DRAFT - DO NOT CITE OR QUOTE
oEPA
EPA/635/R-09/010D
www. ep a. gov/iris
TOXICOLOGICAL REVIEW
OF
CHLOROPRENE
(CAS No. 126-99-8)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
July 2010
NOTICE
This document is an Interagency Science Discussion/Final Agency Review draft. This information
is distributed solely for the purpose of pre-dissemination peer review under applicable information
quality guidelines. It has not been formally disseminated by EPA. It does not represent and should not
be construed to represent any Agency determination or policy. It is being circulated for review of its
technical accuracy and science policy implications.
U.S. Environmental Protection Agency
Washington, DC
July 2010
DRAFT-DO NOT CITE OR QUOTE
-------
1
DISCLAIMER
2 This document is a final draft for review purposes only and has been reviewed in accordance
3 with U.S. Environmental Protection Agency policy and approved for publication. This information is
4 distributed solely for the purpose of pre-dissemination peer review under applicable information
5 quality guidelines. It has not been formally disseminated by EPA. It does not represent and should not
6 be construed to represent any Agency determination or policy. Mention of trade names or commercial
7 products does not constitute endorsement or recommendation for use.
8
July 2010 " DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
TABLE OF CONTENTS
LIST OF TABLES v
LIST OF FIGURES ix
LIST 01 ABBREVIATIONS AM) ACRONYMS xi
FOREWORD xiii
AUTHORS, CONTRIBUTORS, AND REVIEWERS xiv
1. INTRODUCTION 1-1
2. CHEMICAL AM) PHYSICAL IM ORM ATION 2-1
3. TOXICOKINETICS 3-1
3.1. ABSORPTION 3-1
3.2. DISTRIBUTION 3-1
3.3. METABOLISM 3-2
3.4. ELIMINATION 3-12
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS 3-12
4. HAZARD IDENTIFICATION 4-1
4.1 STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL CONTROLS .. .4-1
4.1.1. Chloroprene Exposure and Cancer Effects 4-1
4.1.1.1. Overview 4-1
4.1.1.2. Individual Occupational Studies 4-1
4.1.1.3. Summary and Discussion of Relevant Methodological Issues 4-16
4.1.2. Chloroprene Exposure and Noncancer Effects 4-20
4.1.2.1. Acute-, Short-, and Sub chronic-Duration Noncancer Effects 4-20
4.1.2.2. Chronic Noncancer Effects 4-22
4.2 SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASS AYS IN ANIMALS-
ORAL AND INHALATION 4-24
4.2.1. Oral Exposure 4-24
4.2.2. Inhalation Exposure 4-26
4.3 REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND INHALATION 4-50
4.4 OTHER DURATION- ORENDPOINT-SPECIFIC STUDIES 4-56
4.4.1. Acute and Subchronic Studies 4-56
4.4.2. Immunotoxicity 4-58
4.5 MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF MODE OF ACTION.. 4-59
4.5.1. Mode-of-Action Studies 4-59
4.5.2. Genotoxicity Studies 4-62
4.5.2.1. Bacterial Mutagenicity Assays 4-64
4.5.2.2. Mammalian Cell Assays 4-65
4.5.2.3. In Vivo Bioassays 4-66
4.5.3. Structural Alerts 4-67
4.6 SYNTHESIS OF MAJOR NONCANCER EFFECTS 4-69
4.6.1 Human Studies 4-69
4.6.2. Animal Studies 4-69
4.6.2.1. Oral Exposure 4-69
4.6.2.2. Inhalation Exposure 4-70
4.7 EVALUATION OI CARCINOGENICITY 4-73
4.7.1. Synthesis of Human, Animal, and Other Supporting Evidence 4-74
4.7.1.1. Human 4-74
4.7.1.1.1. Evidence for Causality 4-77
4.7.1.2. Laboratory Animal 4-79
July 2010 iii DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.7.2. Summary of Overall Weight of Evidence 4-80
4.7.3. Mode-of-Action Information 4-83
4.7.3.1. Hypothesized Mode of Action 4-83
4.7.3.2. Experimental Support for the Hypothesized Mode of Action 4-83
4.7.3.3 Conclusions about the Hypothesized Mode of Action 4-86
4.8 SUSCEPTIBLE POPULATIONS AND LIFE STAGES 4-88
4.8.1. Possible Childhood Susceptibility 4-88
4.8.2. Possible Gender Differences 4-89
5. DOSE-RESPONSE ASSESSMENTS 5-1
5 .1 ORAL REFERENCE DOSE (RfD) 5-1
5.2 INHALATION REFERENCE CONCENTRATION (RlC) 5-1
5.2.1. Choice of Principal Study and Critical Effect(s) 5-1
5.2.2. Methods of Analysis 5-4
5.2.3. Exposure Duration and Dosimetric Adjustments 5-5
5.2.4. RfC Derivation—Including Application of Uncertainty Factors 5-9
5.2.5. Previous RfC Assessment 5-10
5.2.6. RfC Comparison Information 5-10
5 .3 UNCERTAINTIES IN THE INHALATION REFERENCE CONCENTRATION 5-11
5.4 CANCER ASSESSMENT 5-16
5.4.1. Choice of Study/Data—with Rationale and Justification 5-16
5.4.2. Dose-Response Data 5-16
5.4.3. Dose Adjustments and Extrapolation Methods 5-18
5.4.4. Oral Slope Factor and Inhalation Unit Risk 5-21
5 5-24
5.4.5 Application of Age-Dependent Adjustment Factors 5-27
5.4.6. Previous Cancer Assessment 5-27
5.4.7. Uncertainties in Cancer Risk Values 5-28
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF HAZARD AND DOSE RESPONSE 6-1
6.1 HUMAN HAZARD POTENTIAL 6-1
6.2 DOSE RESPONSE 6-4
7. REFERENCES 2
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS AND
DISPOSITION 1
Appendix B: Benchmark Dose Modeling Results for the Derivation of the RfC 1
Appendix C: Cancer Dose-Response Modeling 1
July 2010
iv DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 2-1. Physical properties and chemical identity of chloroprene 2-3
Table 3-1. Tissue-to-air partition coefficients for chloroprene 3-2
Table 3-2. Liver microsomal metabolites as a percentage of 1-butanol internal standard 3-3
Table 3-3. Steriochemical 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) 3-6
Table 3-4. Kinetic parameters used to describe the microsomal oxidation of chloroprene 3-8
Table 3-5. Kinetic parameters used to describe the microsomal epoxide hydrolase activity of
(l-chloroethenyl)oxirane 3-9
Table 3-6. Kinetic parameters used to describe the time course of (l-chloroethenyl)oxirane
formation from microsomal oxidation of chloroprene 3-10
Table 3-7. Kinetic parameters used to describe the cytosolic glutathione S-transferase activity
towards (l-chloroethenyl)oxirane 3-11
Table 3-8. Metabolic parameters of chloroprene 3-12
Table 3-9. Physiological parameters used for chloroprene PBPK modeling 3-13
Table 4-1. Standardized mortality ratios (SMRs) for the DuPont Louisville Works cohort
relative to general U.S. population rates 4-3
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 4-5
Table 4-3. Standardized mortality ratios (SMRs) for selected cancer risks relative to general
population rates of Moscow, Russia 4-7
Table 4-4. Selected relative risk (RRs) estimates for the high-exposure group relative to
unexposed factory workers 4-7
Table 4-5. Internal relative risks (RRs) by duration of employment in the high-exposure
category 4-8
Table 4-6. Selected standardized incidence ratios (SIRs) for chloroprene monomer cohort
relative to the general Armenian population 4-9
Table 4-7. Standardized incidence ratios (SIRs) for elevated cancer risks for plant workers
relative to general population rates of Isere, France 4-11
Table 4-8. Standardized mortality ratios (SMRs) at each of four chloroprene production
facilities 4-13
Table 4-9. Relative risks (RRs) for respiratory cancers by cumulative chloroprene exposure 4-15
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 4-17
July 2010
v DRAFT - DO NOT CITE OR QUOTE
-------
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 4-18
Table 4-12. Frequency of chromosomal aberrations in lymphocyte culture cells from
chloroprene production workers 4-24
Table 4-13. Tumor incidence in female BD IV rats treated orally with chloroprene (100 mg/kg)
on GDI7 and in their progeny treated (50 mg/kg) weekly for life (120 weeks) 4-26
Table 4-14. Distribution of tumors in female BD IV rats treated orally with chloroprene (100
mg/kg) on GD17 and their progeny treated (50mg/kg) weekly for life (120 weeks) 4-26
Table 4-15. Survival and body weights of rats in the 16-day inhalation study of chloroprene 4-28
Table 4-16. Incidences of selected nonneoplastic lesions in rats in the 16-day inhalation study
of chloroprene 4-28
Table 4-17. Survival and body weights of mice in the 16-day inhalation study of chloroprene 4-30
Table 4-18. Survival and body weights of rats in the 13-week inhalation study of chloroprene 4-31
Table 4-19. Incidences of selected nonneoplastic lesions in rats in the 13-week inhalation study
of chloroprene 4-32
Table 4-20. Survival and body weights of mice in the 13-week inhalation study of chloroprene 4-33
Table 4-21. Incidences of forestomach lesions in mice in the 13-week inhalation study of
chloroprene 4-34
Table 4-22. 2-Year survival probability estimates for F344/N rats chronically exposed (2
years) to chloroprene by inhalation 4-35
Table 4-23. Incidence and severity of non-neoplastic lesions in F344/N rats chronically
exposed (2 years) to chloroprene by inhalation 4-36
Table 4-24. Incidence of neoplasms in F344/N rats chronically exposed (2 years) to
chloroprene by inhalation 4-37
Table 4-25. 2-Year survival probabilities for B6C3F1 mice chronically exposed (2 years) to
chloroprene by inhalation 4-39
Table 4-26. Incidence and severity of non-neoplastic lesions in B6C3F1 mice chronically
exposed (2 years) to chloroprene by inhalation 4-40
Table 4-27. Incidence of neoplasms in B6C3F1 mice chronically exposed (2 years) to
chloroprene by inhalation 4-41
Table 4-28. Survival-adjusteda neoplasm rates for mice in the 2-year inhalation study of
chloroprene 4-43
Table 4-29. Selected mean relative organ weights of rats exposed for 24 months and hamsters
exposed for 18 months to chloroprene vapor 4-45
Table 4-30. Incidence, site and type of tumor in selected organs and tissues of rats exposed to
chloroprene for 24 months 4-47
Table 4-31. Incidence, site and type of tumor in selected organs and tissues of hamsters
exposed to chloroprene for 18 months 4-48
July 2010
vi DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-32. Summary of epididymal spermatozoal and estrous cycle parameters for rats and
mice in the 13-week study of chloroprene 4-51
Table 4-33. Results of teratology and embryotoxicity studies in rats exposed to chloroprene by
inhalation 4-53
Table 4-34. Incidence of anomalies in litters of rats exposed to chloroprene by inhalation 4-54
Table 4-35. Chloroprene-induced mortality in male rats 4-57
Table 4-36. Genotoxicity assays of chloroprene 4-63
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 4-68
Table 4-38 Quantitative comparison of carcinogenic potency of butadiene and chloroprene in
mice 4-69
Table 4-39. Summary of animal and human tumor data and weight of evidence descriptor for
chloroprene 4-82
Table 5-1. Incidences of nonneoplastic lesions resulting from chronic exposure (ppm) to
chloroprene considered for identification of critical effect 5-3
Table 5-2. Duration adjusted point of departure estimates for best fitting models of the BMD
from chronic exposure to chloroprene 5-6
Table 5-3. Summary of Uncertainties in the Chloroprene noncancer risk assessment 5-15
Table 5-4. Tumor incidence in female and male B6C3F1 mice exposed to chloroprene via
inhalation 5-17
Table 5-5. Tumor incidence in female and male F344 rats exposed to chloroprene via
inhalation 5-18
Table 5-6. Dose-response modeling summary for female mouse tumors associated with
inhalation exposure to chloroprene 5-23
Table 5-7. Dose-response modeling summary for male mouse tumor sites associated with
inhalation exposure to chloroprene 5-24
Table 5-8. Summary of uncertainties in chloroprene cancer unit risk estimate 5-28
Table B-l. Severity scores at control dose and lowest dose showing response for endpoints
considered for critical non-cancer effect 2
Table B-2. Benchmark modeling results for alvelolar epithelial hyperplasia in male F344/N
rats (BMR = 10% extra risk) 3
Table B-3. Benchmark modeling results for alvelolar epithelial hyperplasia in male F344/N
rats (BMR = 5% extra risk) 5
Table B-4. Benchmark modeling results for Alveolar epithelial hyperplasia in female F344/N
rats (BMR = 10% extra risk) 7
Table B-5. Benchmark modeling results for bronchiolar hyperplasia in male B6C3F1 mice
(BMR = 10% extra risk) 10
Table B-6. Benchmark modeling results for bronchi olar hyperplasia in male B6C3F1 mice
(BMR = 5% extra risk) 12
July 2010
vii DRAFT - DO NOT CITE OR QUOTE
-------
Table B-7. Benchmark modeling results for chronic inflammatioin in male F344/N rats (BMR
= 10% extra risk) 14
Table B-8. Benchmark modeling results for olfactory atrophy in male F344/N rats (BMR =
10% extra risk) 17
Table B-9. Benchmark modeling results for olfactory atrophy in male F344/N rats (BMR = 5%
extra risk) 19
Table B-10. Benchmark modeling results for olfactory necrosis in male F344/N rats (BMR =
10% extra risk) 21
Table B-l 1. Benchmark modeling results for olfactory necrosis in male F344/N rats (BMR =
5% extra risk) 23
Table B-12. Benchmark modeling results for olfactory necrosis in female F344/N rats (BMR =
10% extra risk) 25
Table B-13. Benchmark modeling results for olfactory necrosis in female F344/N rats (BMR =
5% extra risk) 29
Table B-14. Benchmark modeling results for olfactory basal cell hyperplasia in female F344/N
rats (BMR = 10% extra risk) 31
Table B-15. Benchmark modeling results for olfactory basal cell hyperplasia in female F344/N
rats (BMR = 5% extra risk) 35
Table B-16. Benchmark modeling results for kidney (renal tubule) hyperplasia in male F344/N
rats (BMR = 10% extra risk) 37
Table B-17. Benchmark modeling results for kidney (renal tubule) hyperplasia in female
F344/N rats (BMR = 10% extra risk) 40
Table B-l8. Benchmark modeling results for forestomach epithelial hyperplasia in male
B6C3F1 mice (BMR = 10% extra risk) 42
Table B-19. Benchmark modeling results for forestomach epithelial hyperplasia in female
B6C3F1 mice (BMR = 10% extra risk) 44
Table B-20. Benchmark modeling results for splenic hematopoietic cell proliferation in female
B6C3F1 mice (BMR = 10% extra risk) 46
Table B-21. Benchmark modeling results for splenic hematopoietic cell proliferation in female
B6C3F1 mice (BMR = 5% extra risk) 48
Table C-l: Tumor incidence, with time to death with tumor: female mice exposed to
chloroprene via inhalation (NTP, 1998) 1
Table C-2: Tumor incidence, with time to death with tumor: male mice exposed to
chloroprene via inhalation (NTP, 1998) 3
Table C-3. Summary of Model Selection and Modeling Results for best-fitting multistage-
Weibull models, using time-to-tumor data for female mice (NTP, 1998)Error! Bookmark not defined.
Table C-4. Summary of human equivalent overall cancer risk values estimated by R/BMDR,
based on male and female mouse tumor incidence (NTP, 1998) 1
July 2010
viii DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure 2-1. The chemical structure of chloroprene 2-1
Figure 3-1. Proposed metabolism of chloroprene 3-4
Figure B-l. Log-logistic model fit for alveolar epitheliala hyperplasia in male F344/N rats
(BMR = 10% extra risk) 3
Figure B-2. Log-logistic model fit for alveolar epitheliala hyperplasia in male F344/N rats
(BMR =5% extra risk) 5
Figure B-3. Log-logistic model fit for alveolar epithelial hyperplasia in female F344/N rats
(BMR = 10% extra risk) 7
Figure B-4. Log-logistic model fit for bronchiolar hyperplasia in male B6C3F1 mice (BMR =
10% extra risk) 10
Figure B-5. Log-logistic model fit for bronchi olar hyperplasia in male B6C3F1 mice (BMR =
5% extra risk) 12
Figure B-6. Log-logistic model fit for olfactory chronic inflammation in male F344/N rats
(BMR = 10% extra risk) 14
Figure B-7. Logistic model fit for olfactory atrophy in male F344/N rats (BMR = 10% extra
risk) 17
Figure B-8. Logistic model fit for olfactory atrophy in male F344/N rats (BMR = 5% extra
risk) 19
Figure B-9. Log-probit model fit for olfactory necrosis in male F344/N rats (BMR = 10% extra
risk) 21
Figure B-10. Log-probit model fit for olfactory necrosis in male F344/N rats (BMR = 5% extra
risk) 23
Figure B-l 1. Log-probit model fit for olfactory necrosis in female F344/N rats (BMR = 10%
extra risk) 25
Figure B-12. Dichotomous hill model fit for olfactory necrosis in female F344/N rats (BMR =
10% extra risk) 27
Figure B-13. Log-probit model fit for olfactory necrosis in female F344/N rats (BMR = 5%
extra risk) 29
Figure B-14. Log-probit model fit for olfactory basal cell hyperplasia in female F344/N rats
(BMR = 10% extra risk) 31
Figure B-l5. Dichotomous model fit for olfactory basal cell hyperplasia in female F344/N rats
(BMR = 10% extra risk) 33
Figure B-l6. Log-probit model fit for olfactory basal cell hyperplasia in female F344/N rats
(BMR = 5% extra risk) 35
Figure B-l7. Log-logistic model fit for kidney (renal tubule) hyperplasia in male F344/N rats
(BMR = 10% extra risk) 37
Figure B-18. Log-probit model fit for kidney (renal tubule) hyperplasia in female F344/N rats
(BMR = 10% extra risk) 40
July 2010
ix DRAFT - DO NOT CITE OR QUOTE
-------
Figure B-19. Multistage model fit for forestomach epithelial hyperplasia in male B6C3F1 mice
(BMR = 10% extra risk) 42
Figure B-20. Multistage model fit for forestomach epithelial hyperplasia in female B6C3F1
mice (BMR = 10% extra risk) 44
Figure B-21. Probit model fit for splenic hematopoietic cell proliferation in female B6C3F1
mice (BMR = 10% extra risk) 46
Figure B-22. Probit model fit for splenic hematopoietic cell proliferation in female B6C3F1
mice (BMR =5% extra risk) 48
Figure C-2. Female mice, hemangiomas and hemangiosarcomas in all organs; high dose
dropped, hemangiosarcomas occurring before termination considered fatal 11
Figure C-3. Female mice, hemangiomas and hemangiosarcomas in all organs; high dose
dropped, all tumors considered incidental. Details below 14
Figure C-4. Female mice, Harderian gland tumors. Details below 17
Figure C-5. Female mice, mammary gland tumors. Details below 20
Figure C-6. Female mice, forestomacFigure C-6: Female mice, forestomach tumors. Details
below 23
Figure C-7. Female mice, hepatocellular adenomas and carcinomas. Details below 26
Figure C-8. Female mice, skin sarcomas. Details below 29
Figure C-9. Female mice, Zymbal's gland tumors. Details below 32
Figure C-10. Male mice, alveolar/bronchiolar tumors. Details below 35
Figure C-l 1. Male mice, hemangiomas and hemangiosarcomas; hemangiosarcomas occurring
before termination considered fatal. Details below 38
Figure C-12. Male mice, hemangiomas and hemangiosarcomas; all tumors considered
incidental. Details below 41
Figure C-13. Male mice, Harderian gland tumors. Details below 44
Figure C-14. Male mice, renal tubule tumors. Details below 47
Figure C-15. Male mice, forestomach tumors. Details below 50
July 2010
x DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
1
2
ACGIH
American Conference of Governmental Industrial Hygienists
3
ADAF
age dependent adjustment factor
4
AEH
alveolar epithelial hyperplasia
5
AIC
Akaike Information Criterion
6
ALP
alkaline phosphatase
7
ALT
alanine aminotransferase
8
BMC
benchmark concentration
9
BMCL
lower bound on the benchmark concentration
10
BMD
benchmark dose
11
BMDL
lower confidence limit on the benchmark dose
12
BMDS
benchmark dose software
13
BMR
benchmark response
14
CASRN
Chemical Abstracts Service Registry Number
15
CI
confidence interval
16
CNS
central nervous system
17
CYP
cytochrome
18
DAF
dosimetric adjustment factor
19
DMSO
dimethyl sulfoxide
20
DNA
deoxyribonucleic acid
21
ED10
effective dose assocated with 10% excess risk
22
EH
epoxide hydralases
23
EPA
U.S. Environmental Protection Agency
24
GD
gestational day
25
GDH
glutamine dehydrogenase
26
GSH
glutathione
27
HEC
human equivalent concentration
28
IARC
International Agency for Research on Cancer
29
ICD
International Classification of Diseases
30
IPCS
International Programme on Chemical Safety
31
IRIS
Integrated Risk Information System
32
LOAEL
lowest-observed-adverse-effect level
33
LOH
loss of heterozygosity
34
M
Molar
35
MLE
maximum likelihood estimate
36
MOA
mode of action
37
MV
minute volume
38
NCEA
National Center for Environmental Assessment
39
NIOSH
National Institute for Occupational Safety and Health
40
NLM
National Library of Medicine
41
NOAEL
no-ob served-adverse-effect level
42
NPSH
nonprotein sulfhydryl
43
NRC
National Research Council
44
NTP
National Toxicology Program
45
OECD
Organisation for Economic Cooperation and Development
46
OSHA
Occupational Safety and Health Administration
47
P
probability value
48
PBPK
physiologically based pharmacokinetic (model)
July 2010
xi DRAFT - DO NOT CITE OR QUOTE
-------
1
PCB
polychlorinated biphenyl
2
PEL
permissible exposure limit
3
POD
point of departure
4
PPm
parts per million
5
PIT
pulmonary
6
R
level of risk
7
RBC
red blood cell
8
RfC
reference concentration
9
RfD
reference dose
10
RGDR
regional gas dose ratio
11
RR
relative risk
12
SA
surface area
13
SD
standard deviation
14
SDH
sorbitol dehydrogenase
15
SIR
standard incidence ratio
16
SMR
standardized mortality ratio
17
TLV
threshold limit value
18
UCL
upper confidence limit
19
UF
uncertainty factor
20
v/v
volume/volume
21
x2
chi squared
July 2010
xii DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
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).
July 2010
xiii DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
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.
Jeffrey Gift, Ph.D.
Karen Hogan, M.S.
Ines Pagan, D.V.M., Ph.D.
John Stanek, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C.
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.
CONTRACTOR SUPPORT
Karla Thrall, Ph.D.
Anthony Fristachi
Nick Heyer, Ph.D.
Paul Hinderliter, Ph.D.
Jessica Sanford, Ph.D.
Battelle Memorial Institute
REVIEWERS
July 2010
xiv DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
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 Assessemnt
Office of Research and Development
July 2010
xv DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
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
July 2010
xvi DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
"3
lifetime. The inhalation RfC (expressed in units of mg/m ) is analogous to the oral RfD, but provides a
continuous inhalation exposure estimate. The inhalation RfC considers toxic effects for both the
respiratory system (portal of entry) and for effects peripheral to the respiratory system (extrarespiratory
or systemic effects). Reference values are generally derived for chronic exposures (up to a lifetime),
but may also be derived for acute (< 24 hours), short-term (> 24 hours up to 30 days), and subchronic
(> 30 days up to 10% of lifetime) exposure durations, all of which are derived based on an assumption
of continuous exposure throughout the duration specified. Unless specified otherwise, the RfD and
RfC are derived for chronic exposure duration.
The carcinogenicity assessment provides information on the carcinogenic hazard potential of
the substance in question and quantitative estimates of risk from oral and inhalation exposure may be
derived. The information includes a weight-of-evidence judgment of the likelihood that the agent is a
human carcinogen and the conditions under which the carcinogenic effects may be expressed.
Quantitative risk estimates may be derived from the application of a low-dose extrapolation procedure.
If derived, the oral slope factor is a plausible upper bound on the estimate of risk per mg/kg-day of oral
"3
exposure. Similarly, an inhalation unit risk is a plausible upper bound on the estimate of risk per |ig/m
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
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
July 2010 1-1 DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
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). Benchmark Dose Technical Guidance Document (U.S. EPA, 2000,
052150), Supplementary Guidance for Conducting Health Risk Assessment of Chemical Mixtures
(U.S. EPA, 2000, 196144), 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 2009.
July 2010
1-2
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
2. CHEMICAL AND PHYSICAL INFORMATION
Beta-chloroprene monomer (C4H5CI) (hereafter referred to as chloroprene) is a volatile,
flammable liquid used primarily in the manufacture of polychloroprene or neoprene rubber (U.S.,
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 polychloroprene solid (trade names include neoprene, Bayprene, etc.),
about 10% is converted to polychloroprene liquid dispersions, a colloidal suspension of
polychloroprene in water (IARC, 1999, 201838). There was one commercial producer of chloroprene
in the United States 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 lb/yr is sold for research applications1.
The total estimated production of polychloroprene from 1986 to 1988 was approximately 250 to 300
million lb (113,000 to 136,000 metric tons), and the volume produced from 1995 to 1996 was
approximately 200 to 250 million lb (90,700 to 113,000 metric tons) (NTP, 2005, 093207)2.
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 lbs (NTP, 2005,
093207). Eight of these plants reported individual atmospheric releases from 2 to 481,871 lbs (NTP,
2005, 093207). Three plants in Kentucky, Texas, and Louisiana, each reporting atmospheric releases
of > 100,000 lbs, accounted for most of the reported chloroprene releases in 1995 . 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.
Figure 2-1. The chemical structure of chloroprene.
1 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 United States; this site both manufactured
the the monomer and converted it to polymer. Chloroprene is used almost exclusively to produce polychloroprene, with
chloroprene monomer sold to only one US company for non-polychloroprene manufacture (1000 lbs in 2008).
2 According to DuPont's public comments, chloroprene production has decreased since 1996 and in 2008, US production
volume was below 40,000 metric tons.
3 According to DuPont's public comments, in 2008, only one chlorprene plant remained open and reported releases of
210,900 lbs. Domestic production and releases have been decreasing (reported 2002 emissions were 356,700 lbs).
July 2010
2-1 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
The starting material for the synthesis of chloroprene is currently 1,3-butadiene in the United
States (Lynch, 2001, 625182). Chloroprene manufacture using butadiene as the starting material
occurs via a two step process consisting of chlorination and subsequent dehydrochlorination reactions.
Initial industrial processes (1930s-70s) 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 as far as having
a chlorine bound to a double-bonded carbon (alkene) backbone. However, chloroprene contains four
carbons arranged with two double bonds. The odor of chloroprene is described as pungent and ether-
like (NLM, 2010, 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 mm Hg 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 *OH (to produce formaldehyde), O3, and NO3 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 or
inhibitors must be added to prevent peroxide formation and consequent spontaneous polymerization;
inhibitors do not reduce dimer formation. 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, chemically 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, 625182). 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 that is released to soil is to leach
into groundwater; 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 the 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
July 2010
2-2
DRAFT - DO NOT CITE OR QUOTE
-------
1 where chloroprene is produced and converted to polychloroprene (Lynch, 2001, 625182)4. The
2 physical and chemical properties of chloroprene are shown in Table 2-1.
Table 2-1. Physical properties and chemical identity of chloroprene
CHLOROPRENE
REFERENCE
CASRN
126-99-8
NLM (2010, 594343)
Synonyms
1,3-butadiene, 2-chloro; chlorobutadiene, 2-chlorobutadiene;
2-chlorobutadiene-1,3; beta-chloroprene
NLM (2010. 594343)
Melting point
-130°C
NLM (2010. 594343)
Boiling point
59.4°C
NLM (2010. 594343)
Density
0.956 at 20°C (relative to the density of H20 at 4°C)
NLM (2010. 594343)
Vapor pressure
215 mm Hg at 25°C
NLM (2010. 594343)
Vapor density
3.0 (air = 1)
NLM (2010. 594343)
Flashpoint (open cup)
-20 °C
NLM (2010, 594343)
Flammability limits
4-20% in air
NLM (2010. 594343)
Water solubility
256-480 mg/L at 20°C
OECD (1998. 624889)
Other solubilities
Miscible with ethyl ether, acetone, benzene; soluble in
alcohol, diethyl ether
NLM (2010. 594343)
Log K0w
2.2
OECD (1998, 624889)
Henry's law constant
5.6 x 10~2 atm/m3-mol at 25°C
NLM (2010. 594343)
Odor threshold
15 ppm (54 mg/m3)
U.S. EPA (2000, 625036)
Molecular weight
88.54
NLM (2010, 594343)
Conversion factors (in air)
1 mg/m3 = 0.276 ppm; 1 ppm = 3.62 mg/m3 at 25°C, 760 torr
NLM (2010. 594343)
Molecular formula
C4H5CI
NLM (2010. 594343)
4 According to DuPont's public comments, as of 2008, occupational exposure potential to chloroprene in the US is limited
to one site in Louisiana; other chloroprene manufacturing facilities exist in Germany, France, Armenia/Azerbaijan, India,
China, and Japan.
July 2010
DRAFT - DO NOT CITE OR QUOTE
2-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
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 states that chloroprene is "rapidly absorbed by the
skin" (Lefaux, 1968, 625192; NLM, 2010, 594343). 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 |il 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
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
July. 2010
3-1
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
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
TISSUE-TO-AIR PARTITION COEFFICIENTS3
Mouse
F344 rat
Wistar rat
Hamster
Humanb
Blood
7.8 ±0.1
7.3 ±0.1
8.0 ±0.5
9.3 ±0.3
4.5 ±0.1
Lung
18.6 ±5.1
13.5 ±1.6
11.2 ± 0.5
9.7 ±0.6
13.3 ±4.1
Liver
9.8 ±0.9
11.5 ± 0.3
10.9 ±0.2
10.5 ±0.5
10.7 ± 1.1
Fat
135.3 ± 1.6
124.0 ±1.5
126.3 ± 1.4
130.1 ±0.9
128.9 ±2.7
Muscle
4.6 ±0.8
4.4 ±0.4
4.0 ±0.3
5.0 ±0.2
4.5 ± 1.0
Kidney
13.7 ±0.6
16.7 ±0.6
9.4 ±0.4
8.2 ±0.3
12.0 ±0.9
aMean ± standard error for three replicates per rodent tissue.
bHuman blood values determined for nine replicates (three subjects, three replicates/subject); human tissue partition
coefficient values were derived from rodents with standard error adjusted to account for the proportion of variation from
each set of rodent data.
Source: 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 B6C3F1 mice, Fisher and Wistar rats, and Syrian hamsters.
July. 2010
3-2
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
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 spectrograph!c 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 B6C3F1 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 |iM between 5-10 minutes after start of incubation with [0.05
|iM]100 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 (1-
chloroethenyl)oxirane most likely involved CYP 2E1, 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
standard
METABOLITE PEAK3
LIVER MICROSOMAL SUSPENSION
B6C3F1 mouseb
F344 rat
Wistar rat
Hamster
Human
1
9.0
12.0
4.0
0.8
1.3
2
0.0
0.1
0.1
0.2
0.1
3
0.8
0.3
0.2
0.8
0.3
4
0.2
0.0
0.1
0.4
0.1
5
0.2
0.3
0.0
0.1
0.0
6
0.6
0.4
0.3
0.3
0.1
aMetabolite 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.
b One vial was used for each species
Source: 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 (lchloroethenyl)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 > B6C3F1 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
July. 2010
3-3
DRAFT - DO NOT CITE OR QUOTE
-------
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.
CI
P450
EH
O
l-(chloroethenyl)
oxirane a
OH
OH
3-chlorobut-3-ene-
l,2-diola
chloroprene P450
CI
EH/H20
2-chloro-2-
ethenyloxirane1
rearrange-
ments
1 -hy dro x y but-3 -en-
2-one
GSH
conjugate
l-chlorobut-3-
en-2-one
2-o ne
X
2-chlorobut-3-
en-l-al
o
1-chlorobutan- l-chloro-2-hydroxy-
but-3-ene
OH
1-hydroxybutan-
2-one
CI
2-chlorobut-2-
en-l-al
0
2-chlo r o-butanal
Figure 3-1. Proposed metabolism of chloroprene.
a R- and S- enantiomers
Source: Adapted from Cottrell et al. (2001, 157445).
July. 2010
3-4
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Comparing metabolism between species, Cottrell et al. (2001, 157445) observed that qualitative
profiles of metabolites from liver microsomes obtained from B6C3F1 mice, Sprague-Dawley or F344
rats, and humans were similar. Microsomal suspensions were prepared by differential centrifugation
from livers pooled from male and female B6C3F1 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 B6C3F1 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 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 (1-
chloroethenyl)oxirane, suggesting that conjugation with glutathione may not be an active
detoxification pathway for the active epoxide metabolite of chloroprene. Glutahione conjugation was
apparent with l-hydroxybut-3-en-2-one, the downstream product of the minor epoxide metabolite of
chloroprene, 2-chloro-2-ethenyloxirane.
July. 2010
3-5
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Table 3-3. Steriochemical 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
FEMALE
Chloroprene
(mM)
Species/straina'b
% R
% S
Chloroprene
(mM)
Species/straina'b
% R
% S
5
Sprague-Dawley
rat
58
42
Sprague-Dawley
rat
10
62
38
10
56
44
20
61
39
20
56
44
30
60
40
30
55
45
40
64
36
40
59
41
5
F344 rat
62
38
F344 rat
10
62
38
10
56
46
20
62
38
20
54
46
30
60
40
30
53
47
40
64
36
40
54
46
5
B6C3F1 mouse
48
52
B6C3F1 mouse
10
47
53
10
47
53
20
46
54
20
45
55
30
47
53
30
47
53
40
47
53
40
46
54
10
Human
43
57
10
Human
43
57
20
43
57
20
44
56
30
43
57
30
42
58
aAverage of three samples per species/strain.
Percentage estimated error ± 1%.
Source: 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
10 |iM 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 |iM and
10 mM chloroprene. For the S-enantiomer, the presence of (l-chloroethenyl)oxirane was detected in
only mouse microsomes after incubations with 10 |iM chloroprene. After incubations with 100 |iM
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 (1-
July. 2010
3-6
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 (1-
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 3mM 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.
Pretreatment 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.
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 (1-
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
July. 2010
3-7
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
and (l-chloroethenyl)oxirane metabolism in the rodent and human liver and lung microsomes. 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 (see 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 (see Table 3-5). Hydrolysis
(Vmax/Km) of (l-chloroethenyl)oxirane in liver and lung microsomes was fastest for humans, followed
by hamsters, rats, and mice..
Table 3-4. Kinetic parameters used to describe the microsomal oxidation of
chloroprene
TISSUE
SPECIES
ACTIVITY OF MICROSOMAL OXIDATION
V a
v max
Km
Vmax/^m
Liver
Mouse
0.23
1.03
224
F344 rat
0.078
0.53
146
Wistar rat
0.11
0.84
125
Hamster
0.29
1.33
218
Human
0.068
0.68
101
Lung
Mouse
0.10
1.5
66.7
F344 rat
-
-
1.3b
Wistar rat
-
-
1.3b
Hamster
-
-
1.3b
Human
-
-
1.3b
a Values derived from modeling of vial headspace concentration time-course data (using liquid-to-air partition coefficient =
0.69) (Himmelstein et al., 2001, 019012). Vmax, (amol/h/mg protein, Km, (imol/1, Vmax/Km, ml/h/mg protein
b The apparent rate of lung metabolism, over the range of biologically relevant concentrations tested, was linear and was
estimated as Vmax/Km
Source: Himmelstein et al. (2004, 625152)
July. 2010
3-8
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Table 3-5. Kinetic parameters used to describe the microsomal epoxide hydrolase
activity of (l-chloroethenyl)oxirane
TISSUE
SPECIES
ACTIVITY OF MICROSOMAL EPOXIDE HYDROLASE
V a
v max
Km
in
Liver
Mouse
0.14
20.9
6.7
F344 rat
0.60
41.5
14.5
Wistar rat
0.64
53.0
12.1
Hamster
2.49
73.8
33.7
Human
3.66
99.7
36.7
Lung
Mouse
0.11
51.5
2.1
F344 rat
0.12
90.9
1.3
Wistar rat
0.16
91.6
1.7
Hamster
1.34
187.6
7.1
Human
0.58
72.2
8.0
a Values derived from modeling of vial headspace concentratioin time-course data (using liquid-to-air partition coefficient =
57.9) (Himmelstein et al., 2001. 019012). Vmax, (amol/h/mg protein, Km, (imol/1, Vmax/Km, ml/h/mg protein
Source: Himmelstein et al. (2004, 625152)
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+, pretreatment of
mouse microsomal preperations 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 (1-
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 (lchloroethenyl)oxirane in the vial headspace. Peak concentrations of
(lchloroethenyl)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 (1-
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.
July. 2010
3-9
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
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 (see 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 2-fold greater compared to humans.
Table 3-6. Kinetic parameters used to describe the time course of (1-
chloroethenyl)oxirane formation from microsomal oxidation of chloroprene
TISSUE
SPECIES
(l-CHLOROETHENYL)OXIRANE FORMATION
V a
v max
Km
in
Ratio of (1 -
chloroethenyl)oxirane/
total (%)b
Liver
Mouse
0.149
36.6
4.1
2
F344 rat
0.184
23.7
7.8
5
Wistar rat
0.148
25.3
5.8
5
Hamster
0.048
9.0
5.4
2
Human
0.108
20.7
5.2
5
Lung
Mouse
0.050
25.0
2.0
3
F344 rat
0.0075
40.4
0.19
15
Wistar rat
0.0082
30.1
0.27
22
Hamster
0.013
81.2
0.16
13
Human
0.024
24.6
0.98
78
a Optimized oxidative rate constants used to describe the amount of (l-chloroethenyl)oxirane derived from total
chloroprene oxidation. Vmax, (imol/h/mg protein, Km, (imol/1, Vmax/Km, ml/h/mg protein
b Vmax/Kmfor (l-chlororethenyl)oxirane formation divided by the Vmax/Km for total chloroprene oxidation (from Table 3-
4) multipled by 100
Source: 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.
July. 2010
3-10
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Table 3-7. Kinetic parameters used to describe the cytosolic glutathione S-
transferase activity towards (l-chloroethenyl)oxirane
TISSUE
SPECIES
ACTIVITY OF CYTOSOLIC GLUTATHIONE ^-TRANSFERASE3
ks
qBS(0)
ks x CBS(0)
Liver
Mouse
0.0015
2.7
0.0040
F344 rat
0.0074
0.92
0.0068
Wistar rat
0.011
0.56
0.0063
Hamster
0.024
0.54
0.0130
Human
0.0017
0.94
0.0016
Lung
Mouse
0.0011
2.01
0.0022
F344 rat
0.0023
0.70
0.0016
Wistar rat
0.0051
0.18
0.00092
Hamster
0.015
0.038
0.00056
Human
0.0028
0.44
0.0012
Note: ks (l/|imol/h/mg cytosolic protein), rate constant CBS(0) ((imol/1) 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
aFirst order reaction of (l-chloroethenyl)oxirane with glutathione was measured as kf = 0.07 h"1 independent of
protein
Source: 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 (B6C3F1), 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 B6C3F1 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. A known volume of concentrated chloroprene
was added to the chamber at the start of the exposures, 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 3-7) was sufficient to explain the in vivo gas uptake data. Inhibition of
uptake was obtained with pre-treatment with 4-methyl pyrazole, indicating the 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.
July. 2010
3-11
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Table 3-8. Metabolic parameters of chloroprene
BIOCHEMICAL PARAMETERS3
SPECIES
Mouse
F344 rat
Wistar rat
Hamster
Liver
Vmax (mg/h/kg BW)
39.2
11.50
15.5
42.8
Km (mg/L)
0.091
0.047
0.075
0.118
Vmax/Km (L/h/kg BW)
431.0
244.0
208.0
363.0
Lung
Vmax (mg/h/kg BW)
1.02
...
...
...
Km (mg/L)
0.13
...
...
...
Vmax/Km (L/h/kg BW)
7.67
0.14
0.14
0.14
aScaled from Himmelstein et al. (2004, 625152) using microsomal protein content.
Source: 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
thioethers (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 there may exist differences in
chloroprene clearance across species. For example, while the fat:air partition coefficient is similar for
all species investigated (see 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 elimantion profiles.
3 5 PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS
Himmelstein et al. (2004, 625154) published a physiologically based toxicokinetic (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 compartments and described by Michaelis-Menten type
July. 2010 3-12 DRAFT - DO NOT CITE OR QUOTE
-------
1 saturable kinetics. Standard physiological values were used to parameterize the model. Tissue-to-
2 blood partition coefficients were calculated from tissue-to-air values and the in vivo metabolic
3 parameters (see Table 3-7) were scaled from in vitro metabolic parameters for total chloroprene
4 metabolism in the liver and lung (Himmelstein et al., 2004, 625152) using microsomal protein content.
5 Microsomal protein contents for the liver differ among species and were obtained from the literature.
6 The microsomal protein content for the lungs was set as equal for all species. Gas uptake was modeled
7 by subtracting the amount taken up by the animal from the chloroprene concentration in the chamber.
8 Physiological and metabolic parameters were not adjusted except for alveolar ventilation and cardiac
9 output as need 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 reponse modeling ^
Body weight (kg)
0.03
0.25
0.25
0.11
70
Ventilation (L/h/kg)
30
21
21
30
16.2
Cardiac output (L/h/kg)
30
18
18
30
16.2
Values for simulation of chamber gas uptake0
Body weight (kg)
0.024-0.034
0.16-0.28
0.20-0.34
0.10-0.18
NA
Ventilation (L/h/kg)
15
10.5
10.5
12
NA
Cardiac output (L/h/kg)
15
9
9
12
NA
Tissue volumes (% body weight)3,11
Liver
5.5
4.0
4.0
4.0
2.6
Fat
5.0
7.0
7.0
7.0
21.4
Rapid perfused
3.5
5.0
5.0
5.0
7.7
Slow perfused
77.0
75.0
75.0
75.0
56.1
Lung
0.73
0.50
0.50
0.50
0.76
Blood flow (% cardiac output)3^
Liver
16.1
18.3
18.3
18.3
22.7
Fat
7.0
7.0
7.0
7.0
5.2
Rapid perfused
51.0
51.0
51.0
51.0
47.2
Slow perfused
15.0
15.0
15.0
15.0
24.9
aParameters 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.
b Values used for the dose-response modeling based on average body weight data from chronic inhalation studies and
ssumption that literature values for ventilation and cardiac output are representative of repear inhalation exposure conditions
0 Values used specifically for simulation of closed chamber gas uptake data
d Tissue volumes and blood flows are calculated by the model with resulting units of liters (L) and L/h, respectively
Source: Himmelstein et al. (2004, 625154)
July. 2010
3-13
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 pretreatment, 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 non
cancer 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 the 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
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
July. 2010 3-14 DRAFT - DO NOT CITE OR QUOTE
-------
1 test the metabolic parameters and validate the model. However, Himmelstein et al. (2004, 625154) did
2 not provide results of sensitivity analyses indicating whether chamber loss was sensitive to
3 metabolism, and therefore it is uncertain whether chamber loss is useful for testing the metabolic
4 parameters used in the model. Also, the chamber data were fit by varying alveolar ventilation and
5 cardiac output. This method does not result in adequate testing of the model and does not validate the
6 scaled in vitro metabolic parameters. Additionally, there are currently no blood or tissue time-course
7 concentration data available for model validation. Therefore, as the model is currently constructed, the
8 PBPK model for chloroprene is inadequate for application for calculation of internal dose metrics or
9 interspecies dosimetry extrapolations.
10
July. 2010
3-15
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
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 (Pell,
1978, 064957; Li et al., 1989, 625181; Bulbulyan et al., 1998, 625105; Bulbulyan et al., 1999, 157419;
Colonna and Laydevant, 2001, 625112). 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
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.
July 2010
4-1 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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 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. Atotal 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.
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 also resulted in some unidentified cancer deaths
that could have been associated with earlier higher exposures. Moreover, as pointed out by Leet and
July 2010
4-2 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
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
TOTAL COHORT
CASES, SMR (95% CIa)
LOW-EXPOSURE
CASES, SMR (95% CI)
HIGH-EXPOSURE
CASES, SMR (95% CI)
All Causes
193,79 (68-91)
102, 82 (67-100)
91,75 (61-92)
All Cancers
51, 107 (80-141)
26, 107 (70-157)
25, 107 (69-158)
Digestive
19, 145 (87-227)
11, 164 (82-294)
8, 125 (54-246)
Biliary/liver
4, 571 (156-1463)
1,250 (6-1393)
3,750 (155-2192)
Trachea, bronchus, lung
17, 106 (62-170)
7, 86 (35-178)
10, 128 (61-236)
Lymphatic, hematopoietic
7, 140 (56-288)
3, 120 (25-351)
4, 160 (44-410)
aCI = confidence interval.
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 and 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 the 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
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
July 2010
4-3 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
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
(see 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.5 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 polymer mechanics. 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).
5 Person-years accrued were not reported in the paper.
July 2010 4-4 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
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
NUMBER OF DEATHS/SMR
ALL CAUSE
LIVER CANCER
LUNG CANCER
Total cohort
16/271*
6/242
2/513
Monomer workshop
8/377
4/482"
1/714
Vinylacetylene operator3
0/—
0/—
0/—
Monomer operator
4/450"
2/465
0/—
Maintenance mechanic3
4/1,290"
2/1,667"
1/5,000"
Neoprene workshop
5/176
2/165
1/556
Polymer operator3
5/394"
2/357
1/1250
Final treatment
0/—
Maintenance mechanic3
0/—
Laboratory
3/319
0/—
0/—
Quality monitor3
1/129
0/—
0/—
Researcher
21/176"
OA-
0/—
Statistical significance: *p < 0.01; **p < 0.05.
a: High-exposure Area
Source: Li et al. (1989, 625181)
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 & 1976-1983) then the SMRs
reported in the study would be overestimated due to a lower expected number of deaths. If mortality
was 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 non-cancer deaths among plant workers
based upon gender, age (± 2 years) and date of death (± 2 years). The authors observed that 16 of the
cancer deaths (30%) were among workers exposed to chloroprene compared to only four of the non-
cancer deaths (7%), yielding an odds ratio of 13 (p < 0.005). Although the average age at death was
July 2010 4-5 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
polychloroprene 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
3 3
from a high of 20 mg/m (gluers in the finishing department) to an intermediate level of 0.4-1 mg/m
(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. APoisson distribution was used to calculate the
95% CIs. One hundred thirty-one (2.5%) workers 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 for lung cancer was
significant in males only.
July 2010
4-6 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Table 4-3. Standardized mortality ratios (SMRs) for selected cancer risks relative
to general population rates of Moscow, Russia.
CAUSE OF DEATH
TOTAL COHORT
CASES, SMR (95% CI)
MEN
CASES, SMR (95% CI)
WOMEN
CASES, SMR (95% CI)
All causes
900; 103 (97-110)
181;121* (104-140)
719; 100 (93-107)
All cancers
265; 122* (107-137)
56; 158* (119-205)
209;115* (100-131)
Liver cancer
10; 240* (110-430)
2; 240 (30-860)
8; 230* (100-460)
Lung cancer
31; 140 (90-200)
17; 170* (100-270)
14; 110 (60-190)
Leukemia
13; 190* (100-330)
2; 190 (20-700)
11;190* (100-350)
* Statistical significance p < 0.05.
Source: 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 (RRs) estimates for the high-exposure group
relative to unexposed factory workers
CAUSE OF DEATH
HIGH-EXPOSURE CASES
HIGH-EXPOSURE RR (95% CI)a
All causes
194
1.23* (1.02-1.49)
Liver cancer
3
4.9 (0.5-47)
Colon cancer
8
2.6 (0.8-7.9)
Kidney cancer
2
3.3 (0.3-37)
Leukemia
5
2.2 (0.6-8.4)
* Statistical significance p < 0.05.
a Reference group is defined as workers with no chloroprene exposure.
Source: Bulbulyan et al. (1998, 625105)
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)
July 2010
4-7 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
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
1-9 YEARS
CASES; RR (95% CI)
10-19 YEARS
CASES; RR (95% CI)
20+ YEARS
CASES; RR (95% CI)
TREND
Liver cancer
1; 2.7 (0.2-45)
1; 8.3 (0.5-141)
1; 45.0 (2.2-903)
p = 0.02
Leukemia
2; 1.3 (0.2-7.3)
2; 3.4 (0.6-19)
1; 8.8 (0.7-66)
p = 0.07
Source: 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 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
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 (3%)
individuals 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.
"3
Before 1980, exposures averaged 5.59-69.80 mg/m (1.54-19.3 ppm) during the summer and 2.30-
249.5 mg/m (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
July 2010
4-8 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
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-
1050). 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
OBSERVED
SIR (95% CI)
All cancers
37
68 (49-94)
Lung cancer
6
53 (24-119)
Liver cancer
6
327* (147-727)
* Statistical significance p < 0.05.
Source: Bulbulyan et al. (1999, 157419)
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: 230-1630). Similar findings were noted in analyses using cumulative exposure,(unit-
years) with a statistically significant SIR of 486 (95% CI: 202-1170) 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.
July 2010 4-9 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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
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
July 2010
4-10 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
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
TOTAL COHORT
CASES, SIR (95% CI)
COHORT EXPOSED BEFORE 1977
CASES, SIR (95% CI)
All Cancers
34, 126 (88-177)
32, 146 (100-206)
Head and Neck
9, 189 (87-359)
8, 209 (90-411)
Larynx
3,243 (50-713)
3,297 (61-868)
Lung
9, 184 (84-349)
8, 199 (86-391)
Liver
1, 136 (4-763)
1, 164 (5-913)
Colon/Rectal
2, 66 (8-239)
2, 79 (10-287)
Source: 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), 3 in those with 11-20 years exposure (SIR = 149) and 1 in those with <_10 years
exposure (SIR = 106). 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: 127-1191), 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
cancer, no additional information was provided in the paper. This study found only one incident 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
July 2010 4-11 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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.
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.
July 2010
4-12 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
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 non-significant,
Table 4-8. Standardized mortality ratios (SMRs) at each of four chloroprene
production facilities
CAUSE OF
DEATH
LOUISVILLE
(L) CASES,
SMR (95% CI)a
MAYDOWN
(M) CASES,
SMR (95% CI)b
PONTCHARTRAIN
(P) CASES, SMR
(95% CI)a
GRENOBLE
(G) CASES,
SMR (95% CI)b
TOTAL
CASES, SMR
(95% CI)
All Causes
2403
74 (71-77)
435
60 (55-67)
102
53 (43-65)
62
65 (50-83)
3002
70 (67-73)
All Cancers
652
75 (69-80)
128
68 (56-80)
34
68 (47-95)
20
59 (36-91)
834
73 (68-78)
Respiratory
Cancers
266
75 (66-85)
48
79 (58-105)
12
62 (32-109)
10
85 (41-156)
336
75 (68-74)
All Cancers:
Exposed
651
74 (69-80)
114
62 (51-75)
26
57 (37-84)
15
59 (33-97)
806
71 (66-76)
Unexposed
1
99 (3-551)
14
126 (69-212)
8
144 (62-285)
5
61 (20-142)
28
108 (72-156)
aLocal county comparisons
National comparisons
Source: 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
(see 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
July 2010 4-13 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
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 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.
July 2010
4-14 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Table 4-9. Relative risks (RRs) for respiratory cancers by cumulative chloroprene
exposure
PLANT
LEVEL 1*
(LOWEST)
N
LEVEL 2
N, RR (95% CI)
LEVEL 3
N, RR (95% CI)
LEVEL 4
N, RR (95% CI)
TREND
Louisville (L)
62, Reference
67, 1.00 (0.71-1.43)
77, 1.32 (0.94-1.88)
60,0.85 (0.58-1.23)
p = 0.71
Maydown (M)
14, Reference
9, 1.65 (0.66-4.15)
12, 1.89 (0.72-4.96)
13,2.28 (0.86-6.01)
p = 0.10
Pontchartrain (P)
3, Reference
3, 1.60 (0.20-12.8)
2,2.90 (0.20-34.1)
4,2.32 (0.30-21.8)
p = 0.34
Grenoble (G)
2, Reference
1,0.61 (0.05-6.76)
4, 2.87 (0.35-39.7)
3,3.14 (0.30-48.0)
p = 0.17
*Chloroprene exposure (in ppm years) levels varied by plant: L (<4.7->164.1); M (<0.04->24.5); P (<0.02->16.2);
G (<0.05->23.9).
Source: 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 (Marsh et al., 2007, 625187; Marsh et al., 2007, 625188; Leonard et
al., 2007, 625179) 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
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
July 2010
4-15 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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 ofRelevant MethodologicalIssues
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 1000 workers, while
the remaining cohorts had fewer than 6000. The most consistent finding was excess liver (Bulbulyan et
al., 1999, 157419; Bulbulyan et al., 1998, 625105; Li et al., 1989, 625181; Leet and Selevan, 1982,
094970) and lung/respiratory system (Marsh et al., 2007, 625188; Colonna and Laydevant, 2001,
625112; Bulbulyan et al., 1999, 157419; Bulbulyan et al., 1998, 625105; Leet and Selevan, 1982,
094970; 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 also 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.
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
July 2010 4-16 DRAFT - DO NOT CITE OR QUOTE
-------
1 quantitative assessments, exposure levels were differentiated based upon job titles and industrial
2 hygiene knowledge of the processes involved. Although measurement error is present in all studies to
3 varying degrees, there is no evidence that this error differed by outcome (i.e., was non-differential) in
4 these studies. Although there are rare exceptions, non-differential misclassification of workers'
5 exposures due to lack of information usually results in an underestimate of the association between
6 exposure and outcome.
7
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 populations9 and
relative risks (RRs) for intermediate and high chloroprene exposures
STUDY
TOTAL COHORT
SMR/SIR (95% CI)1
INTERMEDIATE-
EXPOSURE
SMR/SIR/RRb (95%
CI)
HIGH-EXPOSURE
SMR/SIR /RRb
(95% CI)
(Bulbulvan et al.. 1998. 625105)
140 (90-200)
1.0 (0.4-2.5)c>d
0.8 (0.3-2.4)°'d
(Bulbulvan etal.. 1999. 157419)
50 (16-155)
(Colonna and Lavdevant. 2001. 625112)
184 (84-349)e
125 (15—45l)e
123 (26-361)e
(Leet and Selevan. 1982. 094970)
106 (62-170)
86 (35-178)f
128 (61-236)
(Marsh et al.. 2007. 625187; Marsh et al..
2007. 625188)-Louisville
75 (66-85)
92 (73-115)d
65 (50-84)4
(Marsh et al.. 2007. 625187; Marsh et al..
2007. 625188) - Mavdown
79 (58-105)
97 (50-169)d
113 (60-192)4
(Marsh et al.. 2007. 625187; Marsh et al..
2007. 625188-Pontchartrain) - Pontchartrain
62 (32-109)
96 (12-348)d
85 (23-218)d
(Marsh et al.. 2007. 625187; Marsh et al..
2007. 625188) - Grenoble
85 (41-156)
119 (32-304)d
128 (26-373)4
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
dCumulative chloroprene exposures
"Standardized incidence ratios
fLow-exposure group
July 2010
4-17 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
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
TOTAL COHORT
SMRa (95% CI)
INTERMEDIATE-
EXPOSURE
SMR/RRb (95% CI)
HIGH-EXPOSURE
SMR/RRb (95% CI)
(Bulbulvan et al.. 1998. 625105)
240(110-430)
7.1 (0.8-61)°'d
4.4 (0.4-44)°'d
(Bulbulvan etal.. 1999. 157419)
339 (109-1050)
293 (41-2080)^
486 (202-1170)^
(Colonna and Lavdevant. 2001. 625112)
136 (4-763)e
(Leet and Selevan. 1982. 094970)
571 (156-1463)
250 (6-1393)a
750 (155-2192)a
(Li et al.. 1989. 625181)
482 (N/R, p< 0.01)
(Marsh et al., 2007, 625187; Marsh et al.,
2007,625188) - Louisville
90 (52-144)
5.1 (0.9, 54.5)°'d
3.3 (0.5, 39.3)°'d
(Marsh et al., 2007, 625187; Marsh et al.,
2007. 625188)-Mavdown
24 (1-134)
(Marsh et al.. 2007. 625187; Marsh et al..
2007. 625188) - Pontchartrain
N/R: Not Reported
aSMRs and SIRs calculated relative to external population rates and are reported on a 100-base scale, unless noted all values
are SMRs
Relative to low or unexposed groups
°Relative risk of death from liver cancer
Cumulative chloroprene exposures
"Standardized incidence ratio
Lung Cancer Summary
An increased risk of lung cancer incidence and mortality was observed in a few studies
(Colonna and Laydevant, 2001, 625112; Bulbulyan et al., 1998, 625105; Pell, 1978, 064957; Li et al.,
1989, 625181; Leonard et al., 2007, 625179), 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.
July 2010 4-18 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Liver Cancer Summary
Statistically significant excesses of liver cancers were detected in four studies examining four
cohorts (Bulbulyan et al., 1999, 157419; Bulbulyan et al., 1998, 625105; Li et al., 1989, 625181; Leet
and Selevan, 1982, 094970). 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 1.2 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 (Marsh et al., 2007, 625187; Marsh et al., 2007, 625188; Colonna and
Laydevant, 2001, 625112) 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
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 (see
above).
July 2010 4-19 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
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) (see 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, Sub chronic-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:
currently exposed (workers assigned to the chloroprene polymerization area of the plant as of
July 2010
4-20 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 specificed 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 1200 ppm. The two highest
concentrations (910 and 1200 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 OSHA PEL of 25 ppm for an 8 hour workday. Only one fixed location area air sample
(excluding thoses taken at the drainage trenches) exceeded the OSHA PEL (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 OSHA PEL 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 medical
examination portion of the study, 37 workers were interviewed and demographic and occupational
July 2010
4-21 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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 be 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
-3
from 1-7 mg/m (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. C hronic Xo/tc(//tci'r Kffects
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
alkaline phosphatase were attributed to two workers with abnormally high alkaline phosphatase levels
July 2010
4-22 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
due to bone injury and blood pressue 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 3-fold
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 culture
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/m (0.83-1.93
"3
ppm) chloroprene for 1-4 years; and (2) 8 women aged 19-50 and exposed to 1-4 mg/m (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).
July 2010
4-23 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Table 4-12. Frequency of chromosomal aberrations in lymphocyte culture cells
from chloroprene production workers
CHLOROPRENE
EXPOSURE
# EXAMINED
YEARS
EXPOSED
AGE
RANGE
# CELLS
ANALYZED
PERCENT
ABERRANT (+/-)
PERCENT TYPE
ABERRANT
Chromatid
Chromosome
Chloroprene
Workers
18
....
....
1,666
4.77 (0.57)a
74.4
25.6
Control
9
....
....
572
0.65 (0.56)
100
0
1-4 mg/m3
8
1-20
19-50
648
2.5 (0.49)b
....
....
3-7 mg/m3
20
1-4
19-23
1,748
3.49 (0.51 )a
....
....
Population
Control
181
....
....
28,386
1.19(0.06)
50.3
49.7
a p < 0.001. All values means ± SE
bp < 0.05
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 BD IV rats at a single dose (100 mg/kg
body weight) on gestational day (GD) 17. 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
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
July 2010
4-24 DRAFT - DO NOT CITE OR QUOTE
-------
1 reticulosarcoma, forestomach papilloma, sebaceous basal cell carcinoma) treated weekly with
2 chloroprene were not seen in the vehicle control group. Subcutaneous fibromas were more numerous
3 in chloroprene-treated male rats than in controls. Mammary and ovarian tumors were slightly elevated
4 in chloroprene-treated female rats than in controls.
July 2010
4-25 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-13. Tumor incidence in female BD IV rats treated orally with chloroprene
(100 mg/kg) on GD17 and in their progeny treated (50 mg/kg) weekly for life (120
weeks)
GROUP
NUMBER3
TUMOR BEARING
RATS
NUMBER OF
TUMORS
ANIMALS WITH
MORE THAN ONE
TUMOR
n
%
Total
Per rat
n
%
Treated females
16
9
56.2
14
0.9
5
31.3
Treated progeny
Males
54
15
27.8
18
0.3
3
5.6
Females
62
33
53.2
37
0.6
4
6.5
Control females
14
5
35.7
7
0.5
2
14.3
Control progeny
Males
49
16
32.7
16
0.3
...
...
Females
47
24
51.1
29
0.6
5
10.6
aSurvivors at the time the first tumors were observed.
Source: Ponomarkov and Tomatis (1980, 075453)
Table 4-14. Distribution of tumors in female BD IV rats treated orally with chloroprene
(100 mg/kg) on GD17 and their progeny treated (50mg/kg) weekly for life (120 weeks)
GROUP
ORAL
CAVITY
MAMMARY
OVARY
THYROID
SOFT
TISSUE
PITUITARY
OTHER
n
%
n
%
n
%
n
%
n
%
n
%
n
%
Treated females
1
6.3
6
37.5
2
12.5
—
...
...
...
1
6.3
4a
25.0
Treated progeny
Males
...
...
...
...
...
...
1
1.9
7
13.0
2
3.7
8b
14.8
Females
...
...
25
40.3
9
14.5
1
1.6
...
...
2
3.2
...
...
Control females
1
7.1
4
28.6
...
...
...
...
1
7.1
...
...
lc
7.1
Control progeny
Males
2
4.1
...
...
...
...
...
...
4
8.2
2
4.1
8d
16.3
Females
1
2.1
22
46.8
3
6.4
...
...
...
...
1
...
3e
6.4
a 1 each: uterine squamous cell carcinoma; lung reticulosarcoma; forestomach papilloma; sebaceous basal cell carcinoma.
b leach: 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.
d2 lymphomas; 1 each: lung epidermoid carcinoma; spleen hemangioma; osteosarcoma; mediastinal sarcoma; meningioma;
adrenal cortical adenoma.
e 1 each: stomach fibrosarcoma; lymphoma; uterine adenoma.
Source: Ponomarkov and Tomatis (1980, 075453)
4.2.2. Inhalation Exposure
1 The NTP conducted 16-day, 13-week, and 2-year inhalation exposure studies with chloroprene
2 in F344/N rats and B6C3F1 mice (NTP, 1998, 042076). Results of the 13-week study were reported by
July 2010 4-26 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
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
actual exposure chambers. Histopathology was performed by a study pathologist and reviewed by a
quality assurance pathologist and the Pathology Working Group.
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.
July 2010
4-27 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-15. Survival and body weights of rats in the 16-day inhalation study of
chloroprene
SEX
EXPOSURE (ppm)
SURVIVAL
MEAN BODY WEIGHT (g)
Initial
Final
Change
Male
0
7/10
115 ± 4
139 ± 5
+ 20 ±2
32
10/10
113 ± 4
134 ± 6
+ 20 ±2
80
10/10
118 ± 5
136 ±5
+ 18± 1
200
9/10
114 ± 4
127 ±5
+ 11 ±2**
500
1/10
114 ± 4
104
43
Female
0
9/10
100 ±2
110 ± 3
+ 9 ± 1
32
9/10
100 ±2
109 ±3
+ 8 ± 1
80
9/10
103 ±2
112 ± 2
+ 9 ± 1
200
3/10
101 ±2
101 ±4
+ 4 ± 1**
500
7/10
102 ±2
103 ±3
- 1 ± 1**
3 No standard error calculated due to high mortality
" Significantly different (p <_0.01) from the chamber control group by Williams' or Dunnett's test
Source: NTP (1998, 042076)
1 Minimal to mild olfactory epithelial degeneration was significantly increased in all exposed
2 groups of males and females compared to those in the chamber control groups (Table 4-16). Mild to
3 moderate centrilobular hepatocellular necrosis was observed in male and female rats exposed to 200 or
4 500 ppm. Hematological and clinical chemistry parameters indicated increased serum alanine
5 aminotransaminase (ALT), glutamine dehydrogenase (GDH), and sorbitol dehydrogenase (SDH)
6 activities, as well as anemia and thrombocytopenia (decreased platelet count) in the 200 (female) and
7 500 (male and female) ppm groups, on day 4 only. In females, significant increases in kidney weights
8 (right kidney only) were seen at 80 and 500 ppm, and significantly increased liver weights were seen at
9 200 and 500 ppm.
10
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
10/10
10/10
10/10
10/10
10/10
Degeneration, olfactory epithelium
1/10
(1.0)b
10/10**
(1.0)
10/10**
(1.1)
10/10**
(1.9)
10/10**
(3.8)
Metaplasia, squamous, olfactory
epithelium
0/10
0/10
0/10
1/10
(2.0)
4/10*
(1.8)
Metaplasia, respiratory, olfactory
epithelium
0/10
2/10
(1.0)
5/10*
(1.0)
6/10*
(1.0)
1/10
(2.0)
Metaplasia, squamous, respiratory
epithelium
1/10
(1.0)
1/10
(1.0)
0/10
0/10
7/10
(1.7)
Liver3
10/10
1/10
10/10
10/10
10/10
July 2010 4-28 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
CONTROL
32 ppm
80 ppm
200 ppm
500 ppm
Necrosis, centrilobular
0/10
0/10
0/10
1/10
(2.0)
9/10**
(3.4)
Inflammation, chronic
0/10
0/10
0/10
0/10
1/10
Female
Nose3
10/10
10/10
10/10
10/10
10/10
Degeneration, olfactory epithelium
0/10
9/10**
(1.2)
10/10**
(1.6)
10/10**
(3.4)
10/10**
(3.3)
Metaplasia, squamous, olfactory
epithelium
0/10
1/10
(1.0)
1/10
(1.01)
4/10*
(1.0)
0/10
Metaplasia, respiratory, olfactory
epithelium
0/10
7/10**
(1.0)
8/10**
(1.2)
3/10
(1.0)
7/10**
(1.4)
Metaplasia, squamous, respiratory
epithelium
1/10
(2.0)
1/10
(1.0)
0/10
0/10
4/10
(1.3)
Liver3
10/10
3/10
10/10
10/10
10/10
Necrosis, centrilobular
0/10
0/10
0/10
7/10**
(2.6)
3/10
(2.0)
Inflammation, chronic
0/10
0/10
0/10
2/10
(1.0)
5/10*
(1.0)
a number of animals with tissue examined microscopically.
b Average severity grade of lesions in affected rats: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked
p<°.°5.
" Significantly different (p <_0.01) from the chamber control group by the Fisher's exact test.
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
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.
July 2010
4-29 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
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
10/10
24.7 ±0.5
27.0 ±0.5
+ 2.3 ±0.1
12
10/10
24.8 ±0.5
27.1 ±0.6
+ 2.3 ±0.3
32
10/10
25.3 ±0.3
26.5 ±0.3
+ 1.2 ± 0.3**
80
10/10
24.8 ±0.5
26.1 ±0.6
+ 1.3 ±0.2**
200
0/10
24.2 ±0.4
...
...
Female
0
10/10
19.5 ±0.7
22.6 ±0.5
+ 2.3 ±0.3
12
10/10
20.4 ±0.8
23.1 ±0.4
+ 2.6±0.3
32
10/10
19.9 ±1.0
22.1 ±0.2
+ 1.8 + 0.3
80
10/10
20.1 ±0.8
22.5 ±0.3
+ 2.7 + 0.3
200
0/10
20.0 ±0.6
...
...
"Significantly different (p < 0.01) from the chamber control group by Williams' or Dunnett's test.
Source: NTP (1998, 042076)
A range-finding 13-week inhalation study was conducted by NTP (1998, 042076) (reported by
Melnick et al (1996, 625207)), 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 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.
July 2010
4-30 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
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
10/10
109 ±4
311 ± 9
+ 202 ± 8
5
10/10
119 ± 2*
323 ± 11
+ 204± 10
12
10/10
116 ± 1
306 ±9
+ 190 ±8
32
10/10
117 ± 2
327 ± 11
+ 209± 10
80
10/10
116 ± 1
301 ±8
+ 184 ±7
200
9/10
116 ± 3
304 ±8
+ 185 ±7
Female
0
10/10
102 ±2
191 ±4
+ 89 ±3
5
10/10
101 ± 1
193 ±4
+ 92 ±3
12
10/10
102 ±2
199 ±5
+ 97 ±4
32
10/10
101 ±2
195 ±4
+ 94 ±4
80
10/10
103 ± 1
192 ±3
+ 90 ±3
200
10/10
102 ± 1
183 ±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
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
July 2010
4-31 DRAFT - DO NOT CITE OR QUOTE
-------
1 animals. Total activity was increased in male rats in the 32 and 200 ppm groups. There were no
2 exposure-related effects on motor activity, forelimb/hind-limb grip strength, or startle response.
3 Increased incidences of minimal to mild olfactory epithelial degeneration and respiratory
4 metaplasia occurred in male and female rats exposed to 80 or 200 ppm (Table 4-19). The incidence of
5 olfactory epithelial degeneration in females exposed to 32 ppm was significantly greater than in the
6 chamber control group. No effects were observed in the respiratory epithelium of exposed rats. In
7 female rats exposed to 200 ppm, the incidence of hepatocellular necrosis was significantly greater than
8 in the chamber control group. Variably sized aggregates of yellow or brown material consistent with
9 hemosiderin appeared in small vessels or lymphatics in or near portal triads or in Kupffer cells of male
10 and female rats exposed to 200 ppm and were significantly increased compared with chamber controls.
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
10/10
0/10
10/10
10/10
10/10
10/10
Degeneration, olfactory
epithelium
0/10
...
0/10
3/10
(1.0)b
10/10**
(1.0)
10/10**
(2.0)
Metaplasia, respiratory,
olfactory epithelium
0/10
...
0/10
0/10
4/10*
(1.3)
4/10*
(1.3)
Liver3
10/10
2/10
1/10
1/10
10/10
10/10
Necrosis, centrilobular
0/10
0/10
0/10
01/10
0/10
3/10
(2.0)
Inflammation, chronic
0/10
1/10
(1.0)
0/10
0/10
1/10
(1.0)
2/10
(1.0)
Hemosiderin pigmentation
0/10
0/10
0/10
0/10
0/10
5/10*
(1.6)
Female
Nose3
10/10
0/10
10/10
10/10
10/10
10/10
Degeneration, olfactory
epithelium
0/10
...
0/10
4/10*
(1.0)
9/10**
(1.9)
10/10**
(1.9)
Metaplasia, respiratory,
olfactory epithelium
0/10
...
0/10
0/10
8/10**
(2.0)
9/10**
(2.0)
Liver3
10/10
2/10
5/10
3/10
10/10
10/10
Necrosis, centrilobular
0/10
0/10
0/10
0/10
0/10
5/10*
(1.0)
Inflammation, chronic
2/10
(2.0)
0/10
1/10
(2.0)
0/10
1/10
(2.0)
8/10*
(1.3)
Hemosiderin pigmentation
3/10
(1.0)
0/10
1/10
(3.0)
0/10
0/10
9/10**
(1.7)
3Number of animals with tissue examined microscopically.
b Average severity grade of lesions in affected rats: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked
* Significantly different ( p < 0.05) the chamber control group by Fisher's exact test.
" Significantly different (p < 0.01) from the chamber control group by Fisher's exact test.
Source: NTP (1998, 042076)
July 2010
4-32 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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.
Table 4-20. Survival and body weights of mice in the 13-week inhalation study of
chloroprene
SEX
EXPOSURE (ppm)
SURVIVAL
MEAN BODY WEIGHT (g)
Initial
Final
Change (+)
Male
0
10/10
25.5 ±0.4
35.9 ±0.9
10.5 ±0.7
5
10/10
25.2 ±0.3
35.1 ±0.9
10.0 ±0.7
12
10/10
25.2 ±0.2
34.9 ±0.6
9.7 ±0.6
32
10/10
25.4 ±0.2
36.0 ±0.9
10.6 ±0.9
80
10/10
24.7 ±0.3
32.7 ±0.6*
7.9 ±0.5*
Female
0
10/10
20.4 ±0.2
30.3 ± 1.0
9.9 ±0.9
5
10/10
20.9 ±0.3
32.2 ±0.9
11.3 ± 0.9
12
10/10
20.4 ±0.3
30.1 ±0.6
9.7 ±0.6
32
10/10
20.8 ±0.2
32.6 ±0.8
11.8 ± 0.7
80
10/10
20.5 ±0.2
30.2 ±1.3
9.7 ±1.2
* Significantly different (p < 0.05) from the chamber control group by Williams' or Dunnett's test
Source: NTP (1998, 042076)
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 lead
to direct gastrointestinal exposure to chloroprene.
July 2010 4-33 DRAFT - DO NOT CITE OR QUOTE
-------
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
10/10
3/10
0/10
10/10
10/10
Squamous epithelial hyperplasia
0/10
0/10
...
0/10
4/10*
(1.5)a
Female
Number examined microscopically
10/10
0/10
0/10
10/10
10/10
Squamous epithelial hyperplasia
0/10
...
...
0/10
9/10**
(1.9)
a Average severity grade of lesions in affected mice: 1 = minimal, 2 = mild, 3 = moderate, 4 = marked
* Significantly different (p < 0.05) from the chamber control group by Fisher's exact test.
** p < 0.01.
Source: NTP (1998, 042076)
1 In the 2-year (NTP, 1998, 042076) inhalation study of chloroprene in male and female rats,
2 groups were exposed to target concentrations of 0, 12.8, 32, and 80 ppm chloroprene. The actual
3 chamber concentrations animals were exposed to were 0, 12.8 ± 0.4, 31.7 ± 1.1, and 79.6 ±1.6 and 0,
4 12.7 ± 0.4, 31.9 ± 0.9, and 79.7 ±1.7 ppm chloroprene for rats and mice, respectively. The high-
5 exposure concentration was chosen based on the observation of anemia and hepatocellular necrosis in
6 rats exposed to 200 ppm for 13 weeks. The range of exposures selected included the NOAEL for
7 degenerative olfactory epithelial lesions in the 13 week study. Estimates of 2-year survival
8 probabilities are shown in Table 4-22. Survival of males exposed to 32 or 80 ppm was significantly
9 less than that of the chamber control group.
July 2010
4-34 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Table 4-22. 2-Year survival probability estimates for F344/N rats chronically
exposed (2 years) to chloroprene by inhalation
SEX
STATUS
CONTROL
12.8 ppm
32 ppm
80 ppm
Animals initially in study
50
50
50
50
Moribund
34
40
41
41
Natural deaths
3
1
4
5
Male
Animals surviving to study termination
13
9
5
4
Percent probability of survival at end of
study
26
18
10
8
Mean survival (days)
646
638
609
609
Survival analysis3
p = 0.013
p = 0.615
p = 0.025
p = 0.025
Animals initially in study
50
50
50
50
Moribund
19
21
23
27
Natural deaths
1
1
1
2
-------
1
2
3
4
5
6
7
8
9
10
11
12
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 non-neoplastic lesions in F344/N rats
chronically exposed (2 years) to chloroprene by inhalation
LESION INCIDENCE (SEVERITY)
TISSUE SITE/LESION TYPE
Males (ppm)
Females (ppm)
0
12.8
32
80
0
12.8
32
80
Oral cavity
0/50
0/50
0/50
3/50
Sqamous Cell Hyperplasia
(2.7 f
Thyroid gland
0/50
2/50
4/4 9b
1/50
0/49
0/50
o/so
2/50
Follicular Cell Hyperplasia
(2.0)
(1.8)
(1.0)
\J/ W
(2.5)
Lung
5/50
16/50°
14/49b
25/50°
6/49
22/50°
22/50°
34/50°
Alveolar Hyperplasia
(1.4)
(1.4)
(1.9)
(1.4)
(1.8)
1.4)
(1.5)
(1.3)
Kidney (renal tubules)
14/50
20/50
28/50°
34/50°
6/49
6/50
11/50
21/50°
Hyperplasia
(2.0)
(2.6)
(2.1)
(2.9)
(1.3)
(1.8)
(2.1)
(2.0)
Olfactory
Atrophy
3/50
12/50b
46/49°
48/49°
0/49
1/50
40/50°
50/50°
(1.7)
(1.8)
(2.2)
(3.6)
(1.0)
(1.3)
(2.9)
Basal Cell Hyperplasia
0/50
0/50
38/49°
46/49°
0/49
0/50
17/50°
49/50°
(1.6)
(2.2)
(1.1)
(2.3)
Metaplasia
6/50
5/50
45/49°
48/49°
0/49
1/50
35/50°
50/50°
1.7
(1.0)
(1.8)
(3.1)
(1.0)
(1.0)
(2.7)
Necrosis
0/50
1 l/50b
26/49°
19/49°
0/49
0/50
8/50°
12/50°
(2.0)
(2.0)
(2.2)
(2.0)
(1.3)
Chronic Inflammation
0/50
5/50°
9/49°
49/49°
0/49
0/50
2/50
33/50°
(1.0)
(1.6)
(2.7)
(1.0)
(2.0)
a Severity of lesions graded as: 1= minimal, 2 = mild, 3 = moderate, 4 = marked, average severity reported in
parenthesis
bp < 0.05, ps 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
°p<0.01
Source: NTP (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
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.
July 2010 4-36 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Table 4-24. Incidence of neoplasms in F344/N rats chronically exposed (2 years) to
chloroprene by inhalation
TISSUE SITE/TUMOR TYPE
TUMOR INCIDENCE
Males (ppm)
Females (ppm)
0
12.8
32
80
0
12.8
32
80
Oral cavity
Papillomas or carcinomas
0/50
2/50
5/50a
12/50b
1/49
3/50
5/50
ll/50b
Thyroid gland
Adenomas or carcinomas
0/50
2/50
4/49a
5/50a
1/49
1/50
1/50
5/50
Lung
Adenomas or carcinomas0
2/50
2/50
4/49
6/50
1/49
0/50
0/50
3/50
Kidney (renal tubules)
Adenomas or carcinomas (extended and
standard evaluations combined)
1/50
8/50a
6/5 0b
8/5 0b
0/49
0/50
0/50
4/50
Mammary gland
Fibroadenomas
...
...
...
...
24/49
32/50
36/50a
36/50b
ap < 0.05, ps 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
bp<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 males 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/bronchiolar
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 males and females 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 2 times the normal diameter and were
lined by increased numbers of tubule epithelial cells that partially or totally filled the tubule 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.
July 2010
4-37 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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 females 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.
In the NTP 2-year mouse study, exposure concentrations were 0, 12.8, 32, and 80 ppm. 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 animals were observed twice daily and body weights were recorded
initially, weekly through week 12, 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 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.
July 2010
4-38 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Table 4-25. 2-Year survival probabilities for B6C3F1 mice chronically exposed (2
years) to chloroprene by inhalation
SEX
STATUS
CONTROL
12.8 ppm
32 ppm
80 ppm
Animals initially in study
50
50
50
50
Moribund
15
16
26
34
Natural deaths
3
7
10
3
Animals surviving to study termination
27
27
14
13
Percent probability of survival at end of
study
54
54
28
26
-------
Table 4-26. Incidence and severity of non-neoplastic lesions in B6C3F1 mice
chronically exposed (2 years) to chloroprene by inhalation
LESION INCIDENCE (SEVERITY)
TISSUE SITE/LESION TYPE
Males (ppm)
Females (ppm)
o
12.8
32
80
o
12.8
32
80
Lung
Bronchiolar Hyperplasia
0/50
10/50°
18/50°
23/50°
0/50
15/49°
12/50°
30/50°
(2.0)
(1.7)
(2.2)
(2.0)
(2.2)
(2.2)
Histiocytic Cell Infiltration
7/50
8/50
11/50
22/50°
1/50
14/49°
18/50°
23/50°
(1.6)
(3.3)
(2.5)
(2.9)
(3.0)
(2.0)
(2.3)
(2.4)
Kidney (renal tubule)
Hyperplasia
2/50
16/49°
(1.4)
17/50°
(1.6)
18/50°
(1.6)
--
--
--
--
Mammary Gland
Hyperplasia
—
—
—
—
0/49
1/49
(1.0)
1/50
(1.0)
3/50
(2.0)
Forestomach
Epithelial Hyperplasia
4/50
6/48
7/49
29/50°
4/50
3/49
8/49
27/50°
(3.0)
(1.8)
(2.3)
(2.2)
(2.0)
(3.7)
(1.6)
(2.7)
Olfactory
Suppurative Inflammation
2/50
1/48
4/50
6/50
0/50
1/49
3/49b
4/50°
(2.0)
(1.0)
(1.0)
(1.5)
(1.0)
(1.7)
(1.5)
Atrophy
7/50
8/48
7/50
49/50°
6/50
5/49
4/49
47/50°
(1.1)
(1.4)
(1.1)
(2.5)
(1.2)
(1.2)
(1.3)
(2.0)
Metaplasia
6/50
5/50
5/50
49/50°
2/50
3/49
1/49
44/50°
(1.0)
(1.4)
(1.0)
(2.5)
(1.0)
91.0)
92.0)
(2.0)
Spleen
Hematopoietic Proliferation
26/50
22/49
35/50d
31/50d
13/50
25/49d
42/49d
39/50d
a 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
p < 0.05, ps 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
°p<0.01
d Significantly increased relative to controls, level of significance not reported
Source: NTP (1998, 042076)
1 The incidences of olfactory epithelial atrophy, adenomatous hyperplasia, and metaplasia in
2 males and females exposed to 80 ppm were significantly increased compared to those in the chamber
3 controls. The incidence of suppurative inflammation in females exposed to 32 and 80 ppm was
4 significantly greater than controls. Atrophy and metaplasia of the olfactory epithelium was similar to
5 lesions observed in rats exposed to chloroprene. Adenomas of the respiratory epithelium were present
6 in one female exposed to 32 ppm and one male exposed to 80 ppm.
7 In male mice, a pattern of nonneoplastic liver lesions along with silver-staining helical
8 organisms within the liver was observed, consistent with Helicobacter hepaticus infection.
9 Polymerase chain reaction-restriction fragment length polymorphism based assay confirmed an
10 organism compatible with H. hepaticus. Historically, NTP studies with H. hepaticus associated
July 2010 4-40 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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 B6C3F1 mice chronically exposed (2 years)
to chloroprene by inhalation
TISSUE SITE/TUMOR TYPE
TUMOR INCIDENCE
Males (ppm)
Females (ppm)
0
12.8
32
80
0
12.8
32
80
Lung
Adenomas or carcinomas
13/50
28/50°
36/50°
43/50°
4/50
28/49°
34/50°
42/50°
All Organs
Hemangiomas or hemangiosarcomas
3/50
14/50b
23/50°
21/50°
4/50
6/50
18/5 0b
8/50
Harderian gland
Adenomas or carcinomas
2/50
5/50
10/503
12/50b
2/50
5/50
3/50
9/50a
Kidney (renal tubules)
Adenomas or carcinomas (extended and
standard evaluations combined)
0/50
2/49
3/50a
9/5 0b
...
...
...
...
Mammary gland
Carcinomas
...
...
—
...
3/50
4/50
7/50
12/503
Forestomach
Papillomas or carcinomas
1/50
0/50
2/50
5/50
1/50
0/50
0/50
4/50
Liver
Adenomas or carcinomas
...
...
...
...
20/50
26/49
20/503
30/50°
Skin
Sarcoma
...
...
...
...
0/50
1 l/50b
11/50°
18/50°
Mesentery
Sarcomas
0/50
4/50
8/5 0b
3/50
Zymbal's gland
Carcinomas
0/50
0/50
0/50
3/50
ap < 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
bp<0.01
cp< 0.001
Source: NTP (1998, 042076)
The incidences of Harderian gland adenoma and 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.
July 2010 4-41 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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 one male each exposed to 12.8 ppm
or 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 by using the poly-3 survival-adjusted quantal response
method of Portier and Bailer (1989, 093236). This adjustment accounts for the effects of early
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.
July 2010
4-42 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Table 4-28. Survival-adjusted" neoplasm rates for mice in the 2-year inhalation
study of chloroprene
TISSUE SITE/TUMOR TYPE
MALES (%)
FEMALES (%)
0
12.8
32
80
0
12.8
32
80
Lung
Adenoma or carcinoma
Alveolar/bronchiolar adenoma or
carcinoma
14 J**
29.8**
28.3
63.7**
56.9**
79.2**
66.4**
92.9**
4 6**
9
35.6**
68.3**
53.8**
85.8**
76.0**
96.1**
All Organs
Hemangioma or hemangiosarcoma
2 4**
28.2**
45.2**
43.6**
9.0*
16.0
53.1**
27.7*
Harderian gland
Adenoma or carcinoma
4 7**
12.0
26.3**
32.0**
4.5**
13.5
11.7
31.2**
Kidney (renal tubules)
Adenoma
(single section)
(single + step section)
0*
o**
2.4
4.8
2.8
8.3
8.2
23.9*
...
...
...
...
Mammary gland
Adenoacanthoma or carcinoma
6.7**
12.9
33.7**
42.5**
Forestomach
Squamous cell papilloma or carcinoma
2 4**
0
5.6
13.3
2 3**
0
0
14.6
Liver
Carcinoma
Adenoma or carcinoma
...
...
...
...
9 o**
44.8
28.4*
62.9
47.5**
63.3
58.2**
79 7**
Skin
Sarcoma
0*
27.5**
39.0**
52.6**
Mesentery
Sarcoma
0
10.7*
28.9**
11.0
a Survival-adjusted neoplasm rates were estimated using the Poly-3 survival-adjusted quantal response method of Portier
and Bailer (1989)
In the chamber control column, * indicates a statistically significant trend (p< 0.05) across all exposure groups by the Poly-
3 quantal response test; ** indicates a statistically significant trend at p < 0.01.
In the exposed group columns, * indicates a statistically significant difference (p < 0.05) from the chamber control group
by pairwise comparison; ** indicates a statistically significant difference at p < 0.01.
Source: NTP (1998, 042076)
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 under nitrogen at -
20°C. Phenothiazine (0.01%) was added to prevent oxidation. Afresh 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 were 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
July 2010
4-43 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
and clinical signs and mortality were recorded. Rats and hamsters were weighed immediately before
the first exposure, weekly for the first 8 the 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 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
July 2010
4-44 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
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)
NUMBER1
BW (g)
ADRENALS
BRAIN
KIDNEYS
LIVER
LUNGS
SPLEEN
THYROID
Rats
Males
0
77
494
...
...
0.61
3.09
0.45
0.154
0.0056
10
9
500
...
...
0.68a
3.31
0.37b
0.172
0.0056
50
76
496
...
...
0.64
3.15
0.38b
0.146
0.0056
Females
0
81
308
...
...
0.64
3.00
0.53
0.180
0.0080
10
19
309
...
...
0.65
3.23a
0.45a
0.176
0.0073
50
75
307
...
...
NR2
3.13a
0.45a
0.164b
0.0070b
Hamsters
Males
0
86
101
0.0311
1.10
1.25
5.11
0.85
0.197
...
10
92
101
0.02793
1.11
1.17b
4.75a
0.84
0.190
...
50
92
93
0.0294
1.19°
1.22
4.91
0.90b
0.174b
...
Females
0
60
99
0.0340
1.13
1.48
6.73
1.01
0.253
...
10
74
98
0.0356
1.16
1.50
6.54
0.97
0.269
...
50
72
90
0.0383
1.24°
1.50
6.37
1.01
0.286
...
1 Number at sacrifice
2 Not recorded
a Significant, 0.1 < p < 0.005
b Significant, 0.001 < p < 0.01
0 Significant, p < 0.001
Source : Trochimowicz et al. (1998, 625008)
Gross pathology revealed that lungs from rats exposed at 10 and 50 ppm had markedly lower
incidences of nodular pleural 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
July 2010
4-45 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 < .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
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/m (9.7 ppm).
"3
Several "signs of systemic effect" in male rats were reported at 1.69 ± 0.087 mg/m , including an
July 2010 4-46 DRAFT - DO NOT CITE OR QUOTE
-------
1 increase in a "summation threshold index" (not defined) after 2.5 and 4.5 months, a decrease in the
2 synthesis of hippuric acid from sodium benzoate (Quick's test) at 4.5 months, and an inhibition of gas
3 exchange after 4.5 months. Chloroprene was reported to have had no effect on "the indicators used in
4 the tests" (i.e., summation threshold index, hippuric acid synthesis, and inhibition of gas exchange) in
"3
5 mice at concentrations as high as 35 ± 0.7 mg/m (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
MALES
FEMALES
0 ppm | 10 ppm | 50 ppm
0 ppm | 10 ppm | 50 ppm
Initial number of rats
100
100
100
100
100
100
Number examined
97
13
100
99
24
100
Number tumor-bearingb
51
6
57
66
12
74
Total number primary tumorsb
73/51
6/6
77/57
100/66
13/12
96/71
Hematopoietic system
Lymphoid leukemia
1
0
2
0
0
1
Monocytic leukemia
0
0
1
0
0
0
Kidneys
Lipoma
0
0
1
1
0
1
Adenocarcinoma
0
0
1
0
0
0
Liver
Unidentified
0
0
0
1
0
0
Lungs
Anaplastic carcinoma
0
0
0
1
0
0
Mammary glands
Adenoma
...
...
...
3
1
7
Fibroadenoma
...
...
...
24
6
36
Adenocarcinoma
...
...
...
5
0
3
Papillary carcinoma
...
...
...
1
0
0
Unidentified tumor
...
...
...
1
2
0
Skin
Squamous cell carcinoma
0
0
2
0
0
0
Skin, nasal cavity, maxillary sinus,
Squamous cell carcinoma
0
0
3
1
0
0
Spleen
Hemangio sarcoma
0
0
1
0
0
0
Subcutis, nasal cavity, or maxillary sinus
Reticulum cell sarcoma
0
0
0
0
0
1
Testes
Leydig cell tumor
2
2
4
...
...
...
Testes/epididymides
Mesothelioma
1
0
0
...
...
...
July 2010
4-47 DRAFT - DO NOT CITE OR QUOTE
-------
MALES
FEMALES
Thyroid gland
Parafollicular cell adenoma
Small
6
0
8
11
0
14
Medium/large
3
1
3
3
1
4
Parafollicular cell carcinoma
Small
1
0
0
0
0
0
Large
1
0
0
0
0
0
Follicular adenoma
Small
2
0
2
0
0
3
Large
2
0
1
0
0
0
Papillary carcinoma
0
0
0
0
0
2
Urinary bladder
Transitional cell carcinoma (metastasizing)
0
0
1
0
0
0
Zymbal's gland
Adenoma
0
0
0
0
0
1
a Multiple tumors at one site were counted as one tumor
b Some animals had more than one tumor
Source: 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
MALES
FEMALES
0 ppm
10 ppm
50 ppm
0 ppm
10 ppm
50 ppm
Initial number of hamsters
100
100
100
100
100
100
Number examined
100
97
97
94
93
97
Number tumor bearing3
14
17
20
10
11
15
Total number primary tumors3
15/14
18/17
23/20
11/11
11/11
18/15
Kidney
Cortical adenocarcinoma
2
0
0
0
0
0
Liver
Neoplastic (hepatocellular) nodule
0
1
0
0
0
0
Unidentified tumor-like lesion
0
1
0
0
0
1
Lung tumors
0
0
0
0
0
0
Gallbladder
Cystadenomatous polyp
1
6a
1
1
2
3
Pancreas
Islet-cell adenoma
1
0
2
0
0
0
Islet-cell adenocarcinoma
0
0
0
1
0
1
Stomach
Papilloma
0
0
2
0
0
0
Unidentified papilloma-like lesion
1
1
1
1
2
0
Testes
Leydig-cell tumor
1
0
0
...
...
...
Colon
Adenomatous polyp
0
0
0
2
0
0
July 2010
4-48 DRAFT - DO NOT CITE OR QUOTE
-------
SITE AND TYPE OF TUMOR3
MALES
FEMALES
0 ppm 1 10 ppm 1 50 ppm
0 ppm 1 10 ppm 1 50 ppm
Pituitary
Adenoma
0
0
1
2
0
0
Thyroid gland
Parafollicular cell adenoma
2
0
0
0
2
1
Cystadenoma
1
0
0
0
0
0
Papillary adenoma
0
1
1
1
0
2
Follicular adenoma
2
1
0
1
2
2
Parathyroid
Adenoma
0
0
0
0
1
0
Adrenals
Cortical adenoma
4
1
10
0
0
3
Cortical carcinoma
0
1
0
1
0
1
Pheochromocytoma
0
4b
2
0
0
0
Malignant pheochromocytoma
0
0
2
0
0
0
Ovaries
Granulosa-theca-cell tumor
...
...
...
0
2
1
Parotid salivary glands
Adenoma
0
0
0
0
0
1
Skin
Unidentified tumor-like lesion
0
1
0
0
0
0
Zymbal's gland
Sebaceous adenoma
0
0
1
0
0
0
Depot fat
Lipoma
0
0
0
0
0
1
Nose
Adenoma of Bowman's glands
0
0
0
1
0
0
Adenocarcinoma of Bowman's glands
0
0
0
0
0
1
Bone (ribs)
Osteosarcoma
0
0
0
1
0
0
Abdominal cavity
Reticulum cell sarcoma
1
0
0
0
0
0
a Some animals had more than one tumor
b Significant, p < 0.05 by chi-squared test
Source: Trochimowicz et al. (1998, 625008)
1 Dong et al. (1989, 007520) exposed Kumming albino mice (weaned at 2 weeks age) to 0, 2.9 ±
"3
2 0.3, 19.2 ± 1.9, or 189 ± 13.3 mg/m chloroprene for 4 hours/day, 6 days/week for 7 months. The
3 purity of the chloroprene used to generate the test atmospheres was stated to be 99.8%. Animals were
4 terminated at the end of the exposure period, or when found moribund. Lung tumors were not
5 observed in treated animals before the 6th month of exposure, and were observed to increase in
"3
6 incidence with increasing concentration. The LOAEL for this study was determined to be 2.9 mg/m
7 (8.1% incidence of lung tumors vs. 1.3% in control animals, p < 0.05). Most lung tumors observed
8 were papilloadenomas. Induction of multiple tumors in a single animal was also observed to increase
9 with increasing dose.
July 2010
4-49 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
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
(GD) 17. 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 (see 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.
July 2010
4-50 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Table 4-32. Summary of epididymal spermatozoal and estrous cycle parameters
for rats and mice in the 13-week study of chloroprene
RATS
MICE
0 ppm
5 ppm
32 ppm
200 ppm
0 ppm
12 ppm
32 ppm
80 ppm
n
10
10
9
9
7
8
10
10
Epididymal spermatozoa - males"
Motility (%)
86.73
± 1.04
83.62
± 1.93
82.16
± 1.84
80.04
± 1.99**
79.09
± 1.20
81.07
± 1.13
80.08
± 1.19
80.04
± 1.47
Abnormal sperm (%)
0.70
±0.05
0.78
±0.11
0.73
±0.11
1.02
±0.14
1.49
±0.42
1.30
±0.22
0.98
±0.10
1.36
±0.22
Sperm concentration
(106/g cauda epididymidis)
698
±40
722
±62
689
±46
683
± 25
1,632
± 138
1,447
± 122
1,575
± 104
1,672
± 134
Estrous cycle - females"
Length (days)
5.00
±0.15
4.67
± 0.17b
5.00
± 0.27°
5.33
± 0.17b
4.00
±0.00
4.30
±0.21
4.22
±0.15b
4.13
± .13°
Diestrus stage (% of cycle)
42.9
35.7
44.3
45.7
31.4
31.4
30.0
35.7
Proestrus stage (% of cycle)
15.7
18.6
11.4
17.1
20.0
20.0
22.9
25.7
Estrus stage (% of cycle)
18.6
22.9
20.0
15.7
24.3
24.3
25.7
20.0
Metestrus stage (% of cycle)
22.9
22.9
24.3
20.0
24.3
24.3
21.4
18.6
Uncertain diagnosis stage (%
of cycle)
0.0
0.0
0.0
1.4
....
....
....
....
a Epididymal spermatozoal parameters, and estrous cycle lengths are presented as mean ± standard error.
b Estrous cycle was longer than 12 days or unclear in 1 of 10 animals.
0 Estrous cycle was longer than 12 days or unclear in 2 of 10 animals.
** Significantly different (p < 0.01) from the control group by Shirley's test.
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
"3
developmental parameters. In male rats exposed for 4.5 months to 1.7 mg/m (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/m (0.09 ppm) for 2 months and increased dominant lethal
"3
mutations in germ cells of male and female C57BL/6 mice exposed to 3.5 mg/m (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/m (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.
"3
Exposure to 0.13 mg/m (~ 0.04 ppm) chloroprene was reported to have resulted in increased postnatal
July 2010
4-51 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
"3
mortality. Exposure to 4 mg/m (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
-3
at 0, 1, 10, or 25 ppm (0.28, 2.8, or 6.9 mg/m ) for 4 hours daily, either on GDs 1-12 (embryotoxicity
study) or GDs 3-20 (teratology study). Pregnant rats in these embryotoxicity and teratology studies
were sacrificed and their litters examined on GDs 17 and 21, respectively. Male rats in a separate
"3
reproduction study were exposed to 0 or 25 ppm (0 or 6.9 mg/m ) 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.
July 2010
4-52 DRAFT - DO NOT CITE OR QUOTE
-------
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
21
24
21
19
Pregnancy rate, %
84 (21/25)
96 (24/25)
84 (21/25)
76 (19/25)
Corpora lutea/dam
13 ±3
12 ± 2
12 ± 2
13 ± 2
Implantation sites/dam
10 ± 2
9 ± 3
9 ± 2
11 ± 1
Median preimplantation loss, %
14.7
29.5
20.0
10.0
Live fetuses/litter
9 ± 2
8 ± 3
8 ± 3
10 ± 1
Litters with resorption, %
29 (6/21)
29 (7/24)
62(13/21)*
59(11/19)
Litters totally resorbed
0
0
0
0
Median postimplantation loss
in litters with resorption, %
11.8
16.7
22.0
16.7
Resorptions/litters with
Resorptions
1.3 (8/6)
2.0 (14/7)
1.9(25/13)
1.6 (17/11)
Resorptions/litters total
0.38 (8/21)
0.58(14/24)
1.19(25/21)
0.89(17/19)
Fetal body weight, g
3.76 ±0.28
3.94 ±0.46
3.96 ±0.26
4.04 ±0.27**
Fetal crown-rump length, mm
32.9 ± 1.4
33.7 ± 1.6
33.8 ±0.7**
34.1 ± 1.2**
Embryotoxicity Study
Number of litters
45
43
43
48
Pregnancy rate, %
90 (45/50)
86 (43/50)
88 (43/49)
94 (48/51)
Corpora lutea/dam
15 ± 3
14 ± 3
14 ± 2
13 ±3
Implantation sites/dam
11 ±3
11 ±4
10 ± 4
10 ± 3
Median preimplantation loss, %
20.0
16.2
17.7
16.0
Live fetuses/litter
10 ± 3
9 ± 4
10 ± 3
10 ± 3
Litters with resorption, %
51 (23/45)
51 (22/43)
53 (23/43)
50 (24/48)
Litters totally resorbed
0
1
0
0
Median postimplantation loss
in litters with resorption, %
9.1
12.9
8.3
9.1
Resorptions/litters with
resorptions
1.7 (39/23)
2.1 (47/22)
1.6 (37/23)
1.4 (34/24)
Resorptions/litters total
0.87 (39/45)
1.09 (47/43)
0.86 (37/43)
0.71 (34/48)
* Significantly different (p < 0.05) from the control group by Fisher's exact test.
** Significantly different (p < 0.05) from the control group by an analysis of variance and least significant difference (LSD)
test
Source: Culik et al. (1978, 094969)
1 Culik et al. (1978, 094969) concluded that the statistically significant increase in litters with
2 resorptions observed in the teratology study at 10 ppm was not biologically significant because the
3 increase at 25 ppm was not statistically significant and the effect was not observed in the
4 embryotoxicity study, which had larger numbers of animals per exposure group and was specifically
5 designed to observe such an effect. Further, the control group for the teratology study is the only
6 group in either study (embryotoxicity or teratology) that is far outside of the historical control range for
July 2010
4-53 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
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
CONCENTRATION OF CHLOROPRENE (ppm)
0
1
10
25
Number of litters (fetuses) examined
Gross anomalies
21 (192)
24(191)
21 (172)
19(184)
Soft tissue anomalies
21 (66)
24 (69)
21 (60)
19(62)
Skeletal anomalies
21 (126)
24 (122)
21 (112)
19 (122)
Number of litters (fetuses) affected
Gross anomalies
Runts1
1(1)
0
1(1)
1(1)
Small subcutaneous hematomas
5(5)
9(9)
4(4)
6(10)
Petechial hemorrhages
5(5)
2(6)
3(3)
2(2)
Soft tissue anomalies
Hydronephrosis
8(9)
4(6)
1(1)
5(7)
Subcutaneous edema
0
1(1)
0
0
Skeletal anomalies
Delayed ossification of one or more
sternebrae
17 (58)
15 (39)
13 (33)
14 (45)
14th rudimentary ribs(s) or spur(s)
20 (91)
22 (76)
20 (67)
19(77)
Wavy ribs
4(4)
4(5)
2(3)
3(4)
Bipartite thoracic centra
2(2)
2(3)
2(2)
4(8)
1 Body weight less than control mean weight minus 3 standard deviations
Source: Culik et al. (1978, 094969)
Mast et al. (1994, 625206) exposed groups of 15-16 pregnant New Zealand white rabbits by
"3
inhalation to 10, 40, or 175 ppm chloroprene (36.2 144.8, or 633.5 mg/m ) for 6 hours/day on
gestational days 6-28. Maternal body weights were measured on days 0, 6, 15, 22, and 29 and 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
July 2010 4-54 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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
from 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 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 GD6-28 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/m )
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 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 from 10 weeks (6 hours/day, 5 days/week). In both the Fo and Fi rats, the general
condition, behavior, and signs of possible intoxication were checked daily and all signs of illness or
reaction to exposure were recorded. Individual body weights were recorded weekly during exposure.
In the Fi rats, blood samples were collected from 15 rats/sex/exposure group at an age of 4 weeks and
July 2010 4-55 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
analyzed for hemoglobin concentration. At the end of the exposure period, 10 Fi 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 Fi 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 Fi 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). Fi 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 Fi 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 six male albino rats (from
Charles River laboratories) were exposed to chloroprene by the dermal (200 mg/kg), oral (50 mg/kg),
July 2010
4-56 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
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/m ). 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/m ], 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/m ], 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)
MORTALITY (DEAD/TOTAL)
530
0/6
1,690
0/6
2,280
1/6
3,535
2/6
3,610
2/6
Source: Clary et al. (1978, 064942)
Clary et al. (1978) 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 were 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 sixth and seventh 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
relative weights of all organs except for the heart. The relative organ weights for kidneys were
July 2010
4-57 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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 in 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 the assessment and only the relevant immune data are presented here.
NTP (1998, 042076) observed that thymus weights in adult male and female B6C3F1 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
exposure for a longer period (i.e., 13-week exposure) as part of the same NTP (1998, 042076) study.
July 2010
4-58 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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 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 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 immuntoxicity 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 (Munter, et al., 2002, 625215;
Hurst and Ali, 2007, 625159). 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
July 2010
4-59 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
bioassay that may indicate events in chloroprene-induced neoplasia (Ton et al., 2007, 625004; Sills et
al., 1999. 6249521
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 (1-
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 B6C3F1 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 |iM was detected in mouse liver microsomes between 5-10 minutes after initiation of exposure
with 0.05 |iM (100 ppm) chloroprene. Preliminary data also showed that hydrolysis of (1-
chloroethenyl)oxirane was slowest in the liver microsomes of B6C3F1 mice. Further comparing
metabolism between species, Cottrell et al. (2001, 157445) observed that qualitative profiles of
metabolites from liver microsomes obtained from B6C3F1 mice, Sprague-Dawley or F344 rats, and
humans were similar, with (l-chloroethenyl)oxirane being the major metabolite in all species and
genders. 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). Pretreatment 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. (2002, 625215) investigated the reactivity of the chloroprene metabolite
(l-chloroethenyl)oxirane towards DNA nucleosides and calf thymus DNA in vitro. Adducts were
July 2010
4-60 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 (See Figure 1) 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 (1-
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- vs. 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 B6C3F1 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 definitve 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 due
to non-ra.s mutation mechanisms of genotoxicity or carcinogenicity. Alternatively, differences in
July 2010
4-61 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
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-ra.s 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) reported higher frequencies of K- and H-ras mutations (51%) in chloroprene-induced
forestomach tumors in B6C3F1 mice compared to spontaneous tumors (36%>). The A->T transversion
(CAA->CTA) in H-ra.s 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 B6C3F1 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.
July 2010
4-62 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-36. Genotoxicity assays of chloroprene
TEST SYSTEM
CELLS/STRAIN
TESTED
CONCENTRATIONS
RESULTS3
REFERENCE
Bacterial assays
Salmonella
typhimurium
TA100
0.5 to 8% (volume/
volume) in air
+
Bartsch et al. (1979,
010689)
TA100, TA1535
+
Willems (1980. 625049)
TA98
-
Willems (1980. 625049)
TA100, TA1535
10,000-40,000 ppm
+
Willems (1978. 625048)
TA98, TA1537, TA1538
10,000-40,000 ppm
-
Willems (1980. 625049)
TA100, TA1535, TA1537,
TA98
up to 3,333 (ig/plate
_
NTP (1998. 042076)
TA100
0-5 (imol/plate
-
Westphal et al. (1994,
625047)
TA100
0-5 (imol/plateb
+
Westphal et al. (1994,
625047)
TA100, TA1535, TA97A,
TA98
0 to 69 mMc
+
Himmelstein et al. (2001,
019013)
Mammalian cell assays
Micronucleus
Chinese hamster V79
10% (v/v)
-
Drevon and Kuroki (1979,
010680)
Micronucleus
Chinese hamster V79
0.175 mMc
-
Himmelstein et al. (2001,
019013)
In vivo bioassays
Sex-linked recessive
lethal mutation
Drosophila (Canton-S)
-
Foureman et al. (1994,
065173)
Sex-linked recessive
lethal mutation
Drosophila (Berlin-K)
+
Voeel (1979. 000948)
Sister chromatid
exchange: bone
marrow
B6C3F1 mice
12.8, 32, 80 ppm
-
NTP (1998. 042076); Shelbv
(1990. 624906); Tice (1988.
624981; 1988. 064962)
Chromosomal
aberration: bone
marrow
B6C3F1 mice
12.8, 32, 80 ppm
-
NTP (1998. 042076)
Chromosomal
aberration: bone
marrow
C57BL/6 mice
up to 1 ppm
+
Sanotskii (1976. 063885)
Micronucleus:
peripheral blood
B6C3F1 mice
12.8, 32, 80 ppm
-
NTP (1998. 042076)
Micronucleus: bone
marrow
B6C3F1 mice
-
Shelby and Witt (1995,
624921)
aFor 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).
0 Epoxide metabolite (l-chloroethenyl)oxirane tested
July 2010
4-63 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
July 2010
4-64 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
the same whether chloroprene was stored under air or under an inert gas. The authors speculated that
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 |ig/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, (lchloroethenyl)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 phenobarbitone-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 phenobarbitone, 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.
July 2010 4-65 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
A minimum of 500 binucleated cells were scored for micronuclei. Clastogenicity was determined as
the presence of a dose-dependent increase in the frequency of micronucleated cells, with at least one
concentration producing a 3-fold 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 micronucleated 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 dessicator 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, 0651731 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
(Westphal et al., 1994, 625047; Gahlmann, 1993, 625174).
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
July 2010 4-66 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
were evaluated in male mice exposed by inhalation to chloroprene in the NTP (1998, 042076)
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 micronucleated cells in peripheral blood erythrocytes was not affected when
"3
mice were exposed to chloroprene for 13 weeks to 0, 12.8, 32, or 80 ppm (0, 3.5, 8.8, or 22 mg/m )
(NTP, 1998, 042076; MacGregor et al., 1990, 625184).
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/m (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 B6C3F1 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-l,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
July 2010
4-67 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
intermediate that may mediate its carcinogenic effects; however, there is no evidence of diepoxide
formation in the metabolism of chloroprene. The similarities in the sites of tumor induction in rodents
(see 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 (ED 10 = 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
Mice
Rats
Butadiene
Isoprene
Chloroprene
Butadiene
Isoprene
Chloroprene
Lymphatic/
hematopoietic
M, Fa
M
Circulatory
M, F
M
M, F
Lung
M. F
M
M, F
M
Liver
M, F
M
F
Forestomach
M, F
M
M, F
Harderian
gland
M, F
M,F
M, F
Mammary
gland
F
F
F
M, F
F
Brain
M
Thyroid
F
M,F
Pancreas
M
Testis
M
M
Zymbal's
gland
F
F
Kidney
M
M
M
M, F
Oral Cavity
M,F
July 2010
4-68 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
a M = males, F = females
Source: NTP (1998, 042076); Melnick et al. (1994, 625208); Placke et al. (1996, 624891); U.S.
EPA (2002, 052153)
Table 4-38 Quantitative comparison of carcinogenic potency of butadiene and
chloroprene in mice
Site
Males
Females
Butadiene
Chloroprene
Butadiene
Chloroprene
Lung
2.8a
0.9
0.3
0.3
Harderian gland
4.4
12
12
23
Forestomach
120
70
62
79
Liver
10
1.9
Mammary gland
13
12
a ED 10 values (concentration assocated with 10% excess cancer risk) in ppm
Source: 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 BDIV
rats to a single dose (100 mg/kg) of chloroprene on the 17th day of pregnancy and of their progeny to
July 2010
4-69 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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.
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 (NTP, 1998 [also reported by Melnick et al (1999, 000297)11
two chronic inhalation bioassays (NTP (1998, 042076)[also reported by Melnick et al (1999, 000297)1;
Trochimowicz et al (1998, 625008)) and four reproductive developmental studies (Mast et al., 1994,
625206; Culik et al., 1978, 094969; Appelman and Dreef-van der Meulen, 1979, 064938; Sanotskii,
1976, 063885). These studies associate chloroprene inhalation exposure with toxicity in multiple
organ systems, including respiratory tract, kidney, liver, spleen, and forestomach effects.
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.
July 2010
4-70 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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,
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
July 2010
4-71 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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, 0420761 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
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 2 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.
July 2010
4-72 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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 New Zealand white rabbits
exposed to 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 Fi animals. Exposed Fi 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 is severely lacking in many 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
July 2010
4-73 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
similarities between chloroprene and known human carcinogens, butadiene and vinyl chloride (see
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
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. 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 United States, Russia
(Moscow), Armenia, France, China, and Ireland (Marsh et al., 2007, 625187; Marsh et al., 2007,
625188; Colonna and Laydevant, 2001, 625112; Bulbulyan et al., 1998, 625105; Bulbulyan et al.,
1999, 157419; Romazini et al., 1992, 624896; Li et al., 1989, 625181; Leet and Selevan, 1982,
094970; Pell, 1978, 064957). 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 five-fold increased risk) with chloroprene exposure. Four mortality studies
reported SMRs of 339, 240, 242, 571 when compared to external populations (Bulbulyan et al., 1998,
625105; Bulbulyan et al., 1999, 157419; Li et al., 1989, 625181; Leet and Selevan, 1982, 094970).
July 2010
4-74 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Although sample size and statistical power were limited (thus limiting the precision of risk estimates),
Bulbulyan et al. (1999, 157419; 1998, 625105) 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 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., 1999, 157419;
Bulbulyan et al., 1998, 625105). 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
July 2010
4-75 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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, there were many workers who were exposed during time periods when
chloroprene levels were relatively high who could not be identified or located for inclusion in the
studies. This raises the possibility that the actual number of 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 ascertainment 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 reported
expected counts for liver and lung cancer mortality (Li et al., 1989, 625181; Bulbulyan et al., 1998,
625105; Bulbulyan et al., 1999, 157419). 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 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 1575 male employees (salaried and female employees excluded due to "minimal or no
potential exposure to chloroprene") who were working at the Louisville plant on 6/30/1957. 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
July 2010
4-76 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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.
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 (2005a) 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 to 20 years (Marsh et al.,
2007, 625187; Marsh et al., 2007, 625188; Colonna and Laydevant, 2001, 625112; Bulbulyan et al.,
1998, 625105; 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
July 2010 4-77 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
stasticially significant two- to five-fold increased risk of liver cancer in four cohort studies in China (Li
et al., 1989, 625181). Louisville, United States (Leet and Selevan, 1982, 0949701 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
incidence (RRs = 2.9-4.9, statistically signficant) 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; Li et al., 1989, 625181;
Leet and Selevan, 1982, 094970). 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 investigates 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.
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 highly exposed workers to low or
unexposed workers. 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 (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 in the upper exposure categories (Bulbulyan et al., 1998,
July 2010
4-78 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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. /aftora/orj' Animal
According to the NTP (1998, 042076). there is clear evidence of carcinogenicity in the F344/N
rat and B6C3F1 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 of that chloroprene is more potent in inducing tumors in B6C3F1 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). see 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.
July 2010 4-79 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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 B6C3F1 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 (1-
chloroethenyl)oxirane were as great as 12-fold greater (F344 rats vs. 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 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 B6C3F1 mice in the NTP (1998, 042076) chronic bioassay.
In the only long-term oral cancer study (an F1 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;
July 2010 4-80 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Bulbulyan et al., 1999, 157419; Li et al., 1989, 625181; Leet and Selevan, 1982, 094970) and lung
(Marsh et al., 2007, 625188; Colonna and Laydevant, 2001, 625112; Bulbulyan et al., 1998, 625105;
Bulbulyan et al., 1999, 157419; Leet and Selevan, 1982, 094970; Pell, 1978, 0649571 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., 1999, 157419; Bulbulyan et al., 1998, 625105; Li et al., 1989, 625181; Leet and
Selevan, 1982, 094970). with a two- to more than five-fold 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 workers' 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).
These limitations are further discussed in Section 4.7.1.1.
According to NTP (1998, 042076), there is clear evidence of carcinogenicity in the F344/N rat
and B6C3F1 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 (see Section 4.7.3.2).
July 2010
4-81 DRAFT - DO NOT CITE OR QUOTE
-------
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 B6C3F1 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 male B6C3F1 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 4 occupational epidemiology studies (out of 9 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). Bulbulvan et al. (1998. 625105). and Bulbulvan 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
July 2010 4-82 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
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
(2005, 088823; 2000, 194536)) 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;
U.S. EPA, 2002, 052153; Sills et al., 1999, 624952). 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 Supportfor the f/j'pothe,sizedMode 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 genders 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 CYP 2E1. The oxidation of chloroprene to
(l-chloroethynyl)oxirane is more prevalent in B6C3F1 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 gender, (l-chloroethynyl)oxirane was the major metabolite detected.
One distinct difference between species was the stereospecificity of epoxide metabolites formed. In 2
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
July 2010 4-83 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 (Watson et al., 2001, 625045; U.S. EPA, 2000, 194536).
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., 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., 2001,
624922; Sills et al., 1999, 624952). 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 heptatocellular carcinomas (NTP,
1998, 042076; U.S. EPA, 2000, 194536). Activated K-ras oncogenes were observed in lung tumors,
hepatocellular carcinomas, and lymphomas in B6C3F1 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 B6C3F1 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
substitution mutation {Salmonella typhimurium strains TA100 and TA 1535) assays have been positive
(Willems, 1980, 625049; Bartsch et al., 1979, 010689) while the bacterial frame shift (S. typhimurium
July 2010 4-84 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 tranversions (Sills et al., 2001, 624922; Sills et al., 1999, 624952).
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 (Willems, 1978, 625048; Willems, 1980, 625049; Bartsch et al., 1979, 010689) 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 (Himmelstein et
al., 2001, 019013; Drevon and Kuroki, 1979, 010680). 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
B6C3F1 mice (Sills et al., 2001, 624922; Sills et al., 1999, 624952). 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 (see 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
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
July 2010
4-85 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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-ra.s proto-
oncogenes in chloroprene-induced lung, forestomach, and Harderian gland neoplasms observed in the
NTP (1998) study.
4.7.3.3 Conclusions about t/ie HypothesizedMode 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. (2001, 624922; 1999, 624952) show codon-specific (codons 12, 13, and 61) mutations in the H-
and K-ra.s 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., 2001, 624922; Sills et al., 1999, 624952). These findings provide additional support for
the importance of ras proto-oncogene activation via mutation in the carcinogenesis of chloroprene and
related compounds.
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
July 2010
4-86 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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 vs 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 occured 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 (Sills et al., 2001, 624922; Sills et al., 1999, 624952;
NTP, 1998, 042076). 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., 2001, 624922; Sills et al., 1999, 624952) 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.
There are no data comparing the carcinogenicity of chloroprene after exposure during early life with
the carcinogenicity after exposure during adulthood. Exposure to chloroprene commenced at about
July 2010
4-87 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
6 weeks of age in mice and rats, and continued through adulthood in the 2-year chronic assay (NTP,
1998. 042076V
Therefore, because the weight of evidence supports a mutagenic mode of action for chloroprene
carcinogenicity (see 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) 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
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
July 2010 4-88 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
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 (Xb2 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 (See 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 EPA's Supplemental Guidance for Assessing Susceptibility From Early-Life Exposure
to Carcinogens (U.S. EPA, 2005, 088823).
4.8.2. Possible Gender Differences
In lifetime studies conducted in the rat, mouse, and hamster, chloroprene was not shown to
exhibit any remarkable gender-related differences in effects with the exception of a more pronounced
neoplastic response in B6C3F1 female mice compared to males.
July 2010
4-89 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
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 exposed rats to chloroprene at one dose (50
mg/kg/day) and only qualitatively reported non-cancer 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 (see 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 adverse 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
"3
to the respiratory system (extra-respiratory or systemic effects). It is generally expressed in mg/m .
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 non-
cancer effects. Two inhalation studies investigating portal-of-entry (nasal and pulmonary) and
systemic effects were identified in the literature and considered for the principal study for derivation of
an RfC: a 2-year chronic study in B6C3F1 mice and F344 rats (NTP, 1998, 042076). and a 2-year
chronic study in Wistar rats and Syrian gold hamsters (Trochimowicz et al., 1998, 625008).
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
July 2010
5-1
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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 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 males and 73 females 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 non-cancer 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 day 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
adverse effects at similar exposure levels as the NTP (1998, 042076) study (Trochimowicz et al. (1998,
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.
July 2010
5-2
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
From the NTP (1998, 042076) study, all portal-of-entry and systemic 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).
Table 5-1. Incidences of nonneoplastic lesions resulting from chronic exposure
(ppm) to chloroprene considered for identification of critical effect
SPECIES
TISSUE
ENDPOINT
MALE
FEMALE
0
12.8
32
80
0
12.8
32
80
Mice
Nose
Suppurative
inflammation
--
--
--
--
0/50
1/49
3/49*
4/50**
Lung
Bronchiolar
hyperplasia
0/50
10/50**
18/50**
23/50**
0/50
15/49**
12/50**
30/50**
Kidney
Renal tubule
hyperplasia
2/50
16/49**
17/50**
18/50**
--
--
--
--
Fore-
stomach
Epithelial
hyperplasia
4/50
6/48
7/49
29/50**
4/50
3/49
8/49
27/50**
Spleen
Hematopoietic
cell
proliferation
26/50
22/49
35/50a
31/503
13/50
25/49a
42/49a
39/503
Rats
Nose
Atrophy
3/50
12/50*
46/49**
48/49**
0/49
1/50
40/50**
50/50**
Basal cell
hyperplasia
0/50
0/50
38/49**
46/49**
0/49
0/50
17/50**
49/50**
Metaplasia
6/50
5/50
45/49**
48/49**
0/49
1/50
35/50**
50/50**
Necrosis
0/50
11/50**
26/49**
19/49**
0/49
0/50
8/50**
12/50**
Inflammation,
chronic
0/50
5/50*
9/49**
49/49**
--
--
--
--
Lung
Alveolar
epithelial
hyperplasia
5/50
16/50**
14/49*
25/50**
6/49
22/50**
22/50**
34/50**
Kidney
Renal tubule
hyperplasia
14/50
20/50
28/50**
34/50**
6/49
6/50
11/50
21/50
* p < 0.05
** p < 0.01
-- Endpoint not considered for selection of critical effect
a Reported as statistically significantly greater than controls, but level of significance not reported
Source: NTP (1998, 042076)
July 2010
5-3
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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
amenable to BMD modeling for a number of reasons, including the observation of maximal or near-
maximal reponse 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 under the assumption that it
represents a minimal biologically significant change. 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 (see Table B-l). In the case of endpoints that progressed in incidence as well as
severity (i.e., progression from mild to moderate lesions) from the control dose to the lowest dose
showing reponse, a BMR of 10% was not considerd to be a minimal biologically significant change.
Therefore, for these endpoints, a BMR of 5% was used. All available dichotomous models in the EPA
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 [% ] 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-reponse data across the entire range of doses, whereas the % 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 % 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
July 2010
5-4
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
selected out of the group. When the calculated BMDLs are within a 3-fold 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 BMDLs are not within a
3-fold 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 Bl.
The BMDs and BMDLs associated with an extra risk of 10% or 5% for the best-fitting models
for each endpoint are shown in Table 5-2. NOAELs and LOAELs were used as potential PODs 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 are calculated as follows:
PODadj (ppm) = POD (ppm) x hours exposed per day/24 hours x days exposed per week/7days
"3
RfCs are typically expressed in units of mg/m ; the above ppm value needs to be converted
"3
using the chemical specific conversion factor of 1 ppm = 3.62 mg/m (see Table 2-1) for chloroprene.
Therefore, the final PODAdj values are calculated as follows:
PODadj (mg/m3) = PODadj(ppm) x 3.62 mg/m3/lppm
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 5days/7days
PODadj (ppm) = 0.6 ppm
PODadj (mg/m3) = 0.6 ppm x 3.62 mg/m3/lppm
PODadj (mg/m3) = 2.3 mg/m3
"3
The calculated PODadj (mg/m ) values for all considered endpoints are presented in the last column of
Table 5-2.
July 2010
5-5
DRAFT - DO NOT CITE OR QUOTE
-------
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)
PODadjc
(mg/m3)
Nasal Effects - Olfactory
Atrophy
Rat/male
--
12.8
Logisticd
5
4.9
3.5
2.3
Rat/female
12.8
32
e
-
-
-
8.3
Basal cell
hyperplasia
Rat/male
12.8
32
e
-
-
-
8.3
Rat/female
12.8
32
Log-probitf
10
23.5
19.7
12.7
Metaplasia
Rat/male
12.8
32
e
-
-
-
8.3
Rat/female
12.8
32
e
-
-
-
8.3
Necrosis
Rat/male
--
12.8
Log-probitd
5
5.6
4.5
2.9
Rat/female
12.8
32
Log-probitf
5
24.8
19.7
12.7
Chronic
inflammation
Rat/male
--
12.8
Log-logisticd
10
14.6
9.3
6.0
Suppurative
inflammation
Mouse/female
12.8
32
e
-
-
-
8.3
Lung Effects
Alveolar
hyperplasia
Rat/male
--
12.8
Log-logistic
5
5.4
3.3
2.1
Rat/female
--
12.8
Log-logistic
10
4.9
3.3
2.1
Bronchiolar
hyperplasia
Mouse/male
--
12.8
Log-logistic
5
3.6
2.7
1.7
Mouse/female
--
12.8
g
-
-
-
8.3
Systemic Effects
Kidney (renal
tubules)
hyperplasia
Rat/male
12.8
32
Log-logistic
10
6.5
4.0
2.6
Rat/female
32
80
Log-probit
10
32.5
23.5
15.2
Mouse/male
--
12.8
e
-
-
-
8.3
Forestomach
epithelial
hyperplasia
Mouse/male
32
80
Multistage
10
24.7
20.5
13.3
Mouse/female
32
80
Multistage
10
31.0
19.3
12.5
Splenic
hematopoietic
proliferation
Mouse/male
12.8
32
e
-
-
-
8.3
Mouse/female
--
12.8
Probitd
5
2.1
1.7
1.1
"Best fitting model as determined by goodness-of-fit statistics. Bold numbers indicate which value (BMDL, NOAEL, or
LOAEL) is used in calculation of PODadj
b BMR = benchmark dose response
0 Duration adjusted POD [mg/m3] (PODadj) = POD [ppm] x (3.62 mg/m3/ppm) x (5 days/7days) x (6 hours/24 hours), in
accordance with EPA policy (2002a)
dHigh dose group was dropped in order to obtain adequate model fit
e Did 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 recommend to determine a POD
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 (see Appendix B for details)
g No model fits appropriately according to fit statistics or visual inspection.
Source: NTP (1998, 042076s)
July 2010
5-6
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
The results of BMD modeling indicated that splenic hematopoietic cell proliferation in the
"3
female mouse was the most sensitive endpoint, with a PODAdj value of 1.1 mg/m . Several other
endpoints (olfactory atrophy and necrosis in the male rat, alveolar hyperplasia in male and female rats,
bronchiolar hyperplasia in male mice, and renal tubule hyperplasia in male mice) had somewhat higher
"3
PODadj values, ranging from 1.7 to 2.9 mg/m .
Chloroprene is a relatively water-insoluble, non-reactive gas, with an approximate blood:air
partition coefficient of less than 10 (see Table 3-1), that induces a range of nasal, thoracic, and
systemic non-cancer effects. Water-insoluble, non-reactive 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., non-
respiratory) effects resultant from chloroprene exposure clearly indicates the compound is absorbed
into the bloodstream and distributed throughout the body. However, the pattern of respiratory 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 non-hepatic
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. As it is also observed that chloroprene induces adverse effects in organ
systems distal to the portal-of-entry, consistent with the parent compound's water-insoluble and non-
reactive chemical properties, it is possible that observed nasal and respiratory effects are due to
systemic redistribution of chloroprene to these tissues. Currently, the contribution of either route of
delivery (portal-of-entry vs. systemic distribution) to the induction of nonneoplastic respiratory effects
is unknown.
However, the selected critical effect, splenic hematopoietic cell proliferation, is clearly a
systemic effect and the human equivalent concentration (HEC) for chloroprene was calculated by the
applicaton of the appropriate dosimetric adjustment factor (DAF) 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 gases with
systemic effects, the DAF is expressed as the ratio between the animal and human blood:air partition
coefficients:
July 2010
5-7
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
DAF = (Hb/g)A/(Hb/g)H
where:
(Hb/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 (see
Table 3-1), resulting in a DAF >1, a default value of 1 is substituted (U.S. EPA, 1994, 006488).
Therefore, the HEC for splenic hematopoietic cell proliferation in female mice is calculated as follows:
PODrec (mg/m3) = PODadj (mg/m3) x DAF
= PODadj (mg/m3) x 1.0
= 1.1 mg/m3 x 1.0
= 1.1 mg/m3
"3
Therefore, the PODrec of 1.1 mg/m for the critical effect of splenic hematopoietic cell
proliferation in female mice was selected for the derivation of the RfC for chloroprene.
July 2010
5-8
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
5.2.4. RfC Derivation—Including Application of Uncertainty Factors
-3
A PODhec value of 1.1 mg/m for increased incidence of splenic hematopoietic cell
proliferation in female B6C3F1 mice (NTP, 1998, 042076) was used as the POD to derive the chronic
RfC for chloroprene. A total UF of 100 was applied to this PODhec as described below:
• A 3-fold UFa 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 a 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) (see Section 4.8). Due to this limited data on
variations in susceptibility within the human population, 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 splenic hematopoietic cell proliferation was
selected under an assumption that is represents a minimal biologically significant change.
• A 3-fold UF was used to account for deficiencies in the database. The major 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 day 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 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
July 2010
5-9
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
(1979, 064938) unpublished study exposed F0 and Fi rats to chloroprene, but did not allow
the Fi rats to mate).
Application of this 100-fold composite uncertainty factor yields the calculation of the chronic
RfC for chloroprene as follows:
RfC = PODhec^" UF= 1.1 mg/m3 100 = 1.1 x 10"2 mg/m3 = 1 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.
July 2010
5-10 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
10.00
1.00
0.10
0.01
AHEC
¦ Animal to Human
~ Human Variation
¦ Database
• RfC
I
i
I
A i *
~
I
V ~
0.00
— 2
"aj qj
5 £
CO M~
S Q
B- 2
£ m
_ra q
Si 5
o. ca
QJ
s 3
- E
E o
Figure 5-1. Points of departure (in mg/m3) for selected endpoints with corresponding
applied uncertainty factors and derived sample RfCs (chosen RfC value is circled)
5.3 UNCERTAINTIES IN THE INHALATION REFERENCE CONCENTRATION
As presented earlier in the previous section, the UF approach, following EPA practices and RfC
guidance (U.S. EPA, 1994, 006488), was applied to the PODHec in order to derive the chronic RfC.
Factors accounting for uncertainties associated with a number of steps in the analyses were adopted to
account for extrapolation from an animal bioassay to human exposure, a diverse population of varying
susceptibilities, POD determination methodologies (NOAEL, LOAEL, or BMDL), and to account for
database deficiencies. The following is a more extensive discussion of the uncertainties associated
with the RfC for chloroprene beyond which is described quantitatively in Section 5.2.4. A summary is
provided in Table 5-3.
Choice of endpoint. Sample RfCs considered from the NTP (1998, 042076) chronic inhalation
3 13
study ranged from 6.0 x 10" to 2.0 x 10" mg/m . Sample RfC values primarily depended on whether
olfactory, pulmonary, or systemic effects were considered, what mode of delivery was assumed (i.e.,
portal-of-entry vs. systemic delivery), what BMR was chosen, and whether the PODadj or PODhec was
used for selection of endpoint. The chosen critical effect, increased incidence of splenic hematopoietic
cell proliferation in female mice, was considered to be the most sensitive endpoint because it returned
July 2010 5-11 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
the lowest PODAdj values compared to all other considered endpoints.
In the case of chloroprene, the POD adj was used as the basis for selection of the critical effect
because of uncertainties concerning the application of a portion of the DAF procedures used in
calculating the PODHec HECs are human toxicokinetically equivalent external air concentrations of
the animal PODadj that have been dosimetrically adjusted through application of DAFs, in accordance
with the current RfC methodology (U.S. EPA, 1994, 006488). The DAF for respiratory portal-of-entry
effects is the regional gas dose ratio (RGDR), a ratio of animal and human physiological parameters
that normalizes inhaled doses (expressed as the minute volume) to surface areas of specific regions in
the respiratory tract (i.e., the extrathoracic or nasal, region). The normalization of inhaled dose to
respiratory tract surface area outlined in the RfC methodology is based on three major assumptions:
uniformity in the flow of gas through the respiratory tract region of interest, uniformity and
equivalency (in terms of cell types, relative amount, and distribution) of the surface areas of the
respiratory tract regions subtending the gas flow, and uniformity in the deposition of gas over the
entire surface area of the respiratory tract region of interest. Application of the default DAFs, based on
the ratio of minute volume and respiratory tract surface area between rats and humans, indicates that
humans receive approximately five times the dose in the nasal region compared to rats (i.e., DAF =
0.28). Recent advances have been made in inhalation dosimetry, including analyses investigating the
nature of gas flow through the upper respiratory tract (via flow-dye cast models and computational
fluid dynamics modeling) and the deposition of inhaled toxicants in the respiratory tract (as detailed in
U.S. EPA, 2009, 625038). These analyses have shown that gas flow through the upper respiratory
tract is not uniform in rats or humans, and that uneven gas flow results in uneven distribution of gas
deposition. Further advances in modeling, combining CFD and PBPK models, have allowed for the
calculation of target tissue doses for a number of chemicals in both rats and humans, and therefore for
calculation of alternative DAFs for interspecies extrapolation. Use of these hybrid CFD-PBPK models
generally indicate that the upper respiratory system (i.e., nasal region) of rats and humans receive
nearly equivalent doses at equivalent external exposure concentrations for most gases (i.e., DAFs ~ 1).
For some gases, though, CFD-PBPK modeling indicates that rats receive higher doses in the nasal
region than humans (i.e., DAFs > 1). No such CFD-PBPK model currently exists for chloroprene.
The default DAF for systemic effects is the ratio of animal and human blood:air partition coefficients.
When the animal value is greater than the human value the DAF for systemic effects defaults to one.
Selection of critical effect based on PODhec values calculated using default DAFs, would result in
olfactory effects (either necrosis or atrophy) being identified as the critical effect. Selection of these
two outcomes as critical effects would have lowered the derived RfC approximately 20-40% (8.0 x 10"
3 3 3 3
mg/m for necrosis or 6.0 x 10" mg/m for atrophy) relative to splenic hematopoietic proliferation.
Uncertainty also surrounded the assumed mode of 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
July 2010
5-12 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
category 3 gas, being relatively water insoluble and non-reactive, and would be expected to only elicit
extrarespiratory effects. Medinsky and Bond (2001, 016157) describe an alternative method of
classification that represents the reactivity and water solubility as continuous variables that allow
insight to be gained regarding sites of and conditions of toxicity. Additionally, although not explicitly
included in this scheme, in situ metabolism in the respiratory tract can be considered as a component of
reactivity. In this manner, chloroprene can best be described as both a non-reactive, water insoluble
chemical that is absorbed into the blood stream and induces systemic toxicity, as well as a chemical
that is metabolized into a reactive epoxide within the respiratory tract, inducing portal-of-entry
toxicity. This method of 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, which is expressed in both the olfactory and pulmonary regions of the respiratory tract. In
situ metabolism in the respiratory tract may thus explain a portion of the biological activity of
chloroprene in these regions. It is observed that chloroprene induces adverse effects in organ systems
distal to the portal-of-entry, consistent with its water insoluble and non-reactive chemical properties.
Therefore, it is possible that observed pulmonary effects, such as alveolar hyperplasia in the female rat,
may be explained by blood-borne delivery, rather than air-borne. Currently, the contribution of either
mode of delivery to the sites of observed toxicity is not known. However, splenic hematopoietic cell
proliferation is unequivocally a systemic lesion, and calculation of the PODhec uses a DAF = 1 and
results in a value of 1.1 mg/m .
Choice of pulmonary effects treated as portal-of-entry effects would result in RfCs up to 20-
fold higher than the RfC for splenic hematopoietic cell proliferation. For example, bronchiolar
"3
hyperplasia in male mice, using a BMR of 5% extra risk results in a PODadj =1.7 mg/m . Treating
this endpoint as a portal-of-entry effect dictates that a DAF of 4.1 be applied, resulting in a PODhec of
3 2 3
7.0 mg/m and an ultimate RfC of 7.0 x 10" mg/m . Choice of other systemic effects (i.e., forestomach
epithelial hyperplasia in the male mouse) or pulmonary effects treated as systemic lesions would result
in RfCs up to 15-fold greater.
In summary, for chloroprene, the PODadj was used as the basis for selection of the critical
effect because of uncertainties concerning the application of a portion of the DAF procedures used in
calculating the PODhec, as pointed out by the External Peer Review committee. Based on the PODadj
of 1.1 mg/m , splenic hematopoietic proliferation was chosen as the critical effect, with an ultimate
RfC of 1 x 10"2 mg/m3 being derived.
Choice of model for BMDL derivation. 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 are well-predicted near the BMD (x residual = 0.033). Use of other models
would either increase or decrease the RfC by approximately 50%. However, the probit model was
chosen over these models based on current BMD technical guidance (U.S. EPA, 2000, 052150).
Choice of BMR. There is uncertainty in the selection of the benchmark response (BMR) level.
For increased incidence of splenic hematopoietic cell proliferation in female mice, definitive data do
July 2010
5-13 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
not exist to further inform the selection of what the appropriate BMR should be. However, the
observation was made that the incidence and severity of this lesion increases in low dose animals
compared to control animals; therefore a BMR of 5% extra risk was chosen based on the assumption
that a 5% increase in incidence of this effect is minimally biologically significant.
Statistical uncertainty at POD. For the probit model applied to splenic hematopoietic cell
proliferation in female mice, there is a reasonably small degree of statistical uncertainty at the 5% extra
risk level (the point of departure for derivation of the RfC), with the BMDL being about 20% below
the BMD.
Choice ofbioassay. 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 splenic hematopoietic cell
proliferation in female mice exposed to chloroprene via inhalation for 2 years. Use of other effects that
occurred in another species, F344/N rats, would result in RfCs approximately 40% lower to 20 times
greater than the current 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 3-fold (Bernauer et al., 2003, 625103). Neafsey et al.
(2009, 196814) concluded that evidence for particular CYP2E1 polymorphisms having significant
effect 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.
July 2010
5-14 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-3. Summary of Uncertainties in the Chloroprene noncancer risk
assessment
CONSIDERATION
POTENTIAL
IMPACT3
DECISION
JUSTIFICATION
Choice of endpoint
Use of other
endpoints could t
RfC by up to 20-fold
or I RfC by up to
40%
RfC is based on the
endpoint with the lowest
PODadj , increased
incidence of splenic
hematopoietic cell
proliferation in female
mice
Chosen endpoint is considered to be the most
sensitive (based on PODadj values). Its
observed systemic toxicity is consistent with
the physio-chemical properties of
chloroprene. Selection of the critical effect
was based on the PODadj consistent with
peer reviewer comments.
Choice of model for
BMDL derivation
Other models would
t or | RfC
Probit model used
U.S. EPA (2000, 052150) BMD technical
guidance used to choose model based on
global and local measures of model fit
Choice of BMR
Other BMRs would t
or | RfC
BMR of 5% extra risk
chosen
BMR of 5% extra risk was chosen based on
the assumption that a 5% increase in
incidence of this effect is minimally
biologically significant
Statistical uncertainty
at POD
RfC would be ~ 20%
higher if BMD (vs.
BMDL) were used
BMDL used as POD per
U.S. EPA guidance
(2000, 052150)
Size of bioassay results in sampling
variability; lower bound is 95% confidence
interval on administered exposure
Choice of bioassay
Other bioassays
could t or I RfC
NTP (1998.042076)
used as critical study
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
Choice of species
RfC would t or J, if
based on another
species
Mice chosen
RfC is based on the most sensitive endpoint
(incidence of splenic hematopoietic cell
proliferation) in the most sensitive species
(mouse), based on PODadj
Human population
variability
RfC could t or 4 if a
non-default value of
UF was used
10-fold uncertainty
factor applied to derive
the RfC
10-fold UF, the default value, is applied
principally because of limited data on human
variability or potential susceptible
subpopulations
July 2010
5-15 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
CONSIDERATION
POTENTIAL
IMPACT3
DECISION
JUSTIFICATION
Completeness of the
database
RfC could t or J, if a
different UF for
database limitations
was applied
3-fold uncertainty factor
applied to derive the
RfC
3-fold 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 t = increase, j = decrease
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 (see 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 B6C3F1 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 femal Wistar and Syrian gold 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 (see 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 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), and Zymbal's gland carcinomas (females). These tumors generally appeared earlier
with increasing exposure levels and showed statistically significantly increasing trends with increasing
July 2010 5-16 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
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 female mice and for male mice
in the two higher exposed groups was statistically significantly lower than for the corresponding
control mice. Individual animal data including the time of observation of tumors is provided in Table
C-l.
Rats demonstrated statistically significant increases in tumor incidence at multiple sites as well:
oral cavity, squamous cell papillomas or carcinomas; thyroid gland, follicular cell adenomas or
carcinomas; renal tubule adenomas or carcinomas; alveolar/bronchiolar adenomas or carcinomas
(males); and mammary gland fibroadenomas. 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 B6C3F1 mice exposed to
chloroprene via inhalation
TISSUE
ADMINISTERED CHLOROPRENE
CONCENTRATION (ppm)
Control 12.8 32 80
Females
All organs: hemangioma or hemangiosarcoma
Unadjusted rate
First incidence (days)
4/50
541
6/49
482
18/50
216
8/50
523
Lung: alveolar/bronchiolar adenoma or
carcinoma
Unadjusted rate
First incidence (days)
4/50
706
28/49
447
34/50
346
42/50
324
Liver: hepatocellular adenoma or carcinoma
Unadjusted rate
First incidence (days)
20/50
493
26/49
440
20/50
503
30/50
384
Skin or mesentery: sarcoma
Unadjusted rate
First incidence (days)
0/50
11/49
285
11/50
524
18/50
462
Mammary gland: adenocarcinoma, carcinoma
or adenoacanthoma
Unadjusted rate
First incidence (days)
3/50
527
6/49
440
11/50
394
14/50
336
Forestomach: squamous cell papilloma or
carcinoma
Unadjusted rate
First incidence (days)
1/50
734
0/49
0/50
4/50
576
Harderian gland3: adenoma or carcinoma
Unadjusted rate
First incidence (days)
2/50
527
5/50
621
3/50
524
9/50
467
Zymbal's gland3: carcinoma
Unadjusted rate
First incidence (days)
0/50
0/50
0/50
3/50
565
July 2010
5-17 DRAFT - DO NOT CITE OR QUOTE
-------
Males
All organs: hemangioma or hemangiosarcoma
Unadjusted rate
First incidence (days)
3/50
733
14/50
659
23/50
495
21/50
454
Lung: alveolar/bronchiolar adenoma or
carcinoma
Unadjusted rate
First incidence (days)
13/50
635
28/50
530
36/50
382
43/50
523
Forestomach: squamous cell papilloma or
carcinoma
Unadjusted rate
First incidence (days)
1/50
733
0/48
2/49
733
5/50
587
Harderian gland3: adenoma or carcinoma
Unadjusted rate
First incidence (days)
2/50
596
5/50
701
10/50
596
12/50
589
Kidney: renal tubule adenomas or carcinomas
(extended and standard evaluations combined)
Unadjusted rate
First incidence (days)
0/50
2/49
722
3/50
715
9/50
567
a Tissues were examined histopathologically only if a lesion was observed grossly at necropy
Source: NTP (1998, 042076).
Table 5-5. Tumor incidence in female and male F344 rats exposed to chloroprene
via inhalation
TISSUE
ADMINISTERED CHLOROPRENE
CONCENTRATION (ppm)
Control
12.8
32
80
Females
Oral cavity: papillomas or carcinomas
Unadjusted
First incidence (days)
1/49
687
3/50
681
5/50
588
11/50
660
Thyroid gland: follicular cell adenomas or
carcinomas
Unadjusted
First incidence (days)
1/49
733
1/50
721
1/50
733
5/50
617
Mammary gland: fibroadenomas
Unadjusted
First incidence (days)
24/49
366
32/50
302
36/50
470
36/50
433
Kidney: renal tubule adenomas or
carcinomas (extended and standard
evaluations combined)
Unadjusted
First incidence (days)
0/49
0/50
0/50
4/50
609
Males
Oral cavity: papillomas or carcinomas
Unadjusted
First incidence (days)
0/50
2/50
701
5/50
609
12/50
539
Thyroid gland: follicular cell adenomas or
carcinomas
Unadjusted
First incidence (days)
0/50
2/50
597
4/49
569
5/50
307
Lung: alveolar/bronchiolar adenoma or
carcinoma
Unadjusted
First incidence (days)
2/50
616
2/50
702
4/49
505
6/50
540
Kidney: renal tubule adenomas or
carcinomas (extended and standard
evaluations combined)
Unadjusted
First incidence (days)
1/50
733
8/50
600
6/50
679
8/50
625
a Kaplan-Meier 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).
1 5.4.3. Dose Adjustments and Extrapolation Methods
2 The current EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005, 086237)
3 emphasize that the method used to characterize and quantify cancer risk from a chemical is determined
July 2010
5-18 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
by what is known about the MOA 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
MOA because of DNA reactivity or if another MOA that 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.
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[-(bo + bid + b2d2 + ... + by^) x (t - tof]
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; l 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
parameter estimated in fitting the model, which characterizes the change in response with age. The
parameter to 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's MSW computer software program
(http://epa.gov/ncea/bmds/msw.html), which is based on Weibull models drawn from Krewski et al.
(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. 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
July 2010 5-19 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
of an animal, while fatal tumors are thought to have resulted in animal death. Hemangiosarcomas were
treated as fatal tumors unless observed at terminal sacrifice, in which case they were considered
incidental. Furthermore, these fatal tumors were deemed rapidly fatal, and t0 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, 052153). If
twice the difference in log-likelihoods was less than a % 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)]/[1 - 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 HECs for
the various tumors were calculated by the application of DAFs (see also section 5.2.3). With the
exception of the lung tumors, all tumors were treated as systemic effects. For these sites a DAF of 1.0
was applied as the value for the rat blood:air partition coefficient exceeded the human value. For
alveolar/bronchiolar tumors, the HEC was calculated treating the neoplasms alternatively as portal-of-
entry effects or systemic effects. As there is evidence that chloroprene and/or its metabolite are
distributed systemically (i.e., the observation of tumors in multiple organ systems), there is the
potential that chloroprene is redistributed to the lungs. In this manner, chloroprene may induce lung
tumors as a systemically delivered carcinogen in addition to inducing tumors via inhalation (see
Section 5.2.3 and 5.3 for a additional discussion). However, the contribution of either route of delivery
(i.e., inhalation vs. bloodstream) to the induction of lung tumors is currently unknown. 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. For reactive gases (in this case, the reactive epoxide metabolite of
chloroprene) with portal-of-entry effects in the pulmonary region, the DAF is the regional gas dose
ratio (RGDRpu) and is expressed (in this particular case, for female rats) as follows:
RGDRpu = (MYr/SPU_r)/ (MVh/SPU_h)
July 2010
5-20 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
where:
MVr = B6C3F1 female mouse minute volume (0.053 L/min)6
MVh = human minute volume, (13.8 L/min)
Spu r = surface area of the pulmonary region in rats (.05 m2)
— 2
Spu h = surface area of the pulmonary region in humans (54 cm ),
RGDRPU= (.053/.05)/(13.8/54)
RGDRPU= 4.1
For non-reactive gases (in this case the proportion of inhaled chloroprene not metabolized
within the respiratory tract) with systemic effects, the DAF is expressed as the ratio between the
animal and human blood:air partition coefficients (see calculation in Section 5.2.3). In cases where the
animal blood:air partition coefficient is higher than the human value, resulting in a DAF >1, a default
value of 1 is substituted as for the other systemic effects considered above. Therefore, chloroprene-
induced lung tumors were treated as either point-of-entry lesions using a DAF of 4.1, or as systemic
lesions using a DAF of 1.0 due to a lack of data clarifying whether one or both modes were more likely
to be operating.
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 (see 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^ per |ig/m3, approximately a 50-fold range. The highest unit risk (1.8 x 10 4 per (J,g/m3)
corresponded to lung tumors treated as systemic lesions in female mice, and the lowest unit risk for
female mice corresponded to forestomach tumors, 3.4 x 10 6 per (J,g/m3. The highest unit risk in male
_c o
mice, 8.3 x 10 per |ig/m , also corresponded to lung tumors treated as systemic lesions, and was
approximately twofold lower than in female mice. The lowest unit risk in male mice was for renal
tubule adenomas or carcinomas, at 9.4 x 10 6 |ig/m3. Note that lung tumors were the most sensitive
6 Calculated according to U.S. EPA (1994b): ln(MVr) = b0 + bi x ln(BW). Default minute volume is in L/min; b0 and bi =
species-specific (rat) intercept and coefficient used; body weight in kg. Time-weighted average body weight was 0.289, kg
for female rats
July 2010
5-21 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
response in mice for the structurally related compound 1,3-butadiene as well (U.S. EPA, 2002,
052153).
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
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; see Section 5.4.3.)
were nearly two-fold 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 3-fold lower.
Alternatively, if the lung tumors were strictly portal-of-entry effects, the estimated human
equivalent unit risks associated with this site would be four-fold lower, and the highest unit risk
_C T
estimates would be from female mice liver tumors, at 6.3 x 10 per (j,g/m , and male mice
_5 -3
hemangiomas and hemangiosarcomas (early instances considered fatal), at 4.7 x 10 per (j,g/m .
Concerning the unit risks for the two sites without complete histopathologic evaluation,
Harderian gland and Zymbal's gland (see Section 5.4.2); the female mice Zymbal's gland unit risk was
quite low, at 3.5 x 10 6 per (J,g/m3, virtually identical to the forestomach unit risk in both female and
_5 _C T
male mice. The Harderian gland unit risks were 1.2 x 10 and 1.5 x 10 per (J,g/m , 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. For
male mice, this unit risk was less than twofold higher than that for renal tumors, the lowest male mouse
unit risk.
Given the multiplicity of tumor sites, however, 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
July 2010
5-22 DRAFT - DO NOT CITE OR QUOTE
-------
1 the multistage model, but available biological information (time of tumor observation) would then be
2 ignored.
3
Table 5-6. Dose-response modeling summary for female mouse tumors associated
with inhalation exposure to chloroprene
Power
Parameter
ca
POINT OF DEPARTUREb
UNIT
RISKd
l(ji g/m3)
COMPOSITE
UNIT RISK6
/(Hg/m3)
TUMOR TYPE*
BMR
Modeled from
bioassay (ppm)
Continuous, Human
equivalent0 (jig/m3)
BMDL
BMD
BMDL
BMD
Lung: alveolar/
bronchiolar
adenoma or
3.8
0.1
0.88
1.20
5.69 x 102
2.33 x 103
7.71 x 102
3.16 x 103
1.8 x 10'4
4.3 x 10~5
carcinomaf
All organs:
hemangio-
sarcomas,
hemangiomas8'h
5.9
0.1
5.75
10.1
3.71 x 103
6.52 x 103
2.7 x 10~5
All organs:
hemangio-
sarcomas,
hemangiomas8'1
1.0
0.1
11.1
14.9
7.13 x 103
9.62 x 103
1.4 x 10~5
Mammary gland:
adenocarcinoma,
carcinoma or
adenoacanthoma
1.0
0.1
14.1
20.4
9.06 x 103
1.32 x 104
1.1 x 10~5
2.7 X la4
1.5 x 10"1
Forestomach:
squamous cell
papilloma or
carcinoma
4.1
0.1
46.3
67.8
2.98 x 104
4.37 x 104
3.4 x 10~6
Liver:
hepatocellular
adenoma or
4.2
0.1
2.45
4.24
1.58 x 103
2.73 x 103
6.3 x 10~5
carcinoma
Harderian gland:
adenoma or
2.9
0.1
12.6
27.1
8.13 x 103
1.75 x 104
1.2 x 10~5
carcinoma
Skin: sarcoma
1.6
0.1
7.18
9.49
4.63 x 103
6.11 x 103
2.2 x 10~5
Zymbal's gland:
carcinoma
1.1
0.05
22.5
80.5
1.45 x 104
5.19 x 104
3.5 x 10~6
a Multistage-Weibull model: P(d) = 1 - exp[-(b0 + bid + b2d2 + ... + bkdk) x (t-t0)c], coefficients estimated in terms
of ppm as administered in bioassay; lower stage bi not listed were estimated to be zero. See Appendix C for
modeling details.
b BMD = Concentration at specified extra risk;
BMDL = 95% lower bound on concentration at specified extra risk.
0 Continuous equivalent estimated by multiplying exposures by (6 hours)/(24 hours) x (5 days)/(7 days).
d Unit risk estimated by dividing the BMR by the BMDL.
e Composite unit risk estimate, across all sites listed; see text for method.
f Values in italics indicate BMD/BMDL when lung tumors are treated as systemic lesions.
8 Highest exposure group dropped in order to better characterize low-dose responses.
h Malignancies at early deaths considered fatal
1 All tumors considered incidental
* Tumor incidence data from NTP (1998, 042076)
July 2010
5-23 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-7. Dose-response modeling summary for male mouse tumor sites
associated with inhalation exposure to chloroprene
TUMOR TYPE*
Power
Parameter
ca
BMR
POINT OF DEPARTUREb
UNIT
RISKd
(Hg/rn3)"1
COMPOSITE
UNIT
RISK6
(Hg/rn3)"1
Modeled from
bioassay (ppm)
Continuous, human
equivalent0 (jig/m3)
BMDL
BMD
BMDL
BMD
Lung: alveolar/
bronchiolar
adenoma or
carcinomaf
3.4
0.1
1.86
2.46
1.20 x 103
4.92 x 103
1.59 x 103
6.50 x 103
8.3 x 10-3
2.0 x 10~5
1.4 x Iff4
8.5 x 10~5
All organs:
hemangio-
sarcomas,
Hemangiomas8'11
13.2
0.1
3.34
5.28
2.15 x 103
3.40x 103
4.7 x 10~5
All organs:
hemangio-
sarcomas,
Hemangiomas8'1
3.9
0.1
5.34
7.75
3.44 x 103
4.99 x 103
2.9 x 10"5
Harderian gland:
adenoma or
carcinoma
5.6
0.1
10.5
16.7
6.74 x 103
1.08 x 104
1.5 x 10~5
Kidney: renal
tubule adenomas or
carcinomas
(extended and
standard
evaluations
combined)
6.1
0.1
16.5
26.7
1.06 x 104
1.72 x 104
9.4 x 10~6
Forestomach:
squamous cell
papilloma or
carcinoma
1.3
0.05
22.8
45.1
1.47 x 104
2.91 x 104
3.4 x 10~6
a Multistage-Weibull model: P(d) = 1 - exp[-(b0 + bid + b2d2 + ... + bkdk) x (t-t0)c], coefficients estimated in terms
of ppm as administered in bioassay; lower stage bi not listed were estimated to be zero.
b BMD = Concentration at specified extra risk;
BMDL = 95% lower bound on concentration at specified extra risk.
0 Continuous equivalent estimated by multiplying exposures by (6 hours)/(24 hours) x (5 days)/(7 days).
d Unit risk estimated by dividing the BMR by the BMDL.
e Composite unit risk estimate, across all sites listed; see text for method.
f Values in italics indicate BMD/BMDL when lung tumors are treated as systemic lesions.
8 Highest exposure group dropped in order to better characterize low-dose responses.
h Malignancies at early deaths considered fatal
1 All tumors considered incidental
* Tumor incidence data from NTP (1998, 042076)
1 Following the recommendations of the NRC (1994, 006424) and the current Guidelines for
2 Carcinogen Risk Assessment (U.S. EPA, 2005, 086237) to consider total risk, an upper bound on the
3 composite risk was estimated for all tumor sites in female and male B6C3F1 mice. Note that this
4 upper bound estimate of composite risk describes the risk of developing any combination of the tumor
5 types considered, not just the risk of developing all simultaneously. Statistical methods which can
6 accommodate the underlying distribution of slope factors are optimal, such as through maximum
July 2010
5-24 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
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. Consequently, this analysis used the
same method as in several previous IRIS assessments (e.g., 1,3-butadiene (U.S. EPA, 2002, 052153),
1,2-dibromoethane (U.S. EPA, 2004, 594429), 1,2,3-trichloropropane (2009, 625769)) which involves
assuming asymptotic normality for slope factors. Each 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 BMDs and BMDLs were used to extrapolate to a
lower level of risk (R) where the BMDs and BMDLs were in a linear range. For these data a
_<2
10 risk 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 was also considered and is described below with the
results. The unit risk for each site was then estimated by R/BMDLr, as for the estimates for
each tumor site above.
• The central tendency estimates of unit potency (that is, risk per unit of exposure) at each
BMDr, estimated by R/BMDr, were summed across the sites listed in Table 5-6 for male mice
and similarly across the sites for female mice listed in Table 5-7.
• An estimate of the 95% upper bound on the composite unit risk 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% upper confidence limit (UCL) according to the
following formula:
95% UCL = MLE + 1.645 x SD
(1)
rearranged to:
SD = (UCL - MLE)/1.645
(2)
July 2010
5-25 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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 = SD ) 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 = variance12).
The resulting composite unit risk for all tumor types for female mice was 2.7 x 10"4 per (J,g/m3
(with lung tumors treated as a systemic effect). 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
"3
per (j,g/m 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" risk or near
the PODs.
For male mice the composite unit risk for all tumor types was 1.4 x 10"4 per (J,g/m3 (with lung
tumors treated as a systemic lesion), a 1.7-fold increase compared to the highest unit risk for any
5 3
individual tumor type, 8.3 x 10" per (J,g/m 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" risk or near the PODs.
For estimates in both species, if the lung tumors are primarily site of contact lesions, the
estimated composite risk decreases to 1.5 x 10"4 per (j,g/m3 (females) and 8.5 x 10"5 per (j,g/m3 (males).
Based on the relatively high fat:air partition coefficients (see Section 3.2.) in rodents and humans,
chloroprene is likely to be absorbed rapidly (U.S. EPA, 1994, 006488), consistent with the possibility
that the lung tumors are both portal-of-entry and systemic lesions.
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 (J,g/m3, rounding the
composite risk for female mice above to one significant digit. This unit risk should not be used with
3 3
continuous lifetime exposures greater than 600 (J,g/m (0.6 mg/m ), 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.
July 2010
5-26 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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 (see Section 4.7.3.1), 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 |ig/m3, calculated from data for adult exposures, does not reflect presumed
early-life 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 3-fold adjustments in slope factor are to be
combined with age specific exposure estimates when estimating cancer risks from early life (<16 years
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 (J,g/m3 x 10 x 2yr/70yr = 8.6 x 10"5 per (J,g/m3
Risk for ages 2 through < 16 = 3 x 10^ per (J,g/m3 x 3 x 14yr/70yr = 1.8 x 10"4per (J,g/m3
Risk for ages 16 until 70 = 3 x 10~4 per (J,g/m3 x l x 54yr/70yr = 2.3 x 10"4per (J,g/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= 5.0 x 10"4per ^g/m3
Using the above full lifetime unit risk estimate of 5 x 10"4 per (J,g/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 10"6 -h 5 x 10"4 per (J,g/m3 = 0.002 (J,g/m3
5.4.6. Previous Cancer Assessment
The carcinogenicity of chloroprene has not been evaluated previously for the IRIS program.
July 2010
5-27 DRAFT - DO NOT CITE OR QUOTE
-------
1 5.4.7. Uncertainties in Cancer Risk Values
2 A number of uncertainties underlie the cancer unit risk for chloroprene. These are discussed in
3 the following paragraphs. Specifically addressed is the impact on the assessment of issues such as the
4 use of models and extrapolation approaches, the use of other bioassay data, and the choices made and
5 the data gaps identified. In addition, the use of assumptions, particularly those underlying the
6 Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005, 086237) is explained and the decision
7 concerning the preferred approach is given and justified. Principal uncertainties are discussed below
8 and summarized in Table 5-8.
9 Table 5-8. Summary of uncertainties in chloroprene cancer unit risk estimate
Consideration
Potential Impact3
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 MOA
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 study
Standard design, well conducted,
extensively peer reviewed; carcinogenic
response consistently observed across all 4
species/sex combinations.
Species /gender
combination
Human risk could j
or 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 gender (female
mouse), based on PODhec It was assumed
that humans are as sensitive as the most
sensitive rodent gender/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.
July 2010
5-28 DRAFT - DO NOT CITE OR QUOTE
-------
Consideration
Potential Impact3
Decision
Justification
Cross-species
extrapolation
Alternatives for
lung tumors differ
by 4-fold: human
risk for any site
could | or Low-
dose risk would {
approximately 40%
if lung tumors were
treated as portal-of-
entry effects
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.
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
Statistical
uncertainty at
POD
i risk per unit
concentration 1.2-
fold if BMDio used
rather than BMDLio
BMCL (default approach
for calculating plausible
upper bound)
Limited size of bioassay results in sampling
variability; lower bound is 95% confidence
interval on concentration.
a t = increase, j = decrease
1 Human population variability. The extent of inter-individual variability in chloroprene
2 metabolism has not been characterized. A separate issue is that the human variability in response to
3 chloroprene is also poorly understood. The effect of metabolic variation, including potential
4 implications for differential toxicity, has not been well studied. Although a mutagenic MOA indicates
5 increased early-life susceptibility, there are no data exploring whether there is differential sensitivity to
6 chloroprene carcinogenicity across human life stages. This lack of understanding about potential
7 differences in metabolism and susceptibility across exposed human populations thus represents a
8 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 two-fold 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 are produced in proportion to low-exposures
of chloroprene, the derived unit risk is an unbiased estimate.
Choice of bioassay/species/gender. 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. The calculated composite unit risk is based on
the most sensitive endpoint (risk of any tumor type) in the most sensitive species and gender (female
July 2010 5-29 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
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 (see 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 Point of Departure (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 1.2-fold for lung tumors and for the composite risk estimates.
HEC derivation. A source of uncertainty in the derivation of the HEC comes from whether or
nor chloroprene induces lung tumors due to portal-of-entry or systemic effects. Systemic distribution
of chloroprene is evidenced by the induction of tumors in multiple organs and suggests that
chloroprene may be redistributed back to the lungs and may potentially act as a systemically delivered
carcinogen rather than, or in addition to, a portal-of-entry toxicant. However, the contribution of either
route of delivery (i.e., inhalation vs. 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).
July 2010
5-30 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF HAZARD AND DOSE RESPONSE
6.1 HUMAN HAZARD POTENTIAL
Chloroprene (C4H5CI, 2-chloro-1,3-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 futher 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 gender 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 eparenchymal 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
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
"3
chloroprene associated with these effects ranged from 1-7 mg/m .
July 2010
6-1
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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
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
associate 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
polychloroprene latex conducted in eight cohorts from the United States, 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.
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, mesentery,
July 2010
6-2
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Zymbal's gland. All of these tumor sites showed statistically significantly positive trends with
increasing exposure level (Cochran-Armitage test for trend p < 0.05, most with p < 0.001). 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., 2001,
624922; Sills et al., 1999, 624952).
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 vs. 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
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
July 2010
6-3
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
MOA 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 genotoxocity 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.
A range of portal-of-entry and systemic effects from the NTP study (1998, 042076). including
alveolar epithelial hyperplasia, bronchiolar hyperplasia, pulmonary histiocytic cell 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 respiratory and systemic effect (see Table 5-2). Splenic hematopoietic cell
"3
proliferation in female mice resulted in the lowest PODadj value of approximately 1.1 mg/m . This
POD was then converted into the PODhec by application of the dosimetric adjustment factor (DAF) for
a systemic effect. 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 1 x 10 2 mg/m3.
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 and mesentery, 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
July 2010 6-4 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
concentration [BMDio]) and its lower-bound BMDLio for each selected chloroprene-induced tumor
(see Tables 5-6 and 5-7). Lung tumors, treated as a systemic lesion (see Section 5.4.3 and 5.4.7 for
details), in female mice resulted in the highest inhalation unit risk (1.8 x 10"4 per (J,g/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 |ig/m3 (when lung tumors were considered systemic lesions). 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 (J,g/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 |ig/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 non-neoplastic and neoplastic lesions. In
addition, the key histopathological lesions observed are appropriately described, and suitable statistical
analysis is 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 day 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.
July 2010
6-5
DRAFT - DO NOT CITE OR QUOTE
-------
7. REFERENCES
Acquavella IF; Leonard RC (2001). A review of the epidemiology of 1,3-butadiene and chloroprene. Chem Biol Interact,
135-136: 43-52. doi: 10.1016/S0009-2797(01)00169-7 628495
Appelman LM; Dreef-van der Meulen HC (1979). Reproduction study with beta-chloroprene vapour in rats. Central
Institute for Nutrition and Food Research (CIVO). Zeist, The Netherlands. 6225. 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. Arch Toxicol, 41: 249-277. 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. 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. 625103
Bukowski JA (2009). Epidemiologic evidence for chloroprene carcinogenicity: Review of study quality and its application
to risk assessment. Risk Anal, 29: 1203-1216. doi: 10.111 l/i.l539-6924.2009.01254.x 628496
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. 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. 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. 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 Stat Meth, 34: 187-220. 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.
202141
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
Dourson ML; Knauf LA; Swartout JC (1992). On reference dose (RfD) and its underlying toxicity data base. Toxicol Ind
Health, 8: 171-189. 004400
July 2010
7-2
DRAFT - DO NOT CITE OR QUOTE
-------
Drevon C; Kuroki T (1979). Mutagenicity of vinyl chloride, vinylidene chloride and chloroprene in V79 Chinese hamster
cells. DNA Repair (Amst), 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. 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. 625114
Esmen NA; Kennedy KJ; Hall TA; Phillips ML; Marsh GM (2007). Classification of worker exposures. Chem Biol
Interact, 166: 245-253. 625121
Fajen JM; Ungers LJ (1986). Industrial hygiene walk-through survey report of E.I. duPont deNemours and Companry.
National Institute for Occupational Safety and Health; Centers for Disease Control. Cincinnati, OH. 628500
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. 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. 625124
Gooch JJ; Hawn WF (1981). Biochemical and hematological evaluation of chloroprene workers. J Occup Environ Med, 23:
268-272. 064944
Griebel CP; Halvorsen J; Golemon TB; Day AA (2005). Management of spontaneous abortion. Am Fam Physician, 72:
1243-1250. 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-l, 3-butadiene)-what is the evidence for its carcinogenicity? Clin Toxicol, 13:
153-170. 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. 625243
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. 479853
Hill AB (1965). The environment and disease: Association or causation. J R Soc Med, 58: 295-300. 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-
37.625154
July 2010
7-3
DRAFT - DO NOT CITE OR QUOTE
-------
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. 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: 267-284. 019012
Himmelstein MW; Gladnick NL; Donner EM; Snyder RD; Valentine R (2001). In vitro genotoxicity testing of (1-
chloroethenyl)oxirane, a metabolite of "beta"-chloroprene. Chem Biol Interact, 135: 703-713. 019013
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.
625159
IARC (1999). Re-evaluation of some organic chemicals, hydrazine, and hydrogen peroxide. , 71: 1-1586. 201838
Immels H; Willems M (1978). Initial submission: Dominant lethal assay with beta-chloroprene in male mice with cover
letter dated 09/01/92. Central Institute for Nutrition and Food Research (CIVO). The Netherlands. EPA# 88-920008417;
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.
625160
Jones JH; Young RJ; Selevan S; NIOSH (1975). Walk-through survey of E.I. du Pont de Nemours and Company, Inc.
Louisville, Kentucky. Division of Field Studies and Clinical Investigations. Cincinnati, OH. IWS-87-10; PB89-129845.
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. Toxicol Sci, 3: 140-160. 003194
Leet TL; Selevan SG (1982). Mortality analysis of workers exposed to chloroprene. 094970
Lefaux R (1968). Chloroprene. In 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. 625179
Li SQ; Dong QN; Liu YQ; Liu YG (1989). Epidemiologic study of cancer mortality among chloroprene workers. Biomed
Environ Sci, 2: 141-149. 625181
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. 084126
Lynch M (2001). Manufacture and use of chloroprene monomer. Chem Biol Interact, 135-136: 155-167. 625182
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.
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. 625187
July 2010
7-4
DRAFT - DO NOT CITE OR QUOTE
-------
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. 625188
MARTA ; MTA (1996). Historical control data (1992-1994) for developmental and reproductive toxicity studies using the
Crl:CD (SD)BR rat. Charles River Laboratories. Malvern,
PA.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. U.S. Department of Health and Human
Services. Research Triangle Park, NC. PNL-7458; DE94012384. 625206
McGlothlin JD; Meyer C; Leet TL (1984). Health hazard evaluation report HETA-79-027: E.I. du Pont de Nemours and
Company, Inc. Louisville, Kentucky. National Institute for Occupational Safety and Health. Cincinnati, OH. HETA 79-
027-1459; PB85-186757 . 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) inF344rats and B6C3F(1) mice. Toxicology, 108: 79-91. 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. Cancer Res, 54:
5333-5339. 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(1-3):323-331]. Chem Biol Interact, 168: 169. 625213
Munter T; Cottrell L; Ghai R; Golding BT; Watson WP (2007). The metabolism and molecular toxicology of chloroprene.
Chem Biol Interact, 166: 323-331. 576501
Munter T; Cottrell L; Golding BT; Watson WP (2003). Detoxication pathways involving glutathione and epoxide hydrolase
in the in vitroin vitro metabolism of chloroprene. Chem Res Toxicol, 16: 1287-1297. 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. 625215
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. doi:10.1016/j.mrfmmm.2008.02.007 628501
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. 196814
NLM (2010). Hazardous Substances Data Bank (HSDB). Retrieved 12-FEB-10, from http://toxnet.nlm.nih.gov/cgi-
bin/sis/htmlgen?HSDB. 594343
July 2010
7-5
DRAFT - DO NOT CITE OR QUOTE
-------
NRC (1994). Science and judgment in risk assessment. National Research Council. Washington, DC. 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). Public Health Service, U.S. Department of Health and Human Services.
Washington, DC. 625205
NTP (1998). Toxicology and carcinogenesis studies of chloroprene (CAS No 126-99-8) in F344/N rats and B6C3F1 mice
(inhalation studies). National Toxicology Program. Research Triangle Park, NC. NTP TR 467; NIH PUB 98-3957.
https://ntrl.ntis.gov/search/TRLProductDetail.aspx?ABBR=PB99123671. 042076
NTP (2005). Report on carcinogens. National Toxicology Program. 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 (Organisation for Economic Co-operation and Development) (1998). Screening Information Data Set (SIDS) high
production volume chemicals. United Nations Environment Programme (UNEP) Publications. Nairobi,
Kenya.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. 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
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. , 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. Annu Rev
Pharmacol Toxicol, 25: 567-592. 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
July 2010 7-6 DRAFT - DO NOT CITE OR QUOTE
-------
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. ChemBiol Interact, 135-136: 376-386. 624922
Sills RC; Hong HL; Melnick RL; Boorman GA; Devereux TR (1999). High frequency of codon 61 K-rasras 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 (Ed.),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. 064961
Tarone RE (1975). Tests for trend in life table analysis . Biometrika, 62: 679-690. 624959
Thornton-Manning JR; Dahl AR (1997). Metabolic capacity of nasal tissue: Interspecies comparisons of xenobiotic-
metabolizing enzymes. Mutat Res, 380: 43-59. 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 inlung and brain tumors from B6C3F1 mice exposed to 1,3-butadiene or chloroprene. Chem Biol Interact, 106: 112-
120. 625004
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. 625008
U.S. EPA (1989). Health and environmental effects document for 2-chloro-1,3,-butadiene (chloroprene). U.S. EPA.
Cincinnati, OH. 625024
U.S. EPA (2000). Chloroprene (2-chloro-l,3-butadiene). U.S. EPA. Washington,
DC.http://www.epa. gov/ttn/atw/hlthef/chloropr.html. 625036
U.S. EPA (1985). Summary review of health effects associated with chloroprene: health issue assessment. 017624
U.S. EPA (1986). Guidelines for mutagenicity risk assessment. Risk Assessment Forum, U.S. Enviornmental Protection
Agency. Washington, DC. EPA/630/R-98/003. 001466
U.S. EPA (1986). Guidelines for the health risk assessment of chemical mixtures. Risk Assessment Forum, U.S.
Environmental Protection Agency. Washington, DC. EPA/630/R-98/002. 001468
U.S. EPA (1988). Recommendations for and documentation of biological values for use in risk assessment. Environmental
Criteria and Assessment Office, Office of Health and Environmental Assessment, Office of Research and Development,
U.S. Environmental Protection Agency. Cincinnati, OH. EPA/600/6-87/008.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=34855. 064560
July 2010
7-7
DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA (1991). Guidelines for developmental toxicity risk assessment. U.S. Environmental Protection Agency, Risk
Assessment Forum. Washington, DC. EPA/600/FR-91/001. 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. U.S.
Environmental Protection Agency, Health Effects Division. Washington,
DC.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.
Environmental Criteria and Assessment Office, Office of Health and Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency. Research Triangle Park, NC. EPA/600/8-90/066F.
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. Risk Assessment Forum, U.S.
Environmental Protection Agency. Washington, DC. EPA/630/R-94/007.
http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=42601. 005992
U.S. EPA (1996). Guidelines for reproductive toxicity risk assessment. Risk Assessment Forum, U.S. Environmental
Protection Agency. Washington, DC. EPA/630/R-96/009. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2838.
030019
U.S. EPA (1998). Guidelines for neurotoxicity risk assessment. National Center for Environmental Assessment; Office of
Research and Development; U.S. Environmental Protection Agency. Washington, DC. EPA/630/R-95/001F.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid= 12479. 030021
U.S. EPA (2000). Benchmark dose technical guidance document [external review draft]. Risk Assessment Forum, U.S.
Environmental Protection Agency. Washington, DC. EPA/630/R-00/001. http://www.epa.gov/raf/publications/benchmark-
dose-doc-draft.htm. 052150
U.S. EPA (2000). Science policy council handbook: Risk characterization. Office of Science Policy, Office of Research and
Development, U.S. Environmental Protection Agency. Washington, D.C.. EPA 100-B-00-002.
www. epa. gov/spc/pdfs/rchandbk.pdf. 052149
U.S. EPA (2000). Supplemental guidance for conducting for health risk assessment of chemical mixtures. U.S. EPA.
Washington, DC. 196144
U.S. EPA (2000). Toxicological review of vinyl chloride. U.S. EPA. Washington, DC. EPA/635R-00/004.
http://www.epa.gov/iris. 194536
U.S. EPA (2002). A review of the reference dose and reference concentration processes. Risk Assessment Forum, U.S.
Environmental Protection Agency. Washington, DC. EPA/630/P-02/0002F.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=55365. 088824
U.S. EPA (2002). Health assessment of 1,3-butadiene. U.S. Environmental Protection Agency. Washington, D.C. .
EPA/600/P-98/001F. 052153
U.S. EPA (2004). Toxicological review of 1,2-dibromoethane (CASRN 106-93-4) in support of summary information on
the Integrated Risk Information System (IRIS). U.S. Environmental Protection Agency. Washington,
D.C..http://www.epa.gov/iris/subst/0361 .htm. 594429
U.S. EPA (2005). Guidelines for carcinogen risk assessment, Final Report. Risk Assessment Forum, U.S. Environmental
Protection Agency. Washington, DC. EPA/630/P-03/001F. 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. Risk
Assessment Forum, U.S. Environmental Protection Agency. Washington, DC. EPA/630/R-03/003F.
July 2010
7-8
DRAFT - DO NOT CITE OR QUOTE
-------
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l 16283. 088823
U.S. EPA (2006). A framework for assessing health risk of environmental exposures to children (final). National Center for
Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency. Washington,
DC. EPA/600/R-05/093F. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=158363. 194567
U.S. EPA (2006). Science Policy Council Handbook: Peer Review. Office of Science Policy, Office of Research and
Development, U.S. Environmental Protection Agency. Washington, DC. EPA/100/B-06/002.
http://www.epa.gov/peerreview/pdfs/peer_review_handbook_2006.pdf. 194566
U.S. EPA (2009). Status report: Advances in inhalation dosimetry of gases and vapors with portal of entry effects in the
upper respiratory tract. U.S. Environmental Protection Agency. Washington, DC. EPA/600/R-09/072.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=212131 . 625038
U.S. EPA (2009). Toxicological review OF 1,2,3-trichloropropane (CASRN 96-18-4): In support of summary information
on the Integrated Risk Information System (IRIS). U.S. Environmental Protection Agency. Washington,
DC.http://www.epa.gov/iris/subst/0200.htm. 625769
Vogel E (1979). Mutagenicity of chloroprene, l-chloro-l,3-trans-butadiene, 1,4-dichlorobutene-2 and l,4-dichloro-2,3-
epoxybutane in drosophila melanogaster. DNA Repair (Amst), 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; Muller 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 dimmers in the Salmonella/microsome mutagenicity test.
Central Institute for Nutrition and Food Research (CIVO). Zeist, The Netherlands. R 5712. 625048
Willems MI (1980). Evaluation of beta-chloroprene and four chloroprene dimers in the Ames test by atmospheric exposure
of the tester strains. Central Institute for Nutrition and Food Research (CIVO). Zeist, The Netherlands. R 6392;
OTS0570703; 88-920008421. https://ntrl.ntis.gov/search/TRLProductDetail.aspx?ABBR=OTS0570703. 625049
July 2010
7-9
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS
AND DISPOSITION
The Toxicological Review of chloroprene has undergone a formal external peer review
performed by scientists in accordance with EPA guidance on peer review (U.S. EPA, 2006, 194566).
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 were included in a
separate section of this appendix. There were six external peer reviewers.
EXTERNAL PEER REVIEW PANEL COMMENTS
The reviewers made several editorial suggestions to clarify specific portions of the text. These
changes were incorporated in the document as appropriate and are not discussed further.
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.
General Charge Questions:
1. Is the Toxicological Review logical, clear and concise? Has EPA clearly synthesized the scientific
evidence for noncancer and cancer hazards?
Comment:
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
July 2010
A-1
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
genotoxicity data, but this 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 non-cancer 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 (see 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 (see Section 5.2.2). Specifically, the criteria that were used to determine adequacy of model fit
(global goodness-of-fit p-value, % 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 (see 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 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.
July 2010 A-2 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
2. Please identify any additional studies that should be considered in the assessment of the noncancer
and cancer health effects of chloroprene.
Comment:
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 studies of DuPont plants involved in chloroprene production (Fajen and
Ungers, 1986, 628500; Jones et al., 1975, 625203; McGlothlin et al., 1984, 625204) should be
included in Toxicological review. These walk-through studies 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 vs. 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 studies were added to the discussion of
human health effects of chloroprene exposure (Jones et al., 1975, 625203; McGlothlin et al., 1984,
625204; see Section 4.1.2.1). 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 report was not included in the
Toxicological Review as it primarily dealt with butadiene air monitoring (Fajen and Ungers, 1986,
628500). The two additional health studies were included in the document (Gooch and Hawn,
1981, 064944; Leonard et al., 2007, 625179; see Sections 4.1.1.2 and 4.1.2.1). 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
July 2010 A-3 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 that was included in the assessment. 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; see Section 3.5). 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- vs. water-based adhesives and included multiple coexposures, 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.
Chemical-Specific Charge Questions:
(A) Oral Reference Dose (RfD) for Chloroprene
1. 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:
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
July 2010 A-4 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Review panel, 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.
B) Inhalation Reference Concentration (RfC) for Chloroprene
1. A chronic RfC for chloroprene has been derived from an inhalation toxicity study (NTP, 1998,
042076) investigating non-cancer effects in multiple organ systems. Please comment on 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:
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 non-cancer effects in multiple organ systems in rats and mice exposed
to a wide range of chloroprene. Two reviewers noted that not chosing 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)) 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 (see Section 5.2.1) clarifying the reasons
Trochimowicz et al. (1998, 625008) was not selected as the principal study. Discussion of both the
Gootch and Hawn (1981, 064944) and McGlothlin (1984, 625204) studies was added to the
appropriate sections of Section 4, 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 contained text stating "no human studies are available that would allow for
the quantification of sub-chronic or chronic non-cancer effects".
2. 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
July 2010 A-5 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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:
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
atrophy occurred first. Lastly, the reviewer commented that nasal lesions should not be selected as
the critical effect due to the way the HECs were calculated (see comments below in question B3).
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
3 3
PODhec values were 1.1 mg/m for atrophy and 1.0 mg/m 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 rewritten significantly in response to reviewer comments
regarding question B3 (see below). Specific comments regarding the combination of nasal
olfactory atrophy and necrosis (e.g., poorly explained rationale, incorrect conclusion that atrophy
preceeds necrosis, and the negligible effect combining the lesions has on the PODhec values) were
no longer relevant as the combination of nasal lesions was not performend 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 (see Section 5.2.1). PODs for these
July 2010 A-6 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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.
Figure 5-1 was removed from the document due to the extensive rewriting of Section 5.2 (see
below). Results for statistical tests of trend were not included for non-cancer 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 relates to how chloroprene, as a water insoluble and non-reactive gas, can exert
portal-of-entry effects in the upper and lower respiratory tract.
3. Benchmark dose (BMD) modeling was used to define the point of departure (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
EPA's approach.
Comment:
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
July 2010 A-7 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 be clearer in regard to the derivation of the RfC, and suggested that
EPA provide a clearer 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. One reviewer objected strongly to the approach used to
derive the PODhecs for a number of reasons. First, this reviewer stated that the PODs used in the
-3
calculation of the HECs are very similar (2.1 - 8.3 mg/m ) 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. 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 vs. systemic redistribution
was discussed for cancer effects, but not for non-cancer 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., PODAdj) 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 current 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 under the assumption that it represented a
minimal biologically significant change. 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
July 2010 A-8 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
well as incidence with dose (see Table B-l). In the case of endpoints that progressed in incidence
as well as severity (i.e., progression from mild to moderate lesions) from the control dose to the
lowest dose showing response, a BMR of 10% was not considerd to be a biologically minimal
effect. Therefore, for these endpoints, a BMR of 5% was used for the estimation of BMDs and
BMDLs (see 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 was included in the text (see Section 5.2.2). Specifically,
the criteria that were used to determine adequacy of model fit (global goodness-of-fit p-value, %
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). EPA has also added step-by-step
calculations of the PODadj 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 adverse respiratory and systemic 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 non-reactive, and would be expected to only elicit extrarespiratory effects. Medinsky
and Bond (2001, 016157) described an alternative method of classification that represents the
reactivity and water solubility as continuous variables that allows insight to be gained regarding
sites of and conditions of toxicity. Additionally, although not explicitly included in this scheme, in
situ metabolism in the respiratory tract can be considered as a component of reactivity. In this
manner, chloroprene can best be described as both a non-reactive, water insoluble chemical that is
absorbed into the blood stream and induces systemic toxicity, as well as a chemical that is
metabolized into a reactive epoxide within the respiratory tract, inducing portal-of-entry toxicity.
This method of 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, which is expressed in both the olfactory and pulmonary regions of the respiratory tract.
In situ metabolism in the respiratory tract may thus explain a portion of the biological activity of
chloroprene in these regions. Alternatively, it is observed that chloroprene induced adverse effects
in organ systems distal to the portal-of-entry, consistent with its water insoluble and non-reactive
chemical properties. Therefore, it is possible that observed pulmonary effects, such as alveolar
hyperplasia in the female rat, may be explained by blood-borne delivery, rather than air-borne.
Currently, the contribution of either mode of delivery to the sites of observed toxicity is not known.
July 2010
A-9
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
This discussion of the uncertainty surrounding the mode of delivery to respiratory tissues, included
in Section 5.3, harmonized the non-cancer and cancer derivations.
In order to minimize uncertainty surrounding the application of the default DAFs, the selection
of the critical effect was based on the observed experimental data (i.e., the PODadj) and DAFs
were then applied to the selected critical effect to calculate the HEC.
Given the above changes to the modeling scheme, increased incidence of splenic hematopoietic
proliferation in female mice was chosen as the critical effect based on the observation that this
3 3
endpoint had the lowest PODadj (1.1 mg/m , based on a BMDL05 of 1.7 ppm [6.2 mg/m ]) (see
Section 5.2.3). Using a DAF of 1 (for systemic effects), the calculated PODHec for increased
-3
splenic hematopoietic proliferation was also 1.1 mg/m .
4. 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:
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 3-fold 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 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 BMDL10 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
(see 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 F0 and Fi rats were exposed to chloroprene. However, this study did not involve
the mating of the Fi generation, so developmental effects to the F2 generation could not be
July 2010 A-10 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 including 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 5.2.3 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, and
further, was not needed as a BMR of 5% was ultimately used in the document for derivation of the
RfC based on increased incidence of splenic hematopoietic proliferation in female mice.
(C) Carcinogenicity of Chloroprene
1. Under the EPA's 2005 Guidelines for Carcinogen Risk Assessment (www.epa.gov/iris/backgr-
d.htm), 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:
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 mutagenic metabolites of chloroprene and analogies to related chemical carcinogens with
analogous metabolic pathways that made the determination of "like 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 non-reactive and insoluble in water.
July 2010 A-11 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 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. (Marsh
et al., 2007, 625188), failed to observed an increased risk.
July 2010
A-12
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 (see 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 toxicity studies involving dermal exposure, carcinogenicity by
this route of exposure may be inferred as there is no convincing toxicokinetic data to preclude
absorbtion by this route of exposure, and that rapid absorption of chloroprene through the skin
occurs (NLM, 2010, 594343).
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 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
July 2010 A-13 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 (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 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 warrants 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 reported expected counts for liver and lung cancer mortality in Li et al. (1989,
625181), Bulbulyan et al. (1998, 625105), and Bulbulyan et al. (1999, 157419).
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 that need
to be considered carefully exist in these studies, EPA 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
five-fold increased 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 on individual workers' 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).
These limitations are further discussed in Section 4.7.1.1.
July 2010
A-14
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
2. A two-year inhalation cancer bioassay in B6C3F1 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:
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) was not chosen for selection as the principal
study; the high mortality in the low dose group was identifed as the main reason for not selecting
the study as the principal study (see 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 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 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
July 2010
A-15
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
appropriate species for the inhalation unit risk derivation based on the observation that it was more
sensitive to the carcinogenic effects of chloroprene exposure.
3. A mutagenic 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:
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-ra.s 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 was 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.
4. 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 B6C3F1 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 EPA's approach.
Comment:
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
July 2010 A-16 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
model structure, and provided an extensive example using the mice data. The other reviewers did
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 get one tumor
vs. 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 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 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 (see Sections 5.4.3 and 5.4.4). The suggested alternative
modeling approach incorporating saturating metabolism was a constructive approach that EPA will
consider with regards to future methods developed for human health risk assessment. However, as
noted by the reviewers this model did not currently incorporate time-to-tumor information, and it
appeared complex to do so and was beyond the scope of the Toxicological Review. Also, the
saturating metabolism parameters were not derived from pharmacokinetic data but from empirical
fits to dose and tumor incidence data, so it was as much an empirical model as the multistage-
Weibull. 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. The
multistage-Weibull model did adequately fit the monotonic, supralinear dose-response
relationships seen in the NTP study; EPA retains the analysis in the assessment..
July 2010
A-17
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Additional mouse and human metabolic and toxicokinetic data (Himmelstein et al., 2004,
625152; Himmelstein et al., 2004, 625154) added the document indicated that the metabolic
differences between humans and mice are not as great as previously represented in the document
(see Section 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 was 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
[j,g/m ) was applied to the median cumulative exposure for the Louisville plant, converted to a
3 3
lifetime equivalent continuous concentration (18.35 ppm-yr/70 yr x 3.62 x 10 ((J,g/m )/ppm ~ 950
-3
[j,g/m ), 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 addittional epidemiological studies that observed significant
July 2010 A-18 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
associations between chloroprene exposure and cancer mortalitiy (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.
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.
5. 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:
Four reviewers agreed that alternatively treating lung tumors as portal-of-entry or systemic
effects was appropriate given 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 comments 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 non-reactive 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 active metabolites are formed in and then escape the liver.
Response:
The current modeling approach of treating observed lung tumors as either portal-of-entry or
systemic lesions was maintained in the Toxicological review. Additional discussion regarding the
justification for and application of this approach as it relates to the observed pattern of adverse
respiratory and systemic effects was added to the document (see Sections 5.2.3 and 5.4.3; see
response to question 3 comments above as well). Chloroprene is a water insoluble, non-reactive
July 2010 A-19 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
chemical, and is expected to be absorbed into the bloodstream deep in the respiratory tract and
exert its toxic effect systemically. Indeed, multiple adverse effects were observed distal to the
respiratory tract that supports this assumption. However, lung tissue fractions have been shown to
actively produce the reactive epoxide metabolite of chloroprene when exposed. Therefore, toxic
effects observed in the respiratory tract of exposed animals may be due to either portal-of-entry
effects due to in situ chloroprene metabolism or due to systemic redistribution of chloroprene back
to the respiratory tissues. This additional discussion is consistent with the final 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.
6. 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:
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 (see 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 assessement, was included
in Section 3.3.
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.
July 2010
A-20
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Comment: 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
(see Sections 3.2 and 3.4).
Comment: 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 fat: air 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 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 a further detoxification pathway, further oxidation of (1-
chloroethenyl)oxirane, active in mice, but not in humans, rats, or hamsters. A discussion of fat:air
partition coefficients and body fat percentage was added to the document.
Comment: The text in Section 3.3 should precisely indicate how the estimates for Vmax/Kmj 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.
July 2010
A-21
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Comment: 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.
Comment: 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 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: 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/Kmfor various organs.
Response: The data presented in Table 3-4 is how the data was presented by the authors in the original
reference. 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/Km in other organ systems.
Comment: 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: 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.
July 2010
A-22
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Response: The discussion of the healthy worker survivor effect was expanded in the document (see
Section 4.1.1.3).
Comment: SMRs and SIRs should consistently use basei or baseioo-
Response: The document was revised so that SMRs and SIRs consistently use baseioo throughout the
document.
Comment: It would be useful if as much 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 (see Section 4.1.1.2).
Comment: 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: 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 (see Section 4.7.1.1). Also, the text regarding how the
SMRs may be biased due to the potential non-representativeness of the available local data was
clarified.
Comment: 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?
July 2010 A-23 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Response: No such data were available on how many workers left the study area prior to 1979.
Comment: 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.
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: 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 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 comparions are based on internal populations or other regional/national DuPont
workers.
Comment: 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).
July 2010
A-24
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Comment: 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.
Comment: For later modeling, EPA should report integrated average exposures that were measured,
rather than the nominal target exposures. The difference be 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 (see 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 vs. 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: 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: 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?
July 2010
A-25
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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: Information should be included in the document on how the survival-adjusted neoplasm
rates reported in Table 4-28 were calculated.
Response: Text was added as a footnote in Table 4-28 detailing how survival-adjusted neoplasm rates
were calculated.
Comment: 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 (see Section 4.3).
Comment: 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: 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: With regard to the comparison of carcinogenic potency of chloroprene vs. 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.
July 2010 A-26 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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: 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.
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: 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: 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
July 2010
A-27
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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: 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
results (RR range 2.9-7.1) detected in two other studies for high and intermediate cumulative
exposures (Bulbulyan et al., 1999, 157419; Bulbulyan et al., 1998, 625105)" 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: 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: 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 (see Table 4-11 on page 4-17).
Response: Alcohol cannot be a confounder if it is not both related to the exposure of interest
(chloroprene) and the outcome of interest (liver cancer). There was suggestive evidence of a dose-
response, or in consistent elevated risks in the upper exposure categories, in multiple studies
(Bulbulyan et al., 1999, 157419; Leet and Selevan, 1982, 094970; Bulbulyan et al., 1998, 625105;
Marsh et al., 2007, 625188)
July 2010
A-28
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Comment: What "current understanding" allows for the statement that specificity is "one of the
weaker Hill criteria [szc]"?
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 (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 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."
Comment: 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; this study was not described previously.
Response: This section was rewritten to include discussion of only tumors that were statistically
significantly increased in rats and 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: 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.
July 2010
A-29
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Comment: In Section 4.7.3.1, the document specifies a mutagenic MOA 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.
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: 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.
Comment: 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: 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: 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.
July 2010
A-30
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Response: This text was removed as atrophic and necrotic olfactory lesions were no longer combined
into one endpoint for the purposes of benchmark modeling.
Comment: 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: 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 non-
respiratory 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 RGDR 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 4-fold more
toxicant (1/0.28) than the rat. Is it meant to imply that the metabolic activation rate in the human nose
is 4-fold 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 (see Charge Question B3), the
application of the default DAFs was performed after the critical effect was chosen based on the
observed experimental data (i.e., the PODadj) Given this new basis for choosing the critical effect,
increases in splenic hematopoietic proliferation was chosen over any nasal or pulmonary effect. This
approach reduced the influence that application of a DAF has on the selection of a critical effect as this
effect is unequivocally systemic in nature.
July 2010
A-31
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Increased 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
that posits that the observed toxicity of chloroprene in the respiratory tract may be due systemic
redistribution of chloroprene.
Comment: 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. Appelman and Dreef van
der Meulen (1979, 064938) study is an unpublished report in which F0 and Fi rats were exposed to
chloroprene. However, this study did not involve the mating of the Fi generation, so developmental
effects to the F2 generation could not be assessed.
Comment: 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: 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: The suggested alternative modeling approach incorporating saturating metabolism was a
constructive approach that EPA will consider with regards to future methods developed for human
health risk assessment. However, as noted by the reviewers this model did not incorporate time-to-
tumor information and it was beyond the scope of the Toxicological Review to modify the model to do
so. Also, the saturating metabolism parameters were not derived from pharmacokinetic data but from
empirical fits to dose and tumor incidence data, so it was as much an empirical model as the
multistage-Weibull. 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.
The multistage-Weibull model did adequately fit the monotonic, supralinear dose-response
relationships seen in the NTP study; EPA retained the multistage-Weibull analysis in the assessment.
July 2010 A-32 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Comment: 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 (see Section 5.4.4).
PUBLIC COMMENTS
A. Interpretation of Epidemiological Studies
Comment: 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.
Response: The 2005 US EPA Guidelines 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 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: 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 superceded 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.
July 2010
A-33
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 1575 male employees (salaried and female
employees excluded due to "minimal or no potential exposure to chloroprene") who were
working at the Louisville plant on 6/30/1957. 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 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 warrant the independent analysis of each.
Comment: 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 for liver and lung cancer mortality.
Response: Although some cohorts did report very low expected counts 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))(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 expected number of deaths. This was demonstrated for outcomes
(e.g., liver cancer mortality) in many studies including several of the DuPont plants as were
statistical power limitations when examining cancer-specific effects and exposure-response
July 2010
A-34
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
relationships. Regardless of the study, the 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 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: 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: 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 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
(see 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: 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.
July 2010
A-35
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
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 explicitly 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 ofpersonal 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: 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 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: Vinyl chloride exposure as a potential confounder of the association with
chlorprene 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.
July 2010
A-36
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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.
B. Interpretation of Mode of Action Based on the Mutagenicity and Genotoxicity Data
Comment: 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
non-positive 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
neglible 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)), (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
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
July 2010
A-37
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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-ra.s mutations in rodents
were observed in isoprene-induced lung neoplasms and vinyl chloride-induced heptatocellular
carcinomas (NTP, 1998, 042076; U.S. EPA, 2000, 194536). Activated K-ras oncogenes were
observed in lung tumors, hepatocellular carcinomas, and lymphomas in B6C3F1 mice exposed
to 1,3-butadiene (U.S. EPA, 2002, 052153).
Comment: 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 non-
chloroprene origin.
Response: The Toxicological Review presented the bacterial genotoxicity data as returning
conflicting results, but did note that when positive results were observed they occured in
Salmonella strains that test for point mutations. Assays with Salmonella strains that tested for
frameshift mutations were consistently negative. A guanine adduct was the major adduct
observed (approximately 96% of adducts formed) when the epoxide metabolite of chloroprene
is reacted with calf thymus DNA in 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. 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: A non-genotoxic 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.
July 2010
A-38
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Response: The document 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 (see 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 A to T
transversions at codon 61 (22/37) is inconsistent with the proposed alternative mode of action.
If chloroprene exposure was initiating cytotoxicity with subsequent hyperplasia/regeneration
leading to promotion of pre-existing proto-oncogene mutations, the expectation would be that
no A to 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 unamimously accepted by the External Peer Review panel. 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).
C. Consideration of Species Differences in Toxicokinetics and Target Tissue Dosimetry
Comment: 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. 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 (1-
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 non-cancer and carcinogentic toxicity of
chloroprene in humans.
July 2010
A-39
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Comment: Previous analyses (Himmelstein et al., 2004, 625154) support the use of the
physiologically based pharmacokinetic (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.
Comment: 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.
D. US EPA Decision Points in the Determination of the Inhalation Unit Risk
Comment: 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 portal-of-entry and systemic 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
July 2010 A-40 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
the low- or mid-exposure range in the absence of statistical signficance, 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: 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, US EPA should reconsider
the application of an UF for database uncertainties due to the lack of a multigenerational study.
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 the Fi offspring, and any cumulative effects that may manifest
throughout multiple generations. Additionally, Dourson et al. (1992, 004400) demonstrated
that, when comparing ratios of chronic dog, rat, and mouse studies and
reproductive/developmental studies in rats, the reproductive/developmental studies were useful
in establishing the lowest NOAEL. If one or more bioassays are missing, Dourson et al. (1992,
004400) recommended that an uncertainty factor should be used to account for this uncertainty.
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: In the Draft Review, a proprietary software program (TOXRISK 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.
July 2010
A-41
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Comment: 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: Model selection (goodness-of-fit for arriving at final number of stages) was not
well characterized.
Response: A summary of the model selection decisions was added (see Section 5.4.3).
Comment: 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,
July 2010 A-42 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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: 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 non-cancer and carcinogentic toxicity of chloroprene
in humans.
Comment: 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 US 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 effects" (US EPA, 2000), 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: 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 non-cancer 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, a discussion of how assumptions regarding the
route-of-exposure can effect HECs was added. Therefore, the non-cancer and cancer
quantitative sections have been made consistent in this regard.
July 2010 A-43 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
E. US EPA's Quality Control in Reporting Chloroprene Data
Comment: 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
inadvertantly 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 time on 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 gland and
Harderian gland tumors, which were not evaluated histopathologically in all animals—
Denominators were corrected.
• 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 wass 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
July 2010 A-44 DRAFT - DO NOT CITE OR QUOTE
-------
1 Toxicological Review. The Sanotskii review does not serve as the basis for any quantitative
2 analysis, and only provides data and results that are qualitatively useful in comparison to other
3 study reports included in the Toxicological Review. The limitations of the Sanotskii review are
4 appropriately detailed when the paper is first referenced; it was not necessary to exhaustively
5 delineate the study limitations at every instance the paper is cited in the Toxicological Review.
6
7
8
9
July 2010
A-45
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
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, 2000b) using Benchmark Dose Software 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 critical 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 and female 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). A BMR of 10% extra risk was used for each
endpoint, unless severity scores (see Table B-l) indicated a progression of severity from the control
dose to the lowest dose showing response, in which case a BMR of 5% extra risk was used. A BMR of
5% extra risk was also used in cases involving endpoints with no control response but the presence of
moderate or marked lesions in the lowest dose showing response. 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), x 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 3-fold difference of one another) or the lowest
BMDL when they were not within 3-fold difference of each other.
The critical endpoint selected for the derivation of the chronic RfC was increased incidence of
splenic hematopoietic proliferation in female mice and alveolar hyperplasia in the female rat. The
probit model provided the best fit for this data set. The following tables (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-16) and output are included
immediately after the table.
July 2010
B-1
DO NOT CITE OR QUOTE
-------
1
Table B-l. Severity scores at control dose and lowest dose showing response for
endpoints considered for critical non-cancer effect
Endpoint
0 ppm
12.8 ppm
32 ppm
r
II
in
IV
I
II
III
IV
I
II
III
IV
Male Rats
Alveolar hyperplasia
3
2
0
0
10
5
1
0
b
--
--
--
Kidney hyperplasia
--
--
--
--
--
--
--
--
--
--
--
--
Olfactory atrophy
1
2
0
0
6
3
3
0
--
--
--
--
Olfactory basal cell
hyperplasia
0
0
0
0
0
0
0
0
18
18
2
0
Olfactory metaplasia
2
4
0
0
5
0
0
0
--
--
--
--
Olfactory necrosis
0
0
0
0
5
1
5
0
--
--
--
--
Olfactory chronic
inflammation
0
0
0
0
5
0
0
0
--
--
--
--
Female Rats
Alveolar hyperplasia
3
2
0
1
15
6
1
0
--
--
--
--
Kidney hyperplasia0
--
--
--
--
--
--
--
--
--
--
--
--
Olfactory atrophy
0
0
0
0
1
0
0
0
31
7
2
0
Olfactory basal cell
hyperplasia
0
0
0
0
0
0
0
0
16
1
0
0
Olfactory metaplasia
0
0
0
0
1
0
0
0
34
1
0
0
Olfactory necrosis
0
0
0
0
0
0
0
0
3
2
3
0
Male Mice
Bronchiolar hyperplasia
0
0
0
0
3
5
1
1
--
--
--
--
Kidney hyperplasia
--
--
--
--
--
--
--
--
--
--
--
--
Forestomach epithelial
hyperplasia
0
2
0
2
2
3
1
0
--
--
--
--
Splenic hematopoietic
cell proliferation
2
12
10
2
2
15
5
0
--
--
--
--
Female Mice
Bronchiolar hyperplasia
0
0
0
0
4
8
2
1
--
--
--
--
Forestomach epithelial
hyperplasia
1
2
0
1
0
0
2
1
--
--
--
--
Splenic hematopoietic
cell proliferation
0
8
4
1
3
13
6
3
--
--
--
--
a Severity scores -1 = minimal, II - mild, III - moderate, IV - marked
b Only severity scores in control dose and lowest dose with response used to make determination of severity progression
with increasing dose
0 Severity for single sections and step sections combined not available
Source: NTP (1998)
July 2010
B-2
DO NOT CITE OR QUOTE
-------
Table B-2. Benchmark modeling results for alvelolar epithelial hyperplasia in
male F344/N rats (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
231.042
0.1317
1.698
14.8657
10.0883
Logistic
232.34
0.0775
-0.092
24.4838
19.1571
Log-logistica
230.479
0.1753
1.566
11.4228
7.06934
Log-probit
233.859
0.0363
-0.087
28.604
19.5927
Multistage
231.042
0.1317
1.698
14.8657
10.0883
Probit
232.209
0.0813
-0.126
23.3986
18.2584
Weibull
231.042
0.1317
1.698
14.866
10.0883
Quantal-linear
231.042
0.1317
1.698
14.866
10.0883
Dichotomous hill
231.705
0.1112
-0.1356
5.87477
3.85444
a model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
0.7
Log-Logistic
0.6
0.5
0.4
0.3
0.2
BMDL
BMD
0
10
20
30
40
50
60
70
80
dose
14:17 01/14 2010
Figure B-l. Log-logistic model fit for alveolar epitheliala hyperplasia in male F344/N rats
(BMR = 10% extra risk)
July 2010
B-3
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
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-
Restri ct.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_a~lv_hyper_Ln~l-BMRlO-
Restri ct.pit
Thu Jan 14 14:17:54 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-i ntercept-slope*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
( *** 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
i ntercept
background intercept
1
-0.66
-0.66
1
Variable
background
i ntercept
slope
Parameter Estimates
Estimate
0.1B0984
-4.6B28B
1
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(1 i ke~li hood)
-111.57
-113.24
-121.815
# Param's
4
2
1
Deviance Test d.f.
B.BB902
20.4898
P-value
0.1883
0.0001B4B
AIC:
230.479
Dose
Est._Prob.
Goodness of Fit
Expected
Observed
Size
Sealed
Residual
0.0000
12.8000
B2.0000
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
-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
July 2010
B-4
DO NOT CITE OR QUOTE
-------
1
BMDL =
7.069B4
Table B-3. Benchmark modeling results for alvelolar epithelial hyperplasia in
male F344/N rats (BMR = 5% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
231.042
0.1317
1.698
7.23716
4.91134
Logistic
232.34
0.0775
1.701
13.0617
10.1972
Log-logistica
230.479
0.1753
-0.649
5.41078
3.34864
Log-probit
233.859
0.0363
1.939
19.8906
13.6243
Multistage
231.042
0.1317
1.698
7.23716
4.91134
Probit
232.209
0.0813
1.711
12.3417
9.6301
Weibull
231.042
0.1317
1.698
7.23729
4.91134
Quantal-linear
231.042
0.1317
1.698
7.23729
4.91134
Dichotomous hill
231.705
0.1112
-0.1356
2.63097
1.72618
a model choice based on lowest AIC
Figure B-2. Log-logistic model fit for alveolar epitheliala hyperplasia in male F344/N rats
(BMR = 5% extra risk)
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
BMDL BMD
40
dose
08:09 02/03 2010
July 2010
B-5
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
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. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_a~lv_hyper_Ln~l-BMR05-Restrict.p~lt
Wed Feb 03 08:09:27 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-i ntercept-slope*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
( *** 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
i ntercept
1
-0.66
Variable
background
i ntercept
slope
Estimate
0.130984
-4.63283
1
-0.66
1
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood) # Param's Deviance Test d.f.
-111.57 4
-113.24
-121.815
230.479
3.33902
20.4898
P-value
0.1883
0.0001343
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Sealed
Residual
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
-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
July 2010
B-6
DO NOT CITE OR QUOTE
-------
Table B-4. Benchmark modeling results for Alveolar epithelial hyperplasia in
female F344/N rats (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
245.78
0.0612
1.998
8.0322
5.89582
Logistic
248.949
0.0163
1.99
14.8564
11.9857
Log-logistica
243.677
0.1779
-0.453
4.90719
3.27097
Log-probit
249.954
0.0079
2.446
15.342
10.7468
Multistage
245.78
0.0612
1.998
8.03223
5.89582
Probit
248.806
0.0171
2.006
14.4844
11.8082
Weibull
245.78
0.0612
1.998
8.03223
5.89582
Quantal-linear
245.78
0.0612
1.998
8.03223
5.89582
Dichotomous hill
244.808
0.113
-0.1096
3.08661
2.02512
a model choice based on lowest AIC
Figure B-3. Log-logistic model fit for alveolar epithelial hyperplasia in female F344/N
rats (BMR =10% extra risk)
Log-Logistic Model with 0.95 Confidence Level
BMD
40
dose
14:09 01/14 2010
Log-Logistic
July 2010
B-7
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
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-
Restri ct.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_f_a~lv_hyper_Ln~l-BMRlO-
Restri ct.pit
Thu Jan 14 14:09:02 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-i ntercept-slope*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 = -B.745B2
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** 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
i ntercept
1
-0.62
-0.62
1
Variable
background
i ntercept
slope
Estimate
0.145252
-B.7879B
1
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Indicates that this value is not calculated.
Analysis of Deviance Table
Devi ance
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood)
-118.153
-119.839
-135.512
243.677
# Param's
4
2
1
3.37005
34.7167
Test d.f.
2
3
P-value
0.1854
<.0001
Dose
Est._Prob.
Goodness of Fit
Expected Observed Size
Sealed
Residual
0.0000
12.8000
32.0000
80.0000
ChiA2 = 3.45
0.1453
0.3373
0.5044
0.6960
d.f. = 2
7.117 6.000 49 -0.453
16.866 22.000 50 1.536
25.218 22.000 50 -0.910
34.799 34.000 50 -0.246
P-value = 0.1779
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 4.90719
July 2010
B-8
DO NOT CITE OR QUOTE
-------
1
BMDL =
B.27097
July 2010 B-9 DO NOT CITE OR QUOTE
-------
1
2
Table B-5. Benchmark modeling results for bronchiolar hyperplasia in male
B6C3F1 mice (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
x2 residual
BMD
BMDL
Gamma
192.219
0.1003
1.561
9.962
7.95025
Logistic
206.147
0.0019
2.136
25.582
20.9208
Log-logistica
188.645
0.5085
0.8
7.54241
5.60381
Log-probit
203.779
0.0009
2.278
18.0076
12.7086
Multistage
192.219
0.1003
1.561
9.962
7.95025
Probit
205.312
0.0023
2.094
23.8731
19.6205
Weibull
192.219
0.1003
1.561
9.962
7.95025
Quantal-linear
192.219
0.1003
1.561
9.962
7.95025
Dichotomous hill
190.376
1
-3.22E-06
6.4695
4.24464
Figure B-4. Log-logistic model fit for bronchiolar hyperplasia in male B6C3F1 mice
(BMR = 10% extra risk)
a model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
10 20 30 40 50 60 70 80
dose
14:03 01/14 2010
Log-Logistic
BMDL BMP
July 2010
B-10
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
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-
Restri ct.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_mouse_m_bronch_hyper_Ln~l-
BMR10-Restri ct.pit
Thu Jan 14 14:03:40 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-i ntercept-slope*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
i ntercept
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
i ntercept
1
Variable
background
i ntercept
slope
Estimate
0
-4.21777
1
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
- Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood) # Param's Deviance Test d.f. P-value
-92.1882 4
-93.3224
-113.552
188.645
2.26827
42.7283
0.5186
<.0001
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Sealed
Residual
0.0000
12.8000
32.0000
80.0000
ChiA2 = 2.32
0.0000
0.1586
0.3204
0.5410
0.000
7.932
16.019
27.049
0.000
10.000
18.000
23.000
50
50
50
50
0.000
0.800
0.600
-1.149
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
July 2010
B-11
DO NOT CITE OR QUOTE
-------
Table B-6. Benchmark modeling results for bronchiolar hyperplasia in male
B6C3F1 mice (BMR = 5% extra risk)
Model
AIC
Goodness-of-fit
p-value
% residual
BMD
BMDL
Gamma
192.219
0.1003
0
4.84986
3.87047
Logistic
206.147
0.0019
0.897
14.2582
11.4862
Log-logistica
188.645
0.5085
0
3.57272
2.65444
Log-probit
203.779
0.0009
2.278
12.522
8.83728
Multistage
192.219
0.1003
0
4.84986
3.87047
Probit
205.312
0.0023
0.992
13.136
10.6641
Weibull
192.219
0.1003
0
4.84986
3.87047
Quantal-linear
190.376
1
0
3.6667
0.932026
Dichotomous hill
192.219
0.1003
0
4.84986
3.87047
a model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
0.6
0.5
0.4
0.3
0.2
0.1
0
BMDL BMD
0
10
20
30
40
50
60
70
80
dose
14:04 01/14 2010
Figure B-5. Log-logistic model fit for bronchiolar hyperplasia in male B6C3F1 mice
(BMR = 5% extra risk)
July 2010
B-12
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
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-
Restri ct.pit
Thu Jan 14 14:04:34 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-i ntercept-slope*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
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
i ntercept
intercept 1
Variable
background
i ntercept
slope
Estimate
0
-4.21777
1
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Indicates that this value is not calculated.
Analysis of Deviance Table
Devi ance
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood)
-92.1882
-93.3224
-113.552
188.645
# Param's
4
1
1
2.26827
42.7283
Test d.f.
3
3
P-value
0.5186
<.0001
Dose
Est._Prob.
Goodness of Fit
Expected Observed
Size
Sealed
Residual
0.0000
12.8000
32.0000
80.0000
0.0000
0.1586
0.3204
0.5410
0.000
7.932
16.019
27.049
0.000
10.000
18.000
23.000
50
50
50
50
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
July 2010
B-13
DO NOT CITE OR QUOTE
-------
Table B-7. Benchmark modeling results for chronic inflammatioin in male F344/N
rats (BMR =10% extra risk)
Model
AIC
Goodness-of-fit
p-value
% residual
BMD
BMDL
Gamma
81.4586
0.8964
0.39
15.2489
10.1164
Logistic
87.0594
0.0925
-0.291
23.8087
18.9473
Log-logistica
81.3682
0.9398
0.286
14.6428
9.27776
Log-probit
83.9766
0.2144
1.458
17.7991
13.7362
Multistage
81.4586
0.8964
0.39
15.2489
10.1164
Probit
86.6596
0.1067
-0.345
22.6768
17.7855
Weibull
81.4586
0.8964
0.39
15.2489
10.1164
Quantal-linear
81.4586
0.8964
0.39
15.2489
10.1164
Dichotomous hill
a model choice based on lowest AIC
Log-Logistic Model with 0.95 Confidence Level
0.35
Log-Logistic
0.3
0.25
0.2
0.15
0.1
0.05
BMDL
BMD
0
5
10
15
20
25
30
dose
14:21 01/14 2010
Figure B-6. Log-logistic model fit for olfactory chronic inflammation in male F344/N rats
(BMR = 10% extra risk)
July 2010
B-14
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
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-BMRl0-Restrict. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_inf1ammation_hdd_Ln~l-BMRl0-
Restri ct.pit
Thu Jan 14 14:21:54 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-i ntercept-slope*Log(dose))]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = B
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
intercept
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
i ntercept
1
Variable
background
i ntercept
slope
Estimate
0
-4.88117
1
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(1 i ke~li hood) # Param's Deviance Test d.f.
-B9.62B1 3
-39.6841 1 0.121914 2
-46.4291 1 13.6119 2
P-value
0.9409
0.001107
AIC: 81.3682
Goodness of Fit
Dose Est._Prob. Expected Observed
Size
Sealed
Residual
0.0000
12.8000
32.0000
ChiA2 = 0.12
0.0000
0.0885
0.1954
d.f. = 2
0.000 0.000 50
4.426 5.000 50
9.574 9.000 49
P-value = 0.9398
0.000
0.286
-0.207
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 14.6428
BMDL = 9.27776
July 2010
B-15
DO NOT CITE OR QUOTE
-------
1
July 2010 B-16 DO NOT CITE OR QUOTE
-------
Table B-8. Benchmark modeling results for olfactory atrophy in male F344/N rats
(BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
% residual
BMD
BMDL
Gamma
106.376
NA
0
10.6003
7.99938
Logistic3
105.53
0.2655
-0.597
7.70048
5.97454
Log-logistic
106.376
NA
0
10.81
8.62799
Log-probit
106.376
NA
0
10.9386
8.79455
Multistage
107.65
0.0817
-1.408
6.95763
5.20262
Probit
106.283
0.1555
-0.901
6.91725
5.40111
Weibull
106.376
NA
0
9.95012
7.06875
Quantal-linear
125.166
0
0.459
2.28431
1.80011
Dichotomous hill
0
a model choice based on lowest AIC
Logistic Model with 0.95 Confidence Level
1
0.8
~0
0
"5
a;
<£
c
o
o
ro
0.6
0.4
0.2
0
Logistic
BMDL
BMD
10 15 20 25 30
dose
14:19 01/14 2010
Figure B-7. Logistic model fit for olfactory atrophy in male F344/N rats (BMR = 10%
extra risk)
July 2010 B-17 DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMRlO. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMRl0.p~lt
Thu Jan 14 14:19:23 2010
BMDS Model Run
The form of the probability function is:
P[response] = l/[l+EXP(-i ntercept-s~lope*dose)]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = B
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 model parameter(s) -background
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
intercept slope
intercept 1 -0.85
slope -0.85 1
Variable
i ntercept
slope
Parameter Estimates
Estimate Std. Err.
-B.25094 0.484263
0.179B56 0.026275B
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-4.20007 -2.3018
0.127857 0.230855
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(1 i ke~li hood) # Param's Deviance Test d.f.
-50.1882 3
-50.7651 2 1.15379 1
-100.819 1 101.262 2
P-value
AIC: 105.53
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0.2828
<.0001
Sealed
Residual
0.0000
12.8000
32.0000
ChiA2 = 1.24
0.0373
0.2778
0.9233
d.f. = 1
1.865 3.000 50
13.892 12.000 50
45.243 46.000 49
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
0.847
-0.597
0.406
July 2010
B-18
DO NOT CITE OR QUOTE
-------
Table B-9. Benchmark modeling results for olfactory atrophy in male F344/N rats
(BMR = 5% extra risk)
Model
AIC
Goodness-of-fit
p-value
% residual
BMD
BMDL
Gamma
106.376
NA
0
8.88228
6.24061
Logistic3
105.53
0.2655
0.847
4.90734
3.52532
Log-logistic
106.376
NA
0
9.14915
6.92292
Log-probit
106.376
NA
0
9.51381
7.35396
Multistage
107.65
0.0817
0.377
4.85459
3.12143
Probit
106.283
0.1555
0.977
4.28231
3.10069
Weibull
106.376
NA
0
7.68497
5.01204
Quantal-linear
125.166
0
0.459
1.11208
0.87636
Dichotomous hill
0
a model choice based on lowest AIC
Logistic Model with 0.95 Confidence Level
1
0.8
"O
0
t>
a;
<
c
o
"5
0.6
0.4
0.2
0
10 15 20 25 30
dose
14:20 01/14 2010
Figure B-8. Logistic model fit for olfactory atrophy in male F344/N rats (BMR = 5%
extra risk)
BMDL
BMD
Logistic
July 2010
B-19
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMR05 . (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~log_rat_m_atrophy_hdd_Log-BMR05.p~lt
Thu Jan 14 14:20:15 2010
BMDS Model Run
The form of the probability function is:
P[response] = l/[l+EXP(-i ntercept-s~lope*dose)]
Dependent variable = Effect
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = B
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 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 )
i ntercept
slope
i ntercept
1
-0.85
slope
-0.85
1
Variable
i ntercept
slope
Parameter Estimates
Estimate Std. Err.
-B.25094 0.484263
0.179B56 0.0262753
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-4.20007 -2.3018
0.127857 0.230855
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(1 i ke~li hood) # Param's Deviance Test d.f.
-50.1882 3
-50.7651 2 1.15379 1
-100.819 1 101.262 2
P-value
AIC: 105.53
Goodness of Fit
Dose Est._Prob. Expected Observed Size
0.2828
<.0001
Sealed
Residual
0.0000
12.8000
32.0000
0.0373
0.2778
0.9233
1.865
13.892
45.243
3.000
12.000
46.000
50
50
49
0.847
-0.597
0.406
ChiA2 = 1.24
d.f. = 1
P-value = 0.2655
Benchmark Dose Computation
Specified effect = 0.05
Risk Type = Extra risk
Confidence level = 0.95
BMD = 4.90734
BMDL = 3.52532
July 2010
B-20
DO NOT CITE OR QUOTE
-------
1
2
Table B-10. Benchmark modeling results for olfactory necrosis in male F344/N
rats (BMR =10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
124.435
1
0
6.46561
3.70666
Logistic
130.942
0.0328
1.45
12.1684
9.77545
Log-logistic
124.435
1
0
6.92124
2.96263
Log-probita
122.499
0.9686
0.188
7.98173
6.41755
Multistage
124.435
1
0
5.8893
3.70666
Probit
129.762
0.0494
1.387
11.3581
9.13936
Weibull
124.435
1
0
6.31726
3.70666
Quantal-linear
122.737
0.8622
0
4.75407
3.65317
Dichotomous hill
0
a model choice based on lowest AIC
LogProbit Model with 0.95 Confidence Level
"O
0
-M
O
0
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\1np_rat_m_necrosis_hdd_Lnp-BMRlO.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_m_necrosis_hdd_Lnp-BMRl0.p~lt
Thu Jan 14 14:26:59 2010
BMDS Model Run
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 = B
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 = -B.BB80B
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
intercept
intercept
1
Variable
background
i ntercept
slope
Estimate
0
-3.35871
1
Parameter Estimates
Std. Err.
NA
0.133307
NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-3.61998
-3.09743
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
Devi ance
Log(1 i ke~li hood)
-60.2177
-60.2494
-83.5122
122.499
# Param's
3
1
1
Test d.f. P-value
0.063351
46.5889
0.9688
<.0001
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Sealed
Residual
0.0000
0.0000
0.000
0.000
50
0.000
12.8000
0.2092
10.459
11.000
50
0.188
32.0000
0.5426
26.588
26.000
49
-0.169
ChiA2 = 0.06
d.f. = 2
P-value = 0.9686
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
July 2010
B-22
DO NOT CITE OR QUOTE
-------
BMD =
BMDL =
7.9817B
6.41755
Table B-ll. Benchmark modeling results for olfactory necrosis in male F344/N
rats (BMR =5% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
124.435
1
0
3.68522
1.80454
Logistic
130.942
0.0328
1.45
7.60632
5.7094
Log-logistic
124.435
1
0
4.22667
1.40335
Log-probita
122.499
0.9686
0
5.55031
4.46261
Multistage
124.435
1
0
2.97375
1.80454
Probit
129.762
0.0494
1.387
7.0625
5.28703
Weibull
124.435
1
0
3.49306
1.80454
Quantal-linear
122.737
0.8622
0
2.31445
1.7785
Dichotomous hill
0
* model choice based on lowest AIC
LogProbit Model with 0.95 Confidence Level
"D
0
-M
0
0
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\1np_rat_m_necrosis_hdd_Lnp-BMR05.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_m_necrosis_hdd_Lnp-BMR05.p~lt
Thu Jan 14 14:27:40 2010
BMDS Model Run
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 = B
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 = -B.BB80B
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
intercept
intercept
1
Variable
background
i ntercept
slope
Estimate
0
-3.35871
1
Parameter Estimates
Std. Err.
NA
0.133307
NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-3.61998
-3.09743
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
Devi ance
Log(1 i ke~li hood)
-60.2177
-60.2494
-83.5122
122.499
# Param's
3
1
1
Test d.f. P-value
0.063351
46.5889
0.9688
<.0001
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Sealed
Residual
0.0000
0.0000
0.000
0.000
50
0.000
12.8000
0.2092
10.459
11.000
50
0.188
32.0000
0.5426
26.588
26.000
49
-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
July 2010
B-24
DO NOT CITE OR QUOTE
-------
1
2
BMD =
BMDL =
5.550B1
4.46261
Table B-12. Benchmark modeling results for olfactory necrosis in female F344/N
rats (BMR =10% extra risk)
Model
AIC
Goodness-of-fit
p-value
% residual
BMD
BMDL
Gamma
108.455
0.1223
1.54
33.1378
21.4781
Logistic
114.403
0.0069
2.564
50.8598
41.8856
Log-logistic
108.312
0.1357
1.469
32.4911
20.1388
Log-probita
106.815
0.115
2
35.6629
28.3477
Multistage
108.87
0.1361
1.339
31.2054
20.9166
Probit
113.454
0.0095
2.504
47.668
38.935
Weibull
108.549
0.1241
1.509
32.8886
21.3452
Quantal-linear
106.909
0.2879
1.188
29.5366
20.8661
Dichotomous hill
103.075
1
1.13E-05
30.221
27.5059
a Dichotomous hill model has lowest AIC value, but 2 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.
Figure B-ll. Log-probit model fit for olfactory necrosis in female F344/N rats (BMR =
10% extra risk)
Log Pro bit
BMDL
BMD
LogProbit Model with 0.95 Confidence Level
0.35
0.25
0.15
0.05
10 20 30 40 50 60 70 80
dose
14:16 01/14 2010
July 2010
B-25
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\1np_rat_f_necrosis_Lnp-BMRlO.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_necrosis_Lnp-BMRl0.p~lt
Thu Jan 14 14:16:20 2010
BMDS Model Run
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
background =
intercept =
slope =
(and Specified) Parameter Values
0
-4.83555
1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
i ntercept
intercept 1
Variable
background
i ntercept
slope
Parameter Estimates
Estimate Std. Err.
0 NA
-4.85566 0.142346
1 NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-5.13466
-4.57667
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(1 i ke~li hood)
-49.5375
-52.4076
-64.911
106.815
# Param's
4
1
1
Deviance Test d.f.
P-value
5.74025
30.7469
0.125
<.0001
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Sealed
Residual
0.0000
0.0000
0.000
0.000
49
0.000
12.8000
0.0105
0.527
0.000
50
-0.730
32.0000
0.0823
4.114
8.000
50
2.000
80.0000
0.3179
15.894
12.000
50
-1.183
ChiA2 = 5.93
d.f. = 3
P-value = 0.1150
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
July 2010
B-26
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
BMD = 35.6629
BMDL = 28. 3477
Dichotomous-Hill Model with 0.95 Confidence Level
Dichotomous-Hill
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
BMDL
BMD
0
10
20
30
40
50
60
70
80
dose
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_Dh1 -BMRlO-Restrict. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\dh~l_rat_f_necrosis_Dh~l-BMRlO-Restrict.p~lt
Thu Jan 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 < v <= 1
v is the maximum probability of response predicted by the model,
and v*g is the background estimate of that probability.
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
Default Initial Parameter Values
July 2010
B-27
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
v = -9999
g = -9999
intercept = -9.02B4B
slope = 1.889B8
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -g -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
v intercept
intercept
1
-0.61
-0.61
1
Variable
v
g
intercept
slope
Parameter Estimates
Estimate Std. Err.
0.24 0.060B988
0 NA
-61.6901 1.23084
18 NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
0.121621 0.B58B8
-64.1025
-59.2777
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(1 i ke~li hood) Deviance Test d.f.
-49.5375
-49.5375 3.29863e-006 2
-64.911 30.7469 3
P-value
<.0001
AIC: 103.075
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
12.8000
32.0000
80.0000
0.0000
0.0000
0.1600
0.2400
0.000
0.000
8.000
12.000
0
0
8
12
49 0
50 -0.001284
50 1.131e-005
50 -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
July 2010
B-28
DO NOT CITE OR QUOTE
-------
Table B-13. Benchmark modeling results for olfactory necrosis in female F344/N
rats (BMR =5% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
108.455
0.1223
-1.256
18.8703
10.4563
Logistic
114.403
0.0069
2.564
34.1134
26.6532
Log-logistic
108.312
0.1357
-1.263
18.6016
9.53944
Log-probita
106.815
0.115
2
24.7991
19.7123
Multistage
108.87
0.1361
-1.458
15.6751
10.1829
Probit
113.454
0.0095
2.504
31.4159
24.4034
Weibull
108.549
0.1241
-1.299
18.2634
10.3916
Quantal-linear
106.909
0.2879
-1.528
14.3795
10.1584
Dichotomous hill
103.075
1
1.13E-05
28.5901
26.0761
a Dichotomous hill model has lowest AIC value, but 2 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
0.4
0.35
0.3
0.25
LogProbit
BMDL
BMD
14:17 01/14 2010
Figure B-13. Log-probit model fit for olfactory necrosis in female F344/N rats (BMR:
5% extra risk)
July 2010
B-29
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\1np_rat_f_necrosis_Lnp-BMR05.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_necrosis_Lnp-BMR05.p~lt
Thu Jan 14 14:17:01 2010
BMDS Model Run
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
background =
intercept =
slope =
(and Specified) Parameter Values
0
-4.83555
1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
i ntercept
intercept 1
Variable
background
i ntercept
slope
Parameter Estimates
Estimate Std. Err.
0 NA
-4.85566 0.142346
1 NA
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-5.13466
-4.57667
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(1 i ke~li hood) # Param's Deviance Test d.f. P-value
-49.5375 4
-52.4076 1 5.74025 3 0.125
-64.911 1 30.7469 3 <.0001
106.815
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Sealed
Residual
0.0000
0.0000
0.000
0.000
49
0.000
12.8000
0.0105
0.527
0.000
50
-0.730
32.0000
0.0823
4.114
8.000
50
2.000
80.0000
0.3179
15.894
12.000
50
-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
July 2010
B-30
DO NOT CITE OR QUOTE
-------
1 BMDL = 19.7123
Table B-14. Benchmark modeling results for olfactory basal cell hyperplasia in
female F344/N rats (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
78.8231
0.7502
0.324
22.6607
18.6776
Logistic
83.7749
0.0615
0.81
22.7096
18.8101
Log-logistic
78.3698
0.8754
0.12
24.0671
20.2672
Log-probita
78.083
0.9521
0.093
23.4933
19.7198
Multistage
85.0835
0.155
-1.956
15.3009
13.2469
Probit
83.6185
0.1092
-1.283
22.0007
17.8681
Weibull
81.3562
0.3487
-1.19
20.7516
16.5638
Quantal-linear
120.402
0
0
5.59788
4.52837
Dichotomous hill
77.9075
1
5.70E-07
29.3724
23.7917
a Dichotomous hill model has lowest AIC value, but 2 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
1
0.8
"O
0
"5
a;
<
0.6
0.4
0.2
0
10 20 30 40 50 60 70 80
dose
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)
BMDL
BMD
LogProbit
July 2010
B-31
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\1 np_rat_f_basa~l_hyper_Lnp-BMRlO. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_basa~l_hyper_Lnp-BMRl0.p~lt
Thu Jan 14 14:14:05 2010
BMDS Model Run
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
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
i ntercept
slope
1
-0.99
-0.99
1
Variable
background
i ntercept
slope
Parameter Estimates
Estimate
0
-9.9865
2.7576
Std.
Err.
NA
1.64723
0.456613
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
-13.215
1.86265
-6.758
3.65254
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood)
-36.9537
-37.0415
-126.434
78.083
# Param's
4
2
1
Deviance Test d.f.
P-value
0.175584
178.961
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
0.916
<.0001
Sealed
Residual
0.0000
12.8000
32.0000
80.0000
ChiA2 =0.10
0.0000
0.0016
0.3338
0.9820
d.f. = 2
0.000 0.000 49 0.000
0.078 0.000 50 -0.279
16.691 17.000 50 0.093
49.101 49.000 50 -0.107
P-value = 0.9521
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
July 2010
B-32
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
BMD =
BMDL =
2B.49BB
19.7198
Dichotomous-Hill Model with 0.95 Confidence Level
Dichotomous-Hill
1
0.8
0.6
0.4
0.2
0
BMDL
BMD
0
10
20
30
40
50
60
70
80
dose
14:14 01/14 2010
Figure B-15. Dichotomous 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:\Ch~loroprene\NTP_BMDS\dh~l_rat_f_basa~l_hyper_Dh~l-BMRlO-Restrict. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\dh~l_rat_f_basa~l_hyper_Dh~l-BMRlO-
Restri ct.pit
Thu Jan 14 14:14:07 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 < v <= 1
v is the maximum probability of response predicted by the model,
and v*g is the background estimate of that probability.
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
Default Initial Parameter Values
v = -9999
July 2010
B-33
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
g = -9999
intercept = -16.5503
slope = 4.64205
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -g -slope
have been estimated at a boundary point, or have been specified by the
user,and do not appear in the correlation matrix )
v intercept
i ntercept
1
-0.1
-0.1
1
Variable
v
g
intercept
slope
Parameter Estimates
Estimate Std. Err.
0.98 0.019799
0 NA
-63.0158 0.303295
18 NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
0.941195 1.01881
-63.6102
-62.4213
Model
Full model
Fitted model
Reduced model
AIC:
Analysis of Deviance Table
Log(1 i ke~li hood) Deviance Test d.f.
-36.9537
-36.9537 3.57771e-006 2
-126.434 178.961 3
77.9075
P-value
<.0001
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Scaled
Residual
0.0000
12.8000
32.0000
80.0000
0.0000
0.0000
0.3400
0.9800
0.000
0.000
17.000
49.000
0
0
17
49
49 0
50 -0.001337
50 5.704e-007
50 -1.192e-007
ChiA2 = 0.000002
d.f. = 2
P-value = 1.0000
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
Warni ng:
BMDL =
0.1
Extra risk
0.95
29.3724
BMDL computation is at best imprecise for these data
23.7917
July 2010
B-34
DO NOT CITE OR QUOTE
-------
1
Table B-15. Benchmark modeling results for olfactory basal cell hyperplasia in
female F344/N rats (BMR = 5% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
78.8231
0.7502
-0.597
19.1684
15.0243
Logistic
83.7749
0.0615
-1.253
17.4935
13.1784
Log-logistic
78.3698
0.8754
-0.448
20.879
16.769
Log-probita
78.083
0.9521
-0.279
20.5934
16.7154
Multistage
85.0835
0.155
-1.956
10.676
8.74893
Probit
83.6185
0.1092
-1.283
16.5716
12.2502
Weibull
81.3562
0.3487
-1.19
15.9699
12.0527
Quantal-linear
120.402
0
0
2.72525
2.20457
Dichotomous hill
77.9075
1
5.70E-07
28.1762
22.8227
a Dichotomous hill model has lowest AIC value, but 2 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
1
0.8
"O
0
"5
a;
<
0.6
0.4
0.2
0
10 20 30 40 50 60 70 80
dose
14:14 01/14 2010
Figure B-16. Log-probit model fit for olfactory basal cell hyperplasia in female F344/N
rats (BMR =5% extra risk)
2
July 2010 B-35 DO NOT CITE OR QUOTE
BMDL
BMD
LogProbit
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\1 np_rat_f_basa~l_hyper_Lnp-BMR05 . (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_basa~l_hyper_Lnp-BMR05.p~lt
Thu Jan 14 14:14:47 2010
BMDS Model Run
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
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 )
i ntercept
slope
i ntercept
slope
1
-0.99
-0.99
1
Variable
background
i ntercept
slope
Parameter Estimates
Estimate
0
-9.9865
2.7576
Std.
Err.
NA
1.64723
0.456613
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
-13.215
1.86265
-6.758
3.65254
Model
Full model
Fitted model
Reduced model
AIC:
Dose
Analysis of Deviance Table
Log(1 i ke~li hood) # Param's Deviance Test d.f.
- - 4
2 0.175584 2
1 178.961 3
P-value
-36.9537
-37.0415
-126.434
78.083
Est._Prob.
Goodness of Fit
Expected Observed Size
0.916
<.0001
Sealed
Residual
0.0000
12.8000
32.0000
80.0000
0.0000
0.0016
0.3338
0.9820
0.000
0.078
16.691
49.101
0.000
0.000
17.000
49.000
49
50
50
50
0.000
-0.279
0.093
-0.107
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
July 2010
B-36
DO NOT CITE OR QUOTE
-------
1
2
BMD =
BMDL =
20.5934
16.7154
Table B-16. Benchmark modeling results for kidney (renal tubule) hyperplasia in
male F344/N rats (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
262.742
0.6482
0.091
9.58982
6.61749
Logistic
263.873
0.3712
0.128
14.4291
11.0906
Log-logistica
262.083
0.9017
-0.136
6.52869
3.95681
Log-probit
264.054
0.3356
0.487
17.4209
11.9381
Multistage
262.742
0.6482
0.091
9.58986
6.61749
Probit
263.882
0.3695
0.131
14.3921
11.164
Weibull
262.742
0.6482
0.091
9.58986
6.61749
Quantal-linear
262.742
0.6482
0.091
9.58986
6.61749
Dichotomous hill
20090.6
0
a model choice based on lowest BMDL
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
0.8
0.7
0.6
0.5
0.4
0.3
0.2
BMD
BMDI
0
10
20
30
40
50
60
70
80
dose
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)
July 2010
B-37
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
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. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~ln~l_rat_m_kid_hyper_Ln~l-BMRlO-Restrict.p~lt
Thu Jan 14 14:22:37 2010
BMDS Model Run
The form of the probability function is:
P[response] = background+(l-background)/[l+EXP(-i ntercept-slope*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
( *** 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 -4.07B4B * * *
slope 1 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood)
-128.9B8
-129.042
-1B8.469
262.083
# Param's
4
2
1
Deviance Test d.f.
0.206766
19.0624
Dose
Est._Prob.
Goodness of Fit
Expected Observed Size
P-value
0.9018
0.0002654
Sealed
Residual
0.0000
12.8000
B2.0000
80.0000
ChiA2 = 0.21
0.2808
0.4094
0.5344
0.6954
14.039
20.471
26.718
34.772
14.000
20.000
28.000
34.000
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
50
50
50
50
-0.012
-0.136
0.363
-0.237
July 2010
B-38
DO NOT CITE OR QUOTE
-------
1
July 2010 B-39 DO NOT CITE OR QUOTE
-------
Table B-17. Benchmark modeling results for kidney (renal tubule) hyperplasia in
female F344/N rats (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
% residual
BMD
BMDL
Gamma
200.059
0.6468
0.236
31.181
14.892
Logistic
198.217
0.8346
0.29
31.33
24.9474
Log-logistic
200.048
0.656
0.211
30.79
13.2994
Log-probita
197.991
0.9302
0.269
32.5323
23.5182
Multistage
200.173
0.5712
0.302
31.846
14.7635
Probit
198.217
0.8345
0.222
29.6902
23.4384
Weibull
200.09
0.6245
0.244
31.13
14.8568
Quantal-linear
198.787
0.6299
-0.692
21.0465
14.1492
Dichotomous hill
202.036
NA
0.1984
30.4841
12.4518
a model choice based on lowest AIC
LogProbit Model with 0.95 Confidence Level
LogProbit
0.6
0.5
0.4
0.3
0.2
0.1
BMDL
BMD
0
0
10
20
30
40
50
60
70
80
dose
14:15 01/14 2010
Figure B-18. Log-probit model fit for kidney (renal tubule) hyperplasia in female F344/N
rats (BMR =10% extra risk)
July 2010
B-40
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_kid_hyper_Lnp-BMRlO. (d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\~lnp_rat_f_kid_hyper_Lnp-BMRl0.p~lt
Thu Jan 14 14:15:37 2010
BMDS Model Run
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 = 0.122449
intercept = -4.95177
slope = 1.04703
Asymptotic Correlation Matrix of Parameter Estimates
( *** 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
i ntercept
Variable
background
i ntercept
slope
1
-0.53
-0.53
1
Parameter Estimates
Estimate Std. Err.
0.119059 0.0336048
-4.76379 0.218134
1 NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
0.0531949 0.184924
-5.19132 -4.33625
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Dose
Log(1 i ke~li hood)
-96.9233
-96.9957
-105.132
197.991
# Param's
4
2
1
Deviance Test d.f.
0.144742
16.4183
Est._Prob.
Goodness of Fit
Expected Observed Size
P-value
0.9302
0.0009307
Sealed
Residual
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
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
July 2010
B-41
DO NOT CITE OR QUOTE
-------
1
2
BMD =
BMDL =
B2.5B2B
23.5182
Table B-18. Benchmark modeling results for forestomach epithelial hyperplasia in
male B6C3F1 mice (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
178.784
0.4716
-0.049
39.6884
20.1391
Logistic
177.328
0.5986
-0.825
26.8011
22.1839
Log-logistic
178.762
0.4806
-0.065
39.6607
21.1117
Log-probita
178.805
0.464
-0.013
39.3506
22.7348
Multistage
177.268
0.6124
-0.771
30.167
16.9463
Probit
177.716
0.5004
-0.984
24.635
20.4757
Weibull
178.737
0.491
-0.095
39.8723
19.6367
Quantal-linear
182.602
0.0523
-0.389
13.9921
10.3765
Dichotomous hill
5853.85
0
a model choice based on lowest AIC
Multistage Model with 0.95 Confidence Level
Multistage
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
BMDL
BMD
0
10
20
30
40
50
60
70
80
dose
14:05 01/14 2010
Figure B-19. Multistage model fit for forestomach epithelial hyperplasia in male B6C3F1
mice (BMR = 10% extra risk)
July 2010
B-42
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Multistage Model. (Version: B.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-
Restri ct.pit
Thu Jan 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 = B
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) = 0.000122B6
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
Variable
Background
Beta(l)
Beta(2)
Estimate
0.0832204
0
0.000115775
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Indicates that this value is not calculated.
Analysis of Deviance Table
Devi ance
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood)
-86.1337
-86.6341
-107.064
177.268
# Param's
4
2
1
1.00079
41.8613
Test d.f.
2
3
P-value
0.6063
<.0001
Dose
Est._Prob.
Goodness of Fit
Expected Observed Size
Sealed
Residual
0.0000
12.8000
32.0000
80.0000
ChiA2 = 0.98
0.0832
0.1004
0.1857
0.5630
d.f. = 2
4.161 4.000 50 -0.082
4.821 6.000 48 0.566
9.100 7.000 49 -0.771
28.151 29.000 50 0.242
P-value = 0.6124
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 30.167
BMDL = 16.9463
July 2010
B-43
DO NOT CITE OR QUOTE
-------
1 BMDU = 36.6564
2 Taken together, (16.9463, 36.6564) is a 90 % two-sided confidenceinterval for the BMD
Table B-19. Benchmark modeling results for forestomach epithelial hyperplasia in
female B6C3F1 mice (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
169.362
0.5864
0.147
33.02
19.9556
Logistic
167.998
0.6241
-0.02
29.3493
24.2933
Log-logistic
169.384
0.5729
0.142
32.8973
20.1355
Log-probita
169.261
0.6545
0.071
32.4471
20.7798
Multistage
167.53
0.7935
-0.013
30.9965
19.3466
Probit
168.273
0.5353
-0.198
26.9397
22.3632
Weibull
169.457
0.5344
0.19
33.3943
19.6657
Quantal-linear
173.528
0.0476
-1.415
15.4655
11.4268
Dichotomous hill
5845.73
0
a model choice based on lowest AIC
Multistage Model with 0.95 Confidence Level
Multistage
0.7
0.6
0.5
0.4
0.3
0.2
0.1
BMDI
BMD
0
10
20
30
40
50
60
70
80
dose
13:58 01/14 2010
Figure B-20. Multistage model fit for forestomach epithelial hyperplasia in female
B6C3F1 mice (BMR = 10% extra risk)
July 2010
B-44
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Multistage Model. (Version: B.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-
Restri ct.pit
Thu Jan 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 = B
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 = 0.062B808
Beta(l) = 0
Beta(2) = 0.00011119
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
-0.5
Beta(2)
-0.5
1
Variable
Background
Beta(l)
Beta(2)
Estimate
0.0645849
0
0.000109661
Parameter Estimates
Std. Err.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Indicates that this value is not calculated.
Analysis of Deviance Table
Devi ance
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood)
-81.5287
-81.7648
-102.B17
167.53
# Param's
4
2
1
0.472098
41.577
Test d.f.
2
B
P-value
0.7897
<.0001
Dose
Est._Prob.
Goodness of Fit
Expected Observed
Size
Sealed
Residual
0.0000
12.8000
32.0000
80.0000
0.0646
0.0812
0.1639
0.5363
3.229
3.981
8.033
26.817
4.000
3.000
8.000
27.000
50
49
49
50
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 = 0.95
BMD = 30.9965
BMDL = 19.3466
July 2010
B-45
DO NOT CITE OR QUOTE
-------
BMDU = 37.6172
Taken together, (19.3466, 37.6172) is a 90% two-sided confidence interval for the BMD
Table B-20. Benchmark modeling results for splenic hematopoietic cell
proliferation in female B6C3F1 mice (BMR = 10% extra risk)
Model
AIC
Goodness-of-fit
p-value
X2 residual
BMD
BMDL
Gamma
171.405
NA
0
5.73584
1.90919
Logistic
169.421
0.8993
0.064
4.06642
3.28512
Log-logistic
171.405
NA
0
6.5828
2.43228
Log-probita
171.405
NA
0
6.91076
3.48982
Multistage
171.405
NA
0
4.41391
1.90919
Probit
169.41
0.9466
0.033
4.03306
3.33147
Weibull
171.405
NA
0
5.17994
1.90919
Quantal-linear
170.771
0.2455
0.264
2.34557
1.7616
Dichotomous hill
0
a model choice based on lowest AIC
Probit Model with 0.95 Confidence Level
1
Probit
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
BMDU BMD
0.1
0
5
10
15
20
25
30
dose
14:02 01/14 2010
Figure B-21. Probit model fit for splenic hematopoietic cell proliferation in female
B6C3F1 mice (BMR = 10% extra risk)
July 2010
B-46
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Probit Model. (Version: 3.1; Date: 05/16/2008)
Input Data File: M:\Ch~loroprene\NTP_BMDS\pro_mouse_f_sp1een_hemato_hdd_Pro-BMRl0.(d)
Gnuplot Plotting File: M:\Ch~loroprene\NTP_BMDS\pro_mouse_f_sp1een_hemato_hdd_Pro-
BMR10.pit
Thu Jan 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 = B
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 = -0.64B08B
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
i ntercept
slope
1
-0.73
-0.7B
1
Variable
i ntercept
slope
Model
Full model
Fitted model
Reduced model
AIC:
Estimate
-0.6497BB
0.0534876
Parameter Estimates
Std. Err.
0.16595
0.00913534
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.97499 -0.324476
0.0355826 0.0713925
Analysis of Deviance Table
Log(1 i ke~li hood)
-82.7026
-82.7048
-102.099
169.41
# Param's
3
2
1
Deviance Test d.f. P-value
0.00449095
38.7924
0.9466
<.0001
Goodness of Fit
Dose
Est._Prob.
Expected
Observed
Size
Sealed
Residual
0.0000
12.8000
32.0000
ChiA2 = 0.00
0.2579
0.5139
0.8559
d.f. = 1
12.897 13.000 50 0.033
25.182 25.000 49 -0.052
41.937 42.000 49 0.026
P-value = 0.9466
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 4.03306
BMDL = 3.33147
July 2010
B-47
DO NOT CITE OR QUOTE
-------
1
Table B-21. Benchmark modeling results for splenic hematopoietic cell
proliferation in female B6C3F1 mice (BMR = 5% extra risk)
Model
AIC
Goodness-of-fit
p-value
% residual
BMD
BMDL
Gamma
171.405
NA
0
3.86036
0.929461
Logistic
169.421
0.8993
0.064
2.10908
1.68891
Log-logistic
171.405
NA
0
4.75284
1.34665
Log-probita
171.405
NA
0
5.33285
2.42674
Multistage
171.405
NA
0
2.35161
0.929461
Probit
169.41
0.9466
0.033
2.07526
1.70339
Weibull
171.405
NA
0
3.21493
0.929461
Quantal-linear
170.771
0.2455
0.264
1.14191
0.85761
Dichotomous hill
0
a model choice based on lowest AIC
Figure B-22. Probit model fit for splenic hematopoietic cell proliferation in female
B6C3F1 mice (BMR = 5% extra risk)
Probit Model with 0.95 Confidence Level
Probit
BMDLBMD
15
dose
12:34 01/21 2010
July 2010
B-48
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
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-BMR05.p~lt
Thu Jan 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
user,and do not appear in the correlation matrix )
i ntercept
slope
i ntercept
1
-0.73
slope
-0.73
1
Parameter Estimates
Variable
i ntercept
slope
Estimate
-0.649733
0.0534876
Std. Err.
0.16595
0.00913534
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.97499 -0.324476
0.0355826 0.0713925
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(1 i ke~li hood) # Param's Deviance Test d.f. P-value
-82.7026 3
-82.7048 2 0.00449095 1 0.9466
-102.099 1 38.7924 2 <.0001
169.41
Dose
Est._Prob.
Goodness of Fit
Expected Observed Size
Sealed
Residual
0.0000
12.8000
32.0000
ChiA2 = 0.00
0.2579
0.5139
0.8559
d.f. = 1
12.897 13.000 50 0.033
25.182 25.000 49 -0.052
41.937 42.000 49 0.026
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
July 2010
B-49
DO NOT CITE OR QUOTE
-------
APPENDIX C: CANCER DOSE-RESPONSE MODELING
Table C-l: Tumor incidence, with time to death with tumor: female mice exposed
to chloroprene via inhalation (NTP, 1998)
Week
of
Study
Number of animals with tumors at each site, at specified week of study
Dose
Total
Hemangiomas,
Harderian
Gland0
Zymbal
Gland0
Group
examined
Lung
Hemangiosarcomas
Mammary
Forestomach
Liver
Skin
Incid.a
Fatal®
0
5
1
0
0
0
0
0
0
0
0
0
69
1
0
0
0
0
0
0
0
0
0
70
1
0
0
0
0
0
0
0
0b
0
71
1
0
0
0
0
0
0
1
0
0
76
1
0
0
0
1
1
0
0
0
0
78
1
0
0
1
0
0
0
0
0
0
88
1
0
0
0
0
0
0
0
0
0
91
0
0
0
0
0
0
1
0
0
95
1
0
0
0
0
0
0
0
0
0
97
1
0
0
0
0
0
0
0
0
0
98
1
0
0
0
0
1
0
0
0
0
101
2
1
0
0
0
0
0
0
0
0
105
36
3
3
0
1
lb
1
18
0
0
12.8
41
1
0
0
0
0
0
0
0
1
0
46
2
0b
0b
0
0b
ob
0b
0b
ob
0b
63
1
0
0
0
0
1
0
1
1
0
64
1
1
0
0
0
1
0
0
0
0
69
1
0
0
1
0
0
0
0
0
0
75
1
1
0
0
0
0
0
1
0
0
76
1
0
0
0
0
0
0
1
0
0
78
1
0
0
0
0
1
0
0
0
0
79
3
0
0
0
0
0
0
1
0
0
87
1
0
0
0
0
0
0
1
1
0
89
2
2
0
0
1
0
0
1
0
0
90
1
0
0
0
0
0
0
1
1
0
91
3
2
0
0
0
0
0
0
1
0
97
3
2
0
0
0
1
0
2
1
0
98
1
1
0
0
1
0
0
0
0
0
99
5
4
0
1
0
0
0
2
2
0
100
1
0
0
0
0
0
0
0
0
0
101
1
1
0
1
0
0
0
1
0
0
102
2
2
0
1
0
0
0
1
0
0
103
2
1
0
0
0
1
0
2
1
0
105
16
11
2
0
3
1
0
11
2
0
32
31
1
0
0
1
0
0
0
0
0
0
50
1
1
0
0
0
0
0
0
0
0
54
1
0
0
0
0
0
0
0
0
0
56
1
0
0
1
0
0
0
0
0
0
57
1
0
0
0
0
1
0
0
0
0
61
1
0
0
0
0
1
0
0
0
0
63
1
1
0
1
0
0
0
0
0
0
65
1
1
0
0
0
0
ob
0
0
0
67
1
0
0
0
0
0
0
0
0
0
68
1
0
0
0
0
0
0
0
0
0
70
1
0
0
1
0
0
0
0
0
0
72
2
2
0
1
0
1
0
1
0
0
73
1
0
0
0
0
0
0
1
0
0
74
1
1
0
1
0
0
0
1
0
0
75
2
1
0
0
1
1
0
0
1
0
76
2
2
0
1
0
1
0
1
1
0
77
2
1
0
1
0
0
0
0
0
0
78
1
0
0
0
0
0
0
1
0
0
79
2
2
0
2
0
1
0
0
0
0
82
1
1
0
0
0
1
0
0
0
0
84
2
1
0
1
0
0
0
0
1
0
July 2010
C-1
DO NOT CITE OR QUOTE
-------
Table C-l: Tumor incidence, with time to death with tumor: female mice exposed
to chloroprene via inhalation (NTP, 1998)
Week
of
Study
Number of animals with tumors at each site, at specified week of study
Dose
Total
Hemangiomas,
Harderian
Gland0
Zymbal
Gland0
Group
examined
Lung
Hemangiosarcomas
Mammary
Forestomach
Liver
Skin
Incid.a
Fatal"
86
1
1
0
0
0
0
0
1
0
0
87
3
2
1
1
2
1
0
1
1
0
89
2
2
0
2
0
1
0
1
0
0
90
3
3
0
0
0
0
0
0
1
0
91
3
3
0
1
0
0
0
3
1
0
92
1
1
0
0
0
1
0
0
0
0
93
1
1
0
0
0
0
0
1
1
0
94
3
3
0
0
0
1
0
2
1
0
96
2
1
0
0
0
0
0
2
1
0
97
1
1
0
0
0
0
0
1
1
0
99
1
0
0
1
0
0
0
1
0
0
103
1
1
0
0
0
0
0
1
0
0
105
1
1
1
0
0
0
0
1
1
0
80
1
1
0
0
0
0
0
0
0
0
0
36
1
0
0
0
0
0
0
0
0
0
47
1
1
0
0
0
0
0
0
0
0
48
1
0
0
0
0
1
0
0
0
0
55
1
0
0
0
0
1
0
1
0
0
64
1
0
0
0
0
0
0
0
0
0
65
1
1
0
0
0
1
0
1
0
0
66
1
1
0
0
0
0
0
0
1
0
67
2
1
0
0
1
0
0
2
2
1
70
1
1
0
0
0
1
0
0
0
0
75
4
4
0
1
0
1
0
1
2
0
76
2
2
0
0
0
0
0
1
1
0
77
1
0
0
1
1
0
0
1
0
0
79
1
1
0
0
0
1
0
1
0
0
81
1
1
0
0
0
0
0
0
0
0
83
3
3
0
1
0
0
1
1
2
1
84
1
1
0
0
0
1
0
1
0
0
86
1
1
0
0
0
1
0
1
0
0
87
1
0
0
0
1
0
0
0
1
0
88
2
2
0
0
1
1
1
2
1
0
90
2
2
1
0
0
0
0
1
1
1
91
7
7
1
2
2
4
1
3
4
0
92
1
1
0
0
0
0
1
1
0
0
93
2
2
0
0
0
0
0
2
1
0
94
1
1
0
0
1
0
0
1
0
0
95
2
2
0
0
0
0
0
2
0
0
96
1
1
0
0
0
0
0
1
0
0
97
2
2
0
0
0
0
0
2
1
0
98
1
1
0
0
0
0
0
1
1
0
105
3
3
1
0
2
1
0
3
0
0
' "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.
b Tissue for one animal of total examined was missing or unsuitable for histopathologic examination.
c Harderian 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.
July 2010
C-2
DO NOT CITE OR QUOTE
-------
Table C-2: Tumor incidence, with time to death with tumor:
male mice exposed to chloroprene via inhalation (NTP, 1998)
Number of animals with tumors at each site, at specified week of study
Dose
Week
of
Study
Total
Hemangiomas,
Group
examined
Hemangiosarcomas
Lung
Forestomach
Harderian
Kidney
gland0
Incid.a
Fatal®
65
1
0
0
0
0
0
0
0
77
1
0
0
0
0
0
0
79
1
0
0
0
0
0
0
82
1
0
0
0
0
0
0
86
1
0
0
0
0
1
0
87
3
0
0
0
0
0
0
90
2
0
0
0
0
0
0
91
2
0
0
1
0
0
0
92
1
0
0
0
0
0
0
95
1
0
0
1
0
0
0
96
1
0
0
0
0
0
0
97
1
0
0
1
0
0
0
98
3
0
0
1
0
0
0
103
1
0
0
0
0
0
0
104
1
0
0
0
0
0
0
105
29
3
0
9
1
1
0
12.8
63
1
0
0
0
0
0
0
75
1
0
0
0
0b
0
0
76
1
0
0
1
0b
0
0b
78
1
0
0
0
0
0
0
83
1
0
0
0
0
0
0
84
2
0
0
0
0
0
0
86
1
0
0
0
0
0
0
87
1
0
0
1
0
0
0
88
1
0
0
0
0
0
0
90
1
0
0
0
0
0
0
91
2
0
0
1
0
0
0
92
1
0
0
1
0
0
0
95
1
0
1
0
0
0
0
96
1
0
0
0
0
0
0
98
1
0
1
1
0
0
0
99
3
0
3
1
0
0
0
101
1
0
0
1
0
1
1
102
1
0
1
0
0
0
0
104
1
0
0
1
0
0
0
105
27
8
0
20
0
4
1
July 2010
C-3
DO NOT CITE OR QUOTE
-------
Table C-2: Tumor incidence, with time to death with tumor:
male mice exposed to chloroprene via inhalation (NTP, 1998)
Number of animals with tumors at each site, at specified week of study
Hemangiomas,
Hemangiosarcomas
30
55
1
0
0
1
0
0
0
63
1
0
0
0
0
0
0
68
1
0
0
1
0b
0
0
71
2
0
1
1
0
0
0
72
1
0
1
1
0
0
0
78
1
0
0
0
0
0
0
79
1
0
0
0
0
0
0
81
2
0
0
1
0
0
0
83
1
0
0
1
0
0
0
86
2
0
1
1
0
1
0
87
4
1
1
4
0
2
0
89
2
0
0
1
0
0
0
90
1
0
0
1
0
0
0
91
1
0
1
1
0
1
0
93
1
0
1
1
0
0
0
95
1
0
1
1
0
0
0
96
2
0
1
1
0
0
0
97
2
0
2
2
0
1
0
98
1
0
1
1
0
0
0
99
3
0
2
2
0
0
0
101
2
0
2
1
0
1
0
102
1
0
1
1
0
0
0
103
2
0
0
2
0
0
1
105
14
6
0
10
2
4
2
on
56
1
0
0
0
0
0
0
oU
61
1
0
0
0
0
0
0
65
1
0
1
0
0
0
0
75
2
0
0
2
0
0
0
81
1
0
0
1
0
0
1
83
1
0
1
1
0
0
0
84
1
0
0
1
1
0
1
85
2
0
0
2
1
1
0
86
1
0
0
1
0
0
0
87
2
0
1
1
0
0
0
89
1
0
0
1
0
0
0
90
3
0
1
2
1
0
0
91
3
0
1
3
0
1
0
92
2
0
2
1
1
1
0
93
3
0
0
3
0
0
1
94
2
0
1
2
1
0
0
95
3
0
2
3
0
1
1
96
1
0
0
1
0
0
0
97
2
1
1
2
1
1
0
98
2
0
1
1
0
0
0
99
0
0
0
0
0
1
1
101
2
0
1
2
0
0
1
105
13
7
0
13
0
6
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.
b Tissue for one animal of total examined was missing or unsuitable for histopathologic examination.
c Harderian 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.
July 2010
C-4
DO NOT CITE OR QUOTE
-------
July 2010 C-5 DO NOT CITE OR QUOTE
-------
Table C-3. Summary of Model Selection and Modeling Results for best-fitting multistage-
Weibull models, using time-to-tumor data for female mice (NTP, 1998)
Site
stages
LLa
Para-
meters
AIC
X2"
BMDio
Responses @ppm levels0
Multistage-Weibull model parameters
0
12.8
32
80
c
to
bo
bl
b2
b3
Lung"
4
28
34
42
3
-83.02
5
176.04
NA
1.20
4.1
27.4
32.4
43.1
3.78
0
2.24
E-09
2.04E
-09
Hemangiomas,
hemangio-
sarcomas
(fatal)'
4
6
18
2
-135.848
4
279.70
NA
10.11
3.45
7.4
13.5
5.91
0
1.01
E-13
0
1.27
E-15
1
-138.519
3
283.04
5.342
4.60
3.25
10.6
11.1
5.42
0
9.04
E-13
2.69
E-13
Hemangiomas,
hemangio-
sarcomas
(incidental/
4
6
18
2
-65.8122
4
139.62
NA
14.94
3.7
6.6
17.6
1.00
0
7.9
E-04
0
4.54
E-06
1
-66.9531
3
139.91
2.2818
8.32
3.3
9.3
15.3
1.00
0
7.0
E-04
1.2
E-04
NA
Harderian
glande
2
5
3
9
3
-58.2559
5
126.51
NA
30.09
2.4
3.6
4.1
8.9
2.91
0
7.59
E-08
4.69
E-09
0
1.09
E-13
2
-58.2661
4
124.53
0.0204
27.47
2.3
3.7
4.2
8.8
2.94
0
6.41
E-08
4.57
E-09
1.42
E-12
1
-58.2663
3
122.53
0.0004
27.08
2.3
3.7
4.3
8.7
2.94
0
6.26
E-08
4.60
E-09
Mammary gland
adenomas,
carcinomas,
adena-
canthomas'1
3
6
11
14
3
-87.9599
5
185.92
NA
20.42
3.5
5.9
8.5
15.7
1.00
0
7.4
E-04
4.96
E-05
Forestomachf
1
0
0
4
3
-19.1743
5
48.35
NA
68.44
0.4
0.4
0.4
3.7
4.23
0
3.05
E-ll
0
0
9.76
E-16
2
-19.5963
4
47.19
0.844
67.81
0.5
0.5
0.7
3.4
4.13
0
5.02
E-ll
0
1.09
E-13
1
-20.7578
3
47.52
2.323
90.02
0.5
1
1.1
2.5
3.54
0
8.24
E-10
8.51
E-ll
Hepatocellular
adenomas,
carcinomas'1
20
26
20
30
3
-119.227
5
248.45
NA
4.24
21.6
23
20.7
30.8
4.16
0
2.70
E-09
1.01
E-10
Skin11
0
11
11
18
3
-87.4625
5
184.93
NA
9.50
0
5.6
10.6
22.4
1.56
0
0
7.91
E-06
Zymbal's glandf
0
0
0
3
3
-11.4018
5
32.80
NA
87.88
0
0
0.2
2.8
1.30
0
0
0
0
3.74
E-10
2
-11.7264
4
31.45
0.6492
97.41
0
0.1
0.4
2.5
1.28
0
1.58
E-27
0
2.91
E-08
1
-12.5984
3
31.20
1.744
166.16
0
0.4
0.8
1.9
1.06
0
0
4.72
E-06
" LL=log-likelihood.
b ¦£ = chi-squared statistic for testing the difference between 2 model fits. Calculated from 2 x |(LLj - LLj)| and evaluated for i-j degrees of freedom. In all cases the
difference was evaluated for consecutive numbers of stages; i-j = 1, and the critical chi-squared value was 3.84.
c "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.
11 Model selection rationale was First-order, only available fit
e Model selection rationale was y2 and AIC
fModel selection rationale was AIC
July 2010
C-6
DO NOT CITE OR QUOTE
-------
Table C-4. Summary of Model Selection and Modeling Results for best-fitting multistage-
Weibull models, using time-to-tumor data for male mice (NTP, 1998)
Site
stages
LV
Para-
meters
AIC
X2b
BMDio
Responses @ppm levels0
Multistage-Weibull model parameters
0
12.8
32
80
c
to
bO
bl
b2
b3
Lungd
13
28
36
43
3
-104.927
5
219.86
NA
2.46
14
26.6
33.9
44.6
3.46
0
4.01
E-08
4.46
E-09
Hemangiomas,
hemangio-
sarcomas
(fatal)11
3
14
23
21
3
-201.96
5
413.92
NA
6.05
4.5
11
13.8
22.3
10.2
0
0
3.29
E-22
4.65
E-23
Hemangiomas,
hemangio-
sarcomas
(incidental)d
3
14
23
21
3
-109.463
5
228.93
NA
7.75
5.3
11.1
15.9
27.2
3.87
0
2.09
E-10
Harderian
gland11
2
5
10
12
3
-73.6639
5
157.33
NA
NA
2.3
5.2
7.4
14
5.57
0
3.26
E-13
3.60
E-14
Kidney'
0
2
3
9
3
-40.9478
5
91.90
NA
29.51
0
1.7
3.4
8.9
6.24
0
0
8.87
E-16
0
3.23
E-20
2
-40.96
4
89.92
0.0244
28.87
0
1.7
3.5
8.8
6.23
0
0
8.88
E-16
2.99
E-18
1
-41.0033
3
88.01
0.0866
26.70
0
2
3.7
8.3
6.09
0
0
2.03
E-15
Forestomache
1
0
2
4
3
-28.0952
5
66.19
NA
84.00
0.7
0.8
1.2
4.3
1.91
0
2.05
E-06
5.62
E-09
2.01
E-09
2
-28.0952
4
64.19
0
84.00
0.7
0.8
1.2
4.3
1.91
0
2.05
E-06
5.62
E-09
2.01
E-09
1
-28.3188
3
62.64
0.4472
108.76
0.6
1.1
1.8
3.6
1.79
0
3.03
E-06
2.34
E-07
" LL=log-likelihood.
b ¦£ = chi-squared statistic for testing the difference between 2 model fits. Calculated from 2 x |(LLj - LLj)| and evaluated for i-j degrees of freedom. In all cases the
difference was evaluated for consecutive numbers of stages; i-j = 1, and the critical chi-squared value was 3.84.
c "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.
d Model selection rationale was First-order, only available fit
e Model selection rationale was y2 and AIC
July 2010
C-7
DO NOT CITE OR QUOTE
-------
Incidental Risk: F LUNG 1s
Dose = 32.00
Dose = 80.00
_Q
TO
_Q
O
00
o
CO
o
CM
O
O
O
.Q
TO
_Q
O
00
o
CO
o
CM
O
O
O
Time
Time
Figure C-l. Female mice, alveolar/bronchiolar tumors. Details below.
July 2010
C-8
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\msw\F_LUNG_ls. (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl)}
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
Maximum number of iterations = 16
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial
c
t_0
Parameter Values
B.77778
0 Specified
beta_0 = 2.32179e-009
beta_l = 2.1101Be-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_l
1
-0.99
-1
beta_0
-0.99
1
0.99
beta_l
-1
0.99
1
Parameter Estimates
Variable
Estimate
Std. Err.
c
3.78542
0.978326 1.86793
beta.
.0
2.24014e-009
1
.03745e-
008 -1.80935e-008
beta.
.1
2.03972e-009
8
.87049e-
009 -1.53461e-008
Log(1 i ke~li hood)
# Param
AIC
Fitted Model
-83.02
3
172.04
Data Summary
CLASS
C
F I
U
Total
Expected Response
DOSE
0
46
0 4
0
50
4.10
IB
21
0 28
1
50
27.45
32
16
0 34
0
50
32.39
80
8
0 42
0
50
43.13
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
5.7029
2.25738e-008
1.94256e-008
July 2010
C-9
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
BMDU =
1.19617
0.88B475
1.60092
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
0.114103
0.0865258
0.148645
July 2010
C-10
DO NOT CITE OR QUOTE
-------
Incidental Risk: F_HEM3fatal_2s
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose= 0.00 Dose= 12.80
Time
Time
Dose = 32.00
Time
Figure C-2. Female mice, hemangiomas and hemangiosarcomas in all organs; high dose
dropped, hemangiosarcomas occurring before termination considered fatal
July 2010
C-11
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\msw\F_HEM3fata~l_2s. (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl+beta_2*doseA2)}
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
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 = 2.2479e-010
beta_l = 2.06502e-034
beta_2 = 2.12137e-012
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 )
c
beta_0
beta_2
beta_0
-1
1
0.99
beta_2
-1
0.99
1
Variable
c
beta_0
beta_l
beta_2
Estimate
5.90503
1.01175e-013
0
1.26539e-015
Parameter Estimates
Std. Err.
1.49573
7.08031e-013
NA
8.53103e-015
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
2.97346
-1.28654e-012
-1.54551e-014
8.8366
1.48889e-012
1.79859e-014
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(~li ke~li hood) # Param
Fitted Model -135.848 4
AIC
279.697
Data Summary
CLASS
C
F
I U
Total
Expected
Response
DOSE
0
46
1
3 0
50
3.45
13
43
4
2 1
50
7.40
32
32
16
2 0
50
13.53
Mi ni mum
observation time
for F tumor context =
31
July 2010
C-12
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
BMDU =
10.11B7
5.75142
13.1199
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
3.12363
0.640904
4.05212
July 2010
C-13
DO NOT CITE OR QUOTE
-------
Incidental Risk: F HEM3inc 2s
Dose= 0.00 Dose= 12.80
Dose = 32.00
Time
Figure C-3. Female mice, hemangiomas and hemangiosarcomas in all organs; high dose
dropped, all tumors considered incidental. Details below.
July 2010
C-14
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-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)Ac *
(beta_0+beta_l*doseAl+beta_2*doseA2)}
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
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 Specified
beta_0 = 0.000428228
beta_l = 0
beta_2 = 2.52747e-006
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
1
-0.4
beta_2
-0.4
1
Variable
c
beta_0
beta_l
beta_2
Estimate
1
0.000792254
0
4.54142e-006
Parameter Estimates
Std. Err.
NA
0.000500484
NA
1.85042e-006
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.000188678
9.14653e-007
0.00177319
8.16818e-006
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(~li kelihood) # Param
Fitted Model -65.8122 4
AIC
139.624
Data Summary
CLASS
C
F
I
U
Total
Expected Response
DOSE
0
46
0
4
0
50
3.74
13
43
0
6
1
50
6.57
32
32
0
18
0
50
17.56
July 2010
C-15
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
14.9357
11.0629
19.8583
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
4.61294
2.0194
6.12873
July 2010
C-16
DO NOT CITE OR QUOTE
-------
Incidental Risk: F HARD 1s
Dose = 0.00
Dose = 12.80
_Q
03
_Q
O
CD
O
cm
o
o
o
40 60 80 100
Time
_Q
03
_Q
O
CD
O
CSI
o
o
o
Time
Dose = 32.00
Dose = 80.00
-Q
03
_Q
O
CD
O
CM
O
O
O
03
_Q
O
CD
O
CM
O
O
O
Time
Time
Figure C-4. Female mice, Harderian gland tumors. Details below.
July 2010
C-17
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\msw\F_HARD_ls. (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 120
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
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 2.8BBBB
t_0 = 0 Specified
beta_0 = 1.02152e-007
beta_l = 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_l
c 1-1-1
beta_0 -1 1 0.99
beta_l -1 0.99 1
Variable
c
beta_0
beta_l
Estimate
2.9B861
6.26114e-008
4.59946e-009
Parameter Estimates
Std. Err.
2.46009
7.18253e-007
5.01418e-008
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-1.88B07 7.7603
-1.34514e-006 1.47036e-006
-9.36766e-008 1.02876e-007
Log(1 i ke~li hood) # Param AIC
Fitted Model -58.2663 3 122.533
Data Summary
CLASS
C
F
I
U
Total
Expected
DOSE
0
48
0
2
0
50
2.32
13
45
0
5
0
50
3.71
32
47
0
3
0
50
4.26
80
41
0
9
0
50
8.73
July 2010
C-18
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
BMDU =
27.0825
12.614
85.8726
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
2.58B4
1.20B27
8.04772
July 2010
C-19
DO NOT CITE OR QUOTE
-------
Incidental Risk: F MAMM 1s
Female mice, mammary gland tumors. Details below.
C-20
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\msw\F_MAMM_ls. (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl)}
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
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 1.0B0B
t_0 = 0 Specified
beta_0 = 0.000643678
beta_l = 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_l
beta_0
1
-0.5 7
beta_l
-0.5 7
1
Variable
c
beta_0
beta_l
Estimate
1
0.000740811
4.96148e-005
Parameter Estimates
Std. Err.
NA
0.000512B45
2.12095e-005
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.00026BB68
8.04497e-006
0.00174499
9.11846e-005
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(~li kelihood) # Param AIC
Fitted Model -87.9599 3 181.92
Data Summary
CLASS
C
F
I
U
Total
Expected
DOSE
0
46
0
3
1
50
3.50
IB
43
0
6
1
50
5.93
32
39
0
11
0
50
8.48
80
36
0
14
0
50
15.68
July 2010
C-21
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
MDU =
20.419
14.0543
38.5881
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
1.94776
1.34101
3.71557
July 2010
C-22
DO NOT CITE OR QUOTE
-------
Incidental Risk: F FORST 2s fix
Dose = 0.00
Dose = 12.80
-Q
-Q
o
80 100
Time
Time
Dose = 32.00
Dose = 80.00
-Q
-Q
o
Time
Time
Figure C-6. Female mice, forestomacFigure C-6: Female mice, forestomach tumors.
Details below.
July 2010
C-23
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-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)Ac *
(beta_0+beta_l*doseAl+beta_2*doseA2)}
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
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 = 4.125BB Specified
t_0 = 0 Specified
beta_0 = 5.01708e-011
beta_l = 0
beta_2 = 1.09429e-013
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
1
-0. IB
beta_2
-0. IB
1
Variable
beta_0
beta_l
beta_2
Estimate
5.01701e-011
0
1.0943e-013
Parameter Estimates
Std. Err.
7.09515e-011
NA
8.36829e-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
-8.88924e-011
-5.45854e-014
Upper Conf. Limit
1.89233e-010
2.73445e-013
Log(~li kel i hood) # Param AIC
Fitted Model -19.5963 3 45.1926
Data Summary
CLASS
C
F
I
u
Total
Expected
DOSE
0
49
0
1
0
50
0.46
13
49
0
0
1
50
0.50
32
49
0
0
0
49
0.68
80
46
0
4
0
50
3.35
July 2010
C-24
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
BMDU =
67.812
46.B2B
122.222
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
20.9439
5.69172
36.9312
July 2010
C-25
DO NOT CITE OR QUOTE
-------
Incidental Risk: F LIV 1s
Figure C-7. Female mice, hepatocellular adenomas and carcinomas. Details below.
July 2010
C-26
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: F_LIV_ls.(a)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl)}
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:
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_l = 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_l
beta_0
-1
1
1
beta_l
-1
1
1
Variable
c
beta_0
beta_l
Estimate
4.15974
2.70373e-009
1.01083e-010
Parameter Estimates
Std. Err.
1.3308
1.67272e-008
5.8692e-010
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
1.55141
-3.00809e-008
-1.04926e-009
6.76806
3.54884e-008
1.25142e-009
Log(~li ke~li hood) # Param AIC
Fitted Model -119.227 3 244.454
Data Summary
CLASS
C
F
I
U
Total
Expected
DOSE
0
30
0
20
0
50
21.63
13
22
0
26
1
49
22.96
32
30
0
20
0
50
20.68
80
20
0
30
0
50
30.79
Response
July 2010
C-27
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
BMDU =
4.24297
2.44688
8.51315
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD
BMDL
BMDU
0.404737
0.233408
0.818725
July 2010
C-28
DO NOT CITE OR QUOTE
-------
Incidental Risk: F SKIN 1s
Figure C-8. Female mice, skin sarcomas. Details below.
July 2010
C-29
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-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)Ac *
(beta_0+beta_l*doseAl)}
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:
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_l = 6.08721e-006
user,
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
and do not appear in the correlation matrix )
c beta_l
c 1-1
beta_l -1 1
Parameter Estimates
95.0% Wald Confidence Interval
Upper Conf. Limit
4.02115
9.28561e-005
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Variable
c
beta_0
beta_l
Estimate
1.56405
0
7.77467e-006
Std. Err.
1.25364
NA
4.34097e-005
Lower Conf. Limit
-0.893041
-7.73067e-005
Log(~li kelihood) # Param AIC
Fitted Model -87.4625 3 180.925
Data Summary
CLASS
C
F
I
U
Total
Expected
DOSE
0
50
0
0
0
50
0.00
13
38
0
11
1
50
5.59
32
39
0
11
0
50
10.58
80
32
0
18
0
50
22.43
July 2010
C-30
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
BMDU =
9.48956
7.1844
14.5757
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD
BMDL
BMDU
0.905208
0.665B24
1.4015
July 2010
C-31
DO NOT CITE OR QUOTE
-------
Incidental Risk: F_Zymb_1s05
Dose = 0.00 Dose = 12.80
-Q
-Q
o
00
O -J
o _
o
o
O _|
"T"
20
"T"
40
"T"
60
"T"
80
100
Time
Time
Figure C-9. Female mice, Zymbal's gland tumors. Details below.
July 2010
C-32
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemicals\ch~loroprene\msw\F_Zymb_ls05 . (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl)}
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
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_l = 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_l
1
-1
beta_l
-1
1
Variable
c
beta_0
beta_l
Estimate
1.09674
0
B.90733e-006
Parameter Estimates
Std. Err.
4.17B94
NA
7.24422e-005
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-7.08402 9.27751
-0.0001B8077
0.000145891
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(~li kel i hood)
# Param
AIC
Fi tted
Model
-
12.6107
3
31.2214
Data
. Summary
CLASS
C
F
I
U
Total
Expected Response
DOSE
0
50
0
0
0
50
0.00
13
50
0
0
0
50
0.36
32
50
0
0
0
50
0.76
80
47
0
3
0
50
1.90
July 2010
C-33
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.05
0.9
TIME
104
BMD =
BMDL =
BMDU =
80.5411
22.4657
255.715
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
15.7811
5.75828
50.0819
July 2010
C-34
DO NOT CITE OR QUOTE
-------
Incidental Risk: M LUNG 1s
Dose = 0.00
Dose = 12.80
CO
o
CD
O
csi
o
o
o
_Q
<0
_Q
O
00
o
CD
O
CSI
o
o
o
Time
Time
Figure C-10. Male mice, alveolar/bronchiolar tumors. Details below.
July 2010
C-35
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\msw\M_LUNG_ls. (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl)}
The parameter betas are restricted to be positive
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
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 B. 4
t_0 = 0 Specified
beta_0 = 5.3339e-008
beta_l = 5.89044e-009
user,
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
and do not appear in the correlation matrix )
c
c
beta_0
beta_l
beta_0
-1
1
1
beta_l
-1
1
1
Parameter Estimates
Variable
Estimate
Std. Err.
c
3.46155
1.29734 0.918807
beta.
.0
4.00939e-008
2
.41701e-
007 -4.33631e-007
beta.
.1
4.46048e-009
2
.61655e-
008 -4.6823e-008
Log(1 i ke~li hood)
# Param
AIC
Fitted Model
-104.927
3
215.855
Data Summary
CLASS
C
F I
U
Total
Expected Response
DOSE
0
37
0 13
0
50
14.02
13
22
0 28
0
50
26.63
32
14
0 36
0
50
33.92
80
7
0 43
0
50
44.57
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
6.0043
5.13819e-007
5.5744e-008
July 2010
C-36
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
TIME
INCIDENTAL
EXTRA
0.1
0.9
104
BMD =
BMDL =
BMDU =
2.46168
1.86129
B.465B4
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
TIME
INCIDENTAL
EXTRA
0.01
0.9
104
BMD =
BMDL =
BMDU =
0.2B482
0.178411
0.B218B7
July 2010
C-37
DO NOT CITE OR QUOTE
-------
Incidental Risk: M_HEM_3s
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose = 0.00 Dose = 12.80
CO
o
CO
o
¦§ °
CM
O
o
o
_Q
03
_Q
O
CO
o
CO
o
CM
O
O
O
Time
Time
Dose = 32.00
Dose = 80.00
CO
o
CO
o
CM
o
o
o
n 1 1 r
20 40 60 80
_Q
03
_Q
o
CO
o
CO
o
CM
O
o
o
100
i 1 1 r
20 40 60 80
100
Time
Time
Figure C-ll. Male mice, hemangiomas and hemangiosarcomas; hemangiosarcomas
occurring before termination considered fatal. Details below.
July 2010
C-38
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M_HEM_3s.(a)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl+beta_2*doseA2+beta_3*doseA3)}
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:
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 = 7.33333
t_0 = 0 Specified
beta_0 = 2.92735e-016
beta_l = 1.24661e-017
beta_2 = 5.74518e-040
beta_3 = 1.93026e-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_l
c nan nan
beta_l nan nan
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
c 13.2483 nan nan nan
beta_0 0 NA
beta_l 3.78184e-029 nan nan nan
beta_2 0 NA
beta_3 0 NA
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(~li ke~li hood)
# Param
AIC
Fi tted
Model
-
537.427
5
1084.85
Data
Summary
CLASS
C
F
I
U
Total
Expected Response
DOSE
0
47
0
3
0
50
0.00
13
36
6
8
0
50
8.30
32
27
16
7
0
50
12.19
80
29
13
8
0
50
20.33
Minimum observation time for F tumor context = 65
July 2010 C-39 DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD
BMDL
BMDU
2.58208
B.B4052
5.94514
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD
BMDL
BMDU
0.50B858
0.B18652
0.567106
July 2010
C-40
DO NOT CITE OR QUOTE
-------
Incidental Risk: M HEM 3s inc
Dose= 0.00 Dose= 12.80
Figure C-12. Male mice, hemangiomas and hemangiosarcomas; all tumors considered
incidental. Details below.
July 2010
C-41
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\msw\M_HEM_3s_inc. (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl+beta_2*doseA2+beta_3*doseA3)}
The parameter betas are restricted to be positive
Dependent variable = CLASS
Independent variables = DOSE, TIME
Total number of observations = 10B
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
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
user,
Default Initial Parameter Values
c = B.88235
t_0 = 0 Specified
beta_0 = 1.93573e-009
beta_l = 2.00936e-010
beta_2 = 0
beta_3 = 0
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
and do not appear in the correlation matrix )
c
beta 0
beta 1
c
1
-1
-1
beta_0
-1
1
0.99
beta_l
-1
0.99
1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
c 3.87398 1.89771 0.154536 7.59343
beta_0 2.01294e-009 1.78623e-008 -3.29965e-008 3.70224e-008
beta_l 2.08717e-010 1.80083e-009 -3.32084e-009 3.73828e-009
beta_2 0 NA
beta_3 0 NA
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(1 i ke~li hood)
# Param
AIC
Fi tted
Model
-
109.463
5
228.926
Data
Summary
CLASS
C
F
I
U
Total
Expected Response
DOSE
0
47
0
3
0
50
5.28
13
36
0
14
0
50
11.12
32
27
0
23
0
50
15.86
80
29
0
21
0
50
27.21
July 2010 C-42 DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
BENCHMARK DOSE COMPUTATION
RISK RESPONSE = INCIDENTAL
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
EXTRA
0.01
0.9
TIME
104
TIME
104
BMD
BMDL
BMDU =
7.74767
5.BB82B
12.7663
BMD =
BMDL =
BMDU =
0.73905
0.509228
1.21647
July 2010
C-43
DO NOT CITE OR QUOTE
-------
Incidental Risk: M_HARD_3s
Dose = 0.00
Dose = 12.80
Dose = 32.00 Dose = 80.00
Figure C-13. Male mice, Harderian gland tumors. Details below.
July 2010
C-44
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-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)Ac *
(beta_0+beta_l*doseAl+beta_2*doseA2+beta_3*doseA3)}
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
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 = 1.53577e-010
beta_l = 1.52041e-011
beta_2 = 0
beta_3 = 0
user,
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
and do not appear in the correlation matrix )
c beta_0 beta_l
c 1-1-1
beta_0 -1 1 1
beta_l -1 1 1
Variable
c
beta_0
beta_l
beta_2
beta_3
Estimate
5.57459
3.25883e-013
3.598e-014
0
0
Parameter Estimates
Std. Err.
3.19215
4.84471e-012
5.25235e-013
NA
NA
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-0.681904
-9.16957e-012
-9.93463e-013
11.8311
9.82133e-012
1.06542e-012
Log(~li kel i hood) # Param AIC
Fitted Model -73.6639 5 157.328
Data Summary
CLASS
C
F
I
U
Total
Expected
DOSE
0
48
0
2
0
50
2.29
13
45
0
5
0
50
5.18
32
40
0
10
0
50
7.40
80
38
0
12
0
50
14.04
July 2010
C-45
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
TIME
INCIDENTAL
EXTRA
0.1
0.9
104
BMD =
BMDL =
BMDU =
16.6911
10.4645
35.082
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
TIME
INCIDENTAL
EXTRA
0.01
0.9
104
BMD =
BMDL =
BMDU =
1.59216
0.998471
5.03875
July 2010
C-46
DO NOT CITE OR QUOTE
-------
Incidental Risk: M KIDN 1s
Dose = 0.00
Dose = 12.80
-Q
-Q
o
Time
Time
Figure C-14. Male mice, renal tubule tumors. Details below.
July 2010
C-47
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\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)Ac *
(beta_0+beta_l*doseAl)}
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
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_l = 5.87389e-013
user,
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
and do not appear in the correlation matrix )
c beta_l
c 1-1
beta_l -1 1
Variable
c
beta_0
beta_l
Estimate
6.09231
0
2.03124e-015
Parameter Estimates
Std. Err.
4.64814
NA
4.3366e-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
-8.29646e-014
15.2025
8.70271e-014
Log(~li kelihood) # Param AIC
Fitted Model -41.0033 3 88.0066
Data Summary
CLASS
C
F
I
U
Total
Expected
DOSE
0
50
0
0
0
50
0.00
13
47
0
2
1
50
1.95
32
47
0
3
0
50
3.70
80
41
0
9
0
50
8.34
July 2010
C-48
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.1
0.9
TIME
104
BMD =
BMDL =
BMDU =
26.7011
16.4536
47.1278
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
2.54702
1.56959
4.49547
July 2010
C-49
DO NOT CITE OR QUOTE
-------
Incidental Risk: M FORST 1s
Dose = 0.00 Dose = 12.80
Figure C-15. Male mice, forestomach tumors. Details below.
July 2010
C-50
DO NOT CITE OR QUOTE
-------
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using don1p2-intv, (c) by P. Spellucci
Input Data File: M:\_chemica1s\ch~loroprene\msw\M_FORST_ls. (d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseAl)}
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
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_l = 2.06761e-006
user,
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
and do not appear in the correlation matrix )
c beta_0 beta_l
c 1-1-1
beta_0 -1 1 1
beta_l -1 1 1
Variable
c
beta_0
beta_l
Estimate
1.2938
2.8702e-005
2.79278e-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
-0.00103109
-9.89466e-005
9.31165
0.0010885
0.000104532
Log(~li ke~li hood) # Param AIC
Fitted Model -30.8413 3 67.6827
Data Summary
CLASS
C
F
I
U
Total
Expected
DOSE
0
49
0
1
0
50
0.54
13
48
0
0
2
50
1.20
32
47
0
2
1
50
2.03
80
45
0
5
0
50
4.24
July 2010
C-51
DO NOT CITE OR QUOTE
-------
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.05
0.9
TIME
104
BMD =
BMDL =
BMDU =
45.1225
22.7599
157.OBI
BENCHMARK DOSE COMPUTATION
RISK RESPONSE
RISK TYPE
SPECIFIED EFFECT =
CONFIDENCE LEVEL =
INCIDENTAL
EXTRA
0.01
0.9
TIME
104
BMD =
BMDL =
BMDU =
8.8412B
4.46001
30.7684
July 2010
C-52
DO NOT CITE OR QUOTE
-------
Table C-4. Summary of human equivalent composite cancer risk values
estimated by R/BMDR, based on male and female mouse tumor incidence (NTP,
1998)
Tumor Site
BMDLoi
(ppm)
BMDoi
(ppm)
Risk value® at:
sd
sd2
Proportion of total
variance
BMDoi
(/ppm)
BMDLoi
(/ppm)
Female mice
Lung (systemic dosimetry)
8.65E-02"
1.14E-01
8.76E-02
1.16E-01
1.70E-02
2.88E-04
0.60
Lung (portal-of-entry
dosimetry)
3.57E-01
4.68E-01
2.14E-02
2.80E-02
4.04E-03
1.63E-05
0.08
Hemangiomas,
hemangiosarcomas (fatal)
6.41E-01
3.12E+00
3.20E-03
1.56E-02
7.54E-03
5.68E-05
0.12
0.27
Harderian gland
1.20E+00
2.58E+00
3.87E-03
8.31E-03
2.70E-03
7.28E-06
0.02
0.03
Mammary gland
adenomas, carcinomas,
adenacanthomas
1.34E+00
1.95E+00
5.13E-03
7.46E-03
1.41E-03
1.99E-06
0.00
0.01
Forestomach
5.69E+00
2.09E+01
4.77E-04
1.76E-03
7.78E-04
6.05E-07
0.00
0.00
Hepatocellular adenomas,
carcinomas
2.33E-01
4.05E-01
2.47E-02
4.28E-02
1.10E-02
1.22E-04
0.25
0.58
Skin
6.65E-01
9.05E-01
1.10E-02
1.50E-02
2.42E-03
5.86E-06
0.01
0.03
Zymbal's gland
5.76E+00
1.58E+01
6.34E-04
1.74E-03
6.70E-04
4.50E-07
0.00
0.00
Sum, risk values at BMDra (/ppm):
1.367E-01
7.043E-02
Sum, sd2:
4.829E-04
2.109E-04
Human equivalent sum of risk values (/(|ig/m3))c:
2.122E-04
1.093E-04
Composite
sd":
2.198E-02
1.452E-02
Upper bound on sum of risk estimates (/ppm)e:
1.729E-01
9.432E-02
Human equivalent upper bound on sum of risk estimates (/(|ig/m3)):
2.683E-04
1.464E-04
Male mice
Lung (systemic dosimetry)
1.78E-01
2.35E-01
5.61E-02
4.26E-02
8.19E-03
6.70E-05
0.54
lung (portal-of-entry)
7.31E-01
9.63E-01
1.37E-02
1.04E-02
2.00E-03
3.99E-06
0.07
Hemangiomas,
hemangiosarcomas (fatal)
3.19E-01
5.04E-01
3.14E-02
1.98E-02
7.01E-03
4.92E-05
0.40
0.81
Forestomach
4.46E+00
8.84E+00
2.24E-03
1.13E-03
6.75E-04
4.56E-07
0.00
0.01
Harderian gland
9.98E-01
1.59E+00
1.00E-02
6.28E-03
2.27E-03
5.15E-06
0.04
0.08
Kidney
1.57E+00
2.55E+00
6.37E-03
3.93E-03
1.49E-03
2.21E-06
0.02
0.04
Sum, risk values at BMD0i (/ppm):
7.377E-02
4.157E-02
Sum, sd2:
1.240E-04
6.098E-05
Human equivalent sum of risk values (/(|ig/m3)):
1.145E-04
6.452E-05
Copmposite
sd:
1.114E-02
7.809E-03
Upper bound on sum of risk estimates (/ppm):
9.209E-02
5.442E-02
Human equivalent upper bound on sum of risk estimates (/(|xg/m3)):
1.429E-04
8.445E-05
a Risk value=0.01/POD
'Summary statistics in bold were calculated using the "systemic" entries. The other summary statistics were calculated using the "portal-of-
entry" estimate for lung tumors and all entries for the other tumor sites.
c Human equivalent estimates were adjusted for continuous exposure and converted to ng/m3, by dividing by 6/24 (hours) and 3.62 mg/m3.
11 Composite SD = (Sum, SD2)05
e Upper bound on the composite risk estimate = Sum of MLE cancer risks + 1.645 x Composite SD.
July 2010
C-1
DO NOT CITE OR QUOTE
------- |