC=1.49 mg/hour/kg BW
0.70
C-23
DRAFT - DO NOT CITE OR QUOTE
-------
350
MCA (umol/L)
400
234
MCA (umol/L)
80
70
60
fso
'40-
O 30
LU
20
10 --
0.0 0.2 0.4 0.6 0.8
MCA (umol/L)
1.0 1.2
10 -p
8 --
6 --
4 -L
MCA(1 -5), VMAXC=1.70
0
-2
MCA (umol/L)
s
0
-o
^
O
LU
-1 ^
c _
MCA (1 -5), VMAXC=1.70
•
"••- ,,..«-•*'
>-. ' :,:_'-i-T.»-^-_i T^ '.:i;
MCA (umol/L)
0
O
4 -
3
2 -
1 -
n
-10
-3
-4 -
c
MCA (0.01 - 1), VMAXC=1.70
V ---------,
0 "Or2* - ' -0.4 £>.6 -O8 H> 42
MCA (umol/L)
Figure C-7. Relationship between internal dose metric MCA (time-averaged
arterial blood concentration of carbon tetrachloride) and equivalent
exposure concentration (EC, left panel) and values for % delta for trend lines
(right panel). VMAXC=1.70 mg/hour/kg BW
0.70
C-24
DRAFT - DO NOT CITE OR QUOTE
-------
234
MCA (umol/L)
50
40 --
D)
E
30
20 --
10 --
MCA (0.1 -5), VMAXC=0.40
234
MCA (umol/L)
0.00
0.02
0.04 0.06
MCA (umol/L)
0.08
0.10
f s
"0
_l
O
o:
19 -,
i
12 -
> MCA (0.01 - 5), VMAXC=0.40
i
| >.«*** 23451)
"*"
MCA (umol/L)
8
-o
^
_l
O
0^!
c _,
-1 <
5 _
MCA (0.1 -5), VMAXC=0.40
..
^v'-i-- --2-- --a- V""-«& --is
MCA (umol/L)
4
3
$ 2
0
-a 1
e o
J5 -10:
K -2
-3
-4
5 _
MCA (0.01 - 0.1), VMAXC=0.40
. . ' '
' ' '*' ' . ' . ' '.' ' • ' *
30-- -Q3? 0:04 Ofl6 0.08 0.10
MCA (umol/L)
Figure C-8. Relationship between internal dose metric MCA (time-averaged
arterial blood concentration of carbon tetrachloride) and equivalent rate of
uptake from GI tract to liver (RGIL, left panel) and values for % delta for
trend lines (right panel). VMAXC=0.40 mg/hour/kg BW
0.70
C-25
DRAFT - DO NOT CITE OR QUOTE
-------
0.0
234
MCA (umol/L)
234
MCA (umol/L)
0.00
0.05
0.10
MCA (umol/L)
0.15
0.20
1"
•Q
u9
_l
o
a:
20
4 1
19 -
MCA (0.01 - 5), VMAXC=0.65
>MH,M*«««**«*******«««»«**
[ _.••!** 2 _3 4_ 5 !>
•X
MCA (umol/L)
"0 -1
-D 1
^ 0
cl-1^
5 _
UCA_(Q-2 - 5) VMAXC-CL65
.X*"****""*"*^»»»...
>-^.*"*1- --2-- --5- - -*?-"--.i-.H5- - --(!
MCA (umol/L)
C
4
3
? 2-
« 1
# n
cl-1°^
a: -2
-3
-4
c
MCA (0.01 - 0.2), VMAXC=0.65
•
•
90-- * --^.SS5-"- --Or1-0-- --OH5-- --0.
MCA (umol/L)
20
Figure C-9. Relationship between internal dose metric MCA (time-averaged
arterial blood concentration of carbon tetrachloride) and equivalent rate of
uptake from GI tract to liver (RGIL, left panel) and values for % delta for
trend lines (right panel). VMAXC=0.65 mg/hour/kg BW
•0.70
C-26
DRAFT - DO NOT CITE OR QUOTE
-------
234
MCA (umol/L)
50
^40
01
T3
jj? 30
^ 20
10 --
MCA (0.2-5), VMAXC=1.49
234
MCA (umol/L)
3.5
3.0 -
f 2.5-
2.0 -
• 1.5 -
1.0 --
0.5 -
0.0
_M_CA (0.01_-J12)LVMAXC=149
0.00
0.05
0.10
MCA (umol/L)
0.15
0.20
s^
0
-o
^
_l
o
a:
Art
n
(
on -
• MCA (0.01 -5), VMAXC=0.1.49
,
1 ...«•"" 2 3 4 !>
"••.•**""
MCA (umol/L)
S^
0
-D
£3
_i
O
a:
ACi
o -
(
on
MCA (0.2-5), VMAXC=1.49
•
•f"": """•
31234!)
MCA (umol/L)
1
^ n
c
MCA (0.01 -0.2), VMAXC=1.49 *
.
.
•
MCA (umol/L)
20
Figure C-10. Relationship between internal dose metric MCA (time-
averaged arterial blood concentration of carbon tetrachloride) and
equivalent rate of uptake from GI tract to liver (RGIL, left panel) and values
for % delta for trend lines (right panel). VMAXC=1.49 mg/hour/kg BW "
0.70
C-27
DRAFT - DO NOT CITE OR QUOTE
-------
234
MCA (umol/L)
50
40 -
30 --
20 -
10 --
MCA (0.2-5), VMAXC=1.70
__ _MCA10JJ1-_OJ1VMAXC=170_
3 --
3 -
2 --
2
1
1 --
23456
MCA (umol/L)
0.00 0.05 0.10 0.15
MCA (umol/L)
0.20
"® 5
—i ^ j
CD '5<
MCA (0.01 - 5X_VMAXC=0.1.70
§
V- :: :
r x
MCA (umol/L)
jg 2 -
0 •)
-D 1 "
^ o -
cl-1(
C
MCA (0.2 -5), VMAXC=1.70
,..••••"""""•••••....
>- 2*'* --2-- --3-- --4-*"- !i
**•"*"
MCA (umol/L)
jg 2 -
^ n -
c
MCA (0.01 - 0.2), VMAXC=1.70
•
.
X- '^^'' —0:10- -^15-- -0:?0
MCA (umol/L)
Figure C-ll. Relationship between internal dose metric MCA (time-
averaged arterial blood concentration of carbon tetrachloride) and
equivalent rate of uptake from GI tract to liver (RGIL, left panel) and values
for % delta for trend lines (right panel). VMAXC=1.70 mg/hour/kg BW "
0.70
C-28
DRAFT - DO NOT CITE OR QUOTE
-------
60
50 -
40 -
1 30
O
m
20-
10 -
Vmax = 0.40
fit = 0.49541 *MRAMKL +
15.082*MRAMKL/(18.176 - MRAMKLjj
6 9
MRAMKL
12
0.03
- 0.02
70
0.01 8
E
- o
- -0.01 ^s
- -0.02
-0.03
15
12-
10-
8-
I 6-
O
LLI
4-
2-
0-
C
Vmax =1.49 X
fit = 0.49547* MRAMKL + x'
15.078*MRAMKL/(67.700 - MRAMKL)
if
/
/
/
/
D 3 6 9 12 1
^ 0.03
- 0.02
- 0.01 $
Q.
C
fi)
- 0 Q
3
- -0.01 -gs
- -0.02
- -0.03
5
MRAMKL
E
O
LJJ
\2
(
Vmax = 0.65 ^ /
fit = 0.49544*MRAMKL + ~S~
15.081*MRAMKL/(29.535 - MRAMKL) /
~ /
/
^^
D 3 6 9 12 1
MRAMKL
- 0.01 tfl
a!
c
fi)
n>
3
- n rn
5
1 O
8-
. — .
°- K
a. o -
0
9
c
Vmax = 1.70 y
' /
fit = 0.49543*MRAMKL + X
15.082*MRAMKL/(77.246 - MRAMKL)X
i/
/
./
/
/_
/
i i i i
) 3 6 9 12 1
MRAMKL
n n^
7J
- 0.01 %
a.
D
0
3
n ro
5
Figure C-12. Relationship between internal dose metric MRAMKL (mean
rate of carbon tetrachloride metabolism in the liver) and equivalent exposure
concentration (EC) and values for % delta for trend lines.
C-29
DRAFT - DO NOT CITE OR QUOTE
-------
18
15-
12-
Vmax = 0.40
fit = 0.14773* MRAMKL+
4.4926*MRAMKL/(18.176- MRAMKL)]
6
3-
6 9
MRAMKL
12
-0.03
15
7 R -,
5"
^)
^ A^
O)
_l
— o
o:
s
c
Vmax = 0.65 ;
fit =0.14770*MRAMKL+ - — -/
4.4925*MRAMKL/(29.535 - MRAMKL) /
m /• • •
'•-••—• .•/.-.- : .'
/
^
) 3 6 9 12 1
MRAMKL
n (Y3
- 0.02
- 0.01 »
- o.oo |
^
- -0.01 ^
- -0.02
n (Y3
5
3.6
3
, 2.4 -
i. 1,
_|
I 1-2-
0.6 -
Vmax = 1.49
fit =0.14767*MRAMKL +
4.4929*MRAMKL/(67.711 - MRAMI
6 9
MRAMKL
12
0.02
0.03
0.01
70
S.
c
o.oo j-
3
-0.01 "p
-0.02
-0.03
15
3.6 -i
I 2.4
^
P
£.1.8
(5
0.6
Vmax = 1 .70 *
fit = 0.14767*MRAMKL + -7^
4.4929*MRAMKL/(77.252 - MRA^H
-------
C.4. Sensitivity Analysis
Univariate sensitivity analysis consisted of running the model after perturbing values for
single parameters by a factor of 0.01, in the up and down directions. Parameter sensitivities were
assessed from comparison of standardized sensitivity coefficients:
e^_ ^,,A_/(* + Ax)-/(*-A) x
/(*)
Eq. (1)
where SC is the standardized sensitivity coefficient, f(x) is the output variable at parameter value
x, and ) is the perturbation of x (i.e., O.Olx). Figures C-14 and C-15 show sensitivity coefficients
for each internal dose metric (i.e., MCA, MRAMKB) for the human model. Absolute values of
sensitivity coefficients that were >0.01 are shown in these figures. Conversion to absolute value
removes information on the direction of the change in the output variable, allowing the
magnitudes of the influence of each parameter on the output variable to be directly compared.
Parameters having sensitivity coefficients >0.1 can be considered to be highly influential
parameters. Chemical parameters in this category (i.e., sensitivity coefficient >0.1) are shown in
Table C-12 (indicated with +). The mouse and rat models yielded the same rank order of
sensitivity coefficients as the human model.
Table C-12. Sensitive parameters (indicated with +) in the human model
Parameter
PB
PL
PF
PS
PR
VmaxC
Km
Definition
Blood:air partition coefficient
Liverblood partition coefficient
Fatblood partition coefficient
Slowly-perfused:blood partition coefficient
Readily-perfused:blood partition coefficient
Maximum rate of metabolism (mg/hour-kg B W)
Michaelis-Menten coefficient for metabolism
(mg/L)
Internal Dose Metric
MCA
+
+
+
MRAMKB
+
+
+
+
C-31
DRAFT - DO NOT CITE OR QUOTE
-------
SC - MCA 960 hr
1.0E-02 1.0E-01 1.0E+00
Parameter
QRC
PBLD
QSC
QIC
QPC
QCC
VMAXC
KMX
QFC
VFC
PF
PS
VSC
PR
VRC
VLC
PL
I
I
I
I
I
I
I
I
Figure C-14. Standardized sensitivity coefficients for the MCA dose metric
(average concentration of carbon tetrachloride in blood, umol/L) simulated
with the human carbon tetrachloride PBPK model.
Absolute values of coefficients >0.01 are shown. The simulation was of a
continuous exposure to 2.5 ppm for 980 hours (rank order of sensitivity
coefficients was not dependent on exposure time).
C-32
DRAFT - DO NOT CITE OR QUOTE
-------
SC-MRAMKL960 hr
1.0E-02 1.0E-01 1.0E+00
I i i i i i i i i 1 i i i i i i i i
VLC
PR
QRC
PS
PL
VMAXC
QCC
fc VSC
•5
E VRC
CO
S. QSC
QFC
PF
QPC
VFC
QIC
PBLD
KMX
1
1
1
1
1
1
1
1
1
Figure C-15. Standardized sensitivity coefficients for the MRAMKL dose
metric (average rate of metabolism of carbon tetrachloride umol/hr/kg liver)
simulated with the human carbon tetrachloride PBPK model.
Absolute values of coefficients >0.01 are shown. The simulation was of a
continuous exposure to 2.5 ppm for 980 hours (rank order of sensitivity
coefficients was not dependent on exposure time).
C-33
DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX D. BENCHMARK DOSE MODELING FOR DERIVING THE RfC
MALE RAT:
Incidence data for fatty changes of the liver
Male F344 rats exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25, 125 ppm
BMR=10%
Model
Vmax = 0.4
AIC
X2 /> value"
BMC10
BMCL10
Vmax = 0.65
AIC
XV value3
BMC10
BMCL10
MCA (umol/L)
Gammab
Logistic0
Log-Logistic'
Multistage 1-
degree4'
Probif
Log-probitc
Quantal-linear
Weibullb
144.336
155.104
137.403
142.388
169.521
138.408
142.388
142.388
0.0007
0.0000
0.4355
0.0074
0.0000
0.1761
0.0074
0.0074
0.0793248
0.170834
0.136715
0.0714015
0.22329
0.124953
0.0714017
0.0714016
0.0551873
0.137191
0.0790319
0.0550523
0.17626
0.0755939
0.0550523
0.0550523
144.772
156.51
137.463
142.778
171.234
138.529
142.778
142.778
0.0005
0.0000
0.4087
0.0031
0.0000
0.1581
0.0031
0.0031
0.0689847
0.157857
0.123076
0.0665234
0.21463
0.112257
0.0665234
0.0665235
0.051179
0.126743
0.0707077
0.0511645
0.168317
0.0803264
0.0511645
0.0511645
MRAMKL (umol/hr-kg liver)
Gammab
Logistic'
Log- Logistic0
Multistage 2-
degree11 f
Probit0
Log-probitc
Quantal-linear
Weibullb
137.468
136.747
136.933
137.073
138.891
136.871
151.674
138.997
0.4177
0.3444
0.8012
0.2702
0.0826
0.9538
0.0008
0.1316
3.98707
3.25675
4.56744
3.55184
2.97807
4.27176
1.01942
3.34831
2.6343
2.58557
3.08461
2.02617
2.41619
3.06539
0.831472
2.18252
137.338
136.513
136.996
138.991
138.712
136.872
148.898
138.601
0.4760
0.3671
0.7246
0.0944
0.0728
0.9470
0.0025
0.1751
5.31098
4.60057
6.20422
4.99656
4.23817
5.73628
1.45532
4.4781
3.35649
3.65284
4.00273
2.5022
3.44383
3.97844
1.18412
2.81908
a Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the % test.
b Power restricted to >1.
0 Slope restricted to >1.
dUsed smallest degree polynomial available with an adequate fit; the 2- and 3-degree polynomials provided the same fit as the
1-degree.
e Betas restricted to >0.
f Used smallest degree polynomial available with an adequate fit; the 3-degree polynomial provided the same fit as the 2-
degree.
D-l
DRAFT - DO NOT CITE OR QUOTE
-------
FEMALE RAT:
Incidence data for fatty changes of the liver
Female F344 rats exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25,125ppm
BMR=10%
None of the models in HMDS provided an adequate fit of the female rat data.
Incidence data for fatty changes of the liver
Female F344 rats exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25 ppm [high dose dropped]
BMR=10%
Model
Vmax = 0.4
AIC
X2 /> value3
BMC10
BMCL10
Vmax = 0.65
AIC
X2/> value"
BMC10
BMCL10
MCA (umol/L)
Gammab
Logistic0
Log-Logistic0
Multistage"6
Probit0
Log-probit°
Quantal-linear
Weibullb
92.9928
93.4185
92.9928
2n" degree
92.4089
3rd degree
94.9928
93.6833
92.9928
111.424
92.9928
NA
0.1121
NA
0.2442
NA
0.0968
NA
0.0000
NA
0.187771
0.106984
0.182663
0.123631
0.213915
0.100288
0.174053
0.0363563
0.213201
0.107455
0.0803379
0.111838
0.0851972
0.090506
0.0779817
0.112578
0.0277405
0.102923
92.9928
93.3172
92.9928
2"" degree
92.3049
3rd degree
92.9928
93.5689
92.9928
111.025
92.9928
NA
0.1201
NA
0.2617
NA
0.1043
NA
0.0001
NA
0.170979
0.0979754
0.166144
0.113721
0.195194
0.0919928
0.158234
0.0332712
0.194228
0.0971536
0.0734707
0.101213
0.0775873
0.08177
0.0714911
0.101889
0.0253689
0.0930656
MRAMKL (umol/hr-kg liver)
Gammab
Logistic0
Log- Logistic0
Multistage"6
Probit0
Log-probit°
Quantal-linear
Weibullb
92.9928
99.7262
92.9928
2nd degree
100.7
3r" degree
92.2866
100.988
92.9928
127.034
92.9928
NA
0.0020
NA
0.0039
0.2650
0.0013
NA
0.0000
NA
4.85516
2.45785
4.84705
2.43344
3.76974
2.16088
4.69168
0.817323
5.3798
3.42634
1.90371
3.48106
1.99357
2.82488
1.70134
3.49658
0.634088
3.29131
92.9928
97.8675
92.9928
2ntl degree
98.1134
3r" degree
91.5964
98.8142
92.9928
123.548
92.9928
NA
0.0064
NA
0.0124
0.4421
0.0044
NA
0.0000
NA
6.52318
3.34536
6.48806
3.42266
5.42354
2.98448
6.26103
1.12472
7.27174
4.43018
2.58247
4.51798
2.75565
3.74923
2.34695
4.54001
0.870515
4.24944
a Values <0. 1 fail to meet conventional goodness-of-fit criteria; p value from the
b Power restricted to >1 .
0 Slope restricted to >1.
dUsed smallest degree polynomial available with an adequate fit.
e Betas restricted to >0.
test.
D-2
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat
Dose metric: MCA
Vmax = 0.4 mg/hour/kg BW
0.07
Log-Logistic Model with 0.95 Confidence Level
0.8
O
0.2
Log-Logistic
EJMDL
BMD
0.5
1.5
2
dose
2.5
3.5
10:5910/122007
Logistic Model. (Version: 2.9; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE RAT\MCA-
VMAX=0.4\RAT-FATTYLIVER-MCA-4.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE
RAT\MCA-VMAX=0.4\RAT-FATTYLIVER-MCA-4.pit
Fri Oct 12 10:59:34 2007
BMDS MODEL RUN
The form of the probability function is:
P [response] = background+ (1-background) / [1+EXP (-intercept-slope*Log (dose) ) "_
Dependent variable = FattyLiver
Independent variable = umol/L
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.08
intercept = 1.42536
slope = 1.89476
D-3
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
background intercept slope
background
intercept
slope
-0.077
0.34
-0.077
1
0.54
0.34
0.54
1
Parameter Estimates
Variable
background
intercept
slope
Estimate
0.073606
1.74202
1.97967
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-65.434
-65.7017
-138.619
137.403
# Param's
4
3
1
Deviance Test d.f.
0.535433
146.371
P-value
0.4643
<.0001
Goodness of Fit
Dose
0.0000
0.1280
0.7080
3.8920
Est. Prob.
0.0736
0.1559
0.7614
0.9891
Expected
3 .680
7.796
38.068
49.456
Observed
4
7
39
49
Size
50
50
50
50
Scaled
Residual
0.173
-0.310
0.309
-0.621
Chi 2 =0.61
d.f. = 1
P-value = 0.4355
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 0.136715
BMDL = 0.0790319
D-4
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat
Dose metric: MCA
Vmax = 0.65 mg/hour/kg BW
0.07
Log-Logistic Model with 0.95 Confidence Level
0.8
0-6
0.2
Log-Logistic
EiMDL
BMD
0
11:1210/122007
0.5
1.5 2
dose
2.5
3.5
Logistic Model. (Version: 2.9; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE RAT\MCA-
VMAX=0.65\RAT-FATTYLIVER-MCA-65.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE
RAT\MCA-VMAX=0.65\RAT-FATTYLIVER-MCA-65.plt
Fri Oct 12 11:12:25 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))\
Dependent variable = FattyLiver
Independent variable = umol/L
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.08
intercept = 1.54201
slope = 1.85672
D-5
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
background intercept slope
background 1 -0.05 0.33
intercept -0.05 1 0.6
slope 0.33 0.6 1
Parameter Estimates
Variable
background
intercept
slope
Estimate
0.0733292
1.88323
1.94775
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-65.434
-65.7316
-138.619
137.463
# Param's
4
3
1
Deviance Test d.f.
0.595159
146.371
P-value
0.4404
<.0001
Goodness of Fit
Dose
0.0000
0.1160
0.6530
3 .7750
Est. Prob.
0.0733
0.1568
0.7603
0.9895
Expected
3 .666
7.841
38.017
49.476
Observed
4
7
39
49
Size
50
50
50
50
Scaled
Residual
0.181
-0.327
0.326
-0.661
Chi 2 =0.68
d.f. = 1
P-value = 0.4087
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 0.123076
BMDL = 0.0707077
D-6
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat
Dose metric: MRAMKL
Vmax = 0.4 mg/hour/kg BW
0.07
Logistic Model with 0.95 Confidence Level
0.8
0.2
0
Logistic
BMDL
BMD
0 5
11:1710/122007
10 15
dose
20
25
Logistic Model. (Version: 2.9; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE RAT\MRAMKL-
VMAX=0.4\FATTY_LIVER_MRAMKL-4.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE
RAT\MRAMKL-VMAX=0.4\FATTY_LIVER_MRAMKL-4.pit
Fri Oct 12 11:17:49 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = I/[1+EXP(-intercept-slope*dose)]
Dependent variable = FattyLiver
Independent variable = umol/hr-kgL
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
background = 0 Specified
intercept = -2.35241
slope = 0.249767
Asymptotic Correlation Matrix of Parameter Estimates
D-7
DRAFT - DO NOT CITE OR QUOTE
-------
intercept
slope
*** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept slope
1 -0.82
-0.82 1
Variable
intercept
slope
Parameter Estimates
Estimate Std. Err.
-2.68587 0.383165
0.309634 0.0415113
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-3.43685 -1.93488
0.228273 0.390994
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood)
-65.434
-66.3737
-138.619
136.747
# Param's
4
2
1
Deviance Test d.f.
1.87944
146.371
P-value
0.3907
<.0001
Goodness of Fit
Dose
0.0000
3.8130
12.0920
24.3200
Est. Prob.
0.0638
0.1816
0.7424
0.9922
Expected
3 .191
9.082
37.118
49.609
Observed
4
7
39
49
Size
50
50
50
50
Scaled
Residual
0.468
-0.764
0.609
-0.979
=2.13
d.f. = 2
P-value = 0.3444
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 3.25675
BMDL = 2.58557
D-8
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat
Dose metric: MRAMKL
Vmax = 0.65 mg/hour/kg BW
0.07
0.8
0-6
0.2
0
Logistic Model with 0.95 Confidence Level
Logistic
BMDL 3MD
0 5
11:2310/122007
10 15 20 25 30 35
dose
Logistic Model. (Version: 2.9; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE RAT\MRAMKL-
VMAX=0.65\MRAT_FATTY_LIVER_MRAMKL-65.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\MALE
RAT\MRAMKL-VMAX=0.65\MRAT_FATTY_LIVER_MRAMKL-65.pit
Fri Oct 12 11:23:29 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = I/ [1+EXP (-intercept-slope*dose) "_
Dependent variable = FattyLiver
Independent variable = umol/hr-kgL
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
background = 0 Specified
intercept = -2.28912
slope = 0.166325
Asymptotic Correlation Matrix of Parameter Estimates
D-9
DRAFT - DO NOT CITE OR QUOTE
-------
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept
intercept 1
slope -0.8
slope
-0.8
1
Parameter Estimates
Variable
intercept
slope
Model
Full model
Fitted model
Reduced model
Estimate
-2 .59592
0.207777
Analysis
Log (likelihood)
-65.434
-66 .2567
-138.619
Std. Err. Lower
0.370821
0.0278282
of Deviance Table
# Param's Deviance Test
4
2 1.64536
1 146.371
Conf.
-3 .3
0.15
d.f .
2
3
95.0% Wald Confidence Interval
lit Upper Conf. Limit
! -1.86913
i 0.26232
P-value
0.4393
<.0001
AIC:
136.513
Goodness of Fit
Dose
0.0000
4 .9910
17.6260
36.2660
Est. Prob.
0.0694
0.1738
0.7439
0.9929
Expected
3 .470
8.690
37.195
49.645
Observed
4
7
39
49
Size
50
50
50
50
Scaled
Residual
0.295
-0.631
0.585
-1.085
Chi 2 =2.00
d.f. = 2
P-value = 0.3671
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 4.60057
BMDL = 3.65284
D-10
DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat
Dose metric: MCA
Vmax = 0.4 mg/hour/kg BW
0.07
o
0.4
0.2
0 :
Multistage Model with 0.95 Confidence Level
Multistage
BMDL BMD
0 0.1
11:4210/122007
0.2
0.3 0.4
dose
0.5
0.6
0.7
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE RAT\MCA-
VMAX=0.4\FRAT-FATTYLIVER-MCA-4.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE
RAT\MCA-VMAX=0.4\FRAT-FATTYLIVER-MCA-4.pit
Fri Oct 12 11:42:22 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2) ]
The parameter betas are restricted to be positive
Dependent variable = FattyLiver
Independent variable = umol/L
Total number of observations = 3
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0746099
Beta(l) = 0
Beta(2) = 7.64624
D-ll
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background
Beta(2)
Background
1
-0.21
Beta(2)
-0.21
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0.0951491
0
6.89319
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-43 .4964
-44.2044
-101.707
92.4089
# Param's
3
2
1
Deviance Test d.f.
1.41613
116.422
P-value
0.234
<.0001
Goodness of Fit
Dose
0.0000
0.1280
0.7080
Est. Prob.
0.0951
0.1918
0.9714
Expected
4.757
9.589
48.571
Observed
6
7
49
Size
50
50
50
Scaled
Residual
0.599
-0.930
0.364
Chi^2 = 1.36
d.f. = 1
P-value = 0.2442
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
0.123631
0.0851972
0.148857
Taken together, (0.0851972, 0.148857) is a 90
interval for the BMD
% two-sided confidence
D-12
DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat
Dose metric: MCA
Vmax = 0.65 mg/hour/kg BW
0.07
Multistage Model with 0.95 Confidence Level
0.8
0.6
0.4
0.2
Multistage
0 :
BMDL
BMD
0 0.1
11:4710/122007
0.2
0.3
dose
0.4
0.5
0.6
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE RAT\MCA-
VMAX=0.65\FRAT-FATTYLIVER-MCA-65.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE
RAT\MCA-VMAX=0.65\FRAT-FATTYLIVER-MCA-65.plt
Fri Oct 12 11:47:23 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2) ]
The parameter betas are restricted to be positive
Dependent variable = FattyLiver
Independent variable = umol/L
Total number of observations = 3
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0765787
Beta(l) = 0
Beta(2) = 8.98383
D-13
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background
Beta(2)
Background
1
-0.21
Beta(2)
-0.21
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0.095736
0
8.14699
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-43 .4964
-44.1525
-101.707
92.3049
# Param's
3
2
1
Deviance Test d.f.
1.31215
116.422
P-value
0.252
<.0001
Goodness of Fit
Dose
0.0000
0.1160
0.6530
Est. Prob.
0.0957
0.1896
0.9720
Expected
4.787
9.481
48.599
Observed
6
7
49
Size
50
50
50
Scaled
Residual
0.583
-0.895
0.344
Chi^2 = 1.26
d.f. = 1
P-value = 0.2617
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
0.113721
0.0775873
0.137047
Taken together, (0.0775873, 0.137047) is a 90
interval for the BMD
% two-sided confidence
D-14
DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat
Dose metric: MRAMKL
Vmax = 0.4 mg/hour/kg BW
0.07
<
o
0.8
0.6
0.4
0.2
0 :
Multistage Model with 0.95 Confidence Level
Multistage
giypL
BMD
6
dose
10
12
11:5210/122007
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE
RAT\MRAMKL-VMAX=0.4\FRAT_FATTY_LIVER_MRAMKL-4.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE
RAT\MRAMKL-VMAX=0.4\FRAT_FATTY_LIVER_MRAMKL-4.pit
Fri Oct 12 11:52:42 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2-beta3*dose^3) ]
The parameter betas are restricted to be positive
Dependent variable = FattyLiver
Independent variable = umol/hr-kgL
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0769299
Beta(l) = 0
D-15
DRAFT - DO NOT CITE OR QUOTE
-------
Beta(2) =
Beta(3) =
0.00216647
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l) -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(3)
Background
1
-0.21
Beta(3)
-0.21
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Beta(3)
Estimate
0.0958436
0
0
0.00196673
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-43.4964
-44.1433
-101.707
92.2866
# Param's
3
2
1
Deviance Test d.f.
1.29386
116.422
P-value
0.2553
<.0001
Dose
0.0000
3.8130
12.0920
Est. Prob.
0.0958
0.1892
0.9721
Goodn
Expected
4 .792
9.462
48.603
.ess of Fi
Observed
6
7
49
t
Size
50
50
50
Scaled
Residual
0.580
-0.889
0.340
Chi 2 = 1.24
d.f. = 1
P-value = 0.2650
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
3.76974
2.82488
4.26949
Taken together, (2.82488, 4.26949) is a 90
interval for the BMD
% two-sided confidence
D-16
DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat
Dose metric: MRAMKL
Vmax = 0.65 mg/hour/kg BW
0.07
<
o
0.8
0.6
0.4
0.2
0 :
Multistage Model with 0.95 Confidence Level
Multistage
• BMDL B.MP ~-
0246
11:5710/122007
10 12 14 16 18
dose
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE
RAT\MRAMKL-VMAX=0.65\FRAT_FATTY_LIVER_MRAMKL-65.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\RFC RAT LIVER\FEMALE
RAT\MRAMKL-VMAX=0.65\FRAT_FATTY_LIVER_MRAMKL-65.pit
Fri Oct 12 11:57:06 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2-beta3*dose^3) ]
The parameter betas are restricted to be positive
Dependent variable = FattyLiver
Independent variable = umol/hr-kgL
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0
Beta(3) = 0.000714264
D-17
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l) -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(3)
Background
1
-0.19
Beta(3)
-0.19
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Beta(3)
Estimate
0.101433
0
0
0.000660435
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-43.4964
-43.7982
-101.707
91.5964
# Param's
3
2
1
Deviance Test d.f.
0.603632
116.422
P-value
0.4372
<.0001
Dose
0.0000
4.9910
17.6260
Est. Prob.
0.1014
0.1723
0.9759
Goodn
Expected
5.072
8.613
48.793
.ess of Fi
Observed
6
7
49
t
Size
50
50
50
Scaled
Residual
0.435
-0.604
0.191
Chi 2 =0.59
d.f. = 1
P-value = 0.4421
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
5.42354
3.74923
6.17189
Taken together, (3.74923, 6.17189) is a 90
interval for the BMD
% two-sided confidence
D-18
DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX E. CANCER ASSESSMENT: BMD MODELING OUTPUTS FOR LOW-
DOSE LINEAR EXTRAPOLATION APPROACH
E.I. BMD Analysis
Liver tumors (adenoma + carcinoma)
Female F344 rats exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25, 125 ppm
Multistage: MCA: 2-stage model MRAMKL: 4-stage model
BMR
(extra risk)
Vmax = 0.4
AIC
0.05
61.6602
0.05
63.3399
value"
0.9842
0.6503
BMC
0.609955
9.8151
BMCL
BMR/
BMCL
Vmax = 0.65
AIC val
MCA
(jimol/L)
0.387377
0.129
61.5904 0.9
MRAMKL
(jimol/hr-kg liver)
8.40334
0.00595
62.8343 0.7
ue" BMC BM
316 0.588686 0.35
440 14.582 12.2
BMR/
CL BMCL
4766 0.141
867 0.00407
a Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the %2 test.
Liver tumors (adenoma + carcinoma)
Female F344 rats exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25 ppm
Multistage: 2-stage model
BMR
(extra risk)
Vmax = 0.4
AIC
xV
value"
BMC
BMCL
BMR/
BMCL
Vmax = 0.65
Z3P
AIC value" BMC BM
BMR/
CL BMCL
MCA
(jimol/L)
0.05
24.8957
0.9507
0.655398
0.345984
0.144
24.8889 0.9523 0.604144 0.31
7726 0.157
MRAMKL
Qimol/hr-kg liver)
0.05
25.2825
0.8571
11.5604
6.92352
0.00722
25.1734 0.8831 16.6986 9.76
339 0.00512
a Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the % test.
Note: 3-stage model did not provide a sufficiently improved model fit.
E-l
DRAFT - DO NOT CITE OR QUOTE
-------
Liver tumors (adenoma + carcinoma)
Female BDF1 mouse exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25 ppm
Multistage: MCA: 2-stage model MRAMKL: 2-stage model
BMR
(extra risk)
Fisher
AIC
0.1
117.307
^P
value"
NA
BMC
| 0.10186
BMCL
BMR/
BMCL
MCA
(jimol/L)
0.0467576
2.14
Thrall
x2
AIC vali
| 117.307 N^
P
iea BMC BM
I 0.194624 0.08S
BMR/
CL BMCL
S5305 1.13
MRAMKL
(jimol/hr-kg liver)
0.1
115.912
0.4437
| 9.70893
6.3204
0.0158
| 117.341 | O.K
554 10.4557 l.K
>255 0.0132
a Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the II test.
Note: 3-stage model did not provide a sufficiently improved model fit.
Liver tumors (adenoma + carcinoma)
Female BDF1 mouse exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day,
Doses modeled: 0, 5 ppm
Multistage; 2-stage model
BMR
(extra risk)
Fisher
AIC
0.1
80.6149
0.1
80.6149
x'/»
value"
NA
NA |
BMC BM<
0.101967 | 0.04/
11.6352 5.046
BMR/
CL BMCL
5 days/week)
Thrall
x'/»
AIC value" BMC BM
MCA
(jimol/L)
1224 2.26
80.6149 NA 0.195666 0.08^
MRAMKL
(jimol/hr-kg liver)
31 0.0198
80.6149 NA 14.1982 6.1!
BMR/
CL BMCL
L8621 1.18
788 0.0162
a Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the %2 test.
E-2
DRAFT - DO NOT CITE OR QUOTE
-------
Liver tumors (adenoma + carcinoma)
Male BDF1 mouse exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25 ppm
Note: models could not fit data with all 4 dose groups; highest dose group dropped
BMR = 0.1
Multistage; 3-stage model
BMR
(extra risk)
Fisher
AIC
value"
BMC BMC
BMR/
L BMCL
Thrall
AIC val
,
uea BMC BM
BMR/
CL BMCL
MCA
(jimol/L)
0.1
151.192
0.3562
0.191106 | 0.0636
50 1.57
151.158 0.3
560 0.388392 0.12
2027 0.819
MRAMKL
(fimol/hr-kg liver)
0.1
152.089
0.1864
13.3804 | 7.307
35 0.0137
152.924 0.1
386 14.185 8.82
145 0.0113
1 Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the % test.
E-3
DRAFT - DO NOT CITE OR QUOTE
-------
Pheochromocytomas
Female BDF1 mouse exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25, 125 ppm
Multistage; 2-stage model
BMR =10%
BMR
(extra risk)
Fisher
AIC
0.1
71.4077
value"
0.7947
BMC BM
1.42662 1.13
BMR/
CL BMCL
Thrall
AIC value3 BMC BM
MCA
(jimol/L)
753 0.0879
71.3358 0.8039 2.94801 2.34
BMR/
CL BMCL
113 0.0427
a Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the %2 test.
Note: 3-stage model did not provide a sufficiently improved model fit.
E-4
DRAFT - DO NOT CITE OR QUOTE
-------
Pheochromocytomas
Male BDF1 mouse exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25, 125 ppm
Cancer Multistage
BMR =10%
Cancer Multistage (restricted mode) model did not provide an adequate fit of the male
pheochromocytoma data (1, 2, and 3 stage models provided the same outputs); therefore other models in
BMDS were used (see table below).
BMR
(extra risk)
Fisher
AIC
x2/>
value"
BMC
BMCL
BMR/
BMCL
MCA
Thrall
AIC
x2/>
value"
BMC BM
BMR/
CL BMCL
(jimol/L)
1st, 2nd & 3rd
0.1
139.129
0.0513
0.292123
0.230102
0.435
139.077
0.0488
0.600117 0.47
2644 0.212
1 Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the % test.
Pheochromocytomas
Male BDF1 mouse exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25, 125 ppm
Models other than Multistage
BMR = 0.1
Model
Fisher
AIC
xV
value"
BMC
BMCL
BMR/
BMCL
Thrall
AIC
xV
value"
BMC
BMCL
BMR/
BMCL
MCA (^mol/L)
Gammab
Gamma —
unrestricted
Logistic0
Logistic --
unrestricted
Log-logistic0
Log-logistic -
unrestricted
Probit0
Probit -
unrestricted
Log-probit°
Log-probit —
unrestricted
Quantal-linear
Weibullb
Weibull -
unrestricted
139.129
140.755
161.228
161.228
138.661
138.661
159.808
159.808
141.637
137.136
139.129
139.129
140.513
0.0513
0.0401
0.0000
0.0000
0.0978
0.0978
0.0000
0.0000
0.0044
0.1533
0.0513
0.0513
0.0497
0.292124
0.238028
0.929566
0.929566
0.24731
0.247311
0.851235
0.851235
0.423924
0.264859
0.292124
0.292124
0.226525
0.230102
0.10463
0.75614
0.75614
0.147398
0.130943
0.702221
0.702221
0.340228
0.150882
0.230102
0.230102
0.10562
0.435
0.956
0.132
0.132
0.678
0.764
0.142
0.142
0.294
0.663
0.435
0.435
0.947
139.077
140.587
161.353
161.353
138.467
138.467
159.949
159.949
141.988
136.945
139.077
139.077
140.316
0.0488
0.0428
0.0000
0.0000
0.1050
0.1050
0.0000
0.0000
0.0035
0.1648
0.0488
0.0488
0.0535
0.600118
0.473653
1.9184
1.9184
0.492945
0.492945
1.75643
1.75643
0.867906
0.527758
0.60012
0.60012
0.45102
0.472644
0.204957
1.56019
1.56019
0.297393
0.257935
1.44878
1.44878
0.696011
0.297349
0.472644
0.472644
0.207636
0.212
0.488
0.064
0.064
0.336
0.388
0.069
0.069
0.144
0.336
0.212
0.212
0.482
E-5
DRAFT - DO NOT CITE OR QUOTE
-------
Pheochromocytomas
Male BDF1 mouse exposed to carbon tetrachloride vapor for 104 weeks (6 hours/day, 5 days/week)
Doses modeled: 0, 5, 25, 125 ppm
Models other than Multistage
BMR = 0.1
Model
Fisher
AIC
value"
BMC
BMCL
BMR/
BMCL
Thrall
AIC value3 BMC BM
BMR/
CL BMCL
a Values <0.1 fail to meet conventional goodness-of-fit criteria; p value from the %2 test.
b Power restricted to >1.
0 Slope restricted to >1.
E-f
DRAFT - DO NOT CITE OR QUOTE
-------
Female F344 rat — hepatocellular adenomas + carcinomas (0, 5, 25,125 ppm dose groups)
Dose metric: MRAMKL
Vmax = 0.4 mg/hour/kg BW
0.07
Multistage Cancer Model with 0.95 Confidence Level
0.8
0.6
°-4
ro
0.2
Multistage Cancer
Linear extrapolation
10 15
dose
20
25
10:0010/162007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT LIVER\MRAMKL-
VMAX=0.4\FRAT_LIVER_ADCAR_MRAMKL-4.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT
LIVER\MRAMKL-VMAX=0.4\FRAT_LIVER_ADCAR_MRAMKL-4.pit
Tue Oct 16 10:00:27 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage Cancer model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2-beta3*dose^3-beta4*dose^4)]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 5
Total number of specified parameters = 0
Degree of polynomial = 4
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0
Beta(3) = 0
E-7
DRAFT - DO NOT CITE OR QUOTE
-------
Beta(4) = 6.11699e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l) -Beta(2) -Beta(3)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(4)
Beta(4) 1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Beta(3)
Beta(4)
Estimate
0
0
0
0
5.52689e-006
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-29.6946
-30.67
-109.05
63.3399
# Param's
4
1
1
Deviance Test d.f.
1.95065
158.71
P-value
0.5827
<.0001
Goodness of Fit
Dose
0.0000
3.8130
12 .0920
24.3200
Est. Prob.
0.0000
0.0012
0.1114
0.8554
Expected
0.000
0.058
5.572
42.768
Observed
0
0
3
44
Size
50
50
50
50
Scaled
Residual
0.000
-0.242
-1.156
0.495
=1.64
d.f. = 3
P-value = 0.6503
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.05
Extra risk
0.95
9.8151
8.40334
10.5331
Taken together, (8.40334, 10.5331) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.00595002
DRAFT - DO NOT CITE OR QUOTE
-------
Dose metric: MRAMKL
Vmax = 0.65 mg/hour/kg BW
0.07
0.8
0.6
0-4
0.2
0
Multistage Cancer Model with 0.95 Confidence Level
Multistage Cancer
Linear extrapolation
3MDL
0 5
13:0912/142007
10
15 20
dose
25
30
35
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT LIVER\MRAMKL-
VMAX=0.65\FRAT_LIVER_ADCAR_MRAMKL-65.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT
LIVER\MRAMKL-VMAX=0.65\FRAT_LIVER_ADCAR_MRAMKL-65.pit
Tue Oct 16 10:05:09 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage Cancer model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2-beta3*dose^3-beta4*dose^4) ;
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 5
Total number of specified parameters = 0
Degree of polynomial = 4
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(l) = 0
E-9
DRAFT - DO NOT CITE OR QUOTE
-------
Beta(2) = 0
Beta(3) = 0
Beta(4) = 1.23526e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l) -Beta(2) -Beta(3)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(4)
Beta(4) 1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Beta(3)
Beta(4)
Estimate
0
0
0
0
1.13446e-006
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
- Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-29.6946
-30.4171
-109.05
62.8343
# Param's
4
1
1
Deviance Test d.f.
1.44504
158.71
P-value
0.695
<.0001
Goodness of Fit
Dose
0.0000
4 .9910
17.6260
36 .2660
Est. Prob.
0.0000
0.0007
0.1037
0.8595
Expected
0.000
0.035
5.186
42 .974
Observed
0
0
3
44
Size
50
50
50
50
Scaled
Residual
0.000
-0.188
-1.014
0.418
Chi 2 = 1.24
d.f. = 3
P-value = 0.7440
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.05
Extra risk
0.95
14.582
12 .2867
15.6526
Taken together, (12.2867, 15.6526) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.00406945
E-10
DRAFT - DO NOT CITE OR QUOTE
-------
Female F344 rat — hepatocellular adenomas + carcinomas (0, 5, 25 ppm dose groups)
Dose metric: MRAMKL
,0.07
Vmax = 0.4 mg/hour/kg BW
Multistage Cancer Model with 0.95 Confidence Level
g
t3
ro
0.15
0.1
0.05
Multistage Cancer
Linear extrapolation
BMDL
BMP
6
dose
10
12
08:2310/122007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT LIVER\MRAMKL-
VMAX=0.4\FRAT_LIVER_ADCAR_MRAMKL-4.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT
LIVER\MRAMKL-VMAX=0.4\FRAT_LIVER_ADCAR_MRAMKL-4.pit
Fri Oct 12 08:23:17 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2) ]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 3
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0.00044169
E-ll
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(2)
Beta(2) 1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0
0
0.000383811
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-11.3484
-11.6412
-14.7059
25.2825
# Param's
3
1
1
Deviance Test d.f.
0.585705
6.71498
P-value
0.7461
0.03482
Goodness of Fit
Dose
0.0000
3 .8130
12.0920
Est. Prob.
0.0000
0.0056
0.0546
Expected
0.000
0.278
2.729
Observed
0
0
3
Size
50
50
50
Scaled
Residual
0.000
-0.529
0.169
Chi^2 = 0.31
d.f. = 2
P-value = 0.8571
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.05
Extra risk
0.95
11.5604
6.92352
30.5183
Taken together, (6.92352, 30.5183) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.00722176
E-12
DRAFT - DO NOT CITE OR QUOTE
-------
Dose metric: MRAMKL
Vmax = 0.65 mg/hour/kg BW
0.07
Multistage Cancer Model with 0.95 Confidence Level
0.15
1 0.1
<
o
ro
^ 0.05
0
Multistage Cane
Linear extrapolat
;
^ -~~~~~^
024
on
BMDL
— -^
BMC
-
-
D :
6 8 10 12 14 16 18
dose
08:3510/122007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT LIVER\MRAMKL-
VMAX=0.65\FRAT_LIVER_ADCAR_MRAMKL-65.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE RAT
LIVER\MRAMKL-VMAX=0.65\FRAT_LIVER_ADCAR_MRAMKL-65.pit
Fri Oct 12 08:35:44 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2)]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 3
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0.000206402
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
E-13
DRAFT - DO NOT CITE OR QUOTE
-------
Beta(2)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(2)
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0
0
0.000183949
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-11.3484
-11.5867
-14.7059
25.1734
# Param's
3
1
1
Deviance Test d.f.
0.476667
6.71498
P-value
0.7879
0.03482
Goodness of Fit
Dose
0.0000
4 .9910
17.6260
Est. Prob.
0.0000
0.0046
0.0555
Expected
0.000
0.229
2.777
Observed
0
0
3
Size
50
50
50
Scaled
Residual
0.000
-0.479
0.137
Chi^2 = 0.25
d.f. = 2
P-value = 0.8831
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.05
Extra risk
0.95
16.6986
9.76339
43 .9237
Taken together, (9.76339, 43.9237) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.00512117
E-14
DRAFT - DO NOT CITE OR QUOTE
-------
Female BDF1 mouse - hepatocellular adenomas + carcinomas (0, 5, 25 ppm dose groups)
Dose metric: MRAMKL
Fisher model
Multistage Cancer Model with 0.95 Confidence Level
0.8
1 0.4
ro
0.2
Multistage Cancer
Linear extrapolation
BMPL
10
15
20 25
dose
30
35
40
12:0410/152007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE MOUSE
LIVER\MRAMKL-FISHER\FMOUSE_LIVER_ADCAR_MRAMKL-FISHER.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE MOUSE
LIVER\MRAMKL-FISHER\FMOUSE_LIVER_ADCAR_MRAMKL-FISHER.p1t
Fri Oct 12 08:54:44 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2) ]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 3
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0482072
Beta(l) = 0
Beta(2) = 0.00119035
E-15
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background
Beta(2)
Background
1
-0.38
Beta(2)
-0.38
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0.0693295
0
0.00111772
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-55.6537
-55.9559
-99.1295
115.912
# Param's
3
2
1
Deviance Test d.f.
0.604318
86.9516
P-value
0.4369
<.0001
Dose
0.0000
12.6660
41.6750
Est. Prob.
0.0693
0.2221
0.8664
Goodn
Expected
3 .466
10.883
43.321
.ess of Fi
Observed
4
9
44
t
Size
50
49
50
Scaled
Residual
0.297
-0.647
0.282
Chi 2 =0.59
d.f. = 1
P-value = 0.4437
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
Taken together, (6.3204
interval for the BMD
0.1
Extra risk
0.95
9.70893
6.3204
11.2942
, 11.2942) is a 90
% two-sided confidence
Multistage Cancer Slope Factor =
0.0158218
E-16
DRAFT - DO NOT CITE OR QUOTE
-------
Dose metric: MRAMKL
Thrall model
Multistage Model with 0.95 Confidence Level
0.8
o,
0.4
0.2
0 :
Multistage
Bjyip.L,
BMD
10
15
20 25
dose
30
35
40
45
12:1010/152007
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE
MOUSE LIVER\MRAMKL-THRALL\FMOUSE_LIVER_ADCAR_MRAMKL-THRALL.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS
FEMALE MOUSE LIVER\MRAMKL-THRALL\FMOUSE_LIVER_ADCAR_MRAMKL-THRALL.pit
Fri Oct 12 09:01:03 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2)]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 3
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0162478
Beta(l) = 0
Beta(2) = 0.00110173
Asymptotic Correlation Matrix of Parameter Estimates
E-17
DRAFT - DO NOT CITE OR QUOTE
-------
user,
Background
Beta(2)
( *** The model parameter(s) -Beta(l)
have been estimated at a boundary point, or have been specified by the
and do not appear in the correlation matrix )
Background Beta(2)
1 -0.4
-0.4 1
Parameter Estimates
Limit
Variable
Background
Beta(l)
Beta(2)
Estimate
0.0643165
0
0.000963757
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-55.6537
-56.6705
-99.1295
117.341
# Param's
3
2
1
Deviance Test d.f.
2.03362
86.9516
P-value
0.1539
<.0001
Dose
0.0000
15.4560
43 .5990
Est. Prob.
0.0643
0.2567
0.8502
Goodn
Expected
3.216
12.580
42 .510
ess of Fi
Observed
4
9
44
t
Size
50
49
50
Scaled
Residual
0.452
-1.171
0.590
Chi 2 =1.92
d.f. = 1
P-value = 0.1654
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
10.4557
7.59255
12.107
Taken together, (7.59255, 12.107 ) is a 90
interval for the BMD
% two-sided confidence
E-18
DRAFT - DO NOT CITE OR QUOTE
-------
Female BDF1 mouse - hepatocellular adenomas + carcinomas (0, 5 ppm dose groups)
Dose metric: MRAMKL
Fisher model
Multistage Cancer Model with 0.95 Confidence Level
0.35
0.3
0.25
0.2
0.15
ro
0.1
0.05
Multistage Cancer
Linear extrapolation
BMDL
6
dose
10
12
12:4910/152007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE MOUSE
LIVER\MRAMKL-FISHER\FMOUSE_LIVER_ADCAR_MRAMKL-FISHER.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE MOUSE
LIVER\MRAMKL-FISHER\FMOUSE_LIVER_ADCAR_MRAMKL-FISHER.p1t
Fri Oct 12 09:15:17 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage Cancer model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2)]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 2
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.24898
E-19
DRAFT - DO NOT CITE OR QUOTE
-------
Beta(l) =
Beta(2) =
0.0160225
0.001265
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l) Beta(2)
Background 1 -2.2e-008 8.3e-009
Beta(l) -6e-009 1 -1
Beta(2) -3.2e-009 -1 1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0.08
0.00471969
0.000372627
Std. Err.
* - Indicates that this value is not calculated.
Error in computing chi-square; returning 2
Analysis of Deviance Table
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
Log(likelihood) # Param's Deviance Test d.f. P-value
-37.3075 2
-37.3075 3 2.84217e-014
2.38238
-38.4987 1
AIC: 80.6149
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0.0000
12.6660
Chi^2 = 0.00
0.0800
0.1837
4.000
9.000
d.f. = -1
P-value =
50
49
NA
-1
1
NA
0.1227
Scaled
Residual
-0.000
0.000
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 11.6352
BMDL = 5.04631
BMDU did not converge for BMR = 0.100000
BMDU calculation failed
BMDU = 3.56605e+007
Multistage Cancer Slope Factor =
0.0198165
E-20
DRAFT - DO NOT CITE OR QUOTE
-------
Dose metric: MRAMKL
Thrall model
Multistage Model with 0.95 Confidence Level
0.35
0.3
0.25
0.2
0.15
ro
0.1
0.05
Multistage
, ...BMDL
BMD
10
12
14 16
dose
12:50 10/152007
Multistage Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE MOUSE
LIVER\MRAMKL-THRALL\FMOUSE_LIVER_ADCAR_MRAMKL-THRALL.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE MOUSE
LI VER\MRAMKL - THRALL\ FMOUSE_L I VER_ADCAR_MRAMKL - THRALL .pit
Fri Oct 12 09:17:46 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2) ]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 2
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.24898
Beta(l) = 0.0131302
Beta(2) = 0.000849523
E-21
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l) Beta(2)
Background 1 NA NA
Beta(l) NA NA NA
Beta (2) NA NA NA
NA - This parameter's variance has been estimated as zero or less.
THE MODEL HAS PROBABLY NOT CONVERGED!!!
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0.08 * * *
Beta(l) 0.00386773 * * *
Beta(2) 0.000250241 * * *
* - Indicates that this value is not calculated.
At least some variance estimates are negative.
THIS USUALLY MEANS THE MODEL HAS NOT CONVERGED!
Try again from another starting point.
Error in computing chi-square; returning 2
Analysis of Deviance Table
Model Log(likelihood) # Param's Deviance Test d.f. P-value
Full model -37.3075 2
Fitted model -37.3075 3 2.84217e-014 -1 NA
Reduced model -38.4987 1 2.38238 1 0.1227
AIC: 80.6149
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.0800 4.000 4 50
15.4560 0.1837 9.000 9 49
Chi^2 =0.00 d.f. = -1 P-value = NA
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 14.1982
BMDL = 6.15788
BMDU = 2 .64632e + 014
Taken together, (6.15788, 2.64632e+014) is a 90 % two-sided confidence
interval for the BMD
E-22 DRAFT - DO NOT CITE OR QUOTE
-------
Male BDF1 mouse - hepatocellular adenomas + carcinomas (0, 5, 25 ppm)
Dose metric: MRAMKL
Fisher model
Multistage Cancer Model with 0.95 Confidence Level
1
0.9
§ 0.6
••6
,r 0.5
0.4
0.3
0.2
Multistage Cancer
Linear extrapolation
= ............ BMDL ........... BMD
0
10
15
20
dose
25
30
35
40
12:0312/042007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS MALE MOUSE LIVER\MRAMKL-
FISHER\MMOUSE_LIVER_ADCAR_MRAMKL-FISHER.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS MALE MOUSE
LIVER\MRAMKL-FISHER\MMOUSE_LIVER_ADCAR_MRAMKL-FISHER.pit
Tue Dec 04 12:03:25 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage Cancer model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2-beta3*dose^3)]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.352068
Beta(l) = 0
Beta(2) = 0
Beta(3) = 4.77425e-005
E-23
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l) -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(3)
Background
1
-0.22
Beta(3)
-0.22
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Beta(3)
Estimate
0.41973
0
0
4.39818e-005
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-73.1699
-74.0443
-99.6096
152.089
# Param's
3
2
1
Deviance Test d.f.
1.74874
52.8795
P-value
0.186
<.0001
Dose
0.0000
12.6660
41.6750
Est. Prob.
0.4197
0.4693
0.9760
Goodn
Expected
20.987
23.467
48.798
ess of Fi
Observed
24
20
49
t
Size
50
50
50
Scaled
Residual
0.864
-0.982
0.187
Chi 2 =1.75
d.f. = 1
P-value = 0.1864
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
13.3804
7.30705
15.6428
Taken together, (7.30705, 15.6428) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.0136854
E-24
DRAFT - DO NOT CITE OR QUOTE
-------
Dose metric: MRAMKL
Thrall model
Multistage Cancer Model with 0.95 Confidence Level
Affected
g
2
LL
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
. . ' . . ~ ' ' ' ' ' ' ' :
x^ ~
/X/ H
- <> J^^" i
^^--^^^~^~^ \
\
= BMDL B.MD j
0 5 10 15 20 25 30 35 40 45
dose
13:1212/142007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS MALE MOUSE LIVER\MRAMKL-
THRALL\MMOUSE_LIVER_ADCAR_MRAMKL-THRALL.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS MALE MOUSE
LIVER\MRAMKL-THRALL\MMOUSE_LIVER_ADCAR_MRAMKL-THRALL.pit
Tue Dec 04 12:19:57 2007
BMDS MODEL RUN
Observation # < parameter # for Multistage Cancer model.
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2-beta3*dose^3)]
The parameter betas are restricted to be positive
Dependent variable = IncLiverTumor
Independent variable = umol/hr-kgL
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.317881
Beta(l) = 0
Beta(2) = 0
Beta(3) = 4.21166e-005
E-25
DRAFT - DO NOT CITE OR QUOTE
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(l) -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(3)
Background
1
-0.26
Beta(3)
-0.26
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Beta(3)
Estimate
0.410703
0
0
3.69143e-005
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-73.1699
-74.462
-99.6096
152.924
# Param's
3
2
1
Deviance Test d.f.
2.58426
52.8795
P-value
0.1079
<.0001
Dose
0.0000
15.4560
43 .5990
Est. Prob.
0.4107
0.4858
0.9724
Goodn
Expected
20.535
24.289
48.618
ess of Fi
Observed
24
20
49
t
Size
50
50
50
Scaled
Residual
0.996
-1.214
0.330
Chi 2 =2.57
d.f. = 1
P-value = 0.1086
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
14.185
8.82145
16.5171
Taken together, (8.82145, 16.5171) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.011336
E-26
DRAFT - DO NOT CITE OR QUOTE
-------
BDF1 mouse (female) - pheochromocytomas
Dose metric: MCA
Fisher model
Multistage Cancer Model with 0.95 Confidence Level
o
g
0.6
0.5
0.4
0.3
0.2
0.1
0
Multistage Cancer
Linear extrapolation
BMDL! |BMD
0.5
1
1.5
2.5
dose
09:4910/122007
Multistage Cancer Model. (Version: 1.5; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE
PHEOCHROMOCYTOMAS\FISHER\FMOUSE_PHEOCHROMOCYTOMA_MCA-FISHER.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS FEMALE
PHEOCHROMOCYTOMAS\FISHER\FMOUSE_PHEOCHROMOCYTOMA_MCA-FISHER.plt
Fri Oct 12 09:49:11 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2) ]
The parameter betas are restricted to be positive
Dependent variable = Pheochrom
Independent variable = umol/L
Total number of observations = 4
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: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
E-27
DRAFT - DO NOT CITE OR QUOTE
-------
Beta(l) =
Beta(2) =
0.0548062
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(2)
Beta(2) 1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0
0
0.0517683
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-33.7087
-34.7039
-69.0688
71.4077
# Param's
4
1
1
Deviance Test d.f.
1.99041
70.7202
P-value
0.5744
<.0001
Goodness of Fit
Dose
0.0000
0.1110
0.6030
3 .3150
Est. Prob.
0.0000
0.0006
0.0186
0.4338
Expected
0.000
0.031
0.932
21.259
Observed
0
0
0
22
Size
50
49
50
49
Scaled
Residual
0.000
-0.177
-0.975
0.214
=1.03
d.f. = 3
P-value = 0.7947
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
1.42662
1.13753
1.72224
Taken together, (1.13753, 1.72224) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.08791
E-28
DRAFT - DO NOT CITE OR QUOTE
-------
Dose metric: MCA
Thrall model
Multistage Cancer Model with 0.95 Confidence Level
t>
-------
Asymptotic Correlation Matrix of Parameter Estimates
Beta(2)
( *** The model parameter(s) -Background -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(2)
1
Parameter Estimates
Variable
Background
Beta(l)
Beta(2)
Estimate
0
0
0.0121232
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-33.7087
-34.6679
-69.0688
71.3358
# Param's
4
1
1
Deviance Test d.f.
1.91847
70.7202
P-value
0.5895
<.0001
Goodness of Fit
Dose
0.0000
0.2130
1.2260
6.8560
Est. Prob.
0.0000
0.0005
0.0181
0.4344
Expected
0.000
0.027
0.903
21.285
Observed
0
0
0
22
Size
50
49
50
49
Scaled
Residual
0.000
-0.164
-0.959
0.206
Chi^2 = 0.99
d.f. = 3
P-value = 0.8039
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0.95
2 .94801
2.34113
3.55893
Taken together, (2.34113, 3.55893) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0.0427144
E-30
DRAFT - DO NOT CITE OR QUOTE
-------
BDF1 mouse (male) - pheochromocytomas
Dose metric: MCA
Fisher model
5
0.8
0.7
0.6
T3
-------
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept
slope
intercept
1
-0.092
slope
-0.092
1
Variable
background
intercept
slope
Parameter Estimates
Estimate Std. Err.
0 NA
-0.358995 0.125298
0.694404 0.110458
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.604574
0.47791
-0.113416
0.910899
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood)
-64 .0144
-66.5682
-110.216
137.136
# Param's
4
2
1
Deviance Test d.f.
5.10756
92.4032
Goodness of Fit
=3.75
d.f. = 2
P-value = 0.1533
P-value
0.07779
<.0001
Dose
0.0000
0.1110
0.6030
3 .3150
Est. Prob.
0.0000
0.0297
0.2388
0.6820
Expected
0.000
1.484
11.939
34 .099
Observed
0
0
16
32
Size
50
50
50
50
Scaled
Residual
0.000
-1.237
1.347
-0.637
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 0.264859
BMDL = 0.150882
E-32
DRAFT - DO NOT CITE OR QUOTE
-------
Dose metric: MCA
Thrall model
Probit Model with 0.95 Confidence Level
i Affected
o
"o
ro
LL
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
^^-~~~~^ ^
,^" \
// :
°x/ :
/' 1
-ir :
BMDL ,BMD :
01234567
dose
13:1511/302007
Probit Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS MALE
PHEOCHROMOCYTOMAS\THRALL\MMOUSE_PHEOCHROMOCYTOMA_MCA-THRALL.(d)
Gnuplot Plotting File: G:\CARBON TET\BMD\BMD MODELING 10-2007\TUMORS MALE
PHEOCHROMOCYTOMAS\THRALL\MMOUSE_PHEOCHROMOCYTOMA_MCA-THRALL.plt
Fri Nov 30 13:15:12 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+Slope*Log(Dose)) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Pheochrom
Independent variable = umol/L
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial (and Specified) Parameter Values
background = 0
intercept = -0.965049
slope = 0.776315
Asymptotic Correlation Matrix of Parameter Estimates
E-33
DRAFT - DO NOT CITE OR QUOTE
-------
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
intercept
slope
intercept
1
-0.58
slope
-0.58
1
Variable
background
intercept
slope
Parameter Estimates
Estimate Std. Err.
0 NA
-0.844448 0.153761
0.683918 0.109119
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-1.14581
0.470048
-0.543082
0.897787
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood)
-64 .0144
-66.4723
-110.216
136.945
# Param's
4
2
1
Deviance Test d.f.
4.91585
92.4032
P-value
0.08561
<.0001
Goodness of Fit
Dose
0.0000
0.2130
1.2260
6.8560
Est. Prob.
0.0000
0.0286
0.2404
0.6816
Expected
0.000
1.429
12.019
34.080
Observed
0
0
16
32
Size
50
50
50
50
Scaled
Residual
0.000
-1.213
1.318
-0.631
Chi^2 = 3.61
d.f. = 2
P-value = 0.1648
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
HMD = 0.527758
BMDL = 0.297349
E-34
DRAFT - DO NOT CITE OR QUOTE
-------
E.2. A Bayesian Approach to Modeling Pheochromocytoma Incidence in Male Mice
A Bayesian analysis was conducted utilizing the log-probit model in order to: 1) provide
an alternative to modeling the pheochromocytoma incidence data in male mice using the profile
likelihood method implemented in BMDS; and 2) investigate the distribution of the slope
parameter in the log-probit model.
This Bayesian approach was used to generate a probability distribution of risk estimates.
This formal application of Bayesian methods to the evaluation of uncertainty in dose-response
modeling, although conceptually simple, relies on recent computational advances that allow use
of Markov Chain Monte Carlo (MCMC) methods. The analysis here takes advantage of the
computational power of WinBugs 1.4.1, free software (Spiegelhalter et al., 2003) for the
Bayesian analysis of statistical models using MCMC methods (e.g., Brooks, 1998; Gilks et al.,
1998; Chib and Greenberg, 1995; Casella and George, 1992; Smith and Gelfand, 1992).
More specifically, the use of MCMC methods (via WinBugs) to derive a distribution of
BMDs for the multistage model in BMDS has been recently described by Kopylev et al. (2007).
This same methodology can be straightforwardly generalized to derive a distribution of BMDs
for the log-probit model. For this analysis, diffuse (high variance) Gaussian prior distributions
for both the intercept and slope parameters were used, truncated at zero to exclude negative
parameter values. A uniform (0,1) prior was used for the background parameter. The posterior
distributions of parameters and BMDs are based on three Markov chains of 550,000 simulations
each with a burn-in of 50,000 and thinning rate 10 so that 150,000 total simulations were used
for deriving the posterior distributions of the parameters and the BMDs. Standard practices of
MCMC analysis were followed for verifying convergence using multiple chains and for checking
sensitivity to initial values. The mean and 5th percentile of the posterior distribution provide
estimates of the BMD and the BMDL ("lower bound"), respectively.
Using outputs from the Thrall model and MCA as the dose metric, the BMDio and
BMDLio calculated by this analysis were 0.57568 and 0.3177 jimol/L, respectively; these values
are close to the modeling results generated in BMDS for the log-probit model (BMDio = 0.5278
|imol/L and BMDLio = 0.2973 |imol/L), thus confirming the results of the BMDS analysis.
Additionally, Figure E-l shows the posterior distribution of the slope or shape parameter for the
log-probit model generated by the Bayesian analysis. This graph shows that more than 99% of
the posterior distribution for the shape parameter is <1; whereas in BMDS the slope parameter
for the log-probit model is typically constrained to be >1. Clearly, constraining the slope
parameter in this situation leads to misspecifying the statistical model and should be avoided.
E-3 5 DRAFT - DO NOT CITE OR QUOTE
-------
o
o
o
8
CM
O -1
0.4
0.6 0.8
shape parameter
I
1.0
I
1.2
Figure E-l. Histogram of the shape parameter
E-3 6 DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX F. SOURCE CODE FOR PBPK MODELS
0/0
% File: HUMINH.M
% Programmed by Gary Diamond
% Syracuse Research Corporation, 02/2005
%This run time file implements CCL4.CSL for inhalation exposure
%(human parameters)
%Prepare time history variables
prepare @clear @all
%Set communication interval
CINT=1.;
%Set simulation stop (hr)
TSTOP=17250.;
%Integration error check
!!SETWESITG=.F.
%Air Exposure Parameters
AIRC=2.27; %ppm
AIRON=0.; %hr
AIROFF=1000000.; %17520.; %hr
CIOFF=0.; %ppm
APER1=24.; %hr
AWID1=24.; %hr
APER2=168.; %hr
AWID2=168.; %hr
%Oral Exposure Parameters
RGIL=0.0;
FGIL=1.0;
F-1 DRAFT - DO NOT CITE OR QUOTE
-------
%Human Parameters
BW=70.;
VLC=0.04;
VFC=0.30; %Revised from 0.1 (03/2007)
VSC=0.62;
VRC=0.05;
QCC=15.;
QPC=15.;
QSF=0.74;
QLC=0.25;
QFC=0.06;
QSC=0.18;
QRC=0.51;
PBLD=2.64;
PL=3.14;
PF=79.42;
PS=1.0;
PR=3.14;
%Chemical Parameters
MW=153.8;
VMAXC=1.49;
KMX = 0.25;
VMAXSF=0.7;
Al=0.065;
A2=0.095;
A3=0.84;
Kl=0.25;
K2=0.03;
K3=0.025;
K4=0.;
K5=0.0004;
!! START /NC
F-2 DRAFT - DO NOT CITE OR QUOTE
-------
%Output
HUMl=[_time _day _air _ca _ramkb _mcl _cf _af, _cl];
HUM2=rot90(fliplr(HUM 1));
out=fopen('HUMAN. out', V);
fprintf(out,"%f,%f,%f,%e,%e,%e,%e,%e,%e\n",HUM2);
fclose(out);
F-3 DRAFT - DO NOT CITE OR QUOTE
-------
0/0
% File: HUMOR.M
%
% Programmed by Gary Diamond
% Syracuse Research Corporation, 10/2007
%
%This run time file implements CCL4.CSL for "oral" exposure (RGIL)
%(human parameters)
%Prepare time history variables
prepare @clear @all
%Set communication interval
CINT=24.;
%Set simulation stop (hr)
TSTOP=5000.;
%Integration error check
!!SETWESITG=.F.
%Air Exposure Parameters
AIRC=0.; %ppm
AIRON=0.; %hr
AIROFF=1000000.; %17520.; %hr
CIOFF=0.; %ppm
APER1=24.; %hr
AWID1=24.; %hr
APER2=168.; %hr
AWID2=168.; %hr
%Oral Exposure Parameters
RGILC=10.0;
FGIL=1.0;
%Human Parameters
F-4
DRAFT - DO NOT CITE OR QUOTE
-------
BW=70.;
VLC=0.04;
VFC=0.30; %Revised from 0.1
VSC=0.62;
VRC=0.05;
QCC=15.;
QPC=15.;
QSF=0.74;
QLC=0.25;
QFC=0.06;
QSC=0.18;
QRC=0.51;
PBLD=2.64;
PL=3.14;
PF=79.42;
PS=1.0;
PR=3.14;
%Chemical Parameters
MW=153.8;
VMAXC=0.40;
KMX = 0.25;
VMAXSF=0.7;
Al=0.065;
A2=0.095;
A3=0.84;
Kl=0.25;
K2=0.03;
K3=0.025;
K4=0.;
K5=0.0004;
!! START /NC
%Output
HUMl=[_time _rgil _mca _mramkl];
F-5 DRAFT - DO NOT CITE OR QUOTE
-------
HUM2=rot90(fliplr(HUM 1));
out=fopen('HUMAN.out','w');
fprintf(out,"%e,%e,%e,%e\n",HUM2);
fclose(out);
F-6 DRAFT - DO NOT CITE OR QUOTE
-------
0/0
% File: MOUINH_JF.M
% Programmed by Gary Diamond
% Syracuse Research Corporation, 10/2007
%This run time file implements CCL4.CSL for inhlation exposure
%(mouse parameters, Fisher et al. 2004)
%Prepare time history variables
prepare @clear @all
%Set communication interval
CINT=24.;
%Set simulation stop (hr)
TSTOP=17520.;
%Integration error check
!!SETWESITG=.F.
%Air Exposure Parameters:
AIRC=125.; %ppm
AIRON=0.; %hr
AIROFF= 17520; %hr
CIOFF=0.; %ppm
APER1=24.; %hr
AWID1=6.; %hr
APER2=168.; %hr
AWID2=120.; %hr
%Mouse Parameters:
BW=0.036;
VLC=0.04;
VFC=0.04;
VSC=0.69;
VRC=0.14;
F-7 DRAFT - DO NOT CITE OR QUOTE
-------
QCC=30.;
QPC=30.;
QSF=0.75;
QLC=0.24;
QFC=0.05;
QSC=0.17;
QRC=0.54;
PBLD=3.8;
PL=4.8;
PF=91.4;
PS=2.5;
PR=4.8;
%Chemical Parameters:
MW=153.8;
VMAXC=1.;
KMX = 0.3;
VMAXSF=0.75;
%From Thrall et al 2000
Al=0.065;
A2=0.095;
A3=0.84;
Kl=0.25;
K2=0.03;
K3=0.025;
K4=0.;
K5=0.00042;
!! START /NC
%Output
%MOU=[AIR MCA MRAMKB MCL MRAMKL]
MOUl=[_time _day _air _mramkl _mca _mramkb _mcl _mramkl]; %output matrix
%formating of output for printed comma-dilimited file:
MOU2=rot90(fliplr(MOUl));
F-8 DRAFT - DO NOT CITE OR QUOTE
-------
out=fopen('MOU.out',W);
%fprintf(out,"%f,%f,%f,%f,%f,%f,%An",MOU2);
%fprintf(out,"%f,%f,%f,%e,%e,%e,%e,%e\n",MOU2);
fprintf(out, "%e\n",MOU);
fclose(out);
F-9 DRAFT - DO NOT CITE OR QUOTE
-------
0/0
% File: MOUINHJCT.M
% Programmed by Gary Diamond
% Syracuse Research Corporation, 3/2007
%This run time file implements CCL4.CSL for inhlation exposure
%(mouse parameters, Thrall et al. 2000)
%Prepare time history variables
prepare @clear @all
%Set communication interval
CINT=24.;
%Set simulation stop (hr)
TSTOP=17520.;
%Integration error check
!!SETWESITG=.F.
%Air Exposure Parameters:
AIRC=2.5; %ppm
AIRON=0.; %hr
AIROFF= 17520; %hr
CIOFF=0.; %ppm
APER1=24.; %hr
AWID1=6.; %hr
APER2=168.; %hr
AWID2=120.; %hr
%Mouse Parameters:
BW=0.036;
VLC=0.04;
VFC=0.04;
VSC=0.78;
VRC=0.05;
F-10 DRAFT - DO NOT CITE OR QUOTE
-------
QCC=28.;
QPC=28.;
QSF=0.74;
QLC=0.24;
QFC=0.05;
QSC=0.19;
QRC=0.52;
PBLD=7.83;
PL=2.08;
PF=23.0;
PS=0.61;
PR=2.08;
%Chemical Parameters:
MW=153.8;
VMAXC=0.79;
KMX = 0.46;
VMAXSF=0.7;
Al=0.065;
A2=0.095;
A3=0.84;
Kl=0.25;
K2=0.03;
K3=0.025;
K4=0.;
K5=0.00042;
!! START /NC
%Output
%MOU=[AIR MCA MRAMKB MCL MRAMKL]
MOUl=[_time _day _air _mramkl _mca _mcl]; %output matrix
%formating of output for printed comma-dilimited file:
MOU2=rot90(fliplr(MOUl));
out=fopen('MOU.out','w');
F-11 DRAFT - DO NOT CITE OR QUOTE
-------
fprintf(out,"%f,%f,%e,%e,%e,%e\n",MOU2);
fprintf(out, "%e\n",MOU);
fclose(out);
F-12 DRAFT - DO NOT CITE OR QUOTE
-------
0/0
% File: RATINH.M
%
% Programmed by Gary Diamond
% Syracuse Research Corporation, 02/2005
%
%This run time file implements CCL4.CSL for inhalation exposure
%(rat parameters)
%Prepare time history variables
prepare @clear @all
%Set communication interval
CINT=1.;
%Set simulation stop (hr)
TSTOP=17250.;
%Integration error check
!!SETWESITG=.F.
%Air Exposure Parameters:
AIRC=4.; %ppm
AIRON=0.; %hr
AIROFF= 17520; %hr
CIOFF=0.; %ppm
APER1=24.; %hr
AWID1=6.; %6.; %hr
APER2=168.; %hr
AWID2=120.; %120.; %hr
%Rat Parameters:
BW=0.452;
VLC=0.04;
VFC=0.08;
VSC=0.74;
F- 1 3
DRAFT - DO NOT CITE OR QUOTE
-------
VRC=0.05;
QCC=15.;
QPC=15.;
QSF=0.74;
QLC=0.25;
QFC=0.04;
QSC=0.20;
QRC=0.51;
PBLD=4.52;
PL=3.14;
PF=79.42;
PS=1.0;
PR=3.14;
%Chemical Parameters:
MW=153.8;
VMAXC=0.4;
KMX = 0.25;
VMAXSF=0.7;
Al=0.065;
A2=0.095;
A3=0.84;
Kl=0.25;
K2=0.03;
K3=0.025;
K4=0.;
K5=0.00042:
!! START /NC
%Output
RATl=[_time _day _air _mca _mramkl _mcl]; %output matrix
%formating of output for printed comma-dilimited file:
%RAT2=rot90(fliplr(RAT 1));
F-14 DRAFT - DO NOT CITE OR QUOTE
-------
%out=fopen('RAT. out', V);
%fprintf(out,"%f,%f,%f,%e,%e,%e\n",RAT2);
%fclose(out);
F-15 DRAFT - DO NOT CITE OR QUOTE
-------
PROGRAM: CCL4R
IThis program simulates the pharmacokinetics of carbon tetrachloride
!The program is based on ITRICCL4.ACSL, developed by KD THRALL 9/98; ACSL
code provided to GDiamond, 04/2004
!The above code was translated with minor modifications, by GDiamond, 05/2004
INITIAL
VARIABLE TIME = 0.0 ! Set independent variable to be TIME
ALGORITHM IALG = 2 [Numerical integration algorithm - Gear for stiff systems
CINTERVAL CINT=100. ! Communication interval
NSTP = 1000 !Set initital integration cycle length at CINT/1000
MERROR AL=0.0001 ! Set error tolerance for Gear
! *****BODY AND TISSUE MASSES*****
CONSTANT BW = 0.2 IBody weight (kg)
CONSTANT VLC = 0.04 ILiver fraction of body weight
CONSTANT VFC = 0.08 ! Adipose fraction of body weight
CONSTANT VSC = 0.74 ! Slowly-perfused fraction of body weight
CONSTANT VRC = 0.05 IRapdily-perfused fraction of body weight
VL=VLC*BW ILiver (kg
VF=VFC*BW ! Adipose (kg)
VS=VSC*BW !Slowly-perfused (kg)
VR=VRC*BW IRapidly-perfused (kg)
!*****BLOOD FLOWS*****
CONSTANT QCC=14 ! Cardiac output (L/hr-BWASF)
CONSTANT QPC=14 ! Alveolar ventilation (L/hr-BWASF)
CONSTANT QLC = 0.25 ILiver fraction of cardiac output
CONSTANT QFC = 0.09 ! Adipose fraction of cardiac output
CONSTANT QSC = 0.15 ! Slowly-perfused fraction of cardiac output
CONSTANT QRC = 0.51 IRapidly-perfused fraction of cardiac output
CONSTANT QSF = 0.74 !QC and QP scaling factor (SF)
QC=QCC*BW**QSF!Cardiac output (L/hr)
F-16 DRAFT - DO NOT CITE OR QUOTE
-------
QP=QPC*BW**QSF ! Alveolar ventilation (L/hr)
QL = QLC*QC ILiver (L/hr)
QF = QFC*QC ! Adipose (L/hr)
QS = QSC*QC ! Slowly-perfused (L/hr)
QR = QRC*QC IRapidly-perfused (L/hr)
!*****PARTITION COEFFICIENTS*****
CONSTANT PBLD = 4.52 !Blood:air partition coefficient
CONSTANT PL = 3.14 ILiverblood partition coefficient
CONSTANT PF = 79.42 ! Adipose:blood partition coefficient
CONSTANT PS = 1.0 ! Slowly-perfused:blood partition coefficient
CONSTANT PR = 3.14 !Rapidly-perfused:blood partition coefficient
! * * * * *METABOLISM and EXCRETION* * * * *
CONSTANT MW=153.8 IMolecular weight of CC14
CONSTANT VMAXC = 0.40 ! VMAX for metabolism in liver (mg/hr-BWASF)
CONSTANT KMX = 0.25 !KM for metabolism in liver (mg/L)
CONSTANT VMAXSF=0.7 ! Scaling factor for VMAXC (SF)
CONSTANT Al=0.085 IFraction of metabolism rate to Ml pool
CONSTANT A2=0.095 IFraction of metabolism rate to M2 pool
CONSTANT A3=0.84 IFraction of metabolism rate to M3 pool
CONSTANT Kl=0.123 IRate constant for conversion of Ml to exhaled metabolite
(CO2) (hr-1)
CONSTANT K2=0.03 IRate constant for conversion of M2 to urinary metabolite (hr-1)
CONSTANT K3=0.0252 IRate constant for conversion of M3 to fecal metabolite (hr-1)
CONSTANT K4=0. IRate constant for conversion of M2 to Ml (hr-1)
CONSTANT K5=0.00042 IRate constant for conversion of M3 to Ml (hr-1)
VMAX = 1000*VMAXC*BW**VMAXSF/MW IMaximum rate of metabolism in
liver (umol/hr)
KM = 1000*KMX/MW IMichaelis constant for metabolism in liver (umol/L)
|*****EXPOSURE - AIR*****
CONSTANT AIRC = 1. I Air exposure concentration (ppm)
AIR = AIRC; I Air exposure concentration (ppm)
AIRCM = AIR/24.45 I Air exposure (umol/L)
mgAIR = AIR*MW/24.45 I Air exposure (mg/m3)
F-17 DRAFT - DO NOT CITE OR QUOTE
-------
ugAIR=AIR*MW/24.45 ! Air exposure concentration (ug/L)
CONSTANT TSTOP = 700. ILength of simulaion (hr)
CONSTANT AIRON=0. ITime air exposure starts (hr)
CONSTANT AIROFF=700. ITime air exposure stops (hr)
CONSTANT CIOFF=0. ! Concentration in inhaled air when exposure is off (ppm)
CONSTANT APER1=24. IPulse period 1 for air exposure (e.g., hours in a day)
CONSTANT AWID1=24. IPulse width 1 for air exposure (e.g. 6 hours each day)
CONSTANT APER2=168. IPulse period 2 for air exposure (e.g., hours in a day)
CONSTANT AWID2=168. IPulse width 2 for air exposure (e.g. 6 hours each day)
|*****EXPOSURE - ORAL ******GD 08/2007
CONSTANT RGILC=0. IRate of uptake from GI to liver (umol/hr)
RGIL=RGILC IRate of uptake from GI to liver (umol/hr)
!Use for simuulating
constant rate of uptake from Gl-tract
MGRGIL=RGIL*MW/1000 IRate of uptake from GI to liver (mg/hr)
MGRGILKGD=RGIL*24*MW/(1000*BW) IRate of uptake to liver (mg/kg-day)
CONSTANT GILF=1. I Absorption fraction
I CONSTANT POINTS = 96.
ICINT = TSTOP/POINTS I Sets communication for 96 times in the simulation
END I of INITIAL section of program
DYNAMIC
DERIVATIVE
DAY=TIME/24
YEAR=DAY/365
!*****CONCENTRATION IN INHALED AIR (umol/L)*****
CION = AIRCM*PULSE(AIRON,APER1,AWID1)*PULSE(AIRON,APER2,AWID2)
CI = RSW(TIME.LE. AIROFF,CION,CIOFF)
RAI = QP*(CA/PBLD-CI) IRate inhaled (umol/hr)
CP = CP24.45 I Concentration in chamber (ppm)
I * * * * * AMOUNT TAKEN IN BY ONE ANIMAL (umol)* * * * *
F-l 8 DRAFT - DO NOT CITE OR QUOTE
-------
RIN = QP*CI
AIN = INTEG(RIN,0.0)
I * * * * *LIVER* * * * *
!Use for simulation of constant rate of uptake from Gl-tract (GD 08/2007)
RAL = QL*(CA-CVL)+RGIL-RAM IRate of change in amount (umol/hr)
!Use for simulation of inhalation exposure
!RAL = QL*(CA-CVL)-RAM IRate of change in amount (umol/hr)
AL = INTEG(RAL, 0.0) ! Amount (umol)
CVL = CL/PL ! Concentration in venous blood (umol/L)
CL = AL/VL ! Concentration (umol/L)
AUCCL = INTEG(CL,0.) ! AUC concentration (umol/L x hr)
! Average concentration in liver (umol/L) - MCL(GD 05/2004)
IF (TIME .GT. 0.) THEN
MCL = AUCCL/TIME
ELSE
MCL= 0.
END IF
ugCL = CL*MW/1000 ! Concentration (ug/g)
MugCL = MCL*MW/1000 ! Average concentration (ug/g)
AUCugCL = AUCCL*MW/1000 ! AUC concentration (ug/g x hr)
I * * * * *METABOLIZED * * * * *
RAM = (VMAX*CVL)/(KM+CVL) IRate, total (umol/hr)
RAMKB= RAM/BW IRate, total (umol/hr x kq body)
RAMKL=RAM/VL IRate, total (umol/hr x kg liver)
AUCRAM=INTEG(RAM,0.0) I AUC rate (umol/hr x hr)
I Average rate of metabolism (umol/hr) - MRAM (GD 05/2004)
IF (TIME .GT. 0.) THEN
MRAM = AUCRAM/TIME
ELSE
MRAM=0.
END IF
F-19 DRAFT - DO NOT CITE OR QUOTE
-------
MRAMKB=MRAM/BW ! Average rate, total (umol/hr x kg body) (GD 05/2004)
MRAMKL=MRAM/VL ! Average rate, total (umol/hr x kg liver) (GD 05/2004
RAXA = RAM*A1-K1*AXA+K4*AXU+K5*AXF IRate to air pool (umol/hr)
RAXF = RAM*A3-K3*AXF-K5*AXF IRate to feces pool (umol/hr)
RAXU = RAM*A2-K2*AXU-K4*AXU IRate to urine pool(umol/hr)
AM = INTEG(RAM,0.) I Amount total (umol)
AMK = AM/BW I Amount total (umol/kg bw)
AMKL=AM/VL I Amount total (umol/kg liver)
AXA = INTEG(RAXA,0.) I Amount in air pool (umol)
AXF = INTEG(RAXF,0.) I Amount in feces pool (umol)
AXU = INTEG(RAXU,0.) I Amount in urine pool (umol)
RA = AXA*kl I Rate to air (umol/hr)
RU = AXU*k2 I Rate to urine (umol/hr)
RF = AXF*k3 IRate to feces (umol/hr)
ugRA = RA*MW IRate to air (ug/hr)
ugRU = RU*MW IRate to urine (ug/hr)
ugRF = RF*MW IRate to feces (ug/hr)
CAX = RA/((3/2)*QP) I Concentration of metabolite in exhaled air (umol/L)
CAXM = CAX*24.45 I Concentration of metabolite in exhaled air (ppm)
TJ A T^
RAF = QF*(CA-CVF) IRate of change in amount (umol/hr)
AF = INTEG(RAF, 0.0) I Amount (umol)
CVF = CF/PF I Concentration in venous blood (umol/L)
CF = AF/VF I Concentration (umol/L)
AUCCF = INTEG(CF, 0.) I AUC concentration (umol/L x hr)
I Average concentration in fat (umol/L) - MCF (GD 03/2007)
IF (TIME .GT. 0.) THEN
MCF = AUCCF/TIME
ELSE
MCF=0.
END IF
ugCF = CF*MW/1000 !Concentration(ug/g)
MugCF = MCF*MW/1 000 I Average concentration (ug/g)
F-20 DRAFT - DO NOT CITE OR QUOTE
-------
AUCugCF = AUCCF*MW/1000 ! AUC concentration (ug/g x hr)
!*****SLOWLY PERFUSED TISSUES*****
RAS = QS*(CA-CVS) IRate of change in amount (umol/hr)
AS = INTEG(RAS, 0.0) ! Amount (umol)
CVS = CS/PS !Concentration in venous blood (umol/L)
CS = AS/VS ! Concentration in umol/L
ugCS = CS*MW/1000 !Concentration in ug/g
!*****RICHLY PERFUSED TISSUES*****
RAR = QR*(CA-CVR) IRate of change in amount (umol/hr)
AR = INTEG(RAR,0.0) ! Amount (umol)
CVR = CR/PR ! Concentration in venous blood (umol/L)
CR = AR/VR ! Concentration in umol/L
ugCR = CR*MW/1000 ! Concentration in ug/g
!*****MIXED VENOUS BLOOD*****
CV = (QF*CVF+QL*CVL+QS*CVS+QR*CVR)/QC !Concentration (umol/L)
AUCCV=INTEG(CV,0.) ! AUC concentration (umol/L x hr)
! Average concentration in venous blood (umol/L) - MCV (GD 05/2004)
IF (TIME .GT. 0.) THEN
MCV = AUCCV/TIME
ELSE
MCV=0.
END IF
ugCV = CV*MW/1000 ! Concentration (ug/g)
MugCV = MCV*MW/1000 ! Average concentration (ug/g)
AUCugCV = AUCCV*MW/1000 ! AUC concentration (ug/g x hr)
I * * * * * ARTERIAL BLOOD* * * * *
CA = (QC*CV+QP*CI)/(QC+(QP/PBLD)) ! Concentration(umol/L)
AUCCA=INTEG(CA,0.) ! AUC concentration (umol/L x hr)
! Average concentration in atrerial blood (umol/L) - MCA (GD 05/2004)
F-21 DRAFT - DO NOT CITE OR QUOTE
-------
IF (TIME .GT. 0.) THEN
MCA = AUCCA/TIME
ELSE
MCA=0.
END IF
ugCA=CA*MW/1000 ! Concentration (ug/g)
MugCA=MCA*MW/1000 ! Average concentration (ug/g)
AUCugCA=AUCCA*MW/1000 ! AUC concentration (ug/g x hr)
I * * * * * AMOUNT EXHALED* * * * *
CX=CA/PBLD ! Concentration in alveolar air (umol/L)
CXPPM = (0.7*CX+0.3*CI)*24.45 !Concentration in exhaled air (ppm)
!Total ventilation is
0.7 of alveolar ventilation
RAX= QP*CX !Rae of change in amount (umol/hr)
AX=INTEG(RAX,0.) ! Amount (umol)
!*****NET AMOUNT ABOSORBED*****
DOSEX = AIN-AX !Net amount absorbed(umol)
BODY=AL+AF+AS+AR ! Amount in body (umol)
MAS SB = BODY+AM+AX IMass balance (umol)
END ! of DERIVATIVE section of program
TERMT(TEVIE .GE. TSTOP) ! Termination condition
END ! of DYNAMIC section of program
TERMINAL
END ! of TEMINAL section of program
END ! of program
F-22 DRAFT - DO NOT CITE OR QUOTE
------- |