&EPA
EPA/635/R-09/010F
www.epa.gov/iris
TOXICOLOGICAL REVIEW
OF
CHLOROPRENE
(CAS No. 126-99-8)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
September 2010
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection Agency
policy and approved for publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
TABLE OF CONTENTS
LIST OF TABLES v
LIST OF FIGURES ix
LIST OF ABBREVIATIONS AND ACRONYMS xi
FOREWORD xiii
AUTHORS, CONTRIBUTORS, AND REVIEWERS xiv
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION 3
3. TOXICOKINETICS 6
3.1. ABSORPTION 6
3.2. DISTRIBUTION 6
3.3. METABOLISM 7
3.4. ELIMINATION 18
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS 18
4. HAZARD IDENTIFICATION 22
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL CONTROLS .. 22
4.1.1. Chloroprene Exposure and Cancer Effects 22
4.1.1.1. Overview 22
4.1.1.2. Individual Occupational Studies 22
4.1.1.3. Summary and Discussion of Relevant Methodological Issues 38
4.1.1.3.1. Lung Cancer Summary 40
4.1.1.3.2. Liver Cancer Summary 41
4.1.2. Chloroprene Exposure and Noncancer Effects 42
4.1.2.1. Acute-, Short-, and Sub chronic-Duration Noncancer Effects 42
4.1.2.2. Chronic Noncancer Effects 44
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN ANIMALS-
ORAL AND INHALATION 46
4.2.1. Oral Exposure 46
4.2.2. Inhalation Exposure 48
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND INHALATION 73
4.4. OTHER DURATION-OR ENDPOINT-SPECIFIC STUDIES 79
4.4.1. Acute and Subchronic Studies 79
4.4.2. Immunotoxicity 81
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF MODE OF ACTION... 82
4.5.1. Mode-of-Action Studies 82
4.5.2. Genotoxicity Studies 85
4.5.2.1. Bacterial Mutagenicity Assays 87
4.5.2.2. Mammalian Cell Assays 88
4.5.2.3. In Vivo Bioassays 89
4.5.3. Structural Alerts 90
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS 92
4.6.1. Human Studies 92
4.6.2. Animal Studies 92
4.6.2.1. Oral Exposure 92
4.6.2.2. Inhalation Exposure 93
-------
4.7. EVALUATION OF CARCINOGENICITY 96
4.7.1. Synthesis of Human, Animal, and Other Supporting Evidence 97
4.7.1.1. Human 97
4.7.1.1.1. Evidence for Causality 100
4.7.1.2. Laboratory Animal 102
4.7.2. Summary of Overall Weight of Evidence 103
4.7.3. Mode-of-Action Information 106
4.7.3.1. Hypothesized Mode of Action 106
4.7.3.2. Experimental Support for the Hypothesized Mode of Action 106
4.7.3.3. Conclusions about the Hypothesized Mode of Action 109
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES Ill
4.8.1. Possible Childhood Susceptibility Ill
4.8.2. Possible Sex Differences 112
5. DOSE-RESPONSE ASSESSMENTS 113
5. LORAL REFERENCE DOSE (RfD) 113
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 113
5.2.1. Choice of Principal Study and Critical Effect(s) 113
5.2.2. Methods of Analysis 116
5.2.3. Exposure Duration and Dosimetric Adjustments 118
5.2.4. RfC Derivation-Including Application of Uncertainty Factors 122
5.2.5. Previous RfC Assessment 123
5.2.6. RfC Comparison Information 123
5.3. UNCERTAINTIES IN THE INHALATION REFERENCE CONCENTRATION 124
5.4. CANCER ASSESSMENT 128
5.4.1. Choice of Study/Data-with Rationale and Justification 128
5.4.2. Dose-Response Data 128
5.4.3. Dose Adjustments and Extrapolation Methods 130
5.4.4. Oral Slope Factor and Inhalation Unit Risk 132
5.4.5. Application of Age-Dependent Adjustment Factors 137
5.4.6. Previous Cancer Assessment 138
5.4.7. Uncertainties in Cancer Risk Values 138
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF HAZARD AND DOSE RESPONSE 143
6.1. HUMAN HAZARD POTENTIAL 143
6.2. DOSE RESPONSE 146
6.2.1. Noncancer/Oral 146
6.2.2. Noncancer/Inhalation 146
6.2.3. Cancer/Oral 147
6.2.4. Cancer/Inhalation 147
7. REFERENCES 149
7.1. REFERENCES ADDED AFTER EXTERNAL PEER RE VIEW 157
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS AND
DISPOSITION A-l
APPENDIX B. BENCHMARK DOSE MODELING RESULTS FOR THE DERIVATION OF THE
RFC B-l
APPENDIX C. CANCER DOSE-RESPONSE MODELING C-l
IV
-------
LIST OF TABLES
Table 2-1. Physical properties and chemical identity of chloroprene 5
Table 3-1. Tissue-to-air partition coefficients for chloroprene 7
Table 3-2. Liver microsomal metabolites as a percentage of 1-butanol internal standarda 9
Table 3-3. Stereochemical comparison of relative amounts (percentages) of R- and S-
enantiomers of the major chloroprene metabolite (l-chloroethenyl)oxirane from
liver microsomes compared across species, strains, gender, and chloroprene
concentration (mM) 11
Table 3-4. Kinetic parameters used to describe the microsomal oxidation of chloroprene 14
Table 3-5. Kinetic parameters used to describe the microsomal epoxide hydrolase activity of
(l-chloroethenyl)oxirane 14
Table 3-6. Kinetic parameters used to describe the time course of (l-chloroethenyl)oxirane
formation from microsomal oxidation of chloroprene 16
Table 3-7. Kinetic parameters used to describe the cytosolic glutathione S-transferase activity
towards (l-chloroethenyl)oxirane 17
Table 3-8. Metabolic parameters of chloroprene 18
Table 3-9. Physiological parameters used for chloroprene PBPK modeling 19
Table 4-1. Standardized mortality ratios (SMRs) for the DuPont Louisville Works cohort
relative to general U.S. population rates 24
Table 4-2. Standardized mortality ratios (SMRs) for all cancers, liver and lung cancer among
males exposed to chloroprene relative to general Chinese population rates 26
Table 4-3. Standardized mortality ratios (SMRs) for selected cancer risks relative to general
population rates of Moscow, Russia 28
Table 4-4. Selected relative risk (RR) estimates for the high-exposure group relative to
unexposed factory workers 28
Table 4-5. Internal relative risks (RRs) by duration of employment in the high-exposure
category 29
Table 4-6. Selected standardized incidence ratios (SIRs) for chloroprene monomer cohort
relative to the general Armenian population 30
Table 4-7. Standardized incidence ratios (SIRs) for elevated cancer risks for plant workers
relative to general population rates of Isere, France 32
Table 4-8. Standardized mortality ratios (SMRs) at each of four chloroprene production
facilities 35
Table 4-9. Relative risks (RRs) for respiratory cancers by cumulative chloroprene exposure 37
Table 4-10. Epidemiologic summary results of respiratory system cancers: Standardized
mortality ratios (SMRs) and standardized incidence ratios (SIRs) for the overall
cohort populations relative to external comparison populations and relative risks
(RRs) for intermediate and high chloroprene exposures 39
-------
Table 4-11. Epidemiologic summary results of liver/biliary passage cancers: Standardized
mortality ratios (SMRs) for the overall cohort populations relative to external
comparison populations and SMRs and relative risks (RRs) for intermediate and
high chloroprene exposures 40
Table 4-12. Frequency of chromosomal aberrations in leukocyte cultures from blood cells of
chloroprene production workers 46
Table 4-13. Tumor incidence in female BD-IV rats treated orally with chloroprene
(100 mg/kg) on GDI? and in their progeny treated (50 mg/kg) weekly for life
(120 weeks) 47
Table 4-14. Distribution of tumors in female BD-IV rats treated orally with chloroprene
(100 mg/kg) on GDI? and their progeny treated (50 mg/kg) weekly for life
(120 weeks) 48
Table 4-15. Survival and body weights of rats in the 16-day inhalation study of chloroprene 49
Table 4-16. Incidences of selected nonneoplastic lesions in rats in the 16-day inhalation study
of chloroprene 50
Table 4-17. Survival and body weights of mice in the 16-day inhalation study of chloroprene 51
Table 4-18. Survival and body weights of rats in the 13-week inhalation study of chloroprene 52
Table 4-19. Incidences of selected nonneoplastic lesions in rats in the 13-week inhalation study
of chloroprene 54
Table 4-20. Survival and body weights of mice in the 13-week inhalation study of chloroprene 55
Table 4-21. Incidences of forestomach lesions in mice in the 13-week inhalation study of
chloroprene 56
Table 4-22. Two-year survival probability estimates for F344/N rats chronically exposed
(2 years) to chloroprene by inhalation 57
Table 4-23. Incidence and severity of nonneoplastic lesions in F344/N rats chronically exposed
(2 years) to chloroprene by inhalation 58
Table 4-24. Incidence of neoplasms in F344/N rats chronically exposed (2 years) to
chloroprene by inhalation 59
Table 4-25. 2-year survival probabilities for B6C3Fi mice chronically exposed (2 years) to
chloroprene by inhalation 61
Table 4-26. Incidence and severity of nonneoplastic lesions in B6C3Fi mice chronically
exposed (2 years) to chloroprene by inhalation 63
Table 4-27. Incidence of neoplasms in B6C3Fi mice chronically exposed (2 years) to
chloroprene by inhalation 64
Table 4-28. Survival-adjusted neoplasm rates for mice in the 2-year inhalation study of
chloroprene 66
Table 4-29. Selected mean relative organ weights of rats exposed for 24 months and hamsters
exposed for 18 months to chloroprene vapor 68
Table 4-30. Incidence, site and type of tumor in selected organs and tissues of rats exposed to
chloroprene for 24 months 70
vi
-------
Table 4-31. Incidence, site and type of tumor in selected organs and tissues of hamsters
exposed to chloroprene for 18 months 71
Table 4-32. Summary of epididymal, spermatozoal and estrous cycle parameters for rats and
mice in the 13-week study of chloroprene 74
Table 4-33. Results of teratology and embryotoxicity studies in rats exposed to chloroprene by
inhalation 76
Table 4-34. Incidence of anomalies in litters of rats exposed to chloroprene by inhalation 77
Table 4-35. Chloroprene-induced mortality in male rats 80
Table 4-36. Genotoxicity assays of chloroprene 86
Table 4-37. Sites of increased incidences of neoplasms in the 2 year inhalation studies of 1,3-
butadiene, isoprene, and chloroprene in rats and mice 91
Table 4-38 Quantitative comparison of carcinogenic potency of butadiene and chloroprene in
mice 92
Table 4-39. Summary of animal and human tumor data and weight of evidence descriptor for
chloroprene 105
Table 5-1. Incidences of nonneoplastic lesions resulting from chronic exposure (ppm) to
chloroprene considered for identification of critical effect 116
Table 5-2. Duration adjusted point of departure estimates for best fitting models of the BMD
from chronic exposure to chloroprene 119
Table 5-3. Summary of Uncertainties in the Chloroprene noncancer risk assessment 127
Table 5-4. Tumor incidence in female and male B6C3Fi mice exposed to chloroprene via
inhalation for 2 years 129
Table 5-5. Tumor incidence in female and male F344 rats exposed to chloroprene via
inhalation for 2 years 130
Table 5-6. Dose-response modeling summary for female mouse tumors associated with
inhalation exposure to chloroprene for 2 years 134
Table 5-7. Dose-response modeling summary for male mouse tumor sites associated with
inhalation exposure to chloroprene for 2 years 135
Table 5-8. Summary of uncertainties in chloroprene cancer unit risk estimate 139
Table B-l. Severity scores at control dose and lowest dose showing response for endpoints
considered for critical noncancer effect B-3
Table B-2. Benchmark modeling results for alveolar epithelial hyperplasia in male F344/N rats
(BMR= 10% extra risk) B-4
Table B-3. Benchmark modeling results for alveolar epithelial hyperplasia in male F344/N rats
(BMR = 5% extra risk) B-6
Table B-4. Benchmark modeling results for alveolar epithelial hyperplasia in female F344/N
rats (BMR = 10% extra risk) B-8
Table B-5. Benchmark modeling results for bronchiolar hyperplasia in male B6C3Fi mice
(BMR= 10% extra risk) B-10
VII
-------
Table B-6. Benchmark modeling results for bronchiolar hyperplasia in male B6C3Fi mice
(BMR = 5% extra risk) B-12
Table B-7. Benchmark modeling results for olfactory chronic inflammation in male F344/N
rats (BMR = 10% extra risk) B-14
Table B-8. Benchmark modeling results for olfactory atrophy in male F344/N rats (BMR =
10% extra risk) B-16
Table B-9. Benchmark modeling results for olfactory atrophy in male F344/N rats (BMR = 5%
extra risk) B-18
Table B-10. Benchmark modeling results for olfactory necrosis in male F344/N rats (BMR =
10% extra risk) B-20
Table B-l 1. Benchmark modeling results for olfactory necrosis in male F344/N rats (BMR =
5% extra risk) B-22
Table B-12. Benchmark modeling results for olfactory necrosis in female F344/N rats (BMR =
10% extra risk) B-24
Table B-13. Benchmark modeling results for olfactory necrosis in female F344/N rats (BMR =
5% extra risk) B-28
Table B-15. Benchmark modeling results for olfactory basal cell hyperplasia in female F344/N
rats (BMR = 5% extra risk) B-34
Table B-16. Benchmark modeling results for kidney (renal tubule) hyperplasia in male F344/N
rats (BMR = 10% extra risk) B-36
Table B-l7. Benchmark modeling results for kidney (renal tubule) hyperplasia in female
F344/N rats (BMR = 10% extra risk) B-38
Table B-18. Benchmark modeling results for forestomach epithelial hyperplasia in male
B6C3Fi mice (BMR = 10% extra risk) B-40
Table B-19. Benchmark modeling results for forestomach epithelial hyperplasia in female
B6C3Fi mice (BMR = 10% extra risk) B-42
Table B-20. Benchmark modeling results for splenic hematopoietic cell proliferation in female
B6C3Fi mice (BMR = 10% extra risk) B-44
Table B-21. Benchmark modeling results for splenic hematopoietic cell proliferation in female
B6C3Fi mice (BMR = 5% extra risk) B-46
Table C-l. Tumor incidence, with time to death with tumor: female mice exposed to
chloroprene via inhalation C-l
Table C-2. Tumor incidence, with time to death with tumor: male mice exposed to
chloroprene via inhalation C-4
Table C-3. Summary of model selection and modeling results for best-fitting multistage-
Weibull models, using time-to-tumor data for female mice C-7
Table C-4. Summary of model selection and modeling results for best-fitting multistage-
Weibull models, using time-to-tumor data for male mice C-8
Table C-5. Summary of human equivalent composite cancer risk values estimated by
0.01/BMDoi, based on male and female mouse tumor incidence C-39
viii
-------
LIST OF FIGURES
Figure 2-1. The chemical structure of chloroprene 4
Figure 3-1. Proposed metabolism of chloroprene 10
Figure 5-1. Points of departure (in mg/m3) for selected endpoints with corresponding applied
uncertainty factors and derived sample RfCs (chosen co-critical effects are circled) 124
Figure B-l. Log-logistic model fit for alveolar epithelial hyperplasia in male F344/N rats
(BMR= 10% extra risk) B-4
Figure B-2. Log-logistic model fit for alveolar epithelial hyperplasia in male F344/N rats
(BMR = 5% extra risk) B-6
Figure B-3. Log-logistic model fit for alveolar epithelial hyperplasia in female F344/N rats
(BMR= 10% extra risk) B-8
Figure B-4. Log-logistic model fit for bronchiolar hyperplasia in male B6C3Fi mice (BMR =
10% extra risk) B-10
Figure B-5. Log-logistic model fit for bronchiolar hyperplasia in male B6C3Fi mice (BMR =
5% extra risk) B-12
Figure B-6. Log-logistic model fit for olfactory chronic inflammation in male F344/N rats
(BMR= 10% extra risk) B-14
Figure B-7. Logistic model fit for olfactory atrophy in male F344/N rats (BMR = 10% extra
risk) B-16
Figure B-8. Logistic model fit for olfactory atrophy in male F344/N rats (BMR = 5% extra
risk) B-18
Figure B-9. Log-probit model fit for olfactory necrosis in male F344/N rats (BMR = 10% extra
risk) B-20
Figure B-10. Log-probit model fit for olfactory necrosis in male F344/N rats (BMR = 5% extra
risk) B-22
Figure B-l 1. Log-probit model fit for olfactory necrosis in female F344/N rats (BMR = 10%
extra risk) B-24
Figure B-12. Dichotomous Hill model fit for olfactory necrosis in female F344/N rats (BMR =
10% extra risk) B-26
Figure B-13. Log-probit model fit for olfactory necrosis in female F344/N rats (BMR = 5%
extra risk) B-28
Figure B-14. Log-probit model fit for olfactory basal cell hyperplasia in female F344/N rats
(BMR= 10% extra risk) B-30
Figure B-15. Dichotomous Hill model fit for olfactory basal cell hyperplasia in female F344/N
rats(BMR= 10% extra risk) B-32
Figure B-16. Log-probit model fit for olfactory basal cell hyperplasia in female F344/N rats
(BMR = 5% extra risk) B-34
Figure B-17. Log-logistic model fit for kidney (renal tubule) hyperplasia in male F344/N rats
(BMR= 10% extra risk) B-36
ix
-------
Figure B-18. Log-probit model fit for kidney (renal tubule) hyperplasia in female F344/N rats
(BMR= 10% extra risk) B-38
Figure B-19. Multistage model fit for forestomach epithelial hyperplasia in male B6C3Fi mice
(BMR= 10% extra risk) B-40
Figure B-20. Multistage model fit for forestomach epithelial hyperplasia in female B6C3Fi
mice(BMR= 10% extra risk) B-42
Figure B-21. Probit model fit for splenic hematopoietic cell proliferation in female B6C3Fi
mice(BMR= 10% extra risk) B-44
Figure B-22. Probit model fit for splenic hematopoietic cell proliferation in female B6C3Fi
mice (BMR = 5% extra risk) B-46
Figure C-l. Female mice, alveolar/bronchiolar tumors C-9
Figure C-2. Female mice, hemangiomas and hemangiosarcomas in all organs; high dose
dropped, hemangiosarcomas occurring before termination considered fatal C-l 1
Figure C-3. Female mice, hemangiomas and hemangiosarcomas in all organs; high dose
dropped, all tumors considered incidental C-13
Figure C-4. Female mice, Harderian gland tumors C-15
Figure C-5. Female mice, mammary gland tumors C-17
Figure C-6. Female mice, forestomach tumors C-19
Figure C-7. Female mice, hepatocellular adenomas and carcinomas C-21
Figure C-8. Female mice, skin sarcomas C-23
Figure C-9. Female mice, Zymbal's gland tumors C-25
Figure C-10. Male mice, alveolar/bronchiolar tumors C-27
Figure C-l 1. Male mice, hemangiomas and hemangiosarcomas; hemangiosarcomas occurring
before termination considered fatal C-29
Figure C-12. Male mice, hemangiomas and hemangiosarcomas; all tumors considered
incidental C-31
Figure C-13. Male mice, Harderian gland tumors C-33
Figure C-14. Male mice, renal tubule tumors C-35
Figure C-15. Male mice, forestomach tumors C-37
-------
LIST OF ABBREVIATIONS AND ACRONYMS
ACGIH American Conference of Governmental Industrial Hygienists
ADAF age dependent adjustment factor
AEH alveolar epithelial hyperplasia
AIC Akaike Information Criterion
ALP alkaline phosphatase
ALT alanine aminotransferase
BMC benchmark concentration
BMCL lower bound on the benchmark concentration
BMD benchmark dose
BMDL lower confidence limit on the benchmark dose
BMDS benchmark dose software
BMR benchmark response
CASRN Chemical Abstracts Service Registry Number
CI confidence interval
CNS central nervous system
CYP cytochrome
DAF dosimetric adjustment factor
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid
EDio effective dose associated with 10% excess risk
EH epoxide hydrolases
EPA U.S. Environmental Protection Agency
eV electron volt
GD gestational day
GDH glutamine dehydrogenase
GSH glutathione
GST glutathione S-transferase
HEC human equivalent concentration
IARC International Agency for Research on Cancer
ICD International Classification of Diseases
IPCS International Programme on Chemical Safety
IRIS Integrated Risk Information System
kf non-enzymatic first order glutathione reaction rate
KOW octanol-water partition coefficient
LOAEL lowest-observed-adverse-effect level
LOH loss of heterozygosity
M Molar
MLE maximum likelihood estimate
MOA mode of action
MV minute volume
NCEA National Center for Environmental Assessment
NIOSH National Institute for Occupational Safety and Health
NLM National Library of Medicine
NOAEL no-observed-adverse-effect level
NPSH nonprotein sulfhydryl
NRC National Research Council
XI
-------
NTP National Toxicology Program
OECD Organisation for Economic Cooperation and Development
OSHA Occupational Safety and Health Administration
p probability value
PBPK physiologically based pharmacokinetic (model)
PCB polychlorinated biphenyl
PEL permissible exposure limit
POD point of departure
ppm parts per million
PU pulmonary
R level of risk
RBC red blood cell
RfC reference concentration
RfD reference dose
RGDR regional gas dose ratio
RR relative risk
SA surface area
SD standard deviation
SDH sorbitol dehydrogenase
SIR standard incidence ratio
SMR standardized mortality ratio
TLV threshold limit value
UCL upper confidence limit
UF uncertainty factor
v/v volume/volume
/2 chi squared
XII
-------
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 chloroprene. It is not
intended to be a comprehensive treatise on the chemical or toxicological nature of chloroprene.
The intent of Section 6, Major Conclusions in the Characterization of Hazard and Dose
Response, is to present the major conclusions reached in the derivation of the reference dose, reference
concentration and cancer assessment, where applicable, and to characterize the overall confidence in
the quantitative and qualitative aspects of hazard and dose response by addressing the quality of data
and related uncertainties. The discussion is intended to convey the limitations of the assessment and to
aid and guide the risk assessor in the ensuing steps of the risk assessment process.
For other general information about this assessment or other questions relating to IRIS, the
reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address)
XIII
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER
J.Allen Davis M.S.P.H.
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
AUTHORS
J.Allen Davis, M.S.P.H.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
Jeffrey Gift, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
Karen Hogan, M.S.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C.
Ines Pagan, D.V.M., Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
Reeder Sams II, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
John Stanek, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
XIV
-------
J. Michael Wright, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH
Aparna M. Koppikar, M.D., Ph.D.
E-MAGE Inc.
Los Angeles, CA
CONTRACTOR SUPPORT
Karla Thrall, Ph.D.
Anthony Fristachi
Nick Heyer, Ph.D.
Paul Hinderliter, Ph.D.
Jessica Sanford, Ph.D.
Battelle Memorial Institute
TECHNICAL SUPPORT
Deborah Wales
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
REVIEWERS
This document has been provided for review to EPA scientists, interagency reviewers from
other federal agencies and White House offices, and the public, and has been peer reviewed by
independent scientists external to EPA. A summary and EPA's disposition of the comments received
from the independent external peer reviewers and from the public is included in Appendix A.
INTERNAL EPA REVIEWERS
Ila Cote, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
Lynn Flowers, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
Samantha Jones, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
xv
-------
Connie Meacham, M.S.
National Center for Environmental Assessment
Office of Research and Development
John Vandenberg, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
Debra Walsh, M.S.
National Center for Environmental Assessment
Office of Research and Development
Kate Guyton, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
Jennifer Jinot, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
EXTERNAL PEER REVIEWERS
Herman J Gibb, Ph.D., M.P.H.
Tetra Tech Sciences
Arlington, VA 22201
Dale Hattis, Ph.D.
Clark University
Worcester, MA 01610
Ronald L. Melnick, Ph.D.
Ron Melnick Consulting, LLC
Chapel Hill, NC 27514
John B. Morris, Ph.D.
University of Connecticut
Storrs, CT 06269
Avima M. Ruder, Ph.D.
National Institute for Occupational Safety and Health (NIOSH)
Cincinnati, OH 45226
Richard B. Schlesinger, Ph.D.
Pace University
Pleasantville, NY 10570
XVI
-------
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 chloroprene.
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 a 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 |ig/m3
air breathed.
Development of these hazard identification and dose-response assessments for chloroprene 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, 1986, 001468), Guidelines for Mutagenicity Risk
Assessment (U.S. EPA, 1986, 001466), Recommendations for and Documentation of Biological Values
for Use in Risk Assessment (U.S. EPA, 1988, 064560), Guidelines for Developmental Toxicity Risk
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
-------
Assessment (U.S. EPA, 1991, 008567), Interim Policy for Particle Size and Limit Concentration Issues
in Inhalation Toxicity (U.S. EPA, 1994, 076133), Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994, 006488), Use of the
Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995, 005992), Guidelines for
Reproductive Toxicity Risk Assessment (U.S. EPA, 1996, 030019), Guidelines for Neurotoxicity Risk
Assessment (U.S. EPA, 1998, 030021), Science Policy Council Handbook: Risk Characterization
(U.S. EPA, 2000, 052149), Draft Benchmark Dose Technical Guidance Document (U.S. EPA, 2000,
052150), Supplementary Guidance for Conducting Health Risk Assessment of Chemical Mixtures
(U.S. EPA, 2000, 004421), A Review of the Reference Dose and Reference Concentration Processes
(U.S. EPA, 2002, 088824), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005, 086237).
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
(U.S. EPA, 2005, 088823), Science Policy Council Handbook: Peer Review (U.S. EPA, 2006, 194566),
and A Framework for Assessing Health Risks of Environmental Exposures to Children (U.S. EPA,
2006, 194567).
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 August 2010. It should
be noted that references have been added to the Toxicological Review after the External Peer Review
in response to the reviewers' and public comments. References have also been added for
completeness. These references have not changed the overall qualitative and quantitative conclusions.
See Section 7.1 for a list of these references.
-------
2. CHEMICAL AND PHYSICAL INFORMATION
The monomer 2-chlorobuta-l,3-diene (C/tH5Cl) (hereafter referred to as chloroprene) is a
volatile, flammable liquid used primarily in the manufacture of poly chloroprene (U.S. EPA, 1989,
625024). Polychloroprene rubber is used to make diverse products, such as adhesives, automotive or
industrial parts (e.g., belts/hoses/gaskets), coatings, and dipped goods. While 90% of chloroprene is
used to make poly chloroprene solid (trade names include Neoprene, Bayprene, etc.), about 10% is
converted to poly chloroprene liquid dispersions, a colloidal suspension of poly chloroprene in water
(IARC, 1999, 201838). There was one commercial producer of chloroprene in the United States (U.S.)
in 1995; chloroprene was produced by other plants for on-site use and processing, as a by-product of
vinyl chloride production, or as an impurity in manufacturing processes (NTP, 2005, 093207).
Chloroprene is used almost exclusively to produce polychloroprene, and is sold to only three U.S.
companies for polychloroprene manufacture; less than 20 Ib/yr is sold for research applications2. The
total estimated production of polychloroprene from 1986 to 1988 was approximately 250-300
million Ib (113,000- 136,000 metric tons), and the volume produced from 1995 to 1996 was
approximately 200-250 million Ib (90,700-113,000 metric tons) (NTP, 2005, 093207)3.
There are no known natural occurrences of chloroprene in the environment. The main sources
of releases to the environment are or have been through effluent and emissions from facilities that use
chloroprene to produce polychloroprene elastomers or transport of the product. In 1995, there were
14 facilities reporting releases of chloroprene to the atmosphere totaling 983,888 Ibs (NTP, 2005,
093207). Eight of these plants reported individual atmospheric releases from 2-481,871 Ibs (NTP,
2005, 093207). Three plants in Kentucky, Texas, and Louisiana, each reporting atmospheric releases
of >100,000 Ibs, accounted for most of the reported chloroprene releases in 19954. One of these sites
produced chloroprene, while the other two converted chloroprene to polychloroprene (NTP, 2005,
093207). The chemical structure of chloroprene is shown in Figure 2-1.
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
Through the public comment process, DuPont Performance Elastomers provided updated manufacture, transportation, and emission data. In 2008, there
was one commercial producer of chloroprene in the U.S.; this site both manufactured the monomer and converted it to polymer. Chloroprene is used
almost exclusively to produce polychloroprene, with chloroprene monomer sold to only one U.S. company for nonpolychloroprene manufacture (1,000 Ibs
in 2008).
According to DuPont's public comments, chloroprene production has decreased since 1996 and in 2008, U.S. production volume was below 40,000
metric tons.
According to DuPont's public comments, in 2008, only one chloroprene plant remained open and reported releases of 210,900 Ibs. Domestic production
and releases have been decreasing (reported 2002 emissions were 356,700 Ibs).
-------
H2C. H
H /
C\
Cl CH2
Figure 2-1. The chemical structure of chloroprene.
The starting material for the synthesis of chloroprene is currently 1,3-butadiene in the U.S.
(Lynch, 2001, 625182). Chloroprene manufacture using butadiene as the starting material occurs via a
two step process consisting of chlorination and subsequent dehydrochlorination reactions. Historical
industrial processes (1930s-1970s) for chloroprene manufacture involved the dimerization of acetylene
and then its hydrochlorination to produce chloroprene monomer. Chloroprene is also a structural
analogue of isoprene (2-methyl 1,3-butadiene) and resembles vinyl chloride in that it has a chlorine
bound to a double-bonded carbon (alkene) backbone. However, chloroprene contains four carbons
arranged with two double bonds instead of two carbon atoms. The odor of chloroprene is described as
pungent and ether-like (HSDB, 2009, 594343). Chloroprene is volatile and highly reactive; it is not
expected to bioaccumulate or persist in the environment (OECD, 1998, 624889). Because of its high
vapor pressure (215 mmHg at 25°C), chloroprene is expected to readily volatilize from water and solid
surfaces (NTP, 2005, 093207). Chloroprene vapor has an estimated ionization potential of
8.95 ± 0.05 eV and an estimated half-life in the atmosphere of less than 20 hours (Grosjean, 1990,
625143). Reactions with hydroxyl radicals (*OH) (to produce formaldehyde), Os, and NOs are the
expected pathways of removal, although no experimental data exist (Grosjean, 1991, 625149).
Of particular relevance to any toxicological studies involving chloroprene is its propensity to
spontaneously oxidize and form dimers, peroxides, and other oxygenated species. Stabilizers,
antioxidants, or inhibitors must be added to prevent peroxide formation and consequent spontaneous
polymerization; inhibitors do not reduce dimer formation. Chemically uninhibited chloroprene must
be stored under nitrogen at temperatures below 0°C (e.g., -20°C) to prevent spontaneous
polymerization. If stored at room temperature, uninhibited chloroprene will polymerize to form
various byproducts such as cyclic dimers or open-chain polymers (Stewart, 1971, 010705;
Trochimowicz et al., 1998, 625008). Because these reaction products, if formed, may themselves
account for any observed toxicity, toxicological studies that do not report storage or generation
conditions may yield results that are questionable for their relevance to chloroprene monomer. The
polymerization process has been discussed by Lynch (2001, 646266), Kroshwitz and Howe-Grant
(1993, 010679). Stewart (1971, 010705). and Nystrom (1948, 003695). Additional information on
production and use has been reported by the International Agency for Research on Cancer (IARC,
1999, 201838). Structures have been proposed for some of the chloroprene dimers (Stewart, 1971,
010705); some dimers result upon reaction at room temperature while others result after prolonged
heating.
-------
In addition to volatilization, the potential fate of chloroprene released to soil is likely to leach
into groundwater (NTP, 2005, 093207): however, rapid volatilization into air may mitigate downward
movement into soil. Breakdown via hydrolysis is not likely, as it is only partially soluble in water
(OECD, 1998, 624889). Chloroprene that is released to water may only moderately adsorb to
suspended sediments or particles, and there will be little bioaccumulation in aquatic organisms (log
Kow=2.2).
The occupational exposure potential to chloroprene is limited to facilities in the U.S., Europe,
and Asia where chloroprene is produced and converted to poly chloroprene (Lynch, 2001, 625182)1.
The physical and chemical properties of chloroprene are shown in Table 2-1.
Table 2-1. Physical properties and chemical identity of chloroprene
Chloroprene
CASRN
Synonyms
Melting point
Boiling point
Density
Vapor pressure
Vapor density
Flashpoint (open cup)
Flammability limits
Water solubility
Other solubilities
Log KOW
Henry's law constant
Odor threshold
Molecular weight
Conversion factors (in
air)
Molecular formula
126-99-8
1,3 -butadiene, 2-chloro; chlorobutadiene; 2-chlorobutadiene;
2-chlorobutadiene- 1,3; beta-chloroprene
-130°C
59.4°C
0.956 at 20°C (relative to the density of H2O at 4°C)
215mmHgat25°C
3.0(air=l)
-20 °C
4-20% in air
256-480 mg/L at 20°C
Miscible with ethyl ether, acetone, benzene; soluble in
alcohol, diethyl ether
2.2
5.6 x 10"2 atm/m3-mol at 25°C
15 ppm(54 mg/m3)
88.54
1 mg/m3 = 0.276 ppm; 1 ppm = 3.62 mg/m3 at 25°C, 760 torr
CflsCl
Reference
HSDB (2009, 594343)
HSDB (2009, 594343)
HSDB (2009, 594343)
HSDB (2009, 594343)
HSDB (2009, 594343)
HSDB (2009, 594343)
HSDB (2009, 594343)
OECD (1998, 624889)
HSDB (2009, 594343)
OECD (1998, 624889)
HSDB (2009, 594343)
OECD (1998, 624889)
HSDB (2009, 594343)
U.S. EPA (2000. 625036)
HSDB (2009, 594343)
HSDB (2009, 594343)
HSDB (2009, 594343)
According to DuPont's public comments, as of 2008, occupational exposure potential to chloroprene in the U.S. is limited to one site in Louisiana; other
chloroprene manufacturing facilities exist in Germany, France, Armenia/Azerbaijan, India, China, and Japan.
-------
3. TOXICOKINETICS
No reports are available that address the toxicokinetics of chloroprene in humans by any route
of exposure. Limited information is available for animals regarding the absorption and in vivo
metabolism of chloroprene. No information regarding tissue distribution of chloroprene from animal
studies is available. In vitro studies have been conducted to evaluate the metabolism of chloroprene in
lung and liver tissue fractions from rat, mouse, hamster, and humans (Cottrell et al., 2001, 157445;
Himmelstein et al., 2001, 019013: Himmelstein et al., 2001, 019012: Himmelstein et al., 2004,
625152: Munter et al., 2003, 625214: Munter et al., 2007, 576501: Munter et al., 2007, 625213:
Summer and Greim, 1980, 064961). Hurst and Ali (2007, 625159) evaluated the kinetics of R- and S-
enantiomers of the chloroprene metabolite (l-chloroethenyl)oxirane in mouse erythrocytes. A
physiologically based pharmacokinetic (PBPK) model has been developed to describe changes in
chamber chloroprene concentrations during exposures with mice, rats, and hamsters (Himmelstein et
al., 2004, 625152: Himmelstein et al., 2004, 625154). No in vivo time-course data for blood or tissue
concentration are available for model validation.
3.1. ABSORPTION
Quantitative data on the absorption of chloroprene from any route of exposure have not been
reported. The Hazardous Substances Data Bank (HSDB) states that chloroprene is "rapidly absorbed
by the skin" (HSDB, 2009, 594343: Lefaux, 1968, 625192: NIOSH, 1977, 644450: NIOSH, 1995,
644453). Chronic inhalation studies in B6C3F1 mice and F344/N rats suggest that chloroprene has
multiple nonneoplastic and neoplastic targets (nose and lung, kidney, forestomach, Harderian gland,
skin); therefore, the absorption and systemic distribution via the inhalation route can be inferred (NTP,
1998, 042076).
3.2. DISTRIBUTION
No quantitative in vivo data on the tissue distribution of chloroprene have been reported. As
indicated above, the widespread distribution of chloroprene in vivo following absorption can be
inferred from effects in several target organs (NTP, 1998, 042076). Himmelstein et al. (2004, 625154)
determined tissue-to-air partition coefficients for chloroprene in mouse, F344 rat, Wistar rat, and
hamster tissues by using the vial equilibration method described by Gargas et al. (1989, 063084).
Briefly, gas-tight vials (10 mL) were prepared in triplicate as either reference vials or containing
samples of blood, lung, liver, fat, muscle, or kidney. The vials were sealed and 100 ppm chloroprene
was added after preheating to 37°C for 5 minutes. 100 jil samples were taken at 1.5, 3, and 4.5 hours
from the start of incubation. For measurement of the human blood-to-air partition coefficient, blood
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
-------
samples were drawn from three healthy male subjects and analyzed in triplicate (Himmelstein et al.,
2004, 625154). Results are given in Table 3-1. These tissue-to-air ratios suggest that chloroprene will
be preferentially distributed in adipose tissue, followed by lung, kidney, liver, and muscle. The
relatively low blood:air partition coefficients across species suggests that chloroprene would not likely
be efficiently scrubbed in the upper airways. The partition coefficient values suggest there are no
significant species differences expected in tissue distribution of chloroprene.
Table 3-1. Tissue-to-air partition coefficients for chloroprene
Tissue
Blood
Lung
Liver
Fat
Muscled
Kidney6
Tissue-To-Air Partition Coefficients (Mean ± SE)a
Mouse
7.8±0.1
18.6 ±5.1
9.8 ±0.9
135. 3 ±1.6
4.6 ±0.8
13.7 ±0.6
F344 rat
7.3 ±0.1
13. 5 ±1.6
11.5±0.3
124.0 ± 1.5
4.4 ±0.4
16.7 ±0.6
Wistar rat
8.0 ±0.5
11.2 ±0.5
10.9 ±0.2
126.3 ± 1.4
4.0 ±0.3
9.4 ±0.4
Hamster
9.3 ±0.3
9.7 ±0.6
10.5 ±0.5
130.1 ±0.9
5.0 ±0.2
8.2 ±0.3
Human
4.5±0.1b
13.3±4.1C
10.7 ± 1.1°
128.9 ±2.7C
4.5±1.0C
12.0±0.9C
"Mean ± standard error (SE) for three replicates per rodent tissue.
bHuman chloroprene blood values determined for nine replicates (three subjects, three vials/subject).
°Human tissue partition coefficient values (other than from blood) were derived from rodents; the standard error
was adjusted to account for the proportion of variation from each set of rodent data.
dUsed to represent the slowly perfused tissue group.
eUsed for rapidly perfused tissue group
Source: Used with permission from Oxford University Press, Himmelstein et al. (2004, 625154).
3.3. METABOLISM
The metabolism of chloroprene has been primarily evaluated in vitro with lung and liver tissue
fractions from rat, mouse, hamster, and humans (Cottrell et al., 2001, 157445; Himmelstein et al.,
2001, 019013; Himmelstein et al., 2001, 019012; Himmelstein et al., 2004, 625152; Munter et al.,
2003, 625214; Munter et al., 2007, 576501; Munter et al., 2007, 625213; Summer and Greim, 1980,
064961). In a 1978 review of the older literature, a number of reports suggested that chloroprene
forms peroxides that interact with tissue thiol groups and that the disposition of chloroprene is likely
similar to that of vinyl chloride and vinylidene chloride (Haley, 1978, 010685). This report was the
first to postulate a metabolic profile of chloroprene, including formation of epoxides by cytochrome
P450 (CYP450) enzymes that could give rise to aldehydes and eventually form mercapturic acid
derivatives.
In studies using mouse and human liver microsomes, Bartsch et al. (1979, 010689) showed that
chloroprene was enzymatically converted into a reactive metabolite and postulated that this metabolite
was probably an epoxide. This was based on the finding that 4-(4-nitrobenzyl)pyridine trapped a
volatile metabolite produced during reaction of mouse liver microsomes with chloroprene. The
-------
authors proposed that the epoxidation of the carbon double bonds in chloroprene yields one of two
isomeric oxiranes (or both): 2-chloro-2-ethynyloxirane and/or (l-chloroethenyl)oxirane. A report by
Himmelstein et al. (2001, 019012) was the first to quantitatively identify (l-chloroethenyl)oxirane as
an epoxide metabolite of chloroprene and confirmed the identify of the volatile metabolite reported by
Bartsch et al. (1979, 010689). Microsomal suspensions were isolated through differential
centrifugation of livers pooled from male B6C3Fi mice, Fischer and Wistar rats, and Syrian hamsters.
Human liver microsomal suspensions were prepared from a mixed pool of 15 different individuals.
Chloroprene (800 ppm) was incubated with the microsomal suspensions (1 mg) in sealed vials for all
species. Incubations were stopped after 30 minutes by the addition of cold diethyl ether containing
1-butanol as an internal standard and analyzed using gas chromatography mass spectroscopy.
Himmelstein et al. (2001, 019012) reported that incubation of chloroprene with liver microsomes of all
species resulted in an apparent spectrographic peak that was consistent with (l-chloroethenyl)oxirane
(based on comparison to synthesized (l-chloroethenyl)oxirane standard). Comparisons of the amount
of (l-chloroethenyl)oxirane to the amount of the 1-butanol standard indicated that a greater amount of
(l-chloroethenyl)oxirane was present in B6C3Fi mice and F344 rat liver microsomes, followed by the
Wistar rat, humans, and hamsters (Table 3-2). Additional time course experiments showed that the
decline of chloroprene (from 3 to 0.1 jiM between 5-10 minutes after start of incubation with
[0.05 jiMJlOO ppm chloroprene) from the headspace of mouse liver microsomes coincided with an
increase of (l-chloroethenyl)oxirane (0.01-0.02 |iM). Metabolism of chloroprene into
(l-chloroethenyl)oxirane most likely involved CYP2E1, as evidenced by nearly complete in vitro
inhibition with 4-methylpyrazole hydrochloride.
-------
Table 3-2. Liver microsomal metabolites as a percentage of 1-butanol internal
standard3
Metabolite Peakb
1
2
3
4
5
6
Liver Microsomal Suspension
B6C3F!
mouse0
9.0
0.0
0.8
0.2
0.2
0.6
F344 ratc
12.0
0.1
0.3
0.0
0.3
0.4
Wistar ratc
4.0
0.1
0.2
0.1
0.0
0.3
Hamster0
0.8
0.2
0.8
0.4
0.1
0.3
Human0
1.3
0.1
0.3
0.1
0.0
0.1
"Metabolites as a percentage of 1-butanol internal standard on GC/MS selected ion monitoring (metabolite area/
internal standard area) x 100%.
Vfetabolite peak 1 = (l-chloroethenyl)oxirane. Metabolite peaks 2-5 had insufficient signal to obtain meaningful
spectral data. A tentative spectral match for peak 6 was made as 3-chloro-2-butenal, but was not confirmed.
°One vial for each species was incubated with 800 ppm chloroprene (0.8 umol) for 30 minutes followed by diethyl
ether extraction, 20-fold concentration, and cold on-column injection.
Source: Used with permission from Elsevier Science Ireland Ltd., Himmelstein et al. (2001, 019012).
Further metabolism of (l-chloroethenyl)oxirane was observed in time-course evaluations with
liver microsomes (Himmelstein et al., 2001, 019012). In vitro uptake of (l-chloroethenyl)oxirane from
vial headspace of liver microsomes was observed, with preliminary results indicating that the ranking
of (l-chloroethenyl)oxirane hydrolysis in liver microsomes was as follows: hamsters ~ humans >
Wistar rats > B6C3Fi mice and F344 rats. The uptake of (l-chloroethenyl)oxirane was attributable to
either epoxide hydrolase-mediated hydrolysis or further oxidative metabolism. Time course
experiments demonstrated the uptake of (l-chloroethenyl)oxirane from hepatic cytosol from mice, rats,
or hamsters. Uptake was absent in boiled cytosol, or glutathione depleted cytosol, indicating that
conjugation of (l-chloroethenyl)oxirane to glutathione was enzyme-dependent. The relative activity of
glutathione conjugation was as follows: hamsters > rats > mice (human cytosol was not evaluated).
Studies by Cottrell et al. (2001, 157445) are in agreement with reports from Himmelstein et al.
(2001, 019013; 2001, 019012) and further define the structures and stereochemistry of chloroprene
metabolites from rodent species and humans by comparison with synthetic reference standards. Based
on these studies, the metabolic pathway illustrated in Figure 3-1 was proposed.
-------
Cl
P450
Cl
o
Cl
1 -chlorobut-3-
en-2-one
EH
l-(chloroethenyl)
oxirane a
chloroprene - P450
ci
l-chlorobutan-
2-one
Cl
EH/H,0
2-chloro-2-
ethenyloxiranea
rearrange-
ments
Cl
o
2-chlorobut-3-
en-1 -al
OH
Cl
3-chlorobut-3-ene-
1,2-diol"
O
l-chloro-2-hydroxy-
but-3-ene
GSH
conjugate
l-hydroxybut-3-en-
2-one
0
OH
1 -hy drox y butan-
2-one
Cl
Cl
o
2-chlorobut-2-
en-l-al
2-chloro-butanal
a R- and S- enantiomers.
Source: Adapted with permission from the American Chemical Society, Cottrell et al.
(2001. 157445).
Figure 3-1. Proposed metabolism of chloroprene.
Comparing in vitro metabolism between species, Cottrell et al. (2001, 157445) observed that
qualitative profiles of metabolites from liver microsomes obtained from B6C3Fi mice, Sprague-
Dawley or F344 rats, and humans were similar. Microsomal suspensions were prepared by differential
centrifugation from livers pooled from male and female B6C3Fi mice and Sprague-Dawley rats.
Human liver microsomal suspensions were prepared from a mixed pool of 15 different individuals. In
all species and either gender, (l-chloroethenyl)oxirane was the major metabolite detected. An
important difference among species was in the stereoselectivity of the P-450 mediated formation of R-
and S-enantiomers of (l-chloroethenyl)oxirane in the presence of a epoxide hydrolase inhibitor
(cyclohexene oxide) (Table 3-3). For liver microsomes from both male and female Sprague-Dawley
and F344 rats, there was a distinct enantioselectivity in the mono-epoxidation of chloroprene to
preferentially form the R-enantiomer of (l-chloroethenyl)oxirane. Both female and male B6C3Fi mice
and humans showed slight enantioselectivity in metabolism to the S-enantiomer. In incubations
without an inhibitor of epoxide hydrolase present, (l-chloroethenyl)oxirane was not detected as a
10
-------
metabolite. Instead, 3-chlorobut-3-ene-l,2-diol was observed, indicating that epoxide hydrolase is
effective in the detoxification of the epoxide metabolite of chloroprene. In incubations supplemented
with an epoxide hydrolase inhibitor and glutathione, there was no change in the observed levels of
(l-chloroethenyl)oxirane, suggesting that conjugation with glutathione may not be an active
detoxification pathway for the active epoxide metabolite of chloroprene. Glutathione conjugation was
apparent with l-hydroxybut-3-en-2-one, the downstream product of the minor epoxide metabolite of
chloroprene, 2-chloro-2-ethenyloxirane.
Table 3-3. Stereochemical comparison of relative amounts (percentages) of R- and
S-enantiomers of the major chloroprene metabolite (l-chloroethenyl)oxirane from
liver microsomes compared across species, strains, gender, and chloroprene
concentration (mM)
Male
Chloroprene
(mM)
5
10
20
30
40
5
10
20
30
40
5
10
20
30
40
10
20
30
Species/straina'b
Sprague-Dawley
rat
F344 rat
B6C3F! mouse
Human
%R
58
62
61
60
64
62
62
62
60
64
48
47
46
47
47
43
43
43
%S
42
38
39
40
36
38
38
38
40
36
52
53
54
53
53
57
57
57
Female
Chloroprene
(mM)
10
20
30
40
10
20
30
40
10
20
30
40
10
20
30
Species/straina'b
Sprague-Dawley
rat
F344 rat
B6C3F! mouse
Human
%R
56
56
55
59
56
54
53
54
47
45
47
46
43
44
42
%S
44
44
45
41
46
46
47
46
53
55
53
54
57
56
58
aAverage of three samples per species/strain.
Percentages (estimated error ± 1%) were determined by comparison of peak areas from GC/MS selected ion
monitoring measurements versus those from synthetic standards.
Source: Used with permission from American Chemical Society, Cottrell et al. (2001, 157445).
A further study by this group (Munter et al., 2003, 625214) verified significant differences
between species in the amounts of R- and S-enantiomers of (l-chloroethenyl)oxirane formed in liver
microsomes from rats, mice, or humans without epoxide hydrolase inhibitor present. Microsomal
samples were prepared in the same manner as for Cottrell et al. (2001, 157445). After incubation with
11
-------
10 jiM chloroprene, the relative ratio of the R-enatiomer of (l-chloroethenyl)oxirane formed in mice,
rat, or human microsomes was 20:4:1. This ratio was also observed in incubations with 100 jiM and
10 mM chloroprene. For the S-enantiomer, the presence of (l-chloroethenyl)oxirane was detected in
only mouse microsomes after incubations with 10 jiM chloroprene. After incubations with 100 jiM
chloroprene, S-(l-chloroethenyl)oxirane was detected in rat microsomes, but at levels approximately
10-fold less than observed in mouse microsomes. The formation of S-(l-chloroethenyl)oxirane was
not observed in human microsomes at any incubation concentration. Therefore, in the presence of
epoxide hydrolase, microsomal oxidation of chloroprene to (l-chloroethenyl)oxirane was most
effective in the mouse, and epoxide hydrolase preferentially hydrolyzed the S-enantiomer of
(l-chloroethenyl)oxirane, leading to an accumulation of the R-enantiomer. Levels of detected
3-chlorobut-3-ene-l,2-diol were highest in mouse microsomes compared to rats or humans (which had
similar levels). Additional experiments identified 3 conjugates when racemic (l-chloroethenyl)oxirane
was incubated with glutathione at 37°C in an aqueous phosphate buffer solution, but further indicated
that (l-chloroethenyl)oxirane either did not react with glutathione or did so very slowly in microsomal
incubations with chloroprene. Addition of liver cytosol (containing glutathione transferase) only
marginally affected the formation of glutathione conjugates. Downstream metabolites formed from the
minor epoxide metabolite, 2-chloro-2-ethenyloxirane, were shown to rapidly react with glutathione
even in the absence of glutathione transferase. At all concentrations of chloroprene, the total amount
of glutathione-conjugated metabolites formed in liver microsomes was highest for the mouse, followed
by the rat, and then humans.
Hurst and Ali (2007, 625159) evaluated the kinetics of R- and S-enantiomers of
(l-chloroethenyl)oxirane in mouse erythrocytes. These results implied that
S-(l-chloroethenyl)oxirane was much more quickly detoxified than the R-enantiomer when incubated
with mouse erythrocytes in vitro. The disappearance of S-(l-chloroethenyl)oxirane was blocked when
erythrocytes were preincubated with diethyl maleate, which indicates that rapid removal is dependent
on cellular glutathione. The study by Hurst and Ali (2007, 625159) suggested that the R-enantiomer of
(l-chloroethenyl)oxirane is potentially more toxic because of slower detoxification.
Summer and Greim (1980, 064961) reported that in vitro incubation of hepatocytes isolated
from male Wistar rats with chloroprene decreased cellular glutathione levels to approximately 50% that
of controls after 15 minutes of exposure to 3 mM chloroprene. This effect was dose-dependent and
was observed with exposures to 0.5 and 1.0 mM as well. The limited in vivo rodent studies support the
postulated metabolic pathway for chloroprene. In male Wistar rats (four per experiment) exposed
orally to either 100 or 200 mg/kg chloroprene via gavage (Summer and Greim, 1980, 064961), hepatic
glutathione levels fell to 55 and 39% that of controls three hours after exposure, respectively. These
results indicate that glutathione conjugation plays an active role in the detoxification of chloroprene.
Pre-treatment of rats or hepatocytes with phenobarbital or a poly chlorinated biphenyl (PCB) mixture
(Clophen A50) to induce the mixed-function oxidase enzymes enhanced the GSH depletion effect.
12
-------
Himmelstein et al. (2004, 625152) investigated the in vitro metabolism of chloroprene in
mouse, rat, hamster, and human liver and lung microsomes. Rodent microsomes and cytosol were
prepared from pooled liver and lungs using differential centrifugation. Human microsomes and
cytosol were prepared from pooled individuals as follows: pooled liver microsomes from 15
individuals for experiments involving hydrolysis of (l-chloroethenyl)oxirane, pooled liver microsomes
from 10 individuals for simultaneous measurement of chloroprene and (l-chloroethenyl)oxirane,
pooled lung microsomes from 5 individuals, pooled liver cytosol from 15 individuals, and lung cytosol
from 1 individual. Experiments investigating the microsomal metabolism of chloroprene or
(l-chloroethenyl)oxirane were conducted in closed vials and headspace samples were analyzed using
gas chromatography. A two-compartment closed vial model was developed to describe both
chloroprene and (l-chloroethenyl)oxirane metabolism in the liver and lung microsomes (from rodents
and humans). Liquid-to-air partition coefficients measured in Himmlestein et al. (2001, 019012)
(0.69 ± 0.05 for chloroprene and 57.9 ±1.6 for (l-chloroethenyl)oxirane) were used to calculate liquid
phase concentrations for modeling purposes.
Chloroprene oxidation in liver microsomes for all species was described as a saturable
Michaelis-Menten mechanism. In liver microsomes, the rate (as expressed by Vmax/Km, mL/h/mg
protein) of chloroprene oxidation was faster in the mouse and hamster than in rats or humans
(Table 3-4). Chloroprene oxidation in mouse lung microsomes was also saturable, and oxidation
appeared saturated at all doses in hamsters, rats, and humans; the rate was optimized as Vmax/Km rather
than individual measurements of Vmax or Km for these species (Table 3-4). Chloroprene oxidation in
lung microsomes was much greater (approximately 50-fold) for mice compared with the other species.
Microsomal hydrolysis of (l-chloroethenyl)oxirane also operated via saturable Michaelis-Menten
mechanics, especially in human and hamster liver and lung microsomes (Table 3-5). Hydrolysis
(Vmax/Km) of (l-chloroethenyl)oxirane in liver and lung microsomes was fastest for humans, followed
by rodent species (Table 3-5).
13
-------
Table 3-4. Kinetic parameters used to describe the microsomal oxidation of
chloroprene
Tissue
Liver
Lung
Species
Mouse
F344 rat
Wistar rat
Hamster
Human
Mouse
F344 rat
Wistar rat
Hamster
Human
Activity Of Microsomal Oxidation"
Vmax
0.23
0.078
0.11
0.29
0.068
0.10
-
-
-
-
Km
1.03
0.53
0.84
1.33
0.68
1.5
-
-
-
V max' -"-m
224
146
125
218
101
66.7
1.3b
1.3b
1.3b
1.3b
aValues derived from modeling of vial headspace concentration time-course data (using a liquid-to-air partition
coefficient of 0.69) (Himmelstein et al., 2001, 019012). Vmax, umol/h/mg protein, Km, umol/L, Vmax/Km,
mL/h/mg protein.
bThe apparent rate of lung metabolism, over the range of biologically relevant concentrations tested, was linear
and was estimated as VmaK/Km
Source: Used with permission from Oxford University Press, Himmelstein et al. (2004, 625152).
Table 3-5. Kinetic parameters used to describe the microsomal epoxide hydrolase
activity of (l-chloroethenyl)oxirane
Tissue
Liver
Lung
Species
Mouse
F344 rat
Wistar rat
Hamster
Human
Mouse
F344 rat
Wistar rat
Hamster
Human
Activity Of Microsomal Epoxide Hydrolase
V a
v max
0.14
0.60
0.64
2.49
3.66
0.11
0.12
0.16
1.34
0.58
Km
20.9
41.5
53.0
73.8
99.7
51.5
90.9
91.6
187.6
72.2
* max' -"-m
6.7
14.5
12.1
33.7
36.7
2.1
1.3
1.7
7.1
8.0
aValues derived from modeling of vial headspace concentration time-course data (using a liquid-to-air partition
coefficient of 57.9) (Himmelstein et al., 2001, 019012). Vmax, umol/h/mg protein, Km, umol/L, Vmax/Km,
mL/h/mg protein.
Source: Used with permission from Oxford University Press, Himmelstein et al. (2004, 625152).
14
-------
Further hydrolysis experiments, conducted in the presence or absence of NADP+, demonstrated
oxidation of (l-chloroethenyl)oxirane in mouse liver microsomes, but not in human, rat, or hamster
liver microsomes. When experiments were carried out in the presence of NADP+, pre-treatment of
mouse microsomal preparations with 4-methylpyrazole (4-MP) or 1-aminobenzotriazole (ABT), both
inhibitors of P450 monooxygenase, did not affect hydrolysis but completely inhibited oxidation.
Results were similar when experiments were carried out in the absence of NADP+. Although oxidation
of (l-chloroethenyl)oxirane could potentially produce diepoxides, only 3-chloro-3-butene-l,2,-diol
was detected, in agreement with Cottrell et al. (2001, 157445). The potential for
(l-chloroethenyl)oxirane oxidation was not evaluated in lung microsomes.
The cytochrome P450 dependent oxidation of chloroprene in both liver and lung microsomes
coincided with an increase in (l-chloroethenyl)oxirane in the vial headspace. Peak concentrations of
(l-chloroethenyl)oxirane ranged from 0.01 to 0.1 nmol/mL for liver microsomes, and the greatest
concentration (0.1 nmol/mL) was observed in the mouse due to the faster rate of chloroprene
oxidation compared to the rat, hamster, or human. The chloroprene-dependent formation of
(l-chloroethenyl)oxirane was apparent in mouse lung microsomes with headspace concentrations
approximate to mouse liver microsomes. (l-chloroethenyl)oxirane was detected in rat and hamster
lung microsomes despite lower levels of chloroprene oxidation compared to mice. Only one
detectable value of (l-chlororethenyl)oxirane was recorded in human lung microsomes due to the high
activity of epoxide hydrolase. A satisfactory model fit to (l-chloroethenyl)oxirane formation was
obtained when the oxidative metabolism of chloroprene was split into (l-chloroethenyl)oxirane and
other uncharacterized metabolites, and then the measured epoxide hydrolase kinetics were applied.
Formation of (l-chloroethenyl)oxirane was best modeled as making up only 2-5% of total oxidation of
chloroprene in the liver across all species (Table 3-6). Similar adjustment in lung microsomes
indicated that formation of (l-chlororethenyl)oxirane accounted for 3-22% of total chloroprene
metabolism in rodents, although the adjustment was less robust than for the liver due to limited time
course data. The value of 78% total metabolism for human lung microsomes was most likely an
overestimate due to the rapid removal of (l-chloroethenyl)oxirane by epoxide hydrolase. In the lung,
the rate of (l-chloroethenyl)oxirane formation appeared to be 10-fold greater in mice compared to rats,
and twofold greater compared to humans.
15
-------
Table 3-6. Kinetic parameters used to describe the time course of
(l-chloroethenyl)oxirane formation from microsomal oxidation of chloroprene
Tissue
Liver
Lung
Species
Mouse
F344 rat
Wistar rat
Hamster
Human
Mouse
F344 rat
Wistar rat
Hamster
Human
(l-chloroethenyl)oxirane (CEO) Formation3
vmax
0.149
0.184
0.148
0.048
0.108
0.050
0.0075
0.0082
0.013
0.024
Km
36.6
23.7
25.3
9.0
20.7
25.0
40.4
30.1
81.2
24.6
^max'-l^m
4.1
7.8
5.8
5.4
5.2
2.0
0.19
0.27
0.16
0.98
Ratio of CEO/total
(%)b
2
5
5
2
5
3
15
22
13
78
"Optimized oxidative rate constants used to describe the amount of (l-chloroethenyl)oxirane derived from total
chloroprene oxidation. Vmax, umol/h/mg protein, Km, umol/L, Vmax/Km, mL/h/mg protein.
bVmax/Km for CEO formation divided by the Vmax/Km for total chloroprene oxidation (from Table 3-4) multiplied by
100.
Source: Used with permission from Oxford University Press, Himmelstein et al. (2004, 625152).
Glutathione S-transferase-mediated metabolism of (l-chloroethenyl)oxirane in cytosolic tissue
fractions was described as a pseudo second-order reaction, with rates ranging from 0.0016-0.0130
hour/mg cytosolic protein in liver and 0.00056-0.0022 hour/mg in lung. In the liver the rates were as
follows: hamster > Fischer rat ~ Wistar rat > mouse > human. In the lung cytosol the rates were as
follows: mouse > Fischer rat > human > Wistar rat > hamster. The half-life of the spontaneous first-
order reaction between (l-chloroethenyl)oxirane and glutathione was approximately 10 hours.
16
-------
Table 3-7. Kinetic parameters used to describe the cytosolic glutathione S-
transferase activity towards (l-chloroethenyl)oxirane
Tissue
Liver
Lung
Species
Mouse
F344 rat
Wistar rat
Hamster
Human
Mouse
F344 rat
Wistar rat
Hamster
Human
Activity Of Cytosolic Glutathione S-Transferase3^
ks
0.0015
0.0074
0.011
0.024
0.0017
0.0011
0.0023
0.0051
0.015
0.0028
pBS(O)
2.7
0.92
0.56
0.54
0.94
2.01
0.70
0.18
0.038
0.44
ks x CBS(0)
0.0040
0.0068
0.0063
0.0130
0.0016
0.0022
0.0016
0.00092
0.00056
0.0012
aNote: ks (1/umol/h/mg cytosolic protein), rate constant CBS(0) (umol/L) as initial concentration of protein binding
sites and ks x CBS(0) (h/mg protein) describing enzymatic (l-chloroethenyl)oxirane-glutathione conjugate
formation as a pseudo-second order reaction.
bFirst order reaction of (l-chloroethenyl)oxirane with glutathione was measured as kf = 0.07 h"1 independent of
protein.
Source: Used with permission from Oxford University Press, Himmelstein et al. (2004, 625152).
Himmelstein et al. (2004, 625154) conducted closed-chamber gas uptake exposures to evaluate
chloroprene metabolism rates in rats (Wistar and F344), mice (B6C3Fi), and hamsters (Syrian golden).
The first exposure scenario investigated chemical distribution with or without metabolic inhibition
with 4-methyl pyrazole. Exposure concentrations ranged from 160-240 parts per million (ppm)
chloroprene. Animals (Wistar and F344 rats and B6C3Fi mice, n = 3) were placed in the exposure
chamber 30 minutes prior to exposure. The chamber atmosphere was circulated through the system at
2 L/min and chloroprene concentrations were analyzed by gas chromatography flame ionization
detection for up to six hours. The second exposure scenario measured the uptake of chloroprene over a
range of starting concentrations. Only one rat was used per exposure chamber at one time and
hamsters were substituted for Wistar rats in this second exposure. A known volume of concentrated
chloroprene was added to the chamber at the start of each exposure, with starting concentrations
ranging from 2 to 400 ppm for mice and rats and 10 to 270 ppm for hamsters. APBPK model was
used to describe the decrease in chamber chloroprene concentrations over time by using metabolic
parameters (Vmax, Km) scaled from in vitro studies (Himmelstein et al., 2004, 625152). The in vitro
scaling of total chloroprene metabolism (Table 3-8) was sufficient to explain the in vivo gas uptake
data. Inhibition of uptake was obtained with pre-treatment with 4-methyl pyrazole, indicating the loss
of chamber chloroprene was due to metabolic oxidation via P-450 monooxygenases. Setting Vmax to
zero for liver and lung metabolism allowed the PBPK model to obtain sufficient fit to the observed
inhibition data.
17
-------
Table 3-8. Metabolic parameters of chloroprene
Biochemical Parameters"
Liver
Lung
Vmax(mg/h/kgBW)
Km (mg/L)
Vmax/Km(L/h/kgBW)
Vmax(mg/h/kgBW)
Km (mg/L)
Vmax/Km(L/h/kgBW)
Species
Mouse
39.2
0.091
431.0
1.02
0.13
7.67
F344 rat
11.50
0.047
244.0
—
—
0.14
Wistar rat
15.5
0.075
208.0
—
—
0.14
Hamster
42.8
0.118
363.0
—
—
0.14
aScaled from Himmelstein et al. (2004, 625152) using microsomal protein contents to estimate metabolic
parameters.
Source: Used with permission from Oxford University Press, Himmelstein et al. (2004, 625154).
3.4. ELIMINATION
Limited information is available regarding the elimination of chloroprene in rodents. Summer
and Greim (1980, 064961) exposed male Wistar rats (four per experiment) to 100 or 200 mg/kg
chloroprene by gavage and observed a dose-dependent, nonlinear increase in excreted urinary
thioesters (presumably glutathione conjugates and mercaptic acids). This increase in urinary thioesters
was reversible and levels of urinary thioesters returned to control levels within 24 hours, indicating that
elimination was rapid. At higher concentrations of chloroprene, a decline in the excretion rate of
urinary thioesters was observed
Consideration of physiological and biological factors suggests differences may exist in
chloroprene clearance across species. For example, while the fatair partition coefficient is similar for
all species investigated (Table 3-1), humans have a much greater amount of fat as a percentage of body
weight compared to rodents. This may mean that a greater total amount of chloroprene partitions into
the fat of humans thereby increasing the time necessary to eliminate chloroprene from the body for
humans. Also, it has been shown that metabolic oxidation and hydrolysis rates vary substantially
across species. These differences in enzyme activity may lead to differences in chloroprene body
burdens and elimination profiles.
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS
Himmelstein et al. (2004, 625154) published a PBPK model of chloroprene to describe gas
uptake data and calculate internal dose metrics for use in dose-response analyses. Construction of the
mathematical model was based on physicochemical, physiological, and metabolic parameters for
chloroprene from mouse, rat, hamster, and humans (Table 3-9). The model consisted of distinct
compartments for liver and lung, as well as lumped compartments for fat and slowly and rapidly
perfused tissues. Individual tissues were modeled as homogenous, well-mixed compartments
connected by systemic circulation. Metabolism of chloroprene was localized to the lung and liver
18
-------
compartments and described by Michaelis-Menten type saturable kinetics. Standard physiological
values were used to parameterize the model. Tissue-to-blood partition coefficients were calculated
from tissue-to-air values and the in vivo metabolic parameters (Table 3-8) were scaled from in vitro
metabolic parameters for total chloroprene metabolism in the liver and lung (Himmelstein et al., 2004,
625152) using microsomal protein content. Microsomal protein contents for the liver differ among
species and were obtained from the literature. The microsomal protein content for the lungs was set as
equal for all species. Gas uptake was modeled by subtracting the amount taken up by the animal from
the chloroprene concentration in the chamber. Physiological and metabolic parameters were not
adjusted except for alveolar ventilation and cardiac output as needed to obtain adequate model fit to the
gas uptake data.
Table 3-9. Physiological parameters used for chloroprene PBPK modeling
Physiological Parameters
Species
Mouse
F344 rat
Wistar rat
Hamster
Human
Values for dose response modeling a>b
Body weight (kg)
Ventilation (L/h/kg° 75)
Cardiac output (L/h/kg° 75)
0.03
30
30
0.25
21
18
0.25
21
18
0.11
30
30
70
16.2
16.2
Values for simulation of chamber gas uptake0
Body weight (kg)
Ventilation (L/h/kg° 75)
Cardiac output (L/h/kg° 75)
0.024-0.034
15
15
0.16-0.28
10.5
9
0.20-0.34
10.5
9
0.10-0.18
12
12
NA
NA
NA
Tissue volumes (% body weight)31"1
Liver
Fat
Rapid perfused
Slow perfused
Lung
5.5
5.0
3.5
77.0
0.73
4.0
7.0
5.0
75.0
0.50
4.0
7.0
5.0
75.0
0.50
4.0
7.0
5.0
75.0
0.50
2.6
21.4
7.7
56.1
0.76
Blood flow (% cardiac output)a>d
Liver
Fat
Rapid perfused
Slow perfused
16.1
7.0
51.0
15.0
18.3
7.0
51.0
15.0
18.3
7.0
51.0
15.0
18.3
7.0
51.0
15.0
22.7
5.2
47.2
24.9
Parameters for mouse, rats, and humans drawn from the literature. Hamster ventilation, cardiac output, tissue
volume, and blood flow values were based on the mouse and rat.
bValues used for the dose-response modeling are based on average body weight data from chronic inhalation studies
and the assumption that literature values for ventilation and cardiac output are representative of repeat inhalation
exposure condition.
°Values used specifically for simulation of closed chamber gas uptake data. NA-not applicable.
dTissue volumes and blood flows were calculated by the model with resulting units of liters (L) and L/h, respectively.
Source: Used with permission from Oxford University Press, Himmelstein et al. (2004, 625154).
19
-------
Although the model was used to estimate the chloroprene concentration in each of the defined
compartments (including blood), comparisons of model predictions were limited to experimental
determinations of chloroprene vapor uptake in closed chambers. Inhibition of uptake was achieved
with 4-methyl pyrazole pre-treatment, indicating that the decline of chloroprene chamber concentration
was due to CYP450 monooxygenase-mediated metabolism. The loss in chamber concentration in the
presence of metabolic inhibition represented uptake due to chemical distribution within the animal. A
satisfactory model description for metabolic inhibition was obtained by setting Vmax to zero for both
liver and lung metabolism. Model simulations demonstrated good agreement with chamber uptake
data for a wider range of starting chloroprene concentrations for mice, rats, and hamsters. Scaling of in
vitro metabolic parameters was sufficient to explain the in vivo gas uptake data. The alveolar
ventilation and cardiac output values used to simulate the chamber gas uptake data were lower than the
standard values used in the dose-response modeling. Justification for application of lower alveolar
ventilation and cardiac output values for the gas uptake simulations included decreased ventilation due
to sensory irritation and anesthetic effects. The decision to use standard values as reported in the
literature for the dose-response modeling was that these values more likely represent bioassay
conditions involving chronic, whole-body exposures. Use of a model-calculated internal dose metric
(total chloroprene metabolism/g lung tissue/day) was used in a dose-response analysis of bronchiolar
adenoma/carcinoma in male rodents (NTP, 1998, 042076; Trochimowicz et al., 1998, 625008), and
was found to fit the incidence data much better than the external dose metric. Lastly, the model was
used to calculate exposure concentrations for humans that would result in internal doses equivalent to
the internal dose calculated from the dose-response analysis in rodents.
DeWoskin (2007, 202141) reviewed the chloroprene PBPK model and suggested the following
potential applications of the model for developing an IRIS assessment:
1. Correlate parent compound concentration or total amount metabolized with cancer and
noncancer endpoints in order to determine the relevant mode(s) of action.
2. Investigate observed species differences in the external dose-response relationship.
3. Estimate the human dose-response based on the most relevant internal dose metric for the
proposed mode of action.
4. Use PBPK model parameter distributions to represent variability in intra-population rates of
chemical absorption, distribution, metabolism, and elimination in order to estimate human
variability.
Himmelstein et al. (2004, 625154) addressed the first three of these suggestions in the
application of the PBPK model. DeWoskin (2007, 202141) also notes that in order for a PBPK model
to be applied in the IRIS process, it must be reviewed in detail in regard to the scientific assumptions
used in its construction and application. Currently, the Himmelstein et al. (2004, 625154) model has a
number of limitations. The model currently predicts blood chloroprene and delivery of chloroprene to
20
-------
metabolizing tissues based on metabolic constants and partition coefficients based on in vitro data.
Loss of chamber chloroprene is attributed to uptake and metabolism by test animals and was used to
test the metabolic parameters and validate the model. However, Himmelstein et al. (2004, 625154) did
not provide results of sensitivity analyses indicating whether chamber loss was sensitive to
metabolism, and therefore it is uncertain whether chamber loss is useful for testing the metabolic
parameters used in the model. Also, the chamber data were fit by varying alveolar ventilation and
cardiac output. This method does not result in adequate testing of the model and does not validate the
scaled in vitro metabolic parameters. Additionally, there are currently no blood or tissue time-course
concentration data available for model validation. Therefore, as the model is currently constructed, the
PBPK model for chloroprene is inadequate for application for calculation of internal dose metrics or
interspecies dosimetry extrapolations.
21
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL CONTROLS
Potential for human exposure to chloroprene primarily is via inhalation and perhaps by the
dermal route. This section summarizes studies in occupationally exposed populations published from
1978 to 2008.
4.1.1. Chloroprene Exposure and Cancer Effects
4.1.1.1. Overview
The NTP (1998, 042076: 2005, 093207) described chloroprene as reasonably anticipated to be
a human carcinogen based on evidence of benign and malignant tumor formation at multiple sites in
animals. Evidence in humans for the carcinogenicity was reported to be limited based on consideration
of only two occupational epidemiological studies by Pell (1978, 064957) and Li et al. (1989, 625181).
Rice and Boffetta (2001, 624894) briefly examined evidence from five epidemiologic studies
(Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Colonna and Laydevant, 2001,
625112: Li et al., 1989, 625181: Pell, 1978, 064957). Although several of these earlier
epidemiological studies noted suggestive evidence of an association between chloroprene exposure and
liver cancer risk, study limitations included possible bias from cohort enumeration, follow-up, and
choice of reference population. Other study limitations noted included limited exposure assessment
data, low statistical power and the possible confounding by unmeasured co-exposures (Rice and
Boffetta, 2001, 624894). To date, there have been nine occupational epidemiological studies
conducted covering eight cohorts. It is important to note that where different studies investigated the
same cohort (as with Leet and Selevan (1982, 094970): and Marsh et al. (2007, 625188). which
investigated the Louisville Works cohort), differences in cohort recruitment, follow-up time, and
exposure ascertainment were deemed sufficient to present those study findings independently. This
epidemiological database is reviewed in the following section.
4.1.1.2. Individual Occupational Studies
Pell (1978, 064957) conducted a cohort mortality study in two neoprene (polychloroprene)
manufacturing plants of DuPont. The first cohort ("Louisville Works Cohort") consisted of 1,576 male
workers identified from a roster of wage roll employees in 1957. All workers who were exposed to
chloroprene were followed through December 31, 1974, accruing 26,939 person-years. Workers
terminated before June 30, 1957, were excluded and 17 individuals were lost to follow-up. Causes of
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
22
-------
death were obtained from death certificates and coded according to the 7th and 8th revised editions of
the "International Classification of Diseases" (ICD). Worker exposures to chloroprene were classified
qualitatively as "high," "moderate," "low," and "varied" based on job description. Statistical analyses
were performed using Poisson probability distribution with statistical significance level at p < 0.05.
The general U.S. male population and all male DuPont wage roll employees were used as external and
internal comparison populations, respectively. The study's primary objective was to examine
respiratory system cancer mortality, but mortality from other site-specific cancers was also evaluated.
Among the 193 deaths detected in this cohort, 51 were due to cancer and 16 of those deaths
were due to cancer of the respiratory system. Compared to U.S. rates, the standardized mortality ratios
(SMRs) for all-cause mortality, total cancer mortality and respiratory system cancer mortality were
69.0, 96.6, and 98.4, respectively. Based on the internal comparison, SMRs of 114.0 were detected for
total cancer mortality and 109.6 for respiratory system cancer mortality. The internal comparison
yielded SMRs of 108.7 (15 cases) and 113.2 (12 cases) for respiratory cancer after 15- and 20-year
latency periods, respectively. SMRs were lower for the same latency periods when compared with the
U.S. general population. Thirteen of the 16 deaths due to respiratory system cancer occurred in
smokers, while smoking history was unknown for the other three. Analyses by high-exposure
occupation did not show any significant change in SMRs or any statistically significant trend when
analyzed by years since first exposure. Other cancer deaths that were detected included 19 of the
digestive organs (SMR = 142.9 using an internal comparison) and seven of the lymphatic and
hematopoietic tissues (SMR = 155.6 using an internal comparison). All the SMRs observed in this
study were not statistically significant based on either internal (DuPont) or U.S. general population
mortality rates.
These data were reanalyzed by the National Institute for Occupational Safety and Health
(NIOSH) using a modified life-table analysis (Leet and Selevan, 1982, 094970). Workers were
classified into high and low-exposure categories based on a classification scheme developed by an
industrial hygienist who worked at the plant. Eight hundred and fifty-one workers were allocated to
the high-exposure group and 823 to the low-exposure group, with some workers contributing
person-years in both categories when their exposures or job titles changed. A total of 26,304
person-years were accrued, with 13,606 person-years in the high-exposure and 12,644 in the low-
exposure category. Compared to U.S. population rates, the overall SMR for the total cohort was 79.
Excess deaths were observed for cancers of the digestive system (especially the biliary passages and
liver), the lung, and the lymphatic/hematopoietic system. The only statistically significant SMR, of the
biliary passage and liver, was based on four cases, three from the high-exposure category (Table 4-1).
Of these three deaths, one was due to liver cancer, and the other two to gall bladder cancer. Cancer
mortality data were analyzed with respect to latency and duration of exposures stratified into 10-year
intervals. Statistically significant trends were not observed in either the latency analysis or the years of
presumed chloroprene exposure analysis, but these analyses were based on small numbers.
23
-------
The main limitations of the Pell (1978, 064957) study and the NIOSH reanalysis (Leet and
Selevan, 1982, 094970) include absence of quantitative exposure information and a lack of data on
smoking history and other potential risk factors which precluded further consideration. Exclusion of
workers terminated prior to June 30, 1957, might have resulted in some unidentified cancer deaths that
could have been associated with earlier higher exposures. Moreover, as pointed out by Leet and
Selevan (1982, 094970), the statistical power of the study to detect a significant excess in mortality
was low when the sub-cohort analyses were conducted.
Table 4-1. Standardized mortality ratios (SMRs) for the DuPont Louisville Works
cohort relative to general U.S. population rates
Cause Of Death
All Causes
All Cancers
Digestive
Biliary /liver
Trachea, bronchus, lung
Lymphatic, hematopoietic
Total Cohort
Cases, SMR (95% CF)
193,79(68-91)
51, 107(80-141)
19, 145 (87-227)
4, 571 (156-1463)
17, 106 (62-170)
7, 140 (56-288)
Low-Exposure
Cases, SMR (95% CI)
102, 82 (67-100)
26, 107 (70-157)
11, 164 (82-294)
1, " (-, ")b
7, 86 (35-178)
3, 120(25-351)
High-Exposure
Cases, SMR (95% CI)
91,75(61-92)
25, 107 (69-158)
8, 125 (54-246)
3, 750 (155-2192)
10, 128 (61-236)
4, 160 (44^10)
aCI = confidence interval.
bSMRs were calculated only if the observed number of deaths was greater than one.
Source: Leet and Selevan (1982, 094970).
Pell (1978, 064957) evaluated a second cohort in New Jersey that originally consisted of 270
males ("Chamber Works Cohort") believed to be exposed between 1931 to 1948 in a neoprene
manufacturing facility and followed through December 31, 1974. Follow up was complete for 240
workers. Since historical records were not complete for this cohort, efforts were made to assess
exposures for former employees based largely on memory recall of other employees. The observation
period, during which latency in tumor induction could be analyzed, was 30-40 years from date of first
exposure. Examination of mortality following a long latency period was considered a strength of this
study.
A total of 55 deaths was observed in this cohort. Study exclusions included thirteen deaths
occurring prior to 1957 (the starting point of observation assuming a 15-year latency period) and three
deaths occurring due to heart disease and malignant melanoma among former laboratory personnel
who had little or no exposure. The 39 observed deaths that occurred from 1957 to 1974 were slightly
more than the 37.7 expected using the DuPont comparison population. The 12 observed cancer deaths
were also higher than expected (SMR = 140) but the SMR was not statistically significant. There were
three deaths due to digestive cancer compared to 2.7 expected and four deaths due to lung cancer
compared to 3.0 expected. With five observed cancers of the urinary system (3 bladder and 2 kidney),
the SMR was significantly elevated compared to the DuPont population (SMR = 300; p < 0.01) and
24
-------
compared to the U.S. general population (SMR = 250; p < 0.01). The authors attributed the bladder
cancers to beta-naphthylamine exposure. Biliary and liver cancers were not examined in this study.
Small cohort size, low statistical power, and lack of quantitative exposure data were limitations of this
analysis.
Li et al. (1989, 625181) conducted a cohort mortality study of Chinese employees who worked
in one of three shops with chloroprene exposure (a chloroprene monomer workshop, a neoprene
workshop, and a laboratory) within a larger chemical plant. A cohort of 1,258 employees who had
accrued at least one year of chloroprene-related work prior to June 30, 1980, was identified from an
employee roster. The follow-up period for cancer deaths was from July 1, 1969, through June 30,
1983. Cancer mortality was assessed by searching the death registries at the plant's hospital and the
police substation; cancer diagnoses were verified by review of medical records at the city general
hospitals and cancer hospitals. Exposures were assigned to occupations based upon measured
concentrations in air at work sites and duration of exposure at different sites. When these levels were
not available, exposures were estimated through interviews with workers and administrators. Exposure
assignments took into account movement between exposure areas and were designed to roughly
represent time-weighted average exposure values. Follow up was achieved for 1,213 (96%) cohort
members (955 males and 258 females) and SMRs were calculated using sex- and age-specific
mortality in the local area. A total of 721 (75%) males and 131 (51%) females were exposed for more
than 15 years, while 131 (14%) males and 9 (3%) females were exposed for more than 25 years.
Males had statistically significant (p < 0.005) greater exposure to chloroprene than females based on
>15 years and >25 years of exposure.
Person-years were computed by 5-year categories for the total cohort and for the subgroups
(Table 4-2) starting from July 1, 1969 or when the individual first started working with chloroprene
through June 30, 1983 for live individuals or until their dates of death.l SMRs were calculated using
sex- and age- specific local area rates in 1973-1975. The results presented in Table 4-2 are for male
workers only as all sixteen reported cancer deaths occurred among male workers. The all-cancer SMR
for the male workers was 271 (p < 0.01). Among the 955 males, 464 (49%) were employed in
occupations with high exposures such as maintenance mechanics and monomer/polymer operators.
The SMRs for male workers in several high-exposure areas were statistically significant for liver and
lung cancer mortality. An increased SMR for liver cancer was observed, with four deaths occurring
among monomer workers and two deaths occurring in maintenance mechanics in the neoprene
workshop. Half of the cancers in the monomer shop were primary liver cancers (4 observed, SMR =
482, p < 0.01), with two occurring among the maintenance mechanics (SMR = 1667, p < 0.05).
Person-years accrued were not reported in the paper.
25
-------
Table 4-2. Standardized mortality ratios (SMRs) for all cancers, liver and lung
cancer among males exposed to chloroprene relative to general Chinese population
rates
Exposure Area
Total cohort
Monomer workshop
Vinylacetylene operator
Monomer operator0
Maintenance mechanic0
Neoprene workshop
Polymer operator0
Final treatment
Maintenance mechanic0
Laboratory
Quality monitor0
Researcher
Number Of Deaths/SMR
All Cause
16/27 la
8/377
O/—
4/450b
4/l,290b
5/176
5/3 94b
07—
O/—
3/319
2/l,176b
1/129
Liver Cancer
6/242
4/482b
O/—
2/465
2/l,667b
2/165
2/357
07—
07—
07—
Lung Cancer
2/513
1/714
07—
O/—
l/5,000b
1/556
1/1,250
07—
07—
O/—
""Statistical significance p < 0.01.
bStatistical significance p < 0.05.
°High-exposure area.
Source: Li et al. (1989, 625181: 1990, 644113).
One limitation of the Li et al. (1989, 625181) study was the availability of only three years
(1973-1975) of local area data to calculate SMRs. If these years were not representative of the entire
study period, then the SMRs could be biased. For example, if the general population experienced
higher mortality during the time periods not examined (i.e., 1969-1972 and 1976-1983) then the
SMRs reported in the study would be overestimated due to a lower expected number of deaths. If
mortality were lower during the other time periods not examined, then the reported SMRs would be
overestimated. Lack of quantitative exposure information precluded conducting internal analyses by
latency or duration of exposure. Additionally, there were no data on alcohol use or smoking history
and limited information was available on other potential confounders such as co-exposures to
chloroprene oligomers. The authors did consider potential confounding exposures due to benzene and
anti-ager D (N-phenyl-Z-naphthylamine) but determined that these exposures were limited and not
likely to influence the results. The authors also noted that the chemical plant investigated in the study
used the acetylene process for chloroprene manufacture, and therefore there was no possibility of co-
exposure to l,4-dichloro-2-butene, which is only produced as a by-product using the butadiene process
of chloroprene manufacture.
Li et al. (1989, 625181) also conducted a case-control study for the entire plant. Of 55
observed cancer deaths, 54 were matched with the same number of noncancer deaths among plant
workers based upon gender, age (± 2 years) and date of death (± 2 years). The authors observed that
26
-------
16 of the cancer deaths (30%) were among workers exposed to chloroprene compared to only four of
the noncancer deaths (7%), yielding an odds ratio of 13 (p < 0.005). Although the average age at death
was 12.7 years earlier for the exposed cancer cases relative to the unexposed cancer cases (p < 0.001),
these findings are limited by lack of data on co-exposures and other potential confounders.
Bulbulyan et al. (1998, 625105) examined cancer mortality at a Moscow shoe factory with
exposures to chloroprene from glue and from polychloroprene latex (a colloidal suspension of
poly chloroprene in water). The cohort consisted of 5,185 workers (4,569 women and 616 men)
employed for at least two years during 1960-1976 at specific production departments (i.e., cutting,
fitting, lasting and making, and finishing). Auxiliary departments and management employees were
excluded. Work histories were obtained from the personnel department, and subjects were assigned
exposure levels based on department and job; industrial hygiene measurements of exposure levels were
conducted in the 1970s. The authors provided detailed exposure data by job and department, ranging
from a high of 20 mg/m3 (gluers in the finishing department) to an intermediate level of 0.4-1 mg/m3
(all other jobs in the finishing department and all jobs in the lasting and making department) to the
unexposed (all jobs in the cutting and fitting departments).
The authors concluded that the industrial hygiene data were not systematic enough to assign
quantitative exposures to each worker since the collection of samples varied by location and by
different years. They therefore devised a relative scoring system to assign exposures: workers in the
high-exposure departments were assigned a level of 10, intermediate-exposure - a level of 1, and
unexposed - a level of 0. Cumulative exposures for individual workers were calculated by multiplying
years of exposure by the level of exposure, taking into account changes in job and department. In
addition, workers were classified by their highest exposure category. The authors considered
confounding exposures, including benzene exposures (6-20 ppm) in the high polychloroprene exposure
group during the 1950s, but did not adjust for those exposures in their analysis.
Mortality follow up was conducted from 1979 to 1993 which included 70,328 (62,492 in
females and 7,836 in males) person-years of observation. Thirty-seven percent of cohort members
(female/male distribution not provided) contributing 26,063 person-years were unexposed. Death
certificates were acquired from the National Registry Office Card Index and causes of deaths were
classified using ICD-9. Mortality rates of the general population of Moscow were used for
comparison. For the general population, mortality data for five cancers (liver, kidney, bladder,
pancreas, and malignant neoplasm of mediastinum and rhabdomyosarcoma of the heart) were available
only for 1992-1993. Therefore, the rate of expected deaths among these sites during 1992-1993 was
applied to the entire cohort for the entire period of observation. A Poisson distribution was used to
calculate the 95% CIs. One hundred thirty-one workers (2.5%) were lost to follow up. SMRs were
calculated for the entire cohort and separately for females and males. Among the total cohort, SMRs
were statistically significantly elevated for all cancers, liver cancer and leukemia (Table 4-3). SMRs
for liver cancer and leukemia were statistically significant in females but not in males, while the SMR
27
-------
for lung cancer was significant in males only. The authors suggested that the significant finding for
lung cancer was unlikely to be related to chloroprene exposure.
Table 4-3. Standardized mortality ratios (SMRs) for selected cancer risks relative
to general population rates of Moscow, Russia
Cause Of Death
All causes
All cancers
Liver cancer
Lung cancer
Leukemia
Total Cohort
Cases; SMR (95% CI)
900; 103 (97-110)
265; 122a (107-137)
10; 240a (110-430)
31; 140(90-200)
13; 190a (100-330)
Men
Cases; SMR (95% CI)
181; 121a (104-140)
56; 158a (119-205)
2; 240 (30-860)
17; 170a (100-270)
2; 190 (20-700)
Women
Cases; SMR (95% CI)
719; 100 (93-107)
209; 115a (100-131)
8; 230a (100-460)
14; 110 (60-190)
11; 190a (100-350)
""Statistical significance p < 0.05.
Source: Used with permission from Springer Netherlands, Bulbulyan et al. (1998, 625105).
Internal relative risk (RR) analyses (controlling for gender, age, and calendar period) were
conducted for selected cancers by using multivariate Poisson regression models, with trends evaluated
with the Mantel-extension test. Estimates for liver cancer were relatively imprecise since only one
liver cancer death was observed in the no-exposure category (a low number since this category
included 29% of all observed deaths). Stratified analyses by gender were not reported. Internal
analyses comparing the high-exposure group to the unexposed resulted in statistically significant RRs
for all causes of death (Table 4-4). Although they were not statistically significant largely due to a
small number of cases, elevated RRs ranging from 2.2-4.9 were detected for leukemia and cancers of
the liver, kidney and colon.
Table 4-4. Selected relative risk (RR) estimates for the high-exposure group
relative to unexposed factory workers
Cause Of Death
All causes
Liver cancer
Colon cancer
Kidney cancer
Leukemia
High-Exposure Deaths
194
3
8
2
5
High-Exposure RR (95% CI)a
1.23b (1.02-1.49)
4.9 (0.5-47)
2.6 (0.8-7.9)
3.3 (0.3-37)
2.2 (0.6-8.4)
""Reference group is defined as workers with no chloroprene exposure.
bStatistical significance p < 0.05.
RR = relative risk
Source: Used with permission from Springer Netherlands, Bulbulyan et al. (1998, 625105).
28
-------
Although there were only a few deaths in each group, analysis by categories of duration of
employment among workers with the highest exposure to chloroprene (1-9 years, 10-19 years,
20+ years) relative to no exposure showed a significant trend (p = 0.02) for liver cancer but not for
leukemia mortality (Table 4-5).
The cumulative exposure analysis indicated an increased risk of liver cancer mortality based
upon six deaths in the intermediate-exposure category (10.1-30 unit-years, RR = 7.1, 95% CI: 0.8-61)
and three deaths in the highest exposure category (30.1+ unit years, RR = 4.4, 95% CI: 0.4-44).
Kidney cancer was increased in all cumulative exposure categories but none of the RRs were
statistically significant and no overall trend was observed.
Table 4-5. Internal relative risks (RRs) by duration of employment in the high-
exposure category
Cause Of Death
Liver cancer
Leukemia
1-9 Years
Cases; RR (95% CI)
1; 2.7 (0.2-45)
2; 1.3 (0.2-7.3)
10-19 Years
Cases; RR (95% CI)
1; 8.3 (0.5-141)
2; 3.4 (0.6-19)
20+ Years
Cases; RR (95% CI)
1; 45.0 (2.2-903)
1; 8.8 (0.7-66)
Trend
p = 0.02
p = 0.07
Source: Used with permission from Springer Netherlands, Bulbulyan et al. (1998, 625105)
The most prominent finding in the Bulbulyan et al. (1998, 625105) cohort was 10 deaths
occurring from liver cancer. The authors also detected 11 deaths (3 in males and 8 in females) due to
cirrhosis, a precursor of primary liver cancer, but did not adjust for this as a potential confounder.
Increased mortality due to leukemia was observed in all categories for both cumulative exposure and
duration of employment (with high exposure) but neither trend was statistically significant. The
authors suspected a causal role of chloroprene in the leukemia deaths but could not rule out a possible
role of exposure to benzene. A significant increase in lung cancer was observed among males only,
which may have been due to confounding by smoking. Potential confounding by smoking could not
be examined due to lack of data for this cohort. Pancreatic cancer, which may be smoking related, was
also observed in males only. No excess risk for lung cancer was observed in females or in the total
cohort. Lack of precise quantitative exposure information, no adjustment for confounding risk factors,
and exclusion of deaths prior to 1979 resulting in relatively low statistical power were some of the
limitations of this study. Similar to the Li et al. study (1989, 625181), the minimal data on observed
deaths for some cancers among the general population may have also resulted in biased SMR values if
mortality during these years was not representative of mortality during the entire study period.
Bulbulyan et al. (1999, 157419) conducted a retrospective cohort study of 2,314 workers (1,897
males, 417 females) who had been employed in production departments of a chloroprene monomer
production plant in Yerevan, Armenia, for at least two months between 1940 and 1988 and were alive
29
-------
as of 1979. Mortality was followed from 1979 to 1988, and vital status was accessed through the
Yerevan Address Bureau. Death certificates were coded by using the ICD-9 revision. Sixty-three
individuals (3%) were lost to follow-up. Industrial hygiene exposure measurements of chloroprene
were available both before and after 1980, when production changes led to a dramatic decrease in
exposures. Before 1980, exposures averaged 5.59-69.80 mg/m3 (1.54-19.3 ppm) during the summer
and 2.30-249.5 mg/m3 (0.63-68.9 ppm) during the winter. After 1980 the summer average ranged
from 0.80-3.60 mg/m3 (0.22-0.99 ppm) and concentrations ranged from 0.55-2.10 mg/m3 (0.15-
0.58 ppm) for the winter. Work histories were obtained from the personnel department, including the
start and end of each job, and from the departments of employment. Relative exposure values were
assigned based on either high exposure (production operators: six units before 1980, three units after
1980) or low exposure (other production workers: two units before 1980 and one unit after 1980).
Unexposed workers were assigned a relative exposure value score of zero. SMRs and standardized
incidence ratios (SIRs) were calculated based on comparison rates for the entire Armenian population,
and 95% CIs were also calculated by using a Poisson distribution assumption. Internal RR estimates
were calculated by using multivariate Poisson regression models and adjusting for age, calendar
period, and gender.
A total of 21,107 person-years were contributed by the study population. There were 20 deaths
during the observation period with four due to stomach cancers and three each resulting from liver and
lung cancers. The SMR was statistically significant for liver cancer only (SMR = 339, 95% CI:
109-1,050). Two liver and two lung cancer deaths were identified among males, while one liver cancer
death and one lung cancer death were identified in females. No internal comparisons were included in
the SMR analysis. Cancer incidence data were available for 1979-1990 through the Armenian Cancer
Registry. Several types of cancers (37 cases) were identified, with six liver and six lung cancers (five
each in males) being the most prevalent (Table 4-6). The SIRs for liver cancer were statistically
significant for the total cohort (SIR = 327, 95% CI: 147-727) and for males (SIR = 303, 95% CI:
126-727) when stratified by gender. SIRs below 100 were observed for lung cancer in both the total
cohort as well as among males only.
Table 4-6. Selected standardized incidence ratios (SIRs) for chloroprene monomer
cohort relative to the general Armenian population
Cancer Type
All cancers
Lung cancer
Liver cancer
Observed
37
6
6
SIR (95% CI)
68 (49-94)
53 (24-119)
327a (147-727)
""Statistical significance p < 0.05.
Source: Used with permission from Wiley-Liss, Inc., Bulbulyan et al. (1999, 157419)
30
-------
Internal trend analyses of plant workers showed increasing incidence of liver cancer by
duration of employment with a statistically significant relative risk among chloroprene production
workers who were employed for more than 20 years (4 cases, SIR = 345, 95% CI: 129-920).
Evaluation of liver cancer incidence by duration of employment (<1 year, 1-9 years and 10+ years) in
the high chloroprene exposure groups resulted in a statistically significant SIR in the 10+ years
category (SIR = 612, 95% CI: 230-1,630). Similar findings were noted in analyses using cumulative
exposure,(unit-years) with a statistically significant SIR of 486 (95% CI: 202-1,170) among the five
cases in the highest cumulative exposure category of 40+ units. All six cases of liver cancer in this
study occurred among highly exposed operators. These internal analyses suggest a possible dose-
response relationship between chloroprene exposure and liver cancer incidence.
The authors discussed the strong healthy worker effect observed in this study. In particular,
they suggested that the low SMRs might be due, in part, to potential loss of early cases resulting from
not beginning the follow-up period until 1979. In addition to the incomplete enumeration of health
outcomes among the workers, the authors acknowledged that misclassification might have also
occurred due to incomplete registration of liver cancers in the Armenian registry. Furthermore,
although measurements of chloroprene levels were available, investigators were unable to develop
quantitative estimates and assigned exposure units to the workers depending upon their job description.
The role of potential confounding by alcohol use and smoking could not be examined due to lack of
data. The high incidence (27 in males and 5 in females) of liver cirrhosis, a precursor for liver cancer,
is an unlikely confounder as it is likely an intermediate in the causal pathway precluding statistical
adjustment. There was also little evidence that several other co-exposures (i.e., vinyl acetate, toluidine,
talc, and mercaptans) that were not adjusted for in either the mortality or incidence analyses are liver
carcinogens.
Romazini et al. (1992, 624896) investigated cancer mortality in a retrospective French cohort
study of 660. French chloroprene polymer manufacturing workers (599 males, 61 females) employed
for at least two years at a polychloroprene plant. The follow-up period was from 1966-1989 with 32
observed deaths included in the study; an additional 18 potential study subjects were lost to follow up.
No excess mortality was observed compared to regional rates. In a nested case-control study
comparing era of employment, the authors found that workers exposed to conditions prior to 1977 had
a much higher risk of death compared to those exposed to chloroprene after 1977(odds ratio = 5.34;
95% CI: 1.28-22.3). Similar to other studies, the small size of this cohort and inability to control for
smoking and other potential confounders limited the conclusions that could be drawn from this study.
Colonna and Laydevant (2001, 625112) conducted a cohort cancer incidence study among 533
males who worked a chloroprene production plant in Isere, France, for at least two years between
January, 1966 (when the plant opened) and December, 1997. Cancer incidence cases were traced
through the Isere cancer registry from 1979 (when the registry was founded) through 1997. Workers
who died before 1979 or who left the area were not traced (the number of untraced incident cancers
was not estimated). Work histories were collected and jobs were classified into low, intermediate, and
31
-------
high chloroprene exposure groups based on estimated exposures of <2 ppm, 2-5 ppm, and >5 ppm
respectively. Exposure duration was divided into three groups of < 10 years, 11-20 years and
>20 years. The cohort was divided into two groups, workers employed prior to 1977 and those
employed in 1977 or later, based on lower anticipated exposures following significant changes in
worker protection. SIRs were calculated using the general population rates of Isere as a reference and
confidence intervals were calculated using a Poisson distribution.
A total of 7,950 person-years were accrued. Of the 34 incident cancers, 32 occurred in the
group employed prior to 1977. There were nine lung cancers, nine cancers of the head and neck
(including three laryngeal cancers), and one liver cancer. SIRs were calculated for various cancers
including those occurring in the head and neck, larynx, lung, liver and colon/rectum (Table 4-7). With
the exception of colon/rectum, all of the SIRs exceeded 100 with most of the cases and higher SIRs
noted for earlier periods of first employment (i.e., before 1977).
Table 4-7. Standardized incidence ratios (SIRs) for elevated cancer risks for plant
workers relative to general population rates of Isere, France
Cancer Type
All Cancers
Head and Neck
Larynx
Lung
Liver
Colon/Rectal
Total Cohort
Cases; SIR (95% CI)
34; 126 (88-177)
9; 189 (87-359)
3; 243 (50-713)
9; 184 (84-349)
1; 136 (4-763)
2; 66 (8-239)
Cohort Exposed Before 1977
Cases; SIR (95% CI)
32; 146a (100-206)
8; 209 a (90-4 11)
3; 297 (61-868)
8; 199 a (86-391)
1; 164 (5-913)
2; 79 (10-287)
Source: Used with permission from Elsevier Science Ireland Ltd., Colonna and Laydevant (2001, 625112).
Although none of the SIRs were statistically significant, a trend was observed when the data
were analyzed by duration of exposure. Five lung cancers were reported in workers with >20 years of
exposure (SIR = 257, 95% CI: 84-602), 3 in those with 11-20 years exposure (SIR = 149, 95% CI:
31-436) and 1 in those with < 10 years exposure (SIR = 106, 95% CI: 30-586). No significant
excesses were observed in head and neck cancer by duration of exposure. No trend was detected for
lung cancer incidence in relation to intensity of exposure with SIRs of 463 (95% CI: 127-1,191), 125
(95% CI: 15-451), and 123 (95% CI: 26-361) reported for the low-, intermediate- and high-exposure
categories, respectively.
Increased lung cancer and laryngeal cancer were observed in this study. Given that smoking is
strongly associated with lung cancer, and since seven of the eight lung cancer cases were smokers, the
investigators concluded that the lung cancer excess was unlikely to be due to chloroprene exposure.
Although smoking and alcohol consumption were discussed as strongly associated with laryngeal
32
-------
cancer, no additional information was provided in the paper. This study found only one incident of
liver cancer but noted that liver cancer incidence was likely under-estimated due to difficulties in case
enumeration. Study limitations included lack of precise quantitative exposure information, low cancer
incidence, and reduced power because of elimination of workers who had died or left the area prior to
1979.
More recently, Marsh et al. (2007, 625187) evaluated mortality patterns of four chloroprene
production facilities by using external regional rates and internal comparisons (Marsh et al., 2007,
625188). This study attempted to address the problems identified with earlier studies by conducting a
detailed exposure assessment for both chloroprene and a potential confounding co-exposure, vinyl
chloride monomer (Esmen et al., 2007, 625114; Esmen et al., 2007, 625118; Esmen et al., 2007,
625121; Hall et al., 2007, 625243). As described in detail by Esmen et al. (2007, 625121). a historical
review of processes at all four plants led to the assignment of exposures to 257 unique tasks. Taking
into account shared tasks or rotation between tasks, job title-based exposures to chloroprene were
assigned to one of seven categories, including unexposed (<0.0005 ppm). Vinyl chloride exposures
were assigned to one of five categories, including unexposed (<0.01 ppm) (Esmen et al., 2007,
625118).
Two of the facilities evaluated were in the U.S.-DuPont/Dow plants at Louisville (L),
Kentucky and Pontchartrain (P), Louisiana. The third facility was the Maydown (M) plant in Northern
Ireland, and the fourth facility was the Enichem Elastomer plant in Grenoble (G), France. These plant
cohorts included all employees with possible chloroprene exposure from plant start-up through 2000:
5,507 workers (L), 1,357 workers (P), 4,849 workers (M), and 717 workers (G). Median cumulative
exposures to chloroprene at these plants were 18.35 (L), 0.13 (P), 0.084 (M), and 1.01 (G) ppm-years.
The median average intensity of chloroprene exposure (in ppm) at these plants were: 5.23 (L), 0.0283
(P), 0.160 (M), and 0.149 (G). Vinyl chloride exposures occurred at only two plants, Louisville and
Maydown. Their median cumulative vinyl chloride exposures were 1.54 and 0.094 ppm-years,
respectively. The median average intensity of vinyl chloride exposures were 1.54 and 0.030 ppm,
respectively.
The study period for the cohorts encompassed 52 (L), 41(M), 39 (P), and 34 (G) years resulting
in 197,919 (L), 127,036 (M), 30,660 (P), and 17,057 (G) person-years (Marsh et al., 2007, 625187).
Vital status was assessed using several different sources. A trained nosologist using the ICD codes in
effect at the time of death coded the underlying cause of death. A total of 3,002 deaths had occurred
during the follow-up period in the chloroprene cohorts and cause of death was ascertained for 2,850
individuals (95%). A modified Occupational Cohort Mortality Program was used to conduct statistical
analyses. Independent analyses were conducted for the four facilities for total cancer deaths and
certain site-specific deaths. Person-years at risk were computed for each individual by race, sex, age
group, calendar time, duration of employment, and the time since first employment. SMRs and 95%
CIs were calculated for the total cohort and selected sub-cohorts for each plant.
33
-------
All cause mortality was significantly reduced (compared to local county rates) for each of the
four cohorts (Table 4-8). In addition, each cohort had significantly reduced mortality for all cancers,
and the largest cohort, Louisville, had significantly reduced mortality from respiratory cancers. The
total number of cancer deaths observed at each of the four plants was 652 (L), 128 (M), 34 (P), and
20 (G). Reported respiratory cancer deaths (including bronchus, trachea, and lung) were 266 (L), 48
(M), 12 (P), and 10 (G), while liver cancer deaths were 17 (L), 1 (M), 0 (P), and 1 (G) for each plant.
Compared to the local population rates, fewer deaths than expected from liver cancer were observed in
the Louisville (SMR = 90, 95% CI: 53-144) cohort than expected. All other sites had no more than one
death due to liver cancer. Similar to the healthy worker effect observed in other studies, fewer cancer
deaths were reported in the occupational cohorts compared to general population estimates. An
additional paper by this group (Leonard et al., 2007, 625179) further explored the healthy worker
effect in an analysis of the Louisville and Pontchartrain workers. Compared to the local county
population estimates, SMRs were decreased for all cancers, respiratory cancers, and liver cancers.
However, when comparisons were based on DuPont national and DuPont Region 1 comparison
populations (in order to control for the healthy worker effect), the authors found statistically significant
elevated risks for: all cancers, SMR =111 (DuPont national population only); and respiratory cancer
mortality, SMRs =137 (DuPont national population) and 120 (DuPont Region 1 population). Elevated
SMRs were observed for liver cancer, SMRs = 127 (DuPont national population) and 121 (DuPont
Region 1 population), although these liver cancer risks were smaller than reported in other studies and
were nonsignificant.
34
-------
Table 4-8. Standardized mortality ratios (SMRs) at each of four chloroprene
production facilities
Cause Of
Death
All Causes
All Cancers
Respiratory
Cancers
All Cancers:
Exposed
Unexposed
Louisville
(L) Cases,
SMR
(95% CI)a
2,403
74c(71-77)
652
75 (69-80)
266
75 (66-85)
651
74C (69-80)
1
99(3-551)
Maydown
(M) Cases,
SMR
(95% CI)b
435
60 (55-67)
128
68 (56-80)
48
79 (58-105)
114
62c(51-75)
14
126 (69-212)
Pontchartrain
(P) Cases, SMR
(95% CI)a
102
53 (43-65)
34
68 (47-95)
12
62 (32-109)
26
57C (37-84)
8
144 (62-285)
Grenoble
(G) Cases,
SMR
(95% CI)b
62
65 (50-83)
20
59 (36-91)
10
85 (41-156)
15
59d (33-97)
5
61 (20-142)
Total
Cases, SMR
(95% CI)
3,002
70 (67-73)
834
73 (68-78)
336
75 (68-74)
806
7 1° (66-76)
28
108 (72-156)
aLocal county comparisons.
^National comparisons.
cp<0.01
dp < 0.05
Source: Used with permission from Elsevier Science Ireland Ltd., Marsh et al. (2007, 625187)
When chloroprene exposed and unexposed workers were analyzed separately in this cohort, the
SMRs for all cancers were all significantly reduced for exposed workers at each plant, while they were
generally higher (at or above expected levels for all plants except at Grenoble) for unexposed workers
(Marsh et al., 2007, 625187). The very small number of unexposed workers (n = 28) across all four
plants limits the conclusions that can be drawn based on the crude exposure classification approach
(Table 4-8). In their companion paper (Marsh et al., 2007, 625188), the authors conducted internal RR
analyses of more detailed worker exposure levels at each of these four plants. Exposure-response
trends across quartiles of exposure were examined using a forward stepwise regression modeling
approach to adjust for potential confounding. Analyses were conducted by considering 5- and 15-year
lagged exposures and using white/blue collar as a surrogate for lifetime smoking (due to an inability to
locate complete smoking histories for employees who died from respiratory cancers). Absolute
mortality rates were estimated by calculating exposure category-specific SMRs using external
mortality rates. The internal analyses for all cancers showed increasing RRs with duration of exposure
(<10, 10-19, 20+ years) to chloroprene in plants L and M, but a statistically significant trend
(p < 0.007) was only noted for Plant M. Relative to less than 10 years of exposure, increased RRs
were noted for 10-19 years (RR = 1.53; 95% CI: 1.00-2.34) and 20+ years (RR = 1.78; 95% CI:
1.11-2.84) of exposure. The external comparison consistently showed SMRs less than the internal
analysis (and mostly below 1) for both the plants suggestive of bias due to the healthy worker effect.
This was confirmed by the detection of higher SMRs for all cancer, respiratory cancer and liver cancer
35
-------
mortality in the Louisville and Pontchartrain cohorts based on DuPont national and DuPont Region 1
comparison populations (Leonard et al., 2007, 625179).
The internal analysis for liver cancer could only be conducted in the Louisville cohort, which
included 17 of the 19 observed deaths and also had the highest chloroprene levels (Marsh et al., 2007,
625188). Despite the limited number of deaths, these data show some potential evidence of a dose-
response effect across the four exposure levels (p = 0.09). Although the individual RRs were not
statistically significant, the RRs for the highest three exposure levels were 1.9 (95% CI: 0.21-23.81),
5.1 (95% CI: 0.88-54.64), and 3.3 (95% CI: 0.48-39.26).
As shown in Table 4-9, the results of the internal analyses for respiratory cancers at the three
plants (M, P, G) without worker status adjustment showed higher RRs with increasing cumulative
exposure (Marsh et al., 2007, 625188). The observed trends were not statistically significant but were
based on a small number of respiratory cancers. In contrast, the plant with the most cases (L) showed
little evidence of an exposure-response relationship. The investigators adjusted for the potential
confounding by smoking status in the analyses of lung cancer mortality at Louisville only (due to small
numbers at the other plants) using employment status as a surrogate of blue versus white collar
workers. This decision was justified by the authors based upon this variable being a surrogate for
variables associated with smoking such as education and socio-economic status. It is impossible,
however, to discern whether this surrogate resulted in control for smoking or resulted in an over-
adjustment since work status was so highly correlated with chloroprene exposures.
36
-------
Table 4-9. Relative risks (RRs) for respiratory cancers by cumulative chloroprene
exposure
Plant
Louisville (L)
May down
(M)
Pontchartrain
(P)
Grenoble (G)
Level la
(Lowest);N
62 ; Reference13
14; Reference13
3; Reference13
2; Reference13
Level 2
N; RR (95% CI)
67; 1.00 (0.71-1.43)
9; 1.65(0.66-4.15)
3; 1.60 (0.20-12.8)
1; 0.61 (0.05-6.76)
Level 3
N ; RR (95% CI)
77 ; 1.32 (0.94-1.88)
12; 1.89 (0.72-4.96)
2; 2.90 (0.20-34.1)
4; 2.87 (0.35-39.7)
Level 4
N ; RR (95% CI)
60 ; 0.85 (0.58-1.23)
13; 2.28 (0.86-6.01)
4; 2.32 (0.30-21.8)
3; 3. 14 (0.30^8.0)
Trend
p = 0.71
p = 0.10
p = 0.34
p = 0.17
aChloroprene exposure (in ppm years) levels varied by plant: L (<4.7 to >164.1); M (O.04 to >24.5); P (O.02 to
>16.2);G(<0.05to>23.9).
bReference is the concentration below the lowest concentration measured at each plant.
Source: Used with permission from Elsevier Science Ireland Ltd., Marsh et al. (2007, 625188).
The authors also conducted internal analyses of cancer mortality and vinyl chloride exposure
(the primary co-exposure in this study) at the Louisville plant (Marsh et al., 2007, 625188). They
found inverse associations (many of them statistically significant) between risk of both respiratory and
liver cancer in relation to vinyl chloride exposures; however, these associations were based on limited
numbers of cancer deaths in the vinyl chloride exposure groups. In fact, the vast majority of
respiratory and liver cancers occurred among workers who were unexposed to vinyl chloride. If vinyl
chloride is a negative confounder of the association between chloroprene and liver cancer, then the
reported association between chloroprene and liver cancer would be an underestimate of the
association adjusted for vinyl chloride. However, the authors reported that there was no correlation
between cumulative exposures to vinyl chloride and chloroprene among these workers. Given this, it
is highly unlikely that confounding by vinyl chloride could explain the associations observed between
chloroprene and these cancers.
The recent DuPont studies (Leonard et al., 2007, 625179: Marsh et al., 2007, 625187: Marsh et
al., 2007, 625188) represent some of the more comprehensive studies to date, largely due to exposure
assessment data which allowed for internal comparisons. Although the authors concluded that their
study provided no evidence of cancer risk associated with chloroprene exposures, there was some
evidence that this may in part be due to the healthy worker effect (Leonard et al., 2007, 625179). The
cancer specific findings suggest that the association between chloroprene exposure and liver cancer
mortality risk was smaller but comparable with other studies. There was also some suggestion of
elevated risk of respiratory cancer mortality at the upper two exposure levels in several of the cohorts
(Table 4-9). Although statistical power to detect mortality trends across exposure levels appeared
limited, the relative risks in the upper two exposure groups were all in excess of 1.8 relative to the
37
-------
unexposed populations with the exception of the Louisville plant (Marsh et al., 2007, 625188).
Despite study limitations, findings from this cohort add to the weight of evidence that chloroprene
exposure may be associated with cancer mortality especially when comparisons are based on internal
populations or other regional or national DuPont workers.
4.1.1.3. Summary and Discussion of Relevant Methodological Issues
Nine studies covering eight cohorts were reviewed to assess the relationship between exposure
to chloroprene and cancer incidence and mortality. Four cohorts had fewer than 1,000 workers, while
the remaining cohorts had fewer than 6,000. The most consistent finding was excess liver (Bulbulyan
et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Leet and Selevan, 1982, 094970: Li et al., 1989,
625181) and lung/respiratory system (Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419:
Colonna and Laydevant, 2001, 625112: Leet and Selevan, 1982, 094970: Marsh et al., 2007, 625188:
Pell, 1978, 064957) cancer incidence or mortality (Tables 4-10 and 4-11). The limitations of each of
the aforementioned studies are discussed in this section. Most occupational cohort studies are
historical in nature gathering human subject information from existing records and going back many
years. In general, the constructed databases do not include detailed information on the workers'
individual habits (e.g., tobacco use, alcohol consumption) or pre-existing disease status (i.e.,
hepatitis B infection), and usually only have limited exposure information. These limitations often
limit the ability to control for bias due to confounding variables and to assess the potential for
misclassification of exposure.
One of the limitations of the occupational epidemiologic studies examining chloroprene
exposure is the potential for the healthy worker effect to influence the results. Since occupational
studies involve workers who are healthier than the general population, a reduced mortality risk is often
observed among these populations when compared to external populations. This potential bias was
likely reduced in some studies by using internal comparisons or other study designs such as a nested
case-control study. Internal comparisons however may not completely eliminate the healthy worker
effect as the healthy worker survivor effect (e.g., shorter-term exposed workers having increased
mortality) can also lead to attenuation of effect measures (Arrighi and Hertz-Picciotto, 1994, 625164).
Another limitation of occupational cohort studies is the reliance on death certificates for
outcome ascertainment especially in the mortality studies. Although misclassification of cause of
death can be minimized by the review of medical records or by histological confirmation, this was not
done in any of the studies. Incomplete enumeration of incident cases was another limitation of several
of the studies. This may limit the ability to detect associations as it directly reduces statistical power
through reduced sample sizes. Outcome misclassification can also bias the measures of associations
that were examined. Since there is no direct evidence of substantial misclassification of health
outcomes in these studies, it is difficult to gauge the potential impact of this bias on the reported
findings.
38
-------
Finally, the lack of quantitative exposure assessment is clearly a limiting factor of most
occupational studies; however, they still are able to contribute to the overall qualitative weight of
evidence considerations. In many cases where exposure data were missing or insufficient to provide
quantitative assessments, exposure levels were differentiated based upon job titles and industrial
hygiene knowledge of the processes involved. Although measurement error is present in all studies to
varying degrees, there is no evidence that this error differed by outcome (i.e., was nondifferential) in
these studies. Although there are rare exceptions, nondifferential misclassification of workers'
exposures due to lack of information usually results in an underestimate of the association between
exposure and outcome.
Table 4-10. Epidemiologic summary results of respiratory system cancers:
Standardized mortality ratios (SMRs) and standardized incidence ratios (SIRs) for
the overall cohort populations relative to external comparison populations and
relative risks (RRs) for intermediate and high chloroprene exposures
Study3
Bulbulyan et al. (1998. 625105)
Rnlhnlvan pt Q! (\ QQQ 1 S7J.1 Q^
Colonna and Laydevant (2001, 625112)
Leet and Delevan (1982, 094970)
Marsh etal. (2007, 625187)-Louisville
Marsh et al. (2007, 625187): Marsh et al.
(2007. 625 188)-May down
Marsh et al. (2007, 625187): Marsh et al.
(2007, 625188)-Pontchartrain
Marsh et al. (2007, 625187): Marsh et al.
(2007, 625188)-Grenoble
Total Cohort
SMR/SIR
(95% CI)
140 (90-200)
•SO (\f\ \^\
184 (84-349)e
106 (62-170)
75 (66-85)
79 (58-105)
69 H9 109s!
QZ (A\ \^fC\
Intermediate-
Exposure
SMR/SIR/RRb
(95% CI)
1.0 (0.4-2.5) c'd
125 (15-451)e
86 (35-178/
92 (73-115)d
97 (50-169)d
%H9 348^
1 1Q n? ^04^d
High-Exposure
SMR/SIR /RRb
(95% CI)
0.8 (0.3-2.4)c'd
123 (26-361)e
128 (61-236)
65 (50-84)d
113 (60-192)d
85 f93 918^
170 of> ^7^d
aSMRs and SIRs calculated relative to external population rates and are reported on a 100-base scale, unless noted
all values are SMRs.
bRelative to low or unexposed groups.
'Relative risk of death from lung cancer.
Cumulative chloroprene exposures.
Standardized incidence ratios.
fLow-exposure group.
39
-------
Table 4-11. Epidemiologic summary results of liver/biliary passage cancers:
Standardized mortality ratios (SMRs) for the overall cohort populations relative to
external comparison populations and SMRs and relative risks (RRs) for
intermediate and high chloroprene exposures
Study
Bulbulyan et al. (1998, 6251051
Bulbulyan et al. (1999, 157419)
Colonna and Laydevant (2001, 625112)
Leet and Delevan (1982, 094970)
Li etal. (1989,625181)
Marsh et al. (2007, 625187): Marsh et al.
(2007, 625188) - Louisville
Marsh et al. (2007, 625187): Marsh et al.
(2007, 625 188)-May down
Marsh et al. (2007, 625187): Marsh et al.
(2007, 625188) - Pontchartrain
Total Cohort
SMRa (95% CI)
240(110-430)
339 (109-1,050)
136 C4 763^
571(156-1,463)
482f
90 C59 144s*
94 CI 134s*
Intermediate-
Exposure
SMR/RRb (95% CI)
7.1 (0.8-6 l)c'd
293 (41-2,080)d'e
250 (6-l,393)a
5.1(0.9, 54.5)c'd
High-Exposure
SMR/RRb (95%
CI)
4.4 (0.4^4)c'd
486 (202-1, HO)46
750 (155-2,192)a
3.3 (0.5, 39.3)c'd
aSMRs and SIRs calculated relative to external population rates and are reported on a 100-base scale, unless noted
all values are SMRs.
bRelative to low or unexposed groups.
'Relative risk of death from liver cancer.
dCumulative chloroprene exposures.
Standardized incidence ratio.
fNot reported, p< 0.01.
4.1.1.3.1. Lung Cancer Summary. An increased risk of lung cancer incidence and mortality was
observed in a few studies (Bulbulyan et al., 1998, 625105: Colonna and Laydevant, 2001, 625112:
Leonard et al., 2007, 625179: Li et al., 1989, 625181: Pell, 1978, 064957). although few statistically
significant associations were reported. None of the studies adjusted for smoking because the
investigators either did not have this information available or because the majority of their lung cancer
cases were observed in smokers. Marsh et al. (2007, 625188) used white/blue collar as a surrogate for
smoking habits assuming that blue collar workers smoked more than white collar workers. But due to
small number of deaths in white collar workers the authors reportedly only adjusted the lung cancer
risk for worker type in the Louisville, Kentucky, plant. Since worker pay type is a crude surrogate of
smoking status, it is difficult to rule out the potential confounding effects of smoking. Worker pay
status is also a marker of chloroprene exposure. Therefore, inclusion of this variable in regression
models may result in over-adjustment distorting the relationship between cancer mortality and
chloroprene exposure. A few studies noted higher SMRs for lung cancer among workers exposed to
chloroprene; however, there was not consistent evidence of an exposure-response relationship across
various chloroprene exposure categories.
40
-------
4.1.1.3.2. Liver Cancer Summary. Statistically significant excesses of liver cancers were detected in
four studies examining four cohorts (Bulbulyan et al., 1998, 625105; Bulbulyan et al., 1999, 157419;
Leet and Selevan, 1982, 094970; Li et al., 1989, 625181). Although no statistically significant increase
in the risk of liver cancer (compared to the general population) was detected when the Louisville
cohort was analyzed by Marsh et al. (2007, 625188), the SMRs for liver cancer mortality exceeded 120
when based on comparisons to national and regional DuPont worker populations (Leonard et al., 2007,
625179). The relative risk of liver cancer mortality also increased with increasing cumulative
exposures indicating a potential dose-response trend. In the French (Grenoble/Isere) cohort, there was
only one case of liver cancer or mortality from liver cancer (Colonna and Laydevant, 2001, 625112;
Marsh et al., 2007, 625187; Marsh et al., 2007, 625188) detected, while the Pontchartrain cohort study
had no reported liver cancer deaths (Marsh et al., 2007, 625188). The small numbers of liver cancer
deaths especially in the latter studies precluded further examination of the detailed exposure
information.
Confounding by occupational co-exposures is addressed in some studies but few of these
included direct adjustments for the possible confounders. Some studies have selected workers from
several different processes where the co-exposures might have been different or non-existent in some
processes to help address the potential for confounding. Bulbulyan et al. (1999, 157419) discussed
other possible exposures and concluded that confounding was unlikely, since none of the known co-
exposure chemicals were known to be associated with liver cancer. Marsh et al. (2007, 625188)
conducted a separate analysis with vinyl chloride in the Louisville plant and found that 15 out of 17
liver cancer cases were found in workers who were not exposed to vinyl chloride. The authors also
reported that there was no correlation between cumulative exposures to vinyl chloride and chloroprene
among these workers. Given these data, it is highly unlikely that confounding by vinyl chloride could
explain the association observed between chloroprene and an increased liver cancer risk.
No adjustments for known risk factors for liver cancer, such as alcohol consumption, were
performed in any of the cohorts observing statistically significant increases in liver cancer mortality. If
alcohol consumption was associated with chloroprene exposure, although unlikely, this might be a
source of residual confounding. Other risk factors for liver cancer that were not controlled for,
including hepatitis infection and aflatoxin ingestion, are not likely to be associated with chloroprene
exposure among these occupational cohorts. Although the lack of adjustment for these known risk
factors of liver cancer may be a cause of concern when considering the studies individually, the
consistent observation of increased liver/biliary cancer in multiple heterogeneous occupational cohorts
ameliorates this concern to some degree. Further limitations in these cohorts include the lack of
precise quantitative exposure information, limited statistical power to detect effects due to insufficient
general population mortality data, and incomplete ascertainment of health outcomes. Studies that
41
-------
relied upon comparisons to external population mortality rates are also susceptible to the healthy
worker effect although the potential impact on cancer mortality in these populations is unclear (above).
Primary liver cancer is relatively rare in the U.S. It accounts for approximately 1.3% of new
cancer cases and 2.6% of cancer deaths (Jemal et al., 2003, 625160). There are also few identified
chemicals that have been associated with primary liver cancer, so co-exposures are unlikely to
confound the association between chloroprene exposure and liver cancer mortality. The observation of
an increased risk of liver cancer mortality is fairly consistent and there is some suggestive evidence of
an exposure-response relationship among workers exposed to chloroprene in different cohorts on
different continents (i.e., U.S., China, Russia, and Armenia) (further discussion in Section 4.7.1.1.1 -
Biological Gradient).
4.1.2. Chloroprene Exposure and Noncancer Effects
4.1.2.1. Acute-, Short-, and Subchronic-Duration Noncancer Effects
Nystrom (1948, 003695) reported effects associated with the levels (not specified) of
chloroprene exposure experienced during the start-up of chloroprene production in Sweden. The
author noted a high level of symptoms among workers in two departments, chloroprene polymerization
and distillation, in both the pilot plant and early period of regular production. Over the time period
from 1944-1947, the author conducted a series of employee medical examinations. In the
polymerization department of the production plant, temporary hair loss affected 11 of 12 workers or
90%. The author attributed this to systemic rather than direct skin exposure. Dermatitis was present in
four workers (30%), and all other symptoms evaluated were limited to no more than one worker. In
the distillation department of the production plant, 19 of 21 workers (90%) complained of fatigue and
pressure or pains over the chest, with much lower numbers (3-6 employees) complaining of
palpitations, giddiness, irritability, and dermatitis. No workers experienced loss of hair.
Guided by animal studies and reports from other companies, Nystrom (1948, 003695)
evaluated employees for impaired renal and liver function, basal metabolism, and pulmonary and
cardiovascular abnormalities by conducting general body examination, clinical chemistry of the urine
and blood, and other tests referred to as "special investigations" (including X-rays, electrocardiograms,
and hypoxemia and stress tests). The results of these evaluations were reported in an anecdotal manner
with no qualitative or quantitative (e.g., statistical significance of results) details. Except for increased
symptoms with exercise right after exposure (among distillation department workers), no clear
pathologies were observed. In the pilot plant, where exposures were less controlled, Nystrom (1948,
003695) noted anemia among exposed workers. The author also observed that, when the workers were
educated about the dangers and safety precautions were enforced, the symptoms decreased.
Biochemical and hematological effects of occupational chloroprene exposure of workers in a
chloroprene manufacturing plant were reported by Gooch and Hawn (1981, 064944). The study
investigated exposed and non-exposed workers at the DuPont Louisville Works plant and included any
workers employed as of December 31, 1977. Workers were categorized into three exposure groups:
42
-------
currently exposed (workers assigned to the chloroprene polymerization area of the plant as of
December 31, 1977); not currently exposed (workers with a history of work in the chloroprene
polymerization area of the plant); and never exposed (workers with no history of being assigned to the
chloroprene polymerization area of the plant). Exposure groups were based on a job description
indicating the worker was assigned to the chloroprene polymerization area of the plant. Additionally,
seven employees in supervisory roles familiar with chloroprene manufacture independently rated each
job as "high," "medium," "low," or "varied" in regard to the actual potential for exposure to
chloroprene. At the Louisville plant, all new hires were required to undergo a physical examination
upon employment and at specified intervals thereafter that included clinical chemistry and
hematological analyses, chest x-rays, and pulmonary function tests (Jones et al., 1975, 625203). The
results for tests conducted between 1974 and 1977 were included in the analysis. When clinical
chemistry parameters were compared between exposure groups no effect was seen in currently exposed
workers and those workers never exposed to chloroprene; this lack of effect was also observed when
currently exposed workers with "high" potential for chloroprene exposure were compared to workers
never exposed to chloroprene. Paired analyses (comparisons of clinical chemistry in workers with test
results before and after being assigned to chloroprene manufacture) showed that glucose and
cholesterol values were lower and LDH values were higher in workers after being assigned to
chloroprene manufacture compared to test results before assignment. However, all values were well
within normal ranges, indicating the results were likely due to normal variability and not to any
chemically-related effect. No hematological effects were observed.
In a subsequent NIOSH industrial hygiene investigation of the DuPont Louisville Works plant,
ambient and personal monitoring was conducted to assess worker exposure to chloroprene (McGlothlin
et al., 1984, 625204). Additionally, medical interviews and medical record examinations were
conducted to determine if adverse health outcomes due to workplace exposures could be detected. In
the air quality monitoring portion of the study, personal breathing zone and area air samples were
collected in the manufacturing areas that dealt with both the monomer (chloroprene) and polymer
(polychloroprene). The range of chloroprene air concentrations detected by fixed location area
samples ranged from below detection limits (32 out of 79 total samples) to 1,200 ppm. The two
highest concentrations (910 and 1,200 ppm) were detected at "drainage trenches" and may not have
been representative of normal workday exposures experienced in the manufacture areas. In the
remaining fixed location samples, the average chloroprene concentration (over 6-7 hours) was
5.6 ppm, which was below the OSHAPEL of 25 ppm for an 8-hour workday. Only one fixed location
area air sample (excluding those taken at the drainage trenches) exceeded the OSHAPEL (26 ppm).
Of the 194 personal air samples taken from workers in the monomer and polymer portions of the plant,
103 (54%) exceeded the NIOSH 15-minute recommendation of 1 ppm, 5 (3%) exceeded the ACGIH
TLV of 10 ppm, and only 1 (0.5%) exceeded the OSHAPEL of 25 ppm. It is important to note that the
magnitude of worker exposure detected in this study may not be representative of exposures workers
experience currently due to increased safety procedures and improved manufacturing processes. In the
43
-------
medical examination portion of the study, 37 workers were interviewed and demographic and
occupational information was collected. Smoking histories, medical problems, past illnesses, and
current symptoms were covered in the interviews and any relation to current work exposures was
sought. None of the workers indicated in the interviews that they felt that their current health status
was related to their workplace exposure to chloroprene. Some workers indicated that they had
occasionally experienced lightheadedness and eye, nose, and throat irritation. Workers experiencing
respiratory disease had medical histories indicating heavy smoking, heart disease, or other medical
issues. An examination of medical records for 8 of the 37 workers found that the only significant
problem observed was a large deviation in pulmonary function tests year-to-year that may have been
due to faulty test equipment. In summary, no major health effects were observed in workers involved
in chloroprene manufacture and polymerization even though personal and ambient monitoring
indicated that occupational safety limits were occasionally exceeded.
In a Russian review of the effects of chloroprene, Sanotskii (1976, 063885) noted that medical
examinations of chloroprene production workers had found changes in the nervous system, hepatic and
renal function, cardiovascular system, and hematology. Assessment of exposures in Russian latex and
rubber manufacturing plants showed that chloroprene was the main hazard and that exposures ranged
from 1-7 mg/m3 (0.28-1.93 ppm) in exposed work areas. One of the studies reported in this review
included medical exams of 12 men and 53 women, of whom two-thirds had been employed in a
chloroprene production plant for less than 5 years. Cardiovascular examinations found muffled heart
sounds in 30 workers, reduced arterial pressure in 14, and tachycardia in 9. There was also a reduction
in RBC counts, with hemoglobin substantially below the limit of physiological variation.
Erythrocytopenia, leucopenia, and thrombocytopenia were observed. Increases in vestibular function
disturbance were associated with duration of work.
In another study reviewed by Sanotskii (1976, 063885), women aged 19-23 employed in jobs
with chloroprene exposure for 2-4 years had abnormal diurnal variation in arterial pressure, with
reduced systolic and diastolic components at the end of the workday when compared with controls.
Their pulse rates were considerably higher than those of controls (p < 0.01). Central nervous system
(CNS) function was also affected with lengthening of sensorimotor response to visual cues compared
with controls. Olfactory thresholds increased with duration of employment.
4.1.2.2. Chronic Noncancer Effects
Gooch and Hawn (1981, 064944) investigated the effects on clinical chemistry parameters in
workers chronically exposed to chloroprene (study description above). When currently exposed
workers were compared to never exposed workers stratified by duration of exposure (<1 year, 1-
5 years, 6-10 years, >10 years), cholesterol and alkaline phosphatase were higher in workers exposed
>10 years (cholesterol) and 6-10 years (alkaline phosphatase). This pattern was also observed when
only workers with a "high" potential for exposure were analyzed. When cholesterol values were
adjusted for the age of the workers, no chemically-related effect was observed. The differences seen in
44
-------
alkaline phosphatase were attributed to two workers with abnormally high alkaline phosphatase levels
due to bone injury and blood pressure medication. Therefore, no chemically-related effects were seen
in clinical chemistry parameters in workers chronically exposed to chloroprene.
Chronic effects in exposed workers at an electrical engineering plant were also reported in the
review by Sanotskii (1976, 063885). When compared to 118 unexposed controls, the chloroprene-
exposed cohort (143 workers) exhibited an increased incidence of disturbances of spermatogenesis
after 6-10 years of work and morphological disturbances after 11 years or more. A questionnaire
showed that the rate of spontaneous abortion in the wives of chloroprene workers was more than
threefold greater when compared to the control group. This study presents interpretational difficulties
concerning the level of participation of the exposed workers and their wives, the quantitative
interpretation of the reported sperm abnormalities, and the appropriate matching of exposed and
control populations. In an earlier evaluation of this study, U.S. EPA (1985, 017624) concluded that
recall bias associated with a retrospective questionnaire, such as was used in the study reviewed by
Sanotskii (1976, 063885), was likely, and the likelihood that the study would have discovered a real
increase in the rate of spontaneous abortions was remote, as embryos with chromosomal abnormalities
are spontaneously aborted early in pregnancy. Many spontaneous abortions occur before a woman
recognizes that she is pregnant, with clinical signs of miscarriage often mistaken for heavy or late
menstruation (Griebel et al., 2005, 625142). Thus, U.S. EPA (1985, 017624) concluded that it was not
reasonable to draw conclusions on the possible effect of chloroprene on early fetal losses based on the
Sanotskii (1976, 063885) review. In addition, the EPA suggested that the low participation of male
volunteers available for sperm analysis (9.5% participation, 15/143 workers) indicated that a large
degree of selection bias may have been present. If males with reproductive deficits self-selected
themselves for participation, the meaningful interpretation of the study results may be limited.
The final conclusion of the EPA analysis was that it is not possible to interpret the results in the
Sanotskii (1976, 063885) review with any degree of reliability (U.S. EPA, 1985, 017624). Savitz et al.
(1994, 068186) and Schrag and Dixon (1985, 062573) separately reviewed the study and also
concluded that insufficient methodological details were available to critically evaluate the observation
reported by Sanotskii (1976, 063885).
Sanotskii (1976, 063885) also reported a study of chromosome aberrations in leukocyte
cultures prepared from blood cells of chloroprene production employees. The occurrence of
chromosomal aberrations were significantly higher (p < 0.001) in the exposed group compared to the
control group, as well as elevated compared to reported levels among healthy persons. Similar results
were reported for a different study of two sets of female employees: (1) 20 women aged 19-23 and
exposed to 3-7 mg/m3 (0.83-1.93 ppm) chloroprene for 1-4 years; and (2) 8 women aged 19-50 and
exposed to 1-4 mg/m3 (0.28-1.1 ppm) for 1-20 years. The results of these two studies are shown in
Table 4-12. Insufficient data on analytical methods and exposure ascertainment used in the
investigation of chromosomal aberrations in chloroprene workers preclude drawing conclusions from
the results presented by Sanotskii (1976, 063885).
45
-------
Table 4-12. Frequency of chromosomal aberrations in leukocyte cultures from
blood cells of chloroprene production workers
Chloroprene
Exposure
Air
Concentration
Chloroprene
Workers'
Control"
1-4 (mg/m3)d
3-7 (mg/m3)d
Population
Control1^
Number of
Workers
Examined
18
9
8
20
181
Length of
Service
(Years
Exposed)
— -
1-20
1-4
Age
Range of
Workers
Exposed
— -
19-50
19-23
Number of
Metaphase
Cells
Analyzed
1,666
572
648
1,748
28,386
Aberrant
Cells
(%)
4.77±0.57a
0.65 ±0.56
2.5±0.49b
3.49±0.51a
1.19 ±0.06
Types of Aberrations (%)
Chromatid
74.4
100
— -
50.3
Chromosome
25.6
0
— -
49.7
"Statistical Significance p < 0.001. All values means ± SE.
bStatistical Significance p < 0.05.
"Study 1 (Bochkov et al., 1972, 644818) reviewed in Sanotskii (1976, 063885).
dStudy 2 (Fomenko and Katosova, 1973, 644819) reviewed in Sanotskii (1976, 063885).
Spontaneous chromosome aberrations in normal human leukocyte culture.
Source: Sanotskii (1976, 063885).
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN ANIMALS-
ORAL AND INHALATION
4.2.1. Oral Exposure
The only available long-term animal study using the oral route of administration was part of a
developmental/reproductive study. Ponomarkov and Tomatis (1980, 075453) administered
chloroprene dissolved in olive oil by stomach tube to 17 female Berlin Druckrey (BD-IV) rats at a
single dose (100 mg/kg body weight) on gestational day (GDI7). Progeny from treated females (81
males and 64 females) were treated weekly with 50 mg/kg body weight by stomach tube from the time
of weaning for life (120 weeks). A control group of 14 female rats was treated with 0.3 mL olive oil.
The purity of the chloroprene was reported as 99% with 0.8% 1-chlorobutadiene; storage conditions
were not reported. All survivors were sacrificed at 120 weeks or when moribund and autopsied.
Major organs, as well as those that showed gross abnormalities, were examined histologically.
Litter sizes and preweaning mortality, survival rates, and body weights did not differ between
chloroprene-treated animals and controls. Severe congestion of the lungs and kidneys was observed in
animals treated with chloroprene that died within the first 23-35 weeks of treatment. Multiple liver
necroses were observed in some animals (number not specified) autopsied 80-90 weeks after the onset
of treatment.
Tumor incidences and distribution reported in this study are summarized in Tables 4-13 and
4-14. No statistically significant differences were reported between treated and control rats. However,
several tumors observed in male progeny (intestinal leiomyosarcoma, osteoma, kidney mesenchymal
46
-------
tumor, bone hemangioma, neurinoma of the optic nerve, transition-cell carcinoma of urinary bladder,
and forestomach papilloma) and female dams and progeny (uterine squamous cell carcinoma, lung
reticulosarcoma, forestomach papilloma, sebaceous basal cell carcinoma) treated weekly with
chloroprene were not seen in the vehicle control group. Subcutaneous fibromas were more numerous
in chloroprene-treated male rats than in controls. Mammary and ovarian tumors were slightly elevated
in chloroprene-treated female rats than in controls.
Table 4-13. Tumor incidence in female BD-IV rats treated orally with chloroprene
(100 mg/kg) on GDI? and in their progeny treated (50 mg/kg) weekly for life
(120 weeks)
Group
Treated females
Treated progeny
Males
Females
Control females
Control progeny
Males
Females
Number"
16
54
62
14
49
47
Tumor Bearing Rats
n
9
15
33
5
16
24
%
56.2
27.8
53.2
35.7
32.7
51.1
Number Of Tumors
Total
14
18
37
7
16
29
Per rat
0.9
0.3
0.6
0.5
0.3
0.6
Animals With More
Than One Tumor
n
5
o
J
4
2
—
5
%
31.3
5.6
6.5
14.3
—
10.6
aSurvivors at the time the first tumors were observed.
Source: Used with permission from S. Karger AG, Ponomarkov and Tomatis (1980, 075453).
47
-------
Table 4-14. Distribution of tumors in female BD-IV rats treated orally with
chloroprene (100 mg/kg) on GDI? and their progeny treated (50 mg/kg) weekly for
life (120 weeks)
Group
Treated females
Treated progeny
Males
Females
Control females
Control progeny
Males
Oral
Cavity
n
1
—
1
2
i
%
6.3
—
7.1
4.1
9 1
Mammary
n
6
25
4
99
%
37.5
40.3
28.6
zLfi 8
Ovary
n
2
9
—
T.
%
12.5
—
14.5
—
£ A
Thyroid
n
...
1
1
...
%
...
1.9
1.6
—
Soft Tissue
n
—
7
—
1
4
%
—
13.0
—
7.1
8.2
Pituitary
n
1
2
2
—
2
%
6.3
3.7
3.2
—
4.1
Other
n
4a
8b
—
lc
8d
oe
%
25.0
14.8
—
7.1
16.3
£ A.
aOne each: uterine squamous cell carcinoma; lung reticulosarcoma; forestomach papilloma; sebaceous basal cell
carcinoma.
bOne each: intestinal leiomyosarcoma; osteoma; kidney mesenchymal tumor; bone hemangioma; neurinoma of the
optic nerve; adrenal cortical adenoma; transition-cell carcinoma of urinary bladder; forestomach papilloma.
0 Adrenal cortical adenoma.
dTwo lymphomas; 1 each: lung epidermoid carcinoma; spleen hemangioma; osteosarcoma; mediastinal sarcoma;
meningioma; adrenal cortical adenoma.
eOne each: stomach fibrosarcoma; lymphoma; uterine adenoma.
Source: Used with permission from S. Karger AG, Ponomarkov and Tomatis (1980, 075453).
4.2.2. Inhalation Exposure
The NTP conducted 16-day, 13-week, and 2-year inhalation exposure studies with chloroprene
in F344/N rats and B6C3Fi mice (NTP, 1998, 042076). Results of the 13-week study were reported by
Melnick et al. (1996, 625207), while the cancer results of the 2-year study were discussed separately
by Melnick et al. (1999, 000297) in relation to observations noted with 1,3-butadiene in mice. All
experimental regimes consisted of 6 hours per day, 5 days per week whole-body exposures. Group
sizes were 10 animals/sex/group in the 16-day and 13-week studies and 50 animals/sex/group in the
2-year study. Overall purity of the bulk chloroprene was determined to be approximately 96% by gas
chromatography. Vapor was generated in the 13-week and 2-year studies from chloroprene in an
evaporation flask kept at 66°C (72°C in the 16-day studies) followed by a temperature-controlled
condenser column (to remove less volatile impurities such as chloroprene dimers); the chloroprene
reservoir was kept at dry ice temperature (16-day study) or under nitrogen (13-week and 2-year
studies). The actual concentrations generated from the evaporator flask were within 99% of target
concentrations at the beginning of the exposures and were 95% pure at the end of the exposure period.
Chloroprene was dragged from the evaporator by a metered flow of nitrogen before being injected into
the mixer column, where it was diluted with HEPA- and charcoal-filtered air. Impurities more volatile
than chloroprene, such as chlorobutene, never exceeded more than 0.6% of the desired chloroprene
concentration when sampled from the distribution line, the last sampling point upstream from the
48
-------
actual exposure chambers. Histopathology was performed by a study pathologist and reviewed by a
quality assurance pathologist and the Pathology Working Group.
NTP 16-Day Exposure. In the 16-day study, rats were exposed to target concentrations of 0,
32, 80, 200, or 500 ppm chloroprene (NTP, 1998, 042076). Actual chamber concentrations were 0,
31.1 ± 1.9, 80.7 ± 5.0, 198 ± 10, and 503 ± 24 ppm chloroprene. On day 4, rats were placed in
metabolism cages for 16-hour urine collection. A necropsy was performed on all animals, and
histopathological examinations were performed on controls, 80 ppm female rats, and 200 and 500 ppm
male and female rats. Tissues and organs examined included brain, liver, kidney, lung, bone marrow,
thymus, spleen, and testes. Sperm morphology and vaginal cytology were not evaluated.
Survival and body weights of rats are given in Table 4-15. Only one male in the high-exposure
group (500 ppm) survived. Females in the high-exposure group had a higher survival (7/10) with a
significantly decreased body weight (-6% compared with controls). Significantly decreased body
weight gain was also observed in males and females at 200 ppm, and in females at 500 ppm.
Table 4-15. Survival and body weights of rats in the 16-day inhalation study of
chloroprene
Sex
Male
Female
Exposure (ppm)
0
32
80
200
500
0
32
80
200
500
Survival
7/10
10/10
10/10
9/10
1/10
9/10
9/10
9/10
3/10
7/10
Mean Body Weight (g)
Initial
115±4
113±4
118±5
114±4
114±4
100 ±2
100 ±2
103 ±2
101 ±2
102 ±2
Final
139 ±5
134 ±6
136 ±5
127 ±5
104
110 ±3
109 ±3
112 ±2
101 ±4
103 ±3
Change
(+) 20 ± 2
(+) 20 ± 2
(+) 18 ± 1
(+)ll±2b
(-)4a
(+) 9 ± 1
(+) 8 ± 1
(+) 9 ± 1
(+) 4 ± lb
(-) 1 ± lb
aNo standard error calculated due to high mortality.
bStatistical significance (p < 0.01) from the chamber control group by Williams' or Dunnett's test.
(+) Increased weight.
(-) Decreased weight
Source: NTP (1998. 042076X
Minimal to mild olfactory epithelial degeneration was significantly increased in all exposed
groups of males and females compared to those in the chamber control groups (Table 4-16). Mild to
moderate centrilobular hepatocellular necrosis was observed in male and female rats exposed to 200 or
500 ppm. Hematological and clinical chemistry parameters indicated increased serum alanine
aminotransaminase (ALT), glutamine dehydrogenase (GDH), and sorbitol dehydrogenase (SDH)
activities, as well as anemia and thrombocytopenia (decreased platelet count) in the 200 (female) and
49
-------
500 (male and female)-ppm groups, on day 4 only. In females, significant increases in kidney weights
(right kidney only) were seen at 80 and 500 ppm, and significantly increased liver weights were seen at
200 and 500 ppm.
Table 4-16. Incidences of selected nonneoplastic lesions in rats in the 16-day
inhalation study of chloroprene
Control
32 ppm
80 ppm
200 ppm
500 ppm
Male
Nose3
Degeneration, olfactory epithelium
Metaplasia, squamous, olfactory
epithelium
Metaplasia, respiratory, olfactory
epithelium
Metaplasia, squamous, respiratory
epithelium
Liver3
Necrosis, centrilobular
Inflammation, chronic
10/10
1/10
(1.0)b
0/10
0/10
1/10
(1.0)
10/10
0/10
0/10
10/10
10/10C
(1.0)
0/10
2/10
(1.0)
1/10
(1.0)
1/10
0/10
0/10
10/10
10/10C
(1.1)
0/10
5/10d
(1.0)
0/10
10/10
0/10
0/10
10/10
10/10C
(1.9)
1/10
(2.0)
6/10d
(1.0)
0/10
10/10
1/10
(2.0)
0/10
10/10
10/10C
(3.8)
4/10d
(1.8)
1/10
(2.0)
7/10
(1.7)
10/10
9/10c
(3.4)
1/10
Female
Nose3
Degeneration, olfactory epithelium
Metaplasia, squamous, olfactory
epithelium
Metaplasia, respiratory, olfactory
epithelium
Metaplasia, squamous, respiratory
epithelium
Liver3
Necrosis, centrilobular
Inflammation, chronic
10/10
0/10
0/10
0/10
1/10
(2.0)
10/10
0/10
0/10
10/10
9/10c
(1.2)
1/10
(1.0)
7/10c
(1.0)
1/10
(1.0)
3/10
0/10
0/10
10/10
10/10C
(1.6)
1/10
(1.01)
8/10c
(1.2)
0/10
10/10
0/10
0/10
10/10
10/10C
(3.4)
4/10d
(1.0)
3/10
(1.0)
0/10
10/10
7/10c
(2.6)
2/10
(1.0)
10/10
10/10C
(3.3)
0/10
7/10c
(1.4)
4/10
(1.3)
10/10
3/10
(2.0)
5/10d
(1.0)
3Number of animals with tissue examined microscopically.
bAverage severity grade of lesions in affected rats: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked.
Statistical significance (p < 0.01) from the chamber control group by the Fisher's exact test.
dStatistical significance p < 0.05.
Source: NTP (1998. 042076)
In the mouse portion of the 16-day NTP (1998, 042076) study, target exposure levels were 0,
12, 32, 80, and 200 ppm chloroprene. The actual exposure chamber concentrations were 0, 11.9 ± 0.8,
31.1 ± 2.0, 80.8 ± 5.2, and 301 ± 12 ppm chloroprene. Additional groups of 10 male and 10 female
mice designated for day 5 hematology and clinical chemistry analyses were exposed to the same
50
-------
chloroprene concentrations. Histopathology examinations were performed on chamber controls and 80
and 200 male and female mice as well as on selected target organs in other groups. Tissues and organs
examined were identical to those described for the rat. Survival and body weights for mice are given
in Table 4-17. All male and female animals in the high-concentration group died, exhibiting signs of
narcosis, hepatocellular and thymic necrosis, and hypertrophy of the myocardium. Significantly
decreased body weight gain (compared with controls) was seen in males at 32 and 80 ppm.
Hematological and clinical chemistry parameters in exposed mice were similar to those in the chamber
controls. Increased incidences of multifocal random hepatocellular necrosis and thymic necrosis,
characterized by karyorrhexis of thymic lymphocytes, were observed in male and female mice exposed
to 200 ppm. No histopathological damage was observed in the lungs of exposed mice.
Table 4-17. Survival and body weights of mice in the 16-day inhalation study of
chloroprene
Exposure (ppm)
Survival
Mean Body Weight (g)
Initial
Final
Change
Male
0
12
32
80
200
10/10
10/10
10/10
10/10
0/10
24.7 ±0.5
24.8 ±0.5
25.3 ±0.3
24.8 ±0.5
24.2 ±0.4
27.0 ±0.5
27.1 ±0.6
26.5 ±0.3
26.1 ±0.6
—
(+)2.3±0.1
(+) 2.3 ± 0.3
(+)1.2±0.3a
(+) 1.3±0.2a
—
Female
0
12
32
80
200
10/10
10/10
10/10
10/10
0/10
19.5 ±0.7
20.4 ±0.8
19.9 ±1.0
20.1 ±0.8
20.0 ±0.6
22.6 ±0.5
23.1 ±0.4
22.1 ±0.2
22.5 ±0.3
—
(+) 2.3 ± 0.3
(+) 2.6 ± 0.3
(+) 1.8 ±0.3
(+) 2.7 ± 0.3
—
""Significantly different (p < 0.01) from the chamber control group by Williams' or Dunnett's test.
Source: NTP (1998. 042076X
NTP 13-Week Study. A range-finding 13-week inhalation study was conducted by NTP (1998,
042076) (reported by Melnick et al. (1996, 625207)1 using both mice and rats. In the rat, target
exposure groups were 0, 5, 12, 32, 80, and 200 ppm chloroprene. The actual chamber concentrations
achieved were 0, 5.03 ±0.18, 12.1 ±0.4, 31.9 ± 1.0, 80.2 ± 1.7, and 200 ± 5.0 ppm chloroprene.
Separate groups of 10 male and 10 female rats designated for coagulation studies were exposed to
these concentrations for 2 days. Rats designated for hematology and clinical chemistry tests were first
placed in metabolism cages for 16-hour urine collections. Sperm samples were collected from male
rats at the end of the studies. Samples of vaginal fluid and cells were collected for up to 7 consecutive
days prior to the end of the studies for cytology evaluations. Five male and five female rats were
51
-------
exposed to 0, 5, 32 or 200 ppm for glutathione evaluations. At week 11, all male and female core study
rats were administered neurobehavioral tests measuring the following parameters: forelimb/hind-limb
grip strength, horizontal activity, rearing activity, total activity, tail-flick latency, startle response
latency, and startle response amplitude. Survival and body weights of rats are given in Table 4-18. No
effects on final mean body weights were seen.
Table 4-18. Survival and body weights of rats in the 13-week inhalation study of
chloroprene
Exposure (ppm)
Survival
Mean Body Weight (g)
Initial
Final
Change
Male
0
5
12
32
80
200
10/10
10/10
10/10
10/10
10/10
9/10
109 ±4
119 ±2a
116±1
117±2
116±1
116±3
311±9
323 ±11
306 ±9
327 ±11
301 ±8
304 ±8
(+) 202 ± 8
(+) 204 ± 10
(+) 190 ± 8
(+) 209 ± 10
(+) 184 ± 7
(+) 185 ± 7
Female
0
5
12
32
80
200
10/10
10/10
10/10
10/10
10/10
10/10
102 ±2
101 ±1
102 ±2
101 ±2
103 ±1
102 ±1
191 ±4
193 ±4
199 ±5
195 ±4
192 ±3
183 ±3
(+) 89 ± 3
(+) 92 ± 3
(+) 97 ± 4
(+) 94 ± 4
(+) 90 ± 3
(+)81±3
""Significantly different (p < 0.05) from the chamber control group by Williams' or Dunnett's test.
Source: NTP (1998. 042076)
On day 2, hematocrit values, hemoglobin concentrations, and erythrocyte counts were
increased in males exposed to > 32 ppm and in females exposed to 200 ppm. At week 13, male and
female rats in the 200-ppm groups demonstrated decreased hematocrit values, hemoglobin
concentrations, and erythrocyte counts characterized as normocytic, normochromic anemia.
Thrombocytopenia, evidenced by a reduction in circulating platelet numbers, was observed in male
and female rats in the 200-ppm groups on day 2 and in the females at 80 and 200 ppm on day 22.
Platelet numbers rebounded at study termination in the highest exposure groups for both male and
female rats. Activities of serum ALT, GDH, and SDH were elevated on day 22 in both sexes of the
200-ppm group. However, these increases were transient, and serum activities of the enzyme levels
returned to control levels by the end of the exposure period. At week 13, an alkaline phosphatase
(ALP) enzymeuria occurred in males exposed to > 32 ppm and in females exposed to 200 ppm. In
male rats in the 200-ppm group, proteinuria was seen at week 13. Significant reductions in nonprotein
sulfhydryl (NPSH) concentrations were observed in the livers from male rats exposed to 200 ppm for 1
52
-------
day or 12 weeks, as well as in female rats exposed to 200 ppm for 12 weeks. Nonprotein sulfhydryl
concentrations were reduced in the lung of 200 ppm female rats after 1 day but not after 12 weeks of
exposure to 200 ppm. Significant increases in kidney weights were seen in both male and female rats
at 200 ppm and in females at 80 ppm. In male rats exposed to 200 ppm, sperm motility was
significantly less than that of the chamber control group. Of the neurobehavioral parameters,
horizontal activity was increased in male rats exposed to > 32 ppm compared with chamber control
animals. Total activity was increased in male rats in the 32 and 200-ppm groups. There were no
exposure-related effects on motor activity, forelimb/hind-limb grip strength, or startle response.
Increased incidences of minimal to mild olfactory epithelial degeneration and respiratory
metaplasia occurred in male and female rats exposed to 80 or 200 ppm (Table 4-19). The incidence of
olfactory epithelial degeneration in females exposed to 32 ppm was significantly greater than in the
chamber control group. No effects were observed in the respiratory epithelium of exposed rats. In
female rats exposed to 200 ppm, the incidence of hepatocellular necrosis was significantly greater than
in the chamber control group. Variably sized aggregates of yellow or brown material consistent with
hemosiderin appeared in small vessels or lymphatics in or near portal triads or in Kupffer cells of male
and female rats exposed to 200 ppm and were significantly increased compared with chamber controls.
53
-------
Table 4-19. Incidences of selected nonneoplastic lesions in rats in the 13-week
inhalation study of chloroprene
Control
5 ppm
12 ppm
32 ppm
80 ppm
200 ppm
Male
Nose3
Degeneration, olfactory
epithelium
Metaplasia, respiratory,
olfactory epithelium
Liver3
Necrosis, centrilobular
Inflammation, chronic
Hemosiderin pigmentation
10/10
0/10
0/10
10/10
0/10
0/10
0/10
0/10
—
—
2/10
0/10
1/10
(1.0)
0/10
10/10
0/10
0/10
1/10
0/10
0/10
0/10
10/10
3/10
(1.0)b
0/10
1/10
01/10
0/10
0/10
10/10
10/10C
(1.0)
4/10d
(1.3)
10/10
0/10
1/10
(1.0)
0/10
10/10
10/10C
(2.0)
4/10d
(1.3)
10/10
3/10
(2.0)
2/10
(1.0)
5/10d
(1.6)
Female
Nose3
Degeneration, olfactory
epithelium
Metaplasia, respiratory,
olfactory epithelium
Liver3
Necrosis, centrilobular
Inflammation, chronic
Hemosiderin pigmentation
10/10
0/10
0/10
10/10
0/10
2/10
(2.0)
3/10
(1.0)
0/10
—
—
2/10
0/10
0/10
0/10
10/10
0/10
0/10
5/10
0/10
1/10
(2.0)
1/10
(3.0)
10/10
4/10d
(1.0)
0/10
3/10
0/10
0/10
0/10
10/10
9/10c
(1.9)
8/10c
(2.0)
10/10
0/10
1/10
(2.0)
0/10
10/10
10/10C
(1.9)
9/10c
(2.0)
10/10
5/10d
(1.0)
8/10d
(1.3)
9/10c
(1.7)
3Number of animals with tissue examined microscopically.
bAverage severity grade of lesions in affected rats: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked.
Significantly different (p < 0.01) from the chamber control group by Fisher's exact test.
dSignificantly different (p < 0.05) from the chamber control group by Fisher's exact test.
Source: NTP (1998. 042076).
In the mouse portion of the NTP 13-week inhalation study, the target concentration exposure
groups were 0, 5, 12, 32, and 80 ppm chloroprene. Actual chamber concentrations of 0, 5.02 ± 0.2,
12.1 ± 0.3, 31.9 ± 0.9, and 80.2 ±1.6 ppm chloroprene were achieved. Survival and body weights are
given in Table 4-20. There was no increased mortality in any exposure group. Final mean body
weights in 80 ppm males were significantly decreased compared with controls.
54
-------
Table 4-20. Survival and body weights of mice in the 13-week inhalation study of
chloroprene
Sex
Male
Female
Exposure (ppm)
0
5
12
32
80
0
5
12
32
80
Survival
10/10
10/10
10/10
10/10
10/10
10/10
10/10
10/10
10/10
10/10
Mean Body Weight (g)
Initial
25.5 ±0.4
25.2 ±0.3
25.2 ±0.2
25.4 ±0.2
24.7 ±0.3
20.4 ±0.2
20.9 ±0.3
20.4 ±0.3
20.8 ±0.2
20.5 ±0.2
Final
35.9 ±0.9
35.1±0.9
34.9 ±0.6
36.0 ±0.9
32.7±0.6a
30.3 ±1.0
32.2 ±0.9
30.1 ±0.6
32.6 ±0.8
30.2 ±1.3
Change (+)
10.5 ±0.7
10.0 ±0.7
9.7 ±0.6
10.6 ±0.9
7.9±0.5a
9.9 ±0.9
11.3 ±0.9
9.7 ±0.6
11.8 ±0.7
9.7 ±1.2
""Significantly different (p < 0.05) from the chamber control group by Williams' or Dunnett's test.
Source: NTP (1998. 0420761
Hematology variables were similar to, although more mild than, the 13-week rat study.
Anemia, including decreased hematocrit values and erythrocyte counts, occurred in female mice
exposed to 32 and 80 ppm. Platelet counts were minimally increased in female mice exposed to 32
and 80 ppm, suggesting increased platelet production. No significant organ weight effects were
observed. Sperm morphology and vaginal cytology parameters were similar to those of the chamber
controls. The incidence of squamous epithelial hyperplasia of the forestomach was significantly
increased in male and female mice exposed to 80 ppm (Table 4-21). Preening behavior may have led
to direct gastrointestinal exposure to chloroprene.
55
-------
Table 4-21. Incidences of forestomach lesions in mice in the 13-week inhalation
study of chloroprene
Control
5 ppm
12 ppm
32 ppm
80 ppm
Male
Number examined microscopically
Squamous epithelial hyperplasia
10/10
0/10
3/10
0/10
0/10
—
10/10
0/10
10/10
4/10a
(1.5)b
Female
Number examined microscopically
Squamous epithelial hyperplasia
10/10
0/10
0/10
—
0/10
—
10/10
0/10
10/10
9/10c
(1.9)
""Statistical significance (p < 0.05) from the chamber control group by Fisher's exact test.
bAverage severity grade of lesions in affected mice: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked.
Statistical significance p < 0.01.
Source: NTP (1998. 042076).
NTP 2-Year Exposure (Rat). In the 2-year (NTP, 1998, 042076) inhalation study of
chloroprene in male and female rats, groups were exposed to target concentrations of 0, 12.8, 32, and
80 ppm chloroprene. The actual chamber concentrations to which the animals were exposed, were 0,
12.8 ± 0.4, 31.7 ± 1.1, and 79.6 ±1.6 ppm chloroprene. The high-exposure concentration was chosen
based on the observation of anemia and hepatocellular necrosis in rats exposed to 200 ppm for
13 weeks. The range of exposures selected included the NOAEL for degenerative olfactory epithelial
lesions in the 13 week study. Estimates of 2-year rat survival probabilities are shown in Table 4-22.
Survival of male rats exposed to 32 or 80 ppm was significantly less than that of the chamber control
group.
56
-------
Table 4-22. Two-year survival probability estimates for F344/N rats chronically
exposed (2 years) to chloroprene by inhalation
Sex
-------
The incidences of nonneoplastic and neoplastic lesions observed in rats following 2-year
inhalation exposures to chloroprene are given in Tables 4-23 and 4-24 (NTP, 1998, 042076).
Squamous cell papilloma and combined squamous cell papilloma and squamous cell carcinoma of the
oral cavity (oral mucosa, tongue, pharynx, and gingiva) was significantly increased in male rats
exposed to 32 ppm and male and female rats exposed to 80 ppm compared to those in the chamber
controls. The incidences of these tumors exceeded historical control ranges. Squamous hyperplasia
was observed in three male rats exposed to 80 ppm chloroprene, and was characterized by focal
thickening and folding of the squamous epithelium.
Table 4-23. Incidence and severity of nonneoplastic lesions in F344/N rats
chronically exposed (2 years) to chloroprene by inhalation
Tissue Site/Lesion Type
Oral cavity
Sqamous Cell Hyperplasia
Thyroid gland
Follicular Cell Hyperplasia
Lung
Alveolar Hyperplasia
Kidney (renal tubules)
Hyperplasia
Olfactory
Atrophy
Basal Cell Hyperplasia
Metaplasia
Necrosis
Chronic Inflammation
Lesion Incidence (Severity)
Males (ppm)
0
0/50
0/50
5/50
(1.4)
14/50
(2.0)
3/50
(1.7)
0/50
6/50
1.7
0/50
0/50
12.8
0/50
2/50
(2.0)
16/50C
(1.4)
20/50
(2.6)
12/50b
(1.8)
0/50
5/50
(1.0)
ll/50b
(2.0)
5/50c
(1.0)
32
0/50
4/49b
(1.8)
14/49b
(1.9)
28/50c
(2.1)
46/49c
(2.2)
38/49c
(1.6)
45/49c
(1.8)
26/49c
(2.0)
9/49-
(1.6)
80
3/50
(2.7)a
1/50
(1.0)
25/50c
(1.4)
34/50c
(2.9)
48/49c
(3.6)
46/49c
(2.2)
48/49c
(3.1)
19/49C
(2.2)
49/49°
(2.7)
Females (ppm)
0
0/49
6/49
(1.8)
6/49
(1.3)
0/49
0/49
0/49
0/49
0/49
12.8
0/50
22/50c
1.4)
6/50
(1.8)
1/50
(1.0)
0/50
1/50
(1.0)
0/50
0/50
32
0/50
22/50c
(1.5)
11/50
(2.1)
40/50C
(1.3)
17/50C
(1.1)
35/50c
(1.0)
8/50c
(2.0)
2/50
(1.0)
80
2/50
(2.5)
34/50c
(1.3)
21/50C
(2.0)
50/50C
(2.9)
49/50c
(2.3)
50/50C
(2.7)
12/50C
(1.3)
33/50c
(2.0)
""Severity of lesions graded as: 1= minimal, 2 = mild, 3 = moderate, 4 = marked, average severity reported in
parenthesis.
bStatistical significance p < 0.05, p values correspond to the pairwise comparisons between the chamber controls
and that exposed group. The logistic regression test regards lesions, in animals dying prior to terminal kill, as
nonfatal.
Statistical significance p < 0.01.
Source: NTP (1998. 042076).
The incidences of thyroid gland follicular cell adenoma or carcinoma (combined) in male rats
exposed to 32 or 80 ppm were significantly greater than those in the chamber control group and
exceeded historical control ranges. The incidences of follicular cell adenoma and follicular cell
adenoma or carcinoma combined in female rats exposed to 80 ppm were increased but not significantly
58
-------
greater than those in the chamber controls, although they did exceed the historical control range.
Follicular cell carcinomas destroyed the thyroid gland and occasionally invaded the capsule or adjacent
structures. The incidence of follicular cell hyperplasia was significantly increased in male rats exposed
to 32 ppm. Hyperplasia was characterized by one or a few enlarged follicles with several much
smaller follicles inside and to one side.
Table 4-24. Incidence of neoplasms in F344/N rats chronically exposed (2 years) to
chloroprene by inhalation
Tissue Site/Tumor Type
Oral cavity
Papillomas or carcinomas
Thyroid gland
Adenomas or carcinomas
Lung
Adenomas or carcinomas0
Kidney (renal tubules)
Adenomas or carcinomas (extended and
standard evaluations combined)
Mammary gland
Fibroadenomas
Tumor Incidence
Males (ppm)
0
0/50
0/50
2/50
1/50
—
12.8
2/50
2/50
2/50
8/50a
—
32
5/50a
4/49a
4/49
6/50b
—
80
12/50b
5/50a
6/50
8/50b
—
Females (ppm)
0
1/49
1/49
1/49
0/49
24/49
12.8
3/50
1/50
0/50
0/50
32/50
32
5/50
1/50
0/50
0/50
36/50a
80
ll/50b
5/50
3/50
4/50
36/50b
""Statistical significance p < 0.05, p values correspond to the pairwise comparisons between the chamber controls
and the exposed group. The logistic regression test regards lesions, in animals dying prior to terminal kill, as
nonfatal.
bStatistical significance p < 0.01.
0 Adenomas only in females.
Source: NTP (1998. 042076).
The incidences of alveolar/bronchiolar carcinoma and alveolar/bronchiolar adenoma or
carcinoma (combined) in male rats exposed to 80 ppm were slightly greater than those in the chamber
control group. Although the increase in neoplasms was not statistically significantly increased relative
to control, the incidences exceeded the historical control range. The incidence of alveolar/bronchi olar
adenoma only was increased, though not significantly, in female rats exposed to 80 ppm chloroprene.
Alveolar/bronchiolar carcinomas were solid or papillary, obliterated normal pulmonary structure, and
sometimes invaded the pleura and other adjacent areas. The incidences of alveolar epithelial
hyperplasia (AEH) were significantly greater in all exposed groups of male and female rats compared
with the chamber control groups.
Renal tubule adenoma and hyperplasia were observed in male and female rats. Renal tubule
hyperplasia was distinguished from regenerative epithelial changes commonly observed as a part of
nephropathy and was considered a preneoplastic lesion. Hyperplasia was generally a focal, minimal to
mild lesion consisting of lesions that were dilated approximately two times the normal diameter and
were lined by increased numbers of tubule epithelial cells that partially or totally filled the tubule
59
-------
lumen. Because renal tubule neoplasms are rare in chamber control F344/N rats, additional kidney
sections from male and female control and exposed groups were examined to provide a clearer
indication of the potential effects of chloroprene on the kidney. The combined single- and step-section
incidences of renal tubule hyperplasia in males exposed to 32 and 80 ppm and in females exposed to
80 ppm and the incidences of adenoma and adenoma or carcinoma combined in all exposed males
were significantly greater than those in the chamber controls.
The incidences of multiple fibroadenoma of the mammary gland in all exposed groups of
female rats were greater than in the chamber control group. The incidences of fibroadenoma in
females exposed to 32 and 80 ppm were significantly greater than in the chamber control group.
However, the incidences of fibroadenomas in all exposed females and the chamber control exceeded
the historical control range.
A slight increase in the incidence of transitional epithelium carcinoma of the urinary bladder
was observed in female rats exposed at 80 ppm. In addition, one male exposed at 32 ppm had a
transitional epithelium carcinoma and one male exposed at 80 ppm had a transitional cell papilloma.
No urinary bladder neoplasms have been observed historically in chamber control male or female
F344/N rats.
The incidences of atrophy, basal cell hyperplasia, metaplasia, and necrosis of the olfactory
epithelium in males and females exposed to 32 and 80 ppm and of atrophy and necrosis in males
exposed to 12.8 ppm were significantly greater than those in the chamber control groups. The
incidences of chronic olfactory inflammation were significantly increased in males exposed to 12.8 or
32 ppm and in females exposed to 80 ppm. The incidences of fibrosis and adenomatous hyperplasia of
the olfactory epithelium in males and females exposed to 80 ppm were significantly greater than those
in the chamber controls. Lesions of the nasal cavity were generally minimal to moderate in average
severity. Necrosis of the olfactory epithelium was characterized by areas of karyorrhexis and
sloughing of olfactory epithelium with cell debris in the lumen of the dorsal meatus. Atrophy of the
olfactory epithelium was characterized by decreased numbers of layers of olfactory epithelium and
included loss of Bowman's glands and olfactory axons in more severe cases. Metaplasia was
characterized by replacement of olfactory epithelium with ciliated, columnar, respiratory-like
epithelium. Basal cell hyperplasia was characterized by proliferation or increased thickness of the
basal cell layer in the turbinate and septum. No histopathological effects were observed in the nasal
respiratory epithelium of exposed rats.
NTP 2-Year Exposure (Mouse). In the NTP 2-year mouse study, exposure concentrations
were 0, 12.8, 32, and 80 ppm. The actual chamber concentrations to which the animals were exposed,
were 0, 12.7 ± 0.4, 31.9 ± 0.9, and 79.7 ±1.7 ppm chloroprene. The highest exposure concentration in
the 2-year chronic study was chosen based on the observation of mortality in mice exposed to 200 ppm
chloroprene in the 16-day study. The observation of squamous epithelial hyperplasia in the
forestomach of mice exposed to 80 ppm in the 13-week study was not considered life-threatening. All
mice were observed twice daily and body weights were recorded initially, weekly through week 12,
60
-------
approximately every 4 weeks from week 15 through week 91, and every 2 weeks until the end of the
study. Clinical findings were recorded initially, at weeks 4, 5, 8, 12, every 4 weeks through week 91,
and every 2 weeks until the end of the study. A complete necropsy and a microscopic examination
were performed on all mice as described for the rat portion of the 2-year study. Estimates of 2-year
survival probabilities are shown in Table 4-25.
Table 4-25. 2-year survival probabilities for B6C3Fi mice chronically exposed
(2 years) to chloroprene by inhalation
Sex
-------
multiple alveolar/bronchiolar adenoma and alveolar/bronchiolar carcinoma were increased in all males
and females exposed to chloroprene. The morphology of lung neoplasms was similar in control and
exposed groups. The incidences of bronchiolar hyperplasia in all exposed groups of males and females
were significantly greater than in the chamber control groups. Bronchiolar hyperplasia was
characterized by diffuse thickening of the cuboidal cells lining the terminal bronchioles and in some
cases caused papillary projections into the lumen. The incidences of histiocytic cell infiltration in
males exposed to 80 ppm and in all exposed females were significantly increased relative to chamber
controls. This change consisted of histiocytes within alveolar lumens, usually adjacent to
alveolar/bronchiolar neoplasms.
62
-------
Table 4-26. Incidence and severity of nonneoplastic lesions in B6C3Fi mice
chronically exposed (2 years) to chloroprene by inhalation
Tissue Site/Lesion Type
Lung
Bronchiolar Hyperplasia
Histiocytic Cell Infiltration
Kidney (renal tubule)
Hyperplasia
Mammary Gland
Hyperplasia
Forestomach
Epithelial Hyperplasia
Olfactory
Suppurative Inflammation
Atrophy
Metaplasia
Spleen
Hematopoietic Proliferation
Lesion Incidence (Severity)3
Males (ppm)
0
0/50
7/50
(1.6)
2/50
(2.0)
-
4/50
(3.0)
2/50
(2.0)
7/50
(1.1)
6/50
(1.0)
26/50
12.8
10/50C
(2.0)
8/50
(3.3)
16/49C
(1.4)
-
6/48
(1.8)
1/48
(1.0)
8/48
(1.4)
5/50
(1.4)
22/49
32
18/50C
(1.7)
11/50
(2.5)
17/50C
(1.6)
-
7/49
(2.3)
4/50
(1.0)
7/50
(1.1)
5/50
(1.0)
35/50d
80
23/50c
(2.2)
22/50c
(2.9)
18/50C
(1.6)
-
29/50c
(2.2)
6/50
(1.5)
49/50c
(2.5)
49/50c
(2.5)
31/50d
Females (ppm)
0
0/50
1/50
(3.0)
-
0/49
4/50
(2.0)
0/50
6/50
(1.2)
2/50
(1.0)
13/50
12.8
15/49C
(2.0)
14/49C
(2.0)
-
1/49
(1.0)
3/49
(3.7)
1/49
(1.0)
5/49
(1.2)
3/49
(1.0)
25/49d
32
12/50C
(2.2)
18/50C
(2.3)
-
1/50
(1.0)
8/49
(1.6)
3/49b
(1.7)
4/49
(1.3)
1/49
(2.0)
42/49d
80
30/50C
(2.2)
23/50c
(2.4)
-
3/50
(2.0)
27/50c
(2.7)
4/50c
(1.5)
47/50c
(2.0)
44/50c
(2.0)
39/50d
""Severity of lesions graded as: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked, average severity reported in
parenthesis, average severity not reported for splenic hematopoietic proliferation, see Table B-l for severity scores
for splenic hematopoietic proliferation in male and female mice for 0 and 12.8 ppm.
bStatistical significance p < 0.05, p values correspond to the pairwise comparisons between the chamber controls
and the exposed group. The logistic regression test regards lesions, in animals dying prior to terminal kill, as
nonfatal.
0 Statistical significance p < 0.01.
dSignificantly increased relative to controls, level of significance not reported.
Source: NTP (1998. 042076).
The incidences of olfactory epithelial atrophy, adenomatous hyperplasia, and metaplasia in
males and females exposed to 80 ppm were significantly increased compared to those in the chamber
controls. The incidence of suppurative inflammation in females exposed to 32 and 80 ppm was
significantly greater than controls. Atrophy and metaplasia of the olfactory epithelium was similar to
lesions observed in rats exposed to chloroprene. Adenomas of the respiratory epithelium were present
in one female exposed to 32 ppm and one male exposed to 80 ppm.
In male mice, a pattern of nonneoplastic liver lesions along with silver-staining helical
organisms within the liver was observed, consistent with Helicobacter hepaticus infection.
63
-------
Polymerase chain reaction-restriction fragment length polymorphism based assay confirmed an
organism compatible with H. hepaticus. Historically, NTP studies with H. hepaticus associated
hepatitis showed increased incidences of hemangiosarcoma in male mice. Therefore,
hemangiosarcomas of the liver were excluded from the analyses of circulatory neoplasms in the males
in the chloroprene 2-year study. However, even with this exclusion, the combined occurrence of
hemangioma or hemangiosarcoma at other sites was significantly increased in all males exposed to
chloroprene and in females exposed to 32 ppm. The incidences of neoplasms at other sites were not
considered to have been significantly impacted by the infection with H. hepaticus or its associated
hepatitis. Hepatocellular carcinoma was significantly increased relative to control in all exposed
female mice as was hepatocellular adenoma or carcinoma combined in females exposed to 32 and
80 ppm.
Table 4-27. Incidence of neoplasms in B6C3Fi mice chronically exposed (2 years)
to chloroprene by inhalation
Tissue Site/Tumor Type
Lung
Adenomas or carcinomas
All Organs
Hemangiomas or hemangiosarcomas
Harderian gland
Adenomas or carcinomas
Kidney (renal tubules)
Adenomas or carcinomas (extended and
standard evaluations combined)
Mammary gland
Carcinomas (included multiple)
Forestomach
Papillomas or carcinomas
Liver
Adenomas or carcinomas
Skin
Sarcoma
Mesentery
Sarcomas
Zymbal's gland
Carcinomas
Tumor Incidence
Males (ppm)
0
13/50
3/50
2/50
0/50
—
1/50
43/50
...
12.8
28/503
14/50b
5/50
2/49
...
0/50
37/50
...
32
36/503
23/503
10/50C
3/50c
...
2/50
42/50
...
80
43/503
21/503
12/50b
9/50b
...
5/50
41/49
...
Females (ppm)
0
4/50
4/50
2/50
...
3/50
1/50
20/50
0/50
0/50
0/50
12.8
28/49a
6/50
5/50
...
4/50
0/50
26/49
ll/50b
4/50
0/50
32
34/503
18/50b
3/50
...
7/50
0/50
20/50C
11/503
8/50b
0/50
80
42/503
8/50
9/50c
...
12/50C
4/50
30/503
18/503
3/50
3/50
""Statistical significance p < 0.001.
bStatistical significance p < 0.01.
Statistical significance p < 0.05, correspond to the pairwise comparisons between the chamber controls and that
exposed group. The logistic regression test regards lesions, in animals dying prior to terminal kill, as nonfatal.
Source: NTP (1998, 042076).
64
-------
The incidences of Harderian gland adenoma or carcinoma combined in males exposed to 32 or
80 ppm and females exposed to 80 ppm were significantly greater than in the chamber controls. The
incidences of Harderian gland adenoma or carcinoma combined in these groups exceeded the historical
control range.
Although not significantly increased, the incidence of renal tubule adenoma in males exposed
to 80 ppm was greater than in the chamber control group. The incidence of this rare neoplasm
exceeded the historical control range. The incidences of renal tubule hyperplasia in males exposed to
32 or 80 ppm were significantly greater than in the chamber controls. The morphology for renal tubule
hyperplasia was similar to that observed in rats exposed to chloroprene. The combined single- and
step-section incidence of renal tubule adenoma in males exposed to 80 ppm and the combined
incidences of renal tubule hyperplasia in all groups of exposed male mice were greater than in the
chamber controls.
The incidences of mammary gland carcinoma in females exposed to 80 ppm were significantly
greater than in the chamber control group. The incidences of mammary gland carcinoma in females
exposed to 32 and 80 ppm exceeded the historical control range. Mammary gland hyperplasia was
present in a few females exposed to chloroprene, but was not significantly increased relative to
chamber controls.
The incidence of forestomach squamous cell papilloma in females exposed to 80 ppm was
greater than in the chamber controls but statistically not significant. The incidence observed exceeded
the historical control range. In male and female mice exposed to 80 ppm, the incidences of hyperplasia
of the forestomach epithelium were significantly greater than in chamber controls, and the lesions were
similar to those seen in the 13-week study. Hyperplasia was a focal to multifocal change characterized
by an increase in the number of cell layers in the epithelium.
The incidences of sarcoma of the skin were significantly greater in all exposed female mice
compared with chamber controls. The incidences of sarcomas of the mesentery were increased in all
exposed female mice, with only the mice in the 32 ppm exposure group exhibiting a significant
increase.
Carcinomas of Zymbal's gland were observed in three females exposed to 80 ppm chloroprene,
and two carcinomas had metastasized to the lung. Zymbal's gland carcinomas have not been reported
in the NTP historical database for control female mice.
Single papillary adenomas were observed in the trachea of two mice: one male exposed to
12.8 ppm and one to 32 ppm. These adenomas have not been documented in the NTP historical
database.
The incidences of splenic hematopoietic proliferation in males exposed to 32 and 80 ppm and
in all exposed groups of females were significantly greater than in the chamber controls.
Because of a large number of early deaths of mice exposed to chloroprene for 2-years, survival-
adjusted neoplasm rates were estimated by NTP using the Poly-3 survival-adjusted quantal response
method of Portier and Bailer (1989, 093236). This adjustment accounts for the effects of early
65
-------
mortality on the expression of late-developing neoplasms and provides a clearer indication of
exposure-response relationships for neoplasms induced by chloroprene (Table 4-28). The neoplasm
incidence values provided represent the ratio of the number of animals in an exposure group bearing
the specific neoplasm relative to the adjusted number of animals at risk.
Table 4-28. Survival-adjusted neoplasm rates for mice in the 2-year inhalation
study of chloroprene
Tissue Site/Tumor Type
Lung
Adenoma or carcinoma
Alveolar/bronchiolar adenoma or
carcinoma
All Organs
Hemangioma or hemangiosarcoma
Harderian gland
Adenoma or carcinoma
Kidney (renal tubules)
Adenoma
(single section)
(single + step section)
Mammary gland
Adenoacanthoma or carcinoma
Forestomach
Squamous cell papilloma or carcinoma
Liver
Carcinoma
Adenoma or carcinoma
Skin
Sarcoma
Mesentery
Males (%)a'b'c
0
14.1d
29.8d
2.4d
4.7d
Oe
od
2.4d
12.8
28.3
63.7d
28.2d
12.0
2.4
4.8
0
32
56.9d
79.2d
45.2d
26.3d
2.8
8.3
5.6
80
66.4d
92.9d
43.6d
32.0d
8.2
23. 9e
13.3
Females (%)a'b'c
0
4.6d
9.1d
9.0e
4.5d
6.7d
2.3d
9.0d
44.8
Oe
n
12.8
35.6d
68.3d
16.0
13.5
12.9
0
28.4e
62.9
27.5d
in 7e
32
53.8d
85. 8d
53.1d
11.7
33.7d
0
47.5d
63.3
39.0d
98 Qd
80
76.0d
96. ld
27.7e
31.2d
42.5d
14.6
58.2d
79.7d
52.6d
11 n
""Survival-adjusted neoplasm rates were estimated using the Poly-3 survival-adjusted quanta! response method of
Portier and Bailer (1989, 093236).
bln the chamber control column (0 ppm), statistically significant trends across all exposure groups by the Poly-3
quanta! response test are indicated.
°In the exposed group columns, statistically significant differences from the chamber control group (by pairwise
comparison) are indicated.
dp<0.01.
ep<0.05.
Source: NTP (1998. 042076)
Other Inhalation Studies. In another chronic inhalation study, Trochimowicz et al. (1998,
625008) exposed three groups of 100 Wistar rats and Syrian hamsters of each sex to chloroprene at 0,
10, or 50 ppm for 6 hours/day, 5 days/week for up to 18 months (hamsters) or 24 months (rats).
Chemical purity of the bulk chloroprene was reported to be 99.6%, with less than 50 ppm of dimers as
determined by gas chromatography. Bottles of test material were received weekly and were stored
66
-------
under nitrogen at -20°C. Phenothiazine (0.01%) was added to prevent oxidation. A fresh sample of
chloroprene from cold storage was used to generate the test atmosphere for each day's exposure. To
generate the test atmospheres, bulk material was vaporized with dried and filtered nitrogen at 0°C;
vaporization at this temperature was performed to inhibit the formation of degradation products. The
saturated chloroprene/nitrogen mixture was then directed into the inhalation chamber inlet, where it
was mixed with the main air flow to generate the desired exposure concentration. All animals were
observed daily and clinical signs and mortality were recorded. Rats and hamsters were weighed
immediately before the first exposure, weekly for the first 8 weeks of exposure, and at 4-week intervals
for the remainder of the experiment. During the last 6 months of each study, all animals were
examined once a month for the presence of tumors. Time of tumor appearance, size, location, and
progression were recorded. At study termination, both hamsters and rats were sacrificed by
exsanguination of abdominal aorta. A postmortem examination was conducted during which all major
organs/tissues were examined for gross abnormalities. Gross pathological examinations were
conducted on all animals, including those that died intercurrently or were killed in extremis, unless
advanced autolysis or cannibalism prevented this. The following organs were weighed: adrenals, brain
(hamster), heart, kidneys, liver, lungs with trachea and larynx, ovaries, pituitary, spleen, testes, and
thyroid (rat). The following organs/tissues were preserved and examined microscopically: all gross
lesions, adipose tissue, aorta (rat), epididymides, external auditory canal with Zymbal's glands, eyes,
exorbital lachrymal glands, femur (with knee joint), gastrointestinal tract (esophagus, stomach,
duodenum, jejunum, ileum, cecum, and colon), lungs, lymph nodes (auxiliary, cervical, and
mesenteric), mammary glands, nasal cavity (four transverse sections), pancreas, parotid salivary
glands, preputial glands, prostate, sciatic nerve, seminal vesicles, skeletal muscle, skin, spinal cord,
sternum (bone marrow), sublingual and submaxillary salivary glands, thymus, thyroid with parathyroid
(hamster), urinary bladder, and uterus. Microscopic examinations were performed on all organs from
all control and high-exposure animals, and on the liver, spleen, pituitary gland, thyroid glands,
adrenals, and all grossly visible tumors and tumor-like lesions from the low-exposure animals.
Mortality for rats was low in all groups up to week 72, ranging from 1-3%. During week 72,
however, 87 males and 73 females of the 10-ppm exposure group died overnight from suffocation from
an accidental failure of the exposure chamber ventilation system. For hamsters, mortality was
negatively correlated with the exposure concentration of chloroprene. At the termination of exposure,
survival rates in the 0, 10, and 50-ppm groups were 88, 92, and 93% in males and 63, 75, and 72% in
females, respectively.
Slight but consistent growth retardation was found in male rats (-10%) and female rats (-5%)
in the 50-ppm exposure group. Both male and female hamsters showed a slight growth depression in
the 50-ppm group throughout the study. Rats were not affected by exposure to chloroprene in regard to
appearance or behavior, except that alopecia occurred more frequently in the 50-ppm group than in the
10-ppm group or in the controls. The alopecia varied from small, focal, mostly bilateral bald areas to
severe, diffuse, generalized hair loss. Alopecia was first observed after an exposure period of about
67
-------
10 weeks, but by 25 weeks the incidence and degree of alopecia gradually decreased and in many
animals complete re-growth of hair was observed. No abnormalities were observed in hamsters;
alopecia was occasionally seen in each group during the first 64 weeks of study, regardless of
exposure.
Body and organ weights are given in Table 4-29. In both male and female rats, mean relative
lung weights were significantly lower in both exposure groups than in controls. In females exposed to
50 ppm, the mean relative spleen and thyroid weights were significantly lower. The kidney and
pituitary weights in males exposed to 10 ppm were significantly increased compared with controls,
although this was not observed in the 50 ppm exposure group. In hamsters, both male and female
animals exposed to 50 ppm had significantly higher brain weights compared with controls. Relative
lung weight was significantly higher in males exposed to 50 ppm than in controls.
Table 4-29. Selected mean relative organ weights of rats exposed for 24 months
and hamsters exposed for 18 months to chloroprene vapor
Group
(ppm)
Number3
BW(g)
Adrenals
Brain
Kidneys
Liver
Lungs
Spleen
Thyroid
Rats
Males
0
10
50
77
9
76
494
500
496
—
—
—
—
—
—
0.61
0.68b
0.64
3.09
3.31
3.15
0.45
0.37C
0.38C
0.154
0.172
0.146
0.0056
0.0056
0.0056
Females
0
10
50
81
19
75
308
309
307
—
—
—
—
—
—
0.64
0.65
NRd
3.00
3.23b
3.13b
0.53
0.45b
0.45b
0.180
0.176
0.164C
0.0080
0.0073
0.0070C
Hamsters
Males
0
10
50
86
92
92
101
101
93
0.0311
0.0279b
0.0294
1.10
1.11
1.19e
1.25
1.17°
1.22
5.11
4.75b
4.91
0.85
0.84
0.90C
0.197
0.190
0.174C
—
—
—
Females
0
10
50
60
74
72
99
98
90
0.0340
0.0356
0.0383
1.13
1.16
1.24e
1.48
1.50
1.50
6.73
6.54
6.37
1.01
0.97
1.01
0.253
0.269
0.286
—
—
—
aNumber at sacrifice.
bStatistically significant, 0.1 < p < 0.005.
"Statistically significant, 0.001 < p < 0.01.
dNot recorded.
"Statistically significant, p < 0.001.
Source : Trochimowicz et al. (1998, 625008).
68
-------
Gross pathology revealed that lungs from rats exposed at 10 and 50 ppm had markedly lower
incidences of nodular pieural surfaces, consolidation, and atelectasis (gross changes consistent with,
and characterized as chronic respiratory disease) than did controls. These morphologic indicators of
chronic respiratory disease were seen in 28 of 196 controls, 0 of 37 in the 10-ppm group, and 4 of 200
in the 50-ppm group. The incidence of tumors or tumor-like lesions of the mammary glands was
slightly higher in the exposed animals terminated at the end of the study (10/24 and 34/100 in 10 and
50 ppm, respectively) compared with controls (23/99). These differences were not statistically
significant unless animals that were moribund or dead before the terminal sacrifice were included in
the analysis. No other remarkable differences in gross pathology were seen in rats. Macroscopic
examination of hamsters revealed a slight, concentration-related decrease in the incidence of pale
adrenal glands in males.
The only nonneoplastic lesions in rats were observed in liver and lungs (only the livers of
animals that died accidentally due to a failure in the ventilation system were available for microscopic
examination). The number of female and male rats with one or more small foci of cellular alteration in
the liver was significantly increased in the 50-ppm group than in controls. Mild changes, such as
lymphoid aggregates around bronchi, bronchiole, and blood vessels, were observed in males and
females exposed to 50 ppm. Acute inflammatory processes in the lungs of control and high-dose
animals were observed to be similar.
The only nonneoplastic effect observed in hamsters was a generalized amyloidosis (in the liver,
kidneys, spleen, and adrenals); this effect was lower in incidence in the 50 ppm exposed group
compared with controls.
Tumor incidences for rats and hamsters are shown in Tables 4-30 and 4-31, respectively. Only
mammary gland tumors and squamous cell carcinomas were observed to demonstrate a statistically
significant excess in rats exposed to chloroprene, compared with controls. Mammary tumors were
significantly increased (p < 0.05) in females in the 50-ppm group. The observed increase in mammary
tumors in the high dose animals was due to the inclusion in the analysis of animals that were moribund
or dead before the terminal sacrifice. No difference was observed between control and test group
animals that were sacrificed at the end of the study. The number of mammary tumors per rat was not
different between the 50-ppm group and the control group. The relatively high number of chloroprene-
exposed animals bearing benign fibroadenomas was primarily responsible for the increased incidence
of mammary tumors. Squamous cell carcinomas involving the nasal cavity, sinus maxillaries, subcutis,
and skin were observed in 3 of 100 males of the 50-ppm group and in 1 of 99 females of the control
group. The exact origin of these tumors could not be identified through macroscopic or microscopic
examination. If they originated as skin tumors, the total number of squamous-cell carcinomas of the
skin would have been 5/100 in the 50-ppm group, which would be a statistically significant (p < 0.05)
increase over controls (1/97).
In the hamster, the incidences of cystadenomatous polyps of the gallbladder and
pheochromocytoma were slightly, but significantly, elevated in the males exposed to 10 ppm. All other
69
-------
tumors observed were about equally distributed among test and control groups or occurred in only one
or two hamsters.
Sanotskii (1976, 063885) provided a review of numerous Russian subchronic inhalation studies
of chloroprene (chemical purity and exposure regimen not specified) in rats and mice. According to
Sanotskii (1976, 063885), the studies evaluated the systemic effects of chloroprene exposure in rats
(strain not specified) exposed for 4.5 months to 0.051, 0.15, and 1.69 mg/m3 (0.014, 0.041, and
0.47 ppm) or C57BL/6 mice exposed for 2 months to concentrations as high as 35 mg/m3 (9.7 ppm).
Several "signs of systemic effect" in male rats were reported at 1.69 ± 0.087 mg/m3, including an
increase in a "summation threshold index" (not defined) after 2.5 and 4.5 months, a decrease in the
synthesis of hippuric acid from sodium benzoate (Quick's test) at 4.5 months, and an inhibition of gas
exchange after 4.5 months. Chloroprene was reported to have had no effect on "the indicators used in
the tests" (i.e., summation threshold index, hippuric acid synthesis, and inhibition of gas exchange) in
mice at concentrations as high as 35 ± 0.7 mg/m3 (9.7 ppm).
Table 4-30. Incidence, site and type of tumor in selected organs and tissues of rats
exposed to chloroprene for 24 months
Site And Type Of Tumor3
Initial number of rats
Number examined
Number tumor-bearing13
Total number primary tumorsb
Hematopoietic system
Lymphoid leukemia
Monocytic leukemia
Kidneys
Lipoma
Adenocarcinoma
Liver
Unidentified
Lungs
Anaplastic carcinoma
Mammary glands
Adenoma
Fibroadenoma
Adenocarcinoma
Papillary carcinoma
Unidentified tumor
Skin
Squamous cell carcinoma
Skin, nasal cavity, maxillary sinus,
Squamous cell carcinoma
Spleen
Hemangiosarcoma
Subcutis, nasal cavity, or maxillary sinus
Reticulum cell sarcoma
Testes
Leydig cell tumor
Males
0 ppm
100
97
51
73/51
1
0
0
0
0
0
—
—
—
—
—
0
0
0
0
2
10 ppm
100
13
6
6/6
0
0
0
0
0
0
—
—
—
—
—
0
0
0
0
1
50 ppm
100
100
57
77/57
2
1
1
1
0
0
—
—
—
—
—
2
3
1
0
4
Females
0 ppm
100
99
66
100/66
0
0
1
0
1
1
3
24
5
1
1
0
1
0
0
—
10 ppm
100
24
12
13/12
0
0
0
0
0
0
1
6
0
0
2
0
0
0
0
—
50 ppm
100
100
74
96/71
1
0
1
0
0
0
7
36
3
0
0
0
0
0
1
—
70
-------
Site And Type Of Tumor3
Testes/epididymides
Mesothelioma
Thyroid gland
Parafollicular cell adenoma
Small
Medium/large
Parafollicular cell carcinoma
Small
Large
Follicular adenoma
Small
Large
Papillary carcinoma
Urinary bladder
Transitional cell carcinoma (metastasizing)
Zymbal's gland
Adenoma
Males
0 ppm
1
6
3
1
1
2
2
0
0
0
10 ppm
0
0
1
0
0
0
0
0
0
0
50 ppm
0
8
o
J
0
0
2
1
0
1
0
Females
0 ppm
—
11
3
0
0
0
0
0
0
0
10 ppm
—
0
1
0
0
0
0
0
0
0
50 ppm
—
14
4
0
0
3
0
2
0
1
""Multiple tumors at one site were counted as one tumor.
bSome animals had more than one tumor.
Source: Used with permission from Taylor and Francis, Trochimowicz et al. (1998, 625008).
Table 4-31. Incidence, site and type of tumor in selected organs and tissues of
hamsters exposed to chloroprene for 18 months
Site And Type Of Tumor3
Initial number of hamsters
Number examined
Number tumor bearing3
Total number primary tumors3
Kidney
Cortical adenocarcinoma
Liver
Neoplastic (hepatocellular) nodule
Unidentified tumor-like lesion
Lung tumors
Gallbladder
Cystadenomatous polyp
Pancreas
Islet-cell adenoma
Islet-cell adenocarcinoma
Stomach
Papilloma
Unidentified papilloma-like lesion
Testes
Leydig-cell tumor
Males
0 ppm
100
100
14
15/14
2
0
0
0
1
1
0
0
1
1
10 ppm
100
97
17
18/17
0
1
1
0
63
0
0
0
1
0
50 ppm
100
97
20
23/20
0
0
0
0
1
2
0
2
1
0
Females
0 ppm
100
94
10
11/11
0
0
0
0
1
0
1
0
1
—
10 ppm
100
93
11
11/11
0
0
0
0
2
0
0
0
2
—
50 ppm
100
97
15
18/15
0
0
1
0
3
0
1
0
0
—
71
-------
Site And Type Of Tumor3
Colon
Adenomatous polyp
Pituitary
Adenoma
Thyroid gland
Parafollicular cell adenoma
Cystadenoma
Papillary adenoma
Follicular adenoma
Parathyroid
Adenoma
Adrenals
Cortical adenoma
Cortical carcinoma
Pheochromocytoma
Malignant pheochromocytoma
Ovaries
Granulosa-theca-cell tumor
Parotid salivary glands
Adenoma
Skin
Unidentified tumor-like lesion
Zymbal's gland
Sebaceous adenoma
Depot fat
Lipoma
Nose
Adenoma of Bowman's glands
Adenocarcinoma of Bowman's glands
Bone (ribs)
Osteosarcoma
Abdominal cavity
Reticulum cell sarcoma
Males
0 ppm
0
0
2
1
0
2
0
4
0
0
0
—
0
0
0
0
0
0
0
1
10 ppm
0
0
0
0
1
1
0
1
1
4b
0
—
0
1
0
0
0
0
0
0
50 ppm
0
1
0
0
1
0
0
10
0
2
2
—
0
0
1
0
0
0
0
0
Females
0 ppm
1
2
0
0
1
1
0
0
1
0
0
0
0
0
0
0
1
0
1
0
10 ppm
0
0
2
0
0
2
1
0
0
0
0
2
0
0
0
0
0
0
0
0
50 ppm
0
0
1
0
2
1
0
3
1
0
0
1
1
0
0
1
0
1
0
0
aSome animals had more than one tumor.
bStatistically significant, p < 0.05 by chi-squared test.
Source: Used with permission from Taylor and Francis, Trochimowicz et al. (1998, 625008).
Dong et al. (1989, 007520) exposed Kumming albino mice (weaned at 2 weeks age) to 0,
2.9 ± 0.3, 19.2 ± 1.9, or 189 ± 13.3 mg/m3 chloroprene for 4 hours/day, 6 days/week for 7 months. The
purity of the chloroprene used to generate the test atmospheres was stated to be 99.8%. Animals were
terminated at the end of the exposure period, or when found moribund. Lung tumors were not
observed in treated animals before the 6th month of exposure, and were observed to increase in
incidence with increasing concentration. The LOAEL for this study was determined to be 2.9 mg/m3
(8.1% incidence of lung tumors versus 1.3% in control animals, p < 0.05). Most lung tumors observed
were papilloadenomas. Induction of multiple tumors in a single animal was also observed to increase
with increasing dose.
72
-------
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND INHALATION
Ponomarkov and Tomatis (1980, 075453) administered chloroprene dissolved in olive oil by
stomach tube to 17 female BD-IV rats at a single dose (100 mg/kg body weight) on gestational day
(GDI7). Progeny from treated females (81 males and 64 females) were treated weekly with 50 mg/kg
body weight by stomach tube from the time of weaning for life (120 weeks). A control group of 14
female rats was treated with 0.3 mL olive oil. Litter sizes and preweaning mortality, survival rates, and
body weights did not differ between chloroprene-treated animals and controls (Section 4.2.1 for further
study details).
NTP (1998, 042076) evaluated sperm morphology and vaginal cytology in rats exposed to 0, 5,
32, or 200 ppm and mice exposed to 0, 12, 32, 80 ppm chloroprene for 13 weeks. Methods used were
those described in the NTP's sperm morphology and vaginal cytology evaluations protocol (NTP,
1985, 625205). Table 4-32 is a summary of measured epididymal spermatozoal and estrous cycle
parameters from these 13-week studies. The sperm motility of male rats exposed to 200 ppm was
significantly less than that of controls. This was the only reproductive tissue or estrous cycle
parameter affected, compared with controls, in rats or mice at any exposure level.
73
-------
Table 4-32. Summary of epididymal, spermatozoal and estrous cycle parameters
for rats and mice in the 13-week study of chloroprene
n
Rats
0 ppm
10
5 ppm
10
32 ppm
9
200 ppm
9
Mice
0 ppm
7
12 ppm
8
32 ppm
10
80 ppm
10
Epididymal spermatozoa - males'1
Motility (%)
Abnormal sperm (%)
Sperm concentration
(106/g cauda epididymal
tissue)
86.73
± 1.04
0.70
±0.05
698
±40
83.62
±1.93
0.78
±0.11
722
±62
82.16
±1.84
0.73
±0.11
689
±46
80.04
± 1.99b
1.02
±0.14
683
± 25
79.09
± 1.20
1.49
±0.42
1,632
±138
81.07
±1.13
1.30
±0.22
1,447
±122
80.08
±1.19
0.98
±0.10
1,575
±104
80.04
±1.47
1.36
±0.22
1,672
±134
Estrous cycle -females'1
Length (days)
Diestrus stage (% of cycle)
Proestrus stage (% of
cycle)
Estrus stage (% of cycle)
Metestrus stage (% of
cycle)
Uncertain diagnosis stage
(% of cycle)
5.00
±0.15
42.9
15.7
18.6
22.9
0.0
4.67
±0.17C
35.7
18.6
22.9
22.9
0.0
5.00
± 0.27d
44.3
11.4
20.0
24.3
0.0
5.33
±0.17C
45.7
17.1
15.7
20.0
1.4
4.00
±0.00
31.4
20.0
24.3
24.3
4.30
±0.21
31.4
20.0
24.3
24.3
4.22
±0.15C
30.0
22.9
25.7
21.4
4.13
±.13d
35.7
25.7
20.0
18.6
— -
aEpididymal spermatozoal parameters, and estrous cycle lengths are presented as mean ± standard error.
bSignificantly different (p < 0.01) from the control group by Shirley's test.
°Estrous cycle was longer than 12 days or unclear in 1 of 10 animals.
dEstrous cycle was longer than 12 days or unclear in 2 of 10 animals.
Source: NTP (1998. 042076).
Sanotskii (1976, 063885) reviewed several Russian studies that exposed white rats (strain
unknown) to various concentrations of chloroprene in order to determine the effect on reproductive and
developmental parameters. In male rats exposed for 4.5 months to 1.7 mg/m3 (0.5 ppm) of
chloroprene, reductions in the number of normal spermatogonia, increases in the percentage of dead
spermatozoa, and decreases in spermatozoal motility were reported. These effects were not observed
by NTP (1998, 042076) in F344 rats at much higher concentrations (Table 4-32). Sanotskii (1976,
063885) also reported an increase in the number of seminiferous tubules with desquamating epithelium
in male C57BL/6 mice exposed to 0.32 mg/m3 (0.09 ppm) for 2 months and increased dominant lethal
mutations in germ cells of male and female C57BL/6 mice exposed to 3.5 mg/m3 (1 ppm) for
2 months.
Sanotskii (1976, 063885) also reported on an embryotoxicity study in which pregnant white
rats were exposed during their "whole period of pregnancy." Exposure to 4 mg/m3 (1.1 ppm)
chloroprene was reported to have resulted in an increase of embryonic mortality, a decrease in fetal
weight, and a disturbance in vascular permeability as evidenced by hemorrhaging into body cavities.
74
-------
Exposure to 0.13 mg/m3 (~ 0.04 ppm) chloroprene was reported to have resulted in increased postnatal
mortality. Exposure to 4 mg/m3 (1.1 ppm) chloroprene at various times during pregnancy was reported
to have resulted in cerebral hernia and hydrocephalus.
Culik et al. (1978, 094969) evaluated the embryotoxic, teratogenic, and reproductive toxicity of
chloroprene in rats. Culik et al. (1978, 094969) exposed pregnant CD rats to chloroprene by inhalation
at 0, 1, 10, or 25 ppm (0.28, 2.8, or 6.9 mg/m3) for 4 hours daily, either on GD1-GD12
(embryotoxicity study) or GD3-GD20 (teratology study). Pregnant rats in these embryotoxicity and
teratology studies were sacrificed and their litters examined on GDI7 and GD21, respectively. Male
rats in a separate reproduction study were exposed to 0 or 25 ppm (0 or 6.9 mg/m3) 4 hours daily for
22 days and bred with untreated females for 8 consecutive weeks. The embryotoxicity study included
200 female rats (50 per exposure group), the teratology study included 100 primigravida rats (25 per
exposure group), and the male reproduction study involved 10 male rats (5 per exposure group) and 3
virgin females per male. The test material was reported to be >99.9% pure and was stored under
nitrogen at -20°C in small glass bottles holding one day's supply for generating atmospheres. No
chemical decomposition was observed during the experiment.
In both the embryotoxicity and teratogenicity studies, litter size, average numbers of
implantation sites per litter, and preimplantation losses among exposed females were not significantly
different from those of the controls (Table 4-33). In the teratology study, there was an increase in the
percentage of litters with resorptions that was statistically significant (p < 0.05, Fisher's exact test)
only in the 10 ppm exposure group (62% compared to 29% in the control group). The percentage of
litters with resorptions was also elevated in the 25-ppm group (59%), although this increase in effect
failed to achieve statistical significance. There was no effect on percentage of litters with resorptions
in any exposure group in the larger embryotoxicity study; all groups had approximately 50% of their
litters exhibiting resorption. The number of resorptions per litters with resorptions was not affected in
either study. The more frequently investigated endpoint of number of resorptions per litter (total) was
not reported by the study, but was calculated from the reported data and included in Table 4-33 for
reference. There was a slight, but statistically significant (p < 0.05), increase in the average body
weight of fetuses from dams exposed to chloroprene at 25 ppm in the teratology study. Fetuses from
dams in the teratology study exposed to 10 and 25 ppm chloroprene were significantly (p < 0.05)
longer than the control fetuses. The incidence of minor anomalies (minute subcutaneous hematomas
and petechial hemorrhages) was similar in fetuses from exposed and control dams (Table 4-34). No
major compound-induced or concentration-related skeletal or soft tissue anomalies were found. The
number of unossified sternebrae and unossified thoracic vertebral centers were similar in all groups
regardless of treatment. The combined results of weekly matings for the 8-week reproduction test
indicated that there were no significant effects on reproduction due to chloroprene exposure: the
mating index, average number of pups per litter, viability index, and lactation index were similar for
exposed and control animals.
75
-------
Table 4-33. Results of teratology and embryotoxicity studies in rats exposed to
chloroprene by inhalation
Parameter
Concentration of Chloroprene (ppm)
0
1
10
25
Teratology Study
Number of litters
Pregnancy rate, %
Corpora lutea/dam
Implantation sites/dam
Median preimplantation loss, %
Live fetuses/litter
Litters with resorption, %
Litters totally resorbed
Median postimplantation loss
in litters with resorption, %
Resorptions/litters with
Resorptions
Resorptions/litters total
Fetal body weight, g
Fetal crown-rump length, mm
21
84 (21/25)
13 ±3
10 ±2
14.7
9±2
29 (6/21)
0
11.8
1.3 (8/6)
0.38 (8/21)
3.76 ±0.28
32.9 ± 1.4
24
96 (24/25)
12 ±2
9±3
29.5
8±3
29 (7/24)
0
16.7
2.0 (14/7)
0.58 (14/24)
3.94 ±0.46
33.7 ±1.6
21
84 (21/25)
12 ±2
9±2
20.0
8±3
62 (13/21)3
0
22.0
1.9 (25/13)
1.19(25/21)
3. 96 ±0.26
33.8±0.7b
19
76 (19/25)
13 ±2
11 ±1
10.0
10 ±1
59(11/19)
0
16.7
1.6(17/11)
0.89 (17/19)
4.04 ± 0.27b
34.1±1.2b
Embryotoxicity Study
Number of litters
Pregnancy rate, %
Corpora lutea/dam
Implantation sites/dam
Median preimplantation loss, %
Live fetuses/litter
Litters with resorption, %
Litters totally resorbed
Median postimplantation loss
in litters with resorption, %
Resorptions/litters with
resorptions
Resorptions/litters total
45
90 (45/50)
15 ±3
11±3
20.0
10 ±3
51 (23/45)
0
9.1
1.7 (39/23)
0.87 (39/45)
43
86 (43/50)
14 ±3
11±4
16.2
9±4
51 (22/43)
1
12.9
2.1 (47/22)
1.09 (47/43)
43
88 (43/49)
14 ±2
10 ±4
17.7
10 ±3
53 (23/43)
0
8.3
1.6 (37/23)
0.86 (37/43)
48
94(48/51)
13 ±3
10 ±3
16.0
10 ±3
50 (24/48)
0
9.1
1.4 (34/24)
0.71 (34/48)
""Significantly different (p < 0.05) from the control group by Fisher's exact test.
bSignificantly different (p < 0.05) from the control group by an analysis of variance and least significant difference
(LSD) test.
Source: Used with permission from Academic Press, Inc., Culik et al. (1978, 094969).
Culik et al. (1978, 094969) concluded that the statistically significant increase in litters with
resorptions observed in the teratology study at 10 ppm was not biologically significant because the
increase at 25 ppm was not statistically significant and the effect was not observed in the
embryotoxicity study, which had larger numbers of animals per exposure group and was specifically
designed to observe such an effect. Further, the control group for the teratology study is the only group
76
-------
in either study (embryotoxicity or teratology) that is far outside of the historical control range for
number of resorptions per litter (0.83 ± 0.34) for this strain of rat (MARTA; and MTA, 1996, 625111):
the corresponding control group in the embryotoxicity study had a response rate equivalent to
historical controls. Therefore, if the control group response in the teratology study is abnormally low,
this may indicate that the statistically significant increase seen in the 10-ppm group may be a spurious
observation. Chloroprene exerts an effect on fetal weight and size, as evidenced by increases in both at
higher exposure levels. However, in the absence of other definitive markers of developmental toxicity,
the importance or adversity of this finding remains unclear. Given the lack of a defined dose-response
for litters with resorptions in either the embryotoxicity or teratology study, and that the control group in
the teratology study may be a statistical outlier compared to historic control data, there is no
compelling evidence that chloroprene displays developmental effects in CD rats at exposure levels up
to 25 ppm. Therefore, 25 ppm is identified as the NOAEL for this study.
Table 4-34. Incidence of anomalies in litters of rats exposed to chloroprene by
inhalation
Parameter
Gross anomalies
Soft tissue anomalies
Skeletal anomalies
Gross anomalies
Runts3
Small subcutaneous hematomas
Petechial hemorrhages
Soft tissue anomalies
Hydronephrosis
Subcutaneous edema
Skeletal anomalies
Delayed ossification of one or more
sternebrae
14th rudimentary ribs(s) or spur(s)
Wavy ribs
Bipartite thoracic centra
Concentration of Chloroprene (ppm)
0
1
10
25
Number of litters (fetuses) examined
21 (192)
21 (66)
21 (126)
24(191)
24 (69)
24 (122)
21 (172)
21 (60)
21(112)
19(184)
19 (62)
19 (122)
Number of litters (fetuses) affected
1(1)
5(5)
5(5)
8(9)
0
17 (58)
20 (91)
4(4)
2(2)
0
9(9)
2(6)
4(6)
1(1)
15 (39)
22 (76)
4(5)
2(3)
1(1)
4(4)
3(3)
1(1)
0
13 (33)
20 (67)
2(3)
2(2)
1(1)
6(10)
2(2)
5(7)
0
14 (45)
19 (77)
3(4)
4(8)
aBody weight less than control mean weight minus 3 standard deviations.
Source: Used with permission from Academic Press, Inc., Culik et al. (1978, 094969).
Mast et al. (1994, 625206) exposed groups of 15-16 pregnant New Zealand White (NZW)
rabbits by inhalation to 10, 40, or 175 ppm chloroprene (36.2, 144.8, or 633.5 mg/m3) for 6 hours/day
on GD6-GD28. Maternal body weights were measured on gestational days 0, 6, 15, 22, and 29 and
77
-------
animals were observed twice daily (7 days/week) during the exposure period for signs of illness or
mortality. On GD29, dams were sacrificed and examined for gross tissue abnormalities. Maternal
kidneys and liver were removed and weighed. The uterus was removed and weighed, and the number,
position, and status (live, resorbed, or dead) of implants were recorded. Live fetuses were weighed and
examined for gross, visceral, and skeletal defects. Bulk chemical analysis was performed using
infrared spectroscopy to confirm test material identity. Purity and dimer determinations were
conducted by gas chromatography. Exposure atmospheres were generated by immersing an
evaporation flask containing bulk material in a 150°F water bath and passing a metered flow of
nitrogen through the flask to a condenser. The condenser's temperature was maintained at -2°C in
order to control the chloroprene vapor concentration, and to remove low volatility impurities from the
vapor. From the condenser, the chloroprene vapor was mixed with an appropriate amount of
compressed air in order to achieve the desired exposure concentration. The normal exposure
concentrations in the study were between 98-100% target concentrations, and there was no evidence of
degradation products greater than 0.1% target concentration.
There were no signs of maternal toxicity due to exposure to chloroprene. A few NZW dams in
each group exhibited nasal discharge, vaginal bleeding, and loose stools at various times during the
exposure period. The overall pregnancy rate was 89%, with a range of 80-94% for each exposure
group. The incidence of clinical signs of toxicity was low during the exposure, and dams appeared to
be in excellent health at termination. No exposure-related effects on maternal weight change were
noted. Exposure to chloroprene had no effect on the number of implantations, live pups, or
resorptions. Fetal body, liver, and kidney weights were not affected by exposure. The incidence of
fetal malformations was not affected by exposure to chloroprene. The results of this study indicate that
exposure to chloroprene on GD6-GD28 in rabbits results in no observable developmental toxicity,
therefore the high-exposure group, 175 ppm, was identified as the NOAEL for this study.
In an unpublished report, Appelman and Dreef van der Meulen (1979, 064938) exposed two
successive generations (F0 or FI) of Wistar rats to 0, 10, 33, or 100 ppm (0, 36.2, 119.5, or 362 mg/m3)
chloroprene. In the F0-generation, groups of 25 males and females were exposed to chloroprene for
6 hours/day, 5 days/week for 13 weeks. After the termination of the exposure, the treated animals were
caged and mated with untreated stock animals for 20 days (1 male per 1 female). After the mating
period, the animals were separated: males were sacrificed and their testes were collected and
examined whereas females were caged individually and allowed to birth and rear their litters. After
their litters were weaned, the females were sacrificed and their uteri were collected and examined for
implantation sites. The number of pups in each litter was recorded at birth, as well as the total number
of survivors and total litter weight at postnatal days 1, 3, 14, and 28. Litters containing more than 8
siblings were randomly culled to that number at day 4. From the Fi-litters, 20 males and females were
selected randomly from each exposure group one week after weaning and exposed to the same
concentrations of chloroprene for 10 weeks (6 hours/day, 5 days/week). In both the FO and Fl rats, the
general condition, behavior, and signs of possible intoxication were checked daily and all signs of
78
-------
illness or reaction to exposure were recorded. Individual body weights were recorded weekly during
exposure. In theFl rats, blood samples were collected from 15 rats/sex/exposure group at an age of
4 weeks and analyzed for hemoglobin concentration. At the end of the exposure period, 10 Fl
rats/sex/exposure group were sacrificed and their liver, lungs, and gonads were weighed and examined.
The general condition and behavior of FO rats did not differ between exposure groups. At
100 ppm, slight (less than 10% decrease relative to control), but significant, growth retardation was
observed in males in weeks 3, 6, 7, 8, and 10 and in females from week 2 to termination of exposure
(p < 0.05). There were statistically significant decreases in body weights in both sexes at various time
points in the low and mid-exposure groups compared to controls, but no consistent exposure-related
pattern was observed. No data on food consumption were provided, but the authors note that decreases
in body weight were most likely attributable to occasional shortages in food availability. The
percentage of females (exposed and non-exposed) that successfully mated was not affected by
chloroprene exposure. Sex ratios, mortality during lactation, and resorption quotients were not
significantly altered in any exposure group. The body weight of offspring descended from treated
females and untreated males was statistically reduced in the high-exposure group. Body weights of
offspring descended from treated males and untreated females were not affected.
The general condition and behavior of Fl rats did not differ between exposure groups.
Statistically significant decreases in body weight (greater than 10% reduction compared to control)
were observed in females descended from treated females during week 1 of exposure (p < 0.01), in
males descended from treated males during weeks 4, 6, 7, and 10 (p < 0.01), and in females descended
from treated males during weeks 5 and 6 (p < 0.01). Again, no food consumption data were provided,
precluding a determination of whether these decreases in body weight were related to exposure.
Hemoglobin levels were not affected by exposure. The relative weights of testes from Fl males were
statistically increased in all exposure groups in males descended from treated females (p < 0.05 at 10
and 33 ppm, p < 0.01 at 100 ppm) and at 33 and 100 ppm in males descended from treated males
(p < 0.05). Fl females descended from treated males and exposed to 100 ppm chloroprene had
significantly increased liver (p < 0.01), ovary (p < 0.001), and lung (p < 0.05) weights. Gross and
microscopic histopathological examinations revealed no treatment-related abnormalities in these organ
systems. Given the lack of histopathological findings in any examined organ system, the significant
increases in lung, liver, and gonad weights in Fl males and females are not considered to be adverse.
The NOAEL for this study was identified as 33 ppm based on decreases in body weight during
lactation in pups descended from treated females and untreated males.
4.4. OTHER DURATION- OR ENDPOINT-SPECIFIC STUDIES
4.4.1. Acute and Subchronic Studies
Clary et al. (1978, 064942) conducted a study to investigate the acute and subchronic toxicity
of chloroprene and to determine the dose range for a 2-year chronic inhalation study (chronic study by
Trochimowicz et al. (1998, 625008)) in rats and hamsters. Groups of 6 male albino rats (from Charles
79
-------
River laboratories) were exposed to chloroprene by the dermal (200 mg/kg), oral (50 mg/kg), or
inhalation (2 mg/L [-550 ppm]) routes for 4 hours and sacrificed for histological examinations 14 days
after exposure. This exposure protocol was referred to as a "modified Class B poison test" (extension
of sacrifice from 2-14 days after exposure). A lethal concentration test was also conducted by
exposing male rats to 0, 530, 1,690, 2,280, 3,535, or 3,610 ppm (0, 146, 467, 630, 976, or 997 mg/m3).
The approximate lethal concentration by inhalation (4 hours) in rats was determined to be 2,280 ppm
(Table 4-35). In the 4-week range-finding inhalation study, Wistar rats were exposed to chloroprene at
0, 50, 200, or 800 ppm (actual mean concentrations were 0, 39, 161, or 625 ppm [0, 11, 44, or
173 mg/m3], respectively). A similar study was conducted (after completion of the 4-week rat study)
with Syrian golden hamsters exposed to 0, 40, 160, or 625 ppm (actual mean concentrations were 0,
39, 162, or 630 ppm [0, 11, 45, or 174 mg/m3], respectively). The purity of chloroprene used in this
study was 99.9% with 0.01% phenothiazine added as a polymerization inhibitor. Test atmospheres
were generated by low temperature (0°C) vaporization in nitrogen.
Table 4-35. Chloroprene-induced mortality in male rats
Concentration (ppm)
530
1,690
2,280
3,535
3,610
Mortality (Dead/Total)
0/6
0/6
1/6
2/6
2/6
Source: Used with permission from Elsevier, Inc., Clary et al. (1978, 064942).
Clary et al. (1978, 064942) reported no deaths from dermal, oral, or inhalation administration in
the standard Class B poison test (sacrifice 2 days after the 4-hour exposure period). There was mild to
moderate skin irritation and erythema after the dermal exposure. Irregular respiration, mild
lacrimation, and slight initial weight loss were reported after the inhalation exposure. For the modified
Class B poison test (sacrifice 14 days after the 4-hour exposure period), 2/6 and 3/6 animals died on
the 6th and 7th days, respectively.
In the 4-week range-finding study, exposure to 625 ppm chloroprene was associated with eye
irritation, restlessness, lethargy, nasal discharge, and orange-colored urine in rats and hamsters. Hair
loss was observed in female rats exposed to the two highest exposure groups (161 and 625 ppm).
Increased mortality in rats was observed at the two highest concentrations starting in week 1 (5/10
males and 3/10 females died at 625 ppm; 3/10 males died at 161 ppm at the end of the exposure period,
4 weeks). Mortality was 100% for male and female hamsters in the highest dose group (630 ppm) by
the end of week 1, and 1/10 males and 3/10 females at the mid-exposure (162 ppm) by the end of week
4. One male hamster died in the low-exposure (39 ppm) group by week 4. Decreases in body weight
were observed at all concentrations in rats and at 162 ppm in hamsters. There were changes in the
80
-------
relative weights of all organs except for the heart. The relative organ weights for kidneys were
increased at the 162 ppm exposure level for both male and female hamsters, the 625 ppm level for
male rats, and the 161 and 625 ppm level for female rats. Liver weights were increased in the high-
exposure group in both species except for female hamsters. Male rats exhibited decreased liver
weights at 39 and 161 ppm. Relative lung weights were increased at 625 ppm for male and female
rats. Clary et al. (1978, 064942) noted that these increases in the relative weight of the kidneys, liver,
and lungs may have indicated a direct effect of chloroprene exposure, whereas weight changes in other
organs (spleen, brain, thyroid, and adrenal glands) may have been secondary to decreases in body
weight.
In rats, gross pathological examination of the animals that died during exposure revealed dark,
swollen livers and grayish lungs with hemorrhagic areas. Dark swollen livers were also observed in
several animals exposed to the highest concentration when they were sacrificed at the end of the study.
Microscopic examination revealed slight to severe centrilobular liver degeneration in all male rats and
in 8/10 of the females at the high concentration. This change was also observed in 2/3 male rats
exposed to 161 ppm that died during the study. The kidneys of male and female rats exposed to
625 ppm had enlarged tubular epithelial cells. In addition, one male and one female rat exposed to
625 ppm showed foci of necrotic tubules in the intramedullary area of the kidneys.
In hamsters, the lungs of most of the animals that died within the first 24 hours of exposure (all
animals died after a single exposure to 630 ppm and 1/10 males and 1/10 females at 162 ppm) showed
gray-reddish edematous areas. Fecal and urinary incontinence were observed in 1/10 male and 3/10
females at 630 ppm. The heart of 1/2 females that died on the second day of exposure was pale with
severe myocarditis, and the thoracic cavity contained a considerable amount of fluid. The other female
had a small spleen and a pale liver with a pronounced lobular pattern. Significant body weight
decreases were observed only in the 162-ppm group. Histopathology examinations revealed necrosis
and midzonal degeneration of hepatocytes in most of the survivors of the 162-ppm group. Several
males and females (number not specified) exposed to either 39 or 162 ppm showed irritation of the
mucous membranes of the nasal cavity. This irritation was described as a slight flattening and thinning
of the layer of the olfactory epithelium in the dorsomedial part of the cavity.
4.4.2. Immunotoxicity
There are some laboratory animal data suggesting potential immunomodulatory effects of
chloroprene; however the data are from standard toxicological studies and no targeted
immunotoxicological studies of chloroprene were identified. The studies discussed below were
described in detail previously in this assessment and only the relevant immune data are presented here.
NTP (1998, 042076) observed that thymus weights in adult male and female B6C3Fi mice exposed to
80 ppm chloroprene for 16 days were significantly decreased compared to controls (p < 0.01) and
thymic necrosis, characterized by karyorrhexis of thymic lymphocytes, was observed in both sexes at
200 ppm. No changes in thymus weight or histopathology were reported in mice after chloroprene
81
-------
exposure for a longer period (i.e., 13-week exposure) as part of the same NTP (1998, 042076) study.
Alterations in differential white blood cell counts (i.e., increased leukocyte, neutrophil, and monocyte
numbers) were observed at 500 ppm in male rats after 16 days of exposure and segment neutrophils
were decreased in male rats at 200 ppm after 13 weeks of exposure. In the 2-year chronic portion of
the NTP study, splenic hematopoietic cell proliferation was significantly increased over controls in
male mice at 32 and 80 ppm, and in all exposed females (level of significance not reported).
Hyperplasia of the mediastinal lymph node was observed in females exposed to 32 or 80 ppm
(significance not stated).
Trochimowicz et al. (1998, 625008) observed that mean relative spleen and thymus weights
were significantly (p < 0.01) lower in female Wistar rats exposed to 50 ppm chloroprene for 2 years,
but did not report any accompanying histopathological changes in either organ. Clary et al. (1978,
064942) also observed small spleens in hamsters (qualitative description) and decreased spleen weights
(possibly secondary to decreased body weights) in rats exposed to 625-630 ppm chloroprene for
4 weeks. Sanotskii (1976, 063885) reported that chromosomal aberrations were observed in the bone
marrow of mice exposed to chloroprene and in leukocyte cultures prepared from the blood of exposed
chloroprene production workers.
These findings provide some evidence of immunomodulatory effects of chloroprene in
laboratory animals. The immune-related data for chloroprene include altered lymphoid organ weights
and histopathology, and chromosomal aberrations in bone marrow. However, it has been shown that
changes in lymphoid organ weights and genotoxicity observed in lymphoid organs are both poor
predictors of compound-related changes in immune function (Luster et al., 1992, 084126). The
changes in thymic histopathology reported after 16 days of exposure were not observed with longer
exposure, suggesting no chronic effects. The remaining data on increased hematopoietic cell
proliferation and lymph node hyperplasia are nonspecific effects that are difficult to interpret as
potential immunotoxicity of chloroprene. They may be related to general hematopoietic effects of
chloroprene rather than an effect on the immune system or immune function. In general, measures
such as these (i.e., morphological disturbances) are not clear measures of a chemical's potential to
cause changes in immune function (Putman et al., 2003, 624893). Direct measures of immune
function, such as antibody production to a T-cell dependent antigen, are usually preferred to delineate a
chemical's immunotoxic potential (Luster et al., 1992, 084126; Putman et al., 2003, 624893).
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF MODE OF ACTION
4.5.1. Mode-of-Action Studies
Many of the available studies addressing the mode of action (MO A) of chloroprene have
focused on investigating the metabolic profile for chloroprene including identifying epoxide
metabolites, their reactivity with DNA, and adduct formation in vitro (Hurst and Ali, 2007, 625159;
Munter, et al., 2002, 625215). Other studies have used molecular analysis to study alterations in ras
proto-oncogenes from lung and Harderian gland tumors identified in the NTP (1998, 042076) chronic
82
-------
bioassay that may indicate events in chloroprene-induced neoplasia (Sills et al., 1999, 624952; Ton et
al., 2007, 625004).
The metabolism of chloroprene into reactive epoxides has been primarily evaluated in vitro
with liver and lung tissue fractions from rat, mouse, hamster, and humans. Only a limited number of
studies have investigated the in vivo metabolism of chloroprene. In studies using mouse and human
liver microsomes, Bartsch et al. (1979, 010689) showed that 2-chloro-2-ethynyloxirane and/or
(l-chloroethenyl)oxirane could be intermediates in the biotransformation of chloroprene. Metabolism
of chloroprene into (l-chloroethenyl)oxirane was confirmed by Himmelstein et al. (2001, 019012):
oxidation of chloroprene to (l-chloroethenyl)oxirane was evident in rodent and human liver
microsomes and most likely involved CYP2E1, as evidenced by the near complete in vitro inhibition
with 4-methylpryazole. A comparison across species suggested that a greater amount of
(l-chloroethenyl)oxirane was present in B6C3Fi mice and F344 rat liver microsomes, followed by the
Wistar rat, then humans and hamsters. A maximum concentration of (l-chloroethenyl)oxirane of 0.01-
0.02 jiM was detected in mouse liver microsomes between 5-10 minutes after initiation of exposure
with 0.05 jiM (100 ppm) chloroprene. Preliminary data also showed that hydrolysis of
(l-chloroethenyl)oxirane was slowest in the liver microsomes of B6C3Fi mice. Further comparing
metabolism between species, Cottrell et al. (2001, 157445) observed that qualitative profiles of
metabolites from liver microsomes obtained from B6C3Fi mice, Sprague-Dawley or F344 rats, and
humans were similar, with (l-chloroethenyl)oxirane being the major metabolite in all species and
sexes. Himmelstein et al. (2004, 625152) developed a two-compartment closed vial model to describe
both chloroprene and (l-chloroethenyl)oxirane metabolism in liver and lung fractions from rat
(two strains, F344 and Wistar), mouse, hamster, and humans. Oxidation (Vmax/Km) of chloroprene in
the liver was slightly faster in the mouse and hamster than in rats or humans. However, in lung
microsomes, Vmax/Km was much greater for mice compared with the other species. Conversely,
hydrolysis (Vmax/Km) of (l-chloroethenyl)oxirane in liver and lung microsomes was faster for the
human and hamster, than for rat or mouse. The observation that mice generally metabolized
chloroprene into its epoxide metabolite at equal or faster rates than other species and hydrolyzed the
epoxide more slowly may, in part, explain why mice were observed to be the most sensitive species in
regards to the observed carcinogenicity of chloroprene.
The in vivo rodent studies support the postulated metabolic pathway for chloroprene. For
example, male Wistar rats administered 100 or 200 mg/kg chloroprene by gavage demonstrated a rapid
depletion of hepatic GSH and a dose-dependent increase in excreted urinary thioethers (presumably
GSH-conjugates), which is consistent with in vitro studies using isolated liver hepatocytes (Summer
and Greim, 1980, 064961). Pre-treatment of rats or hepatocytes with phenobarbital or a
polychlorinated biphenyl (PCB) mixture (Clophen A50) to induce the mixed-function oxidase enzymes
enhanced the GSH depletion effect.
Munter et al. (2007, 576501; 2002, 625215) investigated the reactivity of the chloroprene
metabolite (l-chloroethenyl)oxirane towards DNA nucleosides and calf thymus DNA in vitro. Adducts
83
-------
were isolated by reverse-phase chromatography and characterized by their mass spectrometric features.
The reaction of (l-chloroethenyl)oxirane with the nucleoside 2'-deoxyguanosine yielded one major
adduct derived by nucleophilic attack of N-7 guanine on C-3' of the epoxide. In addition, another
chloroprene metabolite 2-chlorobut-2-en-l-al described as an unsaturated aldehyde, yielded 2 major
adducts. Reaction of (l-chloroethenyl)oxirane with 2'-deoxy-adenosine, -cytosine, and -thymine
individually also resulted in adduct formation. When equimolar quantities of all 4 nucleosides were
reacted with (l-chloroethenyl)oxirane simultaneously in a competitive reaction assay, all of the adducts
identified from individual nucleoside reactions were observed and were formed at similar rates. The
reaction of (l-chloroethenyl)oxirane with double stranded calf thymus DNA yielded
N7-(3-chloro-2-hydroxy-3-buten-l-yl)-guanine (dGI) as the major adduct (96% on a molar basis), the
same adduct seen when the chloroprene metabolite was incubated with 2'-deoxyguanosine
individually. N3-(3-chloro-2-hydroxy-b-buten-l-yl)-2'-deoxyuridine (dCI) was also detected. The
reaction of (l-chloroethenyl)oxirane with deoxycytidine in DNA may be significant because such
adducts are difficult to repair and may therefore be implicated in mutagenesis (Koskinen et al., 2000,
010173).
The in vitro reactivity of (l-chloroethenyl)oxirane with hemoglobin (adduct formation) and
enantiomer detoxification (i.e., disappearance of R- versus S-enantiomer from the test system) in vitro
have been investigated by Hurst and Ali (2007, 625159). Mouse (C57BL/6) erythrocytes (RBCs) were
incubated with the R- and S-enantiomers of (l-chloroethenyl)oxirane in vitro. The authors reported a
greater persistence of the R- over the S-enantiomer upon incubation with RBCs in the in vitro system
tested. The authors also reported a greater amount of globin adducts formed with the R- than with the
S-enantiomer.
As part of the 2-year bioassay of chloroprene, NTP (1998, 042076) evaluated possible
oncogene-activating mechanisms for lung and Harderian gland neoplasms in the B6C3Fi mouse at 0,
12.8, 32, and 80 ppm. The results were published by Sills et al. (1999, 624952). After isolation and
amplification of DNA from the neoplasms, H-ras and K-ras mutations were identified. A higher
frequency (80%) of K-ras mutations was detected in chloroprene-induced lung neoplasms than in
spontaneous neoplasms of control mice (30%). The predominant mutation (59% of all mutations;
present in 47% of tumors) was an A->T transversion (CAA->CTA) at K-ras codon 61: 80% (8/10) of
low dose, 71% (10/14) of mid dose, and 18% (4/22) of high dose lung tumors were observed to have
this mutation. This specific mutation was not observed in spontaneously occurring lung neoplasms. A
similar pattern of ras mutations was observed also with isoprene-induced lung neoplasms but not in
those induced by butadiene. Rare point mutations (G->T, A, or C transversions), not seen in
spontaneous lung neoplasms, were detected at codon 12. No consistent morphological pattern
(papillary, solid, or mixed) or type (benign or malignant) of neoplasm was co-observed with specific
K-ras mutations. Although definitive evidence is currently unavailable, there are a number of factors
that may explain the observation of the lower frequency of codon 61 CTA transversions in lung tumors
of high dose animals. In the lung, the lower frequencies in CTA transversions at high doses may be
84
-------
due to non-ras mutation mechanisms of genotoxicity or carcinogen!city. Alternatively, differences in
DNA-adduct formation or induction of repair or removal mechanisms may explain the pattern
observed.
A high incidence (100%) of both K-ras and H-ras mutations was detected in chloroprene-
induced Harderian gland neoplasms, compared with 56% in spontaneous Harderian gland tumors in
control mice, 100% in neoplasms from isoprene-exposed mice, or 69% in neoplasms from butadiene-
exposed mice. The predominant mutation was also a CAA->CTA transversion at K-ras codon 61
(93%), which only occurred in 7% (2/27) spontaneously occurring Harderian gland neoplasms. The
concentration-response was similar across exposure groups. It was suggested that the large number of
ras mutations at A:T base pairs after exposure to chloroprene, isoprene, or butadiene indicated an
interaction with DNA to form adenine adducts that may be important for tumor induction. Sills et al.
(2001, 624922) reported higher frequencies of K- and H-ras mutations (57%) in chloroprene-induced
forestomach tumors in B6C3Fi mice compared to spontaneous tumors (36%). The A->T transversion
(CAA->CTA) in H-ras codon 61 was identified in 29% of the chemically induced forestomach
neoplasms, but was not observed in spontaneous control tumors. Mutations at K-ras codon 61 were
not observed in chloroprene-induced forestomach tumors.
Ton et al. (2007, 625004) evaluated mutations in the K-ras oncogenes and loss of
heterozygosity in the region of K-ras on distal chromosome 6 in lung tumor samples collected from
mice exposed to chloroprene in the NTP 2-year inhalation study. DNA analysis included isolation from
formalin fixed tissue sections, and amplification, cycle sequencing of ras gene and analysis for loss of
heterozygosity (LOH). Chloroprene-induced mouse lung tumors had a high frequency of LOH on
chromosome 6 in the region of K-ras. The correlation between K-ras mutation and loss of the
wildtype allele was high in the tumors examined: of the 19 lung tumors with LOH from B6C3Fi mice
exposed to chloroprene, 16 (84%) of them also had K-ras mutations.
4.5.2. Genotoxicity Studies
This section presents the findings of several genotoxicity studies that are summarized in
Table 4-36.
85
-------
Table 4-36. Genotoxicity assays of chloroprene
Test System
Cells/Strain
Tested
Concentrations
Results3
Reference
Bacterial assays
Salmonella
typhimurium
TA100
TA100, TA1535
TA98
TA100, TA1535
TA98, TA1537, TA1538
TA100, TA1535,
TA1537, TA98
TA100
TA100
TA100, TA1535,
TA97A, TA98
0.5-8% (v/v) in air
10,000^0,000 ppm
10,000^0,000 ppm
up to 3,333 ug/plate
0-5 umol/plate
0-5 umol/plateb
0-69 mMc
+
+
-
+
-
_
-
+
+
Bartschetal. (1979,
010689)
Willems (1980, 625049)
Willems (1980, 625049)
Willems (1978, 625048)
Willems (1980, 625049)
NTP (1998, 042076)
Westphaletal. (1994,
625047)
Westphaletal. (1994,
625047)
Himmelstein et al. (2001,
019013)
Mammalian cell assays
Micronucleus
Micronucleus
Chinese hamster V79
Chinese hamster V79
10% (v/v)
0.175mMc
-
-
Drevon and Kuroki (1979,
010680)
Himmelstein et al. (2001,
019013)
In vivo bioassays
Sex-linked
recessive lethal
mutation
Sex-linked
recessive lethal
mutation
Sister chromatid
exchange: bone
marrow
Chromosomal
aberration: bone
marrow
Chromosomal
aberration: bone
marrow
Micronucleus:
peripheral blood
Micronucleus: bone
marrow
Drosophila (Canton-S)
Drosophila (Berlin-K)
B6C3FJ mice
B6C3FJ mice
C57BL/6 mice
B6C3FJ mice
B6C3FJ mice
12.8, 32, 80 ppm
12.8, 32, 80 ppm
up to 1 ppm
12.8, 32, 80 ppm
-
+
-
-
+
-
-
Foureman et al. (1994,
065173)
Vogel (1979, 000948)
NTP (1998, 042076):
Shelby (1990, 624906): Tice
(1988, 624981: 1988,
064962)
NTP (1998, 042076)
Sanotskii (1976, 063885)
NTP (1998, 042076)
Shelby and Witt (1995,
624921)
Tor bacterial assays, tests were performed in the absence or presence of the exogenous S9 metabolism system. In
all cases of positive mutagenicity (except Westphal et al. (1994, 625047)). addition of S9 mixture enhanced the
observed mutagenicity.
bAged chloroprene distillates tested (in the absence of the exogenous S9 metabolism system).
°Epoxide metabolite (l-chloroethenyl)oxirane tested.
86
-------
4.5.2.1. Bacterial Mutagenicity Assays
Both positive and nonpositive mutagenic responses have been observed in bacterial mutagenic
assays.
Bartsch et al. (1979, 010689) exposed Salmonella typhimurium strain TA100 to 0.5-8%
(volume/volume [v/v]) of chloroprene within sealed desiccators for 4 hours at 37°C in the absence or
presence of the exogenous S9 metabolism system. Batch solutions were freshly prepared before use
and kept at -20°C. Chloroprene purity was 99% and contained a negligible amount of dimers. A
positive mutagenic response that was concentration-dependent was observed without S9 fraction; this
response increased threefold when S9 fractions from either phenobarbital-pretreated or untreated mice
were used.
Willems (1978, 625048; 1980, 625049) found that chloroprene (purity not stated, but sample
was "freshly supplied") was mutagenic with S. typhimurium strains TA100 and TA1535 in the presence
or absence of S9 (mutagenicity was more pronounced in the presence of the S9 fraction), indicating
base pair substitution mutations. Chloroprene, however, was not mutagenic in S. typhimurium strains
TA98, TA1537, and TA1538 indicating a lack of frameshift mutations. Petri plates were incubated at
37°C in desiccators for either 48 or 24 hours, removed, and then incubated for another 24 hours.
Positive controls were used. Four dimers (chemical characterization not stated) were also tested under
the same conditions. Three of the four were mutagenic against both salmonella base pair substitution
strains (TA100 and TA1535).
Westphal et al. (1994, 625047) investigated the mutagenicity of chloroprene with respect to the
compound stability and reactivity with solvents used in the test system. The Ames test was performed
using the S. typhimurium (strain TA100) with or without S9, in gas-tight chambers to prevent
chloroprene volatilization. Chloroprene was freshly distilled from a 50% xylene solution. The
distillates were stored at -20°C and checked for purity immediately before testing. The authors noted
that 2-5% xylenes remained in the chloroprene distillates. Another set of distillates were prepared in
the same manner and stored either under air or under argon and kept at room temperature (referred to
as aging) for 1, 2, or 3 days. Chromatographic analysis of the aged chloroprene revealed the presence
of decomposition products reported to be cyclic dimers. The influence of solvents was also tested in
this study by using either ethanol or dimethyl sulfoxide (DMSO) as vehicles. Propylene oxide (a
volatile direct mutagen) and benzo(a)pyrene were used as positive controls.
Freshly distilled chloroprene dissolved in either DMSO or ethanol as vehicles, with or without
S9, was not mutagenic in TA100. Aged chloroprene had a mutagenic effect on TA100 that increased
linearly with increasing age of the chloroprene distillates. Westphal et al. (1994, 625047) confirmed
these findings by obtaining positive results with 10 additional distillates containing different
proportions (quantitative details not specified) of the decomposition products, without S9. The
mutagenicity of the distillates correlated with the proportion of the decomposition products (which
increased over time in the aged samples). The mutagenicity of aged chloroprene towards TA100 was
the same whether chloroprene was stored under air or under an inert gas. The authors speculated that
87
-------
the mutagenic products in aged chloroprene were less volatile than those in the fresh distillates, thus
remaining in the test medium long enough to cause toxicity.
Addition of GSH, both with and without S9, reduced the mutagenicity of aged chloroprene but
was less effective as the amount of decomposition products increased. Westphal et al. (1994, 625047)
stated that chloroprene diluted in DMSO was markedly more toxic and more mutagenic than
chloroprene dissolved in ethanol, although no data were provided to support this statement.
Chloroprene did not show any evidence of mutagenicity in any of four strains of
S. typhimurium (TA98, TA100, TA1535, or TA1537) tested at concentrations up to 3,333 jig/plate, in
the presence or absence of Aroclor-induced rat or hamster liver S9 fraction (NTP, 1998, 042076).
Himmelstein et al. (2001, 019013) investigated the mutagenicity of chloroprene monoepoxide,
(l-chloroethenyl)oxirane (>98% purity) in Salmonella strains TA100 TA1535, TA97A, and TA98.
Exposures were performed with or without S9 activation in airtight capped glass vials in order to
prevent the loss of the test substance due to volatilization. Test concentrations were 0-69 mM in
DMSO. Cells were preincubated with the test compound for approximately 45 minutes at 37°C and
then plated and allowed to incubate for an additional 48 hours, (l-chloroethenyl)oxirane was
genotoxic in all Salmonella strains tested without Aroclor-induced S9 activation (Himmelstein et al.,
2001, 019013): inclusion of S9 did not enhance the mutagenic effect in any of the tester strains.
Toxicity was noted at >14 mM in plates without S9 and at >34 mM in plates with S9.
4.5.2.2. Mammalian Cell Assays
Chloroprene (99% pure) was evaluated for mutagenic potential in V79 Chinese hamster cells in
the presence of a liver supernatant (SI5 fraction) from phenobarbital-pretreated rats and mice (Drevon
and Kuroki, 1979, 010680). Cells were incubated at 37°C for 5 hours or longer in 2.5 mL of reaction
mixture with or without S15 fraction from mice pretreated with phenobarbital, plus cofactors, either in
liquid suspension or in 0.3 % agar. The petri dishes were placed in a desiccator and exposed to 0, 0.2,
1, 2, and 10% (v/v) chloroprene vapors for 5 hours. Toxicity was evaluated as a measure of plating
efficiency. Mutations were evaluated in terms of resistance to a purine analogue (8-azaguanine) and
ouabain (inhibitor of adenosine triphosphatase in cell membranes). Chloroprene toxicity was observed
at concentrations above 1%; this effect was enhanced with addition of the S15 fraction. The authors
noted that this suggested the formation of a toxic metabolite. No mutations were observed in the
absence or presence of SI5.
Himmelstein et al. (2001, 019013) evaluated the clastogenic potential of the
(l-chloroethenyl)oxirane (>98% purity) using the cytochalasin-B blocked micronucleus test in Chinese
hamster V79 cells without metabolic activation. The V79 cells plated on tissue culture slides were
placed inside sterile bottles filled with culture medium followed by injection of 0-0.943 mM
(l-chloroethenyl)oxirane dissolved in DMSO into the bottles and incubation for 3 hours. Cells were
then transferred to fresh medium containing cytochalasin-B and incubated for an additional 16 hours.
A minimum of 500 binucleated cells were scored for micronuclei. Clastogenicity was determined as
88
-------
the presence of a dose-dependent increase in the frequency of micrenucleated cells, with at least one
concentration producing a threefold increase. Cytotoxicity, reported as a reduction in the number of
binucleated cells, and altered cell morphology were observed starting at 0.175 mM. Although no
clastogenic response (as determined by the above criteria) was noted at concentrations up to
0.175 mM, at least three concentrations induced an increase in the frequency of micrenucleated cells
over control levels and a dose-dependent (although, not monotonic) increase was apparent over the
tested range of concentrations.
4.5.2.3. In Vivo Bioassays
Vogel (1979, 000948) evaluated the in vivo genotoxic potential of chloroprene (99% pure with
negligible dimer content) to induce recessive lethal mutations on the X-chromosome of male
Drosophila melanogaster (wild-type strain Berlin-K). Storage conditions and the elapsed time
between receipt and use were not reported. Chloroprene was dissolved in DMSO and diluted with a
5% sucrose solution to obtain a final concentration of 1% DMSO and the desired experimental
concentration. Adult males (2-3 days old) were treated at 25°C for 1-3 days in sealed beakers placed
in a desiccator to account for the volatility of chloroprene. After mating, the F3 generation was
evaluated for recessive lethality. The increase in the percentage of observed recessive-lethal mutations
was marginal in several experiments and was not concentration dependent. However, when the data
from pooled samples from several experiments (53 lethals in 15,941 X-chromosomes) were compared
with seven control experiments, the difference was statistically significant at p < 0.01. The authors
noted that the possible variation among samples could be related to the instability of chloroprene. Two
different samples of chloroprene were used, one that was highly purified and one that contained several
impurities (chemical characterization not stated). There were no apparent differences in mutagenic
potential between the two samples of chloroprene, suggesting the impurities were not responsible for
the observed genotoxicity.
In a study by Foureman et al. (1994, 065173), chloroprene (purity 50%) dissolved in ethanol
was nonpositive (p > 0.01) for sex-linked recessive lethal mutations in postmeiotic and meiotic germ
cells of adult male D. melanogaster (strain Canton-S) when exposed by either the injection or feeding
route. The investigators suggested that the discrepancy between their nonpositive findings and those of
Vogel (1979, 000948) may be due to (1) differences in purity of the chloroprene sample, (2) differences
between the Berlin-K and Canton-S strains, (3) differences in sample sizes, and (4) possible genetic
drift within the female populations used by the two groups of investigators. Another possibility for the
conflicting results could be that chloroprene in ethanol is less genotoxic than if dissolved in DMSO
(Gahlmann, 1993, 625174: Westphal et al., 1994, 625047).
Cytogenetic tests using chloroprene were nonpositive. In studies performed by Brookhaven
National Laboratories for the NTP (1998, 042076), sister chromatid exchanges and chromosomal
aberrations (bone marrow cells) and the frequency of micronuclei in peripheral blood erythrocytes
were evaluated in male mice exposed by inhalation to chloroprene in the NTP (1998, 042076)
89
-------
bioassay. Results were published separately by Shelby (1990, 624906). Tice (1988, 624981). and Tice
et al. (1988, 064962). Mice were exposed by inhalation to chloroprene at 0, 12.8, 32, 80, or 200 ppm
(0, 3.5, 8.8, 22, or 55 mg/m3) 6 hours/day for 12 days. Mortality was 100% at 200 ppm. There were
no exposure-related effects compared with controls in numbers of sister chromatid exchanges,
chromosomal aberrations, or micronucleus frequency in polychromatic or normochromatic
erythrocytes. Tice (1988, 624981) and Tice et al. (1988, 064962) did report that the mitotic index
(frequency of cells in metaphase) in mouse bone marrow cells was elevated in chloroprene-exposed
animals, with the increase being significant in the 80-ppm group. Tice (1988, 624981), and Tice et al.
(1988, 064962) suggested that the lack of chloroprene-induced genotoxicity in bone marrow may
imply that any carcinogenic activity attributable to chloroprene would likely be localized to tissues
directly exposed to chloroprene (e.g., lung) or to tissues with a high metabolic activity that form
reactive intermediates. Results of the NTP (1998, 042076) demonstrate that carcinogenic activity can
occur at sites distal to the portal-of-entry, so lack of an effect in bone marrow may be due to low
metabolic activity in this tissue.
The frequency of micrenucleated cells in peripheral blood erythrocytes was not affected when
mice were exposed to chloroprene for 13 weeks to 0, 12.8, 32, or 80 ppm (0, 3.5, 8.8, or 22 mg/m3)
(MacGregor et al., 1990, 625184: NTP, 1998, 042076).
Sanotskii (1976, 063885) reported on a study identifying an increase in chromosomal
aberrations in bone marrow cells of mice exposed for 2 months to chloroprene concentrations of
3.5 mg/m3 (1 ppm) and below. The protocol details and information about the purity and storage of
chloroprene were not provided.
Shelby and Witt (1995, 624921) found nonpositive results in vivo in the mouse bone marrow
micronucleus test and in chromosomal aberration tests when male B6C3Fi mice were injected
intraperitoneally with chloroprene in corn oil, three times, at 24-hour intervals. Dose levels, protocol
details, and information about the purity and storage of chloroprene were not provided.
Chloroprene was also tested in a dominant lethal assay with male Swiss mice (Immels and
Willems, 1978, 625176). Groups of 12 males were exposed to 0, 10, or 100 ppm (0, 2.8, or 28 mg/m3)
chloroprene 6 hours/day, 5 days/week for 2 weeks. Immediately after exposure, each male was mated
with two virgin females for seven days. Females were replaced each week for 8 weeks. There was no
sign of dominant lethal mutations or effects on mating performance or fertility.
4.5.3. Structural Alerts
Chloroprene is the 2-chloro analog of 1,3-butadiene, a multiorgan, cross-species carcinogen,
and is structurally similar to isoprene (2-methyl-1,3-butadiene). Inhalation studies have demonstrated
that, similar to butadiene and isoprene, chloroprene is a multisite carcinogen in rats and mice.
Butadiene and isoprene are both metabolized to epoxides and diepoxides that are known mutagens and
are believed to be responsible for their carcinogenicity. Chloroprene is also metabolized to an epoxide
intermediate that may mediate its carcinogenic effects; however, there is no evidence of diepoxide
90
-------
formation in the metabolism of chloroprene. The similarities in the sites of tumor induction in rodents
(Table 4-37) between butadiene, isoprene, and chloroprene provide further evidence for a similar MOA
for these epoxide-forming compounds. A comparison of the carcinogenic potency of butadiene and
chloroprene in mice highlights the general quantitative concordance of their tumorigenic effects
(Melnick and Sills, 2001, 051506). All of the tumorigenic effects (except for chloroprene induced
mammary tumors) exhibited supralinear or linear dose-response curves when fit with a Weibull model.
Chloroprene appeared more potent in the induction of forestomach and lung tumors in male mice and
liver tumors in female mice, whereas butadiene was more potent in inducing Harderian gland tumors in
both male and female mice. However, the female mouse lung was the most sensitive site of
carcinogenicity for both chloroprene and butadiene, and both chemicals seemed equally potent in that
particular neoplasm's induction (EDi0 = 0.3 ppm).
Table 4-37. Sites of increased incidences of neoplasms in the 2 year inhalation
studies of 1,3-butadiene, isoprene, and chloroprene in rats and mice
Site
Lymphatic/
hematopoietic
Circulatory
Lung
Liver
Forestomach
Harderian
gland
Mammary
gland
Brain
Thyroid
Pancreas
Testis
Zymbal's
gland
Kidney
Oral Cavity
Mice
Butadiene
M,Fa
M,F
M. F
M,F
M,F
M, F
F
M
Isoprene
M
M
M
M
M
M, F
Chloroprene
M,F
M,F
F
M,F
M, F
F
F
M
Rats
Butadiene
F
M
F
M
M
F
Isoprene
M,F
M
M
Chloroprene
M
F
M, F
M, F
M, F
aM = males, F = females.
Source: NTP (1998, 042076): Melnick et al. (1994, 6252081: Placke et al. (1996, 6248911: U.S. EPA (2002,
052153).
91
-------
Table 4-38 Quantitative comparison of carcinogenic potency of butadiene and
chloroprene in mice
Site
Lung
Harderian gland
Forestomach
Liver
Mammary gland
Males
Butadiene
2.8a
4.4
120
-
-
Chloroprene
0.9
12
70
-
-
Females
Butadiene
0.3
12
62
10
13
Chloroprene
0.3
23
79
1.9
12
aED10 values (concentration associated with 10% excess cancer risk) in ppm.
Source: Used with permission from Elsevier, Melnick and Sills (2001, 051506).
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS
4.6.1. Human Studies
There is a limited body of information on the nonneoplastic toxicological consequences to
humans who are exposed to chloroprene. In a summary by Nystrom (1948, 003695), chloroprene was
reported to cause respiratory, eye, and skin irritation, chest pains, temporary hair loss, dizziness,
insomnia headache, and fatigue in occupationally exposed workers. Chest pains accompanied by
tachycardia and dyspnea were also reported. In a Russian review (Sanotskii, 1976, 063885) of the
effects of chloroprene, medical examinations of chloroprene production workers revealed changes in
the nervous system (lengthening of sensorimotor response to visual cues and increased olfactory
thresholds), cardiovascular system (muffled heart sounds, reduced arterial pressure, and tachycardia),
and hematology (reduction in RBC counts, decreased hemoglobin levels, erythrocytopenia, leucopenia,
and thrombocytopenia). The ambient concentration of chloroprene in work areas ranged from
1-7 mg/m3 (3.6-25 ppm).
4.6.2. Animal Studies
4.6.2.1. Oral Exposure
The toxic potential of chloroprene by the oral route has been assessed in only one study
(Ponomarkov and Tomatis, 1980, 075453). This was a reproductive study involving exposure of
BD-IV rats to a single dose (100 mg/kg) of chloroprene on the 17th day of pregnancy and of their
progeny to weekly doses (50 mg/kg) for 120 weeks. Animals treated with chloroprene that died within
the first 30 weeks of treatment showed severe congestion of the lungs and kidneys.
92
-------
4.6.2.2. Inhalation Exposure
The database for inhalation toxicity studies in animals on chloroprene includes two range-
finding studies for 16 days and 13 weeks (also reported by Melnick et al., 1999, 000297: NTP, 1998,
042076). two chronic inhalation bioassays (also reported by Melnick et al., 1999, 000297: NTP, 1998,
042076): Trochimowicz et al. (1998, 625008) and four reproductive developmental studies (Appelman
and Dreef-van der Meulen, 1979, 064938: Culik et al., 1978, 094969: Mast et al., 1994, 625206:
Sanotskii, 1976, 063885). These studies associate chloroprene inhalation exposure with toxicity
effects in multiple organ systems, including respiratory tract, kidney, liver, spleen, and forestomach.
Increased mortality was observed in male and female rats exposed to 500 ppm chloroprene for
16 days (NTP, 1998, 042076). In male rats, the mortality reached 90% (9/10), whereas mortality was
lower in females exposed to the same concentration (3/10). In mice exposed to chloroprene for
16 days, all of the males and females in the high-exposure group (200 ppm) died. In the 2-year chronic
bioassay (NTP, 1998, 042076), mortality was increased over controls in male mice exposed to 32 or
80 ppm chloroprene and in females at all exposure concentrations tested. Decreased body weights
were observed in male and female rats exposed for 16 days (> 200 ppm), male mice exposed 16 days
(32 and 80 ppm), and in female mice exposed for 2 years (80 ppm) (NTP, 1998, 042076).
Hematological and clinical chemistry effects were also reported by the NTP (1998, 042076)
study. In rats exposed to chloroprene for 16 days, increases in serum enzyme (ALT, GDH, and SDH)
activities, as well as anemia and thrombocytopenia (decreased platelet count), were observed in the
200- and 500-ppm groups on day 4 of exposure only. In rats exposed to chloroprene for 13 weeks,
minimal increases in hematocrit values, hemoglobin concentrations, and erythrocyte counts were
observed in males exposed to > 32 ppm and in females exposed to 200 ppm on day 2. At week 13,
male and female rats in the 200-ppm group demonstrated decreased hematocrit values, decreased
hemoglobin concentrations, and decreased erythrocyte counts characterized as normocytic,
normochromic anemia. Transient thrombocytopenia, evidenced by a reduction in circulating platelet
numbers, occurred in male and female rats in the 200-ppm group on day 2 and in females at 80 and
200 ppm on day 22. At study termination (13 weeks) increases in platelet numbers were observed at
80 and 200 ppm in exposed males and females. Transient increases in activities of serum enzymes
(ALT, GDH, and SDH) were observed on day 22 in both sexes at 200 ppm. Alkaline phosphatase
enzymeuria was observed in males at > 32 ppm and in females at 200 ppm. In male rats, proteinuria
was observed at 200 ppm. In mice exposed to chloroprene for 13 weeks (NTP, 1998, 042076),
hematological changes were similar to those observed in rats; however, they were less severe.
Minimal anemia, including decreased hematocrit values, erythrocyte counts, and platelet numbers were
observed in female mice exposed to 32 or 80 ppm chloroprene.
Respiratory effects included a number of nasal and pulmonary effects in both rats and mice
exposed to chloroprene (NTP, 1998, 042076: Trochimowicz et al., 1998, 625008). In rats exposed to
chloroprene for 16 days (NTP, 1998, 042076), minimal to mild olfactory epithelial degeneration was
observed in all exposed male and females. Additionally, metaplasia of the olfactory epithelium,
93
-------
characterized as replacement with a simple columnar respiratory-like epithelium, was observed in
males at > 80 ppm and females at > 32 ppm. In rats exposed to chloroprene for 13 weeks (NTP, 1998,
042076), increased incidences of minimal to moderate olfactory epithelial degeneration and olfactory
metaplasia (characterized as replacement with a simple columnar respiratory-like epithelium) occurred
in male and female rats at 80 or 200 ppm. Olfactory epithelial degeneration was observed in female
rats exposed to 32 ppm. In rats exposed to chloroprene for 2 years (NTP, 1998, 042076), the
incidences of atrophy, basal cell hyperplasia, metaplasia, and necrosis of the olfactory epithelium in
males and females were increased at 32 and 80 ppm; atrophy and necrosis were additionally increased
at 12.8 ppm. Necrosis of the olfactory epithelium was characterized by areas of karyorrhexis and
sloughing of olfactory epithelium with cell debris in the lumen of the dorsal meatus. Atrophy of the
olfactory epithelium was characterized by decreased numbers of layers of olfactory epithelium and
included loss of Bowman's glands and olfactory axons in more severe cases. Metaplasia was
characterized by replacement of olfactory epithelium with ciliated, columnar, respiratory-like
epithelium. Basal cell hyperplasia was characterized by proliferation or increased thickness of the
basal cell layer in the turbinate and septum. Increased incidences were observed for chronic
inflammation in males (> 12.8 ppm) and in females (80 ppm), fibrosis and adenomatous hyperplasia of
the olfactory epithelium in males and females (80 ppm), and alveolar/bronchiolar hyperplasia in males
and females in every exposure group. No histopathological changes were observed in the respiratory
tract of mice exposed to chloroprene for either 16 days or 13 weeks. In mice exposed to chloroprene
for 2 years (NTP, 1998, 042076), increases in the incidences of olfactory epithelial atrophy,
adenomatous hyperplasia, and metaplasia were observed in males and females at 80 ppm. Atrophy and
metaplasia of the olfactory epithelium was similar to lesions observed in rats exposed to chloroprene.
Suppurative inflammation was observed in female mice exposed to 32 or 80 ppm. Bronchiolar
hyperplasia was increased in males and females in all exposure groups, whereas pulmonary histiocytic
cellular infiltration was increased in every dose group in females only. Bronchiolar hyperplasia was
characterized by diffuse thickening of the cuboidal cells lining the terminal bronchioles and in some
cases caused papillary projections into the lumen. Histiocytic cellular infiltration consisted of
histiocytes within alveolar lumens, usually adjacent to alveolar/bronchiolar neoplasms. In a second
chronic 2-year bioassay (Trochimowicz et al., 1998, 625008), male and female rats exposed to 50 ppm
chloroprene displayed mild respiratory effects such as lymphoid aggregates around bronchi,
bronchioles, and blood vessels.
Toxicity was also observed in the kidneys and livers of rats and mice exposed to chloroprene.
In rats exposed to chloroprene for 16 days (NTP, 1998, 042076), significant increases in kidney weight
(right kidney only) were seen at 80 and 500 ppm. Mild to moderate centrilobular hepatocellular
necrosis and increased liver weight was also observed in male and female rats exposed to 200 or
500 ppm chloroprene. In rats exposed to chloroprene for 13 weeks (NTP, 1998, 042076), increases in
kidney weight was observed in males at 200 ppm and females at > 80 ppm and the incidence of
hepatocellular necrosis was increased in female rats exposed to 200 ppm. Variably sized aggregates of
94
-------
yellow or brown material, consistent with hemosiderin accumulation, appeared in small vessels or
lymphatics in or near portal triads or in Kupffer cells of male and female rats exposed to 200 ppm.
Increased incidence of kidney (renal tubule) hyperplasia was observed in rats exposed to chloroprene
for 2 years when combined single- and step-sections were analyzed; incidence was increased in males
at > 32 ppm and in females at 80 ppm. Renal tubule hyperplasia was distinguished from regenerative
epithelial changes commonly seen as a part of nephropathy and was considered a preneoplastic lesion.
Hyperplasia was generally a focal, minimal to mild lesion consisting of lesions that were dilated
approximately two times the normal diameter and were lined by increased numbers of tubule epithelial
cells that partially or totally filled the tubule lumen. In rats exposed to chloroprene for 2 years
(Trochimowicz et al., 1998, 625008), the number of rats with one or more small foci of cellular
alteration in the liver was higher in the 50 ppm exposure group than in controls. In males, there was an
increased incidence of hepatocellular lesions described as one or several small clear cell foci in the
50-ppm group. Increased incidences of multifocal random hepatocellular necrosis were observed in
male and female mice exposed to 200 ppm chloroprene for 16 days. In mice exposed to chloroprene
for 2 years (NTP, 1998, 042076), the incidence of kidney (renal tubule) hyperplasia was increased in
males exposed to 32 or 80 ppm when only single-sections were analyzed, and in all groups of exposed
males when single- and step-sections were combined. The morphology of renal tubule hyperplasia in
male mice was similar to that observed in rats.
The reproductive and developmental effects of chloroprene exposure are equivocal. In male
rats exposed to chloroprene for 13 weeks, sperm motility was decreased at 200 ppm, whereas sperm
morphology and vaginal cytology parameters were similar to those in the control groups in exposed
male and female mice. In a study by Culik et al. (1978, 094969), rats were exposed on either
gestational day 1-12 (embryotoxicity study) or 3-20 (teratology study). In the teratology study, an
increase in the percentage of litters with resorptions was observed at 10 and 25 ppm, with only the
change in the 10-ppm group achieving statistical significance relative to controls. An increase in the
percentage of litters with resorptions was not observed in the larger embryotoxicity portion of the study
which was specifically designed to detect such an effect. The equally high numbers of litters with
resorptions (~ 50%) in all experimental groups, including controls, in the embryotoxicity study
correspond well to the level of response observed at 10 and 25 ppm in the teratology study (62% and
59%, respectively). When the potential increase in resorptions is expressed in numbers of resorbed
fetuses per litter, the control group for the teratology study is the only exposure group which falls
outside of the historical control range for this strain of rat (MARTA; and MTA, 1996, 625111). This
suggests that the control group response in the teratology study may be a statistical outlier and that the
finding of a statistically significant increase in litters with resorptions at 10 ppm is spurious.
Chloroprene exposure did result in statistically significant increases in average fetal body weight and
length. No major compound-induced or dose-related skeletal or soft tissue anomalies were observed.
No exposure-related effects on maternal health, number of implantations, live pups, resorptions, fetal
body weight or length, organ weights, or malformations were observed in NZW rabbits exposed to
95
-------
chloroprene (Mast et al., 1994, 625206). In a two-generation reproduction study (Appelman and
Dreef-van der Meulen, 1979, 064938), effects on body weight were observed in the FO and Fl animals.
Exposed Fl males also had smaller testes and females had larger ovaries, livers, and lungs compared to
controls. No histopathological changes were observed in those organs. The general lack of effects in
the above reproductive and developmental studies is not consistent with the many positive effects seen
in previous Russian studies reviewed by Sanotskii (1976, 063885). However, the Sanotskii review
lacks important study details, including the purity of the test substance and experimental design, and is
therefore difficult to interpret with any confidence.
Chloroprene toxicity was observed in a number of additional organ systems. In mice exposed
to chloroprene for 16 days, thymic necrosis, characterized as karyorrhexis of thymic lymphocytes, and
hypertrophy of the myocardium was observed at 200 ppm. In rats exposed for 13 weeks,
neurobehavioral parameters were affected: horizontal activity was increased in male rats exposed to
> 32 ppm compared with chamber control animals and total activity was increased in male rats at 32
and 200 ppm. No exposure-related effects on motor activity, fore/hindlimb grip strength, or startle
response were observed. In mice exposed to chloroprene for 13 weeks, increased incidences of
squamous epithelial hyperplasia of the forestomach were observed in male and female mice exposed to
80 ppm. Preening behavior may have lead to direct gastrointestinal exposure to chloroprene. In mice
exposed for 2 years, the incidence of hyperplasia of the forestomach epithelium was increased in males
and females at 80 ppm. The hyperplastic lesions were similar to those seen in the 13-week study and
consisted of focal to multifocal changes characterized by an increase in the number of cell layers in the
epithelium. The incidence of thyroid follicular cell hyperplasia was increased in male rats exposed to
32 ppm chloroprene for 2 years. Increased splenic hematopoietic cell proliferation was observed in
male mice ( > 12.8 ppm) and female mice ( > 32 ppm) exposed to chloroprene for 2 years.
4.7. EVALUATION OF CARCINOGENICITY
Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005, 086237), there is
evidence that chloroprene is "likely to be carcinogenic to humans" based on (1) statistically significant
and dose-related information from an NTP (1998, 042076) chronic inhalation bioassay demonstrating
the early appearance of tumors, development of malignant tumors, and the occurrence of multiple
tumors within and across animal species; (2) evidence of an association between liver cancer risk and
occupational exposure to chloroprene; (3) suggestive evidence of an association between lung cancer
risk and occupational exposure; (4) the proposed mutagenic mode of action; and (5) structural
similarities between chloroprene and known human carcinogens, butadiene and vinyl chloride
(Table 4-38).
U.S. EPA's Guidelines for Carcinogen Risk Assessment (2005, 086237) indicate that for tumors
occurring at a site other than the initial point of contact, the weight of evidence for carcinogenic
potential may apply to all routes of exposure that have not been adequately tested at sufficient doses.
An exception occurs when there is convincing toxicokinetic data that absorption does not occur by
96
-------
other routes. Information available on the carcinogenic effects of chloroprene via the inhalation route
demonstrates that tumors occur in tissues remote from the site of absorption. Information on the
carcinogenic effects of chloroprene via the oral and dermal routes in humans or animals is limited or
absent (HSDB, 2009, 594343: NIOSH, 1977, 644450: NIOSH, 1995, 644453). Quantitative data
regarding the absorption via any route of exposure are unavailable. However, based on the observance
of systemic tumors following inhalation exposure, and in the absence of information to indicate
otherwise, it is assumed that an internal dose will be achieved regardless of the route of exposure.
Therefore, chloroprene is considered "likely to be carcinogenic to humans" by all routes of exposure.
4.7.1. Synthesis of Human, Animal, and Other Supporting Evidence
4.7.1.1. Human
A number of occupational cohort studies have examined cancer mortality and incidence among
workers exposed to chloroprene monomer and/or polychloroprene latex in the U.S., Russia (Moscow),
Armenia, France, China, and Ireland (Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419:
Colonna and Laydevant, 2001, 625112: Leet and Selevan, 1982, 094970: Li et al., 1989, 625181:
Marsh et al., 2007, 625187: Marsh et al., 2007, 625188: Pell, 1978, 064957: Romazini et al., 1992,
624896). Concern that exposure to chloroprene may result in liver cancer derives principally from its
structural similarity to vinyl chloride, a chemical known to cause liver angiosarcoma in humans.
Exposed workers have included those involved in chloroprene monomer production using both the
acetylene process in which exposure to vinyl chloride was possible and the more recent butadiene
process which does not involve vinyl chloride exposure. Other workers were involved with
handling/sampling of partially finished products such as polychloroprene latex which contains various
amounts of dissolved monomer. Some studies span eras in which little or no worker safety protection
measures were likely used in contrast with years in which process improvements and concern for
worker safety were gradually instituted. Therefore, it is difficult to compare results across studies
given a wide range of exposure variability within and between these cohorts.
Despite these differences in occupational exposure to chloroprene and other chemicals, four of
the cohorts with observed liver/biliary passage cancer cases showed statistically significant
associations (i.e., two- to fivefold increased risk) with chloroprene exposure. Four mortality studies
(Table 4-11) reported SMRs of 339, 240, 482, 571 when compared to external populations (Bulbulyan
et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Leet and Selevan, 1982, 094970: Li et al., 1989,
625181). Although sample size and statistical power were limited (thus limiting the precision of risk
estimates), Bulbulyan et al. (1998, 625105: 1999, 157419) observed significantly elevated relative risk
estimates for liver cancer incidence and mortality among intermediate and highly exposed workers.
The study involving four plants, including the Louisville Works plant included in the Leet and Selevan
(1982, 094970) study by Marsh et al. (2007, 625188). which had the largest sample size and most
extensive exposure assessment, also observed increased relative risk estimates for liver cancer in
relation to cumulative exposure in the plant with the highest exposure levels (trend p-value = 0.09, RRs
97
-------
1.0, 1.90, 5.10, and 3.33 across quartiles of exposure, based on 17 total cases). Although not
statistically significant, these findings are consistent in magnitude with results (RR range: 2.9-7.1)
detected in two other studies for high and intermediate cumulative exposures (Bulbulyan et al., 1998,
625105; Bulbulyan et al., 1999, 157419). Though several studies noted higher SMRs for lung cancer
among workers exposed to chloroprene, the evidence was not considered as strong as liver cancer.
This was mostly due to the inability to adequately control for confounding by smoking status, a strong
risk factor for lung cancer. There was also no evidence of exposure-response relationship across
various chloroprene exposure categories.
One of the strengths of several of the more recent epidemiologic studies was improved
exposure assessment data. These studies utilized industrial hygiene information to determine which
areas or jobs were most likely to have received higher chloroprene exposures. This allowed for
examination of various exposure contrasts and helped reduce the potential for exposure
misclassification. These data allowed for internal analyses to be conducted which should be less
impacted by bias due to the healthy worker effect; however, the potential for healthy survivor effect
remains as noted previously. Despite these improvements, several study limitations added to the
uncertainty in addressing the weight of evidence of the epidemiologic data.
A key limitation of most of the chloroprene studies (and other occupational studies) is the
potential for bias due to the healthy worker effect. Although this may be less of a concern for cancer
mortality outcomes, SMR analyses are based on external comparisons to the general population and
will often result in reduced SMR values for the occupational cohort. Two studies with more advanced
chloroprene exposure assessment conducted internal analyses to reduce this source of bias (Bulbulyan
et al., 1999, 157419: Marsh et al., 2007, 625188). Among these studies, only Bulbulyan et al. (1999,
157419) observed a statistically significant association between chloroprene exposure and liver cancer
mortality. As with most epidemiological research, the potential for bias due to residual confounding is
another limitation that exists in these studies. With respect to liver cancer, the lack of data on alcohol
consumption precluded its examination as a potential confounder, although there is no direct evidence
that alcohol is related to the exposure of interest (i.e., chloroprene). Given the nature of the work
environment for most of the study participants in these occupational studies, there is also the
possibility of co-exposures which may be confounders, although Bulbulyan et al. (1999, 157419)
discussed the known co-exposures at the study facility in Armenia and reported that none were known
liver carcinogens. One study with data on a co-exposure (vinyl chloride) reported evidence of negative
confounding (Marsh et al., 2007, 625188). This would result in an underestimate of the reported
association between chloroprene and liver cancer if adjusted for vinyl chloride which suggests that this
co-exposure was unlikely to explain the association observed between chloroprene and liver cancer in
that population.
An additional limitation in several studies was incomplete enumeration of both incident cases
and deaths. In some studies, former workers exposed to high levels of chloroprene could not be
identified or located for inclusion in the studies. This raises the possibility that the actual number of
98
-------
liver cancer cases might have been higher than indicated from the data on the subset of individuals that
were included in the studies. Another concern in these occupational studies is the reliance on death
certificates for outcome diagnosis in the mortality analyses. Although misclassification of cause of
death can be minimized by the review of medical records or by histological confirmation, this was not
done in any of the studies. The lack of histological review of the liver cancer cases is an important
limitation of the available studies using internal controls. Lastly, another concern in some of the
occupational cohorts is the low expected counts used for liver and lung cancer mortality (Bulbulyan et
al., 1998, 625105: Bulbulyan et al., 1999, 157419: Li et al., 1989, 625181). This could be an
indication of inaccurately applied population rates or incorrect calculation of expected values based on
the selected population mortality rates. Use of very low expected counts of cancer mortality may
result in unstable estimates of effect. Regardless, the results of the studies reporting very low expected
counts of cancer mortality and increased SMRs should not be discounted from the weight of evidence
of the carcinogenicity of chloroprene; these studies do indicate a statistically significant association
across heterogeneous populations and exposure scenarios.
It is also important to note that some of the epidemiology studies investigated the same cohort.
For example, the Marsh et al. (2007, 625187: 2007, 625188) study investigated an employee cohort
from the Louisville Works DuPont plant that was previously investigated in Leet and Selevan (1982,
094970). However, there are a number of differences between the studies that warranted independent
analysis of each. Specifically, Leet and Selevan (1982, 094970) reported that the Louisville cohort
consisted of 1,575 male employees (salaried and female employees excluded due to "minimal or no
potential exposure to chloroprene") who were working at the Louisville plant on June 30, 1957. The
authors further reported that most of the employees had 15 years of potential exposure to chloroprene
(indicating that most had worked at the plant since its opening in 1942). Also, the cohort was followed
until 1974. Marsh et al. (2007, 625187: 2007, 625188) included all workers (male and female) in each
plant with potential exposure to chloroprene from the start of production until 2000. For the Louisville
plant, this included a total of 5,507 workers employed from 1949-1972. The Marsh et al. (2007,
625187: 2007, 625188) analyses started at 1949 to avoid methodological problems associated with the
earlier fifth revision of the ICD and stopped at 1972 for the Louisville plant as that was when they
report chloroprene production stopped at that plant, although chloroprene purification and
polymerization still occurred there according to Leet and Selevan (1982, 094970). Also, there are
important differences in how each study assessed exposure. Leet and Selevan (1982, 094970) used
worker history summaries to classify workers as either "high" or "low" chloroprene exposure, whereas
Marsh et al. (2007, 625187: 2007, 625188) used a more sophisticated approach that considered worker
history summaries and worker exposure profiles to generate quantitative estimates of chloroprene
exposure intensity. Similar differences between Colonna and Laydevant (2001, 625112) and Marsh
et al. (2007, 625187: 2007, 625188) relative to the Isere/Grenoble cohort also warrant independent
analysis of these studies. Therefore, although these studies investigated members of the same cohort, a
number of methodological differences between the studies warrant the independent analysis of each.
99
-------
These epidemiologic study results, when examined in the context of different plant operating
and worker exposure conditions over different time periods and a low number of incident liver cancers,
offer evidence of an association for exposure to chloroprene with an increase of liver cancer in
humans. Despite various limitations (e.g., healthy worker bias, potential co-exposure, and incomplete
enumeration of cases), internal and external comparisons showed consistent evidence of an association
between chloroprene exposures and liver cancer. The associations detected in some studies add
support to the cancer weight of evidence determination.
4.7.1.1.1. Evidence for Causality. The evidence for causality for cancer from the human studies is
summarized in the paragraphs that follow and is based on recommendations from the EPA (2005,
086237) Guidelines for Carcinogen Risk Assessment. These guidelines advocate the use of "criteria"
proposed by Hill (1965, 071664) to assess causality. It should be noted that there exists a number of
methodological limitations of the epidemiologic studies that may preclude drawing firm conclusions
regarding the following criteria. These limitations include lack of control of personal confounders and
risk factors associated with the outcomes in question, imprecise exposure ascertainment resulting in
crude exposure categories, incorrect enumeration of cases leading to misclassification errors, limited
sample sizes, and the healthy worker effect.
Temporality. Exposure must precede the effect for causal inference. Furthermore, and
particularly with cancers, exposure must precede the effect with a sufficient latency to be considered
causal. In all the occupational studies reviewed the chloroprene exposure has preceded effect (either
incidence of or mortality due to liver cancer) with sufficient latency to be considered causally
associated. Several of the studies have specifically evaluated latencies of 15-20 years (Bulbulyan et
al., 1998, 625105: Colonna and Laydevant, 2001, 625112: Marsh et al., 2007, 625187: Marsh et al.,
2007, 625188: Pell, 1978, 064957).
Strength of Association. Refers to the magnitude of measures of association such as the ratio
of incidence or mortality (e.g., SMRs, SIRs, RRs or odds ratios) irrespective of statistical significance.
Studies reporting large, precise risks are less likely to be doing so due to chance, bias, or confounding.
Reports of modest risk, however, do not preclude a causal association and may reflect lower levels of
exposure or an agent of lower potency. When compared to external populations, there was a
statistically significant two- to fivefold increased risk of liver cancer in four cohort studies in China (Li
et al., 1989, 625181), Louisville, KY, (U.S.) (Leet and Selevan, 1982, 094970). Russia (Bulbulyan et
al., 1998, 625105) and Armenia (Bulbulyan et al., 1999, 157419) despite evidence of healthy worker
effect bias. Despite relatively small numbers, there were also suggestive data from the re-analysis of
the Louisville cohort by Marsh et al. (2007, 625188). which found RRs ranging from 1.9-5.1 (not
statistically significant) for cumulative exposures to chloroprene and liver cancer mortality. These data
were consistent in magnitude to two other studies (Bulbulyan et al., 1998, 625105: Bulbulyan et al.,
1999, 157419) examining intermediate and high cumulative exposures to chloroprene and liver cancer
100
-------
incidence (RRs = 2.9-4.9, statistically significant) and mortality (RRs = 4.4-7.1, not statistically
significant), respectively.
Consistency. The observation of the same site-specific effect across several independent study
populations strengthens an inference of causality. Four different studies, examining four independent
cohorts, have shown an association between chloroprene exposure and liver cancer incidence and
mortality (Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Leet and Selevan, 1982,
094970: Li et al., 1989, 625181), while a fifth study showed evidence suggesting an association when
examined in relation to detailed exposure data (Marsh et al., 2007, 625188). It is important to note that
the Marsh et al. (2007, 625188: 2007, 625187) study investigated an employee cohort from the
Louisville Works DuPont plant that was previously investigated in Leet and Selevan (1982, 094970).
However, there are a number of differences between the studies (e.g., different exposure assessment
methodologies) that warrants independent analysis of each. Larger effect estimates for liver cancer
risk have been observed in diverse populations working in chloroprene monomer and polymer
production, neoprene manufacturing, and manufacturing utilizing polychloroprene products in the
U.S., China, Armenia, and Russia. The studies with internal comparisons showed consistently elevated
liver cancer relative risk estimates for intermediate (RR range: 2.9-7.1) and high cumulative risk
exposures (Range: 3.3-4.9) as noted above.
Specificity. As originally intended, this criterion refers to increased inference of causation if a
single site effect as opposed to multiple effects is observed and associated with exposure. Chloroprene
exposure has been found to be associated specifically with increased risk of liver cancer in four cohorts
(Bulbulyan et al. 1998, 1999; Li et al., 1989; Leet and Selevan, 1982). However, based on current
understanding, this is now considered one of the weaker guidelines for causality (for example, many
agents cause respiratory disease and respiratory disease has multiple causes), with some suggesting
that specificity does not confer greater validity to any causal inference regarding the exposure effect
(Rothman and Greenland, (Rothman and Greenland, 1998, 086599).
Biological Gradient. Refers to the presence of a dose-response and/or exposure/duration-
response between a health outcome and exposure of interest. The aforementioned internal analyses for
chloroprene and liver cancer mortality (Bulbulyan et al., 1999, 157419; Leet and Selevan, 1982,
094970) suggest a potential biological gradient by comparing workers exposed to high concentrations
to workers unexposed or exposed to low concentrations. In Bulbulyan et al. (1999, 157419), the SIR
for intermediate cumulative exposure to chloroprene is 293 (95% CI: 41-2080), whereas the SIR for
the high cumulative exposure group is 486 (95% CI: 202-1170). Although these effect estimates are
not statistically significant from one another, the presence of monotonically increasing effects relative
to cumulative exposure is apparent. In Leet and Selevan (Leet and Selevan, 1982, 094970), there is a
dose-response apparent in the author-reported effect of cancer of the liver and biliary passage.
However, if liver cancer was considered separately, this dose-response would disappear as only one of
the three reported cases was liver cancer. The other studies examining exposure-response relationships
do not demonstrate a monotonic increase in risk but have reported consistent elevated risks above 3.3
101
-------
in the upper exposure categories (Bulbulyan et al., 1998, 625105: Marsh et al., 2007, 625188). Some
suggestion of an exposure-response effect has also been observed in comparisons between long-term
employees and short-term employees in the Bulbulyan studies.
Biological Plausibility. Refers to the observed effect having some biological link to the
exposure. Chloroprene has been found to be metabolized by humans and other species to epoxides,
which are known genotoxic metabolites, and has been shown to be a potent (early appearance,
multiplicity, malignancy of observed tumors) carcinogen in mice and rats. In addition, the structurally
related carcinogen, butadiene, is also metabolized to epoxides and produces a tumor profile resembling
that observed with chloroprene.
In summary, the temporality of exposure prior to occurrence of liver cancer, strength of
association, consistency, biological gradient, and biological plausibility provide some evidence for the
carcinogenicity of chloroprene in humans.
4.7.1.2. Laboratory Animal
According to the NTP (1998, 042076), there is clear evidence of carcinogenicity in the F344/N
rat and B6C3Fi mouse due to lifetime inhalation exposure to chloroprene. The mouse is regarded as
the most sensitive species because tumor incidence and multisite distribution were greater than with
the rat. There was decreased survival in chloroprene-exposed rats and mice, and survival in mice was
significantly associated with the burden of neoplastic lesions. Mortality in rats was likely due to overt
toxicity across many organ systems. In rats, statistically significantly increased incidences of
neoplastic lesions occurred in the oral cavity (papillomas or carcinomas, males and females), kidney
(renal tubule adenomas or carcinomas, males), thyroid gland (adenomas or carcinomas, males) and
mammary gland (fibroadenomas, females). In mice, increased incidences in neoplasms occurred in the
lungs (adenomas or carcinomas, males and females), circulatory system (hemangiomas or
hemangiosarcomas, all organs, males and females), Harderian gland (adenomas or carcinomas, males
and females), liver (adenomas or carcinomas, females), skin and mesentery (sarcomas, females),
mammary gland (carcinomas, females), and kidney (renal tubule adenomas or carcinomas, males).
The observation that chloroprene is more potent in inducing tumors in B6C3Fi mice compared to
F344/N rats may be due to species differences in metabolism. The activity of liver or lung microsomal
oxidation of chloroprene and the formation of (l-chloroethenyl)oxirane was generally higher in the
mouse than the rat (Himmelstein et al., 2004, 625152) (Tables 3-4 and 3-6). Additionally, the activity
of epoxide hydrolase in liver microsomes was greater in the rat compared to the mouse (epoxide
hydrolase activity was approximately equal in lung microsomes). The observation that formation of
the reactive epoxide metabolite of chloroprene is greatest in the mouse lung may explain the
observation that chloroprene exposure induces lung tumors in mice, but not rats.
In contrast to the neoplastic findings in the F334/N rat, only small numbers of neoplastic
lesions were observed in Wistar rats or Syrian golden hamsters (Trochimowicz et al., 1998, 625008).
There is no unequivocal explanation for why the results for the rat differ between these two studies.
102
-------
The stability of the bulk material in the NTP (1998, 042076) study was monitored by gas
chromatography, and the material was analyzed for peroxide content. In addition, stabilizer
concentrations were in an acceptable range and no dimer peaks were found in the distribution lines
leading to the exposure chamber. Concentrations of volatile degradation products (e.g.,
1-chlorobutadiene) never exceeded 0.6% of the atmospheric concentration of chloroprene when
sampled from either the distribution line or exposure chamber. In the study in the Wistar rat by
Trochimowicz et al. (1998, 625008), there was no evidence of degradation of the freshly distilled
chloroprene, and dimer concentrations were stated to be less than the limit of detection. Thus, it is
unlikely that the bulk materials or generated atmospheres differed to an extent that would have caused
the differences in results. The discrepancy between the carcinogenicity of chloroprene observed in the
two studies may be due to species and/or strain differences. Himmelstein et al. (2001, 019012)
observed that liver microsomes from B6C3Fi mice and the F344 rats, the two species used in the NTP
(1998, 042076) study, produced more (l-chloroethenyl)oxirane than those from hamsters or Wistar
rats, the two species used in the Trochimowicz et al. (1998, 625008) study. These differences in
production of (l-chloroethenyl)oxirane were as great as 12-fold greater (F344 rats versus hamsters).
However, measurements of Vmax/Kmfor liver microsomal oxidation of chloroprene were approximately
equal for the mouse and hamster, with both being greater than either strain of rat (Himmelstein et al.,
2004, 625152). In lung microsomes, the activity was much greater in the mouse compared to all other
species. The activity of epoxide hydrolase in liver microsomes (Table 3-5) was highest in the hamster,
followed by both rat strains with the mouse having the lowest activity. Epoxide hydrolase activity in
lung microsomes was highest in hamsters, with rats and mice being approximately equal. The
combination of highest rate of oxidation of chloroprene with the slowest rate of epoxide detoxification
in mouse microsomes provides some insight on the observation that the mouse is the most sensitive
species/strain across both studies.
The inhalation study by Dong et al. (1989, 007520) found that a 7-month exposure of the
Kunming strain of albino mice, a strain reported to have a low spontaneous rate of lung tumor
formation, resulted in a chloroprene-associated increase in lung tumors. Although quality assurance
procedures regarding histopathology were not reported, these study results are considered to support
the findings in the B6C3Fi mice in the NTP (1998, 042076) chronic bioassay.
In the only long-term oral cancer study (an Fl generation of inbred BD-IV rats given weekly
doses of 50 mg/kg chloroprene by gavage), no significant neoplastic effects were reported
(Ponomarkov and Tomatis, 1980, 075453). The number of tumor-bearing animals was similar to
controls.
4.7.2. Summary of Overall Weight of Evidence
In the current document, a total of nine studies covering eight cohorts of human subjects
exposed to chloroprene were reviewed to assess the occurrence of cancer. The most consistent
findings across the database were excess cancers of the liver (Bulbulyan et al., 1998, 625105;
103
-------
Bulbulyan et al., 1999, 157419: Leet and Selevan, 1982, 094970: Li et al., 1989, 625181) and lung
(Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Colonna and Laydevant, 2001,
625112: Leet and Selevan, 1982, 094970: Marsh et al., 2007, 625188: Pell, 1978, 064957). The
epidemiologic evidence for increased lung cancer mortality due to chloroprene exposures is limited.
The few studies that reported increased risk were not statistically significant. In addition to a lack of a
consistent association and the small increased risks that were detected, other study limitations, such as
lack of smoking data, limit the ability to determine possible causal associations between lung cancer
and humans exposed occupationally to chloroprene.
There was a statistically significant excess of liver cancers in four of the cohorts reviewed
(Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Leet and Selevan, 1982, 094970: Li et
al., 1989, 625181), with a two- to more than fivefold increased risk in the SMR seen among these
studies. Although no statistically significant increase in risk of liver cancer was detected in the most
recent and comprehensive cohort study involving workers at four plants (Marsh et al., 2007, 625188),
the observed RR increased with increasing cumulative exposure in the plant with the highest exposure
levels, indicating a dose-response trend. Limitations in the existing epidemiological database included
the lack of information on individual worker's habits (i.e., alcohol consumption) needed to control for
potential confounding, incomplete enumeration of incidence and mortality cases, and potential for
biases that may lead to an underestimation of the risk (e.g., the healthy worker effect). See Section
4.7.1.1 for further discussion of these limitations.
According to NTP (1998, 042076), there is clear evidence of carcinogenicity in the F344/N rat
and B6C3Fi mouse due to lifetime inhalation exposure to chloroprene. In rats, increased incidences of
neoplastic lesions primarily occurred in the oral cavity and lung (males only), kidney, and mammary
gland (females). In mice, increased incidences in neoplasms occurred in the lungs, circulatory system
(all organs), Harderian gland, forestomach, liver, skin and mesentery (females only), and kidney (males
only). Additionally, metabolites of chloroprene include DNA-reactive epoxides and a mutagenic mode
of action is proposed based on suggestive results in in vitro bacterial assays and the observation of in
vivo K- and H-ras mutations in animals exposed to chloroprene (Section 4.7.3.2).
104
-------
Table 4-39. Summary of animal and human tumor data and weight of evidence
descriptor for chloroprene
Statistically significant
tumor types
In male F344/N rats, increased incidence of kidney (renal tubule)
adenoma or carcinoma in all dose groups, and oral papilloma or
carcinoma and thyroid adenoma or carcinoma at the two highest
dose groups.
In female F344/N rats, increased incidence of mammary
fibroadenoma at the two highest dose groups and oral papilloma or
carcinoma at the highest dose.
In male B6C3F! mice, increased incidence of lung adenoma or
carcinoma and hemangioma/hemangiosarcoma in all organs in all
dose groups, and Harderian Gland adenoma or carcinoma and
kidney (renal tubule) adenoma or carcinoma at the two highest dose
groups.
In female and male B6C3FJ mice, increased incidence of lung
adenoma or carcinoma and skin sarcoma in all dose groups, liver
adenoma or carcinoma at the two highest dose groups, and
Harderian Gland adenoma or carcinoma and mammary gland
fibroadenomas at the highest dose.
Hemangiomas/hemangiosarcomas in all organs and mesentery
sarcomas were observed in the middle dose.
In humans, significant increases in liver cancer mortality were
observed in four occupational epidemiology studies (out of nine
total studies). Relative risk estimates for liver cancer (while not
statistically significant) increased with increasing exposure,
indicating a dose-response trend.
Rare Tumors
Statistically significant increase in rare kidney (renal tubule)
adenoma in male rats and mice.
Statistically significant increases in primary (assumed) liver cancer
in four cohort studies and lung cancer mortality in two studies in
workers occupationally exposed to chloroprene.
Multiple Studies
• Animals - NTP (1998, 042076)
• Humans - Leet and Selevan (1982, 094970): Li et al. (1989,
625181): Bulbulyan et al. (1998, 625105): and Bulbulyan et al.
(1999, 157419)
Conclusions
Tumors in both sexes of rats and mice.
Decreased time to tumor in both sexes of rats and mice.
Tumors in occupationally exposed workers.
Methodological limitations of the occupational epidemiology studies
(e.g., lack of data on confounders, small sample sizes, and lack of
precise quantitative exposure ascertainment) make it difficult to
draw firm conclusions regarding the human cancer data.
Rare tumors (kidney renal tubule adenomas in animals, primary liver
cancer in humans).
Metabolites include DNA-reactive epoxides and a mutagenic mode
of action is proposed.
Weight of Evidence
characterization
Likely to be carcinogenic to humans.
105
-------
4.7.3. Mode-of-Action Information
4.7.3.1. Hypothesized Mode of Action
The proposed hypothesis is that chloroprene acts via a mutagenic mode of action involving
reactive epoxide metabolites formed at target sites or distributed systemically throughout the body.
DNA-epoxide adduct formation is an effect observed for a number of carcinogens structurally related
to chloroprene, including those with a known mutagenic mode of action (i.e., vinyl chloride; EPA
(2000, 194536; 2005, 088823)) and those for which a preponderance of evidence strongly suggests a
mutagenic mode of action (i.e., isoprene and 1,3-butadiene) (Begemann et al., 2004, 625093; Sills et
al., 1999, 624952: U.S. EPA, 2002, 052153). This hypothesized mode of action is presumed to apply
to all tumor types. Mutagenicity is a well-established cause of carcinogenicity.
4.7.3.2. Experimental Support for the Hypothesized Mode of Action
Compelling evidence for the hypothesized mutagenic mode of action for chloroprene includes:
(1) chloroprene, like butadiene and isoprene, is metabolized to epoxide intermediates and both
compounds are carcinogens; (2) chloroprene forms DNA adducts via its epoxide metabolite;
(3) observation of the genetic alterations (base-pair transversions) in proto-oncogenes in chloroprene-
induced lung, Harderian gland, and forestomach neoplasms in mice and positive results in Salmonella
typhimurium strains that test for base-pair substitution mutations; and (4) similarities in tumor sites and
sensitive species between chloroprene and butadiene in chronic rodent bioassays—NTP (1998,
042076) and Melnick et al. (1999, 000297), respectively. These lines of evidence are elaborated on
below.
Evidence for the formation of reactive epoxide metabolites following exposure to chloroprene
has been observed in both sexes of multiple species. Currently, in vivo data are unavailable for blood
or tissue-specific epoxide metabolism rates or concentrations. However, in studies using mouse and
human liver microsomes, Bartsch et al. (1979, 010689) showed that 2-chloro-2-ethynyloxirane and/or
(l-chloroethenyl)oxirane could be intermediates in the biotransformation of chloroprene. Himmelstein
et al. (2001, 019012) confirmed the identity of the volatile metabolite reported by Bartsch et al. (1979,
010689) as the epoxide (l-chloroethenyl)oxirane. Himmelstein et al. (2001, 019012) reported that the
oxidation of chloroprene to (l-chloroethenyl)oxirane was evident in rodent and human liver
microsomes and most likely involved CYP2E1. The oxidation of chloroprene to
(l-chloroethynyl)oxirane is more prevalent in B6C3Fi mice and F344 rat liver microsomes than in
Wistar rats, humans, or hamsters. Comparing metabolism between species, Cottrell et al. (2001,
157445) confirmed the results of Himmelstein et al. (2001, 019012), and further showed that the
quantitative profiles of metabolites from liver microsomes obtained from mice, rats, and humans were
similar. In all species and either sex, (l-chloroethynyl)oxirane was the major metabolite detected. One
distinct difference between species was the stereospecificity of epoxide metabolites formed. In two
strains of rats (Sprague-Dawley and F344), the R-enantiomer was preferentially formed, whereas this
enantioselectivity was not observed in mice or humans. Hurst and Ali (2007, 625159) reported that the
106
-------
S-(l-chloroethynyl)oxirane enantiomer was more quickly detoxified in mouse erythrocytes than the
R-enantiomer, suggesting that the R-enantiomer may be more toxic due to its slower elimination.
1,3-butadiene exhibits similar biotransformation to reactive epoxide metabolites. Oxidation of
1,3-butadiene to l,2-epoxy-3-butene has been observed in hepatic, lung, and kidney microsomes, as
well as lung tissue and bone marrow, in rats, mice, and humans (U.S. EPA, 2002, 052153). Further
oxidation of l,2-epoxy-3-butene to 1,2,3,4-diepoxybutane has been observed rat, mouse, and human
liver microsomes, as well as in blood and tissues of mice and rats exposed by inhalation to
1,3-butadiene (U.S. EPA, 2002, 052153). Vinyl chloride and isoprene are also readily converted into
their reactive epoxide metabolites; vinyl chloride is converted to chloroethylene epoxide in rats and
isoprene to (2,2')-2-methylbioxirane in rats and mice (U.S. EPA, 2000, 194536: Watson et al., 2001,
625045).
Metabolites of chloroprene have been shown to form DNA adducts when reacted with
nucleosides and double stranded DNA in vitro. Reaction of (l-chloroethenyl)oxirane with the
nucleoside 2'-deoxyguanosine yielded one major adduct derived by nucleophilic attack of N-7 guanine
on C-3' of the epoxide, whereas another metabolite, 2-chlorobut-2-en-l-al, yielded 2 major adducts
(Munter et al., 2007, 576501; Munter, et al., 2002, 625215). The reaction of (l-chloroethenyl)oxirane
with double stranded calf thymus DNA yield the same adduct observed when the chloroprene
metabolite was incubated with 2'-deoxyguanosine individually, (l-chloroethenyl)oxirane also reacted
with deoxycytidine in double stranded DNA to yield an adduct which may be significant as such
adducts are difficult to repair and may therefore be implicated in mutagenesis (Koskinen et al., 2000,
010173).
Evidence for the mutagenic potential of chloroprene has been shown in molecular analysis of
the genetic alteration of cancer genes including the ras proto-oncogenes (Sills et al., 1999, 624952;
Sills et al., 2001, 624922; Ton et al., 2007, 625004), which are alterations commonly observed in
human cancers. Tissues from lung, forestomach, and Harderian gland tumors from mice exposed to
chloroprene in the NTP chronic bioassay (1998, 042076) were shown to have a higher frequency of
mutations in K- and H-ras proto-oncogenes than in spontaneous occurring tumors (Sills et al., 1999,
624952; Sills et al., 2001, 624922). Further, there was a high correlation between K-ras mutations and
loss of heterozygosity in the same chromosome in chloroprene-induced lung neoplasms in mice (Ton et
al., 2007, 625004). Similar increases in the frequencies of K-ras mutations in rodents were observed
in isoprene-induced lung neoplasms and vinyl chloride-induced hepatocellular carcinomas (NTP, 1998,
042076; U.S. EPA, 2000, 194536). Activated K-ras oncogenes were observed in lung tumors,
hepatocellular carcinomas, and lymphomas in B6C3Fi mice exposed to 1,3-butadiene (U.S. EPA,
2002, 052153). Activated K-ras oncogenes have not been found in spontaneously occurring liver
tumors or lymphomas, and are found in only 1/10 spontaneous forming lymphomas in B6C3Fi mice
(U.S. EPA, 2002, 052153).
Although the genetic toxicity database for chloroprene includes numerous studies covering a
range of standard test batteries, their results have been conflicting. In general, bacterial base pair
107
-------
substitution mutation (Salmonella typhimurium strains TA100 and TA 1535) assays have been positive
(Bartsch et al., 1979, 010689: Willems, 1980, 625049) while the bacterial frame shift (S. typhimurium
strains TA 97 and TA 98) assays have been nonpositive (NTP, 1998, 042076: Willems, 1978, 625048:
Willems, 1980, 625049). The observation of positive results in bacterial base pair substitution assays
is in concordance with the finding that mutations in H- and K-ras oncogenes in select neoplasms of
exposed mice manifest in base pair transversions (Sills et al., 1999, 624952: Sills et al., 2001, 624922).
In contrast, other studies (NTP, 1998, 042076) have reported nonpositive results for all bacterial
strains. Westphal et al. (1994, 625047) suggested that decomposition products of chloroprene may be
responsible for the mutagenicity seen in positive tests. Westphal et al. (1994, 625047) exposed bacteria
directly to liquid chloroprene in solution and observed no increase in mutagenicity, whereas positive
tests (Bartsch et al., 1979, 010689: Willems, 1978, 625048: Willems, 1980, 625049) were conducted
by exposure of bacteria to chloroprene in the air. Atmospheric exposures of chloroprene may result in
more degradation products being formed, thereby increasing the mutagenicity of the parent compound.
A positive result with all bacterial strains was observed when exposed to the major epoxide metabolite
of chloroprene, (l-chloroethenyl)oxirane, in solution (Himmelstein et al., 2001, 019013).
Conflicting results (positive in Vogel (1979, 000948): nonpositive in Foureman et al. (1994,
065173)) have also been reported for the in vivo Drosophila melanogaster sex-linked lethal mutation
assay. Differences observed may be due to differences in purity, strain susceptibilities, and sample
size. Chloroprene has been primarily nonpositive in the in vitro micronucleus assay (Drevon and
Kuroki, 1979, 010680: Himmelstein et al., 2001, 019013). in vivo chromosomal damage (NTP, 1998,
042076) assay, and bone marrow micronucleus assays (NTP, 1998, 042076: Shelby and Witt, 1995,
624921). The lack of genotoxic damage induced in bone marrow or blood by chloroprene suggests
that the carcinogenic activity of this chemical may be site specific. The in vivo toxicity of chloroprene
involves a balance of reactive epoxide formation and glutathione- or epoxide hydrolase-dependent
detoxification pathways. These pathways may be enhanced or more active in some tissues, thus
limiting DNA damage in those tissues. Bone marrow was not a target for cancer in the chronic
carcinogenicity bioassays (NTP, 1998, 042076), and the endpoints for chromosomal damage in this
tissue were nonpositive. Evidence for target organ-dependent mutagenicity is further supported by the
findings of K- and H-ras oncogene mutations in lung, forestomach, and Harderian gland neoplasms in
B6C3Fi mice (Sills et al., 1999, 624952: Sills et al., 2001, 624922). However, a positive result with all
bacterial strains was observed with the epoxide intermediate of chloroprene, (l-chloroethenyl)oxirane
(Himmelstein et al., 2001, 019013).
A comparative analysis by Melnick and Sills (2001, 051506) has shown that chloroprene,
isoprene, and butadiene share several tumor sites in rats (mammary gland, thyroid, and kidney) and
mice (hemangiomas and hemangiosarcomas [all organs], lung, liver, forestomach, Harderian gland,
and mammary gland). Similar to butadiene, the female mouse lung was the most sensitive site of
chloroprene carcinogenicity (Section 4.5.3 and Tables 4-24 and 4-27). There are also remarkable
similarities in the potency and shape of the dose response between both compounds. Detailed
108
-------
quantitative analysis (Melnick and Sills, 2001, 051506) has rated butadiene as being of slightly greater
or equal in potency at some of the common sites of tumor induction (mammary gland and Harderian
gland), and more importantly, of equal potency in the induction of the most sensitive tumor, lung
neoplasms in female mice.
In summary, the evidence supports the hypothesized mutagenic mode of action for chloroprene.
A mutagenic mode of carcinogenic action of chloroprene is supported by epoxide metabolite
formation, DNA-adduct formation, observation of in vivo and in vitro mutagenicity, and the well
known structure-activity relationship of similar epoxide-forming carcinogens. Chloroprene has been
found to be metabolized to epoxides by humans and rodents. The hypothesized mutagenic mode of
action is supported by evidence of base pair substitution mutations seen in H- and K-ras proto-
oncogenes in chloroprene-induced lung, forestomach, and Harderian gland neoplasms observed in the
NTP (1998, 042076) study.
4.7.3.3. Conclusions about the Hypothesized Mode of Action
As noted above, the hypothesis is that chloroprene carcinogenicity has a mutagenic mode of
action. This hypothesized mode of action is presumed to apply to all of the tumor types. The key
events in the hypothesized mutagenic mode of action are metabolism to reactive epoxide intermediates
followed by binding to DNA, which leads to mutation. Epoxide-forming agents are generally capable
of forming DNA adducts which in turn have the potential to cause genetic damage, including
mutations; mutagenicity, in turn, is a well-established cause of carcinogenicity. This chain of key
events is consistent with current understanding of the biology of cancer. Further, the mutagenic mode
of action hypothesis is strongly supported by analogy with another epoxide-forming compound,
1,3-butadiene. In addition, although alternative or additional modes of action for chloroprene
carcinogenicity may exist in certain situations (i.e., at high exposure levels), these modes of action
have not been definitively identified or supported by existing evidence.
Strength, Consistency, Specificity of Association. Data from NTP (1998, 042076) and Sills
et al. (1999, 624952: 2001, 624922) show codon-specific (codons 12, 13, and 61) mutations in the H-
and K-ras proto-oncogenes in chloroprene-induced lung, forestomach, and Harderian gland neoplasms.
The high incidence of ras proto-oncogene activation (37/46 lung, 27/27 Harderian gland, 4/7
forestomach) in tumors in treated animals, in contrast with the lower incidence of oncogene activation
in spontaneously occurring tumors (25/82 lung, 15/27 Harderian gland, 4/11 forestomach), provides
support for the role of mutation in the ras oncogene as a precursor to tumor formation in animals
treated with chloroprene. Similar findings of ras oncogene activation for isoprene (11/11 lung, 30/30
Harderian gland, 7/10 forestomach) and 1,3-butadiene (6/9 lung, 20/29 Harderian gland, 20/24
forestomach) were observed in tumors from animals treated with these structurally-related compounds
(Sills et al., 1999, 624952: Sills et al., 2001, 624922). These findings provide additional support for
the importance of ras proto-oncogene activation via mutation in the carcinogenesis of chloroprene and
related compounds.
109
-------
Dose-Response Concordance. High frequencies of K-ras codon 61 CTA mutations were
observed in lung tumors from animals exposed to the low- and mid-dose of chloroprene, but not the
high dose. Similarly high frequencies of K-ras mutations were observed at all doses in Harderian
gland tumors. There are a number of factors that might explain such observations. The higher
frequency of mutations at lower doses in lung neoplasms may indicate the saturation of one or more
metabolic pathways at higher doses or may suggest that non-ras mechanisms of genotoxicity are
operating at those doses. Dose-dependent differences in the mutation profile in the lung and Harderian
gland may be explained by differences in DNA-adduct formation or repair in low doses versus high
doses.
Temporal Relationships. In mice exposed to chloroprene, tumors were observed in a
significant fraction of the exposed animals after 2 years of exposure. DNA-adduct formation and
subsequent ras mutations were most likely early mutagenic events in the development of lung,
Harderian gland, and forestomach neoplasms. The observation that ras mutations occurred in benign
neoplasms in these organ systems (lung and Harderian gland adenomas and forestomach papillomas) is
supportive evidence of this. Additionally, in mice exposed to isoprene for 6 months and then allowed a
6 month recovery period, forestomach neoplasm with ras mutations did not regress (Melnick et al.,
1994, 625208). This suggests that ras mutations may have transformed forestomach epithelial cells at
an early time point and that the transformed cells progressed to neoplasia even after chemical exposure
had been terminated.
Biological Plausibility and Coherence. The biological plausibility of a mutagenic mode of
action for chloroprene is supported by evidence of mutations leading to ras proto-oncogene activation
in tumors from mice treated with chloroprene (NTP, 1998, 042076: Sills et al., 1999, 624952: Sills et
al., 2001, 624922). These studies provide the critical link between the in vitro evidence of
mutagenicity (positive results in S. typhiimurium strains 100 and 1535 that test for point mutations) and
tumor formation in a specific species. Similar findings with the structurally related chemicals
1,3-butadiene and isoprene and the lower incidence of spontaneously occurring tumors displaying ras
mutations in untreated animals (Sills et al., 1999, 624952: Sills et al., 2001, 624922) enhance the
database supporting this particular mode of action for chloroprene.
Additional evidence for the association between mutagenesis and tumor formation is the
observation that chloroprene exposure caused tumors in a wide variety of mouse tissues, including
lung, kidney, Harderian gland, mammary gland, forestomach, liver, skin, mesentery, and Zymbal's
gland (NTP, 1998, 042076). Tumors were also observed in a number of rat tissues, including oral
cavity, thyroid, lung, kidney, and mammary gland. Induction of tumors at multiple sites and in
different species is characteristic of carcinogens acting via mutagenesis (U.S. EPA, 2005, 086237).
Early-Life Susceptibility - According to the Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposures to Carcinogens (U.S. EPA, 2005, 088823) those exposed to
carcinogens with a mutagenic mode of action are assumed to have increased early-life susceptibility.
Data on chloroprene are not sufficient to develop separate risk estimates for childhood exposure.
110
-------
There are no data comparing the carcinogen!city of chloroprene after exposure during early life with
the carcinogenicity after exposure during adulthood. Exposure to chloroprene commenced at about
6 weeks of age in mice and rats, and continued through adulthood in the 2-year chronic assay (NTP,
1998. 042076).
Therefore, because the weight of evidence supports a mutagenic mode of action for chloroprene
carcinogenicity (Section 4.7.3.2), and in the absence of chemical-specific data to evaluate differences
in susceptibility, early-life susceptibility should be assumed and the age-dependent adjustment factors
(ADAFs) should be applied, in accordance with the Supplemental Guidance.
In conclusion, the weight of evidence supports a mutagenic mode of action for chloroprene
carcinogenicity and application of ADAFs to address assumed early-life susceptibility.
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
Bernauer et al. (2003, 625103) investigated cytochrome P450 variability in leukapheresed
samples from 50 humans as an indication of extrahepatic P450 variability via Western blotting and
immunoquantification. CYP2E1 was observed to have a median expression of 0.2 pmol/mg protein
and varied between 0.13 and 0.68 pmol/mg protein. The ratio between the 5th and 95th percentile was
3.3, which was the lowest level of variability in the six P450 isoforms investigated. Additionally,
Neafsey et al. (2009, 196814) identified, in a review of the open literature, a number of CYP2E1
genotypic and phenotypic polymorphisms in a number of human populations, and postulated that the
influence of CYP2E1 polymorphisms on adverse responses in exposed subjects would be expected to
be significant. However, the authors further state that the direction and magnitude of enzyme activity
changes due to polymorphisms is generally not well delineated and ultimately conclude that "the
evidence for particular CYP2E1 polymorphisms having a significant effect on enzyme activity in vivo
is too limited to support the population distribution of CYP2E1 enzyme activity based upon genotype."
They suggest that dietary, lifestyle, and physiological factors may exert substantial influence on
CYP2E1 phenotypes. Additionally, P450 mediated metabolism of chloroprene may be multifactoral,
with multiple individual CYPs playing a role. Thus the expression of one single CYP may not
adequately describe the possible variations within the human population. No data is currently
available on the toxicodynamic variability within the human population.
4.8.1. Possible Childhood Susceptibility
No direct evidence has been found that indicates children are more susceptible to the toxic
effects of chloroprene exposure than adults: exposures of children have not been reported and the
metabolic fate of chloroprene in humans has not been sufficiently characterized. However, there are a
number of issues that, when considered together, suggest that childhood may represent a lifestage with
increased susceptibility to chloroprene effects.
There are indications of reduced metabolic capacity and elimination in children relative to
adults that may be a source of susceptibility. Glutathione levels are rapidly depleted in response to in
111
-------
vitro (rat hepatocytes) and in vivo (Wistar rats) chloroprene exposures, suggesting a GSH-dependent
detoxification pathway (Summer and Greim, 1980, 064961). Additionally, the major metabolite of
chloroprene, (l-chloroethenyl)oxirane, is rapidly detoxified via epoxide hydrolase-mediated hydrolysis
in mouse liver microsomes (Himmelstein et al., 2001, 019012). The levels of both epoxide hydrolase
and glutathione transferase (GST) have been shown to be lower in infants than adults (Ginsberg et al.,
2004, 625124). Epoxide hydrolase is active at birth, but only at 50% of adult function for as long as
2 years. Evidence, although limited, suggests that GSTmu and GSTaB2 may be deficient (40-60% of
adult levels) in early life. This decrement in GST activity is especially relevant as GSTmu is critical to
epoxide conjugation to glutathione. Therefore, as both epoxide hydrolase and certain forms of GST
exhibit decreased activity in early life, newborns and young infants may experience higher and more
persistent blood concentrations of chloroprene and/or its metabolite than adults at similar dose levels.
Compensating mechanisms (i.e., other GST isozymes such as GSTpi) may be active in early life.
Reduced renal clearance in children may be another important source of potential susceptibility.
Excretion of chloroprene in exposed rats occurs through the elimination of urinary thioesters
(presumably glutathione conjugates) (Summer and Greim, 1980, 064961). Data indicating reduced
renal clearance for infants up to 2 months of age may suggest a potential to affect chloroprene
excretion, thus prolonging its toxic effects.
Further, a mutagenic mode of action is proposed for the observed carcinogenicity of
chloroprene (Section 4.7.3). In the absence of chemical-specific data to evaluate the differences
between adults and children, chemicals with such a mode of action are assumed to have increased
early-life susceptibility and age-dependent adjustment factors (ADAFs) should be applied, in
accordance with EPAs Supplemental Guidance for Assessing Susceptibility From Early-Life Exposure
to Carcinogens (U.S. EPA, 2005, 088823).
4.8.2. Possible Sex Differences
In lifetime studies conducted in the rat, mouse, and hamster, chloroprene was not shown to
exhibit any remarkable sex-related differences in effects with the exception of a more pronounced
neoplastic response in B6C3Fi female mice compared to males.
112
-------
5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)
The available data are inadequate to derive an oral RfD for chloroprene. There are no human
data involving oral exposure. The only lifetime oral study available (Ponomarkov and Tomatis, 1980,
075453) exposed rats to chloroprene at one dose (50 mg/kg/day) and only qualitatively reported
noncancer effects.
In summary, this study identified the liver (multiple liver necroses and degenerative lesions of
parenchymal cells), lung (severe congestion), and kidney (severe congestion) as potential target organs
for the oral toxicity of chloroprene; although, the available information was insufficient to characterize
toxicity outcomes or dose-response relationships. A route-to-route extrapolation from available
chronic inhalation data to oral data for the purposes of deriving an RfD was not performed due to the
inadequacies of the current chloroprene PBPK model (Section 3.5).
Therefore, an RfD was not derived due to the significant uncertainty associated with the oral
database for chloroprene and the lack of a validated PBPK model for route-to-route extrapolation.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
RfCs are derived for exposures via the inhalation route. In general, the RfC is an estimate of a
daily exposure to the human population (including susceptible subgroups) that is likely to be without
an appreciable risk of health effects over a lifetime. It is derived from a statistical lower confidence
limit on the benchmark dose (BMDL), a no-observed-adverse-effect level (NOAEL), a lowest-
observed-adverse-effect level (LOAEL), or another suitable point of departure (POD), with
uncertainty/variability factors applied to reflect limitations of the data used. The inhalation RfC 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) effects and systems peripheral
to the respiratory system (extra-respiratory or systemic effects). It is generally expressed in mg/m3.
5.2.1. Choice of Principal Study and Critical Effect(s)
While literature exists on the carcinogenic potential of chloroprene exposure in humans, no
human studies are available that would allow for the quantification of sub-chronic or chronic
noncancer effects. Two inhalation studies were identified in the literature and considered for the
principal study for derivation of an RfC: a 2-year chronic study in B6C3Fi mice and F344 rats (NTP,
1998, 042076), and a 2-year chronic study in Wistar rats and Syrian golden hamsters (Trochimowicz et
al., 1998. 625008).
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
113
-------
The chronic NTP inhalation bioassay (1998, 042076) exposed groups of 50 mice and rats of
each sex to 0, 12.8, 32 or 80 ppm chloroprene for 6 hours/day, 5 days/week for 2 years. This study
observed a range of chloroprene-induced nonneoplastic effects across several organ systems including
the respiratory tract (from the nose to the alveolar region) in both mice and rats, the kidneys of rats and
male mice, the forestomach of male and female mice and the spleen of male and female mice (NTP,
1998, 042076). In addition, many histopathological lesions were significantly increased compared to
controls at the lowest level tested (12.8 ppm), including alveolar epithelial hyperplasia in male and
female rats, bronchiolar hyperplasia in male and female mice, lung histiocytic cell infiltration in female
mice, hematopoietic cell proliferation in the spleen in female mice, and atrophy, necrosis, and chronic
inflammation of the nasal olfactory epithelium in male rats.
Trochimowicz et al. (1998, 625008) exposed three groups of 100 Wistar rats and Syrian
hamsters of each sex to chloroprene at 0, 10, or 50 ppm for 6 hours/day, 5 days/week for up to
18 months (hamsters) or 24 months (rats). Unlike the NTP (1998, 042076) study, this study did not
observe a wide range of nonneoplastic effects in multiple organ systems. Gross pathology revealed
that the lungs from rats exposed at 10 and 50 ppm had markedly lower incidences of pathological
changes consistent with, and characterized as, chronic respiratory disease than did controls. Male
hamsters exhibited a concentration-related decrease in the incidence of pale adrenal glands. The only
remarkable nonneoplastic lesions that statistically increased in male and female rats were observed in
the liver and lungs at 50 ppm: an increase in foci of cellular alteration in the liver, and mild changes,
such as lymphoid aggregates around the bronchi, bronchiole, and blood vessels, in the lungs.
Accidental failure of the exposure chamber ventilation system suffocated 87 male and 73 female rats in
the low-exposure (10 ppm) group during week 72 of exposure, and limited the histopathological
examinations performed in this study. Only the livers of rats that died accidentally were processed for
microscopic examination. No morphological disturbances were noted in the liver of low-exposure
group animals. The only nonneoplastic change seen in hamsters was a generalized amyloidosis (in the
liver, kidneys, spleen, and adrenals) that was lower in incidence in the 50 ppm exposed group
compared with controls.
The chronic NTP (1998, 042076) study was chosen as the principal study for the derivation of
the RfC. Based on the noncancer database for chloroprene, this study demonstrated exposure
concentration-related effects more extensively than any other study. It was a well conducted study that
utilized 50 animals per sex, per exposure group, a range of exposure concentrations based on the
results of preliminary, shorter-duration studies (16 days and 13 weeks), and thoroughly examined the
observed toxicity of chloroprene in two species. Trochimowicz et al. (1998, 625008) was not chosen
as the principal study due to concerns regarding the high mortality observed in the low dose male and
female rats due to the failure in the exposure chamber ventilation system. The high mortality in this
dose group prevented histopathological examination of most organ systems (except for liver samples)
and precluded any firm conclusions on dose-response characteristics from being drawn. Also, a lack of
effects at similar exposure levels as the NTP (1998, 042076) study (Trochimowicz et al. (1998,
114
-------
625008) (see Section 4.7.2.2 for discussion of potential causes of differences in observed toxicity
between the NTP and Trochimowicz studies) was observed and influenced the choice to not select the
Trochimowicz et al. (1998, 625008) as the principal study.
From the NTP (1998, 042076) study, all nonneoplastic lesions that were statistically increased
in mice or rats at the low- or mid-exposure concentration (12.8 or 32 ppm) compared to chamber
controls, or demonstrated a suggested dose-response relationship in the low- or mid-exposure range in
the absence of statistical significance, were considered candidates for the critical effect. Nonneoplastic
effects identified as secondary to neoplastic effects (i.e., histiocytic cell proliferation in mice) were not
considered candidates for the critical effect. The candidate endpoints included olfactory suppurative
inflammation, bronchiolar hyperplasia, kidney (renal tubule) hyperplasia, forestomach epithelial
hyperplasia, and splenic hematopoietic cell proliferation in mice, and olfactory atrophy, olfactory basal
cell hyperplasia, olfactory metaplasia, olfactory necrosis, olfactory chronic inflammation, alveolar
epithelial hyperplasia, and kidney (renal tubule) hyperplasia in rats (Table 5-1).
115
-------
Table 5-1. Incidences of nonneoplastic lesions resulting from chronic exposure
(ppm) to chloroprene considered for identification of critical effect
Species
Mice
Rats
Tissue
Nose
Lung
Kidney
Fore-
stomach
Spleen
Nose
Lung
Kidney
Endpoint
Suppurative
inflammation
Bronchiolar
hyperplasia
Renal tubule
hyperplasia
Epithelial
hyperplasia
Hematopoietic
cell
proliferation
Atrophy
Basal cell
hyperplasia
Metaplasia
Necrosis
Inflammation,
chronic
Alveolar
epithelial
hyperplasia
Renal tubule
hyperplasia
Male
0
a
0/50
2/50
4/50
26/50
3/50
0/50
6/50
0/50
0/50
5/50
14/50
12.8
a
10/50C
16/49C
6/48
22/49
12/50b
0/50
5/50
11/50C
5/50b
16/50C
20/50
32
a
18/50C
17/50C
7/49
35/50d
46/49c
38/49c
45/49c
26/49c
9/49-
14/49b
28/50c
80
a
23/50c
18/50C
29/50c
31/50d
48/49c
46/49c
48/49c
19/49C
49/49°
25/50c
34/50c
Female
0
0/50
0/50
a
4/50
13/50
0/49
0/49
0/49
0/49
a
6/49
6/49
12.8
1/49
15/49C
a
3/49
25/49d
1/50
0/50
1/50
0/50
a
22/50c
6/50
32
3/49b
12/50C
a
8/49
42/49d
40/50C
17/50C
35/50c
8/50c
a
22/50c
11/50
80
4/50c
30/50C
a
27/50c
39/50d
50/50C
49/50c
50/50C
12/50C
a
34/50c
21/50
aEndpoint not considered for selection of critical effect.
bStatistical significance p < 0.05.
Statistical significance p < 0.01.
dReported as statistically significantly greater than controls, but level of significance not reported.
Source: NTP (1998. 042076).
5.2.2. Methods of Analysis
This assessment used benchmark dose (BMD) methodology, where possible, to estimate a POD
for the derivation of an RfC for chloroprene. The use of the BMD methodology was preferred for the
estimation of a POD for many reasons, including consideration of the shape of the entire dose-response
curve and estimation of the experimental variability associated with the calculated dose-response
relationship. Use of BMD methods involves fitting mathematical models to the observed dose-
response data and provides a BMD and its 95% lower confidence limit (BMDL) associated with a
predetermined benchmark response (BMR). The BMDL is then used in lieu of the NOAEL or LOAEL
as the POD for deriving the RfC. The suitability of these methods to determine a POD is dependent on
the nature of the toxicity database for a specific chemical. The data for some endpoints were not
116
-------
amenable to BMD modeling for a number of reasons, including the observation of maximal or near-
maximal response at the lowest dose tested, the failure to achieve an incidence greater than the BMR at
any dose level, or equal incidence in all dose groups. Additionally, even when some datasets were
deemed adequate for BMD modeling, no model provided adequate model fit. In these cases, the
NOAEL/LOAEL approach was used.
A BMR of 10% extra risk is typically chosen as a response level for dichotomous data and is
recommended for the BMR when using dichotomous models to facilitate a consistent basis of
comparison across assessments and endpoints (U.S. EPA, 2000, 052150). For the data from the NTP
(1998, 042076) study, a BMR of 10% extra risk was used initially. In addition to the incidence of the
endpoints, the NTP (1998, 042076) study also reported the severity scores for individual animals in
each dose group, thus making it possible to determine whether the endpoints were increasing in
severity as well as incidence with dose (Table B-l).
Due to the nature and severity of the nasal degenerative effects (i.e., olfactory atrophy and
necrosis), and the proximity of the BMDLio values to the observed LOAEL compared to other
endpoints (Table 5-2), a BMR of 5% was considered to be appropriate for these olfactory endpoints.
The nature of the observed nasal lesions potentially included the loss of Bowman's glands and
olfactory axons in more severe cases. Effects that occur in the underlying lamina propria and basal
layer of the olfactory epithelium may be indicative of more marked nasal tissue injury. For all other
endpoints, a BMR of 10% was chosen as the response level.
All available dichotomous models in the EPA (2009, 200772) BMD software (BMDS version
2.1.1) were fit to the incidence data for lung, nasal, and systemic effects in rats and mice (Table 5-1).
The models selected for each particular endpoint were chosen based on global and local goodness-of-
fit criteria (global p-value and chi-square [%2] residual values, respectively) and visual inspection. The
global goodness-of-fit p-value provides an indication of how well a particular model fits the observed
dose-response data across the entire range of doses, whereas the %2 residual gives an indication of how
well the model fits at the dose group closest to the calculated BMD. A global p-value > 0.1 and %2
residual < |2| is required for a model to be considered as adequately fitting the dose-response data.
Finally, a visual inspection of the dose-response curve is necessary in order to determine whether the
calculated dose-response curve is appropriate (e.g., monotonically increasing). When multiple
appropriately fitting models are identified for a particular endpoint, the "best" model must be selected
out of the group. When the calculated BMDL values are within a threefold difference of one another
for a particular endpoint, indicating a low degree of model-dependence, the model with the lowest
Akaike Information Criterion (AIC) is selected as the best model. The AIC awards the most
parsimonious model so that models with higher numbers of parameters are only selected as the best
fitting model when they significantly improve model fit. When the calculated BMDL values are not
within a threefold difference, model dependence is assumed and the model returning the lowest BMDL
is selected (U.S. EPA, 2000, 052150). Details of the BMD modeling analysis, including all relevant
model-fit criteria and final model selection information, are provided in Appendix B.
117
-------
The BMDs and BMDL values associated with an extra risk of 10% or 5% (endpoint-dependent)
for the best-fitting models are shown in Table 5-2. NOAEL and LOAEL values were used as potential
POD values for the endpoints not deemed appropriate for BMD modeling, or when adequate model fit
could not be achieved by any model.
5.2.3. Exposure Duration and Dosimetric Adjustments
Because an RfC is a measure that assumes continuous human exposure over a lifetime, data
derived from animal studies need to be adjusted to account for the noncontinuous exposure protocols
used in animal studies. In the NTP (1998, 042076) study, rats were exposed to chloroprene for
6 hours/day, 5 days/week for 2 years. Therefore, the duration-adjusted PODs for lung, nasal, and
systemic lesions in rats and mice were calculated as follows:
PODADJ (ppm) = BMDL (ppm) x hours exposed per day724 hours x days exposed per
week/7days
RfCs are typically expressed in units of mg/m3; the above ppm value needs to be converted
using the chemical specific conversion factor of 1 ppm = 3.62 mg/m3 (Table 2-1) for chloroprene.
Therefore, the final PODADJ values were calculated as follows:
PODADj (mg/m3) = PODADj(ppm) x 3.62 mg/mVl ppm
For example, for olfactory atrophy in the male rat, the PODADJ would be calculated as follows:
PODADj(ppm) = 3.5 ppm x 6 hours/24 hours x 5 days/7 days
PODADj(ppm) = 0.6 ppm
Q Q .
PODADj (mg/m3) = 0.6 ppm x 3.62 mg/mVl ppm
PODADJ (mg/m3) = 2.3 mg/m3
The calculated PODADj (mg/m3) values for all considered endpoints are presented in the last
column of Table 5-2.
118
-------
Table 5-2. Duration adjusted point of departure estimates for best fitting models
of the BMD from chronic exposure to chloroprene
Endpoint
Species/
Sex
NOAEL
(ppm)
LOAEL
(ppm)
Model3
BMR
BMDb
(ppm)
BMDLb
(ppm)
PODAD/
(mg/m3)
Nasal Effects -Olfactory
Atrophy
Basal cell
hyperplasia
Metaplasia
Necrosis
Chronic
inflammation
Suppurative
inflammation
Rat/male11
Rat/female
Rat/male
Rat/female
Rat/male
Rat/female
Rat/male
Rat/female
Rat/male
Mouse/female
—
12.8
12.8
12.8
12.8
12.8
—
12.8
-
12.8
12.8
32
32
32
32
32
12.8
32
12.8
32
Logistic"1
e
e
Log-probitf
e
e
Log-probitd
Log-probitf
Log-logistic"1
e
5
~
~
10
~
~
5
5
10
~
4.9
~
~
23.5
~
~
5.6
24.8
14.6
~
3.5
~
~
19.7
~
~
4.5
19.7
9.3
~
2.3
8.3
8.3
12.7
8.3
8.3
2.9
12.7
6.0
8.3
Lung Effects
Alveolar
hyperplasia
Bronchiolar
hyperplasia
Rat/male
Rat/female11
Mouse/male
Mouse/female
~
—
~
~
12.8
12.8
12.8
12.8
Log-logistic
Log-logistic
Log-logistic
g
10
10
10
~
11.4
4.9
7.5
~
7.1
3.3
5.6
~
4.6
2.1
3.6
8.3
Other Organ Systems Effects
Kidney (renal
tubules)
hyperplasia
Forestomach
epithelial
hyperplasia
Splenic
hematopoietic
proliferation
Rat/male
Rat/female
Mouse/male
Mouse/male
Mouse/female
Mouse/male
Mouse/female11
12.8
32
—
32
32
12.8
-
32
80
12.8
80
80
32
12.8
Log-logistic
Log-probit
e
Multistage
Multistage
e
Probitd
10
10
~
10
10
~
10
6.5
32.5
~
24.7
31.0
~
4.0
4.0
23.5
~
20.5
19.3
~
3.3
2.6
15.2
8.3
13.3
12.5
8.3
2.1
aBest fitting model as determined by goodness-of-fit statistics. Bold numbers indicate which value (BMDL, NOAEL, or
LOAEL) is used in calculation of POD ADJ.
bBMR = benchmark dose response; BMD = benchmark dose; BMDL = statistical lower confidence limit of the benchmark
dose.
Duration adjusted PODADJ (mg/m3) = BMDL [ppm] x (3.62 mg/m3/ppm) x (5 days/7days) x (6 hours/24 hours), in
accordance with EPA policy (U.S. EPA, 2002, 088824).
dHigh dose group was dropped in order to obtain adequate model fit.
eDid not model endpoint (reasons include: maximal response in lowest dose showing response over controls, response levels
did not achieve 10% incidence, incidence equal in all doses with response). Therefore, the NOAEL/LOAEL approach is
recommended to determine a PODADJ.
fDichotomous Hill model had lowest AIC, but model output warned that BMDL estimate was "imprecise at best." Therefore,
the model with the next lowest AIC was chosen (Appendix B for details).
8No model fits appropriately according to fit statistics or visual inspection.
hBold indicates critical effects.
Source: NTP (1998. 042076)
119
-------
The results of BMD modeling indicated that olfactory atrophy in the male rat, alveolar
hyperplasia in the female rat, and splenic hematopoietic cell proliferation in the female mouse were the
most sensitive endpoints, with a PODADJ values of 2.3, 2.1, and 2.1 mg/m3, respectively. Each of these
endpoints also represents the most sensitive effect from each category of effects, i.e., nasal, lung, and
systemic effects, thus utilizing toxicity information from multiple organ systems or tissues. The
observation that the PODAoj values from these co-critical effects are similar or identical provides
support for their use in deriving the RfC. Two additional endpoints, specifically renal tubule
hyperplasia in male mice and necrosis in the male rat, had somewhat higher PODAoj values of 2.6 and
2.9 mg/m3, respectively. For the co-critical effects, olfactory atrophy, alveolar hyperplasia, and splenic
hematopoietic cell proliferation, after rounding to one significant figure, the PODADJ resulted in a
value of 2 mg/m3, which was used as the point of departure for deriving the RfC.
PODADj (mg/m3) = 2 mg/m3
Chloroprene is a relatively water-insoluble, nonreactive gas, with an approximate blood:air
partition coefficient of less than 10 (Table 3-1), that induces a range of nasal, thoracic, and systemic
noncancer effects. Water-insoluble, nonreactive chemicals typically do not partition greatly into the
aqueous mucus coating of the upper respiratory system. Rather, they tend to distribute to the lower
portions of the respiratory tract where larger surface areas and the thin alveolar-capillary barrier
facilitate uptake (Medinsky and Bond, 2001, 016157). The observation of systemic (i.e.,
nonrespiratory) effects resultant from chloroprene exposure clearly indicates the compound is absorbed
into the bloodstream and distributed throughout the body. Further, the distribution of lesions (olfactory
effects, but no respiratory mucosal damage) is indicative of a critical role for blood borne delivery and
in situ metabolic activation. The absence of respiratory mucosal injury suggests that direct reactivity
of the parent compound is not likely involved. Rather, the pattern of respiratory tract effects seen
following chloroprene exposure is consistent with what is known about its metabolism and the
expression of cytochrome P450 enzymes in the olfactory mucosa and lower respiratory tract in rats.
The proposed mode of action of chloroprene involves the conversion of the parent compound into its
reactive epoxide metabolite by P450 isoform CYP2E1. The olfactory mucosa of rats has been shown
to specifically express CYP2E1 at levels more similar to hepatic levels than any other nonhepatic
tissue examined (Thornton-Manning and Dahl, 1997, 597688). Himmelstein et al. (2004, 625152)
observed that the microsomal fraction of rat lung homogenates was active in the metabolic oxidation of
chloroprene into (l-chloroethenyl)oxirane at levels between 10-30% that of liver microsomes. In situ
conversion of chloroprene into its highly reactive epoxide metabolite in the olfactory epithelia and
lower respiratory tract may facilitate its uptake in these tissues and explain a portion of its biological
activity in those regions. Evidence for metabolic activation in the respiratory tract combined with the
observation that chloroprene induces effects in organ systems distal to the portal-of-entry, consistent
120
-------
with the parent compound's water-insoluble and nonreactive chemical properties, suggest that
chloroprene's principal mode of action does not involve direct reactivity of the parent compound at the
portal of entry.
Consequently, the selected critical effects, olfactory atrophy, alveolar hyperplasia, and splenic
hematopoietic cell proliferation, are assumed to result primarily from systemic distribution and the
human equivalent concentration (HEC) for chloroprene was calculated by the application of the
appropriate dosimetric adjustment factor (DAF) for Category 3 gases, in accordance with the U.S. EPA
RfC methodology (U.S. EPA, 1994, 006488). DAFs are ratios of animal and human physiologic
parameters, and are dependent on the nature of the contaminant (particle or gas) and the target site
(e.g., respiratory tract or remote to the portal-of-entry) (U.S. EPA, 1994, 006488). For Category 3
gases with systemic effects, the DAF is expressed as the ratio between the animal and human blood:air
partition coefficients:
DAF = (Hb/g)A/(Hb/g)H
where:
(Ftb/g)A = the animal blood:air partition coefficient
(Hb/g)H = the human blood:air partition coefficient
DAF = 7.8/4.5
DAF= 1.7
In cases where the animal blood:air partition coefficient is higher than the human value
(Table 3-1), resulting in a DAF>1, a default value of 1 is substituted (U.S. EPA, 1994, 006488).
Therefore, the HEC for olfactory atrophy, alveolar hyperplasia, and splenic hematopoietic cell
proliferation in male F344/N rats, female F344/N rats, and female B6C3Fi mice, respectively is
calculated as follows:
PC-DHEC (mg/m3) = PODADJ (mg/m3) x DAF
= PODADJ (mg/m3) x 1.0
= 2 mg/m3 x 1.0
= 2 mg/m3
Therefore, the PODnEc of 2 mg/m3 for the co-critical effects of olfactory atrophy, alveolar
hyperplasia, and splenic hematopoietic cell proliferation were selected for the derivation of the RfC for
chloroprene.
121
-------
5.2.4. RfC Derivation-Including Application of Uncertainty Factors
APODHEc value of 2 mg/m3 for increased incidence of olfactory atrophy, alveolar hyperplasia,
and splenic hematopoietic cell proliferation in male F344/N rats, female F344/N rats, and female
B6C3Fi mice, respectively (NTP, 1998, 042076) was used as the PODHEc to derive the chronic RfC for
chloroprene. A total UF of 100 was applied to this PODHEc as described below:
• AnUFAof 3 (101/2 = 3.16, rounded to 3) was used to account for uncertainty in extrapolating
from laboratory animals to humans (i.e., interspecies variability). This uncertainty factor is
comprised of two separate and equal areas of uncertainty to account for differences in the
toxicokinetics and toxicodynamics of animals and humans. In this assessment, toxicokinetic
uncertainty was accounted for by the calculation of a human equivalent concentration by the
application of a dosimetric adjustment factor as outlined in the RfC methodology (U.S. EPA,
1994, 006488). As the toxicokinetic differences are thus accounted for, only the toxicodynamic
uncertainties remain, and an UF of 3 is retained to account for this residual uncertainty.
• A 10-fold UFH was used to account for variation in susceptibility among members of the human
population (i.e., interindividual variability). Only limited information is available to predict
potential variability in human susceptibility, including some data regarding the human variability
in expression of enzymes involved in chloroprene metabolism (e.g., metabolic activation via P450
isoform CYP2E1) (Section 4.8). Due to this limited data on variations in susceptibility within the
human population, a default 10-fold UFH is applied.
• An UFS was not needed to account for subchronic-to-chronic extrapolation because a chronic
inhalation study is being used to derive the chronic RfC.
• An UF for LOAEL-to-NOAEL extrapolation was not applied because the current approach is to
address this factor as one of the considerations in selecting a BMR for benchmark dose modeling.
In this case, a BMR of 5% change in olfactory atrophy and a BMR of 10% change in alveolar
hyperplasia and splenic hematopoietic cell proliferation was selected under an assumption that
these BMR levels represent a minimal biologically significant change for these endpoints.
• An UF of 3 was used to account for deficiencies in the database. The maj or strength of the
database is the observation of exposure-response effects in multiple organ systems in a well-
designed chronic inhalation study that utilized 50 animals per sex per dose group, a range of doses
based on the results of preliminary, shorter-duration studies (16 days and 13 weeks), and thorough
examination of the observed toxicity of chloroprene in two species (rat and mouse). The database
further contains another chronic inhalation bioassay investigating outcomes in another species
(hamster), and well-designed embryotoxicity, teratological, and reproductive toxicity studies. The
database also contains subchronic studies and chronic studies observing potential neurotoxic and
122
-------
immunotoxic effects. A limitation in the database is the lack of a full two-generation reproductive
toxicity study (the Appelman and Dreef van der Meulen (1979, 064938) unpublished study
exposed FO and Fl rats to chloroprene, but did not allow the Fl rats to mate).
Application of this 100-fold composite UF and rounding to one significant digit yields the
calculation of the chronic RfC for chloroprene as follows:
RfC = PODHEc - UF = 2 mg/m3 + 100 = 2 x 10'2 mg/m3
5.2.5. Previous RfC Assessment
The IRIS Program has not previously evaluated the noncancer inhalation toxicity of
chloroprene.
5.2.6. RfC Comparison Information
Figure 5-1 presents PODs, applied UFs, and derived sample RfCs for all of the endpoints from
the chronic inhalation NTP (1998, 042076) study that were modeled with BMDS (version 2.1.1),
including nasal, pulmonary, and systemic effects in male and female rats and mice. Of the considered
studies, the NTP (1998, 042076) study was considered the most suitable to derive an RfC. The
endpoints considered for the critical effects from the NTP (1998) study included any histopathological
lesion that was significantly increased in the lowest dose group relative to controls. The PODs are
based on the BMDL of the best fitting model from BMD modeling and were adjusted for duration and
dosimetry before applications of uncertainty factors.
123
-------
100.00
10.00
i.oo -
0.10 I
0.01
AHEC
• Animal to Human
D Human Variation
d Database
• RfC
'fc fl
t
'.U;
>; s
_c —i
Q. Q
P ^
O E
QJ *H
li
= CD
ro
t/l QJ
ro -=
CQ
SE
I*
« a
ro —i
&i
|i
t/i
i- ^
_ro o
.9 E
o
CQ
CJ
Q. g
ro ^
c ro
E
_CD
5
_Q
CJ
C
-a
QJ o
9: -^
fi
TO CQ
"CD QJ-
-C iyi
'5. o
m E
u ^
I1
t/i
QJ
Q.
-------
considered to be the most sensitive endpoints because they returned the lowest PODADJ values
compared to all other considered endpoints.
Uncertainty exists surrounding the assumed blood-borne delivery of chloroprene (or its reactive
epoxide metabolite) to the target tissues. The current RfC methodology attempts to group chemicals
into one of three discrete categories based on their physio-chemical properties and presumed
toxicokinetics (i.e., regional gas uptake). Using this scheme, chloroprene would be best classified as a
Category 3 gas, being relatively water insoluble and nonreactive, and would be expected to elicit
extrarespiratory effects. The External Peer Reviewers supported the classification of chloroprene as a
Category 3 gas.
If chloroprene's mode of action were considered to be more characteristic of a Category 1 gas,
which would be expected to exhibit direct, portal-of-entry effects, DAF values for the respiratory
endpoints might have been derived differently in accordance with RfC methods (U.S. EPA, 1994,
006488), lowering the RfC for the observed nasal effects and raising the RfC for the observed
pulmonary effects. However, as described in Section 5.2.3, evidence for metabolic activation in the
respiratory tract combined with the observation that chloroprene induces adverse effects in organ
systems distal to the portal-of-entry, consistent with the parent compound's water-insoluble and
nonreactive chemical properties, suggest that chloropene's principal mode of action does not involve
direct reactivity of the parent compound at the portal of entry. In addition, the nasal degenerative
lesions induced by exposure to chloroprene are potentially characterized by the loss of Bowman's
gland in more severe cases. A nasal effect such as this, observed in the lamina prorpia, is indicative of
systemic distribution. Consequently, a DAF of 1 (for systemic effects of Category 3 gas) was applied
in derivation of the RfC. Analysis of the existing inhalation dosimetry modeling database supports the
application of a DAF of 1 to be appropriate (U.S. EPA, 2009, 625038). Application of these models to
gases that have similar physicochemical properties and induce similar nasal effects as chloroprene
estimate DAFs > 1.
Choice of Model for BMDL for Derivations. When the high dose group data was dropped,
the logistic model fit the data for olfactory atrophy in the male rat (global goodness-of-fit p-value =
0.2655). Data points for this endpoint are adequately predicted near the BMD (%2 residuals for control
and low dose group are 0.847 and -0.597, respectively). The log-logistic model fit the data for alveolar
hyperplasia in the female rat adequately (global goodness-of-fit p-value = 0.1779). Data points for this
endpoint are adequately predicted near the BMD (%2 residuals for control and low dose group are
-0.453 and 1.536, respectively). When the high dose group data was dropped, the probit model fit the
data for splenic hematopoietic cell proliferation in the female mouse adequately (global goodness-of-fit
p-value = 0.9466). Data points for this endpoint are well-predicted near the BMD (%2 residuals for
control and low dose group are 0.033 and -0.052, respectively). Use of other models for any of these
endpoints would either increase or decrease the RfC by approximately 50%. However, the selected
models are the most appropriate for RfC derivation based on current Draft Benchmark Dose Technical
Guidance (U.S. EPA, 2000, 052150).
125
-------
Choice of BMR There is uncertainty in the selection of the benchmark response (BMR) level.
Due to the nature and severity of the nasal degenerative effects (i.e., olfactory atrophy and necrosis),
and the proximity of the BMDLio values to the observed LOAEL compared to other endpoints
(Table 5-2), a BMR of 5% was considered to be appropriate for these olfactory endpoints. The
observed nasal lesions were characterized to potentially include the loss of Bowman's glands and
olfactory axons in more severe cases. Effects such as these which occur in the underlying lamina
propria and basal layer of the olfactory epithelium might be indicative of more marked nasal tissue
injury. For all other endpoints, a BMR of 10% was chosen as the response level. The use of a BMR of
5% versus 10% is approximately a twofold difference across all endpoints.
Statistical Uncertainty at POD. For the logistic, log-logistic, and probit model applied to
olfactory atrophy, alveolar hyperplasia, and splenic hematopoietic cell proliferation, respectively, there
is a reasonably small degree of statistical uncertainty at the 5% or 10% extra risk level (the point of
departure for derivation of the RfC), with the BMDL values being about 20-30% below the BMD.
Choice of Bioassay. The NTP (1998, 042076) chronic inhalation study was used for
development of the RfC because it was a well designed study that was conducted in 2 relevant species,
used 50 animals per sex per exposure group, and thoroughly examined a wide-range of appropriate
toxicological endpoints. The other chronic bioassay (Trochimowicz et al., 1998, 625008) was
discounted for use as the principal study due to interpretational difficulties (i.e., high, accidental
mortality in low dose animals resulting from the failure of the ventilation system) and a general lack of
effects at exposure levels similar to those showing effects in the NTP (1998, 042076) study.
Choice of Species. The RfC was based on increased incidence of olfactory atrophy, alveolar
hyperplasia, and splenic hematopoietic cell proliferation in male rats, female rats, and female mice,
respectively, exposed to chloroprene via inhalation for 2 years. Use of other effects observed in
rodents would result in RfCs up to approximately ten times greater than the selected RfC.
Human Population Variability. The extent of inter-individual variation of chloroprene
metabolism in humans has not been well characterized. Expression levels of extrahepatic CYP2E1
have been shown to vary by approximately threefold (Bernauer et al., 2003, 625103). Neafsey et al.
(2009, 196814) concluded that evidence for particular CYP2E1 polymorphisms having significant
effects on enzyme activity in vivo is too limited to support generalized statements on populational
distribution of CYP2E1 activity based on genotype. A number of issues, including lower enzyme
levels and renal clearance in children, potential distribution of chloroprene to breast milk, and the
proposed mutagenic mode of action for chloroprene suggest that childhood may represent a potentially
susceptible lifestage to chloroprene toxicity. The 10-fold default uncertainty value is applied to the
PODHEc primarily due to the limited data on human variability or potential susceptible subpopulations.
126
-------
Table 5-3. Summary of Uncertainties in the Chloroprene noncancer risk
assessment
Consideration
Choice of endpoint
Choice of model
forBMDL
derivation
Choice of BMR
Statistical
uncertainty at
POD
Choice of bioassay
Choice of species
Human population
variability
Completeness of
the database
Potential Impact"
Use of other
endpoints could t
RfC by up to a factor
of 10.
Other models would
t or | RfC.
Other BMR values
would t or | RfC by
a factor of about 2.
RfC would be -20
to 30% higher if
BMD (vs. BMDL)
were used.
Other bioassays
could t or | RfC.
RfC would t or | if
based on another
species.
RfC could t or | if a
nondefault value of
UF was used.
RfC could t or | if a
different UF for
database limitations
was applied.
Decision
RfC is based on effects
with the lowest PODADj,
increased incidence of
olfactory atrophy,
alveolar hyperplasia,
and splenic
hematopoietic cell
proliferation.
Logistic, log-logistic
and probit model used.
BMR of 5% and 10%
extra risk chosen
(endpoint-dependent) .
BMDL used as POD per
U.S. EPA guidance
(2000, 052150).
NTP (1998. 042076)
used as critical study.
Rats and mice chosen.
10-fold uncertainty
factor applied to derive
the RfC.
Threefold uncertainty
factor applied to derive
the RfC.
Justification
Chosen endpoints were considered to be the
most sensitive (based on PODADJ values).
The observed systemic toxicity is consistent
with the physio-chemical properties of
chloroprene. Selection of the co-critical
effects was based on the PODADJ consistent
with peer reviewer comments.
Draft Benchmark Dose Technical Guidance
(U.S. EPA, 2000, 052150) used to choose
models based on global and local measures
of model fit.
BMR of 5% and 10% extra risk (endpoint-
dependent) was chosen based on the
assumption that a 5% or 10% increase in
incidence of the effects represent a
minimally biologically significant effect
and is consistent with external peer
reviewer recommendations.
Size of bioassay results in sampling
variability; lower bound is 95% confidence
interval on administered exposure.
Other bioassays were available but were
discounted as principal study due to lack of
effects or interpretational difficulties. The
chosen bioassay was well-conducted and
reported and resulted in the lowest BMDL
for derivation of RfC.
RfC is based on the most sensitive
endpoints (incidence of olfactory atrophy,
alveolar hyperplasia, and splenic
hematopoietic cell proliferation) in the most
sensitive species (rat and mouse), based on
PODADJ
10-fold UF, the default value, is applied
principally because of limited data on
human variability or potential susceptible
subpopulations.
Threefold UF is applied as a major strength
of the database is the inclusion of a well-
designed chronic inhalation study
investigating effects in multiple species. A
limitation of the data is the lack of a multi-
generational reproductive/developmental
study.
a| = increase, { = decrease.
127
-------
5.4. CANCER ASSESSMENT
5.4.1. Choice of Study/Data-with Rationale and Justification
Both epidemiological and toxicological investigations of chloroprene carcinogenicity were
available. Epidemiological studies of chloroprene provided evidence of associations between liver or
lung cancer risk and occupational exposure to chloroprene (Section 4.7.); however, study limitations
precluded developing quantitative risk estimates from these studies. Two chronic bioassays were
available, NTP (1998, 042076) and Trochimowicz et al. (1998, 625008). In the NTP (1998, 042076)
study, groups of 50 male and female F344 rats and B6C3Fi mice were exposed via inhalation to 0,
12.8, 32, or 80 ppm chloroprene for 6 hours/day, 5 days/week for 2 years. Examination of appropriate
toxicological endpoints in both sexes of rats and mice was included. Tumor incidences were elevated
with increasing exposure level at numerous sites across all sex/species combinations, involving point
of contact in the respiratory system and more distant locations. Trochimowicz et al. (1998, 625008)
studied groups of 100 male and female Wistar and Syrian golden hamsters exposed via inhalation to 0,
10, or 50 ppm chloroprene for 6 hours/day, 5 days/week for up to 18 months (hamsters) or 24 months
(rats). This study was not considered for quantification purposes, due to less pronounced sensitivity in
the tested animals to neoplastic effects at similar exposure levels as in the NTP (1998, 042076) study,
in part associated with high accidental mortality in the low-dose rats (Section 4.2.2. for study details).
The NTP (1998, 042076) study was used for development of an inhalation unit risk.
5.4.2. Dose-Response Data
The NTP (1998, 042076) study incidence data are summarized in Tables 5-4 (mice) and 5-5
(rats). Mice demonstrated statistically significant or biologically noteworthy increases in tumor
incidence at multiple sites: hemangiomas or hemangiosarcomas (all organs), alveolar /bronchiolar
adenomas or carcinomas, forestomach (squamous cell papillomas or carcinomas), Harderian gland
(adenomas and carcinomas), kidney adenomas (males only), skin sarcomas, hepatocellular adenomas
or carcinomas, mammary gland (females only), and Zymbal's gland carcinomas (females only). These
tumors generally appeared earlier with increasing exposure levels and showed statistically significantly
increasing trends with increasing exposure level (by life table test or logistic regression, p < 0.001, as
conducted and reported by NTP). Etiologically similar tumor types, benign and malignant tumors of
the same cell type, were combined for these tabulations because of the possibility that the benign
tumors could progress to the malignant form (U.S. EPA, 2005, 086237). The tumors observed in the
Harderian and Zymbal's glands, however, were confirmed histopathologically only if observed grossly
at necropsy; the corresponding tissues for most mice were not examined histopathologically. Use of
the incidence data from these two sites as reported in Table 5-4 for dose-response analysis may
underestimate the true incidence because other instances were possibly missed, but the sites were
carried through the dose-response analysis in order to consider their relative impact. Survival for all
chloroprene-exposed male and female mice in the two higher exposure groups was statistically
128
-------
significantly lower than for control mice. Individual animal data including the time of observation of
tumors are provided in Tables C-l and C-2.
Rats demonstrated statistically significant or biologically noteworthy increases in tumor
incidence at multiple sites as well: oral cavity (papillomas or carcinomas); thyroid gland (follicular cell
adenomas or carcinomas); renal tubule adenomas or carcinomas; alveolar/bronchiolar adenomas or
carcinomas (males only); and mammary gland fibroadenomas (females only). Overall, rats were not as
sensitive as the mice, and were not considered further for dose-response analysis.
Table 5-4. Tumor incidence in female and male B6C3Fi mice exposed to
chloroprene via inhalation for 2 years
Tissue
Administered Chloroprene Concentration
(ppm)
Control
12.8 | 32
80
Females
All organs: hemangioma or
hemangiosarcoma
Lung: alveolar/bronchiolar adenoma or
carcinoma
Liver: hepato cellular adenoma or
carcinoma
Skin sarcoma
Mammary gland: carcinoma or
adenoacanthoma
Forestomach: squamous cell papilloma or
carcinoma
Harderian gland3: adenoma or carcinoma
Zymbal's gland3: carcinoma
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
4/50
541
4/50
706
20/50
493
0/50
3/50
527
1/50
734
2/50
527
0/50
6/49
482
28/49
447
26/49
440
11/49
285
5/50
440
0/50
5/50
621
0/50
18/50
216
34/50
346
20/50
503
11/50
524
10/50
394
0/50
3/50
524
0/50
8/50
523
42/50
324
30/50
384
18/50
462
14/50
336
4/50
576
9/50
467
3/50
565
Males
All organs: hemangioma or
hemangiosarcoma
Lung: alveolar/bronchiolar adenoma or
carcinoma
Forestomach: squamous cell papilloma
Harderian gland3: adenoma or carcinoma
Kidney: renal tubule adenomas (extended
and standard evaluations combined)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
Unadjusted rate
First incidence (days)
3/50
733
13/50
635
1/50
733
2/50
596
0/50
14/50
659
28/50
530
0/48
5/50
701
2/49
111
23/50
495
36/50
382
2/49
733
10/50
596
3/50
715
21/50
454
43/50
523
4/50
587
12/50
589
9/50
567
3Harderian gland and Zymbal's gland were examined histopathologically only if a lesion was observed grossly at
necropsy.
Source: NTP (1998. 042076).
129
-------
Table 5-5. Tumor incidence in female and male F344 rats exposed to chloroprene
via inhalation for 2 years
Tissue
Administered Chloroprene Concentration
(ppm)
Control
12.8
32
80
Females
Oral cavity: papillomas or carcinomas
Thyroid gland: follicular cell adenomas
or carcinomas
Mammary gland: fibroadenomas
Kidney: renal tubule adenomas or
carcinomas (extended and standard
evaluations combined)
Unadjusted
First incidence (days)
Unadjusted
First incidence (days)
Unadjusted
First incidence (days)
Unadjusted
First incidence (days)
1/49
687
1/49
733
24/49
366
0/49
3/50
681
1/50
721
32/50
302
0/50
5/50
588
1/50
733
36/50
470
0/50
11/50
660
5/50
617
36/50
433
4/50
609
Males
Oral cavity: papillomas or carcinomas
Thyroid gland: follicular cell adenomas
or carcinomas
Lung: alveolar/bronchiolar adenoma or
carcinoma
Kidney: renal tubule adenomas or
carcinomas (extended and standard
evaluations combined)
Unadjusted
First incidence (days)
Unadjusted
First incidence (days)
Unadjusted
First incidence (days)
Unadjusted
First incidence (days)
0/50
0/50
2/50
616
1/50
733
2/50
701
2/50
597
2/50
702
8/50
600
5/50
609
4/49
569
4/49
505
6/50
679
12/50
539
5/50
307
6/50
540
8/50
625
aKaplan-Meier analysis estimated neoplasm incidence rate at the end of the study, involving adjustment for
intercurrent mortality and under the assumption that the observed tumors were fatal.
Source: NTP (1998. 042076).
5.4.3. Dose Adjustments and Extrapolation Methods
The current EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005, 086237)
emphasize that the method used to characterize and quantify cancer risk from a chemical is determined
by what is known about the MO A of the carcinogen and the shape of the cancer dose-response curve.
The dose response is assumed to be linear in the low dose range when evidence supports a mutagenic
MO A because of DN A reactivity, or if another MOAthat is anticipated to be linear is applicable. A
mutagenic mode of carcinogenic action for chloroprene is supported by epoxide metabolite formation,
DNA-adduct formation, observation of in vivo and in vitro mutagenicity, and the well known structure-
activity relationship of similar epoxide-forming carcinogens. The determination of a mutagenic mode
of action is also supported by evidence of base pair substitution mutations seen in H- and K-ras proto-
oncogenes in chloroprene-induced lung, forestomach, and Harderian gland neoplasms observed in the
NTP (1998, 042076) study.
For these reasons, a linear low-dose extrapolation approach was used to estimate human
carcinogenic risk associated with chloroprene exposure.
130
-------
Due to the occurrence of multiple tumor types, earlier occurrence with increasing exposure, and
increased mortality with increasing exposure level, methods that can reflect the influence of competing
risks and intercurrent mortality on site-specific tumor incidence rates are preferred. EPA has generally
used the multistage Weibull model, because it incorporates the time at which death-with-tumor
occurred. The multistage Weibull model has the following form:
P(d) = 1 - exp[-(b0 + bid + b2d2 + ... + bkdk) x (t - t0)c]
where P(d) represents the lifetime risk (probability) of cancer at dose d (i.e., human equivalent
exposure in this case); parameters b; > 0, for i = 0, 1, ..., k; t is the time at which the animal's tumor
status, either no tumor, tumor, or unknown (e.g., missing or autolyzed) was observed; and c is a power
parameter estimated in fitting the model, which characterizes the change in response with age. The
parameter t0 represents the time between when a potentially fatal tumor becomes observable and when
it causes death and is generally set to 0 because of a lack of data to estimate the time reliably, such as
interim sacrifice data. Parameters were estimated using the method of maximum likelihood estimation
(MLE). Note that animals with unknown tumor status contribute to the model fit through the
likelihood function including the respective lengths of time on study without a tumor. The dose-
response analyses were conducted using the U.S. EPA Multistage Weibull (MSW) time-to-tumor
model (http://epa.gov/ncea/bmds/msw.html), which is based on Weibull models drawn from Krewski
etal. (1983,003194).
Other characteristics of the observed tumor types were considered prior to modeling, including
allowance for different, although possibly unidentified, MO As and for relative severity of tumor types.
First, etiologically different tumor types were not combined across sites prior to modeling in order to
allow for the possibility that different tumor types can have different dose-response relationships
because of varying time courses or other underlying mechanisms or factors. Consequently, all the
tumor types listed separately in Table 5-4 were modeled separately (Tables 5-6 and 5-7). A further
consideration allowed by the software program is the distinction between tumor types as being either
fatal or incidental in order to adjust for competing risks. Incidental tumors are those tumors thought
not to have caused the death of an animal, while fatal tumors are thought to have resulted in animal
death. Model results are reported in Tables 5-6 and 5-7 for "All Organ" effects two ways: (1) treatment
of early deaths (prior to final sacrifice) with hemangiosarcomas were treated as fatal tumors (with all
other hemangiomas and hemangiosarcomas as incidental to death); and (2) treatment of all
hemangiosarcomas (and hermangiomas) as incidental when they were observed at terminal sacrifice.
Furthermore, the fatal tumors were deemed rapidly fatal, and to was set equal to 0; the data were
considered insufficient to reliably estimate to in any event, without any interim sacrifice data. Tumors
at all other sites were treated as incidental.
Specific multistage Weibull models were selected for the individual tumor types for each sex,
based on the values of the log-likelihoods according to the strategy used by EPA (U.S. EPA, 2002,
131
-------
052153). If twice the difference in log-likelihoods was less than a x2 with degrees of freedom equal to
the difference in the number of stages included in the models being compared, the models were
considered comparable, and the most parsimonious model (i.e., the lowest-stage model) was selected
contingent on visual fits of the data. In all cases, this was equivalent to selecting the model with the
lowest AIC.
PODs for estimating low-dose risk were identified at doses consistent with the lower end of the
observed data, generally corresponding to 10% extra risk, defined as the extra risk over the background
tumor rate, [P(d) - P(0)]/[l - P(0)]. In some cases the highest observed response was not as high as
10% extra risk. In accordance with the cancer guidelines (U.S. EPA, 2005, 086237), PODs near the
lower end of these data ranges were selected. Next, all PODs were converted to equivalent continuous
exposure levels by multiplying by [(6 hours)/(24 hours)] x [(5 days)/(7 days)], or 0.178, under the
assumption of equal cumulative exposures leading to equivalent outcomes (C x T = k).
Additionally, in accordance with the U.S. EPA (1994, 006488) RfC methodology, the HEC
values for the various tumors were calculated by the application of DAFs. As discussed in Section
5.2.3, due to chloroprene's low water solubility, low reactivity and lesion distribution it is most
appropriately treated as a Category 3 gas for which blood-borne delivery plays a critical role. As was
done for noncancer lesions (Section 5.2.3), all tumors were treated as systemic effects and, since the
blood:air partition coefficient for chloroprene is greater in rats than in humans, a DAF of 1.0 was
applied.
The lifetime continuous inhalation unit risk for humans is defined as the slope of the line from
the POD, the lower 95% bound on the exposure associated with a level of extra risk near the low end
of the data range. Unit risks for each tumor site were calculated by dividing the BMR level (usually
10%) by its corresponding lower bound on the benchmark concentration (BMDLio).
5.4.4. Oral Slope Factor and Inhalation Unit Risk
In the absence of any data on the carcinogenicity of chloroprene via the oral route, or a suitable
PBPK model allowing route-to-route extrapolation, no oral slope factor was derived. An inhalation
unit risk was derived based on the multisite carcinogenic effects of chloroprene observed in mice
exposed via the inhalation route.
First, the results of applying the multistage Weibull models to each elevated female and male
mouse tumor site were evaluated (Tables 5-6 and 5-7, respectively). Human equivalent unit risks
estimated from the mouse tumor sites with statistically significant increases ranged from 3.4 x 10"6 to
1.8 x 10"4 per ug/m3, approximately a 50-fold range. The highest unit risk (1.8 x 10"4 per ug/m3)
corresponded to lung tumors in female mice, and the lowest unit risk (3.4 x 10"6 per ug/m3)
corresponded to forestomach tumors in female mice. The highest unit risk in male mice, 8.3 x 10"5 per
ug/m3), was also for lung tumors, and was approximately twofold lower than in female mice.
Regarding the model fits for hemangiomas or hemangiosarcomas, although there was a
statistically significant increasing trend for both female and male mice, a satisfactory model fit was not
132
-------
possible without dropping the highest exposure group in both cases, whether or not all tumors were
treated as incidental. The incidences in the highest exposure group (80 ppm) were lower than in the
32-ppm group, even after adjusting for intercurrent mortality. However, given the overall tumor
response in both the 32 and 80-ppm groups, fitting the decreased high dose circulatory system tumor
response does not appear relevant to estimating low dose risk. The result of treating
hemangiosarcomas occurring before final sacrifice as rapidly fatal (in combination with hemangiomas;
Section 5.4.3.) were nearly twofold higher than site-specific unit risks for both female (1.9-fold) and
male mice (1.6-fold). The unit risks for hemangiomas or hemangiosarcomas were approximately an
order of magnitude lower than that for lung tumors as systemic lesions in female mice, while for male
mice these unit risks were approximately threefold lower.
Concerning the unit risks for the two sites without complete histopathologic evaluation,
Harderian gland and Zymbal's gland (Section 5.4.2); the female mice Zymbal's gland unit risk was
quite low, at 3.5 x 10"6 per ug/m3, virtually identical to the forestomach unit risk in both female and
male mice. The Harderian gland unit risks were 1.2 x 10"5 and 1.5 x 10"5 per ug/m3, for females and
males, respectively, and were intermediate in the range of available unit risks, along with skin,
mammary gland, and hemangiomas/hemangiosarcomas (all assumed nonfatal) in female mice.
Given the multiplicity of tumor sites, basing the unit risk on one tumor site may underestimate
the carcinogenic potential of chloroprene. An approach suggested in the EPA cancer guidelines would
be to estimate cancer risk from tumor-bearing animals. EPA traditionally used this approach until the
document Science and Judgment in Risk Assessment (NRC, 1994, 006424) made a case that this
approach would tend to underestimate composite risk when tumor types occur in a statistically
independent manner. In addition, application of one model to a composite data set does not
accommodate biologically relevant information that may vary across sites or may only be available for
a subset of sites. For instance, the time courses of the multiple tumor types evaluated varied
substantially, which indicates an association of increasing incidence with time. Fitting a model like the
multistage-Weibull with mechanism-related parameters to composite data would not characterize the
evident range of variation. A simpler empirical model could be used for the composite data, such as
the multistage model, but available biological information (time of tumor observation) would then be
ignored.
133
-------
Table 5-6. Dose-response modeling summary for female mouse tumors associated
with inhalation exposure to chloroprene for 2 years
Tumor Type"
Lung: alveolar/
bronchiolar adenoma
or carcinoma
All organs:
hemangio-sarcomas,
hemangiomasf'h
All organs:
hemangio-sarcomas,
hemangiomas8'1
Mammary gland:
carcinoma or
adenoacanthoma
Forestomach:
squamous cell
papilloma or
carcinoma
Liver: hepatocellular
adenoma or
carcinoma
Harderian gland:
adenoma or
carcinoma
Skin: sarcoma
Zymbal's gland:
carcinoma
Power
Parameter
cb
3.8
5.9
1.0
1.0
4.1
4.2
2.9
1.6
1.1
BMR
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.05
Point Of Departure0
Modeled from
bioassay (ppm)
BMD
1.20
10.1
14.9
20.4
67.8
4.24
27.1
9.49
80.5
BMDL
0.88
5.75
11.1
14.1
46.3
2.45
12.6
7.18
22.5
Continuous, Human
equivalentd (jig/m3)
BMD
7.71 x 102
6.52 x 103
9.62 x 103
1.32 x 104
4.37 x 104
2.73 x 103
1.75 x 104
6.11 x 103
5.19 x 104
BMDL
5.69 x 102
3.71 x 103
7.13 x 103
9.06 x 103
2.98 x 104
1.58 x 103
8.13 x 103
4.63 x 103
1.45 x 104
Unit Risk6
/Oig/m3)
1.8 x 10'4
2.7 x ID'5
1.4 x 10'5
1.1 x ID'5
3.4 x ID'6
6.3 x 10'5
1.2 x ID'5
2.2 x lO'5
3.5 x 1(T6
Composite
Unit Riskf
/Oig/m3)
2.7 x ID'4
aTumor incidence data from NTP (1998, 042076).
bMultistage-Weibull model: P(d) = 1 - exp[-(b0 + b:d + b2d2 + ... + bkdk) x (t-t0)c], coefficients estimated in terms of ppm
as administered in bioassay; lower stage fy not listed were estimated to be zero. See Appendix C for modeling details.
°BMD = Concentration at specified extra risk (benchmark dose); BMDL = 95% lower bound on concentration at specified
extra risk.
Continuous equivalent estimated by multiplying exposures by (6 hours)/(24 hours) x (5 days)/(7 days).
eUnit risk estimated by dividing the BMR by the BMDL.
Composite unit risk estimate, across all sites listed; see text for method.
gHighest exposure group dropped in order to better characterize low-dose responses.
hTreatment of early deaths (prior to final sacrifice) with hemangiosarcomas as fatal, with all other hemangiomas and
hemangiosarcomas as incidental to death.
'All hemangiosarcomas (and hemangiomas) were considered incidental.
134
-------
Table 5-7. Dose-response modeling summary for male mouse tumor sites
associated with inhalation exposure to chloroprene for 2 years
Tumor Type"
Lung: alveolar/
bronchiolar
adenoma or
carcinoma
All organs:
hemangio-
sarcomas,
hemangiomas8'11
All organs:
hemangio-
sarcomas,
hemangiomas8' '
Harderian gland:
adenoma or
carcinoma
Kidney: renal
tubule adenomas
(extended and
standard
evaluations
combined)
Forestomach:
squamous cell
papilloma or
carcinoma
Power
Parameter
cb
3.4
13.2
3.9
5.6
6.1
1.3
BMR
0.1
0.1
0.1
0.1
0.1
0.05
Point Of Departure0
Modeled from
bioassay (ppm)
BMD
2.46
5.28
7.75
16.7
26.7
45.1
BMDL
1.86
3.34
5.34
10.5
16.5
22.8
Continuous, human
equivalentd (jig/m3)
BMD
1.59 x 103
3.40x 103
4.99 x 103
1.08 x 104
1.72 x 104
2.91 x 104
BMDL
1.20 x 103
2.15 x 103
3.44 x 103
6.74 x 103
1.06 x 104
1.47 x 104
Unit Risk6
/Oig/m3)-1
8.3 x ID'5
4.7 x lO'5
2.9 x ID'5
1.5 x 10'5
9.4 x lO'6
3.4 x ID'6
Composite
Unit
Riskf
/ftig/m3)
1.4 x 10'4
"Tumor incidence data from NTP (1998, 042076).
bMultistage-Weibull model: P(d) = 1 - exp[-(b0 + bid + b2d2 + ... + bkdk) x (t-to)°], coefficients estimated in terms
of ppm as administered in bioassay; lower stage fy not listed were estimated to be zero. See Appendix C for
modeling details.
°BMD = Concentration at specified extra risk (benchmark dose); BMDL = 95% lower bound on concentration at
specified extra risk.
dContinuous equivalent estimated by multiplying exposures by (6 hours)/(24 hours) x (5 days)/(7 days).
"Unit risk estimated by dividing the BMR by the BMDL.
Composite unit risk estimate, across all sites listed; see text for method.
8Highest exposure group dropped in order to better characterize low-dose responses.
hTreatment of early deaths (prior to final sacrifice) with hemangiosarcomas as fatal, with all other hemangiomas and
hemangiosarcomas as incidental to death.
1 All hemangiosarcomas (and hemangiomas) were considered incidental.
Consistent with the recommendations of the NRC (1994, 006424) and the current Guidelines
for Carcinogen Risk Assessment (U.S. EPA, 2005, 086237) for the assessment of total risk, an upper
bound on the composite risk for all tumor sites in female and male B6C3Fi mice was estimated. Note
that this upper bound estimate of composite risk describes the risk of developing any combination of
the tumor types considered, not just the risk of developing all simultaneously. Statistical methods
135
-------
which can accommodate the underlying distribution of slope factors are optimal, such as through
maximum likelihood estimation or through bootstrapping or Bayesian analysis. However, these
methods have not yet been extended to models such as the multistage-Weibull model. Summing of
individual upper bound would tend to overestimate the composite upper bound. This analysis involves
assuming asymptotic normality for slope factors. Derivation of the composite unit risk estimate
involved the following steps (detailed in Appendix C):
• It was assumed that the tumor types associated with chloroprene exposure were statistically
independent - that is, that the occurrence of a hemangiosarcoma, say, was not dependent on
whether there was a forestomach tumor. This assumption cannot currently be verified and if
not correct could lead to an overestimate of risk from summing across tumor sites. However,
NRC (1994, 006424) argued that a general assumption of statistical independence of tumor-
type occurrences within animals was not likely to introduce substantial error in assessing
carcinogenic potency from rodent bioassay data.
• The models previously fitted to estimate the BMD values and BMDL values were used to
extrapolate to a lower level of risk (R) where the BMD values and BMDL values were in a
linear range. For these data a 10"2 risk (R = 0.01) was generally the lowest risk necessary.
Although this step appears to differ from the explicit recommendation of the cancer guidelines
(U.S. EPA, 2005, 086237) to estimate cancer risk from a POD "near the lower end of the
observed range, without significant extrapolation to lower doses," this method is recommended
in the cancer guidelines as a method for combining multiple extrapolations. A sensitivity
analysis considering risks nearer the lower end of the observed ranges for each tumor type (not
included in this document) showed that the composite risk was essentially the same (to two
significant digits) whether or not the individual risks were estimated in the region of 10"2 risk or
near the PODs.
• The central tendency estimates of unit potency (that is, risk per unit of exposure) at each
BMDoi, estimated by 0.01/BMD0i, were summed across the sites listed in Table 5-6 for female
mice and similarly across the sites for male mice listed in Table 5-7 (Appendix C, Table C-5).
• The composite unit risk, which is a 95% upper confidence limit (UCL), was calculated by
assuming a normal distribution for the individual risk estimates and deriving the variance of the
risk estimate for each tumor site from its 95% UCL (0.01/BMDL0i) and MLE (0.01/BMD0i)
(Table C-5) according to the following formula:
95% UCL = MLE + 1.645 x SD or
0.01/BMDLoi = 0.01/BMD0i +1.645 xSD (1)
rearranged to:
SD = (0.01/BMDLoi - 0.01/BMD0i)/1.645 (2)
136
-------
where 1.645 is the t-statistic corresponding to a one-sided 95% confidence interval and >120 degrees
of freedom, and the standard deviation (SD) is the square root of the variance of the MLE. The
variances (variance = SD2) for each site-specific estimate were summed across tumor sites to obtain
the variance of the sum of the MLEs. The 95% UCL on the sum of the individual MLEs was
calculated from expression (1) using the variance of the MLE to obtain the relevant SD (SD =
variance172).
As shown in Table 5-6, the resulting composite unit risk for all tumor types for female mice
was 2.7 x 10"4 per ug/m3. Overall, the consideration of the other tumor sites increased the unit risk by
1.5-fold from the highest unit risk for any individual tumor type, 1.8 x 10"4 per ug/m3 for female lung
tumors treated as a systemic lesion. The increase was due largely to the hemangiosarcomas and liver
tumors, with little contribution from the other tumor sites. A sensitivity analysis (not included in this
document) showed that the composite risk was essentially the same (to 2 significant digits) whether or
not the individual risks were estimated in the region of 10"2 risk or near the PODs.
Table 5-7 presents the calculations for the composite unit risk for all tumor types for male mice
of 1.4 x 10"4 per ug/m3 (with lung tumors treated as a systemic lesion), a 1.7-fold increase compared to
the highest unit risk for any individual tumor type, 8.3 x 10"5 per ug/m3 for lung tumors treated as a
systemic lesion. The increase was due almost entirely to the risk associated with the
hemangiosarcomas. As with the composite risk for female mice, there was a trivial difference whether
or not the individual risks were estimated in the region of 10"2 risk or near the PODs.
Based on the analyses discussed above, the recommended upper bound estimate on human
extra cancer risk from continuous lifetime exposure to chloroprene is 3 x 10"4 per ug/m3, rounding the
composite risk for female mice above to one significant digit. This unit risk should not be used with
continuous lifetime exposures greater than 600 ug/m3 (0.6 mg/m3), the human equivalent POD for the
female lung tumors, because the observed dose-response relationships do not continue linearly above
this level and the fitted dose-response models better characterize what is known about the
carcinogenicity of chloroprene. The recommended unit risk estimate reflects the time-to-tumor
dimension of the responses as well as the exposure-response relationships for the multiple tumor sites
in both sexes of mice.
5.4.5. Application of Age-Dependent Adjustment Factors
Because a mutagenic mode of action for chloroprene carcinogenicity is sufficiently supported
by in vivo and in vitro data and relevant to humans (Section 4.7.3.2), and in the absence of chemical-
specific data to evaluate the differences in susceptibility, increased early-life susceptibility is assumed
and the age-dependent adjustment factors (ADAFs) should be applied, as appropriate, along with
specific exposure data in accordance with EPA's Supplemental Guidance for Assessing Susceptibility
From Early-Life Exposure to Carcinogens (U.S. EPA, 2005, 088823). The inhalation unit risk of 3 x
10"4 per ug/m3, calculated from data for adult exposures, does not reflect presumed early-life
137
-------
susceptibility for this chemical. Example evaluations of cancer risks based on age at exposure are
given in Section 6 of the Supplemental Guidance.
The Supplemental Guidance establishes ADAFs for three specific age groups. The current
default ADAFs and their age groupings are 10 for <2 years, 3 for 2 to <16 years, and 1 for 16 years and
above (U.S. EPA, 2005, 088823). The 10-fold and threefold adjustments in slope factor are to be
combined with age specific exposure estimates when estimating cancer risks from early life (<16 years
of age) exposure to chloroprene.
To illustrate the use of the ADAFs established in the Supplemental Guidance (U.S. EPA, 2005,
088823), sample calculations are presented for a lifetime risk estimate for continuous exposure from
birth with a life expectancy of 70 years. The ADAFs are first applied to obtain risk estimates for
continuous exposure over the three age groups:
Risk for birth through <2 yr = 3 x 10"4 per ug/m3 x 10 x 2 yr/70 yr = 8.6 x 10"5 per ug/m3
Risk for ages 2 through <16 = 3 x 10"4 per ug/m3 x 3 x 14 yr/70 yr = 1.8 x 10"4per ug/m3
Risk for ages 16 until 70 = 3 x 10"4 per ug/m3 x 1 x 54 yr/70 yr = 2.3 x 10"4per ug/m3
To calculate the lifetime risk estimate for continuous exposure from birth for a population with
default life expectancy of 70 years, the risk associated with each of the three relevant time periods is
summed:
Risk =8.6 x 10'5+1.8 x 10'4+2.3 x 10'4=5.0x 10'4 per ug/m3
Using the above full lifetime unit risk estimate of 5 x 10"4 per ug/m3 for continuous exposure from
birth to 70 years, the lifetime chronic exposure level of chloroprene corresponding to an extra risk of
1 x 10"6 can be estimated as follows:
1 x lO'6 H- 5 x 10'4 per ug/m3 = 0.002 ug/m3
5.4.6. Previous Cancer Assessment
The carcinogenicity of chloroprene has not been evaluated previously for the IRIS program.
5.4.7. Uncertainties in Cancer Risk Values
A number of uncertainties underlie the cancer unit risk for chloroprene. These are discussed in
the following paragraphs. Specifically addressed is the impact on the assessment of issues such as the
use of models and extrapolation approaches, the use of other bioassay data, and the choices made and
the data gaps identified. In addition, the use of assumptions, particularly those underlying the
Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005, 086237) is explained and the decision
138
-------
concerning the preferred approach is given and justified. Principal uncertainties are discussed below
and summarized in Table 5-8.
Table 5-8. Summary of uncertainties in chloroprene cancer unit risk estimate
Consideration
Potential
Impact8
Decision
Justification
Human
population
variability in
metabolism and
response/
sensitive
subpopulations
Low-dose risk
could t or I to an
unknown extent
Considered qualitatively
No data to support range of human
variability/sensitivity. Mutagenic MO A
indicates potentially increased early-life
susceptibility.
Low-dose
extrapolation
procedure
Unknown; not clear
what departure
from Cancer
Guidelines would
be plausible
Multistage-Weibull
model to determine
POD, linear low-dose
extrapolation from POD
Multistage-Weibull model addresses
competing risks from other tumors and
intercurrent mortality. Mutagenic MO A
supports linear low-dose extrapolation.
Dose metric
Alternatives could
t or I low-dose risk
per unit
concentration by an
unknown extent
Used administered
concentration
Experimental evidence supports a role for
metabolism in toxicity, but actual
responsible metabolites are neither clearly
identified nor quantifiable. Use of
administered concentration provides an
unbiased estimate if proportional to the
actual carcinogen(s).
Bioassay
Unknown; others
unsuitable or
unavailable
NTP (1998, 0420761
Standard design, well conducted,
extensively peer reviewed; carcinogenic
response consistently observed across all
four combinations of species/sex.
Species/sex
combination
Human risk could J,
or t, depending on
relative sensitivity
Multiple sites in female
mice
Unit risk is based on the most sensitive
endpoint (risk of any tumor type) in the
most sensitive species and sex (female
mouse), based on PODnEc- It was
assumed that humans are as sensitive as
the most sensitive rodent sex/species
tested; true correspondence is unknown.
Site concordance for liver tumors for
humans and female mice was observed,
but human data not sufficient to rule out
other types seen in mice or rats.
139
-------
Consideration
Cross-species
extrapolation
Statistical
uncertainty at
POD
Potential
Impact"
Alternatives for
lung tumors differ
by fourfold: human
risk for any site
could I or f. Low-
dose risk would J,
approximately 40%
if lung tumors were
treated as portal-of-
entry effects
I risk per unit
concentration
1.2-fold if BMD10
used rather than
BMDL10
Decision
RfC methodology:
Equal risk per unit of air
concentration for all
sites; for lung also
considered relative
surface areas of affected
region. Treat lung
tumors as systemic
effects.
BMCL (default
approach for calculating
plausible upper bound)
Justification
There are no data to support other
alternatives. There is evidence that
chloroprene is distributed systematically
(observation of tumors at multiple sites),
and correspondingly the possibility that
chloroprene is redistributed to the lungs.
The contribution of one route of delivery
(i.e., inhalation vs. bloodstream) to the
induction of lung tumors is currently
unknown, therefore the derivation
approach that returns the highest unit risk
was used
Limited size of bioassay results in
sampling variability; lower bound is 95%
confidence interval on concentration.
a| = increase, J, = decrease.
Human Population Variability. The extent of inter-individual variability in chloroprene
metabolism has not been characterized. A separate issue is that the human variability in response to
chloroprene is also poorly understood. The effect of metabolic variation, including potential
implications for differential toxicity, has not been well studied. Although a mutagenic MOA indicates
increased early-life susceptibility, there are no data exploring whether there is differential sensitivity to
chloroprene carcinogenicity across human life stages. This lack of understanding about potential
differences in metabolism and susceptibility across exposed human populations thus represents a
source of uncertainty.
Choice of Low-Dose Extrapolation Approach. The MOA is a key consideration in clarifying
how risks should be estimated for low-dose exposure. A multistage Weibull time-to-tumor model was
the preferred model because it can account for differences in mortality and other competing risks
between the exposure groups in the mouse bioassay; however, it is unknown how well this model
predicts low-dose extrapolated risks for chloroprene. Cause of death information was not available for
this model; if available, risk estimates would tend to be slightly higher. For example, treatment of
early deaths (prior to final sacrifice) with hemangiosarcomas as fatal, with all other hemangiomas and
hemangiosarcomas as incidental to death, led to unit risks up to twofold higher than unit risks treating
all hemangiosarcomas (and hemangiomas) as incidental.
Dose Metric. Chloroprene is metabolized to intermediates with carcinogenic potential, most
likely an epoxide. However, data sufficient to estimate quantities were not available. Under the
assumption that the carcinogenic form(s) of chloroprene (or metabolites) are produced in proportion to
low-exposures of chloroprene, the derived unit risk is an unbiased estimate.
140
-------
Choice of Bioassay/Species/Sex. The NTP inhalation bioassay followed an accepted protocol,
was well conducted, and extensively peer reviewed. The carcinogenic response occurs in both species
and sexes of rodents (as well as in humans as observed in occupational epidemiologic cohorts). The
calculated composite unit risk is based on the most sensitive endpoint (risk of any tumor type) in the
most sensitive species and sex (female mouse). There is no information on chloroprene to indicate that
the observed rodent tumors are not relevant to humans. Further, no data exist to guide quantitative
adjustment for differences in sensitivity among rodents and humans. While site concordance generally
is not assumed across species, e.g., due to potential differences in pharmacokinetics, DNA repair, other
protective systems across species and tissues (U.S. EPA, 2005, 086237), it is notable that human-
mouse site concordance was observed for liver tumors. In addition, rat and mouse tumor types
overlapped but included different tumor types observed for each species/sex combination. Human data
were insufficient to rule out the occurrence of these additional tumor types in humans.
Cross-Species Scaling. Another source of uncertainty comes from the interspecies
extrapolation of risk from mouse to human. The two rodent species for which bioassay data were
available— mouse and rat—vary in their carcinogenic responses to chloroprene, in terms of both site
specificity and magnitude of response (Section 4). Ideally, a PBPK model for the internal dose(s) of
the reactive metabolite(s) would decrease some of the quantitative uncertainty in interspecies
extrapolation; however, current PBPK models are inadequate for this purpose (Section 3). Existing
pharmacokinetic models cannot yet adequately explain the species differences in carcinogenic
response, and it is possible that there are pharmacodynamic as well as pharmacokinetic differences
between the mouse and rat with respect to their sensitivities to chloroprene.
While concordance of specific sites between rodents and humans (e.g., liver tumors) tends to
support the relevance of rodent species to humans, lack of specific site concordance (other tumors)
does not diminish concern for human carcinogenic potential. The mouse was the more sensitive
species to the carcinogenic effects of chloroprene exposure. Although the derivation took into account
some known differences between mice and humans in tissue dosimetry (U.S. EPA, 1994, 006488)
differences in anatomy of the upper respiratory tract and resulting differences in absorption or in local
respiratory system effects are sources of uncertainty.
Statistical Uncertainty at the POD. Parameter uncertainty within the chosen model reflects
the limited sample size of the cancer bioassay. For the multistage-Weibull model applied to this data
set, there is a reasonably small degree of uncertainty at the 10% extra risk level (the POD for linear
low-dose extrapolation). Central estimates of risk differed from their upper bounds by about 1.3-fold
for lung tumors and for the composite unit risk estimates (Table C-5).
HEC derivation. A source of uncertainly in the derivation of the HEC comes from whether or
not chloroprene induces lung tumors due to portal-of-entry or systemic effects. Systemic distribution
of chloroprene is consistent with its physiochemical properties and is evidenced by the induction of
tumors in multiple organs and suggests that chloroprene may be redistributed back to the lungs and
may primarily act as a systemically delivered carcinogen. The External Peer Reviewers supported the
141
-------
classification of chloroprene as a Category 3 gas with systemic distribution. The rationale for the
choice of this derivation approach is further explained in Sections 5.2.3 and 5.3. However, the
contribution of either route of delivery (i.e., inhalation versus bloodstream) to the induction of lung
tumors is currently unknown. Treating lung tumors as systemic effects returns the highest composite
unit risk (approximately 60% greater than if lung tumors are treated as portal-of-entry effects).
142
-------
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE
6.1. HUMAN HAZARD POTENTIAL
Chloroprene (C^sCl, 2-chloro-l,3-butadiene, CASRN 126-99-8) is a volatile and flammable
liquid monomer that can be produced by dimerization of acetylene and addition of hydrogen chloride
or by chlorination of 1,3-butadiene. Chloroprene is polymerized to form elastomers for use in the
manufacture of belts, hoses, gloves, wire coatings, tubing, solvents, and adhesives. Chloroprene is also
a structural analogue of isoprene (2-methyl-1,3-butadiene) and resembles vinyl chloride as far as
having a chlorine bound to a double-bonded carbon (alkene) backbone.
Toxicokinetic information on the absorption, distribution, and in vivo metabolism and excretion
of chloroprene and/or its metabolites is nonexistent for humans and limited for animals. Several in
vitro studies have focused on chloroprene metabolism in lung and liver tissue fractions from rat,
mouse, hamster, and humans (Cottrell et al., 2001, 157445: Himmelstein et al., 2001, 019013:
Himmelstein et al., 2001, 019012: Himmelstein et al., 2004, 625152: Hurst and Ali, 2007, 625159:
Munter et al., 2003, 625214: Munter et al., 2007, 576501: Munter et al., 2007, 625213: Summer and
Greim, 1980, 064961). These studies suggest that chloroprene is metabolized via the CYP450 enzyme
system to monoepoxides [(l-chloroethenyl)oxirane and 2-chloro-2-ethynyloxirane], further
metabolized to aldehydes and ketone intermediates and subsequent mercapturic acid derivatives, and
cleared via further oxidation, hydrolysis and/or glutathione conjugation reactions. Similar to
1,3-butadiene, an epoxide metabolite, (l-chloroethenyl)oxirane is considered to be the toxic moiety.
The metabolic profile for chloroprene is qualitatively similar across species. However, in vitro kinetic
studies using tissues from rodents and humans suggest quantitative species and tissue-specific
differences that, if operative in vivo, could contribute to the species, strain, and sex differences
observed in chloroprene-induced effects.
Limited information exists on the noncancer effects of chloroprene due to oral ingestion. In
rats, oral exposures from weaning until death (at 120 weeks) resulted in indices of liver toxicity (liver
necroses and degenerative lesions of the parenchymal cells). No information is available on the oral
toxicity of chloroprene in humans.
Limited information exists on the noncancer effects of chloroprene via the inhalation route in
humans. Chloroprene was reported to cause respiratory, ocular, and dermal irritation, chest pains,
temporary hair loss, dizziness, insomnia, headache, and fatigue. Chest pains accompanied by
tachycardia and dyspnea were also reported. In a Russian review of the effects of chloroprene,
Sanotskii (1976, 063885) reported that medical examinations of chloroprene production workers
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
143
-------
revealed changes in the nervous system (lengthening of sensorimotor response to visual cues and
increased olfactory thresholds), cardiovascular system (muffled heart sounds, reduced arterial pressure,
and tachycardia), and hematology (reduction in red blood cell (RBC) counts, decreased hemoglobin
levels, erythrocytopenia, leucopenia, and thrombocytopenia). The ambient concentration of
chloroprene associated with these effects ranged from 1-7 mg/m3.
The toxic and carcinogenic potential of chloroprene by the inhalation route has been assessed in
several laboratory animal studies, including a rat and mouse subchronic (16 days and 13 weeks) and
2-year chronic inhalation bioassays conducted by NTP (1998, 042076), a subchronic range-finding and
a chronic study in rats and hamsters conducted by Trochimowicz et al. (1998, 625008), an
embryotoxicity and a teratology study by Culik et al. (1978, 094969), and a series of Russian
reproductive and developmental toxicity studies reviewed by Sanotskii (1976, 063885). These studies
associated chloroprene inhalation exposure with respiratory, kidney, liver, splenic, and forestomach
effects. The pulmonary (alveolar and bronchiolar hyperplasia), nasal (olfactory epithelium), and
splenic (hematopoietic cell proliferation) lesions were the most sensitive endpoints in chronically
exposed test animals, having been observed at all the doses tested (12.8-80 ppm) in the NTP (1998,
042076) study of rats and mice. In the chronic study by Trochimowicz et al. (1998, 625008), lesions in
lungs (inflammation, lymphoid aggregates around the bronchi, bronchiole, and blood vessels) and
livers (small foci of cellular alteration) of rats were observed at 50 ppm. Embryotoxicity and fetal
resorptions were reported in the inhalation developmental toxicity study (Culik et al., 1978, 094969).
However, interpretational difficulties obscure whether this effect is an actual outcome or rather a
statistical artifact of an abnormally low background rate in control animals.
The carcinogenic potential of chloroprene in humans has been assessed in a number of
occupational epidemiologic studies among workers exposed to chloroprene monomer and/or
poly chloroprene latex conducted in eight cohorts from the U.S., Russia, Armenia, France, China, and
Ireland. Four cohorts with sufficient numbers of liver/biliary passage cancer cases showed evidence of
association with occupational chloroprene exposure, and reported significantly elevated SMRs when
compared to external populations (Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419:
Leet and Selevan, 1982, 094970: Li et al., 1989, 625181). These measures of association were
observed, even in the presence of the healthy worker effect bias. Several studies were able to use more
advanced exposure assessments and internal reference populations, which should reduce this bias.
These studies showed relatively consistent elevated relative risk estimates among intermediate and
highly exposed workers, despite limited sample size and statistical power (Bulbulyan et al., 1998,
625105: Bulbulyan et al., 1999, 157419: Marsh et al., 2007, 625187). Known risk factors for liver
cancer (e.g., alcohol consumption, hepatitis B infection, etc.) were not controlled for in the studies
observing associations between occupational chloroprene exposure and liver/biliary cancers. Several
studies also reported higher SMRs for lung cancer among workers exposed to chloroprene, although
few of the associations were significant and none of the studies controlled for confounding by smoking
status, a strong indicator of lung cancer.
144
-------
Chloroprene has been shown to induce multisite, malignant tumors in rats and mice in the
2-year NTP (1998, 042076) bioassay. Dose-related increasing trends in tumors were noted in rats at
the following sites: oral cavity, thyroid gland, lung, kidney, mammary gland. Dose related increasing
trends in tumors were noted in mice at the following sites: lung, all organs (hemangiomas and
hemangiosarcomas), Harderian gland, forestomach, kidney, skin, liver, mammary gland, Zymbal's
gland. All of these tumor sites showed statistically significant positive trends with increasing exposure
level (Cochran-Armitage test for trend p < 0.05, most with p < 0.001; data not shown). In addition,
many early deaths and moribund sacrifices were associated with chloroprene-induced neoplasms.
The genetic toxicity database includes numerous studies covering a range of standard
genotoxicity test batteries; however, the results have been conflicting, making it difficult to ascertain
the mutagenic potential of chloroprene. In general, bacterial base pair substitution (S. typhimurium
strains TA100 and TA 1535) mutation assays have been positive (Bartsch et al., 1979, 010689;
Willems, 1980, 625049). while the bacterial frame shift (S. typhimurium strains TA97 and TA98)
mutation assays have been nonpositive (NTP, 1998, 042076: Willems, 1980, 625049). In contrast,
other studies (NTP, 1998, 042076) have reported nonpositive results for all bacterial strains. A positive
result with all bacterial strains was observed with the epoxide metabolite of chloroprene,
(l-chloroethenyl)oxirane (Himmelstein et al., 2001, 019013). Chloroprene has been primarily
nonpositive in in vitro micronucleus assays (Drevon and Kuroki, 1979, 010680; Himmelstein et al.,
2001, 019013), in vivo chromosomal damage assays (1998, 042076), and bone marrow micronucleus
assays (NTP, 1998, 042076: Shelby and Witt, 1995, 624921). Conflicting results positive in Vogel
(1979, 000948); nonpositive in Foureman et al. (1994, 065173)—have been reported for the in vivo
Drosophila sex-linked lethal mutation assay. Further in vivo evidence for the mutagenicity of
chloroprene is the observation that tissues from lung, forestomach, and Harderian gland tumors from
mice exposed to chloroprene in the NTP chronic bioassay (1998, 042076) were shown to have a higher
frequency of mutations in K- and H-ras proto-oncogenes than in spontaneous occurring tumors (Sills
et al., 1999, 624952: Sills et al., 2001, 624922).
There was also a high correlation between K-ras mutations and loss of heterozygosity in the
same chromosome in chloroprene-induced lung neoplasms in mice (Ton et al., 2007, 625004).
Possible explanations for the conflicting mutagenic responses of chloroprene in standard genotoxicity
assays include methods of exposure that do not control for the high volatility of chloroprene (i.e.,
chloroprene is not present in the test system), the presence of more stable (perhaps more toxic)
chloroprene dimers, the use of microsomal inducers that did not elicit a broad range of metabolic
enzymes (specifically, in bacterial assays), and the reactivity (perhaps deactivation) of chloroprene
with treatment vehicle (e.g., DMSO versus ethanol).
The likely MOAfor chloroprene is via mutagenicity involving epoxide metabolites formed at
the target sites. The MOA determination is supported by epoxide metabolite formation, DNA-adduct
formation, observation of in vivo and in vitro mutagenicity, and the well known structure-activity
relationship of similar epoxide-forming carcinogens. Chloroprene has been found to be metabolized to
145
-------
epoxides by humans and rodents. The hypothesized mutagenic mode of action is supported by
evidence of base pair substitution mutations seen in H- and K-ras proto-oncogenes in chloroprene-
induced lung, forestomach, and Harderian gland neoplasms observed in the NTP (1998, 042076) study.
In addition, chloroprene is the 2-chloro analog of 1,3-butadiene. Inhalation studies have
demonstrated that, similar to 1,3-butadiene and isoprene, chloroprene is a multisite carcinogen in rats
and mice. Butadiene and isoprene are metabolized to epoxides and diepoxides which are believed to
be responsible for their carcinogenicity. Chloroprene is also metabolized to epoxide intermediates that,
similarly to butadiene, may mediate its carcinogenic effects. The similarities in the sites of tumor
induction in rodents (mammary gland and thyroid gland in rats, lung, Harderian gland, forestomach,
kidney, and liver in mice) between butadiene and chloroprene provide further evidence for a similar
MO A for these epoxide-forming compounds. In addition, the mouse lung was the most sensitive site
of carcinogenicity for both chloroprene and butadiene. Similar to butadiene, DNA reactivity and
adduct formation have been described for chloroprene. Areas of uncertainty exist in the data
supporting a mutagenic MOA for chloroprene carcinogenicity, more specifically in the genotoxicity
database. There is conflicting evidence in the bacterial genotoxicity assays and generally nonpositive
findings in mammalian in vivo tests, but these results are weighed against the base pair substitution
mutations seen in H- and K-ras proto-oncogenes in chloroprene-induced lung, forestomach, and
Harderian gland neoplasms observed in the NTP (1998, 042076) study.
6.2. DOSE RESPONSE
The chronic inhalation study conducted by NTP (1998, 042076) was considered as the principal
study for both the noncancer and cancer effects of chloroprene exposure.
6.2.1. Noncancer/Oral
The available data are inadequate to derive an oral RfD for chloroprene. There are no human
data involving oral exposure. The only lifetime oral study exposed rats to chloroprene at one dose
(50 mg/kg/day) and only qualitatively reported noncancer effects (Ponomarkov and Tomatis, 1980,
075453).
In summary, this study identifies the liver (multiple liver necroses and degenerative lesions of
parenchymal cells), lung (severe congestion), and kidney (severe congestion) as potential target organs
for the oral toxicity of chloroprene; although, the available information is insufficient to characterize
toxicity outcomes or dose-response relationships. A route-to-route extrapolation from available
chronic inhalation data to oral data for the purposes of deriving an RfD was not performed due to the
inadequacies of the current chloroprene PBPK model (Section 3.5).
6.2.2. Noncancer/Inhalation
The chronic inhalation study conducted by NTP (1998, 042076) was selected as the principal
study for the noncancer effects of chloroprene exposure. A range of effects from the NTP study (1998,
042076), including alveolar epithelial hyperplasia, bronchiolar hyperplasia, pulmonary histiocytic cell
146
-------
infiltration, olfactory epithelial atrophy, olfactory epithelial necrosis, chronic inflammation, kidney
hyperplasia, forestomach hyperplasia, and splenic hematopoietic cell proliferation, were considered as
candidates for the selection of the critical effect for derivation of the RfC. BMD modeling was used to
determine potential PODs for deriving the chronic RfC by estimating the effective dose (benchmark
concentration [BMD]) and it's BMDL at a specified level of response (i.e., BMR) for each selected
chloroprene-induced effect (Table 5-2). Olfactory atrophy in male rats, alveolar hyperplasia in female
rats, and splenic hematopoietic cell proliferation in female mice were selected as co-critical effects.
For these endpoints, after rounding to one significant figure, the PODAoj resulted in a value of
2 mg/m3. This PODADJ was converted into the PODnEc by application of the dosimetric adjustment
factor (DAF) for systemic effects. Application of a 100-fold UF (3 for uncertainty associated with
animal to human differences, 10 for consideration of human variability, and 3 for database
deficiencies) resulted in a chronic RfC of 2 x 10"2 mg/m3.
Confidence in the principal study (NTP, 1998, 042076) is judged to be high as it was a well-
designed study using two test species (rats and mice) with 50 animals per dose group. This study
appropriately characterizes a range of chloroprene-induced nonneoplastic lesions. In addition, the key
histopathological lesions observed are appropriately described, and suitable statistical analysis was
applied to all animal data.
Confidence in the overall database specific to chloroprene is medium to high. The major
strength of the database is the observation of dose-response effects in multiple organ systems in a well-
designed chronic inhalation study that utilized 50 animals per sex per dose group, a range of doses
based on the results of preliminary, shorter-duration studies (16 days and 13 weeks), and thoroughly
examined the observed toxicity of chloroprene in two species (rat and mouse). The database further
contains another chronic inhalation bioassay investigating outcomes in another species (hamster), and
well-designed embryotoxicity, teratological, and reproductive toxicity studies. The database also
contains subchronic studies and chronic studies observing potential neurotoxic and immunotoxic
effects. A major limitation in the database is the lack of a complete two-generation reproductive
toxicity study. Therefore, confidence in the RfC is judged to be medium to high.
6.2.3. Cancer/Oral
In the absence of any data on the carcinogenicity of chloroprene via the oral route, or a suitable
PBPK model allowing route-to-route extrapolation, no oral slope factor was derived.
6.2.4. Cancer/Inhalation
The chronic inhalation study conducted by NTP (1998, 042076) was selected as the principal
study for the cancer effects of chloroprene exposure. Statistically significant increases in tumor
incidence were observed at multiple sites in the mouse (the most sensitive species) in the NTP study:
all organs (hemangiomas and hemangiosarcomas), lung (bronchiolar/alveolar adenomas and
carcinomas), forestomach, Harderian gland (adenomas and carcinomas), kidney (adenomas), skin,
147
-------
liver, and mammary glands. These tumors generally appeared earlier with increasing exposure level
and showed statistically significantly increasing trends with increasing exposure level (by life table test
or logistic regression, p < 0.001). Dose-response modeling was used to determine potential PODs for
deriving the inhalation unit risk by estimating the effective dose at a specified level of response
(benchmark concentration [BMDio]) and its lower-bound BMDLio for each selected chloroprene-
induced tumor (Tables 5-6 and 5-7). Lung tumors, treated as a systemic lesion (Section 5.4.3 and
5.4.7), in female mice resulted in the highest inhalation unit risk (1.8 x 10"4per ug/m3) when modeled
as an individual lesion. When etiologically different tumors were considered together (given the
multiplicity of the tumor sites, basing unit risk on only one tumor site may underestimate the
carcinogenic potential of chloroprene), the resulting composite inhalation unit risk for female mice was
2.7 x 10"4per ug/m3. Based on these modeling results, the upper bound estimate on human extra
lifetime cancer risk from continuous lifetime (adult) exposure to chloroprene is 3 x 10"4per ug/m3.
Application of the ADAFs to account for early-life susceptibility to the proposed mutagenic mode of
action for chloroprene yields an adjusted human lifetime cancer risk of 5 x 10"4 per ug/m3.
148
-------
7. REFERENCES
Appelman LM; Dreef-van der Meulen HC (1979). Reproduction study with beta-chloroprene vapour in rats
(Report No. 6225). Zeist, The Netherlands: Central Institute for Nutrition and Food Research (CIVO).
064938
Arrighi HM; Hertz-Picciotto I (1994). The evolving concept of the healthy worker survivor effect.
Epidemiology, 5: 189-196. 625164
Bartsch H; Malaveille C; Barbin A; Planche G (1979). Mutagenic and alkylating metabolites of halo-ethylenes,
chlorobutadienes and dichlorobutenes produced by rodent or human liver tissues: Evidence for oxirane
formation by P450-linked microsomal mono-oxygenases. Arch Toxicol, 41: 249-277.
http://dx.doi.org/10.1007/BF00296896.010689
Begemann P; Christova-Georgieva NI; Sangaiah R; Koc H; Zhang D; Golding BT; Gold A; Swenberg JA
(2004). Synthesis, characterization, and identification of N7-guanine adducts of isoprene monoepoxides
in vitro. Chem Res Toxicol, 17: 929-936. http://dx.doi.org/10.1021/tx0342565. 625093
Bernauer U; Garritsen H; Heinrich-Hirsch B; Gundert-Remy U (2003). Immunochemical analysis of
cytochrome P450 variability in human leukapheresed samples and its consequences for the risk
assessment process. Regul Toxicol Pharmacol, 37: 318-327. http://dx.doi.org/10.1016/S0273-
2300(03)00012-6.625103
Bochkov NP; Kuleshov NP; Zhurkov VS (1972). [Analysis of spontaneous chromosomal aberrations in a human
leukocyte culture]. Citologiya, 14: 1267-1272. 644818
Bulbulyan MA; Changuina OV; Zaridze DG; Astashevsky SV; Colin D; Boffetta P (1998). Cancer mortality
among Moscow shoe workers exposed to chloroprene (Russia). Cancer Causes Control, 9: 381-387.
http://dx.doi.org/10.1023/A: 1008863516506. 625105
Bulbulyan MA; Margaryan AG; Ilychova SA; Astashevsky SV; Uloyan SM; Cogan VY; Colin D; Boffetta P;
Zaridze DG (1999). Cancer incidence and mortality in a cohort of chloroprene workers from Armenia.
Int J Cancer, 81:31-33. 157419
Clary JJ; Feron VJ; Reuzel PGJ (1978). Toxicity of beta-chloroprene (2-chlorobutadiene-l,3): Acute and
subacute toxicity. Toxicol Appl Pharmacol, 46: 375-384. http://dx.doi.org/10.1016/0041-
008X(78)90082-0. 064942
Colonna M; Laydevant G (2001). A cohort study of workers exposed to chloroprene in the department of Isere,
France. Chem Biol Interact, 135-136: 505-514. http://dx.doi.org/10.1016/80009-2797(01)00185-5.
625112
Cottrell L; Golding BT; Munter T; Watson WP (2001). In vitro metabolism of chloroprene: Species differences,
epoxide stereochemistry and a de-chlorination pathway. Chem Res Toxicol, 14: 1552-1562. 157445
Cox DR (1972). Regression models and life-tables. J Roy Stat Soc B, 34: 187-220.
http://www.istor.org/stable/2985181.008785
Culik R; Kelly DP; Clary JJ (1978). Inhalation studies to evaluate the teratogenic and embryotoxic potential of
"beta"-chloroprene (2-chlorobutadiene-l,3). Toxicol Appl Pharmacol, 44: 81-88. 094969
DeWoskin RS (2007). PBPK models in risk assessment—A focus on chloroprene. Chem Biol Interact, 166: 352-
359. http://dx.doi.0rg/10.1016/j.cbi.2007.01.016. 202141
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
149
-------
Dong QA; Xiao BL; Hu YH; Li SQ (1989). Short-term test for the induction of lung tumor in mouse by
chloroprene. Biomed Environ Sci, 2: 150-153. 007520
Drevon C; Kuroki T (1979). Mutagenicity of vinyl chloride, vinylidene chloride and chloroprene in V79
Chinese hamster cells. Mutat Res, 67: 173-182. 010680
Esmen NA; Hall TA; Phillips ML; Jones EP; Basara H; Marsh GM; Buchanich JM (2007). Chemical process-
based reconstruction of exposures for an epidemiological study: Part II. Estimated exposures to
chloroprene and vinyl chloride. Chem Biol Interact, 166: 264-276.
http://dx.doi.0rg/10.1016/i.cbi.2006.08.010.625118
Esmen NA; Hall TA; Phillips ML; Mars GM (2007). Chemical process based reconstruction of exposures for an
epidemiological study: I. Theoretical and methodological issues. Chem Biol Interact, 166: 254-263.
http://dx.doi.0rg/10.1016/i.cbi.2006.08.009.625114
Esmen NA; Kennedy KJ; Hall TA; Phillips ML; Marsh GM (2007). Classification of worker exposures. Chem
Biol Interact, 166: 245-253. http://dx.doi.Org/10.1016/j.cbi.2006.08.008. 625121
Fomenko VN; Katosova LD (1973). [The results of cytogenic analysis of the peripheral blood in women
workers in contact with chloroprene latex]. In Nauchnye Trudy Akusherkso-Ginekologich-eskaya
(Scientific Publications of Obstetrical and Gynecological Care) (pp. 33-37). Kazan, Russia:
Profpatologiya (Department of Pathophysiology, Kazan State Medical University). 644819
Foureman P; Mason JM; Valencia R; Zimmering S (1994). Chemical mutagenesis testing in Drosophila. X.
Results of 70 coded chemicals tested for the National Toxicology Program. Environ Mol Mutagen, 23:
208-227. http://dx.doi.org/10.1002/em.2850230310. 065173
Gahlmann R (1993). Pros and cons of transgenic mouse mutagenesis test systems. J Exp Anim Sci, 35: 232-243.
625174
Gargas ML; Burgess RJ; Voisard DE; Cason GH; Andersen ME (1989). Partition coefficients of low-molecular-
weight volatile chemicals in various liquids and tissues. Toxicol Appl Pharmacol, 98: 87-99. 063084
Ginsberg G; Slikker W Jr; Bruckner J; Sonawane B (2004). Incorporating children's toxicokinetics into a risk
framework. Environ Health Perspect, 112: 272-283. http://dx.doi.org/10.1289/ehp.6013. 625124
Gooch JJ; Hawn WF (1981). Biochemical and hematological evaluation of chloroprene workers. J Occup Med,
23:268-272. 064944
Griebel CP; Halvorsen J; Golemon TB; Day AA (2005). Management of spontaneous abortion. Am Fam
Physician, 72: 1243-1250. http://www.aafp.org/afp/2005/1001/pl243.html. 625142
Grosjean D (1990). Atmospheric chemistry of toxic contaminants 1. Reaction rates and atmospheric persistence.
J Air Waste Manag Assoc, 40: 1397-1402. 625143
Grosjean D (1991). Atmospheric chemistry of toxic contaminants 5. Unsaturated halogenated aliphatics: Allyl
chloride, chloroprene, hexachlorocyclopentadiene, vinylidene chloride. J Air Waste Manag Assoc, 41:
182-189.625149
Haley TJ (1978). Chloroprene (2-chloro-1,3-butadiene)-- What is the evidence for its carcinogenicity. Clin
Toxicol, 13: 153-170. http://dx.doi.org/10.3109/15563657808988232. 010685
Hall TA; Esmen NA; Jones EP; Basara H; Phillips ML; Marsh GM; Youk AO; Buchanich JM; Leonard RC
(2007). Chemical process based reconstruction of exposures for an epidemiological study: III. Analysis
of industrial hygiene samples. Chem Biol Interact, 166: 277-284.
http://dx.doi.0rg/10.1016/i.cbi.2006.09.004.625243
Hill AB (1965). The environment and disease: Association or causation. Proc R Soc Med, 58: 295-300.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1898525/pdf/procrsmed00196-0010.pdf.071664
Himmelstein MW; Carpenter SC; Evans MV; Hinderliter PM; Kenyon EM (2004). Kinetic modeling of beta-
chloroprene metabolism: II. The application of physiologically based modeling for cancer dose response
analysis. Toxicol Sci, 79: 28-37. http://dx.doi.org/10.1093/toxsci/kfh096. 625154
150
-------
Himmelstein MW; Carpenter SC; Hinderliter PM (2004). Kinetic modeling of beta-chloroprene metabolism: I.
In vitro rates in liver and lung tissue fractions from mice, rats, hamsters, and humans. Toxicol Sci, 79:
18-27. http://dx.doi.org/10.1093/toxsci/kfh092. 625152
Himmelstein MW; Carpenter SC; Hinderliter PM; Snow TA; Valentine R (2001). The metabolism of beta-
chloroprene: Preliminary in-vitro studies using liver microsomes. Chem Biol Interact, 135-136: 267-
284. http://dx.doi.org/10.1016/S0009-2797(01)00214-9. 019012
Himmelstein MW; Gladnick NL; Donner EM; Snyder RD; Valentine R (2001). In vitro genotoxicity testing of
(l-chloroethenyl)oxirane, a metabolite of beta-chloroprene. Chem Biol Interact, 135-136: 703-713.
019013
HSDB (2009). 2-Chloro-1,3-Butadiene: Chemical/physical properties. Retrieved February 12, 2010 from
http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7HSDB. 594343
Hurst HE; Ali MY (2007). Analyses of (l-chloroethenyl)oxirane headspace and hemoglobin N-valine adducts in
erythrocytes indicate selective detoxification of (l-chloroethenyl)oxirane enantiomers. Chem Biol
Interact, 166: 332-340. http://dx.doi.Org/10.1016/j.cbi.2006.04.016. 625159
IARC (1999). IARC monographs on the evaluation of carcinogenic risks to humans: Volume 71: Re-evaluation
of some organic chemicals, hydrazine, and hydrogen peroxide. Lyon, France : IARC Working Group on
the Evaluation of Carcinogenic Risks to Humans; International Agency for Research on Cancer (IARC);
World Health Organization (WHO) . http://monographs.iarc.fr/ENG/Monographs/vol71/volume71.pdf
201838
Immels H; Willems M (1978). Initial submission: Dominant lethal assay with beta-chloroprene in male mice
with cover letter dated 09/01/92 (Report No. 88-920008417; NTIS# OTS0570699). Wilmiinton, DE:
E.I. DuPont deNemours & Company. (NTIS No. PB NTIS# OTS0570699) 625176
Jemal A; Murray T; Samuels A; Ghafoor A; Ward E; Thun MJ (2003). Cancer statistics, 2003. CA Cancer J Clin,
53: 5-26. http://dx.doi.Org/10.3322/canjclin.53.l.5. 625160
Jones JH; Young RJ; Selevan S (1975). Walk-through survey of E.I. du Pont de Nemours and Company, Inc.
Louisville, Kentucky (Report No. IWS-87-10; PB89-129845). Cincinnati, OH: National Institute for
Occupational Safety and Health. 625203
Koskinen M; Calebiro D; Hemminki K (2000). Styrene oxide-induced 2'-deoxycytidine adducts: implications
for the mutagenicity of styrene oxide. Chem Biol Interact, 126: 201-213. 010173
Krewski D; Crump KS; Farmer J; Gaylor DW; Howe R; Portier C; Salsburg D; Sielken RL; Van Ryzin J (1983).
A comparison of statistical methods for low dose extrapolation utilizing time-to-tumour data. Fundam
Appl Toxicol, 3: 140-160. http://dx.doi.org/10.1016/S0272-0590(83)80075-X. 003194
Leet TL; Selevan SG (1982). Mortality analysis of workers exposed to chloroprene (Final report for El DuPont
deNemours & Company).Cincinnati, OH: National Institute for Occupational Safety and Health; Center
for Disease Control; Public Health Service; Department of Health and Human Services. 094970
Lefaux R (1968). Chloroprene. In PP Hopf (Ed.), Practical toxicology of plastics (pp. 69-71). London: Iliffe
Books, Ltd. 625192
Leonard RC; Kreckmann KH; Lineker GA; Marsh G; Buchanich J; Youk A (2007). Comparison of standardized
mortality ratios (SMRs) obtained from use of reference populations based on company-wide registry
cohort to SMRs calculated against local and national rates. Chem Biol Interact, 166: 317-322.
http://dx.doi.0rg/10.1016/i.cbi.2006.09.001.625179
Li SQ; Dong QN; Liu YQ; LiuYG (1989). Epidemiologic study of cancer mortality among chloroprene
workers. Biomed Environ Sci, 2: 141-149. 625181
Li SQ; Dong QN; Liu YQ; LiuYG (1990). Epidemiologic study of cancer mortality among chloroprene workers
[Erattum]. Biomed Environ Sci, 3: 372. 644113
151
-------
Luster MI; Portier C; Pait DG; White KLJ; Gennings C; Munson AE; Rosenthal GJ (1992). Risk assessment in
immunotoxicology. I. Sensitivity and predictability of immune tests. Toxicol Sci, 18: 200-210.
http://dx.doi.org/10.1016/0272-0590(92)90047-L. 084126
Lynch M (2001). Manufacture and use of chloroprene monomer. Chem Biol Interact, 135-136: 155-167.
http://dx.doi.org/10.1016/50009-2797(01)00188-0. 625182
Lynch J (2001). BD monomer and elastomer production processes. Chem Biol Interact, 135-136: 147-153.
http://dx.doi.org/10.1016/80009-2797(01)00187-9. 646266
MacGregor JT; Wehr CM; Henika PR; Shelby MD (1990). The in vivo erythrocyte micronucleus test:
Measurement at steady state increases assay efficiency and permits integration with toxicity studies.
Fundam Appl Toxicol, 14: 513-522. http://dx.doi.org/10.1016/0272-0590(90)90255-I. 625184
Marsh GM; Youk AO; Buchanich JM; Cunningham M; Esmen NA; Hall TA; Phillips ML (2007). Mortality
patterns among industrial workers exposed to chloroprene and other substances: I. General mortality
patterns. Chem Biol Interact, 166: 285-300. http://dx.doi.Org/10.1016/j.cbi.2006.08.011. 625187
Marsh GM; Youk AO; Buchanich JM; Cunningham M; Esmen NA; Hall TA; Phillips ML (2007). Mortality
patterns among industrial workers exposed to chloroprene and other substances: II. Mortality in relation
to exposure. Chem Biol Interact, 166: 301-316. http://dx.doi.Org/10.1016/j.cbi.2006.08.012. 625188
MARTA; MTA (1996). Historical control data (1992-1994) for developmental and reproductive toxicity studies
using the Crl:CD (SD)BR rat.Wilmington, MA:
Middle Atlantic Reproduction and Teratology Association (MARTA); Midwest Teratology Association (
MTA); Charles River Laboratories .
http://www.criver.com/sitecollectiondocuments/rm_rm_r_tox_studies_crlcd_sd_br_rat.pdf 625111
Mast TJ; Evanhoff JJ; Westerberg RL; Weigel RJ; Rommereim RL (1994). Inhalation developmental toxicology
studies: Developmental toxicity of chloroprene vapors in New Zealand White rabbits (Report No. PNL-
7458; DE94012384). Research Triangle Park, NC: U.S. Department of Health and Human Services.
625206
McGlothlin JD; Meyer C; Leet TL (1984). Health hazard evaluation report for chlorprene, dichlorobuadiene,
dichlorobutene, trichlorobutene, and toluene at E.I. du Pont de Nemours and Company, Inc. Louisville,
Kentucky (Report No. HETA 79-027-1459; PB85-186757). Cincinnati, OH: National Institute for
Occupational Safety and Health. 625204
Medinsky MA; Bond JA (2001). Sites and mechanisms for uptake of gases and vapors in the respiratory tract.
Toxicology, 160: 165-172. 016157
Melnick RL; Elwell MR; Roycroft JH; Chou BJ; Ragan HA; Miller RA (1996). Toxicity of inhaled chloroprene
(2-chloro-1,3-butadiene) in F344 rats and B6C3F(1) mice. Toxicology, 108: 79-91.
http://dx.doi.org/10.1016/0300-483X(95)03286-O. 625207
Melnick RL; Sills RC (2001). Comparative carcinogenicity of 1,3-butadiene, isoprene, and chloroprene in rats
and mice. Chem Biol Interact, 135-136: 27-42. 051506
Melnick RL; Sills RC; Portier CJ; Roycroft JH; Chou BJ; Grumbein SL; Miller RA (1999). Multiple organ
carcinogenicity of inhaled chloroprene (2-chloro-1,3-butadiene) in F344/N rats and B6C3F1 mice and
comparison of dose-response with 1,3-butadiene in mice. Carcinogenesis, 20: 867-878. 000297
Melnick RL; Sills RC; Roycroft JH; Chou BJ; Ragan HA; Miller RA (1994). Isoprene, an endogenous
hydrocarbon and industrial chemical, induces multiple organ neoplasia in rodents after 26 weeks of
inhalation exposure. Can Res, 54: 5333-5339.
http://cancerres.aacrjournals.org/cgi/content/abstract/54/20/5333. 625208
Munter T; Cottrell L; Ghai R; Golding BT; Watson WP (2007). Erratum to "The metabolism and molecular
toxicology of chloroprene" [Chem Biol Interact 2007; 166(l-3):323-331]. Chem Biol Interact, 168: 169.
http://dx.doi.0rg/10.1016/j.cbi.2007.03.007.625213
152
-------
Munter T; Cottrell L; Ghai R; Golding BT; Watson WP (2007). The metabolism and molecular toxicology of
chloroprene. Chem Biol Interact, 166: 323-331. http://www.sciencedirect.com/science/article/B6T56-
4K3KC3R-3/2/d2037b7cb89e542cd823e7d8f70b9857. 576501
Munter T; Cottrell L; Golding BT; Watson WP (2003). Detoxication pathways involving glutathione and
epoxide hydrolase in the in vitro metabolism of chloroprene. Chem Res Toxicol, 16: 1287-1297.
http://dx.doi.org/10.1021/tx034107m.625214
Munter, T; Cottrell, L; Hill, S; Kronberg L; Watson WP; Golding BT (2002). Identification of adducts derived
from reactions of (l-chloroethenyl)oxirane with nucleosides and calf thymus DNA. Chem Res Toxicol,
15: 1549-1560. http://dx.doi.org/10.1021/tx020070e. 625215
Neafsey P; Ginsberg G; Hattis D; Johns DO; Guyton KZ; Sonawane B (2009). Genetic polymorphism in
CYP2E1: Population distribution of CYP2E1 activity. J Toxicol Environ Health B Crit Rev, 12: 362-
388. http://dx.doi.org/10.1080/10937400903158359. 196814
NIOSH (1977). Criteria for a recommended standard: Occupational exposure to chloroprene .Washington, DC:
U.S. Department of Health, Education, and Welfare, http://www.cdc.gov/niosh/77-210.html. 644450
NIOSH (1995). Occupational safety and health guideline for beta-chloroprene.Washington, DC: National
Institute for Occupational Safety and Health, http://www.cdc.gov/niosh/docs/81-123/pdfs/0133-rev.pdf
644453
NRC (1994). Science and judgment in risk assessment. Washington, DC: National Research Council (NRC);
National Academy Press . http://www.nap.edu/openbook.php?isbn=030904894X. 006424
NTP (1985). Technical protocol for sperm morphology and vaginal cytology evaluations in toxicity testing for
rats and mice, 10/31/82 version (updated March 1985).Washington, DC: National Toxicology Program
(NTP); National Institutes of Health; Public Health Service; U.S. Department of Health and Human
Services . 625205
NTP (1998). Toxicology and carcinogenesis studies of chloroprene (CAS No 126-99-8) in F344/N rats and
B6C3F1 mice (inhalation studies) (Report No. NTP TR 467; NIH PUB 98-3957). Research Triangle
Park, NC: National Toxicology Program. http://ntp.niehs.nih.gov/?objectid=070A80DO-96F2-115A-
6354967089C17F9B. 042076
NTP (2005). Report on carcinogens (llth edition).Research Triangle Park, NC: National Toxicology Program
(NTP); National Institutes of Health; Public Health Service; U.S. Department of Health and Human
Services . http://ntp.niehs.nih.gov/index.cfm?objectid=32BA9724-FlF6-975E-7FCE50709CB4C932.
093207
Nystrom AE (1948). Health hazards in the chloroprene rubber industry and their prevention: A clinical and
experimental study with special reference to chloroprene as well as oxidation and polymerization
products thereof. J Intern Med, 219: 1-125. 003695
OECD (1998). Screening Information Data Set (SIDS) high production volume chemicals.Nairobi, Kenya:
Organisation for Economic Co-operation and Development (OECD); United Nations Environment
Programme (UNEP) . http://www.inchem.org/documents/sids/sids/Chloroprene.pdf 624889
Pell S (1978). Mortality of workers exposed to chloroprene. J Occup Environ Med, 20: 21-29. 064957
Placke ME; Griffis L; Bird M; Bus J; Persing RL; Cox LA Jr (1996). Chronic inhalation oncogenicity study of
isoprene in B6C3F1 mice. Toxicology, 113: 253-262. 624891
Ponomarkov V; Tomatis L (1980). Long-term testing of vinylidene chloride and chloroprene for carcinogenicity
in rats. Oncology, 37: 136-141. http://dx.doi.org/10.1159/000225422. 075453
Portier CJ; Bailer AJ (1989). Testing for increased carcinogenicity using a survival-adjusted quantal response
test. Toxicol Sci, 12: 731-737. 093236
Putman E; van der Laan JW; van Loveren H (2003). Assessing immunotoxicity: guidelines. Fundam Clin
Pharmacol, 17: 615-626. 624893
153
-------
Rice JM; Boffetta P (2001). 1,3-Butadiene, isoprene and chloroprene: reviews by the IARC monographs
programme, outstanding issues, and research priorities in epidemiology. Chem Biol Interact, 135-136:
11-26.624894
Romazini S; Laydevant G; Lutz J; Colonna M; Menegoz F (1992). Mortality study in occupational exposure to
chloroprene. Arch Mai Prof Environ, 54: 721-725. 624896
Rothman KJ; Greenland S (1998). Modern epidemiology. Philadelphia, PA: Lippincott-Raven Publishers.
086599
Sanotskii IV (1976). Aspects of the toxicology of chloroprene: Immediate and long-term effects. Environ Health
Perspect, 17: 85-93. 063885
Savitz DA; Sonnenfeld NL; Olshan AF (1994). Review of epidemiologic studies of paternal occupational
exposure and spontaneous abortion. Am J Ind Med, 25: 361-383. 068186
Schrag SD; Dixon RL (1985). Occupational exposures associated with male reproductive dysfunction. In R
George; R Okun (Eds.), Annual Review of Pharmacology and Toxicology (pp. 567-592). Palo Alto, CA:
Annual Reviews. 062573
Shelby MD (1990). Results of NTP-sponsored mouse cytogenetic studies on 1,3-butadiene, isoprene, and
chloroprene. Environ Health Perspect, 86: 71-73. 624906
Shelby MD; Witt KL (1995). Comparison of results from mouse bone marrow chromosome aberration and
micronucleus tests. Environ Mol Mutagen, 25: 302-313. 624921
Sills RC; Hong HL; Boorman GA, Devereux TR; Melnick RL (2001). Point mutations of K-ras and H-ras genes
in forestomach neoplasms from control B6C3F1 mice and following exposure to 1,3-butadiene, isoprene
or chloroprene for up to 2-years. Chem Biol Interact, 135-136: 376-386. 624922
Sills RC; Hong HL; Melnick RL; Boorman GA; Devereux TR (1999). High frequency of codon 61 K-ras A~>T
transversions in lung and Harderian gland neoplasms of B6C3F1 mice exposed to chloroprene (2-
chloro-l,3-butadiene) for 2 years, and comparisons with the structurally related chemicals isoprene and
1,3-butadiene. Carcinogenesis, 20: 657-662. 624952
Stewart CA Jr (1971). Dimerization of chloroprene and related dienes. J Am Chem Soc, 93: 4815-4821. 010705
Stewart CA Jr (1993). Chloroprene. In JI Kroschwitz; M Howe-Grant (Eds.), Kirk-Othmer encyclopedia of
chemical technology (pp. 70-78). New York, NY: John Wiley & Sons. 010679
Summer K-H; Greim H (1980). Detoxification of chloroprene (2-chloro-1,3-butadiene) with glutathione in the
rat. Biochem Biophys Res Commun, 96: 566-573. http://dx.doi.org/10.1016/0006-291X(80)91393-5.
064961
Tarone RE (1975). Tests for trend in life table analysis. Biometrika, 62: 679-690.
http://dx.doi.Org/10.1093/biomet/62.3.679.624959
Thornton-Manning JR; Dahl AR (1997). Metabolic capacity of nasal tissue: Interspecies comparisons of
xenobiotic-metabolizing enzymes. MutatRes, 380: 43-59. http://dx.doi.org/10.1016/S0027-
5107(97)00126-7.597688
Tice RR (1988). The cytogenetic evaluation of in vivo genotoxic and cytotoxic activity using rodent somatic
cells. Cell Biol Toxicol, 4: 475-486. 624981
Tice RR; Boucher R; Luke CA; Paquette DE; Melnick RL; Shelby MD (1988). Chloroprene and isoprene:
cytogenetic studies in mice. Mutagenesis, 3: 141-146. 064962
Ton TV; Hong HH; Devereux TR; Melnick RL; Sills RC; Kim Y (2007). Evaluation of genetic alterations in
cancer-related genes in lung and brain tumors from B6C3F1 mice exposed to 1,3-butadiene or
chloroprene. Chem Biol Interact, 166: 112-120. http://dx.doi.Org/10.1016/j.cbi.2006.04.015. 625004
154
-------
Trochimowicz HJ; Loser E; Feron VJ; Clary JJ; Valentine R (1998). Chronic inhalation toxicity and
carcinogenicity studies on B-chloroprene in rats and hamsters. Inhal Toxicol, 10: 443-472.
http://dx.doi.org/10.1080/089583798197628 . 625008
U.S. EPA (2000). Chloroprene (2-chloro-l,3-butadiene) - Fact Sheet Hazard Summary - Created in April 1992;
Revised in January 2000. Retrieved August 31, 2010 from
http://www.epa.gov/ttn/atw/hlthef/chloropr.html. 625036
U.S. EPA (1985). Summary overview of health effects associated with chloroprene: Health issue assessment
(Report No. EPA/600/8-85/01 IF). Research Triangle Park, NC: Environmental Criteria and Assessment
Office; Office of Health and Environmental Assessment; Office of Research and Development; U.S.
Environmental Protection Agency . http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=30001EYD.txt.
017624
U.S. EPA (1986). Guidelines for mutagenicity risk assessment (Report No. EPA/630/R-98/003). Washington,
DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://www.epa.gov/osa/mmoaframework/pdfs/MUTAGEN2.PDF. 001466
U.S. EPA (1986). Guidelines for the health risk assessment of chemical mixtures (Report No. EPA/630/R-
98/002). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=22567. 001468
U.S. EPA (1988). Recommendations for and documentation of biological values for use in risk assessment
(Report No. EPA/600/6-87/008). Cincinnati, OH: U.S. Environmental Protection Agency,
Environmental Criteria and Assessment Office, Office of Research and Development, Office of Health
and Environmental Assessment. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=34855. 064560
U.S. EPA (1989). Health and environmental effects document for 2-chloro-l,3,-butadiene (chloroprene) (Report
No. ECAO-CIN-G037). Cincinnati, OH: U.S. Environmental Protection Agency, Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office.
http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=900R0300.txt. 625024
U.S. EPA (1991). Guidelines for developmental toxicity risk assessment (Report No. EPA/600/FR-91/001).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=23162. 008567
U.S. EPA (1994). Interim policy for particle size and limit concentration issues in inhalation toxicity
studies.Washington, DC: U.S. Environmental Protection Agency, Office of Pesticide Products, Health
Effects Division. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=186068. 076133
U.S. EPA (1994). Methods for derivation of inhalation reference concentrations and application of inhalation
dosimetry (Report No. EPA/600/8-90/066F). Washington, DC: U.S. Environmental Protection Agency,
Office of Research and Development, Office of Health and Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993. 006488
U.S. EPA (1995). The use of the benchmark dose approach in health risk assessment (Report No. EPA/630/R-
94/007). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=42601. 005992
U.S. EPA (1996). Guidelines for reproductive toxicity risk assessment (Report No. EPA/63O/R-96/009).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2838. 030019
U.S. EPA (1998). Guidelines for neurotoxicity risk assessment (Report No. EPA/63O/R-95/00IF). Washington,
DC: U.S. Environmental Protection Agency, Office of Research and Development, National Center for
Environmental Assessment. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=12479. 030021
U.S. EPA (2000). Benchmark dose technical guidance document [external review draft] (Report No.
EPA/630/R-00/001). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://www.epa.gov/raf/publications/benchmark-dose-doc-draft.htm . 052150
155
-------
U.S. EPA (2000). Science policy council handbook: Risk characterization (Report No. EPA 100-B-00-002).
Washington, B.C.: U.S. Environmental Protection Agency, Office of Research and Development, Office
of Science Policy, http://www.epa.gov/osa/spc/pdfs/rchandbk.pdf 052149
U.S. EPA (2000). Supplementary guidance for conducting health risk assessment of chemical mixtures (Report
No. EPA/63O/R-00/002). Washington, DC: Risk Assessment Forum, U.S. Environmental Protection
Agency, http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=20533. 004421
U.S. EPA (2000). Toxicological review of vinyl chloride (Report No. EPA/635R-00/004). Washington, DC: U.S.
EPA. http://www.epa.gov/iris. 194536
U.S. EPA (2002). A review of the reference dose and reference concentration processes (Report No. EPA/630/P-
02/0002F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=55365. 088824
U.S. EPA (2002). Health assessment of 1,3-butadiene (Report No. EPA/600/P-98/001F). Washington, D.C. :
U.S. Environmental Protection Agency. 052153
U.S. EPA (2005). Guidelines for carcinogen risk assessment, final report (Report No. EPA/63O/P-03/00IF,
PB2005-105899). Washington, DC: Risk Assessment Forum; U.S. Environmental Protection Agency.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 16283. 086237
U.S. EPA (2005). Supplemental guidance for assessing susceptibility from early-life exposure to carcinogens
(Report No. EPA/630/R-03/003F). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=160003. 088823
U.S. EPA (2006). A framework for assessing health risk of environmental exposures to children (final) (Report
No. EPA/600/R-05/093F). Washington, DC: U.S. Environmental Protection Agency, Office of Research
and Development, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=158363. 194567
U.S. EPA (2006). U.S. Environmental Protection Agency peer review handbook (Report No. EPA/100/B-
06/002). Washington, DC: U.S. Environmental Protection Agency, Science Policy Council.
http://www.epa.gov/peerreview/pdfs/peer_review_handbook_2006.pdf 194566
U.S. EPA (2009). Benchmark dose software (BMDS). http://www.epa.gov/NCEA/bmds. 200772
U.S. EPA (2009). Status report: Advances in inhalation dosimetry of gases and vapors with portal of entry
effects in the upper respiratory tract (Report No. EPA/600/R-09/072). Washington, DC: U.S.
Environmental Protection Agency, http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=212131.
625038
Vogel E (1979). Mutagenicity of chloroprene, 1-chloro-1,3-trans-butadiene, l,4-dichlorobutene-2 and 1,4-
dichloro-2,3-epoxybutane in Drosophila melanogaster. Mutat Res, 67: 377-381. 000948
Watson WP; Cottrell L; Zhang D; Golding BT (2001). Metabolism and molecular toxicology of isoprene. Chem
Biol Interact, 135-136: 223-238. 625045
Westphal GA; Blaszkewicz M; Leutbecher M; Miiller A; Hallier E; Bolt HM (1994). Bacterial mutagenicity of
2-chloro-1,3-butadiene (chloroprene) caused by decomposition products. Arch Toxicol, 68: 79-84.
625047
Willems MI (1978). Evaluation of beta-chloroprene and five dimers in the Salmonella/microsome mutagenicity
test (Report No. R 5712). Zeist, The Netherlands: Central Institute for Nutrition and Food Research
(CIVO). 625048
Willems MI (1980). Evaluation of beta-chloroprene and four chloroprene dimers in the Ames test by
atmospheric exposure of the tester strains (Report No. R 6392; OTS0570703; 88-920008421). Zeist,
The Netherlands: Central Institute for Nutrition and Food Research (CIVO).
https://ntrl.ntis.gov/search/TRLProductDetail.aspx?ABBR=OTS0570703. 625049
156
-------
7.1. REFERENCES ADDED AFTER EXTERNAL PEER REVIEW1
Acquavella JF; Leonard RC (2001). A review of the epidemiology of 1,3-butadiene and chloroprene. Chem Biol
Interact, 135-136: 43-52. http://dx.doi.org/10.1016/S0009-2797(01)00169-7. 628495
Arrighi HM; Hertz-Picciotto I (1994). The evolving concept of the healthy worker survivor effect.
Epidemiology, 5: 189-196. 625164
Bernauer U; Garritsen H; Heinrich-Hirsch B; Gundert-Remy U (2003). Immunochemical analysis of
cytochrome P450 variability in human leukapheresed samples and its consequences for the risk
assessment process. Regul Toxicol Pharmacol, 37: 318-327. http://dx.doi.org/10.1016/S0273-
2300(03)00012-6. 625103
Bochkov NP; Kuleshov NP; Zhurkov VS (1972). [Analysis of spontaneous chromosomal aberrations in a
human leukocyte culture]. Citologiya, 14: 1267-1272. 644818
Bukowski JA (2009). Epidemiologic evidence for chloroprene carcinogenicity: Review of study quality and its
application to risk assessment. Risk Anal, 29: 1203-1216. http://dx.doi.org/10. llll/j.1539-
6924.2009.01254.x. 628496
DeWoskin RS (2007). PBPK models in risk assessment—A focus on chloroprene. Chem Biol Interact, 166:
352-359. http://dx.doi.0rg/10.1016/j.cbi.2007.01.016. 202141
Fajen JM; lingers LJ (1986). Industrial hygiene walk-through survey report of E.I. duPont deNemours and
Company, Pontchartrain Works, LaPlace, Louisiana (Report No. 147.31 ). Cincinnati, OH: National
Institute for Occupational Safety and Health; Centers for Disease Control. 628500
Fomenko VN; Katosova LD (1973). [The results of cytogenic analysis of the peripheral blood in women
workers in contact with chloroprene latex]. In Nauchnye Trudy Akusherkso-Ginekologich-eskaya
(Scientific Publications of Obstetrical and Gynecological Care) (pp. 33-37). Kazan, Russia:
Profpatologiya (Department of Pathophysiology, Kazan State Medical University). 644819
Gooch JJ; Hawn WF (1981). Biochemical and hematological evaluation of chloroprene workers. J Occup
Med, 23: 268-272. 064944
Heuser VD; de Andrade VM; da Silva J; Erdtmann B (2005). Comparison of genetic damage in Brazilian
footwear-workers exposed to solvent-based or water-based adhesive. Mutat Res Genet Toxicol Environ
Mutagen, 583: 85-94. http://dx.doi.Org/10.1016/j.mrgentox.2005.03.002. 479853
Leonard RC; Kreckmann KH; Lineker GA; Marsh G; Buchanich J; Youk A (2007). Comparison of
standardized mortality ratios (SMRs) obtained from use of reference populations based on
company-wide registry cohort to SMRs calculated against local and national rates. Chem Biol
Interact, 166: 317-322. http://dx.doi.0rg/10.1016/i.cbi.2006.09.001.625179
Li SQ; Dong QN; Liu YQ; Liu YG (1990). Epidemiologic study of cancer mortality among chloroprene
workers [Erattum]. Biomed Environ Sci, 3: 372. 644113
McGlothlin JD; Meyer C; Leet TL (1984). Health hazard evaluation report for chlorprene,
dichlorobuadiene, dichlorobutene, trichlorobutene, and toluene at E.I. du Pont de Nemours and
Company, Inc. Louisville, Kentucky (Report No. HETA 79-027-1459; PB85-186757). Cincinnati,
OH: National Institute for Occupational Safety and Health. 625204
Musak L; Soucek P; Vodickova L; Naccarati A; Halasova E; Polakova V; Slyskova J; Susova S; Buchancova J;
Smerhovsky Z; Sedikova J; Klimentova G; Osina O; Hemminki K; Vodicka P (2008). Chromosomal
aberrations in tire plant workers and interaction with polymorphisms of biotransformation and DNA
repair genes. Mutat Res-Fundam Mol Mech Mutagen, 641: 36-42.
http://dx.doi.0rg/10.1016/j.mrfmmm.2008.02.007.62850i
The unbolded references are those that are only cited in Appendix A.
157
-------
Neafsey P; Ginsberg G; Hattis D; Johns DO; Guyton KZ; Sonawane B (2009). Genetic polymorphism in
CYP2E1: Population distribution of CYP2E1 activity. J Toxicol Environ Health B Crit Rev, 12:
362-388. http://dx.doi.org/10.1080/10937400903158359. 196814
NIOSH (1977). Criteria for a recommended standard: Occupational exposure to chloroprene.Washington,
DC: U.S. Department of Health, Education, and Welfare, http://www.cdc.gov/niosh/77-210.html.
644450
NIOSH (1995). Occupational safety and health guideline for beta-chloroprene.Washington, DC: National
Institute for Occupational Safety and Health, http://www.cdc.gov/niosh/docs/81-123/pdfs/0133-
rev.pdf. 644453
U.S. EPA (1989). Health and environmental effects document for 2-chloro-l,3,-butadiene (chloroprene)
(Report No. ECAO-CIN-G037). Cincinnati, OH: U.S. Environmental Protection Agency, Office of
Health and Environmental Assessment, Environmental Criteria and Assessment Office.
http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=900R0300.txt. 625024
U.S. EPA (2009). Benchmark dose software (BMDS). http://www.epa.gov/NCEA/bmds. 200772
U.S. EPA (2009). Status report: Advances in inhalation dosimetry of gases and vapors with portal of entry
effects in the upper respiratory tract (Report No. EPA/600/R-09/072). Washington, DC: U.S.
Environmental Protection Agency. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=212131.
625038
158
-------
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW
AND PUBLIC COMMENTS AND DISPOSITION
The Toxicological Review of chloroprene (dated September, 2010) has undergone a formal
external peer review performed by scientists in accordance with EPA guidance on peer review
(U.S. EPA, 2006, 194566). An external peer-reviewed workshop was held January 6, 2010. 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 also received
scientific comments from the public. These comments and EPA's responses are included in a separate
section of this appendix. There were six external peer reviewers.
A.l. EXTERNAL PEER REVIEWER 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.
When the external peer reviewers commented on decisions and analyses in the Toxicological
Review under multiple charge questions, these comments were organized under the most appropriate
charge question. In addition, the external peer reviewers made numerous specific comments that were
organized and responded to in a separate section of the section of this appendix. When multiple
reviewers provided specific comments on the same subject, or suggested similar revisions to the
document, their comments were combined, as appropriate.
A.1.1. General Charge Questions
Charge Question 1. Is the Toxicological Review logical, clear and concise? Has EPA clearly
synthesized the scientific evidence for noncancer and cancer hazards?
Comment 1: All six reviewers commented that the Toxicological Review was generally logical, clear,
and concise, although individual reviewers provided suggestions for the improvement of the document
with regards to clarity, transparency and thoroughness. One reviewer commented that a more rigorous
and transparent evaluation of the epidemiological evidence and how it integrated with the entirety of
the chloroprene database should be performed. This reviewer commented that the descriptor of "likely
to be carcinogenic to humans" was justified based on the animal and genotoxicity data, but this
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
A-1
-------
reviewer felt that the human epidemiological data had been overstated. One reviewer commented that
it was not clear why a particular dose-response model was chosen in the quantitative analysis of
noncancer effects if more than one model provided adequate fit. This reviewer also commented that
the rationale for the benchmark response level was not adequately justified. One reviewer commented
that the epidemiology section should have been consolidated (e.g., all studies using a particular cohort,
the Louisville DuPont Works for instance, should be discussed together). This reviewer also
recommended that additional analyses (by age at onset/death with lags) and substudies (nested case-
control) should have been included in the document. This reviewer also commented that additional
studies should be included (this issue is addressed in General Charge Question 2) and that
discrepancies in employee populations included in the studies between epidemiology studies and
NIOSH walk-throughs should be resolved.
Response: Additional information and a more thorough evaluation, integration and discussion of the
epidemiologic database, including individual study limitations, were included in the document to
enhance document completeness, transparency, and clarity (Sections 4.1.1 and 4.7.2). EPA concluded
that the epidemiologic data, considered as a complete database of information with study and
methodological issues taken into account, is generally coherent with the animal and genotoxic data,
and thus supports the conclusion that the most suitable descriptor was "likely to be carcinogenic to
humans."
Additional discussion regarding how the benchmark modeling of noncancer endpoints was
performed and how and why particular models were selected for each endpoint was included in the text
(Section 5.2.2). Specifically, the criteria that were used to determine adequacy of model fit (global
goodness-of-fit p-value, %2 residuals, and visual inspection) were discussed, as well as how the EPA
chose the best model when multiple models appropriately fit the dose-response data for an individual
endpoint (i.e., AIC when no model dependence is assumed, and BMDL otherwise). Additional
discussion and rationale for the chosen BMR levels was included.
The basic structure of the epidemiology section (Section 4.1.1.2; i.e., discussion of earlier
studies first) was retained in the document. The recommendation of additional analyses (by age at
onset/death with lags) and substudies (nested case-control) of existing cohorts was beyond the purpose
and purview of the Toxicological Review and none were included therein. Discrepancies in the study
populations in epidemiological studies and the NIOSH walk-through survey reports were due to study
inclusion criteria. The NIOSH reports enumerated all previous and current employees of the Louisville
Works plant, whereas Marsh et al. (2007, 625187: 2007, 625188) indicated that the study population
was limited to only those employees with a possibility of chloroprene exposure from plant start-up
through 2000. A more complete description of inclusion criteria for the Marsh papers was added to the
document, but no discussion regarding worker numbers contained in the NIOSH reports was deemed
necessary.
A-2
-------
Charge Question 2. Please identify any additional studies that should be considered in the assessment
of the noncancer and cancer health effects of chloroprene.
Comment 1: Four reviewers reported that they were not aware of any additional studies whose
exclusion would significantly impact the Toxicological Review. Two reviewers commented that three
NIOSH walk-through survey reports of DuPont plants involved in chloroprene production (Fajen and
lingers, 1986, 628500: Jones et al., 1975, 625203: McGlothlin et al., 1984, 625204) should be included
in Toxicological Review. These walk-through survey reports included medical assessments of worker
health and industrial hygienic analyses of ambient chloroprene concentrations in manufacture areas.
These two reviewers also commented that two additional health studies, one investigating clinical
chemistry and hematological outcomes (Gooch and Hawn, 1981, 064944) and the other a reanalysis of
the Louisville cohort compared to an external employee database (Leonard et al., 2007, 625179), be
included. One reviewer suggested that two additional reviews of the epidemiology literature at least be
considered for inclusion in the Toxicological Review (Acquavella and Leonard, 2001, 628495:
Bukowski, 2009, 628496). One reviewer suggested that a study detailing the use of a PBPK model for
estimation of rodent versus human delivered doses be included in the document (DeWoskin, 2007,
202141). One reviewer commented that two recent studies of genetic damage in workers potentially
exposed to chloroprene be included (Heuser et al., 2005, 479853: Musak et al., 2008, 628501).
Response: Two of the three suggested NIOSH walk-through survey reports were added to the
discussion of human health effects of chloroprene exposure (Jones et al., 1975, 625203: McGlothlin et
al., 1984, 625204). These studies included both ambient and personal air monitoring of chloroprene
exposures within the Louisville Works DuPont plant as well as a qualitative medical examination.
Although no health effects were associated with chloroprene exposure, these studies provided
information on pre- and post-employment health assessments conducted at the plant, as well as air
monitoring information. The third NIOSH walk-through survey report was not included in the
Toxicological Review as it primarily dealt with butadiene air monitoring (Fajen and lingers, 1986,
628500). The two additional health studies were included in the document (Gooch and Hawn, 1981,
064944: Leonard et al., 2007, 625179). The first was an examination of clinical chemistry and
hematological effects at the Louisville Works plant and found no significant health outcomes
associated with chloroprene exposure. The second study was a re-analysis of cancer mortality data
from the Louisville Works plant compared to external DuPont employee mortality databases in order to
assess the effects of the healthy worker bias. When mortality data from the Louisville Works plant was
compared to employee mortality databases, significant increases in SMRs were observed. These
findings possibly indicated that the protective associations observed when comparing Louisville Works
mortality data to general population databases may have been due to the healthy worker effect. The
two additional reviews of the primary epidemiology literature (Acquavella and Leonard, 2001, 628495:
Bukowski, 2009, 628496) were reviews of primary literature already included in the assessment.
A-3
-------
Therefore, these reviews were not added to the document as the purpose of the Toxicological Review is
to provide information on the EPA's independent review of the epidemiology database. A discussion of
the paper detailing the potential use of a PBPK model was added to the document (DeWoskin, 2007,
202141). The two recent papers reporting on genetic damage in workers exposed to chloroprene were
not added to the document as one was a study of health effects associated with exposure to solvent
versus water-based adhesives and included multiple co-exposures, and the other focused on
lymphocyte chromosomal aberrations due to butadiene exposure (Heuser et al., 2005, 479853; Musak
et al., 2008, 628501). The second study did provide information on genetic polymorphisms in genes
encoding metabolic enzymes, but this was duplicative of background information already provided in
the document.
A.1.2. Chemical-Specific Charge Questions
A.l.2.1. Oral Reference Dose (RfD) for Chloroprene
Charge Question 3. An RfD was not derived for chloroprene. Has the scientific justification for not
deriving an RfD been clearly described in the document? Please identify and provide the rationale for
any studies that should be selected as the principal study.
Comment 1: All six reviewers commented that the rationale for not deriving an RfD, including lack of
an adequate multiple-dose oral animal toxicity study and the lack of any human data on oral exposure
to chloroprene, was suitably described in the document. The reviewers concluded that the scientific
justification was appropriate and the decision to not derive an RfD was well founded. One reviewer
commented that an RfD derivation would be supported if a suitable PBPK model were used for a
route-to-route extrapolation from inhalation to oral data. One reviewer disagreed, and commented a
reliable route-to-route extrapolation via a PBPK model was not supported due to lack of information
on the disposition of chloroprene after inhalation or oral exposures.
Response: A more thorough discussion of the current PBPK model, including its strengths and
weaknesses relevant to route-to-route extrapolations, was included in Section 3.5. EPA concluded that,
based on the available scientific information and consistent with the conclusions of the External Peer
Reviewers, an RfD derivation was not supported. The expanded discussion of the PBPK model was
referenced in this decision not to use a route-route extrapolation for the purpose of deriving an RfD.
A. 1.2.2. Inhalation Reference Concentration (RfC) for Chloroprene
Charge Question 4. A chronic RfC for chloroprene has been derived from an inhalation toxicity study
(NTP, 1998, 042076) investigating noncancer effects in multiple organ systems. Please comment on
A-4
-------
whether the selection of this study as the principal study is scientifically justified. Please identify and
provide the rationale for any other studies that should be selected as the principal study.
Comment 1: All six reviewers concluded that the selection of the NTP (1998, 042076) inhalation
toxicity study as the principal study was scientifically justified as it was a well designed and conducted
study that identified multiple noncancer effects in multiple organ systems in rats and mice exposed to a
wide range of chloroprene. Two reviewers noted that not choosing the Trochimowicz et al. (1998,
625008) study for selection as the principal study was justified, although one of these reviewers
offered that the specific reason for not considering the study was weak and that a more appropriate and
defensible justification would be the high mortality in the low dose animals due to the failure of the
ventilation system. One reviewer noted that two human studies conducted at the Louisville plant
(Gooch and Hawn, 1981, 064944; McGlothlin et al., 1984, 625204) may contain useful information on
subchronic effects in humans. The reviewer also suggested that the limitation of the studies [i.e., lack
of quantitative exposure data in Gootch and Hawn (1981, 064944) and lack of quantitative medical
data in McGlothlin (1984, 625204)1 limit their utility and rule out their selection as the principal study.
Response: Selection of the NTP (1998, 042076) study as the principal study was maintained in the
Toxicological Review. Text was added to the document (Section 5.2.1) clarifying the reasons
Trochimowicz et al. (1998, 625008) was not selected as the principal study. Discussion of both the
Gooch and Hawn (1981, 064944) and McGlothlin (1984, 625204) studies was added to Section
4.1.2.1, including study details, strengths, weaknesses, and findings. Additional text was not necessary
in Section 5 detailing why these studies were not selected as the critical study; Section 5.2.1 contains
text stating "no human studies are available that would allow for the quantification of sub-chronic or
chronic noncancer effects."
Charge Question 5. An increase in the incidence of degenerative nasal lesions in male rats,
characterized by olfactory epithelial atrophy and/or necrosis with increasing severity, was selected as
the critical effect. Please comment on the scientific justification for combining the incidence of
atrophy and necrosis and for selecting this endpoint as the critical effect. Please identify and provide
the rationale for any other endpoints that should be considered in the selection of the critical effect.
Comment 1: Five reviewers commented that the selection of an increase in the incidence of
degenerative nasal lesions (characterized by olfactory epithelial atrophy and/or necrosis) was
reasonable and justified. One reviewer disagreed with selection of degenerative nasal lesions as the
critical effect for a number of reasons. First, this reviewer commented that the rationale for combining
the lesions and the precise way in which they were combined was poorly described. Second, the
reviewer stated that the concept that necrosis precedes atrophy is straightforward and has been
observed for a number of inhaled toxicants, whereas the draft Toxicological Review suggested that
A-5
-------
atrophy occurred first. Lastly, the reviewer commented that nasal lesions should not be selected as the
critical effect due to the way the HEC values were calculated (see comments below in Charge
Question 6).
One reviewer noted that combination of the two lesion types did not make a large difference in
the overall determination as the incidences of each endpoint were equivalent and the calculated
PODnEc values were 1.1 mg/m3 for atrophy and 1.0 mg/m3 for the combined lesions. This reviewer
also commented that the limitation of considering only endpoints that were significantly increased at
the low dose for the critical effect was not justified as it could have inappropriately excluded sensitive
endpoints that may return lower PODs given the nature of the dose-response relationship. This
reviewer commented that kidney (renal tubule) hyperplasia in male mice and rats should be
considered, and that these endpoints, as well as olfactory effects in female rats, female mice, and male
mice, should be included in Figure 5-1. One reviewer commented Table 5-1 did not include p-values
for trend for the dose-response for the various endpoints, but that the relative magnitude of trend
appeared to be greater for atrophy and necrosis combined than for splenic hematopoietic cell
proliferation. One reviewer commented that issues relating to in situ metabolism should be discussed
in more detail, specifically in regard to why upper respiratory effects were selected rather than lower
respiratory effects.
Response: Section 5.2 of the document was significantly rewritten in response to reviewer comments
regarding Charge Question 6 (see below). Specific comments regarding the combination of nasal
olfactory atrophy and necrosis (e.g., poorly explained rationale, incorrect conclusion that atrophy
precedes necrosis, and the negligible effect combining the lesions has on the PODHEc values) are no
longer relevant as the combination of nasal lesions was ultimately not performed for the purposes of
deriving the RfC; all text describing the combination of atrophic and necrotic nasal lesions has been
deleted. In response to the comment regarding endpoint selection criteria, additional endpoints were
considered for selection as the critical effect and modeled (Section 5.2.1). PODs for these endpoints
were determined using either BMD modeling or the NOAEL/LOAEL approach and were included in
Table 5-2. The additional endpoints considered were nasal olfactory basal cell hyperplasia in male and
female rats, nasal olfactory metaplasia in male and female rats, nasal olfactory atrophy in female rat,
nasal olfactory necrosis in female rats, nasal olfactory suppurative inflammation in female mice,
kidney (renal tubule) hyperplasia in male and female rats and male mice, forestomach epithelial
hyperplasia in male and female mice, and splenic hematopoietic cell proliferation in male mice.
Histiocytic cell infiltration was excluded from consideration as NTP (1998, 042076) noted that it was
an effect secondary to lung neoplasms.
Results for statistical tests of trend were not included for noncancer effects in the NTP (1998,
042076) study and thus were not added to Table 5-1. However, the global goodness-of-fit p-values for
each dose-response model fit to the data for each individual endpoint were included in the modeling
results in Appendix B. Discussion of in situ metabolism was included in Section 5.2, specifically as it
A-6
-------
relates to how chloroprene, as a water insoluble and nonreactive gas, can exert effects in the upper and
lower respiratory tract through blood-borne distribution. See the EPA Response to Comment 1 of
Charge Question 6 (pages A-8 and A-9) regarding the selection of the critical effect(s).
Charge Question 6. Benchmark dose (BMD) modeling was used to define the POD for the derivation
of the RfC. The POD was based on increased incidence of degenerative nasal lesions in male rats at a
benchmark response (BMR) of 10% extra risk. Has the BMD approach been appropriately conducted?
Is the BMR selected for use in deriving the POD (i.e., 10% extra risk of degenerative nasal lesions of
less than moderate severity) scientifically justified? Please identify and provide the rationale 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 EPAs approach.
Comment 1: All six reviewers commented that the use of BMD modeling was appropriate to define
the POD for derivation of the RfC. Four of the reviewers specifically commented that the BMD
approach was justified given a number of reasons, particularly that the database appears sufficiently
robust and that BMD modeling is preferred because it takes into consideration all of the dose-response
data and is less impacted by group size. One reviewer commented that use of a PBPK model could
clarify the saturation of metabolism into active metabolites and that this could facilitate dose-response
modeling and lead to a lower POD. One reviewer commented that selection of a BMR of 10% extra
risk was appropriate for degenerative nasal lesions, whereas four reviewers commented that a BMR of
10% was too high. Specifically, one reviewer noted that the NTP study did not identify a NOAEL and
that the severity of nasal lesions seen in the lowest exposure group was greater than minimal. These
four reviewers suggested that a lower BMR be selected for modeling purposes, and specifically
suggested BMRs in the range of 2-5% extra risk. One reviewer noted that, because severity data was
available for individual animals, EPA's categorical regression (CatReg) software could be used to
incorporate severity into the modeling scheme. Two reviewers commented that EPA could provide
more clarification in regard to the derivation of the RfC, and suggested that EPA provide a clear
indication of how and why particular models were selected for the various endpoints and provide a
step-by-step derivation of the RfC in the document.
Five reviewers commented that justification for treating chloroprene as a Category 1 gas and
the impact this had on dosimetric adjustments was not sufficiently justified in the document and that
further justification should be added. Two reviewers specifically indicated that chloroprene should be
classified as a Category 3 gas with regards to the application of a DAF. One reviewer objected
strongly to the approach used to derive the PODHEc values for a number of reasons. First, this reviewer
stated that the PODs used in the calculation of the HEC values are very similar (2.1 - 8.3 mg/m3) and
that nasal lesions were chosen only because the dosimetric adjustment factor (DAF) for nasal effects
was so low. Thus, in the reviewer's opinion, the selection of the nasal lesions as the critical effect was
an artifact of the DAF (RGDR) calculation and not based on the primary experimental observations.
A-7
-------
The reviewer then delineated their concerns relative to the RGDR, stating that the RGDR calculation
was theoretically flawed and discordant with the inhalation dosimetry database. This reviewer also
objected the conclusion that air-borne, rather than blood-borne, chloroprene induces nasal lesions,
stating that it was confusing why a discussion of portal-of-entry effects versus systemic redistribution
was discussed for cancer effects, but not for noncancer effects. This reviewer ultimately provided an
alternative scheme for RfC development: selection of the critical effect based on a POD of a parameter
closer to the observed data (i.e., PODAoj) and then applying the DAF calculation (both portal-of-entry
and systemic for respiratory effects, similar to what was done for cancer effects) to arrive at the HEC.
Response: The global BMD modeling approach was maintained in the document where possible (i.e.,
all endpoints that were considered for the critical effect that were amenable to BMD modeling were
modeled using the 2.1.1 version of BMDS software). When endpoints were not amenable to BMD
modeling, or no adequate model fit could be obtained, the NOAEL/LOAEL approach was used. A
PBPK model was not used in the modeling scheme due to limitations in the currently available, peer-
reviewed model (Himmelstein et al., 2004, 625154). A more detailed discussion of the current PBPK
model for chloroprene was included in Section 3.5 and covers the model structure, the metabolic and
physiological parameters used, and limitations that preclude its use in the Toxicological Review.
The selection of appropriate BMRs for endpoints under consideration for the critical effect was
modified. A BMR of 10% extra risk was used initially. In addition to reporting the incidence of the
endpoints, the NTP (1998, 042076) study also reported the severity scores for individual animals in
each dose group, thus making it possible to determine whether the endpoints were increasing in
severity as well as incidence with dose (Table B-l). For some endpoints (i.e., olfactory atrophy and
necrosis) that progressed in incidence as well as severity (i.e., progression from none or minimal to
mild to moderate lesions) from the control dose to the lowest dose, the majority of the External Peer
Reviewers recommended or indicated that a BMR of 5% or less would be appropriate for derivation of
the POD. Due to the nature and severity of the nasal degenerative effects (i.e., olfactory atrophy and
necrosis), and the proximity of the BMDLio values to the observed LOAEL compared to other
endpoints (Table 5-2), a BMR of 5% was considered to be appropriate for these olfactory endpoints.
The nature of the observed nasal lesions potentially included the loss of Bowman's glands and
olfactory axons in more severe cases. Effects that occur in the underlying lamina propria and basal
layer of the olfactory epithelium may be indicative of more marked nasal tissue injury. For all other
endpoints, a BMR of 10% was chosen as the response level (Section 5.2.2). CatReg software was not
utilized in the modeling scheme due to considerable uncertainty in assigning consistent severity scores
to multiple lesions across organ systems.
Additional discussion regarding how the modeling was performed and how and why particular
models were selected for each endpoint is included in the text (Section 5.2.2). Specifically, the criteria
that were used to determine adequacy of model fit (global goodness-of-fit p-value, %2 residuals, and
visual inspection) are discussed, as well as how the EPA chose the best model when multiple models
A-8
-------
appropriately fit the dose-response data for an individual endpoint (i.e., AIC when no model
dependence is assumed, and BMDL otherwise). EPA has also added step-by-step calculations of the
PODAoj and PODHEc values as well as the final RfC calculation in order to improve clarity in the
methods of RfC derivation.
Additional discussion was added to Section 5.2.3 covering the physio-chemical properties of
chloroprene as they relate to the observed pattern of effects. The current RfC methodology (U.S. EPA,
1994, 006488) attempts to group chemicals into one of three discrete categories based on their physio-
chemical properties and presumed toxicokinetics; using this scheme, chloroprene would be best
classified as a Category 3 gas, being relatively water insoluble and nonreactive, and would be expected
to elicit extrarespiratory effects. This classification is consistent with what is proposed for the mode of
action of chloroprene: conversion of the parent compound into its epoxide metabolite via P450 isoform
CYP2E1. Since CYP2E1 is expressed in both the olfactory and pulmonary regions of the respiratory
tract, in situ metabolism in the respiratory tract may explain a portion of the biological activity of
chloroprene in these regions. However, because of the high potential for blood-borne delivery, as
evidenced by chloroprene's low water solubility, low reactivity and ability to cause systemic effects,
EPA agrees that, in accordance with RfC methods (U.S. EPA, 1994, 006488), chloroprene is most
appropriately treated as a Category 3 gas for the derivation of noncancer and cancer HEC values. The
consistent treatment of chloroprene as a Category 3 gas in the Toxicological Review, as well as
additional discussion of the uncertainty surrounding the mode of delivery to respiratory tissues (e.g., in
Section 5.3), clarifies the noncancer and cancer derivations.
Given the above changes to the modeling scheme, increased incidence of olfactory atrophy,
alveolar hyperplasia, and splenic hematopoietic proliferation in male rats, female rats, and female
mice, respectively, were chosen as the co-critical effects. For these endpoints, after rounding to one
significant figure, the PODAoj resulted in a value of 2 mg/m3 (Section 5.2.4). Using a DAF of 1 (for
systemic effects), the calculated PODnEc was 2 mg/m3.
Charge Question 7. Please comment on the rationale for the selection of the uncertainty factors (UFs)
applied to the POD for the derivation of the RfC. If changes to the selected UFs are proposed, please
identify and provide a rationale(s).
Comment 1: Six reviewers commented that the selection of the uncertainty factors, 10 for human
variation, 3 for animal-to-human extrapolation, and 3 for database deficiencies were reasonable and
consistent with EPA policy. One reviewer commented that application of the threefold database
uncertainty could be the source of some contention, in that it seemed justified considering the absence
of a two-generational reproductive study, but that negative findings for teratogenesis and dominant
lethal effects could be considered an adequate substitute. One reviewer commented that a multi-
generational study was available and should be discussed in regard to the selection of the database
uncertainty factor. One reviewer noted the lack of data on potential neurodevelopmental toxicity or
A-9
-------
long-term effects following perinatal exposure. One reviewer suggested discussion of the uncertainty
surrounding application of the DAFs for effects resulting from airborne delivery (i.e., portal-of-entry
effects) should be discussed. Two reviewers commented that there is probably considerable human
variability in the metabolism of chloroprene due to genetic polymorphisms in the genes coding
metabolizing enzymes and the activity of enzymes. One reviewer suggested that an additional
uncertainty factor of 3-10 be added if the RfC was derived from a BMDLio in the presence of
moderately severe lesions in the low dose.
Response: The current selection and application of uncertainty factors was maintained in the
document (Section 5.2.4). A two-generational reproductive study was not available in the database for
chloroprene. The Appelman and Dreef van der Meulen (1979, 064938) study was an unpublished
report in which FO and Fl rats were exposed to chloroprene. However, this study did not involve the
mating of the Fl generation, so developmental effects to the F2 generation could not be assessed. Lack
of a developmental neurotoxicity study was not considered a sufficient reason to increase the database
uncertainty factor, as there was limited data indicating the neurotoxic or developmental effects of
chloroprene. Therefore, EPA concluded that the application of a database uncertainty factor of 3 be
retained for deriving the RfC. A discussion of the uncertainty surrounding the application of the
default DAFs for portal-of-entry effects was included in Section 5.3 (Uncertainties in the Inhalation
Reference Concentration), but not in the section outlining the application of the actual uncertainty
factors (Section 5.2.4). A concise discussion of the observed variation in CYP2E1 in human
populations was included in Section 5.2.4 supporting the human variation uncertainty factor of 10.
The uncertainty factor of 10 was maintained as it was presumed to account for variations in
susceptibility within the human population. An additional uncertainty factor to account for derivation
of an RfC based on a BMR of 10% for effects showing an increase in severity in the low dose was not
supported by EPA guidance covering uncertainty factors.
A.l.2.3. Carcinogenicity of Chloroprene
Charge Question 8. Under the EPA's 2005 Guidelines for Carcinogen Risk Assessment (2005,
086237) the Agency concluded that chloroprene is likely to be carcinogenic to humans by all routes of
exposure. Please comment on the cancer weight of evidence characterization. Is the cancer weight of
evidence characterization scientifically justified?
Comment 1: Six reviewers commented that the characterization of chloroprene as "likely to be
carcinogenic to humans'" was appropriate and clearly justified based on the animal and genotoxicity
data. Three reviewers commented that the animal data provided ample evidence of carcinogenesis in
both sexes of two rodent species (mouse and rat) at multiple organ sites, many of which were distal to
the point-of-contact. One reviewer commented that there was clear information on the formation of
A-10
-------
mutagenic metabolites of chloroprene and analogies to related chemical carcinogens with analogous
metabolic pathways that made the determination of likely to be carcinogenic" unequivocal. One
reviewer commented that chloroprene was likely to be carcinogenic by all routes of exposure because
its carcinogenicity is likely due to formation of epoxide metabolites, and because P450-mediated
epoxidation of chloroprene can occur in several organs. Another reviewer noted that if there is a
critical role for blood-borne chloroprene, as was assumed for the induction of pulmonary neoplasms,
the possibility of carcinogenicity from multiple routes of exposure is elevated.
One reviewer commented that the mode of action for chloroprene is such that it may not be
carcinogenic via dermal exposure as the parent compound is nonreactive and insoluble in water. One
reviewer noted that there were potential increases in liver tumors in occupationally exposed cohorts
that supported the determination that chloroprene may represent a carcinogenic hazard to humans.
Two reviewers suggested that the strength of the epidemiological data was sufficient to change the
descriptor to "carcinogenic to humans," with one reviewer citing the multiple tumor responses in
animals, the metabolic activation of chloroprene by rat, mouse, and human liver microsomes, the
finding of K-ras mutations in lung neoplasms in mice, and the relatively consistent finding of
increased risk of liver cancer mortality in occupational cohorts. This reviewer felt that the EPA did not
sufficiently justify the "likely to be carcinogenic" over "carcinogenic" descriptor given that many of
the limitations in the epidemiology database (healthy worker effect, etc.) result in underestimations of
risk. This reviewer also commented that EPA's cancer guidelines allow for the determination of
"carcinogenic'" when there is less than convincing epidemiologic evidence, but there is strong animal
carcinogenicity and when the mode of action identified in animals is anticipated to occur in humans.
One reviewer commented that, while the animal and genotoxicity data backed up the current
cancer determination, the epidemiology data did not support that determination and was overstated in
the document. This reviewer commented that the document reported on the evidence of dose-response
for liver cancer in the Marsh et al. (2007, 625188) study, but did not provide the relative risks (and
confidence limits) in each of the exposure categories. This reviewer also commented that the EPA
misrepresented the evidence regarding the presence of dose-response trends in other studies -
responses in the low and high exposure groups are not statistically different (Bulbulyan et al., 1999,
157419), and there is no dose response for liver cancers in the high dose because only one cancer case
was liver cancer (the remaining two cancers were of the gall bladder) (Leet and Selevan, 1982,
094970). This reviewer also commented that known risk factors for liver cancer (hepatitis infection,
alcohol consumption, etc.) were not discussed in sufficient detail given the level of discussion included
for risk factors for lung cancer. This reviewer commented further that discussion of co-exposures and
potential confounding was inadequate. The reviewer provided a list of suggestions in order to increase
the transparency of the presentation of the data on liver cancer in humans, including: discussion on
whether the cohorts that studies investigated (i.e., the Louisville Works cohort investigated by Leet and
Selevan (1982, 094970) and Marsh et al. (2007, 625188) were adequately independent; more complete
presentation of results from Marsh et al. (2007, 625188): and increased discussion regarding the
A-11
-------
variability around central effect measurements based on small numbers of cases in the Bulbulyan et al.
(1998, 625105: 1999, 157419). Li et al. (1989, 625181). and Leet and Selevan (1982, 094970). Lastly,
this reviewer commented that, given the various study limitations in the studies that observed increased
incidence of liver cancer mortality, it is unclear whether an association exists between chloroprene
exposure and liver cancer, especially considered that the best conducted study, Marsh et al. (2007,
625188), failed to observed an increased risk.
Response: The determination that chloroprene is "likely to be carcinogenic to humans" by all routes of
exposure was maintained in the document based on a weight of evidence approach that considered
human epidemiology, animal toxicology, and genotoxicity data (Section 4.7). U.S. EPA's Guidelines
for Carcinogen Risk Assessment (2005, 086237) indicate that for tumors occurring at a site other than
the initial point of contact, the weight of evidence for carcinogenic potential may apply to all routes of
exposure that have not been adequately tested at sufficient doses. An exception occurs when there is
convincing toxicokinetic data that absorption does not occur by other routes. Although there are no
recent toxicity studies involving dermal exposure, carcinogenicity by this route of exposure may be
inferred as there is no convincing toxicokinetic data to preclude absorption by this route of exposure,
and that rapid absorption of chloroprene through the skin occurs (HSDB, 2009, 594343: NIOSH, 1977,
644450: NIOSH, 1995, 644453).
Although there was evidence of increased risk of liver cancer mortality in occupational cohort
studies, EPA concluded that the strength of evidence did not support the cancer descriptor of
"carcinogenic.'" In order for a chemical to be found to be "carcinogenic,'" there either must be
convincing epidemiologic evidence of a causal association or a lesser weight of epidemiologic
evidence that is strengthened by all of the following: (1) strong evidence of an association between
human exposure and cancer, (2) there is extensive evidence of carcinogenicity in animals, (3) the mode
of action has been identified in animals, and (4) the key precursor events that precede the cancer
response in animals are anticipated to occur in humans. EPA: (1) demonstrated throughout the
document that there exists unequivocal evidence of carcinogenicity in animals, (2) provided a plausible
mode of action based on animal and human in vitro metabolic and toxicokinetic studies, and (3)
discussed that the precursor events that occur in animals are reasonably anticipated to occur in humans.
However, EPA concluded that the epidemiologic data, while providing a fairly consistent evidence of
liver cancer mortality (4 studies report statistically significant associations in 4 separate cohorts), did
not support changing the cancer determination to "carcinogenic." This was due to methodological
limitations of the occupational epidemiology studies (e.g., no available data for some potential
confounders which precluded adjustment, limited statistical power due to small sample sizes, and lack
of precise quantitative exposure ascertainment) that made it difficult to draw firm conclusions
regarding the findings of these studies. The most recent and comprehensive studies (Marsh et al.,
2007, 625187: Marsh et al., 2007, 625188) used quantitative exposure ascertainment, and failed to
observe statistically significant relationships between exposure and outcome. These findings did not
A-12
-------
diminish the observations of the four studies that did observe statistically significant associations, but
rather indicated that the epidemiologic database is somewhat equivocal, and did not support changing
the cancer determination from "likely to be carcinogenic."
Additional text regarding the relative risks and confidence limits for each of the exposure
categories for liver cancer in the Louisville cohort from Marsh et al. (2007, 625188) was added to the
document in Section 4.1.1.2. A more thorough discussion of the suggested dose-response relationships
observed in (Bulbulyan et al., 1999, 157419)) and Leet and Selevan (Leet and Selevan, 1982, 094970)
was added to Section 4.7.1.1. This discussion highlights issues surrounding the determination that
there exists a suggestive dose-response relationship in these two studies, even though the responses in
the two exposure categories are not statistically significantly different from one another ((Bulbulyan et
al., 1999, 157419)) and that a dose response only exists when liver and biliary/gall bladder cancers are
grouped together (Leet and Selevan, 1982, 094970). Additional text and discussion was added
throughout the document regarding known risk factors for liver cancer (including hepatitis B infection,
alcohol consumption, and aflatoxin ingestion), and the lack of control for these factors in the
epidemiologic studies observing a statistically significant association between liver cancer and
occupational exposure to chloroprene. Also, a more complete discussion regarding potential co-
exposures to industrial chemicals and the possibility of confounding was added to numerous sections
of the Toxicological Review. A complete evaluation of the independence of the Leet and Selevan
(1982, 094970) and Marsh et al. (2007, 625188) studies was added to Section 4.7.1.1. This evaluation
highlights differences in the methodologies employed by the two studies as well as differences in the
demographics of the sub-sets of the Louisville cohort that were investigated in the studies. EPA
concluded that there exist sufficient differences between these two studies investigating the Louisville
cohort to warrant the independent analysis of each. Additional text was added to Section 4.7.1.1
regarding the variability of the central effect measures based on low expected counts used for liver and
lung cancer mortality in (Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Li et al.,
1989.625181).
Additional text and discussion was added throughout the Toxicological Review regarding
individual study limitations in those studies that observe a statistically significant association between
chloroprene exposure and increased liver cancer mortality. Although limitations exist in these studies,
EPA carefully considered and concluded there is evidence of an association between liver cancer risk
and occupational exposure to chloroprene based on the observation of increased liver cancer mortality
across multiple studies investigating the outcome in heterogeneous populations and exposure
scenarios. This conclusion was based on a consistent two- to more than fivefold increase in risk of
liver cancer mortality in the SMRs observed among these studies. Although no statistically significant
increase in risk of liver cancer was detected in the most recent and comprehensive cohort study
involving workers at four plants (Marsh et al., 2007, 625188), the observed RR increased with
increasing cumulative exposure in the plant with the highest exposure levels, indicating a dose-
response trend. Limitations in the existing epidemiological database included: the lack of information
A-13
-------
on individual workers' habits (i.e., alcohol consumption); the lack of control for potential confounding;
incomplete enumeration of incidence and mortality cases; and potential for biases that may lead to an
underestimation of the risk (e.g., the healthy worker effect). These limitations are further discussed in
Section 4.7.1.1.
Charge Question 9. A 2-year inhalation cancer bioassay in B6C3Fi mice (NTP, 1998, 042076) was
selected as the basis for derivation of an inhalation unit risk (IUR). Please comment on whether the
selection of this study for quantification is scientifically justified. Please identify and provide the
rationale for any other studies that should be selected as the basis for quantification.
Comment 1: Five reviewers commented that the selection of the NTP 2-year inhalation
carcinogenicity bioassay was scientifically justified based on the fact that the study was well-designed
and conducted, the study identified carcinogenic effects in multiple organ systems in rats and mice
exposed to a wide range of chloroprene concentrations, and the study was peer-reviewed. One
reviewer noted that a major strength of this study was the multiple histopathological reviews of lesions
identified in rats and mice. One reviewer commented that a stronger reason than presented in the draft
Toxicological Review for not selecting the Trochimowicz et al. (1998, 625008) study as the principal
study was the high mortality in the low dose animals due to the failure of the ventilation system. One
reviewer commented the dosimetry in terms of an active metabolite may be informed by the
application of a PBPK model. Two reviewers commented that inclusion of lung tumors observed in
mice may be problematic due to greatly increased metabolic activation rate in mice compared to
humans or rats and one of these reviewers commented that a discussion of this should be included in
the document. One reviewer did not comment on the choice of the NTP (1998, 042076) study as
justified, but commented that selection of the mouse as the most appropriate species over the rat was
not adequately explained.
Response: Choice of the NTP (1998, 042076) 2-year inhalation carcinogenicity bioassay as the basis
for derivation of an inhalation unit risk was maintained. Text was added to the document clarifying the
reasons the Trochimowicz et al. (1998, 625008) study was not chosen for selection as the principal
study; the high mortality in the low dose group was identified as the main reason for not selecting the
study as the principal study (Section 5.2.1). A more thorough discussion of the current PBPK model,
including its inadequacies relevant to use in the current Toxicological Review, was included in Section
3.5. Specifically, the current PBPK model was concluded to be inadequate for use to inform dosimetry
in terms of an active metabolite. A more complete and detailed discussion of metabolism and
toxicokinetic differences between species (Himmelstein et al., 2004, 625152; Himmelstein et al., 2004,
625154) was added to Section 3.3, to indicate that differences in epoxide production in the lungs of
mice and humans are not 50-fold, but may be as little as 2- to 10-fold. These additional data also
indicated that in some cases (i.e., glutathione transferase activity) detoxification of the epoxide
A-14
-------
metabolite may be faster in mice than humans. Additionally, the evidence for further oxidation of
(l-chloroethenyl)oxirane in mice, but not in humans, rats, or hamsters was characterized. The mouse
was chosen over the rat as the most appropriate species for the inhalation unit risk derivation based on
the observation that it was more sensitive to the carcinogenic effects of chloroprene exposure.
Charge Question 10. Amutagenic mode of carcinogenic action is proposed for chloroprene. Please
comment on whether the weight of evidence supports this conclusion. Please comment on whether this
determination is scientifically justified. Please comment on data available for chloroprene that may
support an alternative mode(s) of action.
Comment 1: Six reviewers commented that a mutagenic mode of carcinogenic action for chloroprene
was appropriate based on the evidence that chloroprene metabolism operates via P450-mediated
oxidation to a DNA-reactive epoxide metabolite, which is mutagenic in multiple strains of Salmonella,
and the observation of K- and H-ras mutations in tumors obtained from mice exposed to chloroprene.
One reviewer specifically noted that the proposed mode of action was consistent with other epoxide-
forming carcinogens (i.e., 1,3-butadiene). Three reviewers commented that they were not aware of any
scientific data that would support an alternative mode of action. One reviewer commented that while a
mutagenic mode of action may not be the only mode of action, it was clearly one possibility. One
reviewer commented that if it were concluded that a metabolite represented the ultimate toxic species,
the quantitative risk assessment should be discussed in regard to the large differences observed
between mice, rats, and humans.
Response: The proposed mutagenic mode of carcinogenic action for chloroprene was maintained in
the document. A more complete discussion of the metabolic and toxicokinetic differences between
mice, rats, and humans was included in Section 3.3.
Charge Question 11. Data on hemangiomas/hemangiosarcomas (in all organs) and tumors of the lung
(bronchiolar/alveolar adenomas and carcinomas), forestomach, Harderian gland (adenomas and
carcinomas), kidney (adenomas), skin and mesentery, mammary gland and liver in B6C3Fi mice were
used to estimate the inhalation unit risk. Please comment on the scientific justification and
transparency of this analysis. Has the modeling approach been appropriately conducted? Please
identify and provide the rationale for any alternative approaches for the determination of the inhalation
unit risk and discuss whether such approaches are preferred to EPAs approach.
Comment 1: Two reviewers supported the use of a dose-response model which accounted for
differences in survival such as the multistage-Weibull model. One of these reviewers suggested an
alternative modeling approach whereby the assumption of saturating metabolism was incorporated in
the model structure, and provided an extensive example using the mice data. The other reviewers did
A-15
-------
not comment on the dose-response model specifically, with one of these commenting only that the
derivation of the inhalation unit risk could be made clearer in the text.
Four reviewers commented that the scientific justification of combining unit risks for all tumor
types was scientifically justified and conducted, with one noting further, that basing the unit risk
derivation on one tumor type would underestimate the carcinogenic potential of chloroprene. One of
these reviewers suggested further that the results of the animal study should be evaluated to determine
if there are genetic or other factors between animals that determine which animals get one tumor
versus those that get more than one tumor type.
One reviewer commented that the quantitative importance of the mouse lung tumors was
questionable given the differences in metabolic activation between mice and humans. One reviewer
commented that a discussion of site concordance/discordance between mice and humans, and human
relevance of observed rodent tumors, should be included in the document. Two reviewers commented
that a useful analysis would be to compare the unit risk calculated from the animal study to unit risks
calculated from the human epidemiology studies, with one reviewer specifically suggesting that the
Marsh et al. (2007, 625187: 2007, 625188) Louisville cohort be used because it has the most
quantitative exposure information. The other reviewer asked whether it was possible to project human
occupational risks from the unit risk to consider consistency with epidemiologic observations.
A reviewer also commented that discussion should be included as to why an uncertainty factor
for human variability (other than the application of the ADAFs) was not applied to the cancer risk
estimate.
Response: The assessment's cancer risk modeling approach, use of a time-to-tumor model and
subsequent estimation of a composite unit risk for all tumor types in female mice, was maintained and
more thoroughly explained and discussed in the document (Sections 5.4.3 and 5.4.4). The suggested
alternative modeling approach incorporating saturating metabolism appears useful, and is similar to a
model now widely available to BMDS users (the dichotomous Hill model); however, for this situation
it did not appear to have sufficient advantages over the approach EPA used. As noted by the peer
reviewers, this alternative model did not incorporate time-to-tumor information, which they supported
including. Also, the saturating metabolism parameters were not derived from pharmacokinetic data but
from empirical fits to the dose and tumor incidence data. So the alternative model was as much an
empirical model as the multistage-Weibull model. Further, the saturating behavior observed, especially
at the two higher doses, reflected to a large degree the limiting condition that only 100% of the animals
can develop tumors. Thus the saturating behavior could not be attributed solely to metabolism
processes. The multistage-Weibull model did adequately fit the monotonic, supralinear dose-response
relationships seen in the NTP study, and EPA retains the original analysis in the assessment.
Additional mouse and human metabolic and toxicokinetic data (Himmelstein et al., 2004,
625152: Himmelstein et al., 2004, 625154) added to the document indicated that the metabolic
differences between humans and mice are not as great as previously represented in the document
A-16
-------
(Section 3.3). Therefore, the mouse lung tumor data was considered relevant for human risk
estimation and was retained in the modeling approach. The composite risk analysis addressed the risk
of developing any combination of tumors in animals, in order to estimate the risk of developing any
combination of tumors in humans. It was reasonable to assume, given the observed multi-site
carcinogenicity of chloroprene, that induction in tissues specific to humans is possible.
It was unclear what the reviewer was suggesting in regard to evaluating genetic factors that
may influence which animals get more than one tumor type. Given that the animal species used in the
2-year cancer bioassay was an inbred strain of mouse and that all conditions except exposure
concentration were maintained across dose groups, it is unlikely that genetic or other factors other than
dose influenced whether an animal developed one or multiple tumors.
One reviewer suggested and another reviewer concurred that a comparison of the inhalation
unit risk estimates derived in this Toxicological Review (Table 5-7) to unit risks calculated from
human epidemiology studies should be conducted. EPA maintains that unit risk estimates could not be
derived from human epidemiology studies because the available quantitative exposure assessments
were not sufficient for this purpose. However, a comparison of the number of cancer cases predicted
by the mice tumors with those observed in the study with the most thorough exposure assessment (the
Marsh et al. Louisville cohort) was considered in a sensitivity analysis context. Briefly, the unit risk
for composite cancer risk derived from male mice (1.4 x 10"4 per ug/m3) was applied to the median
cumulative exposure for the Louisville plant, converted to a lifetime equivalent continuous
concentration (18.35 ppm-yr/70 yr x 3.62 x 103 (ug/m3)/ppm ~ 950 ug/m3), yielding an upper bound
predicted risk of 0.13 for composite cancer risk. When this risk estimate is applied to the 2282
subjects with known cause of death, the predicted upper bound on the number of cancer cases is -300.
In Louisville, 266 + 17 = 283 deaths due to either respiratory or liver cancer—the cancers of a priori
concern—were reported. Note that the unit risk is an upper bound estimate, and also includes incident
cases as well as deaths.
For the above quantitative comparison, several considerations must be acknowledged with
regards to interpretation of the results. These considerations are: (1) the quantitative exposure
assessment (i.e., cumulative of chloroprene exposure) for the Louisville cohort spanned approximately
3 orders of magnitude; (2) insufficient information regarding whether sufficient latency for subjects to
develop cancer existed; (3) exposure estimates were for the full cohort and likely not applicable to the
subset (with known cause of death) equally well; (4) concerns already elaborated in the Toxicological
Review regarding incomplete ascertainment of incident cases and other deaths possibly involving
cancer; and (5) a quantitative comparison could only be made for Marsh et al. studies (2007, 625187;
2007, 625188) because of the partial availability of exposure information, and not for the additional
epidemiological studies that observed significant associations between chloroprene exposure and
cancer mortality (Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419: Leet and Selevan,
1982, 094970; Li et al., 1989, 625181). Given these considerations, the comparison carried out here
does not demonstrate a striking disagreement between the animal and human data.
A-17
-------
EPA has not developed an Agency-wide policy to apply uncertainty factors to cancer risk
estimates. Therefore, no uncertainty factor to take into account human variability was applied to the
inhalation unit risk.
Charge Question 12. Lung tumors have been alternatively treated as systemic or portal-of-entry
effects in the modeling of cancer endpoints. Please comment on the scientific justification for this
modeling approach. Please comment on whether the rationale for this decision has been transparently
and objectively described. Please comment on data available for chloroprene that may support an
alternative method for modeling the observed lung tumors in mice.
Comment 1: Four reviewers did not object to alternatively treating lung tumors as portal-of-entry or
systemic effects, noting the absence of data suggesting which route of exposure is more relevant to the
carcinogenic effects of chloroprene. However, three of these reviewers also noted that the application
of this approach was not sufficiently discussed in the Toxicological Review and that the text should
provide more elaboration in that regard. One reviewer commented that lung tumors for both male and
female mice appeared to be compatible with systemic saturable metabolic activation and therefore lung
tumors should not be treated as portal-of-entry effects. One reviewer commented that treating
chloroprene-induced lung tumors as either portal-of-entry or systemic effects would be appropriate
given the lack of information only if chloroprene were a gas expected to elicit portal-of-entry effects.
However, this reviewer further commented that the justification for treating chloroprene as a
Category 1 gas and the impact this had on dosimetric adjustments was not sufficiently justified in the
document and that further justification should be added. This reviewer suggested that chloroprene is a
Category 3 gas (i.e., a nonreactive gas expected to elicit its toxicity systemically) and that the DAF
should equal 1 for all observed tumor types. Finally, this reviewer noted that the pattern of respiratory
injury is suggestive of local metabolic activation but that it was possible that active metabolites are
formed in and then escape the liver.
Response: The current modeling approach of treating observed lung tumors as systemic lesions was
maintained in the Toxicological Review. However, several reviewers commented in response to
Charge Question 6 and in response to this charge question (12) that chloroprene is most appropriately
treated as a Category 3 gas. Consistent with the approach employed for the derivation of the RfC in
which noncancer effects due to exposure to chloroprene were evaluated as systemic effects (Section
5.2), likewise the quantitative approach for the cancer assessment has retained the risk estimates based
upon the tumors resulting from systemic distribution of chloroprene. Discussion regarding the
justification for and application of this approach as it relates to the observed pattern of effects was
added to the document (Sections 5.2.3 and 5.4.3; see response to Charge Question 6. comments above
as well). Chloroprene is a water insoluble, nonreactive chemical, and is expected to be absorbed into
the bloodstream deep in the respiratory tract and exert its toxic effect systemically. Indeed, multiple
A-18
-------
effects were observed distal to the respiratory tract which supports this assumption. This discussion is
consistent with the reviewer's comments that the pattern of respiratory injury is suggestive of local
metabolic activation, but that systemically distributed metabolites may be a factor in the observed
carcinogenicity of chloroprene.
Charge Question 13. An oral slope factor (OSF) for cancer was not derived for chloroprene. Is the
determination that the available data for chloroprene do not support derivation of an OSF scientifically
justified?
Comment 1: Five reviewers commented that the determination that there are no available data to
support derivation of an oral slope factor for chloroprene was appropriate. One reviewer commented
that an appropriate PBPK model would allow for a route-to-route extrapolation. One reviewer noted
that the current PBPK model did not seem to be adequate to allow for route-to-route extrapolation.
One reviewer commented that the lack of information on disposition of chloroprene, including the
AUC for the DNA-reactive epoxide metabolite, after oral exposure, did not support a route-to-route
exposure. This reviewer noted that a likely large first-pass liver effect after oral exposure could
significantly alter the systemic distribution of chloroprene and its metabolites compared to inhalation
exposures.
Response: The determination that the chloroprene database did not support the derivation of an oral
slope factor was maintained in the Toxicological Review (Section 5.4.4). A more complete discussion
of the current PBPK model (Himmelstein et al., 2004, 625154), including its strengths and weaknesses
for use in a route-to-route extrapolation in the current assessment, was included in Section 3.3.
A.2. SPECIFIC COMMENTS
This section contains specific comments received from the external peer reviewers and has been
organized so that comments and responses appear sequentially as they relate to the Toxicological
Review.
Comment 1: The data on partition coefficient should be discussed more completely. It is possible to
infer information on tissue distribution from such data. It is also possible to make inferences on
regional respiratory tract absorption from these numbers. A vapor with a blood:air partition coefficient
less than 10 is not likely to be scrubbed efficiently from the airstream in the upper airways.
Response: Additional language was added to the document regarding the partition coefficients for
chloroprene and what inferences could be made regarding the magnitude of those partition coefficients
(Sections 3.2 and 3.4).
A-19
-------
Comment 2: More detail should be provided on the metabolism kinetics for chloroprene. The
information on elucidation of putative metabolites was clear and concise, but the data on kinetics was
incompletely presented data and was very difficult to interpret fully. The meaning of the metabolic and
toxicokinetic data, particularly with respect to rodent-human extrapolations, should be synthesized into
a coherent explanation of species differences in response. Specific areas that need more attention
include species differences in glutathione conjugation with respect to (l-chloroethenyl)oxirane
detoxification and differences in chloroprene clearance among species. Factors that can influence the
clearance of chloroprene include fatair partition coefficients and percentage of body weight as fat.
Response: Extensive additional text regarding the metabolism of chloroprene and the toxicokinetic
differences that exist among species (Himmelstein et al., 2004, 625152; Himmelstein et al., 2004,
625154) was added to section 3.3. These additional discussions indicate that differences in epoxide
production in the lungs of mice and humans are not 50-fold, but may be as little as 2- to 10-fold. These
additional data also indicate that in some cases (i.e., glutathione transferase activity) detoxification of
the epoxide metabolite may be faster in mice than humans. Additionally, there appears to be an
additional step to the detoxification pathway, oxidation of (l-chloroethenyl)oxirane, active in mice, but
not in humans, rats, or hamsters. A discussion of fatair partition coefficients and body fat percentage
was added to the document.
Comment 3: The text in Section 3.3 should precisely indicate how the estimates for Vmax/Km; reported
in Tables 3-4 and 3-5, for lung metabolism were obtained. The mouse-human comparison for lung
metabolism is a particularly important subject; this is a fact that was not adequately considered in the
risk evaluation.
Response: Additional text was added to Section 3.3 clarifying how the estimates of Vmas/Km were
calculated. A detailed discussion of chloroprene metabolism in the mouse and human lung was also
added to the document, as well as extensive discussion on how these differences impacted the risk
evaluation.
Comment 4: The meaning of the ranges given for Vmax/Km for the oxidation of chloroprene should be
described. If these were in fact the ranges of all observations, then the number of observations should
be given.
Response: The ranges previously given in the text were removed and presented only in the
corresponding tables. The ranges that were given were the ranges of values observed across the
species investigated. These values were calculated from pooled microsomal preparations, authors did
not report the number of observations made.
A-20
-------
Comment 5: In Table 3-2, results should be expressed as fraction of total metabolites rather than
relative to butanol standard. Or it could be expressed in terms of absolute rates per unit time per unit
microsomal protein.
Response: The authors of the study (2001, 019012) reported the formation of (l-chloroethenyl)oxirane
relative to butanol standard, and did not present data on the formation of total metabolites or on
absolute rates per unit time per unit microsomal protein. Therefore, reporting the formation of
(l-chloroethenyl)oxirane formation relative to butanol standard was maintained in the document.
Comment 6: Presentation of metabolic data in Table 3-4 was inadequate. No error bars or statements
of how many animals tested independently (or pooled?), or more crucially, how many humans and how
they differ in Vmax/Km for various organs.
Response: The data presented in Table 3-4 is how the data was presented by the authors in the original
reference (Himmelstein et al., 2004, 625152). Additional text was added indicating the results were
from pooled microsomal preparations, and how many human samples were pooled. No other
information was available for human variability in Vmax/Kmin other organ systems.
Comment 7: Values for the major physiological parameters (body weight, cardiac output, and alveolar
ventilation) should be provided.
Response: Those values were added to Table 3-9.
Comment 8: While suitable discussions of the epidemiological data regarding the healthy worker
effect were included in the document, there were no suitable caveats for the "internal" comparisons by
mentioning the distortions expected from the healthy worker survivor effect — that longer exposed
workers with higher cumulative exposures have lower mortality than shorter term workers. This must
be incorporated into the analysis.
Response: The discussion of the healthy worker survivor effect was expanded in the document
(Section 4.1.1.3).
Comment 9: SMRs and SIRs should consistently use basei or base
100-
Response: The document was revised so that SMRs and SIRs consistently use baseioo throughout the
document.
A-21
-------
Comment 10: It would be useful if more information on occupational exposure levels would be
presented in the text. Information on exposure concentrations in addition to cumulative (ppm-year)
exposures would be useful.
Response: Information on the median average intensity of occupational chloroprene (in ppm) was
added to the text (Section 4.1.1.2).
Comment 11: The discussions of both liver and lung cancer would benefit from some attempt at
integrative meta-analysis, combining the effects of multiple studies for reasonably comparable levels
of exposure. This, however, likely depends on obtaining some disaggregated data from the individual
investigators.
Response: Performance of a meta-analysis on liver and lung cancer data was beyond the scope of this
document.
Comment 12: The document indicates that a limitation of the Li et al. (1989, 625181) paper was that
only three years of local area data were used to estimate the expected numbers of deaths which may
not be representative with regard to the period of follow-up of the cohort. An issue not considered is
the stability of the expected rates based on local data. Also, the discussion of how the calculated SMRs
would be biased if the local data for those three years was not representative of the entire period of
follow-up is not clear.
Response: A discussion of the stability of the results reported by studies using low expected counts of
cancer mortality was added to the document (Section 4.7.1.1). Also, the text regarding how the SMRs
may be biased due to the potential nonrepresentativeness of the available local data was clarified.
Comment 13: In Colonna and Laydevant (2001, 625112), if there was any indication of how many
workers died or left the study area prior to 1979, this should be included in the document. Did the
authors have an idea of how much impact this would have on the results?
Response: No such data were available on how many workers left the study area prior to 1979.
Comment 14: It seems odd that of the 652 cancer cases in the Louisville facility, only 1 case was
unexposed (Table 4-8). This might suggest that a large percentage of individuals classified as exposed
were essentially unexposed. The document should provide greater emphasis on the potential impact of
exposure misclassifications.
A-22
-------
Response: The results in Table 4-8 reflect the analysis presented in Marsh et al. (2007, 625187). Text
was added to the document highlighting the small number of unexposed workers across the four
cohorts and limitations to the ability to draw conclusions based on the exposure classification approach
in Marsh et al. (2007, 625187)
Comment 15: It is not difficult to understand why Marsh et al. (2007, 625188) concluded that their
study provided no evidence of cancer risk associated with chloroprene exposures. Table 4-9 on page
4-14 shows little evidence of a dose response. It is inappropriate to conclude, as is done in lines 1-3
on page 4-15, that Marsh et al.'s (2007, 625188) explanations were "not entirely consistent with the
data presented." The authors of this document have chosen one interpretation; the authors of the study
have chosen another interpretation.
Response: The language regarding the interpretation of the Marsh et al.'s (2007, 625188) findings was
revised in the document. Also, discussion of Leonard et al. (2007, 625179) has been included that adds
to the weight of evidence that chloroprene exposure may be associated with cancer mortality,
especially when comparisons are based on internal populations or other regional/national DuPont
workers.
Comment 16: Some of the criticisms of the occupational cohort studies are too harsh. For example,
how often are causes of death verified by histological confirmation or review of medical records?
Incomplete enumeration of incident cases is a criticism that could be leveled at many incident studies.
The statement "that despite the lack of quantitative exposure information, occupational studies are still
able to contribute to the overall qualitative weight of the evidence considerations" states the obvious.
There are numerous examples of studies that have limited or no quantitative exposure information that
have nevertheless contributed to weight of evidence considerations.
Response: It is important to sometimes state the obvious for a broad audience so that readers that are
not experts in epidemiology understand that there is still valuable information that can be gleaned from
the epidemiology literature (i.e., with regard to lack of quantitative exposure information).
Comment 17: In Section 4.1.2.1, the statement "no workers experienced hair loss" is made. This is the
first place where loss of hair is mentioned. Since that is an unusual effect, it would be better to report
the results of the distillation workers after the results of the polymerization workers.
Response: The text was changed so that the results for distillation workers were presented after those
for the polymerization workers.
A-23
-------
Comment 18: For later modeling, EPA should report integrated average exposures that were measured,
rather than the nominal target exposures. The difference is small, as indicated in the discussion, but the
measurements should be used in preference to the target levels in the dose response modeling which
appears later in the document.
Response: The actual average exposure concentrations achieved in the NTP (1998, 042076) study
were added to the document (Section 4.2.2). However, the differences between the target and actual
chamber concentrations were very small. For the 2-year inhalation exposure, the greatest difference
observed between target and actual exposure concentration was 0.9% for rats in the 32 ppm exposure
group (target concentration of 32 ppm versus actual concentration of 31.7 ± 1.1 ppm). Therefore, it
was deemed unnecessary to redo the benchmark modeling with the actual exposure concentrations as
the difference in results would be negligible.
Comment 19: Clarity could be improved in the document if the following were included in the
document: with regard to Table 4-16, the magnitude of injury should be included (i.e., the average
severity score could be added parenthetically in each column); with regard to the lack of
histopathological damage in the lungs of mice in the 16-day study, the text should explicitly state as
such; with regard to the lack of nasal lesions in the respiratory mucosa of rats in the 13-week study, the
text should explicitly state as such (text should differentiate between effects, or lack thereof, observed
in the olfactory and respiratory mucosa throughout the document as necessary); with regard to the
incidence of forestomach lesions in mice in the 13-week study, text should state that preening behavior
might have lead to direct gastrointestinal exposure to chloroprene.
Response: Language regarding these issues was added to the document text and tables where
necessary.
Comment 20: Portions of the text in Section 4.2.2 refer to time to tumor data. Where are these data
and derivation described? Should some discussion of maximum tolerated dose and whether it was
exceeded be included in the text?
Response: The time-to-tumor analysis was detailed in Section 5.4 and complete time-to-tumor data
was added to Appendix C. A discussion regarding maximum tolerated dose and selection of the dose
groups for the chronic 2-year inhalation exposure was added to the text.
Comment 21: Information should be included in the document on how the survival-adjusted neoplasm
rates reported in Table 4-28 were calculated.
A-24
-------
Response: Text was added as a footnote in Table 4-28 detailing how survival-adjusted neoplasm rates
were calculated.
Comment 22: Additional analyses are needed before dismissing the findings of increased resorptions
in the 10 and 25 ppm exposure groups in Culik et al. (1978, 094969).
Response: The uncertainties surrounding these findings, including observation that the control group
in the teratology study falls far outside of the historic control range for this strain of rat leading to
potentially spurious statistical significance, was discussed fully and appropriately. The interpretation
that these data are unreliable was maintained in the document (Section 4.3).
Comment 23: Text in Section 4.5.2.1 alternatively stated that genotoxic activity was observed only in
strains TA97A and TA98 or in all strains tested.
Response: The text was clarified to state that there was evidence of genotoxicity observed in all
Salmonella strains tested, without Aroclor-induced S9 activation.
Comment 24: In Section 4.5.2.3, the hypothesis that chloroprene would only produce tumors in
directly exposed tissues has been disproved by the NTP (1998, 042076) studies which demonstrated
the multiple organ carcinogenicity of this chemical. This statement needs to be removed.
Response: The statement referenced above was taken from Tice (1988, 624981) and Tice et al. (1988,
064962). A clarifying sentence stating that chloroprene has been demonstrated to produce tumors
distal to the portal-of-entry was added, and thus the observed lack of effect in bone marrow may be due
instead to low metabolic activity in this tissue.
Comment 25: With regard to the comparison of carcinogenic potency of chloroprene versus butadiene,
it would be useful to have some quantitative comparison of cancer potency in rodents for these
compounds. A more comprehensive summary of potencies for other and/or all tumors would provide
important background for the quantitative cancer risk analysis. Table 4-37 should be supplemented
with a table giving quantification of the indicated potency for multiple- and all sites.
Response: Table 4-38, which details the relative cancer potencies of chloroprene and butadiene for a
number of tumor sites, was added to the document.
Comment 26: Table 4-37 is very confusing. What was the basis for including data from the rat relative
to "sites of increased incidence" of neoplasms? Listed are many sites in which statistically significant
results were not enumerated in previous portions of the text.
A-25
-------
Response: Table 4-37 compares the incidence of tumors in multiple organ systems in both mice and
rats that were exposed to butadiene, isoprene, or chloroprene. Its purpose is to show the similarity in
tumor profiles for the three structurally related compounds. All of the tumor types listed for
chloroprene have been previously discussed in the text. The lack of previous discussion for the
butadiene and isoprene tumors is logical as this document focuses on chloroprene. Any discussion of
tumor types induced by butadiene and isoprene is appropriately limited to this section, and Section
4.7.3.2, for the sole purpose of comparing tumor profiles as it contributes to the weight of evidence of
the carcinogenic potency of chloroprene operating via a mutagenic mode of action.
Comment 27: In general, the "synthesis" of the inhalation exposure data (Section 4.6) is not a
synthesis but merely a reiteration of the results. Rather than repeat the results study by study, it might
be much preferable to organize this section on the basis of target organ. It could, for example, discuss
the olfactory lesion data in toto, followed by the liver, etc. In this section, it is stated that chloroprene
is associated with reproductive and developmental effects, yet the earlier portions of the text concluded
otherwise.
Response: This section was extensively reorganized according to organ system and the observed
toxicity therein. The discussion on the reproductive and developmental effects of chloroprene
exposure was rewritten to emphasize the interpretation that those effects are equivocal.
Comment 28: Section 4.7 could be better organized. The summary in section 4.7.1 should probably be
moved to the end of the entire section on carcinogenicity. The human data are discussed separately in
an Evidence for Causality section, yet this is not provided for the animal studies. A true synthesis
would discuss Evidence for Causality across studies in all species. This could be integrated with the
discussion in Section 4.7.3.3 on Mode of Action to provide a stronger rationale for effects of
chloroprene
Response: Section 4.7.1 was moved to the end of the section and serves as the summary for the
Evaluation of Carcinogenicity section. While an Evidence for Causality section is included for the
epidemiology data, no such section was needed for the animal data. The new Section 4.7.2 (previously
Section 4.7.1) now serves to summarize all of the cancer data across studies in all species. Section
4.7.3.3 was a summary on the mode of action of chloroprene and the weight of evidence supporting a
mutagenic mode of action, and was thus limited to discussion of the observations that support this
determination.
Comment 29: In Section 4.7.1.1, the statement "Although not statistically significant, these findings
[increased relative risks of liver cancer observed by Marsh et al. (2007, 625188)1 were comparable to
A-26
-------
results (RR range 2.9-7.1) detected in two other studies for high and intermediate cumulative
exposures (Bulbulyan et al., 1998, 625105: Bulbulyan et al., 1999, 157419)" is made. Given that there
could have been considerable differences in exposure, follow-up, duration of exposure, etc. between
the studies, such a statement is probably not justified.
Response: This statement provides perspective on carcinogenic potential across studies. There are
differences between studies, but this comparison reinforces the fact that the results are consistently
elevated across studies.
Comment 30: In Section 4.7.1.1, the statement "only Bulbulyan (1999, 157419) observed a
statistically significant association between chloroprene exposure and liver cancer mortality" suggests
that this was done by an internal analysis, but the increase in liver cancer mortality was observed from
an external analysis.
Response: Bulbulyan (1999, 157419) observed statistically significant associations between
chloroprene exposure and liver cancer mortality based on both external and internal analyses.
Comment 31: Section 4.7.1.1 states "... although there is no direct evidence that alcohol is related to
the exposure of interest (i.e., chloroprene)..." Alcohol may not be related to the exposure of interest,
but that doesn't mean it could not have been a significant confounder. More convincing that alcohol
did not play a confounding role would have been clear evidence of a dose response to chloroprene
since it would be unlikely that alcohol consumption would correlate with chloroprene exposure.
Evidence of a dose response, however, seems equivocal (Table 4-11 on page 4-17).
Response: Alcohol is probably not a confounder based upon the available information. Alcohol
consumption was not related to the exposure of interest (chloroprene) or the outcome of interest (liver
cancer) for these studies. There was suggestive evidence of a chloroprene dose-response, or consistent
elevated risks in the upper exposure categories, in multiple studies (Bulbulyan et al., 1998, 625105:
Bulbulyan et al., 1999, 157419: Leet and Selevan, 1982, 094970: Marsh et al., 2007, 625188).
Comment 32: What "current understanding" allows for the statement that specificity is "one of the
weaker Hill criteria [sic]?"
Response: The criterion of specificity has many requirements and caveats that have been refuted and
deemed invalid by many authors. In particular, Rothman and Greenland (1998, 086599) state
"Specificity requires that a cause lead to a single effect, not multiple effects. This argument has often
been advanced to refute casual interpretations of exposures that appear to relate to myriad effects,
especially by those seeking to exonerate smoking as a cause of lung cancer. Unfortunately the
A-27
-------
criterion is wholly invalid. Causes of a given effect cannot be expected to lack other effects on any
logical ground. To summarize specificity does not confer greater validity to any causal inference
regarding the exposure effect." Therefore, the description of the criterion of specificity has been
modified in Section 4.7.1.1.1.
Comment 33: Section 4.7.1.2 included a listing of increased incidences of tumors, yet the basis for
inclusion in this listing is unclear. Some organs are listed in which the tumor incidence was not
significantly increased. The discussion of species differences (lines 27-31) should include reference to
possible species differences in epoxide hydrolysis rates. Such data are presented earlier and its
absence here is confusing. This section failed to include the most important species difference - the
appearance of lung tumors in mice but not rats. A clear metabolic basis might be provided, given that
the metabolic activation rate in mice appears to be 50-fold higher than the rat. This would also serve to
emphasize the potential role of metabolism relative to carcinogenicity. Epoxide formation is thought to
be important relative to the respiratory tract toxicity/carcinogenicity of naphthalene and styrene and the
same species differences (lung tumors in mice but not in rats) is seen for these vapors. Line 32
includes a reference to Dong et al. (1989, 007520): this study was not described previously.
Response: This section was rewritten to include discussion of only tumors that were biologically
noteworthy or showed a statistically significant increase in rats or mice exposed to chloroprene for
2-years ((NTP, 1998, 042076). A discussion of species differences in metabolism was also included, as
was the fact that lung tumors were induced in mice but not rats. A discussion of Dong (Dong et al.,
1989, 007520) has been included in Section 4.2.1.
Comment 34: Table 4-39 is somewhat confusing. Why was lung cancer mortality listed under "rare
tumors?" The table includes a reference to time to tumor, yet such data were not presented earlier in the
text.
Response: Primary lung cancer in humans is a rare cancer type. Time to tumor information (presented
as survival time) was previously presented in Section 4.2.1, including in the text and in Table 4-25.
Time to tumor data was presented more exhaustively in Tables 5-4 and 5-5, as well as in Appendix C.
Comment 3 5: In Section 4.7.3.1, the document specifies a mutagenic MO A involving the reaction of
epoxide metabolites formed at target sites. Until studies are conducted evaluating blood levels of
epoxide intermediates, it would be inappropriate to impose this target site limitation. It is not known if
epoxide formation occurs in all of the tumor target sites identified in the rodent carcinogenicity studies.
A-28
-------
Response: The sentence was changed to read "... chloroprene acts via a mutagenic mode of action
involving reactive epoxide metabolites formed at target sites or distributed systemically throughout the
body."
Comment 36: In section 4.7.3.2, the statement that in vivo uptake of chloroprene involved the balance
between epoxide formation and detoxification was confusing. Certainly the toxicity depends on the
balance, but it is unlikely that uptake does. Uptake rates depend on the blood and tissue concentration
of parent, downstream conversion of metabolite is not necessarily important in diffusion-based uptake.
Response: The text was changed to reflect that the toxicity of chloroprene involves a balance of
reactive epoxide formation and detoxification.
Comments?: In Section 4.7.3.2, it was stated that there is remarkable similarities in the potency and
shape of the dose response between butadiene and chloroprene. Such data were not presented in
earlier portions of the text.
Response: A discussion of the similarities between the carcinogenic potency and shape of the dose-
response curve of butadiene and chloroprene was added to Section 4.5.3 and Table 4-38 was added to
summarize that data.
Comment 38: In Section 4.7.3.3, it was stated that Melnick et al. (1994, 625208) performed a 6 month
exposure-6 month follow-up study. Where were these data presented?
Response: This study is used in support of the proposed mutagenic mode of action for chloroprene. It
is a study on a structurally related chemical, isoprene, and as such was not previously reported in the
document. It was reported in Section 4.7.3.2 to strengthen the argument that ras mutations observed in
chloroprene-exposed animals were most likely early mutagenic events in the development of
neoplasia.
Comment 39: In Section 5.2.1, the text needs to clearly describe how the atrophy and necrotic data
were combined. It is not certain there are any data indicating nasal olfactory atrophy leads to necrosis
(as stated on lines 5-6). The concept that necrosis may lead to atrophy is quite straightforward
however.
Response: This text was removed as atrophic and necrotic olfactory lesions were no longer combined
into one endpoint for the purposes of benchmark modeling.
A-29
-------
Comment 40: In Table 5-2, DAFs greater than 1 for lung and less than 1 for nasal epithelium deserve
specific discussion.
Response: The application of DAFs was removed from Table 5-2 and moved to the text where a more
complete and in-depth discussion of their calculation and application was included.
Comment 41: Regarding Section 5.2.3, chloroprene is not a Category 1 gas. Its partition coefficient is
only 10; clearly backpressure in nasal tissues controls the uptake process. The presence of
nonrespiratory tract tumors clearly indicates it is absorbed into the bloodstream. This vapor does not
possess the physical chemical characteristics required of Category 1 gases; in my view, it is a
Category 3 gas. The text needs to rigorously support this conclusion with respect to the physical
chemical characteristics of chloroprene relative to those required of Category 1 gases. The presence of
olfactory lesions is not evidence that the toxicant is delivered via the airstream. Numerous compounds
produce selective olfactory injury after parenteral administration. Indeed, the presence of olfactory but
not respiratory nasal mucosal injury might be considered to provide data in support of a blood-borne
mechanism. Naphthalene is one example of this phenomenon. Importantly, the subsequent text
describes in great detail how the lung lesions may be due to blood-delivered rather than air-delivered
chloroprene. The text needs to be consistent.
The RfC methodology is fatally flawed with respect to RR calculation. The derivations of
these equations are based on the faulty assumption that the mass transfer coefficient is uniform
throughout the nose. Dosimetry predictions from RGDR-based evaluations are totally discordant with
the data. "While application of a flawed methodology may be consistent with EPA policy, it certainly
is not consistent with the scientific state-of-the-art." The mode of action is assumed to include
metabolic activation to the epoxide. The RGDR of 0.28 indicates the humans will receive roughly
fourfold more toxicant (1/0.28) than the rat. Is it meant to imply that the metabolic activation rate in
the human nose is fourfold higher than the rat? The use of the RGDR needs to be discussed in light of
the metabolically-based mode of action.
Response: In response to this reviewer's previous comments (Charge Question 6), the application of
the default DAFs was performed assuming chloroprene to be a Category 3 gas. Consequently, a DAF
of 1 (for systemic effects) was used and the resultant PODnEc for was 2 mg/m3.
More in-depth discussion was included in the document regarding the physio-chemical
properties of chloroprene, including how those properties can impact the determination of which
dosimetric adjustments should be applied in calculating the human equivalent dose. Information
regarding the metabolism of chloroprene into the reactive epoxide and potential for this metabolism in
the respiratory tract (expression of CYP2E1 in the olfactory mucosa and microsomal oxidation of
chloroprene in mouse lung homogenates) was also included. Additional discussion was also included
A-30
-------
that posits that the observed toxicity of chloroprene in the respiratory tract may be due to systemic
redistribution of chloroprene.
Comment 42: With regard to the application of uncertainty factors, it may be policy to include a
database limitation factor due to the lack of a two generation study, but it was not scientifically
justified in this case. A multi-generation study does exist. The rationale for the selection of this
uncertainty factor should include this study.
Response: A true multi-generational study for chloroprene does not exist. The Appelman and Dreef
van der Meulen (1979, 064938) study is an unpublished report in which FO and Fl rats were exposed
to chloroprene. However, this study did not involve the mating of the Fl generation, so developmental
effects to the F2 generation could not be assessed.
Comment 43: Table 5-3 does not include a row in the consideration column for database limitation.
Response: Discussion of uncertainty regarding the completeness of the database was added to
Table 5-3.
Comment 44: In view of the saturation of the generation of an active metabolite, and the need to drop
high doses in some cases, there should be an investigation of a Michaelis-Menten transformation of
dose, in lieu of a full PBPK model.
Response: See the EPA Response to Comment 1 of Charge Question 11, pages A-16 through A-l 8.
Comment:45 If variability or uncertainty in slope factors follows a normal distribution, a lognormal
distribution could be used.
Response: The statement referring to variability in slope factors was removed and replaced with a
statement that asymptotic normality was assumed for the slope factors (Section 5.4.4).
A.3. PUBLIC COMMENTS
A.3.1. Interpretation of Epidemiological Studies
Comment 1: The Draft Review did not follow the USEPA approved method to assess epidemiological
data quality, as detailed in the guidelines for the assessment of human cancer risk (U.S. EPA, 2005,
086237). The Draft Review did not assign a study-specific weight to each study cohort to reflect the
quality of the study with regard to the relative strengths and limitations of each study.
A-31
-------
Response: The 2005 U.S. EPA Guide lines for Carcinogen Risk Assessment document (U.S. EPA,
2005, 086237) does provide criteria by which epidemiologic studies, whether providing positive or
negative evidence of association, can be judged in regards to study quality. Specifically, the guidelines
offer a list of characteristics that "are generally desirable in epidemiologic studies." The guidelines
also state that "conclusions about the overall evidence for carcinogenicity from available studies in
humans should be summarized along with a discussion of uncertainties and gaps in knowledge."
However, the guidelines do not support using the suggested criteria as a basis to score studies by an
individual weight for use in a comparison of study quality across multiple studies. As such, a
weighting and comparison scheme as suggested above is not supported by Agency guidance and was
not used in the Toxicological Review. Individual studies were assessed on the basis of study quality in
the document and extensive discussions of study limitations (individually and as part of the overall
weight-of-evidence discussion) were included in the document, in accordance with the 2005 Cancer
Guidelines.
Comment 2: One of the key studies cited by the US EPA as the basis for linking chloroprene exposure
with cancer (Leet and Selevan, 1982, 094970) was superseded by the Marsh et al. (2007, 625187:
2007, 625188) study. The Marsh et al. study of cohorts in the United States, Ireland, and France did
not report an association between exposure to chloroprene and the incidence of either total cancers or
cancers of the lung or liver.
Response: The Marsh et al. (2007, 625187; 2007, 625188) study investigated a employee cohort from
the Louisville Works DuPont plant that was previously investigated in Leet and Selevan (1982,
094970). However, there are a number of differences between the studies that warranted independent
analysis of each. Specifically, Leet and Selevan (1982, 094970) reported that the Louisville cohort
consisted of 1,575 male employees (salaried and female employees excluded due to "minimal or no
potential exposure to chloroprene") who were working at the Louisville plant on June 30, 1957. The
authors further reported that most of the employees had 15 years of potential exposure to chloroprene
(indicating that most had worked at the plant since it's opening in 1942). Also, the cohort was followed
until 1974. Marsh et al. (2007, 625187: 2007, 625188) included "all workers" (male and female) in
each plant with potential exposure to chloroprene from the "start of production" until 2000. For the
Louisville plant, this included a total of 5507 workers employed from 1949-1972. The Marsh et al.
(2007, 625187: 2007, 625188) analyses started at 1949 to "avoid methodological problems associated
with the earlier fifth revision of the ICD" and stopped at 1972 for the Louisville plant as that was when
they report chloroprene production stopped at that plant, although chloroprene purification and
polymerization still occurred there according to Leet and Selevan (1982, 094970). Also, there are
important differences in how each study assessed exposure. Leet and Selevan (1982, 094970) used
worker history summaries to classify workers as either "high" or "low" chloroprene exposure, whereas
Marsh et al. (2007, 625187: 2007, 625188) used a more sophisticated approach that considered worker
A-32
-------
history summaries and worker exposure profiles to generate quantitative estimates of chloroprene
exposure intensity. Therefore, although the two studies investigated members of the same cohort, a
number of methodological differences between the studies warranted the independent analysis of each.
Comment 3: Interpretations of the Chinese, Russian, and Armenian cohorts (Bulbulyan et al., 1998,
625105: Bulbulyan et al., 1999, 157419: Li et al., 1989, 625181) failed to acknowledge the imprecise
and unstable estimates of mortality and incidence ratios due to very low expected counts used for liver
and lung cancer mortality.
Response: Although some cohorts did report very low expected counts used for liver and lung cancer,
some of these same studies demonstrated statistically significant associations that are fairly precise
(e.g., Bulbulyan et al. (1999, 157419). Naturally, studies with a limited number of outcomes and those
that examine exposure-response relationships with few deaths in each cell will have wider confidence
bounds. Given the rarity of the outcomes that were examined (especially in the general population),
the expectation would be a low number of deaths. This was demonstrated for outcomes (e.g., liver
cancer mortality) in many studies including several of the DuPont plants, as there were statistical
power limitations when examining cancer-specific effects and exposure-response relationships.
Regardless of the study, this Toxicological Review has highlighted the issue of imprecision by
presenting confidence intervals and discussions of small sample size throughout the document.
Although, on an individual study basis, there may exist some concern over the potential role of chance
for isolated outcomes that were not replicated in later studies, the consistency of the results indicate
that chance is an unlikely explanation of the results across heterogeneous study populations and
exposure scenarios in several countries. As such, these studies contribute to the weight-of-evidence
characterization of the carcinogenic potential of chloroprene.
Comment 4: The Chinese, Russian, and Armenian studies have limitations and confounders that limit
the interpretation and conclusions of their reported findings
Response: The limitations of each study, including potential confounding, has been discussed
individually and together in multiple sections of the Toxicological Review.
Comment 5: The Draft Review currently gives limited consideration to the Marsh et al. (2007,
625187: 2007, 625188) studies in regard to the overall weight-of-evidence for the association between
chloroprene and cancer mortality.
Response: All studies were judged independently on their individual merits and given full
consideration in the overall weight-of-evidence characterization. The Marsh et al. (2007, 625187:
2007, 625188) studies are discussed in detail in Sections 4.1 and 4.7. In regard to study strengths and
A-33
-------
findings, the studies have been characterized and considered in the overall weight-of-evidence. It is
important to note that, although the Marsh et al. (2007, 625187: 2007, 625188) did not observe
statistically significant associations between chloroprene exposure and cancer mortality, they did
observe elevated risks when internal comparisons were performed (Section 4.1.1.2). Some of these
results were similar in magnitude to findings in other studies which reported more consistent
associations between chloroprene exposure and cancer mortality.
Comment 6: Assessing causality failed to apply methods recommended by the Guidelines for
Carcinogen Risk Assessment (U.S. EPA, 2005, 086237). Specifically, the Draft Assessment does not
explicitly evaluate available epidemiologic quantitative results for potential bias due to systematic
errors (i.e., bias, misclassification, and confounding) and random errors (i.e., the role of chance).
There has been consistent agreement among previous reviews of the epidemiology database for
chloroprene that studies indicating a positive association are of insufficient quality to infer a causal
relationship between chloroprene and cancer mortality.
Response: The Toxicological Review has exhaustively discussed individual study limitations
including a thorough examination of the potential for bias in multiple sections of the document. In
regard to inferring a causal relationship between chloroprene exposure and cancer mortality, no such
definitive determination is made in the Toxicological Review based on the epidemiologic data. In the
discussion of the Evidence of Causality (Section 4.7.1.1.1.), the document states:
It should be noted that there exists a number of methodological limitations of the epidemiologic studies
that may preclude drawing firm conclusions regarding the following criteria. These limitations include
lack of control of personal confounders and risk factors associated with the outcomes in question,
imprecise exposure ascertainment resulting in crude exposure categories, incorrect enumeration of cases
leading to misclassification errors, limited sample sizes, and the healthy worker effect. ... In summary,
the temporality of exposure prior to occurrence of liver cancer, strength of association, consistency,
biological gradient, and biological plausibility provide some evidence for the carcinogenicity of
chloroprene in humans.
Thus, the document makes no definitive claim of a causal relationship between chloroprene
exposure and cancer mortality, but rather explicitly states that there is evidence of an association across
the body of scientific literature.
Comment 7: US EPA interpretation of the potential for lung and liver cancer risks of chloroprene
based on the Marsh et al. (2007, 625187: 2007, 625188) study did not fully consider the impact of
inordinately low death rates for lung and liver cancer among workers in the baseline categories.
Response: Although the authors highlight some "exceedingly" low mortality figures in the "baseline"
exposure levels (i.e., lowest exposure category), comparable numbers of deaths are found in low-,
intermediate-, and some high-exposure groups across different outcomes (those RRs < 1.00 for all
cancers, respiratory and liver cancer mortality). It is unclear why the authors consider any RRs in
A-34
-------
excess of 1.00 to be due to an "exceedingly" low baseline mortality rate. There is little evidence to
suggest that this is not a valid population in which to base comparisons on, and the results of the
internal analyses are preferred given the strong evidence of the healthy worker effect in the SMR
analyses. In addition, given the fact that such strong RRs were detected in healthy workers, one would
be more concerned about potential risk among less healthy populations under similar circumstances.
Comment 8: Vinyl chloride exposure as a potential confounder of the association with chloroprene
exposure and liver cancer in the Marsh et al. (2007, 625187; 2007, 625188) study is not supported
given the lack of correlation between chloroprene and vinyl chloride exposure.
Response: The Toxicological Review has discussed the potential for vinyl chloride to act as a
confounder in detail in Section 4.1.1.2. As noted, since there was no association between cumulative
exposures to vinyl chloride and chloroprene among these workers, vinyl chloride does not meet the
definition of a confounder, and thus any association between chloroprene exposure and cancer
mortality is highly unlikely to be modified by vinyl chloride exposure. The internal analyses of Marsh
et al. (2007, 625187; 2007, 625188) also indicated that there is an inverse association between vinyl
chloride exposure and risk of both respiratory and liver cancers based on limited numbers of cancer
deaths in the vinyl chloride-exposed groups. Therefore, even if vinyl chloride exposures were
positively correlated with chloroprene exposures among workers, any resulting negative confound
would result in attenuation of unadjusted relative risk estimates. That is, associations stronger in
magnitude would be expected if the relative risk estimates for chloroprene and cancer were adjusted
for vinyl chloride exposures.
A-35
-------
A.3.2. Interpretation of Mode of Action Based on the Mutagenicity and Genotoxicity Data
Comment 1: Standard in vivo tests for genotoxicity were negative: chloroprene, unlike butadiene and
isoprene, does not exert genetic toxicity to somatic cells in vivo.
Response: The Toxicological Review describes numerous in vivo genotoxicity tests that return
nonpositive results, including lack of sister chromatid exchange or chromosomal aberrations in bone
marrow and no evidence of micronuclei formation in peripheral blood erythrocytes. However, when
Drosophila melanogaster were exposed to chloroprene (99% pure with negligible dimer content), an
increase in recessive lethal mutations on the X chromosome of male flies was observed (Vogel, 1979,
000948). Similar results were not observed in a similar experiment by Foureman et al. (1994, 065173).
However, there were significant differences between the two experiments that may explain different
findings: (1) differences in purity of the chloroprene sample (99% pure in Vogel (1979, 000948) and
only 50% pure in Foureman et al. (1994, 065173)1 (2) differences between the Berlin-K (Vogel, 1979,
000948) and Canton-S (Foureman et al., 1994, 065173) strains, (3) differences in sample sizes, and
(4) possible genetic drift within the female populations used by the two groups of investigators.
Regardless, the strongest evidence of in vivo genotoxicity is the observation of genetic alteration of
cancer genes including the ras proto-oncogenes (Sills et al., 1999, 624952: Sills et al., 2001, 624922:
Ton et al., 2007, 625004), which are alterations commonly observed in human cancers. Tissues from
lung, forestomach, and Harderian gland tumors from mice exposed to chloroprene in the 2-year NTP
chronic bioassay (1998, 042076) were shown to have a higher frequency of mutations in K- and H-ras
proto-oncogenes than in spontaneous occurring tumors (Sills et al., 1999, 624952; Sills et al., 2001,
624922). Further, there was a high correlation between K-ras mutations and loss of heterozygosity in
the same chromosome in chloroprene-induced lung neoplasms in mice (Ton et al., 2007, 625004).
Similar increases in the frequencies of K-ras mutations in rodents were observed in isoprene-induced
lung neoplasms and vinyl chloride-induced hepatocellular carcinomas (NTP, 1998, 042076; U.S. EPA,
2000, 194536). Activated K-ras oncogenes were also observed in lung tumors, hepatocellular
carcinomas, and lymphomas in B6C3Fi mice exposed to 1,3-butadiene (U.S. EPA, 2002, 052153).
Comment 2: There is a general lack of consistent data for chloroprene-induced point mutations. The
ability of chloroprene to induce point mutations in bacteria is equivocal at best and chloroprene did not
induce mutations in cultured mammalian cells. Conflicting specificities between in vitro bacterial
point mutations (GGG) and DNA adduct induction (preferentially forming guanine adducts when
incubated with calf thymus DNA) and in vivo ras mutations found at tumor sites (A to T transversions)
indicate that in vivo mutations may be of a nonchloroprene origin.
Response: The Toxicological Review presented the bacterial genotoxicity data as returning conflicting
results, but did note that when positive results were observed they occurred in Salmonella strains that
A-36
-------
test for point mutations. Assays with Salmonella strains that tested for frameshift mutations were
consistently negative. Aguanine adduct was the major adduct observed (approximately 96% of
adducts formed) when the epoxide metabolite of chloroprene is reacted with calf thymus DNAin a
cell-free environment. However, when equimolar quantities of all four nucleosides were reacted with
(l-chloroethenyl)oxirane simultaneously in a competitive reaction assay, all of the adducts identified
from individual nucleoside reactions were observed and were formed at similar rates (Munter et al.,
2007, 576501; Munter, et al., 2002, 625215). As stated above, the strongest line of evidence indicating
that chloroprene induced point mutations leading to a carcinogenic response was the observation that
tissues from chloroprene-induced lung, forestomach, and Harderian gland tumors in mice
demonstrated a higher frequency of mutations in K- and H-ras proto-oncogenes than in spontaneous
occurring tumors (NTP, 1998, 042076). Although the majority of these point mutations were A to T
transversions, a number of G transversions were also observed in lung and forestomach tumors.
Another strong indication that the A to T transversion at codon 61 in mouse lung tumors is
chloroprene-induced is that it was not observed in spontaneously occurring tumors in NTP historic
controls.
Comment 3: A nongenotoxic mode of action for chloroprene should be considered. An alternative
mode of action is that chloroprene induces localized cytotoxicity with subsequent induction of
hyperplasia and cell regeneration followed by promotion of pre-existing proto-oncogene mutations.
Response: The Toxicological Review states that there may be alternative modes of action operant in
certain situations (i.e., high dose exposures) that may explain why lung tumors are observed at high
doses when the frequency of ras mutations is less than is observed at lower doses (Section 5.4.1).
However, the scientific evidence indicates that a mutagenic mode of action is a plausible mode of
action with regard to the carcinogenicity of chloroprene. The observation that the majority of ras
mutations in the lungs of chloroprene exposed mice consisted of Ato T transversions at codon 61
(22/37) is inconsistent with the proposed alternative mode of action. If chloroprene exposure were
initiating cytotoxicity with subsequent hyperplasia/regeneration leading to promotion of pre-existing
proto-oncogene mutations, the expectation would be that no Ato T transversions at codon 61 would be
observed as this mutation is not seen in spontaneously occurring lung tumors in historic controls. The
current proposed mutagenic mode of action was unanimously accepted by the External Peer
Reviewers. Additionally, the mutagenicity of chloroprene is proposed in numerous studies cited in the
Toxicological Review, including but not limited to: Munter et al. (2003, 625214): Summer and Greim
(1980, 0649611 Himmelstein et al. (2001, 019013: 2004, 625154: 2004, 625152): Melnick et al. (1999,
000297): Ponomarkov and Tomatis (1980, 075453).
A-37
-------
A.3.3. Consideration of Species Differences in Toxicokinetics and Target Tissue Dosimetry
Comment 1: Significant species differences in metabolism are documented, and the peer reviewed
literature (Cottrell et al., 2001, 157445: Himmelstein et al., 2004, 625152: Munter et al., 2007, 625213:
Munter et al., 2007, 576501) demonstrates that there are significant differences in the metabolism of
chloroprene across species that can impact target tissue dose.
Response: The observed species differences in metabolism were acknowledged and extensively
discussed in the Toxicological Review. While differences in metabolism do exist across species that
could substantially impact target tissue dose, additional discussion added to Section 3.3 indicate that
differences in epoxide production in the lungs of mice and humans are not as great as 50-fold (as once
indicated in a prior draft of the Toxicological Review), but may be as little as 2- to 10-fold
(Himmelstein et al., 2004, 625152: Himmelstein et al., 2004, 625154). These additional data also
indicate that in some cases (i.e., glutathione transferase activity) detoxification of the epoxide
metabolite may be faster in mice than humans. Also, there appears to be an additional detoxification
pathway, oxidation of (l-chloroethenyl)oxirane, that is active in mice, but not in humans, rats, or
hamsters. Therefore, the document clearly and transparently presents data that do indicate that species
differences exist in the metabolic activation of chloroprene; however, these differences are not so great
as to preclude using animal data to estimate the noncancer and carcinogenic toxicity of chloroprene in
humans.
Comment 2: Previous analyses (Himmelstein et al., 2004, 625154) support the use of the PBPK
model.
Response: The use of the PBPK model described in Himmelstein et al. (2004, 625154) in the
Toxicological Review was not supported for a number of reasons discussed throughout the document.
Specifically, the model predicted blood chloroprene and delivery of chloroprene to metabolizing
tissues based on metabolic constants and partition coefficients based on in vitro data. Loss of chamber
chloroprene was attributed to uptake and metabolism by test animals and was used to test the metabolic
parameters and validate the model. However, Himmelstein et al. (2004, 625154) did not provide
results of sensitivity analyses indicating whether chamber loss was sensitive to metabolism, and
therefore it is uncertain whether chamber loss was useful for testing the metabolic parameters used in
the model. Also, the chamber data were fit by varying alveolar ventilation and cardiac output. This
method did not result in adequate testing of the model and did not validate the scaled in vitro metabolic
parameters. Additionally, there were currently no blood or tissue time-course concentration data
available for model validation.
A-38
-------
Comment 3: New data supplied by DuPont at the External Peer Review Meeting 1) support the use of
the quantitative PBPK model, 2) increase confidence in the PBPK model parameters (through refined
liver and lung microsomal metabolic parameters and new kidney microsomal metabolic parameters),
and 3) provide genomic evidence that kinetic differences alone do not influence the production and
retention of reactive metabolites.
Response: At the time of the External Peer Review meeting, the data provided by DuPont had not
been peer-reviewed and as such could not be used as the basis for the use of the PBPK model and the
derivation of the RfC or inhalation unit risk.
A.3.4. U.S. EPA Decision Points in the Determination of the Inhalation Unit Risk
Comment 1: The presentation of datasets to be used to determine the RfC, including the dataset
ultimately selected (i.e., nasal lesions in the male rat) needs additional information. Table 5-1 is
potentially misleading, in that it suggests by omission that nasal effects are only observed in male rats.
Table entries for nasal effects in female rats are listed "not observed," which is incorrect. Also missing
from Table 5-1 are the data for nasal atrophy in male and female mice.
Response: Additional endpoints were added to Table 5-1, including nasal effects observed in female
rats. The criteria for what endpoints were considered for selection of the critical effect were changed
such that all nonneoplastic lesions that were statistically increased in mice or rats at the low- or mid-
exposure concentration (12.8 or 32 ppm) compared to chamber controls, or demonstrated a suggested
dose-response relationship in the low- or mid-exposure range in the absence of statistical significance,
were considered candidates for the critical effect. Table 5-1 was edited to reflect this. Also, nasal
atrophy in male and female mice was not included in Table 5-1 as that endpoint fails to satisfy the
criteria listed above.
Comment 2: A value of 3 for database deficiencies for chloroprene is incorporated in the derivation of
the RfC. However, several lines of evidence suggest that this value may not be needed. First,
chloroprene is not expected to accumulate in tissues such that in a multigenerational study, exposure to
the second generation (F2) would be greater than experienced by the first generation (Fi). Second, the
NOAEL for reproductive toxicity of 100 ppm in the unpublished report by Appelman and Dreef van
der Meulen (1979, 064938) is higher than NOAELs/LOAELs for nasal and systemic effects observed
in the NTP (1998, 042076) study. Based on this comparison of NOAELs/LOAELs, EPA should
reconsider the application of an UF for database uncertainties due to the lack of a multigenerational
study.
A-39
-------
Response: A database uncertainty factor of 3 was maintained in the document due to the lack of a
multigenerational developmental/reproductive study. The lack of a multigenerational precludes the
ability to assess the effects of chloroprene on postnatal maturation and reproductive capacity of the Fl
offspring, and any cumulative effects that may manifest throughout multiple generations. Therefore,
due to the lack of a multigenerational study, there exists residual uncertainty in the chloroprene
database that is accounted for by the current database uncertainty factor of 3.
Comment 3: In the Draft Review, a proprietary software program (TOX_RISK version 5.3) was relied
upon for the time-to-tumor dose-response modeling. This software is no longer available to the
general public, and adversely affected the transparency of the dose-response model. Simpler models
provided in BMDS should be used instead.
Response: The time-to-tumor dose-response modeling was redone using EPA's Multistage Weibull
(MSW) time-to-tumor model. This model is free and available to the general public at:
www.epa.gov/ncea/bmds/dwnldu.html. Use of this model removed any previous issues with the
transparency of the modeling approach.
Comment 4: EPA's assumption that hemangiosarcomas were the only fatal tumor type did not appear
to be consistent with the data, in that the pattern of responses should have been different if
hemangiosarcomas had impacted the occurrence of other tumors. Incidence of these tumors dropped at
the high dose, suggesting that other tumors caused deaths before the hemangiosarcomas could have
developed. This modeling approach was not viable without considering lethality assumptions further.
Response: EPA agrees that earlier deaths likely impacted the incidence of circulatory system tumors;
that is why the multistage-Weibull model was used. However, the designation of some tumors as fatal
did not automatically imply that they occurred earlier than the rest of the tumors. The multistage-
Weibull model addressed the time of death for each animal as recorded; the fatal designation impacted
only the magnitude of the risk estimate for that tumor type and is not a data input for the analysis of
other tumor types. Designation of individual tumor occurrences as fatal (as appropriate) will tend to
increase unit risk estimates. As shown in the document, analyses of fatal and incidental circulatory
system tumors showed a roughly twofold range in unit risks between treating all tumors as incidental
or all as fatal; the more representative value is likely between those two extremes. However, without
specific causes of death for each animal in this study, it is difficult to consider the impact of this issue
more thoroughly. The uncertainty discussion was expanded to include these points.
Comment 5: Model selection (goodness-of-fit for arriving at final number of stages) was not well
characterized.
A-40
-------
Response: A summary of the model selection decisions was added (Section 5.4.3).
Comment 6: Unit risks from multiple tumor types should not be summed in the determination of the
composite unit risk for carcinogenicity. Given the considerable overlap in tumor incidence data among
animals, EPA's assumption that the tumors are independent leads to an overstatement of the
carcinogenic potential of chloroprene. EPA's method has no precedent in final IRIS assessments, and
is statistically flawed. The most appropriate approach for derivation of the unit risk for chloroprene if
animal data are used is to rely upon the most sensitive tumor endpoint (i.e., lung tumors) in the most
sensitive species.
Response: Basing the inhalation unit risk on only one tumor type when chloroprene has been shown to
induce tumors in multiple organ systems in two species of rodents would most likely result in an
underestimation of the human carcinogenic potential of chloroprene. The basis for considering the
tumor types statistically independent was clarified. Briefly, the commenter's demonstration of the
overlap of tumors focused on the overlap of tumors at the high doses, where there is insufficient
information to determine whether the tumors are independent or not, since high rates of response have
to overlap regardless of their independence. Note that at the lowest exposure, only 9 of 36 female mice
with tumors had more than one tumor. The composite unit risk describes the risk for much lower
exposures where the risk of multiple tumors is trivial.
Concerning the statistical method, the document was revised to clarify that it is an approximate
approach. The document cited two final IRIS assessments that have used this method, and a third has
been added; all were externally peer reviewed.
Comment 7: Because the mode of action proposed for chloroprene in the Draft Review is dependent
upon target tissue dose, it is critical that the HEC values take into consideration important species
differences in metabolism.
Response: Additional discussion of toxicokinetics included throughout the Toxicological Review
clearly and transparently present data that do indicate that species differences exist in the metabolic
activation of chloroprene. However, these differences are not so great as to preclude using animal data
to estimate the noncancer and carcinogenic toxicity of chloroprene in humans.
Comment 8: The points of departure for two tumor types (lung and liver) in the female mice appear to
fall considerably below the range of observation (i.e., by more than a factor of 3), and therefore are
inconsistent with U.S. EPA guidelines for benchmark modeling.
Response: The selected PODs are in fact consistent with the cited guidance. The BMRs are within the
observable range, "the range of doses for which toxicity studies have reasonable power to detect
A-41
-------
effects" (U.S. EPA, 2000, 052150), since 10% is within the sensitivity of typical cancer bioassays,
such as this one. Use of a BMR which falls within the actual range of responses observed in this study
leads to a trivial difference in the estimated PODs for these two derivations.
Comment 9: The lung tumor response data was assessed in the Draft Review assuming the responses
were either portal-of-entry effects or systemic effects. This approach is internally inconsistent with the
noncancer assessment in which the nasal atrophy/necrosis and lung hyperplasia in rodents were
attributed as portal-of-entry effects.
Response: In the current derivation of the RfC, chloroprene is treated as a Category 3 gas that causes
systemic effects from blood-borne distribution or in-situ metabolic activation. The noncancer and
cancer quantitative sections have been made consistent in this regard.
A.3.5. U.S. EPA Quality Control in Reporting Chloroprene Data
Comment 1: In comparing information provided in the Draft Review to that in the primary literature, a
number of inconsistencies were noted. In particular, commenters interpreted dose-response modeling
to have inappropriately included animals without histopathologic evaluation for particular sites.
Commenters also noted inconsistencies within the document. In addition, information on the
production of chloroprene in the Draft Review is not current and there are issues in attempting to
duplicate some of the quantitative analyses.
Response: All editorial corrections regarding data reporting were made where needed. The majority
of data discrepancies noted suggested that risks may have been underestimated:
• One tumor response, a forestomach tumor in a high dose male mouse, had been
inadvertently omitted from dose-response modeling; all relevant analyses have been
revised.
• Animals noted with missing tissues, but included in dose-response analyses were
included correctly; time-to-tumor modeling takes into account the time on the study
without appearance of a tumor. If they had been included in a simpler dichotomous
model, such as the multistage model, an underestimate of risk would have resulted. In
the instance of an animal on study for 3 days, EPA concluded there was likely little
impact including or excluding that animal. For purposes of accountability, these
animals were included in the analyses.
Other discrepancies noted:
• Number of animals considered at risk for dose-response analysis of Zymbal's gland and
Harderian gland tumors, which were not evaluated histopathologically in all animals—
Denominators were corrected.
A-42
-------
• Differences in ToxRisk output suggesting different time value inputs—Time values had
been input as week of study, not weeks on study. Since this was done consistently
throughout the data sets, no substantive difference was expected. The input data were
included in the assessment.
Information on the physical/chemical properties of chloroprene was corrected. Information
provided to EPA by DuPont regarding current production and manufacturing levels and processes was
added to the Toxicological Review. Information previously in the Toxicological Review was retained
in the document to give a complete description of historical and current production levels and
processes. The Sanotskii (1976, 063885) reference was retained in the Toxicological Review: although
there are concerns with the methodologies used by the studies cited in the Sanotskii review, these
concerns have been detailed appropriately in the Toxicological Review. The Sanotskii review does not
serve as the basis for any quantitative analysis, and only provides data and results that are qualitatively
useful in comparison to other study reports included in the Toxicological Review. The limitations of
the Sanotskii review are appropriately detailed when the paper is first referenced; it was not necessary
to exhaustively delineate the study limitations at every instance the paper is cited in the Toxicological
Review.
A-43
-------
APPENDIX B. BENCHMARK DOSE MODELING RESULTS FOR THE
DERIVATION OF THE RFC
Benchmark Dose (BMD) modeling was performed to identify the point of departure for
the derivation of the chronic RfC for chloroprene. The modeling was conducted in accordance
with the draft EPA guidelines (U.S. EPA, 2000, 052150) using Benchmark Dose Software (2009,
200772) Version 2.1.1 (BMDS). The BMDS model outputs for the derivation of the chronic RfC
are attached. The doses used in modeling the individual endpoints, and reported as BMDs and
BMDLs, are in ppm.
The following effects were modeled using BMDS: alveolar epithelial hyperplasia (male
and female rats), bronchiolar hyperplasia (male and female mice), olfactory epithelial chronic
inflammation (male rats), olfactory epithelium atrophy (male rats), olfactory epithelial necrosis
(male and female rats), olfactory basal cell hyperplasia (female rats), kidney (renal tubule)
hyperplasia (male and female rats), forestomach epithelial hyperplasia (male and female mice)
and splenic hematopoietic cell proliferation (female mice). Due to the nature and severity of the
nasal degenerative effects (i.e., olfactory atrophy and necrosis), and the proximity of the
BMDLio values to the observed LOAEL compared to other endpoints (Table 5-2), a BMR of 5%
was considered to be appropriate for these olfactory endpoints. The nature of the observed nasal
lesions potentially included the loss of Bowman's glands and olfactory axons in more severe
cases. Effects that occur in the underlying lamina propria and basal layer of the olfactory
epithelium may be indicative of more marked nasal tissue injury. For all other endpoints, a BMR
of 10% was chosen as the response level (Table B-l). The endpoint being modeled specified
which set of models, either continuous (linear, polynomial, power, and Hill) or dichotomous
(gamma, logistic, multi-stage, probit, quantal-linear, quantal-quadratic, Weibull, and
dichotomous Hill), would be utilized. Model eligibility was determined by assessing the
goodness-of-fit using a value of a = 0.1 (i.e., p-value > 0.1), ^ scaled residuals, visual fit, and
consideration of model parameter estimates. Once all appropriately fitting models were
identified, final model selection was based on either the Akaike Information Criterion (AIC)
when the BMDL estimates for all appropriately fitting models were sufficiently close (i.e., within
threefold difference of one another) or the lowest BMDL when they were not within threefold
difference of each other.
The co-critical endpoints selected for the derivation of the chronic RfC were olfactory
atrophy in the male rat, alveolar hyperplasia in the female rat, and splenic hematopoietic
proliferation in female mouse. The probit model provided the best fit for this data set. Tables
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental
Research Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science
assessments such as the Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
B-1
-------
B-l through B-21 are summaries of the modeling results for all considered endpoints. The best
fitting model for each endpoint is indicated in bold and the model plot (Figures B-l through
B-l6) and output are included immediately after the table.
B-2
-------
Table B-l. Severity scores at control dose and lowest dose showing response
for endpoints considered for critical noncancer effect
Endpoint
0 ppm
r
II
III
IV
12.8 ppm
I
II
III
IV
32 ppm
I
II
Ill
IV
Male Rats
Alveolar hyperplasia
Kidney hyperplasia
Olfactory atrophy
Olfactory basal cell
hyperplasia
Olfactory metaplasia
Olfactory necrosis
Olfactory chronic
inflammation
3
-
1
0
2
0
0
2
-
2
0
4
0
0
0
-
0
0
0
0
0
0
-
0
0
0
0
0
10
-
6
0
5
5
5
5
-
3
0
0
1
0
1
-
3
0
0
5
0
0
-
0
0
0
0
0
b
-
-
18
-
-
-
-
-
-
18
-
-
-
-
-
-
2
-
-
-
-
-
-
0
-
-
-
Female Rats
Alveolar hyperplasia
Kidney hyperplasia0
Olfactory atrophy
Olfactory basal cell
hyperplasia
Olfactory metaplasia
Olfactory necrosis
3
-
0
0
0
0
2
-
0
0
0
0
0
-
0
0
0
0
1
-
0
0
0
0
15
-
1
0
1
0
6
-
0
0
0
0
1
-
0
0
0
0
0
-
0
0
0
0
~
-
31
16
34
3
~
-
7
1
1
2
~
-
2
0
0
3
~
-
0
0
0
0
Male Mice
Bronchiolar
hyperplasia
Kidney hyperplasia
Forestomach epithelial
hyperplasia
Splenic hematopoietic
cell proliferation
0
-
0
2
0
-
2
12
0
-
0
10
0
-
2
2
3
-
2
2
5
-
3
15
1
-
1
5
1
-
0
0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Female Mice
Bronchiolar
hyperplasia
Forestomach
epithelial hyperplasia
Splenic hematopoietic
cell proliferation
0
1
0
0
2
8
0
0
4
0
1
1
4
0
3
8
0
13
2
2
6
1
1
3
-
-
-
-
-
-
-
-
-
-
-
-
""Severity scores -1 = minimal, II - mild, III - moderate, IV - marked
bOnly severity scores in control dose and lowest dose with response used to make determination of severity
progression with increasing dose
'Severity for single sections and step sections combined not available
Source: NTP (1998, 042076)
B-3
-------
Table B-2. Benchmark modeling results for alveolar epithelial hyperplasia in
male F344/N rats (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
231.042
232.34
230.479
233.859
231.042
232.209
231.042
231.042
231.705
Goodness-of-
fit p-value
0.1317
0.0775
0.1753
0.0363
0.1317
0.0813
0.1317
0.1317
0.1112
X2 residual
1.698
-0.092
1.566
-0.087
1.698
-0.126
1.698
1.698
-0.1356
BMD
14.8657
24.4838
11.4228
28.604
14.8657
23.3986
14.866
14.866
5.87477
BMDL
10.0883
19.1571
7.06934
19.5927
10.0883
18.2584
10.0883
10.0883
3.85444
aBold indicates model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
T3
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_a~lv_hyper_Ln~l-BMRlO-
Restrict.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_a~lv_hyper_Ln~l-BMRlO-
Restrict.pit
Thu Dan 14 14:17:54 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-s~lope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
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 = 0.1
intercept = -4.4782
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
the
user,and do not appear in the correlation matrix )
background intercept
background 1 -0.66
intercept -0.66 1
Parameter Estimates
95.0% Wald Confidence
Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf.
Limit
background 0.130984 * * *
intercept -4.63283 * * *
slope 1 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(1 ike~lihood) # Param's Deviance Test d.f. P-value
Full model -111.57 4
Fitted model -113.24 2 3.33902 2 0.1883
Reduced model -121.815 1 20.4898 3 0.0001343
AIC: 230.479
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0 . 0000
12 . 8000
32.0000
80 . 0000
0.1310
0.2272
0.3373
0.5113
6.549
11.360
16.526
25.564
5.000
16.000
14.000
25.000
50
50
49
50
Scaled
Residual
-0.649
1.566
-0.763
-0.160
ChiA2 = 3.48 d.f. = 2 P-value = 0.1753
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 11.4228
BMDL = 7.06934
B-5
-------
Table B-3. Benchmark modeling results for alveolar epithelial hyperplasia in
male F344/N rats (BMR = 5% extra risk)
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
231.042
232.34
230.479
233.859
231.042
232.209
231.042
231.042
231.705
Goodness-of-
fit p-value
0.1317
0.0775
0.1753
0.0363
0.1317
0.0813
0.1317
0.1317
0.1112
X2 residual
1.698
1.701
-0.649
1.939
1.698
1.711
1.698
1.698
-0.1356
BMD
7.23716
13.0617
5.41078
19.8906
7.23716
12.3417
7.23729
7.23729
2.63097
BMDL
4.91134
10.1972
3.34864
13.6243
4.91134
9.6301
4.91134
4.91134
1.72618
aBold indicates model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
T3
%
ID
ro
LL
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0 -BMDL BMD
70
80
08:09 02/03 2010
Figure B-2. Log-logistic model fit for alveolar epithelial hyperplasia in male
F344/N rats (BMR = 5% extra risk).
B-6
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_a~lv_hyper_Ln~l-BMR05-Restrict.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_a~lv_hyper_Ln~l-BMR05-
Restrict.pit
Wed Feb 03 08:09:27 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-s~lope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
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 = 0.1
intercept = -4.4782
slope = 1
the
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
background
intercept
background
1
-0.66
intercept
-0.66
1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
0.130984
-4.63283
1
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood)
-III.57
-113.24
-121.815
230.479
# Param's
4
2
1
Deviance Test d.f.
3.33902
20.4898
P-value
0.1883
0.0001343
Goodness of Fit
Dose
0.0000
12 . 8000
32.0000
80.0000
Est._Prob.
0.1310
0.2272
0.3373
0.5113
Expected
6.549
11.360
16.526
25.564
Observed
5.000
16 . 000
14 . 000
25.000
Size
50
50
49
50
Scaled
Residual
-0.649
1.566
-0.763
-0.160
ChiA2 = 3.48
d.f. = 2
P-value = 0.1753
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 5.41078
BMDL = 3.34864
B-7
-------
Table B-4. Benchmark modeling results for alveolar epithelial hyperplasia in
female F344/N rats (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
245.78
248.949
243.677
249.954
245.78
248.806
245.78
245.78
244.808
Goodness-of-
fit p-value
0.0612
0.0163
0.1779
0.0079
0.0612
0.0171
0.0612
0.0612
0.113
X2 residual
1.998
1.99
-0.453
2.446
1.998
2.006
1.998
1.998
-0.1096
BMD
8.0322
14.8564
4.90719
15.342
8.03223
14.4844
8.03223
8.03223
3.08661
BMDL
5.89582
11.9857
3.27097
10.7468
5.89582
11.8082
5.89582
5.89582
2.02512
aBold indicates model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
T3
0)
ro
LL
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0 -BMDL BMD
70
80
14:09 01/14 2010
Figure B-3. Log-logistic model fit for alveolar epithelial hyperplasia in
female F344/N rats (BMR = 10% extra risk).
B-8
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_f_a~lv_hyper_Ln~l-BMRlO-
Restrict.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_f_a~lv_hyper_Ln~l-BMRlO-
Restrict.pit
Thu Dan 14 14:09:02 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-s~lope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
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 = 0.122449
intercept = -3.74532
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
the user,and do not appear in the correlation matrix )
background intercept
background 1 -0.62
intercept -0.62 1
Parameter Estimates
95.0% Wald Confidence
Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf.
Limit
background 0.145252 * * *
intercept -3.78793 * * *
slope 1 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log(1 ike~lihood) # Param's Deviance Test d.f. P-value
Full model -118.153 4
Fitted model -119.839 2 3.37005 2 0.1854
Reduced model -135.512 1 34.7167 3 <.0001
AIC: 243.677
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0 . 0000
12 . 8000
32.0000
80 . 0000
0.1453
0.3373
0.5044
0.6960
7.117
16.866
25.218
34.799
6.000
22.000
22.000
34.000
49
50
50
50
Scaled
Residual
-0.453
1.536
-0.910
-0.246
ChiA2 = 3.45 d.f. = 2 P-value = 0.1779
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 4.90719
BMDL = 3.27097
B-9
-------
Table B-5. Benchmark modeling results for bronchiolar hyperplasia in male
B6C3Fi mice (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
192.219
206.147
188.645
203.779
192.219
205.312
192.219
192.219
190.376
Goodness-of-
fit p-value
0.1003
0.0019
0.5085
0.0009
0.1003
0.0023
0.1003
0.1003
1
y? residual
1.561
2.136
0.8
2.278
1.561
2.094
1.561
1.561
-3.22E-06
BMD
9.962
25.582
7.54241
18.0076
9.962
23.8731
9.962
9.962
6.4695
BMDL
7.95025
20.9208
5.60381
12.7086
7.95025
19.6205
7.95025
7.95025
4.24464
aBold indicates model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
T3
ID
ro
LL
0.6
0.5
0.4
0.3
0.2
0.1
Log-Logistic
BMDL
BMD
10
20
30
40
dose
50
60
70
80
14:03 01/14 2010
Figure B-4. Log-logistic model fit for bronchiolar hyperplasia in male
B6C3Fi mice (BMR = 10% extra risk).
B-10
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~ln~l_mouse_m_bronch_hyper_Ln~l-BMRlO-
Restrict.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_mouse_m_bronch_hyper_Ln~l-
BMRlO-Restrict.pIt
Thu Dan 14 14:03:40 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-s~lope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
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
intercept = -4.24694
slope = 1
the
intercept
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept
1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
-4.21777
1
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
- Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log (1 i ke~li hood) # Param's Deviance Test
Full model -92.1882 4
Fitted model -93.3224 1 2.26827
Reduced model -113.552 1 42.7283
Dose
0 . 0000
12 . 8000
32.0000
80 . 0000
AIC:
Est._Prob.
0.0000
0.1586
0.3204
0.5410
188.645
Goodness of Fit
Expected Observed
0.000
7.932
16.019
27.049
0.000
10.000
18.000
23.000
Size
50
50
50
50
d.f. P-valu
3 0.
3 <.0
Scaled
Residual
0.000
0.800
0.600
-1.149
ChiA2 =2.32
d.f. = 3
P-value = 0.5085
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 7.54241
BMDL = 5.60381
B-11
-------
Table B-6. Benchmark modeling results for bronchiolar hyperplasia in male
B6C3Fi mice (BMR = 5% extra risk)
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
192.219
206.147
188.645
203.779
192.219
205.312
192.219
190.376
192.219
Goodness-of-
fit p-value
0.1003
0.0019
0.5085
0.0009
0.1003
0.0023
0.1003
1
0.1003
X2 residual
0
0.897
0
2.278
0
0.992
0
0
0
BMD
4.84986
14.2582
3.57272
12.522
4.84986
13.136
4.84986
3.6667
4.84986
BMDL
3.87047
11.4862
2.65444
8.83728
3.87047
10.6641
3.87047
0.932026
3.87047
aBold indicates model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
T3
ID
0.6
0.5
0.4
0.3
0.2
0.1
70
80
14:04 01/14 2010
Figure B-5. Log-logistic model fit for bronchiolar hyperplasia in male
B6C3Fi mice (BMR = 5% extra risk).
B-12
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~ln~l_mouse_m_bronch_hyper_Ln~l-BMR05-
Restrict.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_mouse_m_bronch_hyper_Ln~l-BMR05-
Restrict.pit
Thu Dan 14 14:04:34 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-s~lope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
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
intercept = -4.24694
slope = 1
the
intercept
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept
1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
-4.21777
1
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
- Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log (1 i ke~li hood) # Param's Deviance Test
Full model -92.1882 4
Fitted model -93.3224 1 2.26827
Reduced model -113.552 1 42.7283
Dose
0 . 0000
12 . 8000
32.0000
80 . 0000
AIC:
Est._Prob.
0.0000
0.1586
0.3204
0.5410
188.645
Goodness of Fit
Expected Observed
0.000
7.932
16.019
27.049
0.000
10.000
18.000
23.000
Size
50
50
50
50
d.f. P-valu
3 0.
3 <.0
Scaled
Residual
0.000
0.800
0.600
-1.149
ChiA2 =2.32
d.f. = 3
P-value = 0.5085
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 3.57272
BMDL = 2.65444
B-13
-------
Table B-7. Benchmark modeling results for olfactory chronic inflammation
in male F344/N rats (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
81.4586
87.0594
81.3682
83.9766
81.4586
86.6596
81.4586
81.4586
Goodness-of-
fit p-value
0.8964
0.0925
0.9398
0.2144
0.8964
0.1067
0.8964
0.8964
X2 residual
0.39
-0.291
0.286
1.458
0.39
-0.345
0.39
0.39
BMD
15.2489
23.8087
14.6428
17.7991
15.2489
22.6768
15.2489
15.2489
BMDL
10.1164
18.9473
9.27776
13.7362
10.1164
17.7855
10.1164
10.1164
aBold indicates model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
T3
ID
0.35
0.3
0.25
0.2
0.15
0.1
0.05
Log-Logistic
BMDL
BMD
10
15
dose
20
25
14:21 01/14 2010
30
Figure B-6. Log-logistic model fit for olfactory chronic inflammation in male
F344/N rats (BMR = 10% extra risk).
B-14
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_inf1ammation_hdd_Ln~l-BMRlO-
Restrict.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_inf1ammation_hdd_Ln~l-BMRlO-
Restrict.pit
Thu Dan 14 14:21:54 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-s~lope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
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
intercept = -4.79799
slope = 1
the
intercept
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept
1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
-4.88117
1
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
- Indicates that this value is not calculated.
Analysis of Deviance Table
Deviance
Model
Full model
Fitted model
Reduced model
AIC:
Log(~li kelihood)
-39.6231
-39.6841
-46.4291
81.3682
# Param's
3
1
1
0.121914
13.6119
Test d.f.
2
2
P-value
0.9409
0.001107
Goodness of Fit
Dose
0 . 0000
12 . 8000
32.0000
Est._Prob.
0.0000
0.0885
0.1954
Expected
0.000
4.426
9.574
Observed
0.000
5.000
9.000
Size
50
50
49
Scaled
Residual
0.000
0.286
-0.207
ChiA2 =0.12
d.f. = 2
P-value = 0.9398
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 14.6428
BMDL = 9.27776
B-15
-------
Table B-8. Benchmark modeling results for olfactory atrophy in male
F344/N rats (BMR =10% extra risk)
Model
Gamma
Logistic"
Log-logistic
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
106.376
105.53
106.376
106.376
107.65
106.283
106.376
125.166
Goodness-of-
fit p-value
NA
0.2655
NA
NA
0.0817
0.1555
NA
0
X2 residual
0
-0.597
0
0
-1.408
-0.901
0
0.459
0
BMD
10.6003
7.70048
10.81
10.9386
6.95763
6.91725
9.95012
2.28431
BMDL
7.99938
5.97454
8.62799
8.79455
5.20262
5.40111
7.06875
1.80011
aBold indicates model choice based on lowest AIC
Logistic Model with 0.95 Confidence Level
T3
ID
ro
LL
0.8
0.6
0.4
0.2
14:19 01/14 2010
30
Figure B-7. Logistic model fit for olfactory atrophy in male F344/N rats
(BMR = 10% extra risk).
B-16
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMRlO.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMRl0.p~lt
Thu Dan 14 14:19:23 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
The form of the probability function is:
P[response] = l/[l+EXP(-intercept-s~lope*dose)]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
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 = -2.84277
slope = 0.164779
the
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept slope
intercept
slope
1
-0.85
-0.85
1
Parameter Estimates
Interval
Variable
Limit
intercept
2.3018
slope
0.230855
Estimate
-3.25094
0.179356
95.0% wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
0.484263 -4.20007
0.0262753 0.127857
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) # Param
-50.1882 3
-50.7651 2
-100.819 1
105.53
's Deviance Test d.f.
P-value
1.15379
101.262
Goodness of Fit
ChiA2 = 1.24
d.f. = 1
P-value = 0.2655
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 7.70048
BMDL = 5.97454
1
2
0.2828
<.0001
Dose
0 . 0000
12 . 8000
32.0000
Est._Prob.
0.0373
0.2778
0.9233
Expected
1.865
13.892
45.243
Observed
3.000
12 . 000
46.000
Size
50
50
49
Scaled
Residual
0.847
-0.597
0.406
B-17
-------
Table B-9. Benchmark modeling results for olfactory atrophy in male
F344/N rats (BMR = 5% extra risk)
Model
Gamma
Logistic"
Log-logistic
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
106.376
105.53
106.376
106.376
107.65
106.283
106.376
125.166
Goodness-of-
fit p-value
NA
0.2655
NA
NA
0.0817
0.1555
NA
0
X2 residual
0
0.847
0
0
0.377
0.977
0
0.459
0
BMD
8.88228
4.90734
9.14915
9.51381
4.85459
4.28231
7.68497
1.11208
BMDL
6.24061
3.52532
6.92292
7.35396
3.12143
3.10069
5.01204
0.87636
aBold indicates model choice based on lowest AIC
Logistic Model with 0.95 Confidence Level
T3
ID
ro
LL
0.8
0.6
0.4
0.2
14:20 01/14 2010
30
Figure B-8. Logistic model fit for olfactory atrophy in male F344/N rats
(BMR = 5% extra risk).
B-18
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMR05.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMR05.p~lt
Thu Dan 14 14:20:15 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
The form of the probability function is:
P[response] = l/[l+EXP(-intercept-s~lope*dose)]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
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 = -2.84277
slope = 0.164779
Asymptotic Correlation Matrix of Parameter Estimates
the
C *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept slope
intercept
slope
Interval
Variable
Limit
intercept
2.3018
slope
0.230855
1
-0.85
-0.85
1
Parameter Estimates
Estimate Std. Err.
-3.25094 0.484263
0.179356 0.0262753
95.0% wald Confidence
Lower Conf. Limit Upper Conf.
-4.20007
0.127857
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) # Param
-50.1882 3
-50.7651 2
-100.819 1
105.53
's Deviance Test d.f.
1.15379
101.262
Goodness of Fit
ChiA2 = 1.24
d.f. = 1
P-value = 0.2655
1
2
P-value
0.2828
<.0001
Dose
0.0000
12 . 8000
32.0000
Est._Prob.
0.0373
0.2778
0.9233
Expected
1.865
13.892
45.243
Observed
3.000
12 . 000
46.000
Size
50
50
49
Scaled
Residual
0.847
-0.597
0.406
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 4.90734
BMDL = 3.52532
B-19
-------
Table B-10. Benchmark modeling results for olfactory necrosis in male
F344/N rats (BMR =10% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
124.435
130.942
124.435
122.499
124.435
129.762
124.435
122.737
Goodness-of-
fit p-value
1
0.0328
1
0.9686
1
0.0494
1
0.8622
X2 residual
0
1.45
0
0.188
0
1.387
0
0
0
BMD
6.46561
12.1684
6.92124
7.98173
5.8893
11.3581
6.31726
4.75407
BMDL
3.70666
9.77545
2.96263
6.41755
3.70666
9.13936
3.70666
3.65317
aBold indicates model choice based on lowest AIC
LogProbit Model with 0.95 Confidence Level
T3
ID
0.7
0.6
0.5
0.4
0.3
0.2
0.1
14:26 01/14 2010
30
Figure B-9. Log-probit model fit for olfactory necrosis in male F344/N rats
(BMR = 10% extra risk).
B-20
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_m_necrosis_hdd_Lnp-BMRlO.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_m_necrosis_hdd_Lnp-BMRl0.p~lt
Thu Dan 14 14:26:59 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+S~lope*Log(Dose)) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Effect
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
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 = -3.33803
slope = 1
the
intercept
Interval
Variable
Limit
background
intercept
3.09743
slope
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept
1
Parameter Estimates
95.0% Wald Confidence
Estimate
0
-3.35871
1
Std. Err.
NA
0.133307
NA
Lower Conf. Limit Upper Conf.
-3.61998
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Deviance Test d.f.
Log(likelihood)
-60.2177
-60.2494
-83.5122
122.499
# Param's
3
1 0.063351 2
1 46.5889 2
Goodness of Fit
P-value
0.9688
<.0001
Dose
0.0000
12 . 8000
32.0000
Est._Prob.
0.0000
0.2092
0.5426
Expected
0.000
10.459
26.588
Observed
0.000
11.000
26.000
Size
50
50
49
Scaled
Residual
0.000
0.188
-0.169
ChiA2 = 0.06 d.f. = 2
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 7.98173
BMDL = 6.41755
P-value = 0.9686
B-21
-------
Table B-ll. Benchmark modeling results for olfactory necrosis in male
F344/N rats (BMR = 5% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
124.435
130.942
124.435
122.499
124.435
129.762
124.435
122.737
Goodness-of-
fit p-value
1
0.0328
1
0.9686
1
0.0494
1
0.8622
X2 residual
0
1.45
0
0
0
1.387
0
0
0
BMD
3.68522
7.60632
4.22667
5.55031
2.97375
7.0625
3.49306
2.31445
BMDL
1.80454
5.7094
1.40335
4.46261
1.80454
5.28703
1.80454
1.7785
aModel indicates model choice based on lowest AIC
LogProbit Model with 0.95 Confidence Level
T3
ID
0.7
0.6
0.5
0.4
0.3
0.2
0.1
14:27 01/14 2010
30
Figure B-10. Log-probit model fit for olfactory necrosis in male F344/N rats
(BMR = 5% extra risk).
B-22
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_m_necrosis_hdd_Lnp-BMR05.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_m_necrosis_hdd_Lnp-BMR05.p~lt
Thu Dan 14 14:27:40 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+S~lope*Log(Dose)) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Effect
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
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 = -3.33803
slope = 1
the
intercept
Interval
Variable
Limit
background
intercept
3.09743
slope
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept
1
Parameter Estimates
95.0% Wald Confidence
Estimate
0
-3.35871
1
Std. Err.
NA
0.133307
NA
Lower Conf. Limit Upper Conf.
-3.61998
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) # Param's Deviance Test d.f.
-60.2177 3
-60.2494 1 0.063351 2
-83.5122 1 46.5889 2
P-value
0.9688
<.0001
122.499
Goodness of Fit
Dose
0.0000
12 . 8000
32.0000
Est._Prob.
0.0000
0.2092
0.5426
Expected
0.000
10.459
26.588
Observed
0.000
11.000
26.000
Size
50
50
49
Scaled
Residual
0.000
0.188
-0.169
ChiA2 = 0.06
d.f. = 2
P-value = 0.9686
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 5.55031
BMDL = 4.46261
B-23
-------
Table B-12. Benchmark modeling results for olfactory necrosis in female
F344/N rats (BMR =10% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hillb
AIC
108.455
114.403
108.312
106.815
108.87
113.454
108.549
106.909
103.075
Goodness-of-fit
p-value
0.1223
0.0069
0.1357
0.115
0.1361
0.0095
0.1241
0.2879
1
X2 residual
1.54
2.564
1.469
2
1.339
2.504
1.509
1.188
1.13E-05
BMD
33.1378
50.8598
32.4911
35.6629
31.2054
47.668
32.8886
29.5366
30.221
BMDL
21.4781
41.8856
20.1388
28.3477
20.9166
38.935
21.3452
20.8661
27.5059
aBold indicates selected model.
bDichotomous Hill model has lowest AIC value, but two of its parameters were estimated at their respective bounds,
and the resulting model fit was highly suspect upon visual inspection. The model output warned that the BMDL
calculation was "at best imprecise for these data." Therefore, the model with the next lowest AIC (i.e., the log-probit)
model was selected.
LogProbit Model with 0.95 Confidence Level
T3
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_necrosis_Lnp-BMRlO.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_necrosis_Lnp-BMRl0.p~lt
Thu Dan 14 14:16:20 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+S~lope*Log(Dose)) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Effect
Independent variable = Dose
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 (and Specified) Parameter Values
background = 0
intercept = -4.83555
slope = 1
the
intercept
Variable
Limit
background
intercept
4.57667
slope
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept
1
Parameter Estimates
Estimate Std. Err.
0 NA
-4.85566 0.142346
1 NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf.
-5.13466
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model Log (likelihood) # Param's Deviance Test
Full model -49.5375 4
Fitted model -52.4076 1 5.74025
Reduced model -64.911 1 30.7469
AIC: 106.815
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0 . 0000
12 . 8000
32.0000
80 . 0000
0.0000
0.0105
0.0823
0.3179
0.000
0.527
4.114
15.894
0.000
0.000
8.000
12 . 000
49
50
50
50
d.f. P-value
3 0.125
3 <.0001
Scaled
Residual
0.000
-0.730
2.000
-1.183
ChiA2 = 5.93
d.f. = 3
P-value = 0.1150
0.1
Benchmark Dose Computation
Specified effect =
Risk Type = Extra risk
Confidence level = 0.95
35.6629
BMD =
BMDL =
28.3477
B-25
-------
Dichotomous-Hill Model with 0.95 Confidence Level
T3
ID
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
Dichotomous-Hill
BMDL
BMD
10
20
30
40
dose
50
60
70
80
14:16 01/14 2010
Figure B-12. Dichotomous Hill model fit for olfactory necrosis in female
F344/N rats (BMR =10% extra risk).
Dichotomous Hill Model. (Version: 1.0; Date: 09/24/2006)
Input Data File: M:\Ch~loroprene\NTP_BMDS\dh~l_rat_f_necrosis_Dh~l-BMRlO-Restrict.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\dh~l_rat_f_necrosis_Dh~l-BMRlO-
Restrict.pit
Thu Dan 14 14:16:22 2010
BMDS Model Run
The form of the probability function is:
P[response] = v*g +(v-v*g)/[l+EXP(-intercept-s~lope*Log(dose))]
where: 0 <= g < 1, 0= 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
Default Initial Parameter Values
v = -9999
g = -9999
intercept = -9.02343
slope = 1.88938
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -g -slope
B-26
-------
the
v
intercept
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
v intercept
1
-0.61
-0.61
1
Parameter Estimates
Interval
Variable
Limit
v
0.35838
g
intercept
59.2777
slope
Estimate
0.24
0
-61.6901
18
Std. Err.
0.0603988
NA
1.23084
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
0.121621
-64.1025
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(likelihood) Deviance Test d.f. P-value
-49.5375
-49.5375 3.29863e-006 2 1
-64.911 30.7469 3 <.0001
AIC: 103.075
Goodness of Fit
Dose
0 . 0000
12 . 8000
32.0000
80 . 0000
Est._Prob.
0.0000
0.0000
0.1600
0.2400
Expected
0.000
0.000
8.000
12 . 000
Observed
0
0
8
12
Size
49
50
50
50
Scaled
Residual
0
-0.001284
1.131e-005
-2.788e-006
ChiA2 = 0.000002
d.f. = 2
P-value = 1.0000
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 30.221
Warning: BMDL computation is at best imprecise for these data
BMDL =
27.5059
B-27
-------
Table B-13. Benchmark modeling results for olfactory necrosis in female
F344/N rats (BMR = 5% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hillb
AIC
108.455
114.403
108.312
106.815
108.87
113.454
108.549
106.909
103.075
Goodness-of-
fit p-value
0.1223
0.0069
0.1357
0.115
0.1361
0.0095
0.1241
0.2879
1
X2 residual
-1.256
2.564
-1.263
2
-1.458
2.504
-1.299
-1.528
1.13E-05
BMD
18.8703
34.1134
18.6016
24.7991
15.6751
31.4159
18.2634
14.3795
28.5901
BMDL
10.4563
26.6532
9.53944
19.7123
10.1829
24.4034
10.3916
10.1584
26.0761
aBold indicates selected model.
bDichotomous Hill model has lowest AIC value, but two of its parameters were estimated at their respective
bounds, and the resulting model fit was highly suspect upon visual inspection. The model output warned that the
BMDL calculation was "at best imprecise for these data." Therefore, the model with the next lowest AIC (i.e.,
the log-probit) model was selected.
LogProbit Model with O.95 Confidence Level
T3
«
0
ro
LL
O.4
O.35
O.3
O.25
O.2
O.15
O.1
O.O5
7O
SO
14:17 O1/14 2O1O
Figure B-13. Log-probit model fit for olfactory necrosis in female F344/N
rats (BMR = 5% extra risk).
B-28
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_necrosis_Lnp-BMR05.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_necrosis_Lnp-BMR05.p~lt
Thu Dan 14 14:17:01 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+S~lope*Log(Dose)) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Effect
Independent variable = Dose
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 (and Specified) Parameter Values
background = 0
intercept = -4.83555
slope = 1
the
intercept
Interval
Variable
Limit
background
intercept
4.57667
slope
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept
1
Parameter Estimates
95.0% Wald Confidence
Estimate
0
-4.85566
1
Std. Err.
NA
0.142346
NA
Lower Conf. Limit Upper Conf.
-5.13466
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model Log (likelihood) # Param's Deviance Test
Full model -49.5375 4
Fitted model -52.4076 1 5.74025
Reduced model -64.911 1 30.7469
AIC: 106.815
Goodness of Fit
Dose
0 . 0000
12 . 8000
32.0000
80 . 0000
Est._Prob.
0.0000
0.0105
0.0823
0.3179
Expected
0.000
0.527
4.114
15.894
Observed
0.000
0.000
8.000
12 . 000
Size
49
50
50
50
d.f. P-value
3 0.125
3 <.0001
Scaled
Residual
0.000
-0.730
2.000
-1.183
ChiA2 = 5.93
d.f. = 3
P-value = 0.1150
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 24.7991
BMDL = 19.7123
B-29
-------
Table B-14. Benchmark modeling results for olfactory basal cell hyperplasia
in female F344/N rats (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hillb
AIC
78.8231
83.7749
78.3698
78.083
85.0835
83.6185
81.3562
120.402
77.9075
Goodness-of-
fit p-value
0.7502
0.0615
0.8754
0.9521
0.155
0.1092
0.3487
0
1
X2 residual
0.324
0.81
0.12
0.093
-1.956
-1.283
-1.19
0
5.70E-07
BMD
22.6607
22.7096
24.0671
23.4933
15.3009
22.0007
20.7516
5.59788
29.3724
BMDL
18.6776
18.8101
20.2672
19.7198
13.2469
17.8681
16.5638
4.52837
23.7917
aBold indicates selected model.
bDichotomous Hill model has lowest AIC value, but two of its parameters were estimated at their respective
bounds, and the resulting model fit was highly suspect upon visual inspection. The model output warned that the
BMDL calculation was "at best imprecise for these data." Therefore, the model with the next lowest AIC (i.e.,
the log-probit) model was selected.
LogProbit Model with 0.95 Confidence Level
T3
%
ID
0.8
0.6
0.4
0.2
70
80
14:14 01/14 2010
Figure B-14. Log-probit model fit for olfactory basal cell hyperplasia in
female F344/N rats (BMR = 10% extra risk).
B-30
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_basa~l_hyper_Lnp-BMRlO.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_basa~l_hyper_Lnp-BMRl0.p~lt
Thu Dan 14 14:14:05 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+S~lope*Log(Dose)) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Effect
Independent variable = Dose
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 (and Specified) Parameter Values
background = 0
intercept = -8.5284
slope = 2.39417
the
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
intercept slope
intercept
slope
Interval
Variable
Limit
background
intercept
6.758
slope
3.65254
1
-0.99
-0.99
1
Parameter Estimates
Estimate
0
-9.9865
2.7576
Std. Err.
NA
1.64723
0.456613
95.0% wald Confidence
Lower Conf. Limit Upper Conf.
-13.215
1.86265
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Deviance Test d.f.
Log(likelihood)
-36.9537
-37.0415
-126.434
78.083
# Param's
4
2
1
P-value
0.175584
178.961
0.916
<.0001
Dose
0.0000
12 . 8000
32.0000
80 . 0000
Est._Prob.
0.0000
0.0016
0.3338
0.9820
Goodness of Fit
Expected Observed Size
0.000
0.078
16.691
49.101
0.000
0.000
17.000
49.000
49
50
50
50
Scaled
Residual
0.000
-0.279
0.093
-0.107
ChiA2 =0.10
d.f. = 2
P-value = 0.9521
B-31
-------
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 23.4933
BMDL = 19.7198
T3
%
ID
ro
LL
0.8
0.6
0.4
0.2
Dichotomous-Hill Model with 0.95 Confidence Level
Dichotomous-Hill
BMDL
BMD
10
20
30
40
dose
50
60
70
80
14:14 01/14 2010
Figure B-15. Dichotomous Hill model fit for olfactory basal cell hyperplasia
in female F344/N rats (BMR = 10% extra risk).
Dichotomous Hill Model. (Version: 1.0; Date: 09/24/2006)
Input Data File: M:\Ch1oroprene\NTP_BMDS\dh~l_rat_f_basa~l_hyper_Dh~l-BMRlO-Restrict.Cd)
Gnupl9t Plotting File: M:\Ch~loroprene\NTP_BMDS\dh~l_rat_f_basa~l_hyper_Dh~l-BMRlO-
Restrict.plt
Thu Dan 14 14:14:07 2010
BMDS Model Run
x/'s/'s/'s/'s/'s/'s/'s/'s/'s/'s/'s/'s/'s/'s/'s/'s/'^
The form of the probability function is:
P[response] = v*g +(v-v*g)/[l+EXP(-intercept-s~lope*Log(dose))]
where: 0 <= g < 1, 0= 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
Default Initial Parameter Values
v = -9999
g = -9999
intercept = -16.5503
slope = 4.64205
B-32
-------
the
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -g -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
v intercept
intercept
1
-0.1
-0.1
1
Parameter Estimates
Interval
Variable
Limit
V
1.01881
g
intercept
62.4213
slope
Estimate
0.98
0
-63.0158
18
Std
0.
0.
. Err.
019799
NA
303295
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
0.941195
-63.6102
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(~li kelihood)
-36.9537
-36.9537
-126.434
77.9075
Goodness of
Dose
0.0000
12 . 8000
32.0000
80 . 0000
Est
0
0
0
0
._Prob
.0000
.0000
.3400
.9800
Deviance Test
3.57771e-006
178.961
Fit
Expected
0
0
17
49
.000
.000
.000
.000
Observed
0
0
17
49
d.f. P-value
2 1
3 <.0001
Size
49
50
50
50
Scaled
Residual
0
-0.001337
5.704e-007
-1.192e-007
ChiA2 = 0.000002
d.f. = 2
P-value = 1.0000
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 29.3724
Warning: BMDL computation is at best imprecise for these data
BMDL =
23.7917
B-33
-------
Table B-15. Benchmark modeling results for olfactory basal cell hyperplasia
in female F344/N rats (BMR = 5% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hillb
AIC
78.8231
83.7749
78.3698
78.083
85.0835
83.6185
81.3562
120.402
77.9075
Goodness-of-
fit p-value
0.7502
0.0615
0.8754
0.9521
0.155
0.1092
0.3487
0
1
X2 residual
-0.597
-1.253
-0.448
-0.279
-1.956
-1.283
-1.19
0
5.70E-07
BMD
19.1684
17.4935
20.879
20.5934
10.676
16.5716
15.9699
2.72525
28.1762
BMDL
15.0243
13.1784
16.769
16.7154
8.74893
12.2502
12.0527
2.20457
22.8227
aBold indicates selected model.
bDichotomous Hill model has lowest AIC value, but two of its parameters were estimated at their respective
bounds, and the resulting model fit was highly suspect upon visual inspection. The model output warned that the
BMDL calculation was "at best imprecise for these data." Therefore, the model with the next lowest AIC (i.e.,
the log-probit) model was selected.
LogProbit Model with 0.95 Confidence Level
T3
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_basa~l_hyper_Lnp-BMR05.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_basa~l_hyper_Lnp-BMR05.p~lt
Thu Dan 14 14:14:47 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
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 = Effect
Independent variable = Dose
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 (and Specified) Parameter Values
background = 0
intercept = -8.5284
slope = 2.39417
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the
intercept
slope
user,and do not appear in the correlation matrix )
intercept slope
1
-0.99
-0.99
1
Parameter Estimates
Interval
Variable
Limit
background
intercept
6.758
slope
3.65254
Estimate
0
-9.9865
2.7576
Std. Err.
NA
1.64723
0.456613
95.0% wald Confidence
Lower Conf. Limit Upper Conf.
-13.215
1.86265
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
Model Log (likelihood) # Param's Deviance Test
Full model -36.9537 4
Fitted model -37.0415 2 0.175584
Reduced model -126.434 1 178.961
AIC: 78.083
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0.0000 0.0000 0.000 0.000 49
12.8000 0.0016 0.078 0.000 50
32.0000 0.3338 16.691 17.000 50
80.0000 0.9820 49.101 49.000 50
d.f. P-va~lui
2 0
3 <.0i
Scaled
Residual
0.000
-0.279
0.093
-0.107
0.916
ChiA2 =0.10
d.f. = 2
P-value = 0.9521
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 20.5934
BMDL = 16.7154
B-35
-------
Table B-16. Benchmark modeling results for kidney (renal tubule)
hyperplasia in male F344/N rats (BMR =10% extra risk)
Model
Gamma
Logistic
Log-logistic"
Log-probit
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
262.742
263.873
262.083
264.054
262.742
263.882
262.742
262.742
20090.6
Goodness-of-
fit p-value
0.6482
0.3712
0.9017
0.3356
0.6482
0.3695
0.6482
0.6482
X2 residual
0.091
0.128
-0.136
0.487
0.091
0.131
0.091
0.091
0
BMD
9.58982
14.4291
6.52869
17.4209
9.58986
14.3921
9.58986
9.58986
BMDL
6.61749
11.0906
3.95681
11.9381
6.61749
11.164
6.61749
6.61749
aBold indicates model choice based on lowest BMDL.
Log-Logistic Model with 0.95 Confidence Level
T3
0.8
0.7
0.6
0.5
0.4
0.3
0.2
Log-Logistic
: BMDL
BMD
10
20
30
40
dose
50
60
70
80
14:22 01/14 2010
Figure B-17. Log-logistic model fit for kidney (renal tubule) hyperplasia in
male F344/N rats (BMR = 10% extra risk).
B-36
-------
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_kid_hyper_Ln~l-BMRlO-Restrict.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_kid_hyper_Ln~l-BMRlO-
Restrict.pit
Thu Dan 14 14:22:37 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-intercept-s~lope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
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 = 0.28
intercept = -4.0785
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
C *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
the
user,and do not appear in the correlation matrix )
background intercept
background 1 -0.65
intercept -0.65 1
Parameter Estimates
95.0% wald Confidence
Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf.
Limit
background 0.280771 * * *
intercept -4.07343 * * *
slope 1 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model Log (likelihood) # Param's Deviance Test
Full model -128.938 4
Fitted model -129.042 2 0.206766
Reduced model -138.469 1 19.0624
AIC: 262.083
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0 . 0000
12 . 8000
32.0000
80 . 0000
0.2808
0.4094
0.5344
0.6954
14.039
20.471
26.718
34.772
14 . 000
20.000
28.000
34.000
50
50
50
50
d.f. P-value
2 0.9018
3 0.0002654
Scaled
Residual
-0.012
-0.136
0.363
-0.237
ChiA2 = 0.21 d.f. = 2 P-value = 0.9017
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 6.52869
BMDL = 3.95681
B-37
-------
Table B-17. Benchmark modeling results for kidney (renal tubule)
hyperplasia in female F344/N rats (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
200.059
198.217
200.048
197.991
200.173
198.217
200.09
198.787
202.036
Goodness-of-
fit p-value
0.6468
0.8346
0.656
0.9302
0.5712
0.8345
0.6245
0.6299
NA
X2 residual
0.236
0.29
0.211
0.269
0.302
0.222
0.244
-0.692
0.1984
BMD
31.181
31.33
30.79
32.5323
31.846
29.6902
31.13
21.0465
30.4841
BMDL
14.892
24.9474
13.2994
23.5182
14.7635
23.4384
14.8568
14.1492
12.4518
aBold indicates model choice based on lowest AIC.
LogProbit Model with 0.95 Confidence Level
T3
ID
ro
LL
0.6
0.5
0.4
0.3
0.2
0.1
LogProbit
BMDL
BMD
10
20
30
40
dose
50
60
14:15 01/14 2010
70
80
Figure B-18. Log-probit model fit for kidney (renal tubule) hyperplasia in
female F344/N rats (BMR = 10% extra risk).
B-38
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_kid_hyper_Lnp-BMRlO.Cd)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_kid_hyper_Lnp-BMRl0.p~lt
Thu Dan 14 14:15:37 2010
BMDS Model Run
NJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/SJ/^
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 = Effect
Independent variable = Dose
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 (and Specified) Parameter Values
background = 0.122449
intercept = -4.95177
slope = 1.04703
the
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
user,and do not appear in the correlation matrix )
background intercept
background
intercept
1
-0.53
-0.53
1
Parameter Estimates
Interval
Variable Estimate Std. Err.
Limit
background 0.119059 0.0336048
0.184924
intercept -4.76379 0.218134
4.33625
slope 1 NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
95.0% wald Confidence
Lower Conf. Limit Upper Conf.
0.0531949
-5.19132
Model Log (likelihood) # Param's Deviance Test
Full model -96.9233 4
Fitted model -96.9957 2 0.144742
Reduced model -105.132 1 16.4183
AIC: 197.991
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0 . 0000
12 . 8000
32.0000
80 . 0000
0.1191
0.1309
0.2046
0.4286
5.834
6.543
10.231
21.428
6.000
6.000
11.000
21.000
49
50
50
50
d.f. P-valu
2 0.
3 0.000
Scaled
Residual
0.073
-0.228
0.269
-0.122
ChiA2 =0.14
d.f. = 2
P-value = 0.9302
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 32.5323
BMDL = 23.5182
B-39
-------
Table B-18. Benchmark modeling results for forestomach epithelial
hyperplasia in male B6C3Fi mice (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
178.784
177.328
178.762
178.805
177.268
177.716
178.737
182.602
5853.85
Goodness-of-
fit p-value
0.4716
0.5986
0.4806
0.464
0.6124
0.5004
0.491
0.0523
X2 residual
-0.049
-0.825
-0.065
-0.013
-0.771
-0.984
-0.095
-0.389
0
BMD
39.6884
26.8011
39.6607
39.3506
30.167
24.635
39.8723
13.9921
BMDL
20.1391
22.1839
21.1117
22.7348
16.9463
20.4757
19.6367
10.3765
aBold indicates model choice based on lowest AIC.
Multistage Model with 0.95 Confidence Level
T3
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Multistage
BMDL
BMD
10
20
30
40
dose
50
60
70
80
14:05 01/14 2010
Figure B-19. Multistage model fit for forestomach epithelial hyperplasia in
male B6C3Fi mice (BMR = 10% extra risk).
B-40
-------
Multistage ivtodel. (Version: 3.0; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\mst_mouse_m_fore_hyper_Mst-BMRlO-
Restrict.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\mst_mouse_m_fore_hyper_Mst-BMRlO-
Restrict.pit
Thu Dan 14 14:05:47 2010
BMDS Model Run
The form of the probability function is:
P[response] = background + (l-background)*[l-EXP(-betal*doseAl-beta2*doseA2)]
The parameter betas are restricted to be positive
Dependent variable = Effect
Independent variable = Dose
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 = 0.0745999
Beta(l) = 0
Beta(2) = 0.00012236
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
-0.48
Beta(2)
-0.48
1
Parameter Estimates
Interval
Variable Estimate Std. Err.
Limit
Background 0.0832204 *
Beta(1) 0 *
Beta(2) 0.000115775 *
* - Indicates that this value is not calculated.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(likelihood) # Param's Deviance Test d.f.
-86.1337 4
P-value
-86.6341
-107.064
177.268
2
1
1.00079
41.8613
0.6063
<.0001
Dose
0.0000
12 . 8000
32.0000
80.0000
Est._Prob.
0.0832
0 . 1004
0.1857
0.5630
Goodness of Fit
Expected Observed Size
4.161
4.821
9.100
28.151
4.000
6.000
7.000
29.000
50
48
49
50
Scaled
Residual
-0.082
0.566
-0.771
0.242
ChiA2 = 0.98
d.f. = 2
P-value = 0.6124
Benchmark Dose Computation
Specified effect =
Risk Type = Extra
0.1
risk
Confidence level = 0.95
BMD = 30.167
BMDL = 16.9463
BMDU = 36.6564
Taken together, (16.9463, 36.6564)
is a 90 % two-sided confidence interval for the BMD
B-41
-------
Table B-19. Benchmark modeling results for forestomach epithelial
hyperplasia in female B6C3Fi mice (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
169.362
167.998
169.384
169.261
167.53
168.273
169.457
173.528
5845.73
Goodness-of-
fit p-value
0.5864
0.6241
0.5729
0.6545
0.7935
0.5353
0.5344
0.0476
X2 residual
0.147
-0.02
0.142
0.071
-0.013
-0.198
0.19
-1.415
0
BMD
33.02
29.3493
32.8973
32.4471
30.9965
26.9397
33.3943
15.4655
BMDL
19.9556
24.2933
20.1355
20.7798
19.3466
22.3632
19.6657
11.4268
"Bold indicates model choice based on lowest AIC.
Multistage Model with 0.95 Confidence Level
T3
ID
0.7
0.6
0.5
0.4
0.3
0.2 -
0.1
13:58 01/14 2010
70
80
Figure B-20. Multistage model fit for forestomach epithelial hyperplasia in
female B6C3Fi mice (BMR = 10% extra risk).
B-42
-------
Multistage ivtodel. (Version: 3.0; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\mst_mouse_f_fore_hyper_Mst-BMRlO-
Restrict.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\mst_mouse_f_fore_hyper_Mst-BMRlO-
Restrict.pit
Thu Dan 14 13:58:16 2010
BMDS Model Run
The form of the probability function is:
P[response] = background + (l-background)*[l-EXP(-betal*doseAl-beta2*doseA2)]
The parameter betas are restricted to be positive
Dependent variable = Effect
Independent variable = Dose
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 = 0.0623808
Beta(l) = 0
Beta(2) = 0.00011119
Asymptotic Correlation Matrix of Parameter
( *** The model parameter(s) -Beta(l)
have been estimated at a boundary point,
the
Estimates
or have been specified by
user,and do not appear in the correlation matrix )
Background Beta(2)
Background 1 -0.5
Beta(2) -0.5 1
Parameter Estimates
95.0% Wald
Interval
Variable Estimate Std. Err.
Limit
Confidence
Lower Conf. Limit Upper Conf.
Background
Beta(l)
Beta(2)
0.0645849
0
0.000109661
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood)
-81.5287
-81.7648
-102.317
167.53
# Param's
4
2
1
Deviance Test d.f.
P-value
0.472098
41.577
0.7897
<.0001
Dose
0.0000
12 . 8000
32.0000
80.0000
Est._Prob.
0.0646
0.0812
0.1639
0.5363
Goodness of Fit
Expected Observed Size
3.229
3.981
8.033
26.817
4.000
3.000
8.000
27.000
50
49
49
50
Scaled
Residual
0.443
-0.513
-0.013
0.052
ChiA2 = 0.46
d.f. = 2
P-value = 0.7935
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level =
BMD =
BMDL =
BMDU =
Taken together, (19.3466,
0.95
30.9965
19.3466
37.6172
37.6172) is a 90% two-sided confidence interval for the BMD
B-43
-------
Table B-20. Benchmark modeling results for splenic hematopoietic cell
proliferation in female B6C3F! mice (BMR = 10% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
171.405
169.421
171.405
171.405
171.405
169.41
171.405
170.771
Goodness-of-
fit p-value
NA
0.8993
NA
NA
NA
0.9466
NA
0.2455
X2 residual
0
0.064
0
0
0
0.033
0
0.264
0
BMD
5.73584
4.06642
6.5828
6.91076
4.41391
4.03306
5.17994
2.34557
BMDL
1.90919
3.28512
2.43228
3.48982
1.90919
3.33147
1.90919
1.7616
aBold indicates model choice based on lowest AIC.
Probit Model with 0.95 Confidence Level
T3
ID
0.3
0.2
0.1
14:02 01/14 2010
30
Figure B-21. Probit model fit for splenic hematopoietic cell proliferation in
female B6C3Fi mice (BMR = 10% extra risk).
B-44
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\pro_mouse_f_sp1een_hemato_hdd_Pro-BMRlO.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\pro_mouse_f_sp1een_hemato_hdd_Pro-
BMRlO.pIt
Thu Dan 14 14:02:47 2010
BMDS Model Run
The form of the probability function is:
P[response] = CumNorm(Intercept+S~lope*Dose) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Effect
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
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 (and Specified) Parameter Values
background = 0 Specified
intercept = -0.643083
slope = 0.0528681
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
slope
if you
Interval
Variable
Limit
intercept
0.324476
slope
0.0713925
Model
Full model
Fitted model
Reduced model
AIC:
1
-0.73
-0.73
1
Parameter Estimates
95.0% wald Confidence
Estimate
-0.649733
0.0534876
Std. Err.
0.16595
0.00913534
Lower Conf. Limit
-0.97499
0.0355826
Upper Conf.
Analysis of Deviance Table
Deviance
Log(~li kelihood)
-82.7026
-82.7048
-102.099
169.41
# Param's Deviance Test d.f. P-value
3
2 0.00449095 1 0.9466
1 38.7924 2 <.0001
Dose
0 . 0000
12 . 8000
32.0000
Est._Prob.
0.2579
0.5139
0.8559
Goodness of Fit
Expected Observed Size
12.897
25.182
41.937
13.000
25.000
42.000
50
49
49
Scaled
Residual
0.033
-0.052
0.026
ChiA2 = 0.00
d.f. = 1
P-value = 0.9466
0.1
Benchmark Dose Computation
Specified effect =
Risk Type = Extra risk
Confidence level = 0.95
BMD = 4.03306
BMDL = 3.33147
B-45
-------
Table B-21. Benchmark modeling results for splenic hematopoietic cell
proliferation in female B6C3Fi mice (BMR = 5% extra risk)
Model
Gamma
Logistic
Log-logistic
Log-probita
Multistage
Probit
Weibull
Quantal-linear
Dichotomous Hill
AIC
171.405
169.421
171.405
171.405
171.405
169.41
171.405
170.771
Goodness-of-
fit p-value
NA
0.8993
NA
NA
NA
0.9466
NA
0.2455
X2 residual
0
0.064
0
0
0
0.033
0
0.264
0
BMD
3.86036
2.10908
4.75284
5.33285
2.35161
2.07526
3.21493
1.14191
BMDL
0.929461
1.68891
1.34665
2.42674
0.929461
1.70339
0.929461
0.85761
aBold indicates model choice based on lowest AIC.
Probit Model with 0.95 Confidence Level
T3
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
12:34 01/21 2010
30
Figure B-22. Probit model fit for splenic hematopoietic cell proliferation in
female B6C3Fi mice (BMR = 5% extra risk).
B-46
-------
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\pro_mouse_f_sp1een_hemato_hdd_Pro-BMR05.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\pro_mouse_f_sp1een_hemato_hdd_Pro-
BMROS.pIt
Thu Dan 21 12:34:25 2010
BMDS Model Run
The form of the probability function is:
P[response] = CumNorm(Intercept+S~lope*Dose) ,
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Effect
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 3
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 (and Specified) Parameter Values
background = 0 Specified
intercept = -0.643083
slope = 0.0528681
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background
have been estimated at a boundary point, or have been specified by
the
intercept
slope
user,and do not appear in the correlation matrix )
intercept
1
-0.73
slope
-0.73
1
Parameter Estimates
Interval
Variable
Limit
intercept
0.324476
slope
0.0713925
Estimate
-0.649733
0.0534876
Std. Err.
0.16595
0.00913534
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-0.97499
0.0355826
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
LogO i kelihood)
-82.7026
-82.7048
-102.099
169.41
# Param's
3
2
1
Deviance Test d.f.
P-value
0.00449095
38.7924
ChiA2 = 0.00
d.f. = 1
P-value = 0.9466
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 2.07526
BMDL = 1.70339
1
2
0.9466
<.0001
Dose
0 . 0000
12 . 8000
32.0000
Est._Prob.
0.2579
0.5139
0.8559
Goodness of Fit
Expected Observed Size
12.897
25.182
41.937
13.000
25.000
42.000
50
49
49
Scaled
Residual
0.033
-0.052
0.026
B-47
-------
APPENDIX C. CANCER DOSE-RESPONSE MODELING
Table C-l. Tumor incidence, with time to death with tumor: female mice exposed
to chloroprene via inhalation
Dose
Group
0
12.8
Week
of
Study
5
69
70
71
76
78
88
91
95
97
98
101
105
41
46
63
64
69
75
76
78
79
87
89
90
91
97
98
99
100
101
102
103
105
Total
examined
2
2
36
1
2
1
1
1
1
1
1
o
J
1
2
1
o
J
3
1
5
1
1
2
2
16
Number Of Female Animals With Tumors At Each Site, At Specified Week Of Study
Lung
0
0
0
0
0
0
0
0
0
0
0
1
3
0
ob
0
1
0
1
0
0
0
0
2
0
2
2
1
4
0
1
2
1
11
Hemangiomas,
Hemangiosarcomas
Incid.a
0
0
0
0
0
0
0
0
0
0
0
0
3
0
ob
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
Fatal3
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
0
Harderian
Glandc
0
0
0
0
1
0
0
0
0
0
0
0
1
0
ob
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
3
Mammary
0
0
0
0
1
0
0
0
0
0
1
0
lb
0
ob
1
1
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
1
1
Forestomach
0
0
0
0
0
0
0
0
0
0
0
0
1
0
ob
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Liver
0
0
0
1
0
0
0
1
0
0
0
0
18
0
ob
1
0
0
1
1
0
1
1
1
1
0
2
0
2
0
1
1
2
11
Skin
0
0
ob
0
0
0
0
0
0
0
0
0
0
1
ob
1
0
0
0
0
0
0
1
0
1
1
1
0
2
0
0
0
1
2
Zymbal
Glandc
0
0
0
0
0
0
0
0
0
0
0
0
0
0
ob
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Note: Hyperlinks to the reference citations throughout this document will take you to the NCEA HERO database (Health and Environmental Research
Online) at http://epa.gov/hero. HERO is a database of scientific literature used by U.S. EPA in the process of developing science assessments such as the
Integrated Science Assessments (ISA) and the Integrated Risk Information System (IRIS).
C-1
-------
Dose
Group
32
Week
of
Study
31
50
54
56
57
61
63
65
67
68
70
72
73
74
75
76
77
78
79
82
84
86
87
89
90
91
92
93
94
96
97
99
103
105
Total
examined
2
2
2
2
1
2
1
2
1
o
J
2
o
5
o
3
i
i
3
2
1
1
1
1
Number Of Female Animals With Tumors At Each Site, At Specified Week Of Study
Lung
0
1
0
0
0
0
1
1
0
0
0
2
0
1
1
2
1
0
2
1
1
1
2
2
3
3
1
1
3
1
1
0
1
1
Hemangiomas,
Hemangiosarcomas
Incid."
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
Fatal3
1
0
0
1
0
0
1
0
0
0
1
1
0
1
0
1
1
0
2
0
1
0
1
2
0
1
0
0
0
0
0
1
0
0
Harderian
Glandc
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
Mammary
0
0
0
0
1
1
0
0
0
0
0
1
0
0
1
1
0
0
1
1
0
0
1
1
0
0
1
0
1
0
0
0
0
0
Forestomach
0
0
0
0
0
0
0
ob
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Liver
0
0
0
0
0
0
0
0
0
0
0
1
1
1
0
1
0
1
0
0
0
1
1
1
0
3
0
1
2
2
1
1
1
1
Skin
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
1
0
1
0
1
1
0
1
1
1
1
0
0
1
Zymbal
Glandc
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
C-2
-------
Dose
Group
80
Week
of
Study
1
36
47
48
55
64
65
66
67
70
75
76
77
79
81
83
84
86
87
88
90
91
92
93
94
95
96
97
98
105
Total
examined
2
1
4
2
o
J
2
1
7
1
2
1
2
1
2
1
3
Number Of Female Animals With Tumors At Each Site, At Specified Week Of Study
Lung
0
0
1
0
0
0
1
1
1
1
4
2
0
1
1
3
1
1
0
2
2
7
1
2
1
2
1
2
1
3
Hemangiomas,
Hemangiosarcomas
Incid."
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
1
Fatal3
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
1
0
0
0
0
0
2
0
0
0
0
0
0
0
0
Harderian
Glandc
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
1
1
0
2
0
0
1
0
0
0
0
2
Mammary
0
0
0
1
1
0
1
0
0
1
1
0
0
1
0
0
1
1
0
1
0
4
0
0
0
0
0
0
0
1
Forestomach
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
1
1
0
0
0
0
0
0
0
Liver
0
0
0
0
1
0
1
0
2
0
1
1
1
1
0
1
1
1
0
2
1
3
1
2
1
2
1
2
1
3
Skin
0
0
0
0
0
0
0
1
2
0
2
1
0
0
0
2
0
0
1
1
1
4
0
1
0
0
0
1
1
0
Zymbal
Glandc
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
a"Incid.," or Incidental, denotes tumors not concluded to have caused the death of the animal. Fatal denotes tumors
considered to have caused the death of the animal.
bTissue for one animal of total examined was missing or unsuitable for histopathologic examination.
cHarderian gland and Zymbal's gland were examined histopathologically only if a lesion was observed grossly at
necropsy; instances of "0" for these tissues indicate only that no tumor was seen grossly, for dose-response modeling
purposes.
Source: NTP (1998, 042076).
C-3
-------
Table C-2. Tumor incidence, with time to death with tumor: male mice exposed to
chloroprene via inhalation
Dose
Group
0
12.8
Week
of
Study
65
77
79
82
86
87
90
91
92
95
96
97
98
103
104
105
63
75
76
78
83
84
86
87
88
90
91
92
95
96
98
99
101
102
104
105
Total
Examined
3
2
2
o
J
29
2
2
3
27
Number Of Male Animals With Tumors At Each Site, At Specified Week Of Study
Hemangiomas,
Hemangiosarcomas
Incid.a
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
o
J
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8
Fatal3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
3
0
1
0
0
Lung
0
0
0
0
0
0
0
1
0
1
0
1
1
0
0
9
0
0
1
0
0
0
0
1
0
0
1
1
0
0
1
1
1
0
1
20
Forestomach
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
Ob
Ob
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Harderian
gland0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
4
Kidney
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Ob
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
C-4
-------
Dose
Group
32
Week
of
Study
55
63
68
71
72
78
79
81
83
86
87
89
90
91
93
95
96
97
98
99
101
102
103
105
Total
Examined
2
2
1
2
4
2
1
1
1
1
2
2
1
3
2
1
2
14
Number Of Male Animals With Tumors At Each Site, At Specified Week Of Study
Hemangiomas,
Hemangiosarcomas
Incid.a
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
6
Fatal3
0
0
0
1
1
0
0
0
0
1
1
0
0
1
1
1
1
2
1
2
2
1
0
0
Lung
1
0
1
1
1
0
0
1
1
1
4
1
1
1
1
1
1
2
1
2
1
1
2
10
Forestomach
0
0
ob
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
Harderian
gland0
0
0
0
0
0
0
0
0
0
1
2
0
0
1
0
0
0
1
0
0
1
0
0
4
Kidney
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
C-5
-------
Dose
Group
80
Week
of
Study
56
61
65
75
81
83
84
85
86
87
89
90
91
92
93
94
95
96
97
98
99
101
105
Total
Examined
2
2
1
2
1
3
3
2
3
2
3
1
2
2
0
2
13
Number Of Male Animals With Tumors At Each Site, At Specified Week Of Study
Hemangiomas,
Hemangiosarcomas
Incid.a
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
7
Fatal3
0
0
1
0
0
1
0
0
0
1
0
1
1
2
0
1
2
0
1
1
0
1
0
Lung
0
0
0
2
1
1
1
2
1
1
1
2
3
1
3
2
3
1
2
1
0
2
13
Forestomach
0
0
0
0
0
0
1
1
0
0
0
1
0
1
0
1
0
0
1
0
0
0
0
Harderian
gland0
0
0
0
0
0
0
0
1
0
0
0
0
1
1
0
0
1
0
1
0
1
0
6
Kidney
0
0
0
0
1
0
1
0
0
0
0
0
0
0
1
0
1
0
0
0
1
1
3
a"Incid." , or Incidental, denotes tumors not concluded to have caused the death of the animal. Fatal denotes tumors
considered to have caused the death of the animal.
bTissue for one animal of total examined was missing or unsuitable for histopathologic examination.
cHarderian gland was examined histopathologically only if a lesion was observed grossly at necropsy; instances of "0" for
these tissues indicate only that no tumor was seen grossly, for dose-response modeling purposes
Source: NTP (1998. 042076).
C-6
-------
Table C-3. Summary of model selection and modeling results for best-fitting
multistage-Weibull models, using time-to-tumor data for female mice
Site
Lung
Hemangiomas,
hemangiosarcomas
(fatal)
Hemangiomas,
hemangiosarcomas
(incidental)
Harderian gland
Mammary gland
carcinomas,
adenoacanthomas
Forestomach
Hepatocellular
adenomas,
carcinomas
Skin
Zymbal's gland
Stages
ld
2
1
2
1
o
J
2
1
1
o
3
2
1
1
1
o
J
2
1
LLa
-83.020
-135.848
-138.519
-65.812
-66.953
-58.256
-58.266
-58.266
-87.960
-19.174
-19.596
-20.772
-119.227
-87.463
-11.402
-11.726
-12.611
x2"
5.34
2.28
0.02
0.00
0.84
2.35
0.65
1.77
AIC
172.04
279.70
283.04
139.62
139.91
126.51
124.53
122.53
181.92
48.35
45.19
45.54
244.45
180.93
32.80
31.45
31.22
Responses @ ppm levels0
0 12.8 32 80
4 28 34 42
4.1 27.5 32.4 43.1
4 6 18 —
3.45 7.4 13.5 —
3.25 10.6 11.1 —
4 6 18
3.7 6.6 17.6 —
3.3 9.3 15.3 —
2539
2.4 3.6 4.1 8.9
2.3 3.7 4.2 8.8
2.3 3.7 4.3 8.7
3 6 11 14
3.5 5.9 8.5 15.7
1004
0.4 0.4 0.4 3.7
0.5 0.5 0.7 3.4
0.5 1.0 1.1 2.5
20 26 20 30
21.6 23.0 20.7 30.8
0 11 11 18
0.0 5.6 10.6 22.4
0003
0 0 0.2 2.8
0 0.1 0.4 2.5
0.0 0.4 0.8 1.9
Model Selection
Rationale
One-stage model was
only available fit
(highest dose group
dropped)
X2, lowest AIC
Lowest AIC
Lowest AIC
One-stage model was
only available fit
Lowest AIC
One-stage model was
only available fit
One-stage model was
only available fit
Lowest AIC
aLL=log-likelihood.
b%2 = chi-squared statistic for testing the difference between 2 model fits.
Calculated from 2 x | (LL; - LLj) , and evaluated
for the identified model fit relative to the fit with one less stage (unless it was a one-stage model). In all cases 1 degree of
freedom was associated with the test, and the critical chi-squared value was 3.84.
'"Responses" describes the number of animals with each tumor type: Observed responses are in italics, and expected
responses (predicted by each model fit) are given to one decimal place for comparison with the observed data.
Bold indicates the best-fitting model for each endpoint. Outputs for best-fitting models are included in the following pages.
Source: Data modeled from: (NTP, 1998, 042076).
C-7
-------
Table C-4. Summary of model selection and modeling results for best-fitting
multistage-Weibull models, using time-to-tumor data for male mice
Site
Lung
Hemangiomas,
hemangiosarcomas
(fatal)
Hemangiomas,
hemangiosarcomas
(incidental)
Harderian gland
Kidney
Forestomach
Stages
ld
1
1
1
3
2
1
2
1
LLa
-104.927
-537.427
-109.463
-73.664
-40.948
-40.960
-41.003
-30.404
-30.841
^
—
—
—
—
0.02
0.09
—
0.88
—
AIC
215.86
1084.85
228.93
157.33
91.90
89.92
88.01
68.81
67.68
Responses
0 ] 12.8
13 28
14.0 26.6
3 14
0.0 8.3
3 14
5.3 11.1
2 5
2.3 5.2
0 2
0 1.7
0 1.7
0.0 2.0
; o
0.7 0.8
0.5 1.2
@ ppm levels0
32
36
33.9
23
12.2
23
15.9
10
7.4
3
3.4
3.5
3.7
2
1.2
2.0
80
43
44.6
21
20.3
21
27.2
12
14.0
9
8.9
8.8
8.3
5
4.3
4.2
Model Selection Rationale
One-stage model was
only available fit
One-stage model was
only available fit
One-stage model was
only available fit
One-stage model was
only available fit
Lowest AIC
(highest stage model
available)
Lowest AIC
aLL=log-likelihood.
b%2 = chi-squared statistic for testing the difference between 2 model fits. Calculated from 2 x | (LL; - LLj) , and
evaluated for the identified model fit relative to the fit with one less stage (unless it was a one-stage model). In all cases 1
degree of freedom was associated with the test, and the critical chi-squared value was 3.84.
'"Responses" describes the number of animals with each tumor type: Observed responses are in italics, and expected
responses (predicted by each model fit) are given to one decimal place for comparison with the observed data.
Bold indicates the best-fitting model for each endpoint. Outputs for best-fitting models are included in the following
pages.
Source: Data modeled from NTP (1998, 042076).
C-8
-------
Incidental Risk: F LUNG 1s
Dose = 0.00
Dose = 12.80
..&1
^
ro
0
0.
i i
CO
CD
CD _
CD
CD ~
CM
CD
O
^^^*
° "V — i 1 1 " "i ""i '
0 20 40 60 80 100
.0
ro
_a
o
oq
CD
cp
o
CNI
CD
p
o
Time
20 40 60 80 100
Time
Dose = 32.00
Dose = 80.00
.a
CO
.0
o
CO
o
cp
d
CN
CD
CD
CD
20
I
40
I
60
.a
ro
.0
o
oq
cb
cp
CD
CN
CD
CD
CD
80 100
I
20
I
40
I
60
80 100
Time
Time
Figure C-l. Female mice, alveolar/bronchiolar tumors. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M: \_chemicals\chloroprene\msw\F_LTJNG_ls . (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)^c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 112
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
User specifies the following parameters:
t 0 0
C-9
-------
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 3.77778
t_0 = 0 Specified
beta_0 = 2.32179e-009
beta 1 = 2.11013e-009
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c
beta_0
beta 1
1
-0.99
-1
beta_0
-0.99
1
0.99
beta_l
-1
0.99
1
Variable
c
beta_0
beta 1
Estimate
3.78542
2.24014e-009
2 .039726-009
Parameter Estimates
Std. Err.
0.978326
1.037456-008
8.870496-009
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
1.86793 5.7029
-1.809356-008 2.25738e-008
-1.534616-008 1.942566-008
Log(likelihood)
Fitted Model -83.02
# Param
3
AIC
172.04
DOSE
0
13
32
80
46
21
16
Data Summary
CLASS
F I
4
28
34
42
U Total Expected Response
50
50
50
50
4 .10
27.45
32 .39
43 .13
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
1.19617
0.883475
1.60092
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
0.114103
0.0865258
0.148645
C-10
-------
Incidental Risk: F_HEM3fatal_2s
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose= 0.00 Dose= 12.80
Q
CL
l_I _
oo
ci
CD
ci
ci ~
CM
ci
O
ci
O OO OBO^p
, _ ^» ifjgl 1
1 1 1 1 1 1
0 20 40 60 80 100
&
^
S
Q
CL
i_i _
oo
ci
CD
ci
^
ci ~
CM
ci
O
-000
^^Si
_z_as-^^af
I I I I I I
0 20 40 60 80 100
Time
Time
Dose= 32.00
s.
CL
oq
ci
CD
ci
CM
ci
q
ci
\ \ \ I I
20 40 60 80 100
Time
Figure C-2. Female mice, hemangiomas and hemangiosarcomas in all organs; high
dose dropped, hemangiosarcomas occurring before termination considered fatal.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M:\_chemicals\chloroprene\msw\F_HEM3fatal_2s.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l+beta_2*dose^2) '
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 84
Total number of records with missing values = 0
Total number of parameters in model = 5
Total number of specified parameters = 1
Degree of polynomial = 2
C-11
-------
User specifies the following parameters:
t_0 = 0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 4.25
t_0 = 0
beta_0 = 2.2479e-010
beta_l = 2.06502e-034
beta 2 = 2.12137e-012
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -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 0
beta 2
c
beta_0
beta 2
Variable
c
beta_0
beta_l
beta 2
Estimate
5.90503
1.01175e-013
0
1.265396-015
Parameter Estimates
Std. Err.
1.49573
7.08031e-013
NA
8.531036-015
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
2.97346 8.8366
-1.286546-012 1.488896-012
-1.545516-014
1.798596-014
Log(likelihood)
Fitted Model -135.848
# Param
4
AIC
279.697
DOSE
0
13
32
46
43
32
Data Summary
CLASS
1
4
16
U Total Expected Response
50
50
50
3 .45
7.40
13 .53
Minimum observation time for F tumor context =
31
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
10.1137
5 .75142
13 .1199
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
3 .12363
0.640904
4 .05212
C-12
-------
Incidental Risk: F HEMSinc 2s
Dose = 0.00
Dose= 12.80
ro
.a
o
cq
CD
cp
CD
Ol
CD
20
F
40
60
Time
\^
80
100
_
ro
.0
o
oq
CD
cp
CD
CNI
CD
CD
CD
20
40 60
Time
\ \
80 100
Dose= 32.00
ro
e
CL
oq
CD
cp
CD
01
CD
CD
CD
20
\
40
60
Time
I
80
100
Figure C-3. Female mice, hemangiomas and hemangiosarcomas in all organs; high
dose dropped, all tumors considered incidental. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: F_HEM3inc_2s.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)^c *
(beta_0+beta_l*dose^l+beta_2*dose^2) '
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 84
Total number of records with missing values = 0
Total number of parameters in model = 5
Total number of specified parameters = 1
Degree of polynomial = 2
User specifies the following parameters:
t 0 0
C-13
-------
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 1.13333
t_0 = 0
beta_0 = 0.000428228
beta_l = 0
beta 2 = 2.52747e-006
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -c ~t_0 -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 0
beta 2
beta_0
beta 2
1
-0.4
Variable
c
beta_0
beta_l
beta 2
Estimate
1
0.000792254
0
4.541426-006
Parameter Estimates
Std. Err.
NA
0.000500484
NA
1.850426-006
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.000188678
9.14653e-007
0.00177319
! .168186-006
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -65.8122 4 139.624
DOSE
0
13
32
46
43
32
Data Summary
CLASS
F I
4
6
18
U Total Expected Response
50
50
50
3 .74
6 .57
17.56
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
14 .9357
11.0629
19.8583
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
4 .61294
2 .0194
6 .12873
C-14
-------
Incidental Risk: F HARD 1s
Dose = 0.00
Dose= 12.80
>%
=
s
Q
CL
CD
ci
ci ~
-
CM
ci
O
ci
1 1 1 1 1 1
0 20 40 60 80 100
&
Q
CL
CD
ci
ci ~
CM
ci
O
ci "1
^^^
1 1 1 1 1 1
0 20 40 60 80 100
Time
Time
Dose = 32.00
Dose = 80.00
cp
ci
!Q o
&
o
CL
-------
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 2.83333
t_0 = 0 Specified
beta_0 = 1.02152e-007
beta 1 = 7.3281e-009
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c
beta_0
beta 1
1
-1
-1
beta_0
-1
1
0.99
beta_l
-1
0.99
1
Variable
c
beta_0
beta 1
Estimate
2.93861
6.26114e-008
4.599466-009
Parameter Estimates
Std. Err.
2.46009
7.18253e-007
5.01418e-008
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-1.88307 7.7603
-1.345146-006 1.470366-006
-9.36766e-008 1.028766-007
Log(likelihood) # Param AIC
Fitted Model -58.2663 3 122.533
DOSE
0
13
32
80
48
45
47
41
Data Summary
CLASS
F I
U Total Expected Response
50
50
50
50
2 .32
3 .71
4 .26
8 .73
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
27.0825
12 .614
85.8726
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
2 .5834
1.20327
8 .04772
C-16
-------
Incidental Risk: F MAMM 1s
Dose = 0.00
Dose= 12.80
.8
Q
CL
CO
ci
c\i
ci
q
ci
I
20
40 60
Time
I
80
100
.8
Q
CL
CO
ci
CN
ci
q
ci
I
20
\
40
60
Time
\
80 100
Dose= 32.00
Dose = 80.00
_Q
CO
_Q
0
CO
ci
C\l
ci
q
ci
\
20
40 60 80 100
Time
_Q
CO
_Q
O
CO
ci
CN
ci
q
ci
20 40 60 80 100
Time
Figure C-5. Female mice, mammary gland tumors. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M:\_chemicals\chloroprene\msw\F_MAMM_ls.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 126
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
User specifies the following parameters:
t_0 = 0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
C-17
-------
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 1.0303
t_0 = 0 Specified
beta_0 = 0.000643678
beta 1 = 4.34581e-005
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -c ~t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
beta_0
beta 1
beta_0
1
-0.57
beta_l
-0.57
1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
c 1 NA
beta_0 0.000740811 0.000512345 -0.000263368 0.00174499
beta_l 4.96148e-005 2.12095e-005 8.04497e-006 9.11846e-005
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -87.9599 3 181.92
DOSE
0
13
32
80
46
43
39
36
Data Summary
CLASS
F I
0 3
0 6
0 11
0 14
U Total Expected Response
50 3.50
50 5.93
50 8.48
50 15.68
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
20.419
14 .0543
38 .5881
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD
BMDL
BMDU
=
=
=
=
Computation
Incidental
Extra
0.01
0.9
104
1.94776
1.34101
3 .71557
C-18
-------
Incidental Risk: F FORST 2s fix
Dose = 0.00
Dose= 12.80
—
£
Q
CL
in
ci
o
ci
in
o -
ci
o
o
•
<-> 1 1 1 1 1 1
0 20 40 60 80 100
in
ci
o
= o
&
p in
- §-
o
o
ci
I I I I I I
0 20 40 60 80 100
Time Time
Dose = 32.00
Dose = 80.00
&
Q
CL
ci
o
ci
in
o _
o
R
ci
1 1 1 1 1 1
0 20 40 60 80 100
in
ci
= o
.a
&
Q in
a. 9
o
8 _|
ci
I
20
Time
40 60
Time
I
80
100
Figure C-6. Female mice, forestomach tumors. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: F_FORST_2s_fix.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l+beta_2*dose^2) '
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 118
Total number of records with missing values = 0
Total number of parameters in model = 5
Total number of specified parameters = 2
Degree of polynomial = 2
User specifies the following parameters:
c = 4.1253
t 0 0
C-19
-------
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 4.12533 Specified
t_0 = 0
beta_0 = 5.01708e-011
beta_l = 0
beta 2 = 1.09429e-013
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -c ~t_0 -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 0
beta 2
beta_0
beta 2
1
-0.13
-0.13
1
Variable
beta_0
beta_l
beta 2
Estimate
5.01701e-011
0
1.09436-013
Parameter Estimates
Std. Err.
7.09515e-011
NA
8 .368296-014
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-8.889246-011 1.892336-010
-5.458546-014
2 .734456-013
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -19.5963 3 45.1926
DOSE
0
13
32
80
49
49
49
46
Data Summary
CLASS
F I
U Total Expected Response
50
50
49
50
0.46
0.50
0.68
3 .35
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
67.812
46 .323
122 .222
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
20.9439
5.69172
36 .9312
C-20
-------
Incidental Risk: F LIV 1s
Dose = 0.00
Dose= 12.80
&
oq
ci
cp
ci
c\|
ci
q
ci
20
40 60
Time
\ I
80 100
&
Q
CL
oq
ci
cp
ci
c\i
ci
q
ci
20
40 60
Time
I I
80 100
Dose= 32.00
Dose = 80.00
&
Q
CL
oq
ci
cp
ci
c\i
ci
q
ci
20
40 60
Time
\ I
80 100
&
Q
CL
100
Figure C-7. Female mice, hepatocellular adenomas and carcinomas. Details
below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: F_LIV_ls.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 129
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
User specifies the following parameters:
C-21
-------
t 0
0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 4.25
t_0 = 0 Specified
beta_0 = 1.77794e-009
beta 1 = 6.82109e-011
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c
beta_0
beta 1
1
-1
-1
beta_0
-1
1
1
beta_l
-1
1
1
Variable
c
beta_0
beta 1
Estimate
4.15974
2 .703736-009
1. 01083e-010
Parameter Estimates
Std. Err.
1.3308
1.672726-008
5.8692e-010
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
1.55141 6.76806
-3.008096-008 3.548846-008
-1.049266-009 1.251426-009
Log(likelihood) # Param AIC
Fitted Model -119.227 3 244.454
DOSE
0
13
32
80
30
22
30
20
Data Summary
CLASS
F I
20
26
20
30
U Total Expected Response
50 21.63
49 22.96
50 20.68
50 30.79
Benchmark Dose
Risk Response
Risk Type
Specified effect
Confidence level
Time
HMD
BMDL
BMDU
Computation
Incidental
Extra
0.1
0.9
104
4.24297
2.44688
8.61316
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
0.404737
0.233408
0.818725
C-22
-------
Incidental Risk: F SKIN 1s
Dose = 0.00
Dose= 12.80
.8
2
0.
00
ci
CD
ci
CN|
ci
I
20
40 60
Time
I
80
100
.8
Q
CL
oq
ci
cp
ci
CNJ
ci
q
ci
I
20
\
40
60
Time
\
80 100
Dose= 32.00
Dose = 80.00
_Q
ro
g
Q.
oq
ci
cp
ci
CNI
ci
q
ci
20 40 60 80 100
Time
_a
ro
g
Q.
oq
ci
cp
ci
CNI
ci
q
ci
• • • •
20 40 60 80 100
Time
Figure C-8. Female mice, skin sarcomas. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: F_SKIN.(d)
Wed Feb 17 15:09:24 2010
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 121
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
User specifies the following parameters:
C-23
-------
t 0
0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 1.61905
t_0 = 0 Specified
beta_0 = 4.01488e-023
beta 1 = 6.08721e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta 1
c
beta 1
1
-1
-1
1
Variable
c
beta_0
beta 1
Estimate
1.56405
0
7.77467e-006
Parameter Estimates
Std. Err.
1.25364
NA
4 .340976-005
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.893041 4.02115
-7.73067e-005
9.28561e-005
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -87.4625 3 180.925
DOSE
0
13
32
80
50
38
39
32
Data Summary
CLASS
F I
0
11
11
18
U Total Expected Response
50
50
50
50
0.00
5.59
10.58
22 .43
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Specified effect = 0.1
Confidence level = 0.9
Time
HMD
BMDL
BMDU
104
9.48956
7.18444
14.5757
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Specified Effect = 0.01
Confidence Level = 0.9
BMD =
BMDL =
BMDU =
104
0.905208
0.665324
1.4015
C-24
-------
Incidental Risk: F_Zymb_1s05
Dose = 0.00
Dose= 12.80
00
o
1 s
Ql °
o
o _
\^
20
40
\^
60
Time
\ \
80 100
00
o _
_
ro
.0
o
o
o _
20 40 60 80 100
Time
Dose= 32.00
Dose = 80.00
oo
o _
ro _..
S S_i
A- o
o
o _
oo
o _
ro _..
^ S-i
A- o
o
o _
20
40 60
Time
80 100
20 40 60 80 100
Time
Figure C-9. Female mice, Zymbal's gland tumors. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M:\_chemicals\chloroprene\msw\F_Zymb_ls05.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)^c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 119
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
User specifies the following parameters:
t 0 0
C-25
-------
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 1.09677
t_0 = 0 Specified
beta_0 = 3.72225e-028
beta 1 = 3.90719e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta 1
c
beta 1
-1
1
Variable
c
beta_0
beta 1
Estimate
1.09674
0
3.907336-006
Parameter Estimates
Std. Err.
4.17394
NA
7.24422e-005
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-7.08402 9.27751
-0.000138077
0.000145891
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -12.6107 3 31.2214
DOSE
0
13
32
80
50
50
50
47
Data Summary
CLASS
F I
U Total Expected Response
50
50
50
50
0.00
0.36
0.76
1.90
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.05
0.9
104
80.5411
22 .4657
255.715
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
15.7811
5.75828
50.0819
C-26
-------
Incidental Risk: M LUNG 1s
Dose = 0.00
Dose= 12.80
_
ro
_a
o
ro
cb
cp
CD
o
o
20 40 60 80 100
Time
ro
o
o_
cp
CD
CNI
CD
o
CD
20
\
40
60
Time
I I
80 100
Dose= 32.00
Dose= 80.00
_
ro
.a
o
ro
CD
cp
CD
Ol
CD
CD
CD
I
20
40
I
60
Time
I
80
100
_
ro
.a
o
ro
cb
cp
CD
01
CD
CD
CD
I
20
\
40
60
Time
I
80
100
Figure C-10. Male mice, alveolar/bronchiolar tumors. Details below.
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)~c *
(beta 0+beta l*dose^l)}
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 100
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
C-27
-------
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-OOE:
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 3.4
t_0 = 0 Specified
beta_0 = 5.3339e-008
beta 1 = 5.89044e-009
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
1
-1
-1
beta 0
-1
1
1
beta 1
-1
1
1
Parameter Estimates
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
0.918807 6.0043
-4.33631e-007 5.13819e-007
-4.6823e-008 5.5744e-008
Log(likelihood) # Param
Fitted Model -104.927 3
AIC
21J
U Total Expected Response
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD
BMDL
BMDU
=
=
Computation
Incidental
Extra
0.1
0.9
104
2 .46168
1 .86129
3 .46534
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
0.23482
0.178411
0.321837
C-28
-------
Incidental Risk: M_HEM_3s
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose = 0.00 Dose= 12.80
ro
o
0.
00
CD
CD _
CD
CD ~
CM
CD
O
CD
0
I I I I I I
0 20 40 60 80 100
ro
o
0.
Time
00
CD
CD _
CD
CD ~
CM
CD
O
CD ~~l
r o
_
ro
.a
o
01
CD
CD
CD
I
20
\
40
60
Time
I
80
100
Figure C-ll. Male mice, hemangiomas and hemangiosarcomas;
hemangiosarcomas occurring before termination considered fatal. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M_HEM_3s.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)^c *
(beta_0+beta_l*dose^l+beta_2*dose^2+beta_3*dose^3)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 103
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 1
Degree of polynomial = 3
User specifies the following parameters:
C-29
-------
t 0
0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c
t_0
beta_0
beta_l
beta 2
7.33333
0
2.927356-016
1.246616-017
5.74518e-040
Specified
beta 3 = 1.930266-021
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0 -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 )
c beta 1
c
beta 1
nan
nan
nan
nan
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
13.2483
0
3.781846-029
0
0
Parameter Estimates
Std. Err.
nan
NA
nan
NA
NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
nan nan
nan
nan
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -537.427 5 1084.85
DOSE
0
13
32
80
47
36
27
29
Data Summary
CLASS
0
6
16
13
U Total Expected Response
50
50
50
50
0.00
8 .30
12 .19
20.33
Minimum observation time for F tumor context =
65
Benchmark Dose
Risk Response
Risk Type
Specified effect
Confidence level
Time
HMD
BMDL
BMDU
Computation
Incidental
Extra
0.1
0.9
104
5.28208
3.34052
5.94514
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
0.503858
0.318652
0.567106
C-30
-------
Incidental Risk: M HEM 3s inc
Dose = 0.00
Dose= 12.80
>%
Probabi
CD
o
—
ci ~
CM
O
O
ci
I I I I I I
0 20 40 60 80 100
CD
O
>% ~~
1 d"
Q
Q- CM
O
O
ci
___^^
1 1 1 1 1 1
0 20 40 60 80 100
Time Time
Dose= 32.00
Dose = 80.00
CD
ci
g
Q.
q
ci
20 40 60
Time
\ I
80 100
CD
ci
g
Q-
q
ci
20 40 60 80 100
Time
Figure C-12. Male mice, hemangiomas and hemangiosarcomas; all tumors
considered incidental. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M:\_chemicals\chloroprene\msw\M_HEM_3s_inc.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)^c *
(beta_0+beta_l*dose^l+beta_2*dose^2+beta_3*dose^3)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 103
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 1
Degree of polynomial = 3
User specifies the following parameters:
C-31
-------
t 0
0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c
t_0
beta_0 =
beta_l =
beta_2 =
beta 3 =
3.88235
0
1.935736-009
2.009366-010
0
0
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -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 0
beta 1
c
beta_0
beta 1
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
3.87398
2.01294e-009
2.087176-010
0
0
Parameter Estimates
Std. Err.
1.89771
1.786236-008
1.800836-009
NA
NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
0.154536 7.59343
-3.299656-008 3.702246-008
-3.320846-009 3.738286-009
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood)
Fitted Model -109.463
# Param
5
AIC
228 .926
DOSE
0
13
32
80
47
36
27
29
Data Summary
CLASS
F I
3
14
23
21
U Total Expected Response
50
50
50
50
5.28
11.12
15.86
27.21
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
7.74767
5.33823
12 .7663
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
0.73905
0.509228
1.21647
C-32
-------
Incidental Risk: M HARD 3s
Dose = 0.00
Dose= 12.80
>,
—
!Q
s
Q
CL
d
CO
ci
CM
ci
ci ~
o
-s-^3*^
° ~h 1 1 1 1 r^
0 20 40 60 80 100
Time
CO
ci
s
Q
CL
q
ci
I
20
\
40
60
Time
\
80 100
Dose= 32.00
Dose = 80.00
_Q
&
Q
CL
CO
ci
CN
ci
q
ci
20
40 60
Time
\ I
80 100
>, co
i= ci
!Q
& CM
S. o
Q_
q
ci
20 40 60
Time
I I
80 100
Figure C-13. Male mice, Harderian gland tumors. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M_HARD_3s.(d)
Wed Feb 24 14:48:16 2010
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)^c *
(beta_0+beta_l*dose^l+beta_2*dose^2+beta_3*dose^3) '
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 106
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 1
Degree of polynomial = 3
User specifies the following parameters:
t_0 = 0
Maximum number of iterations = 16
C-33
-------
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c
t_0
beta_0 =
beta_l =
beta_2 =
beta 3 =
4 .25
0
1.535776-010
1.520416-011
0
0
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -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 0
beta 1
c
beta_0
beta 1
1
-1
-1
-1
1
1
-1
1
1
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
5.57459
3.258836-013
3.5986-014
0
0
Parameter Estimates
Std. Err.
3.19215
4 .844716-012
5.252356-013
NA
NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.681904 11.8311
-9.169576-012 9.821336-012
-9.934636-013 1.065426-012
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -73.6639 5 157.328
DOSE
0
13
32
80
48
45
40
38
Data Summary
CLASS
F I
2
5
10
12
U Total Expected Response
50
50
50
50
2 .29
5.18
7.40
14 .04
Benchmark Dose
Risk Response
Risk Type
Specified effect =
Confidence level =
Time
HMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.1
0.9
104
16 .6911
10.4645
35.082
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
1.59216
0.998471
5.03875
C-34
-------
Incidental Risk: M KIDN 1s
Dose = 0.00
Dose= 12.80
>, °
.51 CM _
s
p
o
o _
1 1 1 1 1 1
0 20 40 60 80 100
ci
>, °
.51 CM _
~ o
£
Q o
Q. •"- -
ci
O
ci
_fr^^
1 1 1 1 1 1
0 20 40 60 80 100
Time
Time
Dose = 32.00
Dose = 80.00
8 -I
ci
£ ° -I
5 °
&
e o
Q. •<-
ci
8 -I
ci
T I I I I I
0 20 40 60 80 100
8 J
ci
f 8-I
!Q °
&
e o
Q. •<-
ci
8 J
ci
n i i i i i
0 20 40 60 80 100
Time
Time
Figure C-14. Male mice, renal tubule tumors. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M:\_chemicals\chloroprene\msw\M_KIDN_ls.(d)
Tue May 11 10:57:53 2010
title = Chloroprene: Male mice, kidney adenomas, source = NTP 1998, chemical = CHLOROPRENE, mol.wgt
88.5, route = AIR (ppm), expt.length = 104, life.length = 104, dose.avg.factor = 1
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 106
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
C-35
-------
User specifies the following parameters:
t_0 = 0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 4.85714
t_0 = 0 Specified
beta_0 = 0
beta 1 = 5.87389e-013
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta 1
c
beta 1
1
-1
Variable
c
beta_0
beta 1
Estimate
6.09231
0
2.03124e-015
Parameter Estimates
Std. Err.
4.64814
NA
4 .33666-014
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
-3.01787 15.2025
-8.296466-014
! .702716-014
Log(likelihood) # Param
Fitted Model -41.0033 3
AIC
88.0066
DOSE
0
13
32
80
50
47
47
41
Data Summary
CLASS
F I
U Total Expected Response
50
50
50
50
0.00
1.95
3 .70
8 .34
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD
BMDL
BMDU
=
_
=
=
Computation
Incidental
Extra
0.1
0.9
104
26 .7011
16 .4536
47.1278
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD
BMDL
BMDU
=
_
=
=
Computation
Incidental
Extra
0.01
0.9
104
2 .54702
1.56959
4 .49547
C-36
-------
Incidental Risk: M FORST 1s
Dose = 0.00
Dose= 12.80
_
ro
_a
o
CO
o _
§ J
o
o
o _
20 40 60 80 100
Time
_
ro
_a
o
CO
o _
§ J
o
o
o _
I
20
40 60
Time
\
80
100
Dose= 32.00
Dose = 80.00
.a
ro
_a
o
oo
o
p
o
o
o _
20
40 60
Time
I
80
100
.a
ro
_a
o
oo
o
p
o
o
o _
I
20
40
I
60
\
80
100
Time
Figure C-15. Male mice, forestomach tumors. Details below.
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: M:\_chemicals\chloroprene\msw\M_FORST_ls.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)*c *
(beta_0+beta_l*dose^l)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 106
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 1
Degree of polynomial = 1
C-37
-------
User specifies the following parameters:
t_0 = 0
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 1.36
t_0 = 0 Specified
beta_0 = 2.11777e-005
beta 1 = 2.06761e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c
beta_0
beta 1
1
-1
-1
beta_0
-1
1
1
beta_l
-1
1
1
Variable
c
beta_0
beta 1
Estimate
1.2938
2.8702e-005
2.792786-006
Parameter Estimates
Std. Err.
4.09082
0.000540721
5.19088e-005
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-6.72406 9.31165
-0.00103109 0.0010885
-9.89466e-005 0.000104532
Log(likelihood) # Param AIC
Fitted Model -30.8413 3 67.6827
DOSE
0
13
32
80
49
48
47
45
Data Summary
CLASS
F I
U Total Expected Response
50
50
50
50
0.54
1.20
2 .03
4 .24
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.05
0.9
104
45.1225
22 .7599
157.031
Benchmark Dose
Risk Response
Risk Type
Specified Effect =
Confidence Level =
Time
BMD =
BMDL =
BMDU =
Computation
Incidental
Extra
0.01
0.9
104
8 .84123
4 .46001
30.7684
C-38
-------
Table C-5. Summary of human equivalent composite cancer risk values estimated
by 0.01/BMDoi, based on male and female mouse tumor incidence
Tumor Site
BMDoi
(ppm)
BMDLoi
(ppm)
1% Risk estimate" at:
BMDoi
(/ppm)
BMDLoi
(/ppm)
SDb
SD2
Proportion
of Total
Variance
Female Mice
Lung (systemic
dosimetry)
Hemangiomas,
hemangiosarcomas
(fatal)
Harderian gland
Mammary gland;
carcinomas,
adenoacanthomas
Forestomach
Hepatocellular
adenomas,
carcinomas
Skin
Zymbal's gland
1.14E-01
3.12E+00
2.58E+00
1.95E+00
2.09E+01
4.05E-01
9.05E-01
1.58E+01
8.65E-02
6.41E-01
1.20E+00
1.34E+00
5.69E+00
2.33E-01
6.65E-01
5.76E+00
Sum of MLE cancer risks at BMD0i (/ppm):
Human equivalent sum of
risk estimates (/(ug/m3))0:
8.76E-02
3.20E-03
3.87E-03
5.13E-03
4.77E-04
2.47E-02
1.10E-02
6.34E-04
1.367E-01
2.122E-04
Upper bound on sum of risk estimates at BMD0i (/ppm)e:
Human equivalent upper bound on sum of
risk estimates (/(ug/m3))f:
1.16E-01
1.56E-02
8.31E-03
7.46E-03
1.76E-03
4.28E-02
1.50E-02
1.74E-03
1.729E-01
2.683E-04
1.70E-02
7.54E-03
2.70E-03
1.41E-03
7.78E-04
1.10E-02
2.42E-03
6.70E-04
Sum, SD2:
Composite SDd
2.88E-04
5.68E-05
7.28E-06
1.99E-06
6.05E-07
1.22E-04
5.86E-06
4.50E-07
4.829E-04
2.198E-02
0.60
0.12
0.02
0.00
0.00
0.25
0.01
0.00
Male Mice
Lung (systemic
dosimetry)
Hemangiomas,
hemangiosarcomas
(fatal)
Forestomach
Harderian gland
Kidney
2.35E-01
5.04E-01
8.84E+00
1.59E+00
2.55E+00
1.78E-01
3.19E-01
4.46E+00
9.98E-01
1.57E+00
Sum of MLE cancer risks at BMD0i (/ppm):
Human equivalent sum of
risk estimates (/(ug/m3))0:
4.26E-02
1.98E-02
1.13E-03
6.28E-03
3.93E-03
7.377E-02
1.145E-04
Upper bound on sum of risk estimates at BMD0i (/ppm)e:
Human equivalent upper bound on sum of
risk estimates (/(ug/m3))f:
5.61E-02
3.14E-02
2.24E-03
l.OOE-02
6.37E-03
9.209E-02
1.429E-04
8.19E-03
7.01E-03
6.75E-04
2.27E-03
1.49E-03
Sum, SD2:
Composite SDd
6.70E-05
4.92E-05
4.56E-07
5.15E-06
2.21E-06
1.240E-04
1.114E-02
0.54
0.40
0.00
0.04
0.02
al% risk estimate = 0.01/POD = (0.01/BMDm ) or (0.01/BMDLOT.).
bSD = ((0.01/BMDLoi -BMDoi)/1.645).
°Equals sum of MLE cancer risks at BMD0i(/ppm) divided by 6/24 (hours), 5/7 (days), and 3.62E+03 ug/m3.
Composite SD = (Sum, SD2)05.
eEquals (sum of MLE cancer risks at BMD0i(/ppm) + (1.645 x composite SD)).
fEquals upper bound on the sum of risk estimates at BMDLM (/ppm) divided by: 6/24 (hours); 5/7 (days);
and3.62E+03 ug/m3.
Bold indicates summary values.
Source: Data modeled from (NTP, 1998. 042076)
C-39
------- |