DRAFT - DO NOT CITE OR QUOTE
EPA/635/R-10/006
www.epa.gov/iris
£EPA
TOXICOLOGICAL REVIEW
OF
BENZO[a]PYRENE
(CAS No. 50-32-8)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
June 2011
NOTICE
This document is an Interagency Science Consultation draft. This information is distributed
solely for the purpose of pre-dissemination peer review under applicable information quality
guidelines. It has not been formally disseminated by EPA. It does not represent and should not
be construed to represent any Agency determination or policy. It is being circulated for review
of its technical accuracy and science policy implications.
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document is a preliminary draft for review purposes only. This information is
distributed solely for the purpose of pre-dissemination peer review under applicable information
quality guidelines. It has not been formally disseminated by EPA. It does not represent and
should not be construed to represent any Agency determination or policy. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
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DRAFT - DO NOT CITE OR QUOTE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
39
40
41
42
43
44
45
46
47
48
49
50
CONTENTS TOXICOLOGICAL REVIEW OF BENZO[a]PYRENE (CAS No. 50-32-8)
LIST OF TABLES VII
LIST OF FIGURES X
LIST OI ABBREVIATIONS AM) ACRONYMS XII
FOREWORD XVI
AUTHORS, CONTRIBUTORS, AND REVIEWERS XVII
1. INTRODUCTION 1
2. CHEMICAL AM) PHYSICAL INFORMATION 3
3. TOXICOKINETICS 9
3.1. ABSORPTION 9
3.1.1. Inhalation Exposure 9
3.1.1.1. Inhalation Exposure in Humans 9
3.1.1.2. Inhalation Exposure in Animals 11
3.1.2. Oral Exposure 17
3.1.2.1. Oral Exposure in Humans 18
3.1.2.2. Oral Exposure in Animals 18
3.1.3. Dermal Exposure 20
3.1.3.1. Dermal Exposure in Humans 20
3.1.3.2. Dermal Exposure in Animals 23
3.1.4. Other Types of Exposure 25
3.2. DISTRIBUTION 26
3.2.1. Inhalation Exposure 26
3.2.2. Oral Exposure 28
3.2.3. Dermal and Other Exposures 29
3.3. METABOLISM 32
3.3.1. Phase I Metabolism 34
3.3.1.1. CYP450-dependentReactions 34
3.3.1.2. Non-CYP-related metabolic pathways 38
3.3.2. Phase II Metabolism 40
3.3.3. Tissue-specific Metabolism 43
3.3.3.1. Respiratory Tract Tissues 43
3.3.3.2. GI Tract and Liver Tissues 47
3.3.3.3. Skin 48
3.3.3.4. Reproductive Tissues and Fetal Metabolism 50
3.3.3.5. Other Tissues 53
3.4. ELIMINATION 53
3.4.1. Inhalation Exposure 53
3.4.2. Oral Exposure 56
3.4.3. Dermal Exposure and Other Exposure Routes 57
3.5. PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODELS 59
4. HAZARD IDENTIFICATION 64
4.1. HUMANS STUDIES 64
4.1.1. Sources of Human Exposure 64
4.1.2. Biomonitoring of benzo[a]pyrene Exposure and Effects In Humans 64
in
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
4.1.2.1. Urinary Excretion of benzo[a]pyrene 65
4.1.2.2. Adduct Formation with DNA and Protein 66
4.1.2.2.1. DNA Adduct measures in blood 69
4.1.2.2.2. Adduct measures in reproductive tissues 72
4.1.2.3. BenzofaJpyrene-JraiwceJ Cytogenetic Damage 73
4.1.3. Epidemiologic Findings in Humans 75
4.1.3.1. Molecular Epidemiology and Case-Control Cancer Studies 76
4.1.3.2. Cohort Cancer Studies 77
4.1.3.3. Noncancer Disease Caused by benzo[a]pyrene 81
4.2. SUB CHRONIC AND CHRONIC STUDIES AND CANCER BIO ASSAYS IN
ANIMALS—ORAL, INHALATION, AND DERMAL 85
4.2.1. Oral 85
4.2.1.1. Subchronic Studies 85
4.2.1.2. Chronic Studies and Cancer Bioassays 93
4.2.2. Inhalation 106
4.2.2.1. Short-term and Subchronic Studies 106
4.2.2.2. Chronic Studies and Cancer Bioassays 106
4.2.3. Dermal Exposure 108
4.2.3.1. Skin-Tumor Initiation-Promotion Assays 108
4.2.3.2. Carcinogenicity Dermal Bioassays 109
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL, INHALATION, AND
DERMAL 118
4.3.1. Oral 118
4.3.2. Inhalation 124
4.4. OTHER DURATION OR ENDPOINT-SPECIFIC STUDIES 129
4.4.1. Acute Neurological Studies 129
4.4.2. Immunological Studies 133
4.4.2.1. Oral Exposure Immunological Studies 133
4.4.2.2. Inhalation Exposure Immunological Studies 134
4.4.2.3. Other Exposure Route Immunological Studies 134
4.4.2.4. Other Exposure Route DevelopmentalImmunotoxicy 135
4.4.3. Cancer Bioassays (Other Routes of Exposure) 135
4.4.4. Atherogenesis Studies 136
4.4.5. Reproductive Studies (Other Routes of Exposure) 138
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MOA 139
4.5.1. Genotoxicity 140
4.5.2. Metabolic pathways 161
4.5.3. Mechanistic Studies- Mutagenesis and Tumor Initiation 164
4.5.3. Tumor Promotion and Progression 166
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS 171
4.6.1. Oral 171
4.6.2. Inhalation 177
4.6.3. Dermal 180
4.6.4. Mode-of-Action Information 180
4.6.4.1. Forestomach Lesions from Oral Exposure 180
4.6.4.2. Immune System Effects 180
4.6.4.3. Developmental and Reproductive Toxicity Effects 182
4.7. EVALUATION OF CARCINOGENICITY 185
4.7.1. Summary of Overall Weight-of-Evidence 185
iv
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence 185
4.7.3. Mode-of-Action Information 187
4.7.3.1. Hypothesized MO A 187
4.7.3.2. Experimental Support for the Hypothesized MO A 188
4.7.3.2. Other Possible MO As 190
4.7.3.4. Conclusions About the Hypothesized Mode ofA ction 191
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES 193
4.8.1. Possible Childhood Susceptibility 193
4.8.2. Possible Gender Differences 196
4.8.3. Genetic Polymorphisms 198
5. DOSE-RESPONSE ASSESSMENTS 201
5.1. ORAL REFERENCE DOSE (RID) 201
5.1.1. Choice of Principal Study and Critical Effect—with Rationale and Justification ... 201
5.1.2. METHODS OF ANALYSIS 205
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs) 209
5.1.4. Previous RfD Assessment 210
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 210
5.2.1. Choice of Principal Study and Critical Effect—with Rationale and Justification ... 210
5.2.2. Methods of Analysis- Adjustment to a Human Equivalent Concentration (HEC) ... 212
5.2.3. RfC Derivation- Including Application of Uncertainty Factors (UFs) 214
5.2.4. Previous RfC Assessment 215
5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE AND INHALATION
REFERENCE CONCENTRATION 216
5.4. CANCER ASSESSMENT 217
5.4.1. Oral Exposure—Oral Slope Factor 217
5.4.1.1. Choice of Study/Data—with Rationale and Justification—Oral Exposure 217
5.4.1.2. Dose-response Data—Oral Exposure 218
5.4.1.3. Dose Adjustments and Extrapolation Method(s)—Oral Exposure 220
5.4.1.4. Oral Slope Factor Derivation 223
5.4.2. Inhalation Exposure—Inhalation Unit Risk 225
5.4.2.1. Choice of Study/Data—with Rationale and Justification—Inhalation Exposure
225
5.4.2.2. Dose-response Data—Inhalation Exposure 226
5.4.2.3. Dose Adjustments and Extrapolation Method(s)—Inhalation Exposure 227
5.4.2.4. Inhalation Unit Risk Derivation 227
5.4.3. Dermal Exposure—Dermal Slope Factor 227
5.4.3.1. Choice of Study/Data—with Rationale and Justification—Dermal Exposure . 227
5.4.3.2. Dose-response Data—Dermal Exposure 228
5.4.3.3. Dose Adjustments and Extrapolation Method(s)—Dermal Exposure 234
5.4.3.4. Dermal Slope Factor Derivation 235
5.4.3.5. Dermal Slope Factor Cross Species Scaling 236
5.4.4. Application of Age-Dependent Adjustment Factors 238
5.4.5. Uncertainties in Cancer Risk Values 240
5.4.5.1. Oral Slope Factor 240
5.4.5.2. Inhalation Unit Risk 240
5.4.5.3 Dermal Slope Factor 241
5.5.5. Previous Cancer Assessment 242
v
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE 244
6.1. HUMAN HAZARD POTENTIAL 244
6.2. DOSE RESPONSE 244
6.2.1. RID 244
6.2.2. RfC 246
6.2.3. Cancer 247
6.2.3.1. Cancer—Oral 247
6.2.3.2. Cancer—Inhalation 248
6.2.3.3. Cancer—Dermal 248
7. REFERENCES 250
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS
AND DISPOSITION 287
APPENDIX B. BENCHMARK DOSE MODELING RESULTS FOR NONCANCER 288
APPENDIX C. ADDITIONAL CALCULATIONS FOR THE RFC 327
APPENDIX D. TIME-TO-TUMOR MODELING FOR THE ORAL SLOPE FACTOR 330
APPENDIX E. TIME-TO-TUMOR MODELING FOR THE INHALATION UNIT RISK. .. 360
APPENDIX F. BENCHMARK MODELING FOR THE DERMAL SLOPE FACTOR 367
APPENDIX G. ADDITIONAL INFORMATION IN SUPPORT OF THE DERMAL SLOPE
FACTOR 395
APPENDIX H. ALTERNATIVE APPROACHES FOR CROSS-SPECIES SCALING OF THE
DERMAL SLOPE FACTOR 398
vi
DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 2-1. Physical properties and chemical identity of benzo[a]pyrene 4
Table 2-2. Benzo[a]pyrene Concentrations in Air 5
Table 2-3. Benzo[a]pyrene Levels in Food 6
Table 2-4. Levels of benzo[a]pyrene in Soil 7
Table 3-1. Distribution of benzo[a]pyrene in selected tissues of male rats following i.v. dosing
with 2 mg/kg 30
Table 3-2. Species differences in tracheobronchial benzo[a]pyrene metabolism 44
Table 3-3. Excretion of benzo[a]pyrene metabolites in several animal species 55
Table 3-4. Biliary excretion of benzo[a]pyrene metabolites in several species 56
Table 4-1. Studies of PAH-DNA adducts in human populations or tissues exposed to PAHs ... 67
Table 4-2. Exposure-related effects in male Wistar rats exposed to benzo[a]pyrene by gavage 5
days/week for 5 weeks 87
Table 4-3. Exposure-related effects in Wistar rats exposed to benzo[a]pyrene by gavage 5
days/week for 5 weeks 90
Table 4-4. Means ± SDa for liver and thymus weights in Wistar rats exposed to benzo[a]pyrene
by gavage 5 days/week for 90 days 92
Table 4-5. Incidences of exposure-related neoplasms in Wistar rats treated by gavage with
benzo[a]pyrene, 5 days/week, for 104 weeks 95
Table 4-6. Incidences of alimentary tract tumors in Sprague-Dawley rats chronically exposed to
benzo[a]pyrene in the diet or by gavage in caffeine solution 98
Table 4-7. Incidence of nonneoplastic and neoplastic lesions in female B6C3Fi mice fed
benzo[a]pyrene in the diet for up to 2 years 100
Table 4-8. Tumor incidence in oral exposure rodent cancer bioassays with limitations for
describing dose-response relationships for lifetime exposure to benzo[a]pyrene 102
Table 4-9. Incidence of respiratory and upper digestive tract tumors in male hamsters treated for
life with benzo[a]pyrene by inhalation 107
Table 4-10. Numbers of animals with pharynx and larynx tumors in male hamsters exposed by
inhalation to benzo[a]pyrene for life 108
Table 4-11. Skin tumor incidence and time of appearance in male C57L mice dermally exposed
to benzo[a]pyrene for up to 103 weeks 109
Table 4-12. Skin tumor incidence and time of appearance in male SWR, C3HeB, and A/He mice
dermally exposed to benzo[a]pyrene for life or until a skin tumor was detected 110
Table 4-13. Tumor incidence in female Swiss mice dermally exposed to benzo[a]pyrene for up
to 93 weeks Ill
Table 4-14. Skin tumor incidence in female NMRI and Swiss mice dermally exposed to
benzo[a]pyrene 112
Table 4-15. Skin tumor incidence in female NMRI mice dermally exposed to benzo[a]pyrene
114
Table 4-16. Skin tumor incidence in female NMRI mice dermally exposed to benzo[a]pyrene
114
Table 4-17. Skin tumor incidence and time of appearance in female CFLP mice dermally
exposed to benzo[a]pyrene for 104 weeks 115
Table 4-18. Skin tumor incidence in female NMRI mice dermally exposed to benzo[a]pyrene for
life 116
Table 4-19. Skin tumor incidence in male C3H /HeJ mice dermally exposed to benzo[a]pyrene
for 24 months 117
vii
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-20. Means ± SD for ovary weight in female SD-rats 119
Table 4-21. Reproductive effects in male and female CD-I F1 mice exposed in utero to
benzo[a]pyrene 121
Table 4-22. Effect of prenatal exposure to benzo[a]pyrene on indices of reproductive
performance in F1 female NMRI mice 123
Table 4-23. Pregnancy outcomes in female F344 rats treated with benzo[a]pyrene on GDs 11-21
by inhalation 125
Table 4-24. In vitro genotoxicity studies of benzo[a]pyrene in non-mammallian cells 141
Table 4- 25. In vitro genotoxicity studies of benzo[a]pyrene in mammalian cells 143
Table 4-26. In vivo genotoxicity studies of benzo[a]pyrene 148
Table 4-27. NOAELS and LOAELs for noncancer effects in animals repeatedly exposed to
benzo[a]pyrene by the oral route 172
Table 4-28. NOAELS and LOAELs for noncancer effects in animals repeatedly exposed to
benzo[a]pyrene by the inhalation route 179
Table 5-1. Means ± SD for ovary weight in female SD-rats 204
Table 5-2. Summary of BMDs and BMDLs for modeled noncancer effects following oral
exposure 207
Table 5-3. Pregnancy outcomes in female F344 rats treated with benzo[a]pyrene on GDs 11-21
by inhalation 212
Table 5-4. Incidence data for tumors in Wistar rats exposed to benzo[a]pyrene by gavage, 5
days/week for 104 weeks 220
Table 5-5. Incidence data for tumors in female B6C3Fi mice exposed to benzo[a]pyrene in the
diet for 104 weeks 220
Table 5-6. Human equivalent PODs and oral slope factors derived from multistage-Weibull
modeling of tumor incidence data at multiple tissue sites in Wistar rats and B6C3Fi mice
exposed to benzo[a]pyrene orally for 2 years 223
Table 5-7. Incidence of tumors in male hamsters exposed by inhalation to benzo[a]pyrene for
life 226
Table 5-8. Skin tumor incidence, benign or malignant, in C57L male mice dermally exposed to
benzo[a]pyrene 232
Table 5-9. Skin tumor incidence, benign or malignant in female Swiss or NMRI mice dermally
exposed to benzo[a]pyrene 233
Table 5-10. Skin tumor incidence, benign or malignant, in female CFLP mice dermally exposed
to benzo[a]pyrene 234
Table 5-11. Skin tumor incidence, benign or malignant, in male C3H /HeJ mice dermally
exposed to benzo[a]pyrene(Sivak et al., 1997) 234
Table 5.13. Application of ADAFs to benzo[a]pyrene cancer risk following a lifetime (70-year)
oral exposure 239
Table 5-14. Application of ADAFs to benzo[a]pyrene cancer risk following a lifetime (70-year)
inhalation exposure 239
Table 5-15. Application of ADAFs to benzo[a]pyrene cancer risk following a lifetime (70-year)
dermal exposure 240
Table B-l. Liver weight (±SD)a in male F344 rats administered benzo[a]pyrene by gavage for
90 days 288
Table B-2. BMD modeling results for increased liver weight in male rats, with BMR=10%... 288
Table B-3. Means ± SDa for thymus weight in male Wistar rats exposed to benzo[a]pyrene by
gavage 5 days/week for 90 days 292
Table B-4. Model predictions for decreased thymus weight in male Wistar rats—90 days 292
Table B-5. Means ± SDa for thymus weight in female Wistar rats exposed to benzo[a]pyrene by
viii
DRAFT - DO NOT CITE OR QUOTE
-------
gavage 5 days/week for 90 days 297
Table B-6. Model predictions for decreased thymus weight in female Wistar rats—90 days .. 297
Table B-7. Means ± SD for thymus weight in male Wistar rats exposed to benzo[a]pyrene by
gavage 5 days/week for 35 days 302
Table B-9. Exposure-related effects in male Wistar rats exposed to benzo[a]pyrene by gavage 5
days/week for 5 weeks 307
Table B-10. Model predictions for decreased spleen B-cells in male Wistar rats—35 days 307
Table B-l 1. Dose-response data for forestomach hyperplasia in Wistar rats and B6C3Fi rats
orally exposed to benzo[a]pyrene for 2 years 312
Table B-12. Summary of BMDs and BMDLs from the best fitting model forestomach
hyperplasia—oral exposure 313
Table B-13. Model predictions for forestomach hyperplasia in male Wistar rats in a 2-year study
313
Model predictions for forestomach hyperplasia in female Wistar rats in a 2-year
316
Model predictions for forestomach hyperplasia in female B6C3F1 mice in a 2-year
319
Means ± SDa for ovary weight in female SD-rats 323
Model predictions for decreased ovary weight in female SD-rats—60 days 323
Table D-l. Tumor incidence data, with time to death with tumor; male rats exposed by gavage to
Benzo[a]pyrene (Kroese et al., 2001) 330
Table D-2. Tumor incidence data, with time to death with tumor; female rats exposed by gavage
to Benzo[a]pyrene (Kroese et al., 2001) 332
Table D-3. Tumor incidence data, with time to death with tumor; female mice exposed to
Benzo[a]pyrene via diet (Beland and Culp, 1998) 334
Table D-4. Derivation of HEDs to use for BMD modeling of Wistar rat tumor incidence data
from Kroese et al. (2001) 336
Table D-5. Derivation of HEDs for BMD modeling of B6C3F1 female mouse tumor incidence
data from Beland and Culp (1998) 336
Table D-6. Summary of Model Selection and Modeling Results for best-fitting multistage-
Weibull models, using time-to-tumor data for rats (Kroese et al., 2001) 337
Table F-l. Summary of model selection and modeling results for best-fitting multistage models,
for multiple data sets of skin tumors in mice following dermal benzo[a]pyrene exposure . 367
Table G-l: Exposure methods for selected lifetime dermal exposure mouse cancer bioassays for
benzo[a]pyrene-induced skin tumors 396
Table H-l. Alternative approaches to cross-species scaling 401
Table B-l4
study..
Table B-l5
study..
Table B-l6
Table B-l7
IX
DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure 2-1. Structural formula of benzo[a]pyrene 3
Figure 3-1. Metabolic pathways for benzo[a]pyrene 33
Figure 3-2. The stereospecific activation of benzo[a]pyrene 35
Figure 4-1. Interaction of PAHs with the AhR 168
Figure 4-2. NOAELs and LOAELs for selected noncancer effects from repeated oral exposure
to benzo[a]pyrene 175
Figure 4-3. Proposed principal pathways and key events in the benzo[a]pyrene carcinogenic
MOA 188
Figure B-l. Fit of polynomial model to data on increased liver weight in male Wistar rats—90
days 289
Figure B-2. Fit of linear model (nonconstant variance) to data on decreased thymus weight in
male Wistar rats—90 days 293
Figure B-3. Fit of linear model (constant variance) to data on decreased thymus weight in
female Wistar rats—90 days 298
Table B-8. Model predictions for decreased thymus weight in male Wistar rats—35 days 302
Figure B-4. Fit of linear model (constant variance) to data on decreased thymus weight in male
Wistar rats—35 days 303
Figure B-5. Fit of linear model to data on decreased spleen B-cells in male Wistar rats—35
days 308
Figure B-6. Fit of log logistic model to data on forestomach hyperplasia in male Wistar rats in a
2-year study 314
Figure B-7. Fit of log logistic model to data on forestomach hyperplasia in female Wistar rats in
a 2-year study 317
Figure B-8. Fit of log logistic model to data on forestomach hyperplasia in female B6C3Fi mice
in a 2-year study 320
Figure B-9. Fit of linear/polynomial (1°) model to data on decreased ovary weight 324
Figure C-l. Human Fractional Deposition 327
Figure C-2. Rat Fractional Deposition 329
Figure F-l. Fit of multistage model to skin tumors in C57L mice exposed dermally to
benzo[a]pyrene (Poel, 1959); graph and model output 368
Figure F-2. Fit of multistage model to skin tumors in female Swiss mice exposed dermally to
benzo[a]pyrene (Roe et al., 1970); graph and model output 370
Figure F-3. Fit of multistage model to skin tumors in female NMRI mice exposed dermally to
benzo[a]pyrene (Schmidt et al., 1973); graph and model output 372
Figure F-4. Fit of multistage model to skin tumors in female Swiss mice exposed dermally to
benzo[a]pyrene (Schmidt et al., 1973); graph and model output 374
Figure F-5. Fit of multistage model to skin tumors in female NMRI mice exposed dermally to
benzo[a]pyrene (Schmahl et al., 1977); graph and model output 376
Figure F-6. Fit of multistage model to skin tumors in female NMRI mice exposed dermally to
benzo[a]pyrene (Habs et al., 1980); graph and model output 378
Figure F-8. Fit of multistage model to skin tumors in female CFLP mice exposed dermally to
benzo[a]pyrene (Grimmer et al, 1983); graph and model output 383
Figure F-9. Fit of multistage model to skin tumors in female CFLP mice exposed dermally to
benzo[a]pyrene (Grimmer et al., 1984); graph and model output 385
Figure F-9. Fit of log-logistic model to skin tumors in female CFLP mice exposed dermally to
benzo[a]pyrene (Grimmer et al., 1984); graph and model output 388
x
DRAFT - DO NOT CITE OR QUOTE
-------
Figure F-l 1. Fit of multistage model to skin tumors in female CFLP mice exposed dermally to
benzo[a]pyrene (Grimmer et al., 1984), highest dose dropped; graph and model output.... 390
Figure F-12. Fit of multistage model to skin tumors in female CFLP mice exposed dermally to
benzo[a]pyrene (Sivak et al., 1997); graph and model output 393
XI
DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
1-OH-Py
1-hydroxypyrene
3-MC
3 -m ethyl cholanthrene
8-OHdG
8-hydroxydeoxyguanosine
ADAFs
age-dependent adjustment factors
AFC
antibody forming cells
Ah
aryl hydrocarbon
AHH
aryl hydrocarbon hydroxylase
AhR
Ah receptor
AhRE
AhR-responsive element
AhRR
AhR repressor
AIC
Akaike's Information Criterion
AKR
aldo-keto reductase
ALT
alanine aminotransferase
Arnt
Ah receptor nuclear translocator
AST
serum aspartate transaminase
ATSDR
Agency for Toxic Substances and Disease Registry
AUC
area under the curve
benzo[a]pyrene
benzo[a]pyrene
BeP
benzo[e]pyrene
BMD
benchmark dose
BMDL
benchmark dose, 95% lower bound
BMDS
Benchmark Dose Software
BMR
benchmark response
BPDE
benzo[a]pyrene-7,8-diol-9,10-epoxide
BPQ
benzo[a]pyrene-7,8-quinone
BRCA
breast cancer antigen
BrdU
bromodeoxyuridine
BSM
benzene-soluble matter
BUN
blood urea nitrogen
CA
chromosomal aberrations
CASRN
Chemical Abstracts Service Registry Number
CAT
chloramphenicol acetytransferase
CB
carbon black
CDK
cyclin-dependent kinase
CFU-GM
colony forming unit-granulocyte macrophage
CHL
Chinese hamster lung cells
CHO
Chinese hamster ovary cells
CI
confidence interval
CNS
central nervous system
Con A
Concanavalin A
CONSAAM
Conversational SAAM
COX
cyclooxygenase
CPDB
Cancer Potency Database
cSt
centi-Stoke
CTPV
coal tar pitch volatiles
CYP
cytochrome
xii DRAFT - DO NOT CITE OR QUOTE
-------
CYP450
cytochrome P450
dG
deoxyguanosine
dG-N2-BPDE
10P-(deoxyguanosin-N2-yl)-7p,8a,9a-trihydroxy-7,8,9,10-tetrahydro-
benzo[a]pyrene
DHH
dihydrodiol dehydrogenase
DMBA
7,12-dimethylbenzanthracene
DMSO
dimethyl sulfoxide
DNCB
2,4-dinitrochlorobenzene
DRE
dioxin-responsive element
ED
effective dose
EGFR
epidermal growth factor receptor
EH
epoxide hydrolase
ELISA
enzyme-linked immunosorbent assay
EPC
endothelial progenitor cells
EpRE
electrophile (or antioxidant) response element
ER
estrogen receptor
EROD
7-ethoxyresorufin-O-deethylase
ETS
environmental tobacco smoke
Fe2C>3
ferrous oxide
FEL
frank effect level
Ga203
gallium oxide
GD
gestational day
GGT
y-glutamyl transferase
GI
gastrointestinal
GJIC
gap junctional intercellular communication
GLP
good laboratory practice
GM-CSF
granulocyte-macrophage colony stimulating factor
GNMT
glycine N-methyltransferase
GP
glycophorin
GSH
reduced glutathione
GST
glutathi one- S -transferase
hAR
human androgen receptor
HED
human equivalent dose
HF
human fibroblasts
HFC
high-frequency cells
HL
human lymphocytes
HPLC
high-performance liquid chromatography
hprt
hypoxanthine guanine phosphoribosyl transferase
Hsp
heat shock protein
IARC
International Agency for Research on Cancer
IC50
half maximal inhibitory concentration
IFN
interferon
Ig
immunoglobulin
IGF
insulin-like growth factor
IHD
ischemic heart disease
IL
interleukin
y-inf
gamma-interferon
i.p.
intraperitoneal
IRIS
Integrated Risk Information System
Xlll
DRAFT - DO NOT CITE OR QUOTE
-------
i.v.
intravenous
KLH
keyhole limpet hemocyanin
ko
knock-out
LALN
lung-associated lymph nodes
LC-MS
liquid chromatography-mass spectrometry
LDH
lactate dehydrogenase
LH
luteinizing hormone
LOAEL
lowest-observed-adverse-effect level
LPS
lipopolysaccharide
MAP
mitogen-activated protein
MCHC
mean cell hemoglobin concentration
MCP
monocyte-chemoattractant protein
M-CSF
macrophage colony stimulating factor
MGP
manufactured gas plant residue
MLR
mixed lymphocyte response
MMAD
mass median aerodynamic diameter
MN
micronucleus
MOA
mode of action
MPO
myeloperoxidase
NADH
nicotinamide adenine dinucleotide phosphate
NAT
N-acetyl transferase
NER
nucleotide excision repair
NF
naphthoflavone
NK
natural-killer
NMDA
N-methyl-D-aspartate
NO
nitrous oxide
NOAEL
no-ob served-adverse-effect level
NQO
NADPH:quinone oxidoreductase
NSAID
non-steroidal anti-inflammatory drug
NTP
National Toxicology Program
OR
odds ratio
PAH
polycyclic aromatic hydrocarbon
PBMC
peripheral blood mononuclear cell
PBPK
physiologically-based pharmacokinetic
PCNA
proliferating cell nuclear antigen
PCR
polymerase chain reaction
PHA
phytohemagglutinin
PHS
prostaglandin H synthase
PMN
polymorphonuclear leukocyte
PND
postnatal day
p.o.
per os
POD
point of departure
Py
pyrene
RBC
red blood cell
RfC
reference concentration
RfD
reference dose
RN
reaction network
ROS
reactive oxygen species
RR
relative risk
xiv
DRAFT - DO NOT CITE OR QUOTE
-------
RT-PCR
real-time or reverse transcriptase PCR
s.c.
subcutaneous
SAAM
Simulation, Analysis and Modeling
SAM
S-adenosylmethionine
see
squamous cell carcinoma
SCE
sister chromatid exchanges
SCE-H
SCE heterogeneity index
SD
standard deviation
SEM
standard error of the mean
SIR
standardized incidence ratio
SLRL
sex-linked recessive lethal
SMR
standardized mortality ratio
SNP
single nucleotide polymorphisms
SPF
specific pathogen-free
SRBC
sheep red blood cell
SSB
single strand break
TCDD
2,3,7,8-tetrachlorodibenzo-p-dioxin
TEF
toxicity (or toxic) equivalency factor
TGF
transforming growth factor
TK
thymidine kinase
TNF
tumor necrosis factor
TPA
12-O-tetradecanoylphorb ol-13 -acetate
TWA
time-weighted average
UDP
uridine diphosphate
UDPGA
UDP glucuronic acid
UDS
unscheduled DNA synthesis
UF
uncertainty factor
UGT
UDP-dependent glucuronosyltransferase
Vmax
maximum substrate turnover velocity
vSMC
vascular smooth muscle cell
WBC
white blood cells
WHO
World Health Organization
WT
wild type
WTC
World Trade Center
XP
xeroderma pigmentosum
XPA
xeroderma pigmentosum group A
XV
DRAFT - DO NOT CITE OR QUOTE
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to
benzo[a]pyrene (benzo[a]pyrene). It is not intended to be a comprehensive treatise on the
chemical or toxicological nature of benzo[a]pyrene.
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).
xvi
DRAFT - DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER
Kathleen Newhouse
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
CONTRIBUTORS
Christine Cai
Gene Ching-Hung Hsu
Glinda Cooper
Martin Gehlhaus
Scott Glaberman
Karen Hogan
Margaret Pratt
John Schaum
John Stanek
Suryanarayana Vulimiri
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
CONTRACTING SUPPORT
Heather Carlson-Lynch
Peter McClure
Kelly Salinas
Joe Santodonato
Julie Stickney
SRC, Inc.
North Syracuse, NY
George Holdsworth
Lutz W. Weber
Oak Ridge Institute for Science and Education
Oak Ridge, TN
Janusz Z. Byczkowski
Consultant
Fairborn, OH
Andrew Maier
Lynne Haber
Bernard Gadagbui
xvii
DRAFT - DO NOT CITE OR QUOTE
-------
Toxicology Excellence for Risk Assessment
Cincinnati, OH
Glenn Talaska
University of Cincinnati
Cincinnati, OH
REVIEWERS
This document has been provided for review to EPA scientists.
INTERNAL EPA REVIEWERS
Stephen Nesnow
National Health and Environmental Effects Research Laboratory
US Environmental Protection Agency
Research Triangle Park, NC
Rita Schoeny
Office of Water
U.S. Environmental Protection Agency
Washington, DC
xviii
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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
benzo[a]pyrene (benzo[a]pyrene). 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 (MOA). 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
"3
deleterious effects during a 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 exposure. Similarly, a plausible inhalation unit risk is
an upper bound on the estimate of risk per (j,g/m air breathed.
There is evidence in humans and animal studies, demonstrating an increased incidence of
skin tumors with increasing dermal exposure to polycyclic aromatic hydrocarbons (PAHs)
mixtures including benzo[a]pyrene or to benzo[a]pyrene alone. Thus this assessment for
benzo[a]pyrene derives a dermal slope factor; a quantitative risk estimate that is a plausible
upper bound on the estimate of risk per (j,g/day of dermal exposure.
Development of these hazard identification and dose-response assessments for
benzo[a]pyrene 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
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
the following: Guidelines for the Health Risk Assessment of Chemical Mixtures (U.S. EPA,
1986a), Guidelines for Mutagenicity Risk Assessment (U.S. EPA, 1986b), Recommendations for
and Documentation of Biological Values for Use in Risk Assessment (U.S. EPA, 1988),
Guidelines for Developmental Toxicity Risk Assessment (U.S. EPA, 1991), Interim Policy for
Particle Size and Limit Concentration Issues in Inhalation Toxicity (U.S. EPA, 1994a), Methods
for Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry
(U.S. EPA, 1994b), Use of the Benchmark Dose Approach in Health Risk Assessment (U.S. EPA,
1995), Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for
Neurotoxicity Risk Assessment (U.S. EPA, 1998), Science Policy Council Handbook. Risk
Characterization (U.S. EPA, 2000a), Benchmark Dose Technical Guidance Document (U.S.
EPA, 2000b), Supplementary Guidance for Conducting Health Risk Assessment of Chemical
Mixtures (U.S. EPA, 2000c), A Review of the Reference Dose and Reference Concentration
Processes (U.S. EPA, 2002), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens
(U.S. EPA, 2005b), Science Policy Council Handbook: Peer Review (U.S. EPA, 2006a), and A
Framework for Assessing Health Risks of Environmental Exposures to Children (U.S. EPA,
2006b).
The literature search strategy employed for this compound was based on the Chemical
Abstracts Service Registry Number (CASRN) and at least one common name. Any pertinent
scientific information submitted by the public to the IRIS Submission Desk was also considered
in the development of this document. The relevant literature was reviewed through December,
2010.
2
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
2. CHEMICAL AND PHYSICAL INFORMATION
Benzo[a]pyrene is a five-ring polycyclic aromatic hydrocarbon (PAH) (Figure 2-1). It is
a pale yellow crystalline solid with a faint aromatic odor. It is relatively insoluble in water and
has low volatility. Benzo[a]pyrene is released to the air from both natural and anthropogenic
sources and removed from the atmosphere by photochemical oxidation; reaction with nitrogen
oxides, hydroxy and hydroperoxy radicals, ozone, sulfur oxides, and peroxyacetyl nitrate; and
dry deposition to land or water. In air, benzo[a]pyrene is predominantly adsorbed to particulates
but may also exist as a vapor at high temperatures (NLM, 2010). The structural formula is
presented in Figure 2-1. The physical and chemical properties of benzo[a]pyrene are shown in
Table 2-1.
Benzo[a]pyrene
Figure 2-1. Structural formula of benzo[a]pyrene.
There is no known commercial use for benzo[a]pyrene and it is only produced as a
research chemical. It is found ubiquitously in the environment primarily as a result of
incomplete combustion emissions. It is released to the environment via both natural sources
(such as forest fires) and anthropogenic sources including stoves/furnaces burning fossil fuels
(especially wood and coal), motor vehicle exhaust, cigarettes, and various industrial combustion
processes (ATSDR, 1995). Benzo[a]pyrene is also found in soot and coal tars. Mahler et al.
(2005) has reported that urban run-off from asphalt-paved car parks treated with coats of coal-tar
emulsion seal could account for the majority of PAHs in many watersheds (Mahler et al., 2005).
Occupational exposure to PAHs occurs primarily through inhalation and skin contact during the
production and use of coal tar and coal tar-derived products, such as roofing tars, creoste and
asphalt (IARC, 2010). Chimney sweeping can result in exposure to benzo[a]pyrene
contaminated soot (ATSDR, 1995). As shown below in Table 2-2, benzo[a]pyrene exposure can
also occur to workers involved in the production of aluminum, coke, graphite, and silicon
carbide.
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
Table 2-1. Physical properties and chemical identity of benzo[a]pyrene
CASRN 50-32-8
Synonyms
Benzo[d,e,f]chrysene;
3,4-benzopyrene,
3,4-benzpyrene; benz[a]pyrene;
benzo[a]pyrene; BP
http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7CHEM
Melting point
179-179.3°C
O'Neil et al. (2001)
Boiling point
310-312°C at 10 mm Hg
O'Neil et al. (2001)
Vapor pressure, at 20°C
5 x 10"7 mmHg
Verschueren (2001)
Density
1.351 g/cm3
I ARC (1973)
Flashpoint (open cup)
No data
Water solubility at 25°C
1.6-2.3 x 10"3 mg/L
ATSDR (1995); Howard and Meylan (1997)
Log Kow
6.04
Verschueren (2001)
Odor threshold
No data
Molecular weight
252.32
O'Neil et al. (2001)
Conversion factors3
1 ppm = 10.32 mg/m3
Verschueren (2001)
Empirical formula
C2oH12
http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7CHEM
aCalculated based on the ideal gas law, PV= «RT at 25°C: ppm = mg/m3 x 24.45 ^ molecular weight.
Inhalation
ATSDR (1995) reports average indoor concentrations of benzo[a]pyrene as 0.37 to 1.7
3 3
ng/m for smokers and 0.27 to 0.58 ng/m for non-smokers. Naumova et al. (2002) measured
PAHs in 55 nonsmoking residences in three urban areas during June 1999-May 2000. Mean
indoor benzo[a]pyrene levels ranged from 0.02 to 0.078 ng/m . They also reported outdoor
benzo[a]pyrene levels ranging from 0.025 to 0.14 ng/m . They concluded that indoor levels of
the 5-7 ring PAHs (such as benzo[a]pyrene) were dominated by outdoor sources and observed an
average indoor/outdoor ratio of approximately 0.7. Mitra and Wilson (1992) measured
benzo[a]pyrene air levels in Columbus, OH and found elevated indoor levels in homes with
smokers. They measured an average outdoor air concentration of 1.38 ng/m and indoor
3 3
concentrations of 0.07 ng/m for homes with electrical utilities, 0.91 ng/m for homes with gas
3 3
utilities, 0.80 ng/m for homes with gas utilities and a fireplace, 2.75 ng/m for homes with gas
-3
utilities and smokers, and 1.82 ng/m for homes with gas utilities, smokers, and a fireplace.
Mitra and Ray (1995) evaluated data on benzo[a]pyrene air levels in Columbus, OH and reported
3 3
an average of 0.77 ng/m inside homes and 0.23 ng/m outdoors. Park et al. (2001) measured
ambient levels of benzo[a]pyrene in Seabrook, TX during 1995-1996. Based on continuous
"3
measurements over this period, they found an average of 0.05 ng/m (vapor plus particulate).
Parke et al. (2001) also reports average ambient air levels in ng/m from other studies conducted
earlier as 1.0 for Chicago, 0.19 for Lake Michigan, 0.01 for Chesapeake Bay and 0.02 for Corpus
Christie, TX. Petry et al. (1996) conducted personal air sampling during 1992 at 5 workplaces in
4
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Switzerland: carbon anode production, graphite production, silicon carbide production, bitumen
paving work, and metal recycling. These data are summarized in Table 2-2.
Santodonato et al. (1981) estimated adult daily intake from inhalation as ranging from 9
to 43 ng/day. EC (2002) reported the following benzo[a]pyrene air levels in Europe during the
3 3
1990's: rural areas: 0.1-1 ng/m and urban areas: 0.5 - 3 ng/m . They estimated the mean
intake via inhalation for an adult non-smoker as 20 ng/day. The data from exposure studies by
Naumova et al. (2002) suggest typical inhalation intakes may be lower (probably due in part to
the focus on nonsmoker residences). These data suggest that air exposures are typically less than
3 3
0.14 ng/m which would result in 2 ng/day assuming a 13 m /day inhalation rate (adult average
based onUSEPA, 1997).
Table 2-2. Benzo|alpyrene Concentrations in Air
Setting
Year
n
Concentration
(ng/m3)
reference
Outdoor - Urban
Los Angeles, CA
1999-2000
19
0.065
Naumova et al., 2002
Houston, TX
1999-2000
21
0.025
Naumova et al., 2002
Elizabeth, NJ
1999-2000
15
0.14
Naumova et al., 2002
Seabrook, TX
1995-1996
NA
0.05
Park et al. 2001
Columbus, OH
1986-1987
8
0.23
Mitra and Ray, 1995
Indoor Residential
Los Angeles, CA
1999-2000
19
0.078
Naumova et al., 2002
Houston, TX
1999-2000
21
0.020
Naumova et al., 2002
Elizabeth, NJ
1999-2000
15
0.055
Naumova et al., 2002
Columbus, OH
1986-1987
8
0.77
Mitra and Ray, 1995
Columbus, OH
10
0.07-2.75
Mitra and Wilson,
1992
Homes with smokers
0.37-1.7
ATSDR 1995
Homes without smokers
0.27-0.58
ATSDR 1995
Occupational:
Aluminum production
30 -530
ATSDR 1995
Coke production
150-672 0
8000
ATSDR 1995
Petry et al., 1996
Carbon anode production -
Switzerland
1992
30
1100
Petry et al., 1996
Graphite production -
Switzerland
1992
16
83
Petry et al., 1996
SiC production - Switzerland
1992
14
36
Petry et al., 1996
Metal recovery - Switzerland
1992
5
14
Petry et al., 1996
Bitumen paving - Switzerland
1992
9
10
Petry et al., 1996
NA = Not Available
Airborne intake of benzo[a]pyrene in the environment predominantly occurs via
inhalation of insoluble carbonaceous particles (e.g. soot, diesel particles) to which organic
compounds, such as PAHs, are adsorbed. Reliable, quantitative measurements of the percent
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
absorption of benzo[a]pyrene from insoluble particles are not available; however, studies in
experimental animals indicate that benzo[a]pyrene is readily absorbed from carbonaceous
particles following inhalation exposure (Gerde et al., 2001; Hood et al., 2000).
Oral
The processing and cooking of foods is viewed as the dominant pathway of PAH
contamination in foods (as reviewed by Bostrom, 2002). Among the cooking methods that lead
to PAH contamination are the grilling, roasting and frying of meats. Raw meat, milk, poultry
and eggs will normally not contain high levels of PAH due to rapid metabolism of these
compounds in the species of origin. However, some marine organisms, such as mussels and
lobsters are known to adsorb and accumulate PAH from water, which may be contaminated, for
example by oil spills. Vegetables and cereal grains can become contaminated primarily through
aerial deposition of PAHs present in the atmosphere (Li 2009).
Kazerouni et al. (2001) measured benzo[a]pyrene in a variety of commonly consumed
foods collected from grocery stores and restaurants in Maryland (analyzed as a composite from
4-6 samples of each food type). The foods were tested after various kinds of cooking. These
results are reported in Table 2-3. The concentrations were combined with food consumption
data to estimate intake. The intakes of the 228 subjects ranged from approximately 10 to 160
ng/d with about 30% in the 40 to 60 ng/day range. The largest contributions to total intake were
reported as bread cereal and grain (29%) and grilled/barbecued meats (21%).
Kishikawa et al. (2003) measured benzo[a]pyrene levels in cow milk, infant formula and
human milk from Japan. They report the following means: cow milk - 0.03 ng/g (n=14), infant
formula - 0.05 ng/g (n=3) and human milk - 0.002 (n=51).
From the surveys conducted in six EU countries, the mean or national-averaged dietary
intake of benzo[a]pyrene for an adult person was estimated in the range 0.05 to 0.29 (J,g/day
(European Commission[EC], 2002). In the UK, average intakes on a ng kg"1 day"1 basis were
estimated for the following age groups: adults - 1.6, 15 to 18 years - 1.4, 11 to 14 years - 1.8, 7
to 10 years - 2.6, 4 to 6 years - 3.3 and toddlers 3.1-3.8. The major contributors were the oils
and fats group (50%), cereals (30%) and vegetables (8%) (EC, 2002). The contribution from
grilled foods appeared less important in Europe than the U.S. because grilled foods are consumed
less often (EC, 2002).
Table 2-3. Benzo|alpyrene Levels in Food
Concentration (ng/g)
Meat
Fried or broiled beef
0.01-0.02
Grilled beef
0.09-4.9
Fried or broiled chicken
0.08-0.48
Grilled chicken
0.39-4.57
6
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Fish
0.01-0.24
Smoked fish
0.1
Bread
0.1
Breakfast Cereals
0.02-0.3
Vegetable Oil
0.02
Eggs
0.03
Cheese
<0.005
Butter
<0.005
Milk
0.02
Fruit
0.01-0.17
Source: Kazerouni et al., 2001
Estimates of oral bioavailability from animal studies range from about 30-70% (Ramesh
et al, 2001b; Cavret et al., 2003; Hecht et al., 1979). Direct information regarding absorption of
benzo[a]pyrene in humans is limited. One study indicated near 100% absorption of
benzo[a]pyrene in eight subjects exposed to benzo[a]pyrene through the ingestion of charbroiled
meat (Hecht et al., 1979). Other dietary factors likely influence the oral absorption of
benzo[a]pyrene. In experimental animals, a high fat diet appears to increase absorption of
benzo[a]pyrene whereas a high fiber or protein rich diet appears to decrease absorption
(Kawamura et al., 1988; O'Neill et al., 1991; Mirvish et al., 1981).
Dermal
The general population can be exposed dermally to benzo[a]pyrene when contacting soils
or materials which contain benzo[a]pyrene such as soot or tar. Exposure can also occur via the
use of dermally applied pharmaceutical products which contain coal tars, including formulations
used to treat conditions such as eczema and psoriasis (IARC, 2010).
PAHs are commonly found in all types of soils. ATSDR (1995) reported benzo[a]pyrene
levels in soil for a variety of settings: 2-1,300 |ig/kg in rural areas, 4.6 - 900 |ig/kg in
agricultural areas and 165-220 |ig/kg in urban areas and 14,000-159,000 |ig/kg at contaminated
sites (before remediation). The soil levels for all land uses appear highly variable. The levels
are affected by proximity to roads/combustion sources, use of sewage sludge derived
amendments on agricultural lands, particle size and organic carbon content. Wilke (2000)
reports that PAH levels in soils have generally increased during the 1900's and that sediment
studies suggest some declines may have occurred since the 1970's. An illustration of
benzo[a]pyrene levels in soil is presented in Table 2-4.
Table 2-4. Levels of
jenzofalpyrene in Soil
Reference
Location
Land Type
Concentration
Mean (jig/kg)
Butler et al, 1984
UK
Urban
1165
Vogt et al. 1987
Norway
Industrial
321
Norway
Rural
14
Yang et al. 1991
Australia
Residential
363
7
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Maliszewska, 1996
Poland
Agricultural
22
Trapido, 1999
Estonia
Urban
106
Estonia
Urban
398
Estonia
Urban
1113
Estonia
Urban
1224
Estonia
Rural
6.8
Estonia
Rural
15
Estonia
Rural
27
Estonia
Rural
31
Nam et al., 2008
UK
Rural
46
Norway
Rural
5.3
Mielke et al. 2001
New Orleans
Urban
276
Nadal et al, 2004
Spain
Industrial-
chemical
100
Spain
Industrial-
petrochemical
18
Spain
Residential
56
Spain
Rural
22
Maliszewska, 2009
Poland
Agricultural
30
Wilkce, 2000
Various temperate
Arable
18
Various temperate
Grassland
19
Various temperate
Forest
39
Various temperate
Urban
350
Bangkok
Urban-tropical
5.5
Brazil
Forest-tropical
0.3
A number of studies have measured dermal absorption of benzo[a]pyrene from soil (Turkall et
al. 2008; Moody et al, 2007; Roy and Singh, 2001; Roy et al, 1998; Wester et al, 1990; Yang et
al, 1989). These studies utilize in vitro and in vivo testing in a variety of animal species and in
vitro testing with human skin samples. . The absorption percentages of benzo[a]pyrene from
soil, as tested in vitro for human skin, ranged from 0.9 to 15% (Moody et al., 2007; Roy et al.,
1998; Wester et al., 1990). However, major methodological differences between these studies
exist including whether the amount of benzo[a]pyrene left in the skin depot was included as part
of the absorbed fraction or whether only benzo[a]pyrene or its metabolites passing into the
receptor fluid was quantified.
These studies of benzo[a]pyrene absorption from soil suggest that reduced absorption of
benzo[a]pyrene occurs with increasing organic carbon content and clay content of the soils.
They also indicate that dermal absorption increases as soil aging decreases (ie. contact time
between soil and chemical).
8
DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
3. TOXICOKINETICS
Benzo[a]pyrene is one of the most extensively studied PAH compounds. Numerous
primary reports and secondary reviews are present in the scientific literature that describe the
toxicokinetics of benzo[a]pyrene following oral, inhalation, and dermal exposures.
Benzo[a]pyrene is absorbed following exposure by inhalation, oral, and dermal routes. The rate
and extent of absorption are dependent upon the exposure medium. For example, bioavailability
of benzo[a]pyrene is dependent on vehicle characteristics and adsorption to particles. The
presence of benzo[a]pyrene in body fat, blood, liver, and kidney and the presence of
benzo[a]pyrene metabolites in serum and excreta indicate wide tissue distribution.
Benzo[a]pyrene metabolism occurs in essentially all tissues, with high metabolic capacity in the
liver and significant metabolism in tissues at the portal of entry (lung, skin, and gastrointestinal
[GI] tract) and in reproductive tissues. Stable metabolic products identified in body tissues and
excreta are very diverse and include phenols, quinones, and dihydrodiols. These classes of
metabolites are typically isolated as glucuronide or sulfate ester conjugates in the excreta but can
also include glutathione conjugates formed from quinones or intermediary epoxides. The
primary route of metabolite elimination is in the feces, particularly following exposure by the
inhalation route. To a lesser degree benzo[a]pyrene metabolites are eliminated via urine.
Overall, benzo[a]pyrene is eliminated quickly with a biological half-life of several hours.
3.1. ABSORPTION
3.1.1. Inhalation Exposure
3.1.1.1. Inhalation Exposure in Humans
The absorption of benzo[a]pyrene is frequently assessed by identification of
benzo[a]pyrene metabolites in the urine of people exposed to emissions from combustion
processes. Because of the nature of these processes, oral and dermal exposures are likely to
accompany exposure through the inhalation route, rendering estimates of inhalation-only
exposure rather imprecise. For these reasons, quantitative estimates of absorption via the
respiratory tract cannot be derived from these studies. Nevertheless, the observation of
benzo[a]pyrene metabolites (as well as DNA adducts) in tissues and excreta of exposed humans
provides qualitative evidence for benzo[a]pyrene absorption, at least some of which is likely to
be via the respiratory tract.
Becher and Bj0rseth (1983) studied urinary excretion of 11 PAHs in highly exposed
Norwegian aluminum smelter workers (11 exposed: 7 smokers, 4 nonsmokers; 9 controls:
5 smokers, 4 nonsmokers). The authors compared urinary excretion of parent compound to that
of hydroxy-metabolites (determined as parent compound following a chemical reduction
procedure) and found that, in the case of benzo[a]pyrene, 92% of the total excreted were
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
metabolites. These values differed widely for the other 10 PAHs, several of which could be
detected only following the reduction procedure. Exposed nonsmoking workers excreted 0.104
±0.154 jag total benzo[a]pyrene per mmol creatinine (range: 0.002-0.37; parent compound plus
metabolites) in urine (0.0153 ±0.016 without one extreme outlayer), while smoking workers
excreted 0.025 ± 0.016 [j,g/mmol creatinine (range: 0.01-0.063); the difference was not
significant. Total PAH excretion in the urine of aluminum plant workers was also higher in
nonsmokers than in smokers (6.61 ± 3.59 vs. 5.65 ± 2.31 [j,g/mmol creatinine). Air
concentrations of PAHs in the aluminum reduction plant were typically 100 (J,g/m . The authors
concluded that neither high occupational exposure to PAHs nor smoking status provided accurate
determinants for PAH body burdens and that interindividual differences in absorption or
metabolism played a major role.
Grimmer et al. (1994) measured PAH metabolites in the 24-hour urines of four coke oven
workers whose exposure to PAHs had been monitored with personal air samplers on
4 consecutive workdays. They observed a correlation between benzo[a]pyrene amounts
extracted from the sampler filters and benzo[a]pyrene-9,10-dihydrodiol concentrations in urine.
Urinary concentrations following similar levels of exposure, however, varied by a factor of about
5 among the four workers, which the authors attributed to differences in genetically determined
metabolism. One of the central findings in that study was that only a very small fraction of the
inhaled benzo[a]pyrene (0.013%) was recovered from urine, suggesting poor pulmonary
absorption, poor metabolism, or that urine is not a major route for excretion of benzo[a]pyrene.
In the case of phenanthrene and pyrene (Py), percentages recovered from urine were at least
fivefold higher.
Giindel et al. (2000) studied the urinary excretion of metabolites of eight PAHs, among
them benzo[a]pyrene, in 19 workers at a fireproof stone manufacturing plant in Germany, and
provided concentrations in the air to which the workers were exposed. In the case of
"3
benzo[a]pyrene, the median for personal air samplers was 1.07 (J,g/m (range: 0.043-2.96), and
the median for stationary air sampling was 1.31 (J,g/m (range: 0.63-5.41). Other PAH air
3 3
concentrations ranged from 0.11 (J,g/m (dibenz[a,h]anthracene) to 4.85 (J,g/m (chrysene). The
median for urinary excretion of 3-OH-benzo[a]pyrene (the only benzo[a]pyrene metabolite
evaluated) was 1.58 ng/mmol creatinine (range: 0.34-22.6). This was by far the lowest level of
PAH metabolites found in the urine of exposed workers; for comparison, phenanthrene, which
showed almost the same median concentration as benzo[a]pyrene in personal air samplers
"3
(1.08 (J,g/m ), produced a total of 679 ng/mmol creatinine in the form of metabolites (range:
205-4,700). (The author's values were given as (j,g/g creatinine and recalculated using a mol. wt.
of 113.12 for creatinine. Values for phenanthrene metabolites were obtained by addition of
urinary concentrations of four metabolites.) The authors pointed out that they were not able to
detect a correlation between the levels of individual PAH exposures and urinary excretion of
related metabolites.
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
Wu et al. (2002) found a statistically significant correlation between trans-anti-
benzo[a]pyrene-tetrol in the urine of coke oven workers and PAH concentrations in benzene
extracts obtained from personal air monitoring devices. These workers were exposed to a variety
of PAHs, including benzo[a]pyrene. The results were not influenced by smoking or alcohol
consumption habits. However, genetic factors had some influence on urinary trans-anti-
benzo[a]pyrene-tetrol levels (e.g., workers homozygous for the cytochrome (CYP)l Al Mspl
variant displayed 27% higher urinary tetrol levels than did workers heterozygous or wild type
[WT] for this variant). There was also a statistically significant correlation between urinary
levels of ^ram,-a«ft'-benzo[a]pyrene-tetrol and 1-hydroxypyrene (1-OH-Py), a metabolite not
derived from benzo[a]pyrene, but from Py metabolism that is frequently used for assessment of
PAH exposure.
Hecht et al. (2003) attempted to establish a procedure for the assessment of PAH
exposure by measurement of urinary metabolites of phenanthrene. They compared levels of
r-l,t-2,3,c-4-tetrahydroxy-l,2,3,4-tetrahydrophenanthrene to those of 1-OH-Py or
r-7,t-8,9,c-10-tetrahydroxy-7,8,9,10-tetrahydrobenzo[a]pyrene (/ra/7.s-£//7//-benzo[a]pyrene tetrol)
in the urine of psoriasis patients treated with coal tar, of coke oven workers, and of smoking or
nonsmoking unexposed control persons. They demonstrated statistically significant correlations
among all three metabolites in the urine of coke oven workers but not in psoriasis patients,
despite the fact that the latter had 13- to 94-fold higher urinary metabolite concentrations. In
controls, only trans-anti-PheT vs. 1-OH-Py was assessed, and the correlation was statistically
significant. The authors emphasized that urinary concentrations of trans-anti-benzo[a]pyrene
tetrol were 8,000-19,000 times lower than those of r-l,t-2,3,c-4-tetrahydroxy-l,2,3,4-
tetrahydrophenanthrene. In a similar attempt, Ariese et al. (1994) tried to establish a correlation
between urinary 1-OH-Py and 3-OH-benzo[a]pyrene and found a significant correlation in
unexposed controls but not in exposed coke oven workers, who displayed significantly elevated
1-OH-Py levels but no corresponding elevation of 3-OH-benzo[a]pyrene.
The available data from human exposure studies provide qualitative evidence of
benzo[a]pyrene absorption via the respiratory tract and also indicate that, in comparison with
other PAHs, benzo[a]pyrene is absorbed in an unpredictable fashion. Most authors appear to
assume that, in occupational settings, exposure occurs predominantly by the inhalation route.
Occupational inhalation exposure to air contaminated with high levels of benzo[a]pyrene does
not necessarily result in correspondingly elevated excretion of benzo[a]pyrene metabolites.
These qualitative observations in humans are supported by inhalation and instillation
toxicokinetic studies in animals.
3.1.1.2. Inhalation Exposure in Animals
Gerde et al. (1993a) conducted a series of studies on the disposition of benzo[a]pyrene in
the respiratory tract of dogs. Seven-year-old female beagle dogs (n = 3) were exposed to a bolus
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
of 77 mg aerosolized benzo[a]pyrene crystals (particle size <1 (j,m) injected directly into the
trachea of the animals during a single breath. Blood levels of benzo[a]pyrene in the ascending
aorta and right atrium were monitored to evaluate the rate of appearance of benzo[a]pyrene in the
systemic circulation. Benzo[a]pyrene concentrations in the blood built up rapidly and peaked at
1.8 minutes. Over the 15-minute time period after dosing, approximately 92% of the
administered benzo[a]pyrene was cleared from the lungs. The half-life for lung clearance was
2.4 minutes. Comparing lung clearance rates for benzo[a]pyrene and the less lipophilic
phenanthrene, which is cleared more quickly, the authors concluded that the clearance rate was
limited by diffusion through the alveolar septa for highly lipophilic compounds such as
benzo[a]pyrene. A half-life of 2 hours was estimated for clearance of absorbed benzo[a]pyrene
from the blood. These data demonstrate that absorption of benzo[a]pyrene from the pulmonary
portion of the respiratory tract of animals (as opposed to the tracheal, bronchial, and bronchiolar
portions) is nearly complete and very rapid.
In a second study in this series, Gerde et al. (1993b) determined the disposition of
benzo[a]pyrene in the conducting airways by instilling benzo[a]pyrene (dissolved in 20 [xL saline
and administered as a mist) into a main stem bronchus or distal portion of the trachea of female
beagle dogs (n = 4). Benzo[a]pyrene retained in mucus was determined from lavage samples to
assess the degree to which benzo[a]pyrene was transferred from the mucus into the bronchial
epithelium. Roughly 34% of the benzo[a]pyrene was retained in the mucus collected 1 minute
after instillation. Benzo[a]pyrene in the mucus cleared at the same rate as inert particles (90
mm/minute), suggesting transport via the mucociliary escalator. The estimated half-life for
clearance of benzo[a]pyrene from the mucus to the respiratory epithelium was 9.5 minutes.
Therefore, the mucociliary escalator acted like a rather shallow pool for benzo[a]pyrene.
Absorption into the airway walls was evaluated by instilling benzo[a]pyrene solutions into the
upper bronchial tree and measuring benzo[a]pyrene and metabolites in tissues. Approximately
20% of the benzo[a]pyrene dose penetrated into the epithelium of the left and right bronchus,
respectively, within 45 minutes and cleared from the main stem bronchi with a half-time of 1.4
hours. Benzo[a]pyrene metabolites were also measured in main stem airway segments (trachea
and bronchi were dissected into approximately 1-cm-long segments). The percent of
benzo[a]pyrene recovered as parent compound was 62 and 33% after 45 minutes and 1.5 hours,
respectively. The pattern of metabolites varied across these segments. Benzo[a]pyrene tetrols
ranged from 9 to 37% and benzo[a]pyrene-9,10-diol ranged from 3 to 18% of the administered
dose. About 8% of the administered dose was covalently bound to tissues. The authors
concluded that the significantly higher retention time in the bronchi, as compared to pulmonary
epithelium, made the conducting portion of the respiratory tract a possible target of
benzo[a]pyrene toxicity.
In the third paper of the series, Gerde et al. (1993c) used the results from the previous
two studies to evaluate benzo[a]pyrene dosimetry in the respiratory tract. The authors concluded
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
that benzo[a]pyrene uptake occurs via diffusion-limited transport through the lung epithelium.
Uptake by the alveolar epithelium took only a few minutes while uptake by the airway
epithelium took hours, due to the much thicker air/blood barrier in the conducting airways. This
would result in longer residence time in the respiratory tract tissues, where benzo[a]pyrene can
be metabolized to reactive metabolites, making the conducting airways an important target for
benzo[a]pyrene.
This series of studies by Gerde et al. (1993a, b, c) assessed the absorption of
benzo[a]pyrene in the respiratory tract using an aerosolized solution of benzo[a]pyrene in saline.
However, other studies in animals assessed the bioavailability of benzo[a]pyrene in the lung
using benzo[a]pyrene adsorbed onto particle substrates. Gerde et al. (2001) evaluated the
bioavailability of diesel soot-adsorbed benzo[a]pyrene in three 1-year-old beagle dogs (n = 3).
Soot particles were denuded by toluene extraction and benzo[a]pyrene was adsorbed onto the
soot as a surface coating. The dogs were exposed to a single, 220 mL bolus of aerosolized
benzo[a]pyrene-coated diesel soot (mass median aerodynamic diameter [MMAD] 1.3 ± 0.2 (j,m),
followed by 90 mL of clean air to facilitate delivery of benzo[a]pyrene to the alveolar region,
and arterial and venous blood samples were taken at intervals over a 1-hour period. In separate
tests, the amount of benzo[a]pyrene deposited was determined to be 36 ± 20 jag (n = 6). The
concentration of benzo[a]pyrene in the blood peaked at about 2 minutes and the first half-life of
absorption was approximately 4 minutes. Although one dog received approximately seven times
the dose of benzo[a]pyrene than the other two, probably due to variability in the aerosol
generation technique, the fractional retention of benzo[a]pyrene in the lung was similar in all
three dogs, indicating first-order absorption kinetics. The initial absorption was rapid; <10% of
the dose remained in the lung after 30 minutes. However, there was a small fraction of
benzo[a]pyrene that remained adsorbed to the soot even after 5.6 months in the lungs, when the
fraction of material coating the particles had decreased to approximately 16% of what would be a
monolayer of benzo[a]pyrene molecules deposited on the soot particles. The authors suggested
that the small portion of tightly adsorbed benzo[a]pyrene reflected limited high-energy binding
sites that cover only a fraction of the soot particle surface. Benzo[a]pyrene was further released
from particles transported to the lymph nodes to approximately 10% of a monolayer coating,
which may reflect the more reactive chemical environment provided by alveolar macrophages.
Only 30%) of the benzo[a]pyrene that remained bound to particles was present as parent
compound. Based on these results, most of the adsorbed benzo[a]pyrene was readily released
from diesel soot into the systemic circulation, mostly as parent benzo[a]pyrene (with only a
minor portion as metabolites), while a small fraction was released from the soot at a much slower
rate.
Ramesh et al. (2001a) conducted a toxicokinetic study of inhaled benzo[a]pyrene in F344
rats. The rats were exposed for a single 4-hour period via nose-only inhalation to aerosol
"3
concentrations of 0.1, 1.0, or 2.5 mg/m of benzo[a]pyrene adsorbed to carbon black (CB)
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
particles. The particle size distribution was monodisperse and largely in the respirable range (the
reported MMAD was 1.7 [j,m with a geometric standard deviation [SD] of 0.085 suggesting
that the results reflect absorption from the entire lung, since particles of the size distribution used
here are expected to be deposited in all respiratory tract regions. Plasma and lung tissue
concentrations of benzo[a]pyrene and metabolites were evaluated at 30, 60, 120, and 240
minutes postexposure. The plasma benzo[a]pyrene concentration peaked at 1 hour postexposure,
and 65% of the inhaled aerosol was cleared from the lung at 2 hours postexposure, presumably at
all dose levels, though this was not stated explicitly in the study. There was a significant
difference in the time course of plasma levels between male and female rats. Female plasma
benzo[a]pyrene levels were about one-third lower at 30 minutes, about 28% higher at 1 hour, and
marginally lower than the male levels at later time points. This study is limited by the fact that
administered aerosol concentration was reported instead of deposited dose, plasma samples were
not collected during the 4-hour exposure period, and the study could not distinguish between
absorption from the respiratory tract and mucociliary clearance followed by absorption from the
gut.
Rapid absorption through the lungs was also shown following intratracheal
"3
administration of 1 (J,g/kg body weight [ H]-benzo[a]pyrene dissolved in triethylene glycol in
male Sprague-Dawley rats (Weyand and Bevan, 1986). Elimination of radiolabel from the lung
was biphasic with half-lives of 5 and 116 minutes. The highest levels of liver radiolabel,
equivalent to 21% of the administered dose, were found within 10 minutes after exposure,
suggesting rapid absorption from the upper respiratory tract. The authors noted that it was
unlikely that the appearance of radiolabel in organs was due to GI tract absorption after
mucociliary clearance because the tracheal cannula was left in place for the entire experiment
and levels of radioactivity in the stomach increased only slowly. Based on a comparison of
benzo[a]pyrene concentrations in the blood following intratracheal administration versus
intravenous (i.v.) dosing, the authors calculated the pulmonary bioavailability of benzo[a]pyrene
as 57%). A significant degree of metabolism occurred in the lungs (as measured by the
concentration of metabolites in lung), suggesting that benzo[a]pyrene absorption into the
systemic circulation is limited by first-pass metabolism in the lung.
Petridou-Fischer et al. (1988) applied 10 [xL aliquots of [14C]-benzo[a]pyrene in a
gelatin: saline solution over a 2-hour period to the ethmoid and maxillary nasal turbinates of two
female cynomolgus monkeys and four male beagle dogs to assess differences in benzo[a]pyrene
disposition in portions of the nose. The dose of benzo[a]pyrene was not provided, but using the
total radioactivity administered (93 jaCi per animal) and the specific activity of the radiolabeled
benzo[a]pyrene that was used (39 mCi/mmol), a total administered dose of 0.6 mg/animal can be
calculated. (98 jaCi x 1 mmol/39 mCi x 1 mCi/1,000 jaCi x 252 mg benzo[a]pyrene/mmol =
0.6 mg/animal.) No radioactivity was found in blood (collected over 2 hours in dogs and 3 hours
in monkeys), and very little radioactivity was identified in excreta. Urinary excretion reached 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
38
maximum of 0.69% of the administered dose in dogs and 0.07% in monkeys, while in feces a
maximum of 6.42% of the administered dose was recovered in dogs and 1.17% in monkeys over
a period of 48 hours. These results suggest only limited systemic absorption of benzo[a]pyrene
from the nasal turbinates under the test conditions used. The results of this study are in contrast
to more traditional inhalation or intratracheal instillation experiments, which have demonstrated
significant absorption via other portions of the respiratory tract following inhalation or
intratracheal instillation. Little radioactivity was recovered from the mucus, blood, or excreta,
suggesting that benzo[a]pyrene and its metabolites were sequestered in the nasal tissues.
Several studies demonstrated rapid desorption of benzo[a]pyrene bound to particles.
However, the adsorption matrix can impact the bioavailability of inhaled benzo[a]pyrene. Leung
et al. (1988) reported that benzo[a]pyrene adsorbed on diesel soot particles and suspended in
buffer was transferred to microsomes in vitro far less efficiently than free benzo[a]pyrene. The
authors concluded that benzo[a]pyrene transfer to microsomes depends on the lipid content of
the particles rather than on protein in the medium. Microsomes may enhance the slow transfer of
benzo[a]pyrene from particles, which may become an important source of exposure with long
retention times. No metabolism of benzo[a]pyrene adsorbed to particles was detected in this
study, suggesting that particle-bound benzo[a]pyrene serves as a slow release source of
benzo[a]pyrene to the respiratory tract. These findings are consistent with the report by Gerde et
al. (2001) that a slow-release phase follows the initial rapid desorption of benzo[a]pyrene from
diesel soot. Furthermore, Gerde and Scholander (1989) found in an in vitro study that the release
from carrier particles was the rate-limiting step in the absorption of benzo[a]pyrene by the
bronchial epithelium.
The absorption of inhaled benzo[a]pyrene may also be affected by the size of the particle
to which it is adsorbed. Elimination of benzo[a]pyrene from the lungs of mice was investigated
following intratracheal administration of benzo[a]pyrene crystals (0.5-1.0 [j.m in size) or
benzo[a]pyrene-coated carbon particles (0.5-1.0 [j,m or 15-30 (j,m) (Creasia et al., 1976).
Approximately 50% of the benzo[a]pyrene crystals were cleared within 1.5 hours and >95%
were cleared within 24 hours of treatment. In contrast, benzo[a]pyrene clearance was
approximately 50% after 36 hours following exposure to benzo[a]pyrene absorbed onto small
carbon particles. With larger carbon particles, desorption was even slower, requiring 4-5 days to
release 50% of bound benzo[a]pyrene. The difference in absorption rate for small versus large
carbon particles suggests an influence of particle area surface on the rate of desorption.
The deposition, retention, and bioavailability of benzo[a]pyrene as a pure aerosol or
adsorbed onto gallium oxide (67Ga203) particles was investigated by Sun et al. (1982). Male and
"3
female F344 rats were exposed nose-only for 30 minutes to atmospheres containing 0.6 mg/m
[3H]-benzo[a]pyrene absorbed onto 67Ga203 or to 1.0 [j,g/L neat (i.e., the pure chemical) [3H]-
benzo[a]pyrene (MMADs were reported as approximately 0.1 [j,m in both cases). Radiolabel was
detected in the esophagus, stomach, small and large intestines, cecum, liver, kidney and blood;
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
however, the time to reach peak tissue concentrations differed considerably between the two
exposure regimens. Based on the total amount excreted, 22% of the inhaled dose of
benzo[a]pyrene on Ga2C>3 was released over 16 days, but only 10% of the inhaled dose of pure
benzo[a]pyrene was released. Since the amount excreted can reflect differences in absorption,
i.e., uptake via pulmonary epithelium vs. ingestion of cleared particles, and hence alternative
tissue distribution and metabolism, this result cannot be used quantitatively to estimate
bioavailability of inhaled benzo[a]pyrene. The study established that benzo[a]pyrene adsorbed
to particles had a longer respiratory tract retention period. For benzo[a]pyrene coated on Ga2C>3,
"3
1 day was required to clear 90% of the [ H]-benzo[a]pyrene lung and trachea burdens that were
present 30 minutes after exposure. In contrast, only 1.5 and 4 hours were required to clear 90%
of pure benzo[a]pyrene burdens from the lung and trachea, respectively. A different effect was
observed in the nose, where clearance of 90% benzo[a]pyrene coated on Ga2C>3 required 7 hours
as compared to 20 hours for pure benzo[a]pyrene aerosol. Thereafter, clearance curves for
particle-bound and neat benzo[a]pyrene were similar; the authors attributed this to the absence of
a mucociliary escalator in the nose. Inhalation of benzo[a]pyrene on Ga2C>3 also increased the
dose of the compound and its metabolites to the stomach, liver, and kidneys, which the authors
suggest may have resulted from mucociliary clearance with subsequent ingestion.
benzo[a]pyrene absorption from the respiratory tract may also be affected by the
characteristics of the vehicle. Following intratracheal administration in hydrophilic triethylene
glycol, approximately 70% of the benzo[a]pyrene administered was excreted within 6 hours by
male Sprague-Dawley rats (Bevan and Ulman, 1991). In contrast, 58.4% and 56.2% of
administered benzo[a]pyrene were excreted within a 6-hour period when the lipophilic solvents
ethyl laurate and tricaprylin, respectively, were the vehicles.
Pregnant Wistar rats were exposed head-only for 95 minutes on gestational day (GD) 17
3 3
to 200, 350, 500, 650, or 800 mg/m of a [ H]-benzo[a]pyrene microcondensate generated from
heated pure material (Withey et al., 1993). Particle sizes ranged from 0.61 to 0.88 [j,m MMAD.
Immediately following exposure (no time estimate was provided), blood radiolabel
concentrations varied >eightfold over the fourfold dose range (2.66 ± 0.51 vs. 21.96 ± 1.37 j_ig/g
at lowest and highest dose, respectively). Six hours after exposure, blood radiolabel
concentrations had decreased two- to fourfold from the earlier observation but retained a >10-
fold difference over the dose range (0.74 ± 0.12 vs. 9.56 ±2.1 jag/g). Therefore, the difference in
the ratio of blood levels to dose range was not likely due to an initial rapid phase of
benzo[a]pyrene absorption only at high exposures. However, the study authors did not provide
an explanation for this finding. Radiolabel was detected at 0 and 6 hours in all maternal tissues
and in fetuses examined, indicating that systemic absorption had occurred.
Hood et al. (2000) exposed male and timed pregnant female Sprague-Dawley rats to
-3
100 (J,g/m benzo[a]pyrene-CB aerosol (nose only for 4 hours on GD 15), and collected blood
was analyzed at 30, 60, 120, 180, and 240 minutes for concentrations of benzo[a]pyrene. The
16
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
benzo[a]pyrene aerosol particle distribution was trimodal with a significant portion of particles
<1 [j,m in size. The particle size distribution was expected to result in deposition across all
regions of the respiratory tract, including the pulmonary region. Following exposure, blood
benzo[a]pyrene levels peaked at 30 minutes (the first time point reported), with females
exhibiting approximately a 1.7-fold higher peak concentration than males. At later time points,
the female benzo[a]pyrene blood concentrations were similar to those observed in males. By
240 minutes postexposure benzo[a]pyrene in blood had diminished to <5% of the peak level.
The authors did not report a mass balance to allow for the determination of the percentage of
dose that was absorbed. Furthermore, benzo[a]pyrene metabolites were not measured.
Nevertheless, the appearance of benzo[a]pyrene in the blood at the earliest time point measured
is consistent with the conclusion that benzo[a]pyrene is rapidly absorbed from the respiratory
tract. Although the peak came earlier than in the Ramesh et al. (2001a) study, the same trend of
gender differences was observed, with females displaying higher peak blood levels than males.
In summary, although quantitative estimates of human lung absorption are not available,
existing toxicity and biological monitoring studies suggest that benzo[a]pyrene is absorbed in the
respiratory tract, albeit rather poorly, following inhalation exposure of humans. The evidence
suggests, however, that in humans it is difficult to establish a relationship between
benzo[a]pyrene exposure and urinary excretion of its metabolites due to large interindividual
variation, most likely the result of different genetic makeups and varying background exposures.
Numerous controlled studies indicate that benzo[a]pyrene is well absorbed in animals following
inhalation or intratracheal instillation. In general, the animal studies show that benzo[a]pyrene is
absorbed rapidly (within minutes) and extensively. The rate of absorption varies across regions
of the respiratory tract, with more rapid absorption in the pulmonary regions and slower
absorption in the conducting airways and nose. In some studies in rats, blood benzo[a]pyrene
peak levels at early exposure time points differed between females and males, but leveled out at
later time points. Quantitative estimates of benzo[a]pyrene absorption from the respiratory tract
are difficult to derive because the contribution of absorption from the GI tract following
mucociliary clearance and metabolism of benzo[a]pyrene in the respiratory tract itself are often
difficult to determine. Another complication in interpreting these studies is that benzo[a]pyrene
absorption from the lung depends on the characteristics of the exposure vehicle or the nature of
the particle to which benzo[a]pyrene is adsorbed. In general the data indicate that
benzo[a]pyrene is released to a greater extent from hydrophilic vehicles than from lipophilic
solvents; particle-bound benzo[a]pyrene is released more slowly than the neat compound; and
desorption from large particles is slower than from smaller particles. Because much of the
environmental benzo[a]pyrene is adsorbed onto particles of other materials, the effect of the
carrier particle is highly relevant to environmental exposures.
3.1.2. Oral Exposure
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
3.1.2.1. Oral Exposure in Humans
In a study with eight volunteers who ingested broiled meat containing approximately
8.6 jag of benzo[a]pyrene, the concentration of benzo[a]pyrene in feces was below detection
limits (<0.1 (j.g/person) (Hecht et al., 1979). Although the analytical method used in this study
assessed only parent compound, not fecal metabolites, the result can be interpreted as indicating
that the ingested benzo[a]pyrene was absorbed completely from the GI tract. In addition, studies
were conducted to assess DNA adduct levels or benzo[a]pyrene metabolites in humans exposed
to PAHs by the oral route. In general, these human dietary studies are not adequate to develop
quantitative estimates of oral bioavailability; in one case no measurable relationship between
benzo[a]pyrene intake and internal dose measure was found (Scherer et al., 2000).
3.1.2.2. Oral Exposure in Animals
The bioavailability of benzo[a]pyrene was evaluated in F344 rats dosed by gavage with
100 mg/kg benzo[a]pyrene dissolved in peanut oil and sacrificed 0, 0.5, 1.0, 2.0, 4.0, 8.0, 24, 48,
or 72 hours after dosing (Ramesh et al., 2001b). Blood, liver, reproductive tissues, urine and
feces were analyzed for benzo[a]pyrene and metabolites. Plasma benzo[a]pyrene levels peaked
at 8 hours and fecal levels at 2.4 hours postexposure. Lipophilic metabolites of benzo[a]pyrene
peaked at 2 hours in liver, 8 hours in feces, 24 hours in blood and lung, and 48 hours
postexposure in urine. Water-soluble metabolites reached their maxima at 4 hours in liver and
lung, 8 hours in blood and feces, and 48 hours in urine. Some of the disposition patterns
displayed minor peaks at earlier time points. Based on comparison with plasma levels following
i.v. injection, the oral bioavailability was estimated by the authors as approximately 40%;
however, the details of this determination were not presented.
Foth et al. (1988) conducted several experiments to determine the oral bioavailability of
benzo[a]pyrene. In male Sprague Dawley rats, the areas under the blood concentration-time
curve (AUCs) following oral versus i.v. bolus doses of benzo[a]pyrene (dissolved in Krebs
Ringer buffer with 4% bovine serum albumin) were compared. The oral bioavailability was
"3
estimated as 7.8 and 11.5% for doses of 3.2 and 4.0 nmol [ H]-benzo[a]pyrene per rat
(approximately 1.8-2.7 and 2.2-3.4 (J,g/kg, respectively, calculated as 3.2-4 nmol/rat x
0.252 [j,g/nmol (0.30-0.46) kg body weight as stated by the authors). This result may reflect
limited systemic absorption of benzo[a]pyrene at low doses. In contrast to this result, analysis of
benzo[a]pyrene concentrations in arterial blood and bile after continuous intraduodenal infusion
of radiolabeled benzo[a]pyrene (in the same vehicle as above) showed that approximately 40%
of the administered dose was absorbed by the duodenum over a 240-minute period (Foth et al.,
1988). In a similar experimental design, bile- and duodenum-cannulated male Sprague-Dawley
rats were given [ H]-benzo[a]pyrene in corn oil with and without exogenous bile (Rahman et al.,
1986). The absorption of benzo[a]pyrene was estimated from the cumulative recovery of
radioactivity in the bile and urine over 24 hours. This study showed that absorption of
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
benzo[a]pyrene was enhanced by bile as absorption in the presence of endogenous bile only was
22.9% of that when exogenous bile was administered.
Cavret et al. (2003) reported that absorption of benzo[a]pyrene (as measured by the
percentage of orally administered radioactivity appearing in the portal blood in 24 hours) was
30.5% in pigs fed 1 L of 4% fat milk containing 235 [j,g/L [14C]-benzo[a]pyrene (this
corresponds to an oral dose of approximately 6 (J,g/kg based on the reported body weight of 40
kg). The level of radioactivity increased rapidly between 1 and 6 hours, with maximum uptake
between 3 and 6 hours after dosing.
In male F344 rats administered [14C]-benzo[a]pyrene in peanut oil via gavage at doses
from 0.04 to 4.0 [j,mol/rat (approximately 0.03-0.04 mg/kg and 3.4-4 mg/kg, respectively,
calculated as 0.04-4 [j,mol/rat x 0.252 mg/[j,mol 0.25-0.30 kg body weight as stated by the
authors) approximately 85% of the radiolabel was recovered in the feces and 1—3% in the urine
after 168 hours (Hecht et al., 1979). Because radiolabel in feces may represent unabsorbed as
well as absorbed parent compound that is subsequently eliminated via biliary excretion,
bioavailability cannot be estimated from this study. However, the percent of radioactivity
recovered as parent benzo[a]pyrene was small (ranging from 6 to 13% of the administered
radioactivity), suggesting that a minimum of 73% of the administered dose was absorbed (i.e.,
total in urine + total in feces - feces as benzo[a]pyrene).
Dietary matrices may have an impact on the absorption of benzo[a]pyrene from the GI
tract. For example, intestinal absorption of benzo[a]pyrene was enhanced in rats when the
compound was solubilized in lipophilic compounds such as triolein, soybean oil, and high-fat
diets, as compared with fiber- or protein-rich diets (O'Neill et al., 1991; Kawamura et al., 1988).
This may be relevant for the absorption of benzo[a]pyrene from charbroiled meats and other fatty
foods.
O'Neill et al. (1990) assessed the intestinal absorption of [14C]-benzo[a]pyrene given to
rats by gavage in olive oil with regular rat chow or with low-fat diets high or low in either fiber
or beef protein (used to represent human diets). Benzo[a]pyrene and its metabolites were
recovered from feces, where they had been trapped by microcapsules given by gavage 2 hours
prior to benzo[a]pyrene. Benzo[a]pyrene was absorbed from rat chow differently than from
representative human diets. Dietary fiber decreased the availability of benzo[a]pyrene, as
evidenced by the appearance of lower metabolite amounts in the GI tract, while the beef-
enriched diet affected absorption to result in increased formation of 1,6- and 3,6-benzo[a]pyrene
diones. Urinary excretion of benzo[a]pyrene was decreased in rats given the high fiber diet but
not the beef-enriched diet. The total amount of benzo[a]pyrene excreted in feces and the
feces/urine ratio were increased by the high-fiber diet but not by the beef-enriched diet. These
results indicated that bioavailability of benzo[a]pyrene from the GI tract is affected by the type
of diet and that bioavailability studies in animals, using typical laboratory animal chow, may not
appropriately model the situation with varied human diets.
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
The potential influence of diet on PAH bioavailability was investigated also by Wu et al.
(1994). Female mice were fed either gel or powder diets containing coal tar with detectable
levels of benzo[a]pyrene, Py, and other PAHs. Urine samples were collected on the first,
seventh, and fourteenth day of treatment. Measurement of the amount of the Py metabolite 1-
OH-Py in the urine showed that the diet matrix did not influence the bioavailability of the PAHs.
Stavric and Klassen (1994) administered radiolabeled benzo[a]pyrene dissolved in various
vehicles (water, corn oil, liquid paraffin, or 50% ethanol) by gavage and monitored intestinal
absorption in bile-cannulated rats. The animals were fed diets with or without added carbon
particles and typical food components, such as quercetin or chlorogenic acid. They observed that
aqueous vehicles, quercetin, chlorogenic acid, or carbon particles reduced biliary excretion of
benzo[a]pyrene, while lipid media such as corn oil increased it strongly. The authors postulated
that absorption of benzo[a]pyrene from food was affected by its solubility in the vehicle, by
physical adsorption, and/or by adduction of benzo[a]pyrene to certain food ingredients. On the
other hand, Mirvish et al. (1981) observed that varying the corn oil content of a synthetic diet
containing 100 j_ig/g benzo[a]pyrene had little influence on fecal excretion of unmetabolized
benzo[a]pyrene. However, addition of 5% wheat bran to the synthetic diet, or using standard lab
chow, increased the fecal excretion of parent compound 13-fold. They suggested that the
insoluble dietary fiber sequestered benzo[a]pyrene in the GI tract.
In summary, absorption of ingested benzo[a]pyrene was demonstrated qualitatively in
exposed humans by the excretion of metabolites or the presence of DNA adducts. However,
these studies are not sufficient to determine the rate and extent of absorption from the GI tract in
humans. Animal studies have produced variable results, in part due to different study designs.
Standard approaches in animal studies suggest that the oral bioavailability ranges from 10 to
40%. However, some studies have indicated that the standard diets of laboratory animals may
not model the human oral exposure to benzo[a]pyrene appropriately. No data on species or
gender-based differences in absorption were identified.
3.1.3. Dermal Exposure
Several studies in humans and experimental animals have investigated the dermal
absorption of benzo[a]pyrene. Benzo[a]pyrene metabolites or DNA adducts were measured in
humans exposed dermally to benzo[a]pyrene-containing mixtures in biological monitoring
studies. These studies provide only qualitative support for assessing the rate and degree of
dermal absorption of benzo[a]pyrene through human skin. However, some studies provide
quantitative information on the degree of benzo[a]pyrene absorption through the skin in
volunteers or in explanted viable skin samples from tissue donors.
3.1.3.1. Dermal Exposure in Humans
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
Van Rooij et al. (1993) demonstrated differences in absorption rates of PAHs in coal tar
ointment at various skin sites in volunteers. A dose of 2.5 mg/cm of coal tar ointment, which
consisted of 10% coal tar in a vehicle of zinc oxide paste, was applied to 24 cm of skin at the
forehead, shoulder, volar forearm, palmar side of the hand, groin, or ankle and allowed to stand
for 45 minutes before removing the residue. Dermal absorption of PAH was determined through
measuring the disappearance of PAH fluorescence from the skin surface. Absorption rate
constants ranged from 0.036 to 0.135/hour, across application sites, suggesting that 20-56% of
the dose would be absorbed within 6 hours. A 69% difference in dermal absorption rates
between anatomical sites was reported, while only a 7% difference between individual
volunteers was observed. The skin on the shoulder absorbed the greatest dose of PAH, followed
by the forehead, forearm, and groin, with the ankles and palms absorbing the least amount.
There was, however, no correlation between anatomical site of exposure and excretion of 1-OH-
Py in urine. The authors estimated that 0.3-1.4%) of the PAH dose (assessed as 1-OH-Py)
became systemically available, although systemic measurement of benzo[a]pyrene or its
metabolites was not reported.
Quantitative comparisons of dermal penetration of human skin and other mammalian
species have been conducted. Kao et al. (1985) compared benzo[a]pyrene permeation through
skin in short-term organ cultures using skin harvested from mice, rats, rabbits, guinea pigs,
marmosets, and human donors. Two test systems were used, a dynamic one where skin samples
were held on top of chambers flushed with fresh organ culture medium at a constant rate and a
static one where skin samples were incubated on a filter disk in a culture dish over culture
medium at 36°C. The culture medium was a modified minimal essential medium with Earle's
salts and 10%> fetal calf serum. In addition, all experiments were conducted with fresh,
metabolically viable skin and with metabolically nonviable skin (previously frozen or poisoned
"3
with cyanide). Using the static system, [ H]-benzo[a]pyrene dissolved in acetone was applied to
full-thickness skin samples (2.5 [j,g/cm ) and medium collected after 24 hours to be assayed for
"3
[ H]-benzo[a]pyrene and metabolites. The overall penetration rate of benzo[a]pyrene was about
2.6% of the dose in 24 hours from viable human skin and only about 0.5% from nonviable skin.
After 24 hours penetration through viable skin, 52% of the radioactivity in medium consisted of
water-soluble benzo[a]pyrene metabolites and 18% was parent compound. By contrast, after
penetrating through nonviable skin, 90% of the recovered radioactivity was parent compound,
and the ratio of water- to lipid-soluble metabolites was much lower. Results in marmoset, rat,
and rabbit were similar. Skin from the mouse allowed more than 10% of the dose to penetrate,
while that of guinea pig let only a negligible percentage of the dose penetrate. In all species,
metabolism was an important determinant of permeation, with very low rates observed in
nonviable skin.
Using the dynamic test system, Kao et al. (1985) studied the influence of dermal
metabolism on the rate of penetration of benzo[a]pyrene in more detail. Comparing rat to mouse
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
skin they again found a 10-fold difference in the amount penetrating in 16 hours through viable
skin, approximately 0.7% of the dose in rat skin vs. approximately 7% in murine skin, but most
of this species difference disappeared when previously frozen skin was used or potassium
cyanide was added to the medium to destroy metabolic capability. An additional factor was
responsiveness to aryl hydrocarbon (Ah)-receptor agonists: dermal penetration in responsive
mice given an inducing dose of 3-methylcholanthrene (3-MC) was more than twice that in
noninduced Ah-responsive mice, and the latter was similar to nonresponsive mice (see Section
3.3 for further discussion of Ah-responsiveness). These results show that functional enzyme
systems facilitate penetration of benzo[a]pyrene or its metabolites.
Wester et al. (1990) compared skin penetration of [14C]-benzo[a]pyrene dissolved in
acetone versus benzo[a]pyrene adsorbed onto soil using skin from human cadavers. Skin was
dermatomed to 500 [j,m and stored in medium at 4°C to preserve viability. The authors used a
dynamic system with undiluted human serum as the receptor fluid and 24-hour exposure.
Radiolabeled benzo[a]pyrene was applied to the skin, and concentrations of radioactivity in the
receptor fluid, skin, and surface skin wash were determined. For benzo[a]pyrene applied in
acetone, 23.7% of the applied dose was recovered in the skin, 0.09% was recovered in the
receptor fluid, and 53.0% was recovered in the surface wash. The amount of radiolabel
recovered from exposure to benzo[a]pyrene in soil was 1.4% of the dose in skin and 0.01% in
receptor fluid, indicating that the exposure matrix greatly impacted dermal absorption. The
considerable difference in penetration through human skin observed in this study (only 0.09% of
the dose per 24 hours) and that of Kao et al. (1985) (2.6% of the dose) is at least in part a result
of differences in experimental design, such as full thickness (approximately 1.5 mm) vs.
dermatomed (0.5 mm) skin and synthetic receptor fluid with 10% serum vs. pure human serum.
Together, these studies suggest that benzo[a]pyrene is absorbed into human skin but does not
penetrate through the skin rapidly.
In a second part of the same study, Wester et al. (1990) also evaluated the dermal
penetration of benzo[a]pyrene in female rhesus monkeys in vivo. Radiolabeled benzo[a]pyrene
was applied to a 12 cm area of the abdominal skin that was protected by a nonocclusive cover.
The material was maintained on the skin for 24 hours, after which time the skin area was
washed. Urine was collected from the animals during the initial 24-hour exposure period and 6
days after. Skin penetration was determined as the ratio of urinary radiolabel for topically
exposed animals compared to monkeys administered the same dose by i.v. injection. When
benzo[a]pyrene was dissolved in acetone, 51% of the applied dose was absorbed as compared to
13.2% when benzo[a]pyrene was applied in soil. These results further stress the fact that the
dermal absorption of benzo[a]pyrene depends on the vehicle of administration (pure substance
vs. contaminated environmental material).
Also using human tissues, van der Bijl and van Eyk (1999) compared the mean flux of
benzo[a]pyrene across vaginal and buccal mucosa samples from human donors. No significant
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
difference between the two tissue types was observed, and steady-state flux was very low,
2 2
approximately 0.01% of the dose per cm /minute (400 cpm/cm /minute in a figure in the paper,
or 4,000 dpm [assuming 10% [ H]-liquid scintillation counting efficiency] out of a 17.33 jaCi
1 2
dose, or 3.8 x 10 dpm, equals about 0.01% of the dose/cm /minute). These results indicate that
benzo[a]pyrene can be absorbed from and metabolized by the female reproductive tract.
However, the results cannot be compared to those obtained with human skin. First, there was the
difference in absorbing surface, namely, the quasi-dry, mainly lipophilic stratum corneum vs.
fresh epithelium. Second, there were also fundamental differences in experimental design:
although van der Bijl and van Eyk (1999) used a dynamic system, the tissues had been frozen
previously, benzo[a]pyrene was administered in aqueous solution in a large compartment on top
of the tissue sample, receptor fluid was a simple buffer, and incubation was performed at 20°C.
Potter et al. (1999) also demonstrated that the exposure matrix can affect the dermal
absorption of benzo[a]pyrene in mice in vivo as well as in human skin in vitro. In particular, the
uptake of benzo[a]pyrene decreased as the viscosity of the oil product used as vehicle increased.
Although mouse skin absorbed more radiolabel than human skin, the trend of decreasing
absorption with increasing viscosity found in the mouse was also present in human skin.
Vehicles tested included mineral oils (32-198 centi-Stoke [cSt], measure of viscosity), residual
aromatic extracts (5,160-5,400 cSt), and bitumens (0.65-69 x 106 cSt). Most likely, differences
in diffusion coefficients for benzo[a]pyrene in the various vehicles would adequately describe
the differences in dermal absorption.
3.1.3.2. Dermal Exposure in Animals
Yang et al. (1989) compared dermal absorption of benzo[a]pyrene in crude oil adsorbed
to soil (particle size of 150 [j,m or less) versus benzo[a]pyrene in the crude oil alone. In the in
"3
vivo portion of the study, test materials spiked with 100 ppm [ H]-benzo[a]pyrene were placed
over a 7-cm portion of shaved dorsal skin of female Sprague-Dawley rats. Urine and feces were
collected daily over 4 days, and at the end of the experiment animals were sacrificed and
radioactivity in tissues was determined. The total percent of the absorbed dose recovered in
excreta was greater for benzo[a]pyrene in crude oil (35.3%) than oil adsorbed to soil
(approximately 9.2%). While the latter number refers to soil applied to the skin as a monolayer,
when six times the amount of soil, and thus a sixfold dose, was applied, the absolute amount of
benzo[a]pyrene absorbed remained virtually the same. The authors also found that using viable
skin samples in vitro gave almost identical results to those obtained in vivo.
Ng et al. (1992) examined the percutaneous absorption of radiolabeled benzo[a]pyrene in
the hairless guinea pig. A single dose of 28 jag benzo[a]pyrene dissolved in 50 [xL acetone was
applied to 4 cm of dorsal skin and covered with a protective pad for 24 hours. Urine and feces
were collected at 6 and 12 hours and then daily for the next 7 days postexposure. Approximately
34% of the radiolabel was absorbed and eliminated in 24 hours. Most excretion had occurred by
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
day 3 and then continued at a lower rate to reach about 73% by day 7. Benzo[a]pyrene in the
skin wash accounted for 11% of the applied dose. Comparison with the amount excreted
following intramuscular injection (which reached about 85% in 7 days) suggested that the dermal
bioavailability of benzo[a]pyrene was high (about 85%). Ng et al. (1992) noted that, in an in
vitro study, only metabolites were present in the receptor solution, indicating that metabolism in
the skin preceded systemic absorption. This is consistent with the findings of Kao et al. (1985)
that dermal absorption may be aided by a high metabolic capacity of the skin but leaves open the
question how much of the reactive metabolites escape from the dermal compartment.
In an earlier study, Kao et al. (1984) used viable mouse skin to study the dose
dependence of dermal absorption of benzo[a]pyrene. They administered 1, 2, 4, and 6 jag
benzo[a]pyrene per cm full-thickness, shaved skin incubated at 36°C for 24 hours. The
percentage of absorbed benzo[a]pyrene dose decreased with increasing dose: 24, 17, 12, and 7%>
with 1, 2, 4, and 6 [j,g/cm , respectively. The authors also found that, with doses of 4 and
6 [j,g/cm , the total amount of penetrated benzo[a]pyrene-derived radioactivity remained similar,
indicating saturation of the skin's absorption and metabolic capacity. When the donor mice were
pretreated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) to induce cytochrome P450
(CYP450) enzymes in the skin, amounts of penetrated benzo[a]pyrene increased to 38, 33, 30,
and 19%) of the dose, respectively. In addition to the increased rate of absorption, a shift in
metabolic pattern was observed among the benzo[a]pyrene metabolites in the medium. CYP450
induction caused a large increase in metabolite conjugates at the expense of polar metabolites
and benzo[a]pyrene diols, and a strong increase in covalent binding of benzo[a]pyrene to skin
DNA.
Kao et al. (1988) also investigated the influence of skin hair on the rate of benzo[a]pyrene
penetration. Using the same dynamic system, six strains of haired and two strains of hairless
mice were administered a dose of 1 [j,g/cm benzo[a]pyrene. After 16 hours of incubation,
between 4.4 and 9.4%> of the dose had penetrated through viable skin of haired mice but only 2-
2.9% of the dose through skin of hairless mice. Since the authors found that permeation of
testosterone, which was used as a comparison compound with similar physiochemical properties
as benzo[a]pyrene, was the same in haired and hairless mice, 70%> of the dose, they concluded
that skin appendages have considerable metabolic capacity for benzo[a]pyrene. Extending these
findings to humans implies that in the assessment of dermal absorption of benzo[a]pyrene,
presence of dense hair may play an additional role.
Morse and Carlson (1985) compared in vivo dermal absorption of benzo[a]pyrene in
BALB/c and SENCAR mice to find an explanation for the higher sensitivity of the latter strain
toward benzo[a]pyrene-induced skin tumors. Animals were treated with 50 mg/kg
benzo[a]pyrene dissolved in acetone. Levels of benzo[a]pyrene in skin of BALB/c mice did not
reach higher peak levels, but stayed up to three times higher than in SENCAR mice between 6
and 24 hours postexposure. The levels of benzo[a]pyrene-derived radioactivity in liver, lung,
24
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
and stomach also peaked at comparable levels, but stayed at up to twice the levels in SENCAR
mice between 6 and 48 hours posttreatment, compared with BALB/c mice. There were no strain
differences in these tissues when benzo[a]pyrene was administered orally. The metabolic
capacity of the skin was assessed by measuring DNA binding of benzo[a]pyrene in skin.
Six hours after exposure DNA-binding of benzo[a]pyrene was higher in SENCAR than in
BALB/c mice, but by 24 hours after topical administrations, DNA adducts were higher in
BALB/c than in SENCAR mice, possibly indicating differences in metabolism. Kao et al.
(1988) had included the same two strains of mice in one of their studies and found comparable
results, with 9.2% of the dose penetrating over 16 hours in BALB/c and 4.4% in SENCAR mice.
However, the results of Morse and Carlson (1985) did not provide the expected toxicokinetic
explanation for the difference in skin tumor sensitivity between the two mouse strains.
In summary, due to the use of coal tar products for medicinal purposes, quantitative data
on dermal absorption of benzo[a]pyrene in humans are more abundant than for other routes of
exposure. The animal-to-human differences are significant. Generally, mice have greater
absorption than humans, followed by rats and rabbits. The 24-hour penetration values range
from 1 to 3% in viable human skin; much greater amounts of material are absorbed into the skin
but do not readily permeate through it. Dermal absorption of PAH is strongly dependent on
anatomical site (69% difference across six sites), while inter-individual variation is much smaller
(7% difference across nine volunteers). Furthermore, dermal absorption is highly dependent on
underlying metabolic activity of the skin. The vehicle of exposure also impacts dermal
absorption, with vehicles that absorb benzo[a]pyrene, such as soil, or vehicles with low diffusion
coefficients (high viscosity) decreasing the rate and degree of absorption.
3.1.4. Other Types of Exposure
Ewing et al. (2006) used isolated, perfused rat lungs and delivered benzo[a]pyrene coated
onto silica carrier particles (average size: 3.5 (j,m) at mean doses of 2.2, 36, and 8,400 ng to each
lung within <1 minute. Perfusate was collected for 77 minutes thereafter. Lungs and perfusates
were analyzed for benzo[a]pyrene and metabolites. Absorption was strongly dose-dependent: at
the low and mid exposure levels benzo[a]pyrene concentration increased rapidly in the perfusate
to reach a maximum within <5 minutes, then decreased over the remaining observation period.
At the high dose, benzo[a]pyrene in the perfusate reached the maximum at about 30 minutes
after exposure and stayed at a constant level from there on, i.e., the absorption of benzo[a]pyrene
proceeded at zero order until all deposited solid benzo[a]pyrene was dissolved. The mass
balances for benzo[a]pyrene equivalents in lung vs. perfusate were lung ca. one-third perfusate at
the low dose, lung = perfusate at the mid dose, and lung about twice that of perfusate at the high
dose. At the low exposure level metabolism was apparently able to convert most of the parent
compound, while at the highest exposure level most of the absorbed benzo[a]pyrene remained
unmetabolized even at the end of the experiment. The results suggest that, at low doses (2.2
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
ng/lung) benzo[a]pyrene is absorbed very efficiently in rat lung (two-thirds absorbed in
77 minutes), while at higher doses the rate of absorption decreases markedly, either because of
diffusion limitation, or by saturation of local metabolism.
3.2. DISTRIBUTION
No adequate quantitative studies of benzo[a]pyrene tissue distribution in exposed humans
were identified. Obana et al. (1981) observed low levels of benzo[a]pyrene in liver and fat
tissues from autopsy samples. However, prior exposure histories were not available for the
donors. Nevertheless, the identification of benzo[a]pyrene metabolites or DNA adducts in
tissues and excreta of PAH-exposed populations suggest that benzo[a]pyrene is widely
distributed.
3.2.1. Inhalation Exposure
Numerous studies evaluated the disposition of benzo[a]pyrene or its metabolites
following inhalation or instillation in the respiratory tract in animals. The distribution of
benzo[a]pyrene following inhalation exposure was shown to be similar in various species. Male
and female F344 rats were exposed nose-only for 30 minutes to atmospheres containing
3 3 67 3 3
0.6 mg/m [ H]-benzo[a]pyrene adsorbed onto Ga2C>3 or to 1.0 mg/m neat [ H]-
benzo[a]pyrene (Sun et al., 1982). There were qualitative differences in time and amount of
absorption between the exposure regimens. With either exposure regimen, high tissue levels of
radiolabel were found in the small and large intestines and cecum, followed by liver, kidney, and
blood. Levels in the upper GI tract (esophagus and stomach) differed, with significant levels of
radioactivity in the stomach following exposure to [3H]-benzo[a]pyrene adsorbed onto 67Ga203
but only minimal levels in the stomach following exposure to neat benzo[a]pyrene aerosol.
Twelve hours after exposure, the highest tissue concentrations of radiolabel were found in the
lower GI tract for both regimes. The differences in distribution of radiolabel likely reflect the
relative contribution of two mechanisms for the delivery of benzo[a]pyrene-related radioactivity:
for particle-adsorbed benzo[a]pyrene, mucociliary clearance followed by ingestion may be a
significant factor, while aerosolized benzo[a]pyrene is likely to be more readily absorbed in the
respiratory tract only.
Other inhalation studies measured levels of administered radioactivity as benzo[a]pyrene
parent compound or metabolites in whole blood, plasma or lung, but not in any other tissues
(Ramesh et al., 2001a; Gerde et al., 1993a, b). The rapid appearance of benzo[a]pyrene and
metabolites in the blood is consistent with the conclusion that benzo[a]pyrene is readily
bioavailable following exposure by the inhalation route. The degree to which absorbed
benzo[a]pyrene or metabolites was delivered to target tissues was not determined in these
studies.
26
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Weyand and Bevan (1986) examined benzo[a]pyrene disposition and metabolism in male
Sprague-Dawley rats following intratracheal instillation of 1 (J,g/kg [ H]-benzo[a]pyrene
dissolved in triethylene glycol. The amount of radioactivity in various organs was determined at
timed intervals between 5 and 360 minutes after dosing. Peak levels of radioactivity as the
percentage of the administered dose per organ/tissue as well as time profiles were as follows:
• Early peak levels: lungs (59.5% at 5 minutes, declining 10-fold over 360 minutes); blood
(3-4% between 5 and 15 minutes, declining only twofold by 360 minutes); spleen (0.5%
at 5 minutes, declining very slowly); liver (20.8% at 10 minutes, gradually declining);
heart/thymus (1.6% at 10 minutes, barely declining from 0.5% after 30 minutes).
• Medium term peak levels: carcass (27.1% at 60 minutes, mostly stable at around
approximately 25% between 10 and 90 minutes and at approximately around 22%
thereafter); kidney (2.4% at 90 minutes but rather stable at approximately around 2%
between 10 and 360 minutes); testes (1.3% at 90 minutes but rather stable at
approximately around 1% between 15 and 360 minutes).
• Late peak levels: stomach (6.9% at 120 minutes, decreasing slowly); intestinal contents
(44.7%) at 360 minutes, increasing over time period); intestines (14.9% at 360 minutes,
continually increasing).
These profiles are consistent with rapid uptake and delivery of benzo[a]pyrene to well-
perfused tissues, followed by clearance from the tissues via metabolism and excretion
(particularly in the feces). The profile of radioactivity appearing in the GI tract suggests partial
removal of benzo[a]pyrene from the lungs via mucociliary escalator. Total recovery of
radiolabel at 5 minutes was 96% of the dose and 104% at 360 minutes, indicating complete
recovery from the tissues.
Weyand and Bevan (1986) also investigated benzo[a]pyrene metabolites in tissues.
Quinones were at highest concentrations in both lung and liver 5 minutes after instillation,
accounting for 12 and 7% of radiolabeled extractable material, respectively. Benzo[a]pyrene
disposition was also investigated in male rats with and without biliary cannulas. Distribution
patterns among organs were similar, though the amount excreted in bile and intestinal contents
was 77%) of the administered dose in cannulated rats and 53% in animals that were not
cannulated. The intestinal contents carried lower fractions of the administered dose as thioether
and glucuronic acid conjugates than the bile, indicating enterohepatic recirculation of
benzo[a]pyrene metabolites.
A comparative intratracheal instillation study with Sprague-Dawley rats, Gunn rats,
guinea pigs, and hamsters gave results (Weyand and Bevan, 1987) qualitatively similar to those
reported by Weyand and Bevan (1986). Doses of 0.16 or 350 jag [ H]-benzo[a]pyrene per animal
were administered intratracheally, and the distribution of benzo[a]pyrene-derived radioactivity
was determined in various tissues. Relative amounts of radiolabel recovered per gram of tissue
27
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
at 6 hours were: lung (1.7%) > kidney (0.76%) ~ liver (0.67%) > testes (0.21%) = spleen
(0.2P/o) = heart (0.21%) ~ GI tract (0.19%) > stomach (0.13%) > carcass (0.071%). This pattern
of distribution was qualitatively similar among all species tested at both doses, but the relative
values differed between species. For example, in Sprague-Dawley rats the liver burden
represented 0.61% of the recovered radiolabel/gram of tissue, while this value was 2.16% for
hamsters, 0.35% for guinea pigs, and 1.02% for Gunn rats after the 0.16 [j.g/animal dose. The
study did not include the intestine or its contents, where likely the majority of radioactivity was.
When benzo[a]pyrene was instilled intratracheally into mice, Schnizlein et al. (1987) found that
radioactivity in the lung declined steadily throughout the 144-hour investigation period, while
levels in liver, spleen, and intestine peaked at 8 hours after dosing, decreased slowly until
24 hours, and declined rapidly thereafter. It was also noted that as the lung burden of the
radiolabel decreased, radioactivity increased in lung-associated lymph nodes (LALN) over the 6-
day study, suggesting distribution of benzo[a]pyrene or its metabolites via the lymph.
"3
Bevan and Ulman (1991) administered 1 (J,g/kg [ H]-benzo[a]pyrene intratracheally to
male Sprague-Dawley rats in three different liquid vehicles. At 6 hours after dosing, 56.2, 58.4,
and 70.5%) of the dose delivered from tricaprylin, ethyl laurate, and triethylene glycol,
respectively, were recovered from bile. Recovery from whole lung after 6 hours was 13.0% of
the administered dose for tricaprylin and 15.6% for ethyl laurate but only 2.6% for triethylene
glycol, indicating that pulmonary absorption of benzo[a]pyrene was more efficient from a less
hydrophobic vehicle than from a highly hydrophobic one. Among the other organs, kidney and
liver retained rather high levels of radioactivity (around 2 and 5%, respectively, for the whole
organs). Two percent or less of the administered dose was recovered from intestine and its
contents in this study.
Pregnant Wistar rats were exposed for 95 minutes on GDs 17 to 200, 350, 500, 650, or
3 3
800 mg/m [ H]-benzo[a]pyrene generated as a microcondensate from heated pure material
(Withey et al., 1993). Immediately (time not specified) following exposure, the ranking of
benzo[a]pyrene concentrations was maternal lung > blood > liver > kidney > fat > fetus. When
total metabolites (as measured by detection of radiolabel) were measured immediately following
dosing, the ranking was maternal lung > blood > liver > kidney > fetus > fat. Six hours after
exposure, benzo[a]pyrene concentrations were fat > lung > kidney > liver > blood > fetus, while
total metabolite concentrations were lung = fat > kidney > liver = blood > fetus. Concentrations
of benzo[a]pyrene and metabolites in the GI tract were not reported. This study is consistent
with other studies in showing wide tissue distribution of benzo[a]pyrene. In addition, the results
also demonstrated placental transfer of benzo[a]pyrene and its metabolites.
3.2.2. Oral Exposure
Saunders et al. (2002) evaluated the neurotoxicity of benzo[a]pyrene in male F344 rats
following a single gavage dose in peanut oil at doses of 0, 25, 50, 100, or 200 mg/kg body
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
weight (10/sex/dose). Benzo[a]pyrene and metabolite concentrations were monitored for up to
96 hours after administration in plasma, cerebellum, and cerebral cortex. Unmetabolized
benzo[a]pyrene was observed in brain tissue only at the two highest doses, peaking at 2 to
4 hours and then gradually decreasing. Total metabolite concentrations in plasma as well as
brain tissue peaked at 2 hours, remained at similar levels until 6 hours and then gradually
decreased. By 96 hours after dosing benzo[a]pyrene or its metabolites had dropped to trace
levels. The distribution of metabolites shifted over the observation period, with diol metabolites
(4,5-, 7,8-, and 9,10-benzo[a]pyrene diols) predominating for the first 12 hours and hydroxy
metabolites (3-OH, 9-OH-benzo[a]pyrene) predominating at later time points. The distribution
of metabolites was similar in plasma and brain.
Neubert and Tapken (1988) administered a single 12 mg/kg oral dose of [14C]-
benzo[a]pyrene to groups of five pregnant NMRI:Han-mice on GDs 11, 12, and 13 to determine
whether placental transfer occurred. Six, 24, and 48 hours after treatment radiolabel was found
in the maternal lung, liver, and kidney (between 5 and 17% of the dose were recovered per gram
tissue at 6 hours post dosing, decreasing to 0.5-1.3% by 48 hours). Radiolabel was also found in
the placenta and embryonic liver at one to two orders of magnitude less than that found in
maternal tissues. Similar results were found with five pregnant albino rats that received a single
oral dose of 200 mg/kg benzo[a]pyrene in sunflower oil (Shendrikova and Aleksandrov, 1974).
Three hours after treatment, fetal levels of radiolabel were approximately 10% of the maternal
concentration.
Taken together it is apparent that, in rats, benzo[a]pyrene can penetrate the blood barrier,
and in pregnant mice or rats it crosses the placental barrier and reaches the fetus. Otherwise,
distribution data on benzo[a]pyrene following oral administration are insufficient to establish a
cogent picture of organ or tissue distribution.
3.2.3. Dermal and Other Exposures
Some studies have evaluated the distribution of benzo[a]pyrene and its metabolites
following dermal or other dose routes. Morse and Carlson (1985) sought to determine whether
differences in toxicokinetic parameters could explain the difference in tumor response between
SENCAR (high susceptibility) and BALB/c mice (low susceptibility). The mice were
administered radiolabeled benzo[a]pyrene via dermal application or orally, and the time course
of radioactivity levels was assessed in several organs. Following topical application of
benzo[a]pyrene, organ levels of radioactivity were generally 1.5-4 times higher in SENCAR
than in BALB/c mice. Radioactivity levels were liver > lung ~ stomach, with an approximately
twofold difference. DNA adduct levels were liver > lung > stomach, with a threefold difference.
Following oral dosing, tissue radioactivity levels differed little between the two strains and
essentially followed the same pattern of distribution. However, DNA adduct levels following
oral dosing were much higher than after topical administration, approximately 6 times higher in
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
stomach and liver and 10 times higher in lung. RNA and protein binding showed patterns
similar to DNA binding. This study seems to indicate that the route of administration exerts little
influence on the tissue distribution of benzo[a]pyrene.
Moir et al. (1998) also measured the toxicokinetics of benzo[a]pyrene in male Wistar rats
dosed i.v. with 2, 6, or 15 mg/kg of [14C]-benzo[a]pyrene dissolved in an emulphor/saline
emulsion. The concentrations of both benzo[a]pyrene-derived radioactivity and parent
compound were determined in blood, adipose tissue, lung, liver, kidney, heart, spleen, brain, and
testes as well as in urine and feces. The concentrations in all tissues examined except lung
appeared to follow a similar pattern of smoothly increasing and decreasing curves, while the lung
data were rather erratic (this pattern may reflect temporary trapping of lipid vesicles from the
benzo[a]pyrene emulsion in the fine lung vessels). The authors also extracted selected tissues
and determined parent compound levels by high-performance liquid chromatography (HPLC).
Peak concentrations of [14C]-benzo[a]pyrene equivalents and of parent compound and time to
peak for the 2 mg/kg dose group are given in Table 3-1. Similar patterns were observed at
higher doses, with notable exceptions noted below.
Table 3-1. Distribution of benzo[a]pyrene in selected tissues of male rats
following i.v. dosing with 2 mg/kg
Tissue/organ"
Blood
Adiposeb
Kidney0
Liver
Lungd
Total tissue radioactivity
Peak level (ng/g tissue)
4.27 ±0.25
2.31 ±0.87
8.94 ± 1.21
20.55 ±2.20
40.54 ±4.85
Peak time (min)
5
120
5 and 20
5
5 and 120
Parent compound
Peak level ((ig/g tissue)
2.64 ±0.84
3.96 ± 1.92
7.68 ± 1.23
11.33 ± 4.66
5.17 ± 1.15
Peak time (min)
5
120
5 to 20
5
5
"Mean ± SD, n = 3-4.
bAdipose tissue levels showed broad maxima between 20 and 480 min; the highest values are shown.
°Total radioactivity levels in kidney changed little between 30 and 480 min.
dParent compound lung had another maximum of 14.65 ± 1.67 ng/g at 120 min; a second maximum at this time
point was observed at the higher doses, too, but those did not exceed the value of the first maximum at 5 min.
Source: Moir et al. (1998).
At all exposure levels blood maintained the lowest initial concentrations (with some
exceptions in white adipose tissue discussed below). Half-lives for benzo[a]pyrene parent
compound in blood after the 2 and 6 (J,g/kg doses were 36 and 25 minutes, respectively, while no
parent compound could be detected at time points >480 minutes. Following the 15 (J,g/kg dose,
blood showed an initial benzo[a]pyrene decline similar to the lower doses, but benzo[a]pyrene
could be recovered up to the end of the study at 32 hours post dosing, and with these additional
data a second elimination phase half-life for benzo[a]pyrene was estimated at 408 minutes. The
30
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
final half-lives for elimination of benzo[a]pyrene "metabolites" (total benzo[a]pyrene-derived
radioactivity minus parent compound) from blood were 666, 848, and 179 minutes for the 2, 6,
and 15 mg/kg doses, respectively. It is fair to speculate that the second phase of elimination is
the result of redistribution of parent compound (and subsequent metabolism) from a large, deep
compartment such as adipose tissue and/or enterohepatic circulation.
Similar to the results in the lung, erratic patterns of radioactivity occurred in adipose
tissue. Peaks and dips in radioactivity levels in lung occurred roughly with a pattern opposite to
that in adipose tissue. This might suggest occasional redistribution of benzo[a]pyrene parent
compound between a pool of vesicles trapped in the lung and adipose tissue. Data derived from
these tissues should therefore be viewed with caution. Moir et al. (1998) noted that, as a general
rule, increasing tissue radioactivity levels following administration of [14C]-benzo[a]pyrene
reflected metabolite accumulation. Kinetics models were fit to the data to derive rate constants
for clearance of benzo[a]pyrene. Many organs displayed rapid uptake phases for
benzo[a]pyrene: liver (uptake complete at the first time point, 5 minutes) kidney, brain, testis,
and adipose tissue, which was followed by a bi-exponential decline (except in adipose tissue).
Half-lives for benzo[a]pyrene parent compound elimination changed with the dose administered
(half-life in minutes at doses of 2, 6, and 15 (J,g/kg, respectively): liver (28, 163, and 281),
kidney (498, 456, and 389), and adipose tissue (239, 945, and 781). The following organs were
evaluated for the 2 (J,g/kg dose only: brain (2,326 minutes), heart (25 minutes), and spleen (38
minutes); the latter, most likely due to nondetectable radioactivity levels at later time points,
reflect the initial elimination phase only. Elimination half-lives of benzo[a]pyrene metabolites in
liver and several other organs were in the range of 10-15 hours and independent of dose,
suggesting first-order elimination. Overall, the results of Moir et al. (1998) suggest a pattern
consistent with initial wide distribution determined by blood flow, with lipophilicity and rates of
metabolism contributing to temporal patterns of organ levels after the initial distribution period.
Some studies showed that reactive metabolites of benzo[a]pyrene are transported in the
blood and may be distributed to tissues incapable of benzo[a]pyrene metabolism. Ginsberg and
Atherholt (1989) evaluated DNA adduct formation after intraperitoneal (i.p.) administration of
benzo[a]pyrene in mice. Serum of benzo[a]pyrene-treated mice incubated with splenocytes or
salmon sperm DNA resulted in adduct formation suggesting that reactive benzo[a]pyrene
metabolites were systemically distributed and available for interaction with target tissues. DNA
adduct levels formed in vivo were highest in liver, lung, and spleen, with levels in kidney and
stomach significantly lower.
Taken together, the limited human data and few toxicokinetic studies in animals
demonstrate that benzo[a]pyrene and its metabolites are widely distributed throughout the body.
Inhalation of particle-bound or pure benzo[a]pyrene results in significant levels of
benzo[a]pyrene and metabolites in numerous tissues. Distribution of inhaled benzo[a]pyrene and
its metabolites to the GI tract is a result of both mucociliary clearance of particulates from the
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
lung and of biliary excretion following metabolism. Following absorption, benzo[a]pyrene and
metabolites are initially found predominantly in the highly perfused tissues, such as the lung,
liver, and kidneys. Lower amounts are distributed to other tissues, including the male
reproductive organs, central nervous system, and adipose tissue. Despite its high lipophilicity
benzo[a]pyrene has no specific affinity for lipid-rich tissues, most likely because of its rapid
metabolism to more hydrophilic compounds. There is some indication that the distribution of
benzo[a]pyrene is not much influenced by the route of administration. Reactive benzo[a]pyrene
metabolites are also distributed via blood where they may form protein adducts or reach tissues
themselves unable to form reactive benzo[a]pyrene metabolites. Benzo[a]pyrene or metabolites
are also transferred to the fetus at concentrations one to two orders of magnitude lower than
those found in maternal tissues (Withey et al., 1993; Neubert and Tapkin, 1988; Shendrikov and
Aleksandrov, 1974).
3.3. METABOLISM
The metabolism of benzo[a]pyrene is a critical aspect of the assessment of its potential
toxicity because for many endpoints reactive metabolites are likely to contribute to the toxic
response. Numerous reviews on the metabolism of benzo[a]pyrene are available (Miller and
Ramos, 2001; WHO, 1998; ATSDR, 1995; Conney et al., 1994; Grover, 1986; Levin et al.,
1982; Gelboin, 1980). Key concepts have been adapted largely from these reviews and
supplemented with recent findings.
benzo[a]pyrene is metabolized extensively by Phase I reactions to form numerous
oxidative or reactive metabolites that are targets for further metabolism through diverse phase II
reactions. Many of the enzymes involved in benzo[a]pyrene metabolism are isoenzymes within
gene families, the members of which have varying metabolic specificities. Many of these
enzymes are encoded by genes that show functional polymorphism. Many of the critical
enzymes are inducible by a variety of agents, including benzo[a]pyrene itself, and, therefore,
studies that evaluate benzo[a]pyrene kinetics following single short-term exposures may not be
representative of the kinetics of benzo[a]pyrene following longer-term exposure conditions or
exposure to mixtures. Metabolism of benzo[a]pyrene is species-, strain-, and organ-system-
specific. There are age- and gender-related differences in the expression of many of the key
enzymes that must be considered.
The metabolism of benzo[a]pyrene has been extensively studied using both in vivo and in
vitro models, and a schematic representation of metabolic pathways is provided in Figure 3-1.
Only Phase I reaction products are shown. Phase II reactions include glutathione conjugation of
diol epoxides, sulfation and glucuronidation of phenols, and reduction of quinones by
NADPH:quinone oxidoreductase (NQO)with subsequent conjugation.
32
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1-OHBaP
Ba P 4,5-transdiol
BaP 7,8-oxide
BaP 1,6-hydroquinone BaP 1,6-semiquinone BaP 1,6-quinone
T BaP 3.6 1
Lsemiquinone J
Ba P 3,6- hydroquinon e
BaP 7,8-transdiol
T BaP 6.12 1
LsemiquinoneJ
< >
BaP 6.12-quinone BaP 6,12-hydroquinone
Ba P 7,8 -d iol-9,10- epoxide
7-OH BaP
6-ojco-BaP radical
Source: Miller and Ramos (2001).
Figure 3-1. Metabolic pathways for benzo[a]pyrene.
Some enzymes involved in the metabolism of benzo[a]pyrene are highly inducible
(although benzo[a]pyrene itself is a relatively weak inducer compared to other environmental
pollutants, such as certain dioxins and polychlorinated biphenyls). The cellular mechanisms
underlying the inducibility of these enzymes, as well as the relative potency of various inducers,
has been reviewed in detail (Miller and Ramos, 2001; Whitlock, 1999; Nebert et al., 1993).
Inducibility of genes that metabolize benzo[a]pyrene is termed genetic responsiveness and the
genetic locus that imparts inducibility is designated the Ah locus, so named for the enzyme
activity aryl hydrocarbon hydroxylase (AHH) (now known to be catalyzed by enzyme activities
from the CYP1 family. The Ah gene encodes a cytosolic receptor that regulates the inducible
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
expression of genes that encode Phase I (CYP1A and CYP1B isoforms) enzymes and may
interact coordinately in regulating genes that encode isoforms of the Phase II enzymes uridine
diphosphate (UDP)-dependent glucuronosyltransferases [UGTs], GSTs, NQOl) that metabolize
the products of Phase I metabolism. The role of these enzyme systems in benzo[a]pyrene
metabolism is discussed below. Due to the inducibility of benzo[a]pyrene metabolism,
interpretation of toxicity studies should consider whether the studies were conducted in species
and strains that have inducible metabolism, whether duration of exposure was sufficient to
induce benzo[a]pyrene metabolism, and whether or not known inducers were administered.
3.3.1. Phase I Metabolism
3.3.1.1. CYP450-dependent Reactions
Phase I reactions are catalyzed primarily by the mixed function oxidase system of
enzymes associated with CYP450 to form arene oxides. This initial phase I metabolic reaction
of benzo[a]pyrene is carried out primarily by the inducible activities of CYP1A1 or CYP1B1 and
the constitutively expressed and somewhat inducible CYP1A2, depending on the tissue. Other
CYP isoforms may also catalyze the initial oxidation reactions. NAPDH:CYP reductase is an
important cofactor for this reaction as a supplier of redox equivalents (Byczkowski and Gessner,
1989). In addition to the CYP450 enzymes there are also a number of oxidoreductases that are
not typically considered Phase I enzymes, yet can play an important role in the oxidative
metabolism of benzo[a]pyrene (see below).
The isomerization of the arene oxides to their respective phenol metabolites is thought to
be a nonenzymatic process; however, physical/chemical studies have shown that rearrangement
is susceptible to catalysis by amines (Johnson and Bruice, 1975). This would suggest that
rearrangements in vivo could be catalyzed by the amino or sulfhydryl groups of proteins.
Typically, a single phenolic isomer tends to be produced and the direction of regioselective ring
opening is predictable based on the relative stability of the two possible cationic intermediates
(Fu et al., 1978). In accordance with these predictions, exclusively 3-OH-benzo[a]pyrene but not
the other possible phenol metabolite, 2-OH-benzo[a]pyrene, is formed from the 2,3-oxide of
benzo[a]pyrene (Yang et al., 1977). The other monophenol metabolites of benzo[a]pyrene
include 1-, 6-, 7-, and 9-OH-benzo[a]pyrene. Benzo[a]pyrene epoxide formation may yield both
phenols and dihydrodiols. Arene oxides that are poor substrates for epoxide hydrolase (EH) or
the less stable ones that rearrange rapidly, such as benzo[a]pyrene 2,3-oxide, are less likely to
yield dihydrodiols.
The arene oxides can be hydrolyzed by EH to form dihydrodiols (Oesch, 1980). The
dihydrodiols may be further metabolized by CYPs to form diol epoxides, which are the DNA-
reactive metabolites that have been the subject of most studies. In particular, much of the study
of oxidative products of benzo[a]pyrene metabolism has been done for the 7,8-oxide, since it is a
precursor to the potent DNA-binding metabolites. The metabolism of benzo[a]pyrene, as well as
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
of PAHs in general, proceeds with a high degree of stereoselectivity. Since most aromatic bonds
are prochiral, their epoxidation catalyzed by CYPs often results in optically active products.
Liver microsomes from rats stereospecifically oxidize the 7,8-bond of benzo[a]pyrene to yield
almost exclusively the (+)-benzo[a]pyrene-(7,8)-oxide (see Figure 3-2). Each enantiomer of the
7,8-oxide is stereospecifically converted by EH to a different dihydrodiol via attack of water at
the 8-position. The (+)-benzo[a]pyrene-7,8-oxide gives rise to the (-)-benzo[a]pyrene-7,8-
dihydrodiol, while the (-)-benzo[a]pyrene-7,8-oxide yields the (+)-benzo[a]pyrene-7,8-di-
HO
t
OH
Epoxide Hydrolase |_|q
OH
(-)-BP-7,8-diol
(+)-BP-7,8-oxide
Mixed Function
2 Oxidase System
(+)-BP-7R,8S-diol-9S, 10R-epoxide
Mixed Function (+) anti BPDE
Oxidase System o s
(-)-BP-7R, 8S-diol-9R, 10S-epoxide
(-) syn BPDE
O *
Epoxide Hydrolase
(-)-BP-7,8-oxide
OH
(+)-BP-7,8-diol
HO M
OH
Mixed Function (+)-BP-7S,8R-diol-9S,10R-epoxide
(+) syn BPDE
Oxidase System
C
HO W
OH
(-)-BP-7S,8R-diol-9R, 10S-epoxide
(-) anti BPDE
hydrodiol.
Source: Grover (1986).
Figure 3-2. The stereospecific activation of benzo[a]pyrene.
Further metabolism of the (-)-benzo[a]pyrene-7,8-dihydrodiol enantiomer by rat CYP
enzymes preferentially yields (+)-benzo[a]pyrene-7R,8S-diol-9S,10R-epoxide [(+)-anti-BPDE],
which is believed to be the most potent carcinogen among the four stereoisomers (Figure 3-2).
Formation of these stereoisomers does not occur at equimolar ratios, and the ratios differ
between biological systems. For example, in a study with rabbit livers, purified microsomes
oxidized the (-)-benzo[a]pyrene-7,8-dihydrodiol to isomeric diol epoxides in a ratio ranging from
1.8:1 to 11:1 in favor of the (+)-anti-BPDE isomer (Deutsch et al., 1979).
Another important factor in evaluating variability in the metabolic activation of
benzo[a]pyrene is the degree to which functional polymorphisms plays a role. Schwarz et al.
(2001) used recombinant CYP1A1 allelic variants to determine catalytic activity in vitro.
Catalytic activity of the variant allele 1A1.4 was 70%, and that of variant allele 1A1.2 was only
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
50% of the WT allele, CYP1A1.1. Km values were generally lower for variants than for the
WT. Each variant produced BPDE, with the activity of CYP1A1.1 > 1.2 > 1.4. The formation
of diol epoxides was stereospecific, with the allelic variants producing about three times the
amount of (±)-anti-BPDE isomers, the suspected ultimate carcinogens, as compared to the
noncarcinogenic stereoisomers (±)-syn-BPDE. Wu et al. (2002) found no relationship between
benzo[a]pyrene metabolite formation and the CYP1A1 Mspl polymorphism. The identification
and characterization of CYP polymorphisms has been the subject of numerous reviews (e.g.,
Wormhoudt et al., 1999).
Several studies have attempted to clarify the question of which CYP isozyme is
predominantly responsible for the metabolism of benzo[a]pyrene. The studies used knock-out
(ko) animals in which either one of the isozymes in question, CYP1 Al 1A2, or 1B1, or the Ah
receptor (AhR) had been removed or inactivated (CYP 1A1 and 1B1 levels respond to AhR
induction, while 1A2 is expressed constitutively). Kleiner et al. (2004) measured DNA adduct
formation in the epidermis of 1B1 1 , 1 A2 ; , AhR , and WT mice. Six [3H]-PAHs were
administered topically in one dose of 10-2000 nmol and animals were sacrificed after 24 hours.
Absence of CYP1A2 had very little effect on benzo[a]pyrene adduct formation; in 1B1 mice
adducts were about 150% of WT (not significant), while in AhR they were only 27% of WT.
These findings differed considerably with other PAHs. The benzo[a]pyrene-DNA adduct was
identified as being derived from (+)-anti-BPDE. The authors concluded that 1 Al was the
primary CYP to metabolize benzo[a]pyrene, and that 1B1 rather serves detoxification.
Sagredo et al. (2006) conducted a similar experiment, but with various types of AhR
knock out mice. AhR+/+, , and mice were treated once with 100 mg/kg benzo[a]pyrene by
gavage. Twenty-four hours after treatment gene expression for CYP1 Al, 1B1, and AhR was
measured in lung, liver, spleen, kidney, heart, and blood by real-time or reverse transcriptase
PCR (RT-PCR). CYP1 Al expression was increased following benzo[a]pyrene treatment in +/+
and +/~ mice (generally higher in heterozygotes), but mice expressed no 1A1. There was a low
level of basic 1B1 expression in all three genotypes that was inducible by benzo[a]pyrene in
lung, but not in liver in AhR+/+ and , but not at all in either tissue of AhR mice. Expression
of 1A1 was 25-40 times that of 1B1. There was an AhR gene-dose-response relationship for the
basal CYP1A1 expression, i.e., +/+ > +/- > -/-, but no such dependence was seen for 1B1.
Protein adduct levels were spleen > liver > lung > heart > plasma > kidney. The tissue levels of
adducts showed an inverse relationship with AhR gene dose, i.e., -/- > +/- > +/+. Similarly, the
levels of unmetabolized benzo[a]pyrene and of free benzo[a]pyrene-tetrol metabolites were
higher in all organs of AhR mice, as compared with AhR+ mice. Also, in the livers of AhR
mice the levels of the less carcinogenic tetrol II were much higher than those of the tetrol I,
opposite to the situation in AhR+ mice. The authors suggested that the high levels of free
benzo[a]pyrene metabolites in AhR mice were the result of delayed bioactivation, and that a
powerful AhR-independent pathway for benzo[a]pyrene metabolism must exist. The authors
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
explained the very high levels of benzo[a]pyrene adducts in organs outside the liver as the result
of slow detoxification of the agent in the liver of AhR mice, allowing high concentrations of the
parent compound to reach distant tissues. These findings establish important roles in
benzo[a]pyrene metabolism for both CYP1A1 and 1B1, but they do not clarify which enzyme is
responsible for biological activation, and which for detoxification.
Uno et al. (2006) investigated the finding that CYP1A1 (1 Al ) knock out mice are more
sensitive than WT animals to the toxic effects of orally administered benzo[a]pyrene. They
produced a series of C57BL/6-based single- or double knock out mice, 1A2^~, 1B1 1 , 1A1/1BU
and 1A2/1B1 Benzo[a]pyrene was administered in the feed at 1.25, 12.5, or 125 mg/kg for
18 days (this dose is well tolerated by WT C57BL/6 mice for 1 year, but lethal within 30 days to
the 1 Al ^ type). Steady-state blood levels of benzo[a]pyrene, reached within 5 days of
treatment, were -25 times higher in lAl_/~ and -75 times higher in lAl/lBl^ than in WT mice,
while in the other knock out types clearance differed little from that in WT animals.
Pretreatment of the animals with TCDD to induce the CYP1 enzyme family resulted in decreased
benzo[a]pyrene peak blood levels and AUCs in WT, but doubled peak blood levels and vastly
increased AUCs in the two 1 Altypes. A lower-than-WT benzo[a]pyrene clearance was
observed only in the two 1A1 types. DNA adduct levels, measured by [ P]-postlabeling in
liver, spleen, and bone marrow, were highest in the 1 Almice at the higher doses, and in the
1A1/1B1mice at the mid dose only. Only 1 Al_/~ mice, but not the other genotypes, showed
signs of severe toxicity. In conclusion, the authors of the study painted a rather complex picture
of how the three CYP1 family enzymes affect the toxicokinetics of benzo[a]pyrene.
Detoxification of benzo[a]pyrene is mostly achieved by 1 Al, probably not only in liver, but also
in the intestine. Second, in spleen and bone marrow 1B1 brings about metabolic activation of
benzo[a]pyrene, which, in the absence of 1 Al, results in damage to the immune system. The
authors suggested that tissue-specific expression of 1A1 and 1B1, respectively, determine an
organism's susceptibility to benzo[a]pyrene toxicity and, possibly, carcinogenicity.
Van Lipzig et al. (2005) conducted experiments concerning potential estrogenic activity
of mono- and dihydroxylated metabolites of benzo[a]pyrene. Estrogen receptor (ER) affinity
and estrogenic activity were tested in T47D human breast adenocarcinoma cells (for
experimental details see Section 4.5.2). Benzo[a]pyrene metabolite mixtures were generated
using (3-naphthoflavone ((3-NF)-induced rat liver microsomes. Several hydroxylated
benzo[a]pyrene metabolites had measurable estrogenic activity. In an attempt to identify
enzymes involved in benzo[a]pyrene hydroxylation, the researchers added a specific CYP1A2
inhibitor to the metabolic activation mixture and found a -15-fold increased estrogenic activity
of the benzo[a]pyrene metabolite mix. They suggested that 1A2 inhibition drove the metabolism
of benzo[a]pyrene towards formation of more estrogenic metabolites. These findings impart a
role to CYP1A2 in the metabolism of benzo[a]pyrene.
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
To summarize, there is experimental evidence to suggest that three members of the CYP1
family, 1A1, 1A2, and 1B1, contribute to the metabolism of benzo[a]pyrene. However, the
available data do not attribute precise roles to either of these enzymes. There is evidence that
their expressions, and thus activities, are organ- or tissue-specific and that local isozyme
activities determine not only the ratio of toxification vs. detoxification, but also the pattern of
highly toxic vs. less toxic metabolites.
3.3.1.2. Non-CYP-related metabolic pathways
Other bioactivation processes independent of CYP-mediated diolepoxide formation have
been demonstrated for benzo[a]pyrene. One-electron oxidation (via CYPs or peroxidases) can
generate radical cations, which, in turn, can generate benzo[a]pyrene quinones (see bottom
portion of Figure 3-1). These metabolites may generate DNA damage through redox cycling or
the formation of depurinating adducts (McCoull et al., 1999; Cavalieri et al., 1990). Kim et al.
(2000) treated male Sprague-Dawley rats with 20 mg/rat benzo[a]pyrene by i.p. injection and
reported that the pattern of lipid peroxidation and increase in antioxidant enzymes correlated
with formation of quinone metabolites of benzo[a]pyrene. Direct i.v. injection of
benzo[a]pyrene and a series of metabolites confirmed the quinone metabolites indeed to be
associated with the observed increase in lipid peroxidation. Secondary oxidation of 6-OH-
benzo[a]pyrene can generate hydroquinone and quinone metabolites at the 1,6-, 3,6-, or 6,12-
positions. Quinone formation may be catalyzed by dihydrodiol dehydrogenases (DHHs), such as
aldo-keto reductase (AKR) (Penning et al., 1999). Tsuruda et al. (2001) showed in vitro that the
expression of rat DHH in human breast cancer MCF-7 cells generated 7,8-benzo[a]pyrene-diones
from benzo[a]pyrene-7,8-diol.
Mallet et al. (1991) treated tetradecanoylphorbol acetate-stimulated human
polymorphonuclear leukocytes (PMNs) with (±)-trans-7,8-dihydroxy-7,8-dihydro-
benzo[a]pyrene. They found that the cells were able to transform the benzo[a]pyrene-diol into
the diolepoxide and tetrols with a stereochemical anti/syn ratio of six. The kinetics of the
reaction suggested that hydrogen peroxide or a ferryl-oxygen-transfer were involved. Because
myeloperoxidase (MPO) uses hydrogen peroxide in its reaction, the authors inhibited this
enzyme specifically with azide and found that the formation of tetrols from benzo[a]pyrene-diol
was reduced. Thus, MPO is able to execute the metabolic activation of benzo[a]pyrene.
Byczkowski and Kulkarni (1990) observed that benzo[a]pyrene diol can be cooxygenated
during lipid peroxidation to form the diolepoxide. To reduce interference from CYP-catalyzed
reactions they used term human placental microsomes, which are low in CYP450. Peroxidative
conditions were created by a redox cycling system comprised of partially chelated ferrous ions
and NADPH:CYP450 oxidase; this system quadrupled the formation of malonic dialdehyde (a
measure of lipid peroxidation) compared with placental microsomes alone. The peroxidative
system increased overall metabolism of benzo[a]pyrene and benzo[a]pyrene-7,8-diol by 27-
38
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
28%, compared with placental microsomes alone. The amounts of individual metabolites were
changed to various extents in the presence of peroxidative conditions; the most striking
observation was the more than doubled formation of trans-anti-benzo[a]pyrene-tetrol (the
proximate carcinogen) and an almost tripled binding to protein. There was a highly significant
correlation (p < 0.0005) between malondialdehyde and trans-anti-benzo[a]pyrene-tetrol
formation. The authors pointed out that by means of this metabolic pathway the human fetus
could be exposed to BPDE despite the absence of pronounced CYP450 activity in term placenta.
Similarly, a redox cycling system based on vanadate-IV ions was able to increase formation of
trans-anti-benzo[a]pyrene-tetrol five- to sixfold in the presence of term human placental
microsomes (Byczkowski and Kulkarni, 1992).
Flowers et al. (1996) investigated the role of radical and reactive oxygen species (ROS)
formation in benzo[a]pyrene-quinone-induced redox cycling, a reaction involving DHH, which
can oxidize BP-diol to benzo[a]pyrene-7,8-dione (BPQ). BPQ is mutagenic and genotoxic. In
isolated rat hepatocytes BPQ was incorporated covalently into DNA (30 ± 17 adducts/106 base
pairs) while extensive DNA fragmentation took place. DNA fragmentation was also observed in
hepatocytes treated with BP-diol; the effect was partially abolished when an inhibitor of DHH
was added to the reaction. Hepatocytes treated with either BP-diol or BPQ produced superoxide
anion radical, formation of which could be blocked by DHH inhibitors. In an in vitro experiment
it was shown that BPQ at 0.05-10 |iM caused DNA strand scission in the presence of NADPH
and CuCl2, suggesting that redox-cycling took place. DNA strand scission was prevented by
catalase and hydroxyl radical scavengers but not by superoxide dismutase. The authors
concluded that DHH metabolizes (+/-)-anti-BPDE to BPQ, which in turn causes extensive DNA
fragmentation via the generation of ROS.
In a subsequent study, Flowers et al. (1997) provided more detail on the redox cycling
reaction of BPQ. They showed that the reaction required the presence of NADPH (1 mM) and
2+
Cu (10 |iM), During the reaction superoxide anion radicals, benzo[a]pyrene semiquinone
radicals, hydroxyl radicals, and H2O2 were formed. Hydroxyl radical scavengers, such as
mannitol, sodium benzoate, or formic acid prevented the redox cycling (as assessed by DNA
strand scission), as did the Cu+ chelators bathocuproine or neocuproine. The results were
interpreted as indicating that redox cycling of benzo[a]pyrene quinone involves a Cu+-catalyzed
Fenton reaction.
Other peroxidases such as prostaglandin H synthase (PHS) may also generate radical
cations from benzo[a]pyrene (Parman and Wells, 2002). Marnett (1990) reviewed the role of
PHS in benzo[a]pyrene metabolism. Peroxy radicals epoxidize the procarcinogen
benzo[a]pyrene-7,8-diol via PHS to the epoxide BPDE. Stereochemical experiments have
allowed the distinction between peroxide-mediated and CYP450-mediated epoxide formation
from benzo[a]pyrene. Thus, peroxy radical-dependent epoxidation of benzo[a]pyrene-7,8-diol
occurs in rat liver microsomes, mouse skin homogenates, cultured fibroblasts, cultured hamster
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
trachea, and freshly isolated mouse epidermal cells. Peroxy radical-generated metabolites are
predominant in uninduced animals, while in B-NF-induced animals the CYP450-produced
metabolites prevail. There are other pathways by which peroxides can oxidize benzo[a]pyrene-
7,8-diol because inhibition of PGH synthase with non-steroidal anti-inflammatory drug
(NSAIDs) does not prevent BPDE formation.
Redox-active quinones are formed through the oxidative metabolism of benzo[a]pyrene,
particularly at the 6-position (see Figure 3-1). NQOl is an important enzyme for the
detoxification of reactive quinones. Joseph and Jaiswal (1998) reported that NQOl expression
inhibited the formation of benzo[a]pyrene-quinone adducts with DNA and mutagenicity in vitro.
There are also potential ring opening mechanisms for benzo[a]pyrene. Stansbury et al. (2000)
used activated polymorphonuclear monocytes or a reconstituted in vitro system to generate a
ring-opening dialdehyde metabolite from the benzo[a]pyrene-7,8-diol-9,10-epoxide (BPDE) that
can react with DNA to generate unique DNA adducts that are different from the ones formed by
BPDE.
Nordling et al. (2002) identified a novel benzo[a]pyrene metabolite (7-oxo-
benz[d]anthracene-3,4-dicarboxylic acid anhydride) in the urine of benzo[a]pyrene-treated rats.
Follow-up in vitro experiments with this compound found it to be weakly mutagenic in
salmonella strain TA102, able to induce DNA stand breaks in FIT-29 cells, capable of inducing
cytotoxicity via an apoptotic mechanism, and increasing gene expression through a
cyclooxygenase (COX)-2 promoter in HCT 116 cells. These results demonstrate that novel
benzo[a]pyrene metabolites have toxicological properties distinct from those of the better studied
BPDE.
The roles of PHS-2 (now mostly called COX-2), MPO, NQOl, and other enzymes in the
oxidative metabolism of benzo[a]pyrene may be crucial, but not enough data are as yet available
to attempt a quantitative comparison with the CYP-mediated pathways. This may be of some
importance in the assessment of cancer risks because not only CYP450 isozymes, but also MPO
and NQOl exhibit gene polymorphism. This is an area where much more research is necessary.
3.3.2. Phase II Metabolism
The reactive products of phase I metabolism are subject to the action of several phase II
conjugation and detoxification enzyme systems that display preferential activity for specific
oxidation products of benzo[a]pyrene. These phase II reactions play a critical role in protecting
cellular macromolecules from binding with reactive benzo[a]pyrene diolepoxides, radical
cations, or ROS. Therefore, the balance between Phase I activation of benzo[a]pyrene and its
metabolites and detoxification by Phase II processes is an important determinant of toxicity.
The diol epoxides formed from benzo[a]pyrene metabolism are not usually found as
urinary metabolites, rather they are detected as adducts of nucleic acids or proteins if not further
metabolized. Detoxification of the diol epoxide metabolites of benzo[a]pyrene is through
40
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
rearrangement to form tetrols or via conjugation with glutathione. Early work by Gozukara et al.
(1981) demonstrated that exogenous EH added to benzo[a]pyrene-treated human monocytes
reduced DNA binding, suggesting a role for this enzyme in the inactivation of diol epoxides.
Furthermore, nonenzymatic hydration of diol epoxides proceeds rapidly in aqueous media in the
absence of EH to yield tetrol products via cis or trans addition. A second mechanism for the
detoxification of reactive diol epoxides formed from benzo[a]pyrene is through glutathione
(GSH) conjugation. This Phase II reaction is catalyzed by GSTs. GSTs are a family of enzymes
with varying substrate specificity and distribution among tissues (reviewed in Hayes and
Strange, 2000; Eaton and Bammler, 1999; Hayes and Pulford, 1995). Primary isoforms of
relevance for conjugation of BPDE include the a, %, and 9 isoforms (GSTA, GSTM, GSTP,
GSTT, respectively).
Numerous studies using human GSTs expressed in mammalian cell lines have
demonstrated the ability of GST to metabolize benzo[a]pyrene diol epoxides. For example, Dreij
et al. (2002) demonstrated that GST isozymes, including alpha class GSTA1-1, GSTM1-1, and
GSTP1-1 isoforms, had significant catalytic activity toward benzo[a]pyrene-derived diol
epoxides. Robertson et al. (1986) incubated isolated human GST isoforms with GSH and
(±)anti-BPDE to assess differences in their catalytic properties. Maximum substrate turnover
velocity (Vmax) values for a, |i, and n were 38, 570, and 825 nmol mg"1 minute"1, and Km values
were 28, 27, and 54 [xM, respectively. Rojas et al. (1998) reported that no BPDE adducts were
formed in GSTM 1-positive cells, but adducts were present in GSTM1-negative cells. This body
of in vitro studies suggests that GST is an important detoxification mechanism for
benzo[a]pyrene-derived epoxides. This compelling evidence for a role of GSTs in protecting
against reactive benzo[a]pyrene metabolites has triggered several molecular epidemiology
studies. However, recent studies on the impact of polymorphism on adduct levels in PAH-
exposed human populations did not succeed in showing clear relationships between CYP1 Al,
EH, or GSTM1 polymorphisms and DNA (Hemminki et al., 1997) or blood protein adduct
formation (Pastorelli et al., 1998).
Conjugation with glucuronide is another important detoxification mechanism for
oxidative benzo[a]pyrene metabolites. Most of the phenolic metabolites of benzo[a]pyrene are
further metabolized by glucuronidation or sulfation, and significant portions of total metabolites
in excreta or tissues can be recovered in this form. Bevan and Sadler (1992) administered a
single 2 (J,g/kg benzo[a]pyrene dose by intratracheal instillation to male Sprague-Dawley rats and
assessed the benzo[a]pyrene metabolite profiles in bile after 6 hours. Identified metabolites were
31.2% quinol diglucuronides, 30.4% thioether conjugates, 17.8% monoglucuronide, 6.2% sulfate
conjugates, and 14.4% unconjugated metabolites.
The UGTs are a family of enzymes that catalyze the conjugation of UDP-glucuronide
with endogenous substrates (e.g., bilirubin) as well as xenobiotics (reviewed in Guillemette,
2003). UGT isoforms as well as their allelic variants show different patterns of tissue
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
distribution and catalytic activity toward benzo[a]pyrene-derived phenols and diols. For
example, Fang and Lazarus (2004) assessed the ability of the allelic variants UGT1A1 and 1A9
in human liver to catalyze glucuronidation of benzo[a]pyrene-7,8-diol. Microsomes from
subjects homozygous for the UGT1A1*28 variant (present in approximately 12% of the
Caucasian population) had approximately twofold lower UGT1A1 protein levels than liver
microsomes from individuals with the WT allele. Addition of a UGT1A9 inhibitor to the
incubation decreased benzo[a]pyrene-glucuronide formation three- to sixfold, suggesting that
UGT1A9 also has significant catalytic activity toward benzo[a]pyrene-7,8-diols. With UGT1A9
activity blocked, glucuronide formation in UGT1 Al*28 homozygotes was significantly lower
than in WT and heterozygous individuals. The apparent Km for this reaction did not differ
among microsomes from allelic variants.
UGT activity also shows significant interindividual variability. Hu and Wells (2004)
evaluated glucuronidation of benzo[a]pyrene metabolites in human lymphocytes (HL) in vitro.
The degree to which glucuronide conjugates were formed varied over 200-fold (percent of
metabolites as glucuronide conjugates ranged from 0.01 to 5% of total benzo[a]pyrene
metabolites). Incubation of lymphocytes with benzo[a]pyrene, benzo[a]pyrene-7,8-diol, or
benzo[a]pyrene-4,5-diol resulted in covalent binding to protein with, in the case of
benzo[a]pyrene, a 220-fold inter-individual variability. Addition of the UGT substrate, UDP
glucuronic acid (UDPGA), lowered the inter-individual variability to 143-fold. For
benzo[a]pyrene or its diols there was a statistically highly significant relationship between
increase in glucuronidation and decrease in protein binding. Cytotoxicity also was inversely
correlated to conjugation of diols and diones, suggesting that glucuronidation is an important
pathway for protection from chemically reactive benzo[a]pyrene metabolites.
Sulfation, normally a defoliation process, can produce a DNA-damaging intermediate in
the case of benzo[a]pyrene. It was shown that in rat or mouse liver cytosolic sulfotransferase (in
the presence of 3'-phosphoadenosine 5'-phosphosulfate) catalyzes formation of sulfates of
benzo[a]pyrene-7,8,9,10-tetrahydro-7-ol, benzo[a]pyrene-7,8-dihydrodiol, and benzo[a]pyrene-
7,8,9,10-tetrol. All three sulfates were tested for their ability to bind to DNA, but only the
benzo[a]pyrene-7,8,9,10-tetrahydro-7-ol-sulfate formed DNA adducts (Surh and Tannenbaum,
1995).
Although not specific for benzo[a]pyrene, there is now considerable evidence that genetic
polymorphisms of the GST, UGT, and EH genes impart an added risk to humans for developing
cancer. Of some significance to the assessment of benzo[a]pyrene may be that smoking, in
combination with genetic polymorphism at several gene loci (for detail, see Section 4.8.3),
increases the risk for bladder cancer (Moore et al., 2004; Choi et al., 2003; Park et al., 2003) and
lung cancer (Alexandrie et al., 2004; Lin et al., 2003). Also, Leng et al. (2004), according to the
English abstract of a paper in Chinese, showed that coke oven workers (who are exposed to
PAHs, including benzo[a]pyrene) homozygous at the P187S site of the NQOl gene or carrying
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
the null variant of the GSTM1 gene had a significantly increased risk of chromosomal damage in
peripheral blood lymphocytes, while the risk was much lower than controls in subjects with a
variant allele at the HI 13 Y site of the EH gene.
3.3.3. Tissue-specific Metabolism
3.3.3.1. Respiratory Tract Tissues
benzo[a]pyrene treatment has been associated with the induction of respiratory tract
tumors. This finding is consistent with the ability of the lung to metabolize benzo[a]pyrene,
which has been demonstrated in numerous studies. Ewing et al. (2006) investigated the
hypothesis that lung cancer following PAH induction may be a result of slow absorption and
extensive metabolism in the thick respiratory epithelia. The researchers used isolated, perfused
rat lung to investigate these processes. Benzo[a]pyrene was coated onto 3.5 [j,m silica carrier
particles at concentrations to deliver an average of 2.2, 36, and 8,400 ng to each lung within <1
minute. Perfusate was collected for 77 minutes thereafter. Both perfusates and lungs were
analyzed for benzo[a]pyrene and metabolites. Absorption and metabolism were both strongly
dose-dependent: at the low and mid exposure levels benzo[a]pyrene concentration increased
rapidly in the perfusate to a maximum after <5 minutes, then decreased over the remaining
observation period. At the high dose, benzo[a]pyrene in perfusate reached the maximum at
about 30 minutes after exposure and stayed at a constant level from there on, i.e., the absorption
of benzo[a]pyrene proceeded at zero order until all deposited solid benzo[a]pyrene was
dissolved. The mass balances for benzo[a]pyrene equivalents in lung vs. perfusate were lung ca.
one-third perfusate at the low dose, lung = perfusate at the mid dose, and lung about twice that of
perfusate at the high dose. At the low exposure level metabolism was apparently able to convert
most of the parent compound, while at the highest exposure level most of the absorbed
benzo[a]pyrene remained unmetabolized even at the end of the experiment. The authors pointed
out that these findings may explain why many attempts have failed to inducer lung cancer in
animals using high-dose particle inhalation protocols. The results further confirm that
benzo[a]pyrene metabolism is organ specific.
Autrup et al. (1980) compared the metabolic capacity of tracheobronchial tissues in
culture among several species, including humans, mice, rats, hamsters and bovines. Results from
this study are summarized in Table 3-2. Benzo[a]pyrene was metabolized extensively in tissues
from all species tested, with lower amounts of metabolites identified in rats and nonresponsive
mice. Patterns of metabolism differed among the species but showed formation of a complex
array of metabolites, including phenols, diols, tetrols, and quinones. Data summarized from the
study suggest that under the conditions tested: (1) upper respiratory tract tissues for all species
were able to metabolize benzo[a]pyrene, (2) the degree of phase II conjugation products was
greatest in humans, followed by hamsters, Ah-responsive mice, bovines, rats, and Ah-
nonresponsive mice, (3) multiple phase II conjugation pathways were operative in tissues of all
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
species, although the relative proportions of conjugate formation varied, and (4) conducting
airway tissues from all species were able to metabolize benzo[a]pyrene into DNA-reactive
metabolites, with DNA binding greatest in hamster trachea followed by human and bovine
bronchus.
DNA binding in human, rat, and mouse tissue was similar but considerably higher in
hamster (Table 3-2). The results were quite variable among individuals, a 33-fold difference in
human bronchus, a fivefold variation in human trachea, and a threefold difference in bovine
bronchus, but minimal variation among individuals of the laboratory animal species. Overall,
these results show that human lung tissue metabolizes benzo[a]pyrene in a manner that is
qualitatively similar to that of species that are susceptible to lung tumors, although some
quantitative differences in specific metabolic pathways are observed (Autrup et al., 1980).
Table 3-2. Species differences in tracheobronchial benzo[a]pyrene
metabolism
Species
Total
metabolites"
Ratio organic/water
soluble metabolites
Percent water soluble metabolites as
sulfate esters, glucuronides, and
glutathione conjugates
DNA
bindingb
C57B1/6N mouse
1.00 ±0.25
1.4
31/30/39
10
DBA/2N mouse
0.40 ±0.13
0.5
28/27/44
10
CD rat
0.65 ±0.10
0.5
31/13/56
10
Syrian golden hamsters
1.34 ± 0.13
1.2
20/17/63
26
Bovine bronchus
0.93 ±0.12
0.12
24/29/38
16
Human trachea
1.09 ±0.48
2.5
56/12/32
11
Human - main-stem
bronchus
1.33 ±0.72
1.7
44/7/51
16
Human secondary and
tertiary bronchus
1.75 ±0.82
2.3
44/6/51
16
aMean ± SD in pmol/(xg DNA.
bMean ± SD in pmol/mg DNA; results are from trachea in mouse, rat, and hamster.
Source: Autrup et al. (1980).
In vitro studies with human bronchial epithelial and lung tissue showed that
benzo[a]pyrene is metabolized to the 7,8- and 9,10-diols and, to a lesser extent, to the 4,5-diol
and 3-OH metabolites (Autrup et al., 1978; Cohen et al., 1976). The metabolites were identified
as glutathione and sulfate conjugates; no glucuronide metabolites were found. The ability of
human tissues to metabolize benzo[a]pyrene has also been demonstrated in lung-derived cell
lines. Kiefer et al. (1988) demonstrated that benzo[a]pyrene was metabolized in vitro in the
human lung cancer cell line NCI-H322. These cells were also able to form benzo[a]pyrene-7,8-
diol, suggesting that human lung cells are able to generate carcinogenic metabolites of
benzo[a]pyrene. Approximately 30% of the detected metabolites were water-soluble, about 30%
44
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
of which were glutathione conjugates. Sulfates, but not glucuronide conjugates, were also
detected.
A complement of enzymes for the oxidative metabolism of benzo[a]pyrene in the lungs
has been identified in both humans and animal tissues, and these activities are inducible with
prior exposure. Wei et al. (2001) evaluated CYP1A1 levels in fresh lung tissue from nine human
donors. CYP1 Al and CYP1A2 were present at variable levels in lung tissues based on mRNA,
protein levels, enzyme activity, and ability of S9 fractions to induce mutagenicity in an Ames
assay. CYP1B1 was not identified. The authors emphasized that, in contrast to some previous
studies, they were able to identify CYP1A2 in human lung (Wei et al., 2001). In a subsequent
study, Wei et al. (2002) lung tissue from 27 human donors was evaluated for CYP1A status.
CYP1 Al and CYP1A2 transcripts were present at variable levels in nearly all samples and were
inducible by benzo[a]pyrene treatment. Mean inducible CYP1A2 levels were roughly four times
lower than CYP1A1. Microsomes prepared from these tissues pretreated with benzo[a]pyrene
resulted in a threefold increase in DNA adduct formation, while pretreatment with the potent
AhR ligand TCDD increased benzo[a]pyrene-DNA adduct formation to 24-fold over controls.
This result shows that that CYP1A activity was highly inducible via the AhR pathway in human
lung tissues.
These results using human donor tissues are consistent with the body of literature
demonstrating the induction of benzo[a]pyrene metabolism in lungs of rodents. For example,
Vainio et al. (1976) compared the metabolism of intratracheally-instilled benzo[a]pyrene in the
isolated perfused rat lung of both control and rats induced with 3-MC. Pretreated rat lungs had
"3
increased covalent binding of [ H]-benzo[a]pyrene to lung tissue, decreased appearance of
unmetabolized benzo[a]pyrene in perfusion liquid, and increased formation of water soluble
metabolites. Bompart and Clamens (1990) assessed AHH activity in male Sprague-Dawley rats
given 2 mg/kg benzo[a]pyrene by i.p. injection weekly for 30 weeks. Every third week, five
animals were sacrificed and lung and liver microsomes were prepared for determination of AHH
activity (as measured by formation of the 3-OH metabolite of benzo[a]pyrene). Control levels of
AHH in lung were much lower than in liver. AHH activity in the lung increased with repeated
benzo[a]pyrene doses until approximately week 15, when it reached levels approximately
eightfold over controls. Even after this induction, lung AHH activity was still approximately 30-
fold lower than in liver. Benzo[a]pyrene treatments had no effect on AHH activity in the liver in
this study.
Petridou-Fischer et al. (1988) instilled radiolabeled benzo[a]pyrene into the nasal
turbinates (ethmoid and maxillary) of monkeys and dogs. Metabolic activity in these tissues was
demonstrated by the formation of diverse metabolic products (phenols, diols, tetrols, and
quinones). No region-specific metabolism was identified. Even though ethmoid versus
maxillary turbinates contain different CYP activities, the pattern of metabolites was qualitatively
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
similar. Differences were not compared quantitatively due to the small number of animals used
(four dogs and two monkeys).
Bond et al. (1988) showed differential levels of Phase I and Phase II activities in
respiratory tract regions of dogs. Benzo[a]pyrene metabolism was greater in nasal tissue than in
lung tissue. Metabolic activities towards benzo[a]pyrene in various areas of the upper
respiratory tract ranged from 5 to 15 pmol/mg protein/minute; however, in the ethmoid turbinates
they reached approximately 45 pmol/mg-minute. The was no difference in the regional patterns
of metabolite formation. Although CYP isozyme activity was similar in liver and lung,
benzo[a]pyrene metabolism in respiratory tract tissues was about one-tenth that in liver (one-
third for the ethmoid turbinates). EH activity was highest in the lower generations of the
conducting airways, followed by liver then nasal tissues. GST activity was highest in liver,
followed by nasal tissues. UGT activity was more evenly distributed among lung regions, and
was similar to levels in the liver. These data show that in dogs the nasal region and lungs have
greater metabolic capability for benzo[a]pyrene and its metabolites than the conducting airways,
and levels of metabolism are generally similar to those observed in the liver.
Dahl et al. (1985) evaluated benzo[a]pyrene metabolism in the respiratory tract tissues
from Syrian hamsters. All regions of the respiratory tract had metabolic activity as assessed by
the formation of benzo[a]pyrene metabolites. Activity was highest in the nasal tissues on a per
gram tissue basis, with similar activities observed in esophagus, forestomach, trachea, larynx,
and lungs. Total metabolism on a per organ basis was highest for the lung and trachea. Similar
results were obtained for lung and nasal tissue of rats that had inhaled benzo[a]pyrene (Wolff et
al., 1989).
Persson et al. (2002) showed that [ H]-benzo[a]pyrene instilled nasally in female
Sprague- Dawley rats was taken up in nasal structures (sustenacular cells and Bowman's glands).
Transport of benzo[a]pyrene or metabolites (as determined by radiography) via axons of
olfactory neurons to the olfactory bulb was identified, indicating uptake into nasal structures and
transport to central nervous system (CNS) structures via neurons.
Weyand and Bevan (1986) examined benzo[a]pyrene disposition and metabolism in male
Sprague-Dawley rats following intratracheal instillation of 1 (J,g/kg body weight [ H]-
benzo[a]pyrene. The overall concentration of benzo[a]pyrene metabolites in lung and liver
decreased over the 360-minute period, with a shift from predominately lipid-soluble metabolites
to an increasing component of water-soluble metabolites at later times. In the lung,
benzo[a]pyrene metabolites represented 47.7% of the administered dose in the organic and
11.9% in the aqueous fraction, respectively, at 5 minutes, while at 360 minutes the corresponding
fractions were 2.16 and 1.83%), respectively. The metabolite profile determined in the organic
phase at 360 minutes was as follows: conjugates or polyhydroxylated compounds, 16.3%>; 9,10-
diol, 4.95%; 4,5-diol, 2.60%; 7,8-diol, 3.11%; 1,6-quinone, 2.99%; 3,6-quinone, 4.09%; 6,12-
quinone, 2.23%; 9-OH, 2.26%; 3-OH, 4.59%; and benzo[a]pyrene, 20.0%. Over the 360-minute
46
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
period, notable shifts in the relative proportion of benzo[a]pyrene and metabolite levels in the
lung indicated a net increase in the contribution of conjugates and tetrols, stable to decreasing
levels of quinones, increasing levels of diols and phenols, and decreasing levels of parent
benzo[a]pyrene. A similar pattern was observed for concentrations of benzo[a]pyrene
metabolites in the liver, except that quinone levels also increased in this tissue over the 6-hour
period. This identification of significant levels of quinone metabolites early in the lungs is
consistent with in vitro studies, such as that of Autrup et al. (1980) in which rat lung had a high
capacity to form quinones originating from the oxidation of benzo[a]pyrene at the 6 position.
Most of the metabolism studies have focused on the role of the CYP1 isoforms on
benzo[a]pyrene metabolism in the lung. However, other CYPs may also be important for
benzo[a]pyrene metabolism among species. For example, Shultz et al. (2001) reported that
benzo[a]pyrene was metabolized using recombinant CYP2F2 from mouse lung in a cell-free in
vitro assay, although metabolic capability was less than that of other isoforms. CYP2F2
expression in mouse lung airways was greater than that in tracheal parenchyma, showing region-
specific metabolic differences.
3.3.3.2. GI Tract and Liver Tissues
Fontana et al. (1999) reported that in healthy volunteers fed diets enriched with char-
grilled meat, CYP1A1 and 1A2 activities were induced in the liver and CYP1A1 protein levels
were increased in small intestine biopsies. No change in CYP3 A4 or 3 A5 levels was observed.
DNA adducts in peripheral blood cells were inversely correlated to CYP1A levels in enterocytes
and CYP1A2 levels in liver. These findings point to the presence of AhR ligands, such as
benzo[a]pyrene, in char-grilled meat and lend further support to the complexity of
benzo[a]pyrene metabolism by CYP isozymes.
"3
In a human hepatoma cell line (HepG2) incubated with [ H]-benzo[a]pyrene,
radiolabeled metabolites were recovered primarily in the medium (88.4% of the radioactive
material) (Diamond et al., 1980). Sixty-four percent of the metabolites were unidentified water-
soluble metabolites. Chloroform extractable metabolites (36% of the radiolabel) included 7,8-
diol, 9,10-diol, quinones, 3-OH (16% combined) metabolites, and unmetabolized benzo[a]pyrene
(20%). The cell lysate contained the same metabolites, but the proportions of the 3-OH
metabolite and parent compound were relatively higher. Enzymatic treatment for conjugate
formation did not change recovery of radioactivity, suggesting that at least this tumor cell line
did not extensively form phenol products. The authors noted that the HepG2 cell line did not
utilize the major phenol detoxification pathway of rodent cell cultures.
Similar results were obtained with human hepatocytes in culture (Monteith et al., 1987).
Following incubation for 24 hours, the primary metabolites of [ H]-benzo[a]pyrene were
unidentified highly polar, water-soluble conjugates. The next four most prevalent metabolites
consisted of 3-OH-benzo[a]pyrene and the 4,5-, 9,10- and 7,8-diols. As the dose of
47
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
benzo[a]pyrene was increased from 10 to 100 pmol, the amount of metabolites increased in a
linear fashion, suggesting that the capacity of human hepatocytes to metabolize benzo[a]pyrene
was not saturated at benzo[a]pyrene concentrations up to 100 [j,mol. 3-MC induced rat
microsomes convert benzo[a]pyrene to the BPDE approximately 10 times faster than
microsomes from uninduced rats, likely through the induction of CYP1A1 (Keller et al., 1987).
The rate-limiting step appeared to be competition for CYP1 Al binding sites between parent
benzo[a]pyrene and the benzo[a]pyrene-7,8-diol.
The metabolism of benzo[a]pyrene in the GI tract and liver has been studied. Zheng et al.
(2002) determined the expression of UGT isozyme levels and activity toward benzo[a]pyrene
from human tissue samples of the liver, lung, and regions of the aerodigestive tract (tongue,
tonsil, floor of the mouth, larynx, esophagus). Glucuronidation of phenolic benzo[a]pyrene
metabolites was detected in all aerodigestive tract tissues examined and, of the isoforms tested,
activity was identified for UGT1A7, 1A8, and 1A10. No UGT expression or glucuronidation
activity was detected in lung tissue.
Bentsen-Farmen et al. (1999a) measured CYP1A1 induction and DNA adduct levels in
Wistar rats given i.p. doses of 3-MC and followed by a single dose of benzo[a]pyrene, or a single
benzo[a]pyrene dose without pretreatment. 3-MC pretreatment increased CYP1 Al activity in
the liver and DNA adduct levels in both liver and lung, with significant correlation to CYP1 Al
activity in lung but not in liver. The results indicate the difficulty in using DNA adducts as a
biomarker in short-term exposures. The authors further reported that the study results were
highly dependent on the analytical technique used.
Ramesh et al. (2001b) exposed F344 rats for up to 90 days to benzo[a]pyrene in the diet
at doses of 0, 5, 50, or 100 mg/kg-day. AHH activity (as measured by formation of the 3-OH-
benzo[a]pyrene) in the liver was increased in both males and females (approximately twofold
higher in females at the end of the study) in a dose- and duration-dependent manner.
Granberg et al. (2000) treated female NMRI mice with [ H]-benzo[a]pyrene by i.v.
injection and identified tissue-bound radiolabel in the lung, liver, and cardiovascular endothelial
cells. The levels of tissue-bound radioactivity were correlated to CYP1 Al, as determined by
7-ethoxyresorufin-O-deethylase (EROD) activity. Pretreatment with the AhR ligand P-NF
increased tissue binding in lung and heart cells.
3.3.3.3. Skin
Indirect evidence for metabolism of benzo[a]pyrene in skin is plentiful and includes
numerous studies of dermally exposed humans or animals with subsequent detection of
benzo[a]pyrene metabolites in tissues or excreta. These types of studies provide qualitative
evidence for dermal metabolism but are confounded by potential contribution of metabolism in
other tissues following systemic circulation. However, direct measurements of benzo[a]pyrene
metabolism have been made in in vitro models using human skin or skin cells.
48
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Hall and Grover (1988) also showed that human skin samples can metabolize
benzo[a]pyrene. In skin samples from 11 subjects, the majority of benzo[a]pyrene-7,8-diol
formed from benzo[a]pyrene was recovered as the (-) enantiomer. However, the stereospecific
formation of tetrol metabolites was highly variable, suggesting significant inter-individual
variability in the relative formation of the DNA reactive epoxide metabolites from
benzo[a]pyrene. Bowman et al. (1997) assessed levels of the 7,8,9,10-tetrol metabolite in the
urine of 43 psoriasis patients treated with coal tar medication. Urinary benzo[a]pyrene
metabolites were detected in 40% of patients vs. 10% of matched controls. Amounts of
metabolites detected were highly variable, but precise measures of applied dose were not
available. These studies show that human skin can metabolize benzo[a]pyrene to diverse
products, the metabolism of benzo[a]pyrene is inducible and variable with regard to
stereospecific formation of downstream metabolites, and there may be significant interindividual
variation.
Merk et al. (1987) used hair follicles from volunteers to study benzo[a]pyrene
metabolism and the effect of coal-tar-containing shampoos. Hair follicles were able to form
numerous metabolites of benzo[a]pyrene. The results with coal-tar-exposed individuals showed
that AHH activity was inducible in these cells and inhibition of CYP activity decreased the
metabolism of benzo[a]pyrene and the formation of DNA binding activity. Alexandrov et al.
(1990) demonstrated that hair follicles from human subjects (10 healthy female smokers and 10
healthy female nonsmokers) were able to generate benzo[a]pyrene tetrols from diol epoxide
precursors. There was stereospecific metabolism with most of the tetrol formed from (-)-
benzo[a]pyrene-7,8-diol consistent with extensive formation of (±)-anti-BPDE rather than the
(±)-syn-BPDE metabolite. Agarwal et al. (1991) reported that human melanocytes treated in
culture formed diverse benzo[a]pyrene metabolites, including dihydrodiols, hydroxyl
compounds, quinones, and their glucuronide and sulfate conjugates.
Kao et al. (1985) investigated the metabolism of benzo[a]pyrene in skin from several
species, including humans. In the case of human skin, TCDD induction could not be studied, but
the authors investigated the influence of metabolic viability of the skin on metabolism. Skin
2 14
samples were treated with 2 [j,g/cm [ C]-benzo[a]pyrene in acetone and were incubated for
24 hours. Medium under the skin was then extracted with ethyl acetate but was not subjected to
hydrolysis to identify conjugate formation. Fifty-two percent of the radioactivity in culture
medium below viable human skin was composed of water-soluble metabolites, 8% were lipid-
soluble polar metabolites, 17% were diols, 1.2% were monophenols, 2.5% were quinones, and
18%) were parent compound. By contrast, previously frozen, nonviable skin allowed mostly
parent compound to pass (50% of radioactivity in the medium) and considerably smaller portions
of water-soluble and polar metabolites and diols to pass, but, surprisingly, relatively increased
portions of monophenols and quinines to pass. Results for skin from other species—marmoset,
rabbit, rat, mouse—were similar: water-soluble metabolites varied between 55 and 77% of the
49
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
radioactivity in the medium; polar metabolites, 5.8-13.2%; diols, 7.9-15.2%; monophenols, 0.7-
1.8%; quinones, 0.6-2.1%; and parent compound, 0% (mouse), to 6.7% (rabbit). Metabolites
found below nonviable skin from animal species were not reported.
Animal studies are consistent with the metabolism data in human skin. Ng et al. (1992)
used an in vitro culture system with skin from female hairless guinea pigs and concluded that the
degree of penetration of benzo[a]pyrene was dependent on metabolism, since the collection of
the administered radioactivity was much greater for viable than nonviable skin. At the lowest
dose tested (32.1 nmol/cm ), 37% of the applied radioactivity was recovered in the receptor fluid
within 24 hours, 84% of which were metabolites, including OH derivatives, dihydrodiol diones,
and conjugates. The tetrol metabolite of the DNA-reactive epoxide accounted for 2.56% of the
"3
administered dose. MacNicoll et al. (1980) assessed metabolism of [ H]-benzo[a]pyrene in
Swiss mouse skin maintained in short-term organ culture. After incubation for 24 hours,
radioactivity derived from benzo[a]pyrene consisted of 3 benzo[a]pyrene-equivalents of lipid-
soluble metabolites, 147 benzo[a]pyrene-equivalents water-soluble metabolites, and 5
benzo[a]pyrene-equivalents bound to skin. The high proportion of water-soluble metabolites
indicates that benzo[a]pyrene was readily metabolized.
Kao et al. (1984) evaluated metabolism of benzo[a]pyrene as a factor in the dermal
penetration of benzo[a]pyrene in mouse skin. Skin samples formed predominantly polar
metabolites and diols (approximately 20% of the radioactivity each), with small shares
(approximately 1% of the penetrated material) of conjugates, monophenols, and quinones.
Parent compound was recovered at about 1.5%. TCDD induction changed the portion of
conjugates to almost 8%, at the expense of polar metabolites, diols, and parent compound, but
did not affect monophenol or quinone formation. Investigation of the radioactivity left in the
skin revealed that it was mostly parent compound (almost 50% of the dose in uninduced and a
little more than 30% in induced skin), with about 12-16% water-soluble metabolites and polar
metabolites, diols, monophenols, and quinones decreasing in that order from -5% of the dose to
<2%.
3.3.3.4. Reproductive Tissues and Fetal Metabolism
Several toxicity studies have demonstrated the ability of benzo[a]pyrene to impair
reproductive function in male and female rodents as well as induce developmental toxicity.
Therefore, the ability of benzo[a]pyrene to be metabolized in tissues that affect reproduction or
development is of interest. Williams et al. (2000) reported the presence of CYP1 Al, CYP1A2,
and CYP1B1 transcripts in prostate tissue from human donors exposed in short-term organ
culture and the activation of benzo[a]pyrene as indicated by DNA adduct formation. Primary
human prostate cells have also been demonstrated to metabolize benzo[a]pyrene in vitro (Martin
et al., 2002).
50
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Bao et al. (2002) exposed a human endometrium epithelial cell line, which expresses low
constitutional levels of EROD activity (used as a measure of CYP1 activity), to benzo[a]pyrene
(1 mM) in vitro and observed a 12-fold induction of CYP1A1. Specific inhibitors of CYP1B1
and CYP1A2 had no effect on EROD activity, suggesting little metabolic contribution from these
two isoforms in endometrial tissues, while the CYP1 Al inhibitor a-NF (100 nM) inhibited
EROD activity more than 60%. Melikian et al. (1999) measured DNA adducts in cervical
epithelial and stromal tissues in smokers and nonsmokers. Increased levels of adducts in the
tissues of smokers suggested that delivery of benzo[a]pyrene to the cervical tissues with
subsequent metabolism may occur, although transport of reactive metabolites generated in other
tissues was not specifically ruled out.
Ramesh et al. (2003, available only as abstract) investigated benzo[a]pyrene metabolism
-3
in multiple organ systems in rats exposed by nose-only inhalation to 75 (J,g/m benzo[a]pyrene
adsorbed on CB 4 hours/day for 60 days. AHH activity and benzo[a]pyrene metabolism were
reported for various tissues. A diverse spectrum of metabolites was identified, including
dihydrodiols and 3- and 9-monophenols. The concentrations of BPDE were highest in testes,
providing mechanistic support for previously observed effects of benzo[a]pyrene on male
reproductive parameters in animal toxicity studies.
Exposure to reactive benzo[a]pyrene metabolites may be a concern both in utero and
during lactation. Wu et al. (2003) measured the generation of benzo[a]pyrene metabolites in F1
generation pups and determined the mRNA development profile for the AhR in the absence and
presence of subacute exposure concentrations of benzo[a]pyrene in preweaning rats. Pregnant
"3
F344 rats were exposed to 25, 75, or 100 (J,g/m benzo[a]pyrene aerosols via nose-only
inhalation, 4 hours/day for 10 days (GDs 11-21). Benzo[a]pyrene metabolites and mRNA and
protein expression profiles of AhR and CYP1 Al were analyzed in the cerebral cortex,
hippocampus, liver, and plasma. Plasma and cerebral cortex benzo[a]pyrene levels on postnatal
day (PND) 0 changed with the dose. Benzo[a]pyrene decreased steadily with time in these
tissues, always reflecting the administered dose, reaching nondetectable levels by PND 30. In
plasma, diols represented approximately 60% of the total metabolites over the period of
20 PNDs. In the cerebral cortex, diols represented close to 80% of metabolites soon after
parturition, decreasing to approximately 40% by PND 20. Benzo[a]pyrene-7,8-diol represented
a maximum of approximately 25% of the recovered metabolites in plasma and approximately
30%) of recovered metabolites in the cerebral cortex.
There was a statistically significant (p < 0.05) up-regulation of AhR mRNA, a subsequent
induction of CYP1 Al mRNA, and a significant increase in CYP1 Al protein levels in pup livers
"3
at 100 (j,g/m compared to unexposed controls (data were presented only for PND 60 and the
"3
100 (J,g/m concentration). The AhR mRNA expression profile in the developing cerebral cortex
and hippocampus indicated up-regulation of AhR during the first 3 postnatal weeks at all
concentrations, although these differences were not statistically significant due to large
51
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
individual variation. At the high concentration, AhR mRNA abundance was more than twice
that of controls by PND 30. However, this up-regulation of AhR mRNA was not accompanied
by a concomitant up-regulation of CYP1A1 mRNA in the CNS. In fact, expression of CYP1A1
in these tissues was very low or absent during the 1-month postnatal period after exposure to
-3
100 (j,g/m of benzo[a]pyrene (the only dose tested). Based on these findings, the authors
suggested that the up-regulation of the AhR may increase the potential for benzo[a]pyrene
neurotoxicity via the activation of CYP450 in liver and the subsequent deposition of toxic
metabolites in the developing CNS. The results do not suggest that benzo[a]pyrene metabolism
by CYP1A1 is induced in the developing CNS.
Pregnant Swiss Webster mice were administered radiolabeled benzo[a]pyrene by i.v.
injection, and maternal and fetal levels of benzo[a]pyrene and radiolabel were determined. Fetal
tissue levels of radiolabel increased while maternal tissues decreased. The level of
benzo[a]pyrene decreased in fetal tissue during this period, suggesting the accumulation of
benzo[a]pyrene metabolites. Increasing the capacity of maternal plasma to bind benzo[a]pyrene
by administering a benzo[a]pyrene antiserum decreased fetal accumulation of radiolabel,
suggesting that bioavailability had decreased (McCabe and Flynn, 1990). Other investigators
have also demonstrated the ability of benzo[a]pyrene to be transported to the fetus (Neubert and
Tapken, 1988; Shendrikova and Aleksandrov, 1974). Martin et al. (2000) exposed exfoliated
cells from breast milk of human donors to benzo[a]pyrene in vitro. Treated cells showed
increased DNA single strand breaks (SSB) relative to controls, indicating that these cells can
activate benzo[a]pyrene. Taken together, the results from these various experiments suggest that
placental and lactational transfer of benzo[a]pyrene and active metabolites may be of concern.
The developmental expression patterns of metabolizing enzymes can be an important
determinant of childhood susceptibility. Numerous in vivo mechanistic tumor screening assays
have been conducted in newborn mice (see Section 4.4.1 on mechanistic cancer studies), and the
observed increases in tumors provide evidence for the ability of young animals to metabolize
"3
benzo[a]pyrene to DNA reactive metabolites. Melikian et al. (1989) administered [ H]-
"3
benzo[a]pyrene or [ HJ-BPDE via i.p. injection to CD-I mice on PND 1, 8, or 15 and formation
of benzo[a]pyrene metabolites was determined in lung and liver over time periods ranging from
2 minutes to 24 hours after the last dose. In the lung, metabolites included diones, quinones, and
phenols, and metabolite levels in the lung were higher on day 1 than on days 8 or 15. In liver, a
different spectrum of metabolites was observed, dominated by unidentified polar metabolites.
The percentage of radiolabel as metabolites was increased on days 8 and 15, perhaps reflecting
greater inducibility of benzo[a]pyrene metabolism in liver versus the lung. Formation of
glucuronide or sulfate conjugates was greater than glutathione conjugates in both tissues. The
total amount of benzo[a]pyrene metabolized by Phase II enzymes was greater in liver than in
lung on day 1 but was similar on days 8 and 15. Age-dependent expression of genes that encode
52
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
metabolizing enzymes has been the subject of reviews (e.g., Cresteil, 1998) that are addressed in
Section 4.4.
3.3.3.5. Other Tissues
"3
Moore et al. (1982) exposed organ cultures of human and rat bladder tissues to [ H]-
benzo[a]pyrene for 24 hours and evaluated metabolite profiles. Benzo[a]pyrene was metabolized
by bladder tissues of both species. Total mean amount of metabolites formed was higher in
human than rat bladders (twofold). A similar spectrum of diverse metabolites was generated,
although relative proportions varied. Formation of benzo[a]pyrene-7,8-diol was similar in both
species and similar levels of DNA binding was observed. These results indicate that bladder
tissues have metabolic capability for benzo[a]pyrene and that the metabolic capacity is similar
between humans and rats.
Several other target tissues for benzo[a]pyrene toxicity have been found to have the
capacity to metabolize benzo[a]pyrene. For example, Moorthy et al. (2003) treated mouse aortic
smooth muscle cells with benzo[a]pyrene, 3-OH-benzo[a]pyrene, or benzo[a]pyrene-3,6-quinone
in vitro. Several DNA adducts were identified that were attributed to the 3-OH and BPQ
metabolites. Benzo[a]pyrene treatment increased CYP1B1 but not CYP1A1 in these cells. The
authors suggest that CYP1B1 activates benzo[a]pyrene to 3-OH and BPQ metabolites, which
induce the DNA damage responsible for changes that are important precursors for
atherosclerosis.
3.4. ELIMINATION
3.4.1. Inhalation Exposure
Human studies of benzo[a]pyrene exposure have generally been limited to individuals
exposed to coke oven emissions, coal tars, or other products containing a mixture of PAHs and
are of limited value in assessing urinary benzo[a]pyrene exposure biomarkers. Numerous
studies have evaluated 1-OH pyrene in urine as a general marker for PAH exposure, but, because
1-OH pyrene is not a metabolite of benzo[a]pyrene, these data are not directly useful in
evaluating the toxicokinetics of benzo[a]pyrene. Some studies have unsuccessfully attempted to
quantify exposure to benzo[a]pyrene via measurement of parent compound or its metabolites.
Bentsen-Farmen et al. (1999b) compared air concentrations of PAHs to PAH metabolites in the
urine of 17 electrode paste plant workers and detected 1-OH-Py but no benzo[a]pyrene
metabolites despite the fact that benzo[a]pyrene in personal air samples showed mean exposure
"3
levels of 0.3 (j,g/m . Waidyanatha et al. (2003) also attempted to measure PAH exposure via
urinary metabolite levels of coke oven workers; several PAHs but no benzo[a]pyrene were
detected. In other cases, exposure and urinary metabolites were positively correlated (Hecht et
al., 2003; Wu et al., 2002; Giindel et al., 2000; Grimmer et al., 1993; Becher and Bj0rseth, 1983).
53
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Wu et al. (2002) reported a correlation between benzo[a]pyrene-tetrols and total PAH exposure
in coke oven workers.
Several studies evaluated the elimination of benzo[a]pyrene in animals following
exposure via the respiratory tract. Petridou-Fischer et al. (1988) applied 10 j_iL aliquots of [14C]-
benzo[a]pyrene in a gelatin:saline solution over a 2-hour period to the ethmoid and maxillary
nasal turbinates of monkeys and dogs to assess differences in benzo[a]pyrene disposition in
portions of the nose. Urine levels reached a maximum of 0.69% of the administered dose in
dogs and 0.07% in monkeys, while fecal levels reached a maximum of 6.42% of the dose in dogs
and 1.17%) in monkeys over 48 hours. The pattern of metabolites in the excreta was consistent
with results from other respiratory tract deposition studies, showing that benzo[a]pyrene is
excreted preferentially in the feces. Wolff et al. (1989) investigated the effects of nose-only
exposure of male and female F344 rats to unlabeled benzo[a]pyrene for 4 weeks, followed by a
single exposure to [14C]-benzo[a]pyrene. Four weeks after treatment with the radiolabel, the rats
had eliminated approximately 96%> of the total dose in the feces. The mean half-life was
calculated as 22 hours for feces and 28 hours for urine. Sun et al. (1982) exposed male and
female F344 rats via nose-only inhalation for 30 minutes to atmospheres containing 0.6 [j,g/L
3 3
[ H]-benzo[a]pyrene absorbed onto 67Ga203 or to 1.0 [j,g/L neat [ H]-benzo[a]pyrene and
measured levels of radioactivity in tissues and excreta. With either exposure, benzo[a]pyrene
excretion declined rapidly over the first 3 days after exposure, with little additional excretion
occurring by 5 days postexposure. Excretion in feces was much greater than in urine for either
exposure regimen. Following exposure of rats to neat benzo[a]pyrene, feces accounted for 86%>
of total excreted radioactivity and urine for 14%>. Metabolite profiles were not evaluated. Wang
et al. (2003) exposed B6C3Fi mice for 10 days to asphalt fumes in an inhalation chamber at a
-3
concentration of approximately 180 mg/m ; the benzo[a]pyrene content of the fumes was not
reported. Benzo[a]pyrene metabolites measured in the urine of exposed mice were in the ng/100
mL range and were identified at the following mass ratios (with benzo[a]pyrene arbitrarily set as
1): 7,8,9,10-benzo[a]pyrene-tetrol, 14; benzo[a]pyrene-7,8-diol epoxides, 17; benzo[a]pyrene-
7,8-diol, 3.6; 3-OH-benzo[a]pyrene, 10.
"3
In a study investigating the disposition of benzo[a]pyrene, 1 (J,g/kg of the [ H]-labeled
compound dissolved in triethylene glycol was instilled into the trachea of male Sprague-Dawley
rats (Weyand and Bevan, 1986). Approximately 60%> of the administered dose was associated
with the intestine and intestinal contents, and 2.2% of the radiolabel was recovered in the urine
6 hours after dosing. The relative contribution of water- and lipid-soluble metabolites in the
intestinal contents was 18.2 and 26.5% of the administered dose, respectively. The amounts of
metabolites determined in the organic phase 6 hours after the end of exposure were as follows
(percent of total in the organic phase): conjugates or polyhydroxylated compounds (24.1%),
9,10-diol (7.75%), 4,5-diol (8.43%), 7,8-diol (5.73%), 1,6-quinone (7.13%), 3,6-quinone
(7.85%), 6,12-quinone (7.10%), 9-OH (2.24%), 3-OH (4.66%), benzo[a]pyrene (1.55%). The
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
total amounts of metabolites in feces and urine were not reported in this study, and, due to
enterohepatic circulation, the intestinal content concentrations may not reflect the final amounts
of each metabolite in the excreta. Bevan and Sadler (1992) administered a single intratracheal
instillation of benzo[a]pyrene (2 (J,g/kg) to male Sprague-Dawley rats and assessed metabolite
profiles in bile after 6 hours. Relative metabolite levels were 31.2% diglucuronides, 30.4%
thioether conjugates, 17.8% monoglucuronides, 6.2% sulfate conjugates, and 14.4%
unconjugated metabolites.
Weyand and Bevan (1987) studied the differences in benzo[a]pyrene elimination among
rats, hamsters, and guinea pigs. Male Sprague-Dawley and Gunn rats (the latter a strain
genetically deficient in bilirubin glucuronidation), male Syrian golden hamsters, and male
"3
Dunkin-Hartley guinea pigs were given single intratracheal doses of 0.16 jag or 350 jag [ H]-
benzo[a]pyrene per animal. Tissue levels and the amount of radiolabel excreted into the urine
and bile 6 hours after treatment are given in Table 3-3. Urinary excretion of benzo[a]pyrene
metabolites amounted only to a small fraction of biliary excretion in all tested species. The
results suggested that the metabolic capacity for benzo[a]pyrene became saturated in guinea pigs
and Sprague-Dawley and Gunn rats but not in hamsters at the 350 [j,g/animal dose (note: rats
weighed 200-250 grams, average high dose 1.56 mg/kg; hamsters 100-140 grams, 2.92 mg/kg;
and guinea pigs 600-850 grams, 0.48 mg/kg), indicating that hamsters, who received about 6
times the dose of guinea pigs, command a metabolic system that handles benzo[a]pyrene very
well. Moir et al. (1998) administered benzo[a]pyrene i.v. to Wistar rats and recovered, at 8 hours
after a 2 mg/kg dose, 4.27% of the dose in urine but only 0.06% in feces, confirming the fact of
enterohepatic circulation of benzo[a]pyrene metabolite conjugates (Hirom et al., 1983; Weyand
and Bevan, 1986).
Table 3-3. Excretion of benzo[a]pyrene metabolites in several animal species
Dose
Route of
excretion
Gunn rat"
Sprague-Dawley
rat"
Syrian golden
hamster"
Dunkin-Hartley
guinea piga
0.16 (xg/animal
Urine
0.98 ±0.16
2.21 ± 1.1
3.74± 1.2
2.22 ± 1.2
Bile
59.4 ± 1.07
70.3 ±2.0
54.6 ±3.6
71.7 ±1.9
350 (xg/animal
Urine
1.44 ± 1.82
1.68 ±0.6
2.53 ±0.4
1.27 ±0.4
Bile
30.8 ± 1.51
55.0 ±2.0
52.9 ±2.7
47.9 ±4.9
aValues are percent of the applied dose at 6 hours after intratracheal instillation.
Source: Weyand and Bevan (1987).
The pattern of metabolites in bile was also reported; results are compiled in Table 3-4.
At each dose thioether conjugates predominated but to a varying extent in each species. Guinea
pigs evidently used preferentially thioether conjugation for biliary benzo[a]pyrene elimination,
with the other conjugates (or nonconjugated metabolites) making up a small portion of biliary
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
excretion. At the high dose a shift in the relative proportion of metabolites was observed, with
thioethers decreasing, glucuronides and sulfates increasing, and total nonconjugated remaining
unchanged in both rat strains and hamster. The data also indicate that glucuronidation and
sulfation became saturated at the high dose only in guinea pigs, while thioether formation
capacity became overwhelmed by the high dose in all species. Glucuronide levels in the bile of
Gunn rats were about one-half those of Sprague-Dawley rats; this strain also excreted fewer
nonconjugated metabolites and appeared to have a generally lower metabolic capacity for
benzo[a]pyrene.
Table 3-4. Biliary excretion of benzo[a]pyrene metabolites in several
species
Metabolite group
Dose
(Hg/animal)
Gunn rat"
Sprague-
Dawley rat"
Syrian golden
hamster"
Dunkin-Hartley
guinea piga
Nonconjugated
0.16
4
11.5
4.1
1.7
350
2
8.3
4.2
1.3
Glucuronides
0.16
7
12.9
11.9
4.5
350
9
18.7
18.1
0.2
Sulfate conjugates
0.16
8
3
3.7
1.9
350
8
5.9
6.9
1.1
Thioether conjugates
0.16
40
42.9
33.8
54.1
350
12
22.1
23.7
45.4
aValues are percent of the applied dose at 6 hours after intratracheal instillation.
Source: Weyand and Bevan (1987).
"3
Following intratracheal administration of 1 (J,g/kg body weight [ H]-benzo[a]pyrene to
male Sprague-Dawley rats in hydrophilic triethylene glycol, 70.5% of the administered
benzo[a]pyrene was excreted into bile within 6 hours (Bevan and Ulman, 1991). In contrast,
benzo[a]pyrene excretion was 58.4 and 56.2% in the same time period when the lipophilic
solvents ethyl laurate and tricaprylin, respectively, were the vehicles. Benzo[a]pyrene (in
tri ethylene glycol) excretion in bile was described as biphasic with half-lives of 31 and 100
minutes, respectively. Individual metabolite concentrations were not monitored.
3.4.2. Oral Exposure
Only limited data were identified on the elimination of benzo[a]pyrene following
exposure by the oral route. Although the dietary/oral route is likely to be the predominant route
of exposure for the general population not occupationally exposed, it has received very little
attention (Stavric and Klassen, 1994). In the only study investigating this issue in humans, the
concentration of benzo[a]pyrene was below detection limits (<0.1 (j,g/person) in the feces of
56
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
eight volunteers who had ingested broiled meat containing approximately 8.6 jag of
benzo[a]pyrene (Hecht et al., 1979).
Ramesh et al. (2001b) evaluated benzo[a]pyrene disposition in F344 rats dosed via
gavage with 100 mg/kg benzo[a]pyrene dissolved in peanut oil. Recovery of unmetabolized
benzo[a]pyrene from feces reached a maximum of 35% of the dose at 24 hours but was very low
at 48 hours and thereafter. Lipid-soluble metabolites in feces reached approximately 28% of all
metabolites by 8 hours after dosing then declined sharply. Lipid-soluble metabolites in urine
reached 35—40% of total by 48 hours then declined to approximately 10% by 72 hours. Diol
metabolites were most numerous in feces, while phenols predominated in urine.
Hecht et al. (1979) conducted a study of fecal excretion of benzo[a]pyrene and its
metabolites in rats. In male F344 rats administered [14C]-benzo[a]pyrene via gavage (0.04, 0.4,
or 4 (j,mol/animal), approximately 85% of radiolabel was recovered in feces, and 1-3% in urine
after 168 hours. The portion of benzo[a]pyrene recovered from feces as parent compound
ranged from 13 to 6% of the administered dose within 48 hours. In rats fed charcoal-broiled
hamburger containing 52.7 jag benzo[a]pyrene/kg meat, 11% of the benzo[a]pyrene was excreted
unchanged in feces.
3.4.3. Dermal Exposure and Other Exposure Routes
Bowman et al. (1997) detected benzo[a]pyrene-tetrols in 40% of the urine samples from
psoriasis patients treated with coal tar medication, as compared to only 10% of those of controls.
No specific measures of applied dose were available.
In a dermal absorption study, Yang et al. (1989) evaluated the recovery of [ H]-
benzo[a]pyrene, 100 ppm in crude oil applied topically, in urine and feces of female Sprague-
Dawley rats. Total recovery of applied radioactivity over 96 hours was 5.3% in urine and 27.5%
in feces. Individual metabolite concentrations were not measured. Ng et al. (1992) examined the
percutaneous absorption of radiolabeled benzo[a]pyrene in the hairless guinea pig. Following a
single application of 28 jag benzo[a]pyrene (dissolved in 50 [xL acetone applied to 4 cm of
dorsal skin), approximately 34% of the administered radiolabel was eliminated within 24 hours.
Most excretion had occurred by day 3 and continued slowly to reach 73% by day 7. Relative
amounts of benzo[a]pyrene or metabolites in urine versus feces were not reported.
Moir et al. (1998) dosed male Wistar rats i.v. with 2, 6, or 15 mg/kg of [14C]-
benzo[a]pyrene and examined excretion in urine and feces over 32 hours. At 8 hours after
injection, urinary excretion was 4.3, 2.6, and 3.2% of the dose, while fecal excretion was only
0.06, 5.6, and 0.43% of the 2, 6, and 15 mg/kg doses, an indication of enterohepatic circulation.
The amount of radioactivity excreted in the urine after 32 hours was similar in each dose group
(6-7%) of the administered dose). However, the proportion of the administered dose in the feces
was dose-dependent. At the low dose (2 mg/kg) fecal excretion accounted for 26% of the dose,
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
while at the mid- and high-doses, fecal excretion accounted for 56 and 50% of the dose,
respectively. No measurement of specific metabolites or whole body clearance was conducted.
Following i.v. administration of 3 (iinol [14C]-benzo[a]pyrene to male New Zealand white
rabbits, approximately 30% of the dose was recovered in the bile and 12% in the urine within
6 hours after treatment (Chipman et al., 1982). Excretion in the bile was biphasic, with estimated
half-lives of 0.27 and 4.62 hours. Further analysis of metabolite profiles in bile and urine were
conducted. Treatment of bile or urine with P-glucuronidase or aryl sulfatase increased the
amount of radioactivity recovered in ethyl acetate extracts, suggesting that sulfate and
glucuronide conjugation are important contributors to benzo[a]pyrene metabolites in excreta.
Analysis of these extracts revealed the primary metabolite as 9,10-diol, with lower amounts of
numerous other metabolites identified (diols, quinones, and monophenols).
Likhachev et al. (1992) measured the excretion of benzo[a]pyrene-7,8-diol and 3-OH-
benzo[a]pyrene in L10 rats given a single i.p. dose of 200 mg/kg benzo[a]pyrene. Urine and
feces were collected over a 15-day period and then again at 30 days after exposure for another 5
days. Both metabolites were excreted in feces at two to three times the amount excreted in urine.
Metabolites in urine and feces decreased steadily over the 15-day postexposure period, but no
more metabolites were detected at 30 days. A comparative study of metabolism in male Macaca
fascicularis monkeys (five animals) and male L10 rats (four animals) was also performed. Both
species were given a single i.p. dose of 100 mg/kg benzo[a]pyrene, and levels of benzo[a]pyrene
metabolites in feces were evaluated over 8 days. Total benzo[a]pyrene metabolites were
significantly lower in monkeys than in rats, but the results were hampered by infrequent feces
collection in monkeys. Monkeys had an approximately fourfold higher benzo[a]pyrene-7,8-diol
to 3-OH-benzo[a]pyrene ratio in feces than rats. Together these data suggest that monkeys have
lower excretion of benzo[a]pyrene (and perhaps lower metabolism), and formation of the
proximate carcinogenic metabolite is higher than in rats.
In another part of this study (Likhachev et al., 1992), benzo[a]pyrene metabolism and
excretion were assessed following multiple exposures. L10 rats were given i.p. injections of
1 mg/kg benzo[a]pyrene every 11th day for 10 treatments. Benzo[a]pyrene metabolites were
measured in excreta for 8 days after 1, 5, and 10 treatments. After the single dose, excretion in
feces was favored over urine, and benzo[a]pyrene-7,8-diol was the predominant metabolite. The
rate of metabolite excretion was decreased considerably after 5 and 10 doses, respectively. The
authors attributed this finding to age-related changes in metabolic capacity or "exhaustion of
enzymes." They also compared metabolite profiles for individual rats with their tumor responses
and based on the results suggested a link between tumor latency and elimination of
benzo[a]pyrene metabolites (although small sample size and variability limited the power of the
analysis). A direct correlation between benzo[a]pyrene-7,8-diol excreted in urine and tumor
latency was observed. It was postulated that higher excretion of reactive benzo[a]pyrene
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
36
37
38
metabolites caused fewer of these metabolites to bind to DNA targets and thus increased tumor
latency.
In addition to benzo[a]pyrene and its metabolites, adducts of benzo[a]pyrene with
nucleotides have also been identified in feces and urine of animals but only as a small fraction of
the administered dose. Autrup and Seremet (1986) administered i.p. doses of 0, 10, 50, or
100 (J,g/kg tritiated benzo[a]pyrene to male Wistar rats and collected urine over 24-hour periods
for 72 hours. The level of BPDE adducts with guanine detected in urine was dose-dependent.
The authors reported that, at the high dose, 0.15% of the administered benzo[a]pyrene dose was
excreted in the urine as this adduct within 48 hours. Rogan et al. (1990, only study abstract was
available for review) reported that 0.02% of a benzo[a]pyrene dose was excreted as an adduct
with guanine in urine and feces over 5 days.
Overall, the data on benzo[a]pyrene elimination in humans are too limited to estimate
quantitative rates of elimination. The situation is further complicated by the existence of
multiple pathways in which many of the key enzymes exhibit polymorphisms. In the context of
biomonitoring studies, benzo[a]pyrene metabolites have been detected in the urine of exposed
humans, but the fecal excretion has not been investigated in any detail. The animal data are
consistent in showing that feces are the primary route of elimination of benzo[a]pyrene, while
urinary excretion plays a lesser role. A diverse array of metabolites, as well as parent
benzo[a]pyrene, is found in the excreta. Enterohepatic circulation of benzo[a]pyrene metabolite
conjugates has been demonstrated in animals and may exist in humans as well, but its impact on
the toxicokinetics of benzo[a]pyrene is not understood.
Similar considerations apply to the toxicokinetics of benzo[a]pyrene as a whole. A few
animal studies that have attempted comprehensive assessments of benzo[a]pyrene metabolism.
Although a lot is known from ex vivo studies about which humans tissue can metabolize
benzo[a]pyrene to what extent, comprehensive studies to estimate overall metabolic capacities or
the balance between Phase I and Phase II metabolism in humans have not been conducted.
There are also no comprehensive data on the tissue distribution of benzo[a]pyrene in humans, not
counting a number of studies that have reported post mortem levels in a few tissues in the
absence of any exposure assessments. Thus, despite the fact the humans are universally exposed
to benzo[a]pyrene, there is a great need for more knowledge concerning its toxicokinetics.
3.5. PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODELS
Several toxicokinetic or pharmacokinetic models of benzo[a]pyrene have been developed
for rodents (rat and hamster), but none has been calibrated in humans. Bevan and Weyand
(1988) performed compartmental pharmacokinetic analysis of distribution of radioactivity in
male Sprague-Dawley rats, following the intratracheal instillation of benzo[a]pyrene to normal
and bile duct-cannulated animals (Weyand and Bevan, 1987, 1986). The authors used the
Simulation, Analysis and Modeling (SAAM) and Conversational SAAM (CONSAAM)
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
computer programs to model the disposition of labeled benzo[a]pyrene and its metabolites. A
good fit to the experimental data was achieved with 10 assumed compartments corresponding to
lung, blood, liver, other tissues (reflecting the sum of radioactivity in kidney, stomach, testis,
spleen, heart, and thymus), carcass (reflecting the sum of radioactivity in skin, fat, bones,
muscle, and blood in those organs), and several additional hypothetical compartments that linked
the central compartment with intestines (including their contents) and urine. Enterohepatic
circulation and intestinal secretion were both included in the model. The model allowed
calculation of linear rate constants for moving radioactivity among compartments (in minute"1),
representing the probability per unit time that radioactivity from one compartment would be
transferred to another. The model adequately simulated disposition of benzo[a]pyrene and its
metabolites, measured as total radioactivity in blood, organs and excreta, under the assumption
that the kinetics of benzo[a]pyrene and its metabolites are the same. The authors emphasized
differences between their approach in building the implicit SAAM model and the approach used
to build explicit PBPK models. Roth and Vinegar (1990) reviewed the capacity of the lung to
impact the disposition of chemicals and used benzo[a]pyrene as a case study. A PBPK model
was presented based on data from Wiersma and Roth (1983a, b) and was evaluated against tissue
concentration data from Schlede et al. (1970). The model was structured with compartments for
arterial blood, venous blood, lung, liver, fat, and slowly as well as rapidly perfused tissues.
Metabolism in liver and lung was estimated using kinetic data from control rats and rats
pretreated with 3-MC to induce benzo[a]pyrene metabolism. Benzo[a]pyrene binding was
accounted for in blood, liver, and lung, but only one binding component was used, which, as the
authors suggested, resulted in a rather poor fit for liver. The model was built based on blood and
tissue concentrations measured over 5 hours in rats given a dose of 117 nmol/kg by i.v. injection
into the arterial circulation or the venous blood supply of the liver. The model was tested against
the data of Schlede et al. (1970). The number of data points was limited (six time points for the
venous circulation, liver, and fat and only a single time point for the lung). The model predicted
the data of Schlede et al. (1970) reasonably well, although deviations were apparent for several
compartments. Most notably, the model overpredicted concentrations in the induced lung,
uninduced and induced liver, and induced fat compartments. The results of PBPK simulations
showed that induction of metabolizing enzymes increased the amount of benzo[a]pyrene cleared
by the lungs relative to the liver.
An interesting result of the Roth and Vinegar (1990) modeling study was that even
though the metabolic clearance of benzo[a]pyrene in the lungs was low in comparison to the
liver, under simulated pathological conditions of reduced hepatic blood flow, a substantial
metabolic clearance was carried on by well perfused lungs. This illustrates that changes in tissue
perfusion can shift the organ pattern of benzo[a]pyrene metabolic clearance in vivo, and
consequently, might also shift the organ pattern of benzo[a]pyrene-induced disease, such as
cancer. The authors emphasized that their PBPK model with the results of simulations should
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
not be taken as a definite model for benzo[a]pyrene but that it could be used for designing further
studies of benzo[a]pyrene metabolic clearance. They also suggested that it would be appropriate
to incorporate into the PBPK model a description of the appearance of various metabolites of
benzo[a]pyrene.
Moir et al. (1998) conducted a pharmacokinetic study on benzo[a]pyrene to obtain data
for model development. Rats were injected with [14C]-benzo[a]pyrene at doses of 2, 6, or
15 mg/kg and blood, liver, fat, and richly perfused tissue were sampled at 15 time points from 5
minutes to 32 hours after dosing. Moir (1999) then described a model for lung, liver, fat, richly
and slowly perfused tissues, and venous blood, with saturable metabolism occurring in the liver.
The fat and richly perfused tissues were modeled as diffusion-limited, while the other tissues
were flow-limited. The model was developed using the 15 mg/kg dose group data from the Moir
et al. (1998) study and validated using the 2 and 6 mg/kg data. The model predicted the blood
benzo[a]pyrene concentrations well, although it overestimated the 6 mg/kg results at longer
times (>100 minutes). The fat and richly perfused tissue benzo[a]pyrene levels were reproduced
by the model fairly well at the two highest dose levels but underestimated at the 2 mg/kg dose.
The model produced a poor fit to the liver data, underestimating benzo[a]pyrene tissue levels at
the two lowest doses. At the high dose the model overestimated the liver benzo[a]pyrene
concentration at times <200 minutes but afterwards underestimated them increasingly more. It
appeared that the model could not accommodate a slow elimination phase at times beyond 200
minutes for the mid and high dose levels. The model simulations were also compared to data of
Schlede et al. (1970) who had injected rats with 0.056 mg/kg body weight of benzo[a]pyrene.
Again the model predicted blood and fat benzo[a]pyrene concentrations quite well but
underestimated liver benzo[a]pyrene concentrations.
Moir (1999) suggested that the poor prediction of fat benzo[a]pyrene concentrations after
injection of 2 mg/kg benzo[a]pyrene in their own study (Moir et al., 1998) was due to analytical
error, causing the model to fail. This explanation is credible because the measured fat
benzo[a]pyrene concentrations following injection of 2 mg/kg benzo[a]pyrene were
indistinguishable from those obtained following injection of 6 mg/kg benzo[a]pyrene (Moir et
al., 1998). The author also speculated that the poor fit to liver benzo[a]pyrene concentration data
was due to binding of benzo[a]pyrene to receptors in the liver. The model predicted blood and
fat benzo[a]pyrene concentrations fairly accurately over a wide range of doses (0.056-
15 mg/kg), but it was limited by the failure to accurately predict liver benzo[a]pyrene
concentrations. The model included only one saturable metabolic pathway, and only parent
chemical concentrations were used to establish the model. Metabolites were not modeled.An
attempt to scale the rodent PBPK model to humans, relevant to risk assessment of oral exposures
to benzo[a]pyrene, was presented by Zeilmaker et al. (1999a, b). The PBPK model for
benzo[a]pyrene was derived from an earlier model for TCDD in rats (Zeilmaker and van
Eijkeren, 1997). The structure of the mainly perfusion-limited PBPK model included
61
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
compartments for blood, adipose tissue (with diffusion limitation), slowly and richly perfused
tissues, and the liver (Figure 3-3). However, there was no separate compartment for the lung.
The liver compartment featured the AhR-dependent CYP450 induction mechanism and DNA
adduct formation as a marker for formation of genotoxic benzo[a]pyrene metabolites. It was
assumed that DNA adduct formation and the bulk benzo[a]pyrene metabolism were mediated by
two different metabolic pathways. The model was experimentally calibrated in rats with the data
for EROD and formation of DNA adducts in the liver after i.v. administration of a single dose
and per oral (p.o.) administration of a single or repeated doses of benzo[a]pyrene (Zeilmaker et
al., 1999a).
In order to scale this rat PBPK model to human, Zeilmaker et al. (1999b) assumed
identical values for several parameters in rats and humans, respectively: benzo[a]pyrene tissue
partition coefficients, AhR concentration in liver, rate constant for the decay of the
benzo[a]pyrene-CYP450 complex, half-life of the CYP450 protein, fraction and rate of GI
absorption of benzo[a]pyrene, and rates of formation and repair of DNA adducts in liver. The
basal CYP450 activity in humans was assumed to be lower than that in rat liver (ranging from
almost absent up to equal to that in the rat). The mechanism of AhR-dependent induction of
CYP450 dominated the simulated benzo[a]pyrene-DNA adduct formation in the liver. The
results of PBPK model simulations indicated that the same dose of benzo[a]pyrene administered
to rats or humans might produce one order of magnitude higher accumulation of DNA adducts in
human liver when compared with the rat (Zeilmaker et al., 1999b).
Even though the model of Zeilmaker et al. (1999b) represents a major improvement in
predictive modeling of benzo[a]pyrene toxicokinetics, the results of modeling, which included
interspecies extrapolation, bear significant uncertainties. As emphasized by the authors, the
conversion of benzo[a]pyrene to its mutagenic and carcinogenic metabolites could not be
explicitly modeled in human liver because no suitable experimental data were available.
According to the authors, to improve the model would require direct measurements of basal
activities of CYP1A1 and CYP1A2 and formation of benzo[a]pyrene-DNA adducts in human
liver. Moreover, despite the prior results of the study by Roth and Vinegar (1990), the metabolic
clearance of benzo[a]pyrene in the lungs was not addressed. Also, the toxicokinetic modeling by
Zeilmaker et al. (1999b) addressed only one pathway of benzo[a]pyrene metabolic activation, a
single target organ (the liver), and one route of administration (oral).
For modeling and predicting of the health outcomes of exposures to benzo[a]pyrene, a
mechanistically accurate PBPK model needs to follow, over time and in several target organs,
the rate of accumulation of benzo[a]pyrene-DNA adducts and/or the distribution and fate of
benzo[a]pyrene metabolites (e.g., BPDE) that bind to DNA and other macromolecules.
Alternatively, a stable derivative of the "ultimately carcinogenic" metabolite (e.g.,
benzo[a]pyrene trans-anti-tetrol) may be used as an internal dose surrogate. Calibration of such
a model requires quantitation of these benzo[a]pyrene metabolites in biological samples, which,
62
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
in turn, would require refined and sufficiently sensitive analytical methods. Therefore, while the
metabolic pattern of benzo[a]pyrene has been relatively well characterized qualitatively in
animals, both in vitro and in vivo, the quantitative kinetic relationships between overlapping and
sometimes competing metabolic reactions in potential target organs, essential for meaningful
PBPK modeling, are yet not well defined.
Furthermore, the potential for scaling this model to humans without complete
reparameterization is questionable. According to Liao (2004), the recombinant human CYP1A1
has a higher activity for the formation of benzo[a]pyrene-hydroxy metabolites and a lower
activity for the formation of benzo[a]pyrene-diones, when compared with recombinant rat
CYP1A1. The RN-PBPK model overestimated the benzo[a]pyrene concentration in the liver,
but after 4 hours, it underestimated benzo[a]pyrene concentration in fat. In contrast to the
findings of Zeilmaker et al. (1999a), the adipose tissue compartment was modeled only as
perfusion-limited. Tissue-blood partition coefficients for benzo[a]pyrene and its metabolites
used in the RN-PBPK model were estimated but not validated experimentally or otherwise.
Also, no adjustment was made for using in vitro enzyme activities, especially when measured in
recombinant proteins as was done in the calibration of this model, although they typically differ
from those in vivo.
The published PBPK models for benzo[a]pyrene were evaluated to determine whether the
existing models could be used to extrapolate from rats to humans or for a route to route
extrapolation from oral exposure to the inhalation route. The focus was concentrated on models
for the inhalation route since several well conducted studies by the oral route exist from which to
derive toxicity values. No full PBPK model for the inhalation route was identified. It was
concluded that at present, none of the published models allow for computation of the resulting
internal doses used in the only cancer bioassay available for the inhalation route (Thyssen et al.
1981), nor is there a model for humans that simulates the typical inhalation exposure to
benzo[a]pyrene on poorly soluble carbonaceous particles.
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
37
38
4. HAZARD IDENTIFICATION
4.1. HUMANS STUDIES
4.1.1. Sources of Human Exposure
Although it has no commercial production or significant uses, benzo[a]pyrene is
nevertheless a ubiquitous environmental contaminant resulting from the incomplete combustion
of organic matter. Essentially all humans experience repeated exposure to benzo[a]pyrene,
which can begin in utero and continue throughout life. The magnitude of benzo[a]pyrene
exposure depends on several factors related to lifestyle (e.g., diet, tobacco smoking), occupation,
and living conditions (e.g., urban versus rural setting, domestic heating and cooking methods).
A distinguishing feature concerning benzo[a]pyrene exposure is that environmental sources
always occur as complex mixtures, which may consist of numerous PAHs, including
heterocyclic and nonheterocyclic forms as well as aza arenes and nitro-substituted PAHs
(reviewed in Bostrom et al., 2002; WHO, 1998; ATSDR, 1995; IARC, 1973). Many of these
complex mixture components are carcinogens, and some can exceed the carcinogenic potency of
benzo[a]pyrene, as observed in animal bioassays.
With the exception of certain occupational exposure sources such as aluminum
production and the conversion of coal to coke and coal tar, the major contributor to total PAH
(and thus benzo[a]pyrene) exposure in nonsmokers is the diet (Cogliano et al., 2008; Straif et al.,
2005; Bostrom et al., 2002; Cenni et al., 1993; Andersson et al., 1983; Bjorseth et al., 1978a, b).
In particular, charbroiled and grilled meats, and certain grain products, are important
determinants of PAH exposure in most populations. In very limited populations receiving
clinical coal tar treatment of the skin for conditions such as psoriasis, acute dermal exposure to
benzo[a]pyrene may greatly exceed that received from both occupational and nonoccupational
sources (Godschalk et al., 2001; Pavanello et al., 1999).
4.1.2. Biomonitoring of Benzo[a]pyrene Exposure and Effects In Humans
Quantitative exposure assessment becomes very challenging for complex mixture
components such as benzo[a]pyrene, where individual exposures vary depending on background
concentrations in the environment, lifestyle factors, and occupation. An alternative to measuring
concentrations in various environmental media, human biomonitoring focuses on biomarkers of
internal dose which may also serve as mechanistic indicators of early biological response
following exposure to a genotoxic agent. Carcinogen exposure biomarkers are generated by
uptake and metabolic processes which may result either in detoxification and excretion or in
bioactivation to reactive forms that can bind covalently with DNA and other macromolecules.
Following exposure to complex PAH mixtures, biomonitoring of benzo[a]pyrene uptake can
involve evaluation of urinary metabolites of benzo[a]pyrene or surrogates, DNA adducts in
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
37
38
peripheral blood cells such as leukocytes and lymphocytes, protein adducts in hemoglobin and
serum albumin, and cytogenetic damage in lymphocytes (Gyorffy et al., 2008; Godschalk et al.,
2003; Bostrom et al., 2002; Scherer et al., 2000). Biomarkers offer advantages over traditional
benzo[a]pyrene exposure concentration monitoring because they can be quantitative indicators of
an individual's environmental exposure and internal carcinogen dose irrespective of time, route
of exposure, and inter-individual metabolic differences. Associations between increased cancer
risk and levels of specific biomarkers, such as benzo[a]pyrene-DNA adducts in target or
surrogate tissues, also provide important mechanistic information in studies of human disease
etiology.
Considerable progress has been achieved in the application of biomarker methods for
human biomonitoring of carcinogens (Vineis and Perera, 2007; Perera and Weinstein, 2000;
Poirier et al., 2000). However, further validation is needed before using biomarkers to routinely
predict disease risk. Successful application of biomarkers in benzo[a]pyrene exposure
assessment currently depends on several assumptions: 1) urinary metabolites, whether derived
from benzo[a]pyrene or surrogate PAH compounds such as Py or phenanthrene, provide
information about recent benzo[a]pyrene exposures; 2) DNA and protein adduct measurements
in easily accessible tissues serve as surrogate biomarkers of long-term benzo[a]pyrene exposure
and internal dose for less accessible target tissues (e.g., lung); 3) and cytogenetic biomarkers
reflect early biological effects that correlate with relevant preclinical events occurring at target
sites, and with biomarkers of biologically effective benzo[a]pyrene dose. Evidence is
accumulating to support these assumptions for carcinogenic PAHs, in general, and for
benzo[a]pyrene, in particular, although conflicting and variable results among some studies limit
their application for dose-response assessment. Important issues to address include differences
in biomarker assay sensitivity and specificity, acquired and inherited variations in PAH
bioactivation, detoxification, and DNA repair, and uncertain carcinogen intake(Gyorffy et al.,
2008; Divi et al., 2002; Poirier et al., 2000; Scherer et al., 2000; Santella et al., 1994).
4.1.2.1. Urinary Excretion of Benzo[a]pyrene
Based on animal studies, urinary excretion of benzo[a]pyrene is a minor pathway of
elimination compared with elimination in the feces. The hydroxylated metabolites of
benzo[a]pyrene, 3-hydroxy-benzo[a]pyrene (3-OH-benzo[a]pyrene) and 9-hydroxy-
benzo[a]pyrene, have been identified in the urine of humans exposed to PAH, although they are
often not detectable in human urine despite known exposure to benzo[a]pyrene (Rossella et al.,
2009; Hecht, 2002; Bentsen-Farmen et al., 1999b). In numerous human studies, the
hydroxylated metabolite of Py, 1-OH-Py, is typically used as a surrogate indictor of internal
exposure to carcinogenic PAH mixtures based on the assumption that levels of 1-OH-Py in urine
correlate with exposure to individual PAH compounds of higher molecular weight, including
benzo[a]pyrene. Although urinary 1-OH-Py appears suitable as a sensitive biomarker for total
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
37
38
PAH exposure, results are variable for the correlation of 1-OH-Py with the benzo[a]pyrene
metabolite, 3-OH-benzo[a]pyrene, in urine (Forster et al., 2008; Godschalk et al., 1998a). In a
study of urinary PAHs and their hydroxylated metabolites in 55 coke oven workers, 1-OH-Py
was found in 100% of the urine samples, whereas 3-OH-benzo[a]pyrene was always below the
quantification limit (Rossella et al., 2009). In addition, reported correlations between urinary 1-
OH-Py levels and PAH-DNA adducts in white blood cells (WBCs) are conflicting, although in
one study, 3-OH-benzo[a]pyrene levels in urine were significantly correlated with specific
benzo[a]pyrene diol epoxide adducts in skin DNA following dermal application of coal tar
ointments (Gyorffy et al., 2008; Godschalk et al., 1998a).
Buckley et al. (1995) found that excretion of benzo[a]pyrene metabolites in residents
(non-smokers not employed in high PAH occupational environments) of Phillipsburg, New
Jersey, correlated better with ingestion of benzo[a]pyrene from food rather than with
environmental inhalation exposures. Benzo[a]pyrene metabolites in urine were measured
following "reverse metabolism," a procedure that involves enzymatic hydrolysis of
benzo[a]pyrene conjugates followed by chemical conversion of all hydroxylated benzo[a]pyrene
metabolites back to the parent compound for subsequent thin-layer chromatography and
scanning spectrofluorometry. Buckley et al. (1995) estimated pulmonary uptake at 11 and 2.3 ng
benzo[a]pyrene/person/day for winter and summer, respectively, based on 24-hour personal air
measurements. The median intake of benzo[a]pyrene from the diet was estimated at 176 ng/day.
The median urinary excretion of benzo[a]pyrene and metabolites was 121 and 129 ng/day in
winter and summer, respectively. Based on multiple regression analyses of estimated inhaled
and ingested doses, change in urinary excretion of benzo[a]pyrene was only marginally
predictive of benzo[a]pyrene exposure. Most of the variation in urinary benzo[a]pyrene
excretion was explained by the ingested dose. These results confirm an earlier study of
occupationally exposed (aluminum plant) workers using the same reverse metabolism procedure
to measure urinary PAH elimination (Becher and Bj0rseth, 1983). Although a significant
difference in urinary PAH excretion was seen in nonoccupationallly exposed smokers versus
nonsmokers, PAH metabolite levels in urine of aluminum plant workers did not reflect the large
difference in inhalation exposure relative to controls (Becher and Bj0rseth, 1983).
Giindel et al. (2000) investigated PAH exposure and urinary metabolite excretion in
19 workers from a fireproof stone factory. Along with other PAH metabolites in urine, they
measured 3-OH-benzo[a]pyrene levels following airborne exposures to benzo[a]pyrene ranging
"3
from 0.043 to 5.41 (J,g/m based on personal and stationary air sampling. The study authors
failed to identify a correlation between benzo[a]pyrene inhalation exposure concentration and
urinary 3-OH-benzo[a]pyrene excretion, nor between 1-OH-Py and 3-OH-benzo[a]pyrene levels
in urine.
4.1.2.2. Adduct Formation with DNA and Protein
66
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
A large body of literature supports DNA and protein adducts as biomarkers of a
biologically effective dose (ED) for exposures to some DNA reactive human carcinogens
(Gyorffy et al., 2008; Vineis and Perera, 2007; Hecht, 2004; Godschalk et al., 2003; Bostrom et
al., 2002; Perera and Weinstein, 2000; Poirier et al., 2000). DNA adduct formation is considered
to be a necessary early event in tumor formation for many such compounds, particularly large,
bulky carcinogens such as benzo[a]pyrene and PAHs in general. In PAH exposed populations,
there is evidence demonstrating the formation of DNA adducts with tobacco-derived PAH
including benzo[a]pyrene, not only sites directly exposed, but also in distant organs and the
peripheral blood (see Table 4-1).
Table 4-1. Studies of PAH-DNA adducts in human populations or tissues
exposed to PAHs
Reference
Study description
Arnould et al. (1998)
DNA adducts from leukocytes from heavy smokers
Arnould et al. (1999)
DNA adducts in leukocytes from workers from plant producing carbon
electrodes
Arnould et al. (2000)
DNA adducts in leukocytes of coke oven workers
Assennato et al. (1993)
DNA adducts in peripheral blood leukocytes in coke oven workers
Bartsch et al. (1999)
PAH DNA adduct levels in lung parenchyma of coke oven workers and
smokers
Bartsch et al. (1999)
DNA adduct formation in smokers, tobacco chewers, coke oven workers
Bhattacharya et al. (2003)
benzo|alpyrene-DNA adducts in human urine
Casale et al. (2001)
benzo[a]pyrene adducts in urine of cigarette smokers and women exposed
to household smoke
Galati et al. (2001)
DNA adducts in sera of humans exposed to PAHs
Gallagher et al. (1993)
DNA adducts in blood cells, placental syncytial nuclei, placental tissue
homogenates, and lung cells following exposure to cigarette or coal smoke
Godschalk et al. (1998a)
DNA adducts in alveolar macrophages and subpopulations of white blood
cells in smokers
Godschalk et al. (1998b)
DNA adducts in biopsies of treated skin and in WBCs along with levels of
urinary 1 -OH-pyrene in psoriasis patients being treated with coal tar
Hemminki et al. (1997
DNA adduct formation in exposed humans
Izzotti et al. (1991, 1992)
DNA adducts in pulmonary alveolar macrophages following exposure to
cigarette smoke
Li et al. (2001)
DNA adduct levels in human peripheral blood lymphocytes of SCC patients
and controls
Li et al. (2002)
BPDE-DNA adducts in human breast tissues
Lodovici et al. (1998)
benzo[a]pyrene levels in autoptic lungs of smokers and nonsmokers
Lodovici et al. (1999)
DNA adduct formation in human white blood cells from smokers and
nonsmokers
Mancini et al. (1999)
DNA adducts in cervical cells from cigarette smokers
Melikian et al. (1999)
DNA adducts in epithelial and stromal cervical tissue samples from women
smokers and self-reported nonsmokers after hysterectomy for nonmalignant
conditions
Paleologo et al. (1992)
BPDE-DNA adducts in the white blood cells of patients treated with coal tar
preparations
Pavanello et al. (1999)
DNA adducts in mononuclear white blood cells of coke oven workers and
chimney sweeps
Rojas et al. (2000)
PAH DNA adduct levels in lung parenchyma of coke oven workers and
smokers
Scherer et al. (2000)
benzo[a]pyrene adducts of hemoglobin and albumin in smokers and
67
DRAFT - DO NOT CITE OR QUOTE
-------
Reference
Study description
nonsmokers
Schoket et al. (1993)
DNA adducts in lymphocytes in aluminum plant workers
Schwartz et al. (2003)
DNA analysis (content, damage, cell cycle, and apoptosis) in smokers and
nonsmokers
Shinozaki et al. (1999)
DNA adduct formation in aging smokers and non-smokers
Van Delft et al. (1998)
WBC-DNA adducts in workers in a carbon electrode manufacturing facility
Wiencke et al. (1990)
benzo|alpyrene DNA adducts and SCEs in human lymphocytes
Zenzes et al. (1999a)
DNA adducts in sperm cells from smokers and nonsmokers
Zenzes et al. (1999b)
DNA adducts in embryos from smoking couples
Zhang et al. (1995)
DNA adduct levels in oral mucosa cells from nonsmokers and smokers
68
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Measurement of DNA adducts in target tissues (e.g., lung) can be difficult and invasive;
however, readily accessible nucleated tissue (e.g., WBCs, sperm cells, cervical cells) and
proteins (e.g., hemoglobin and albumin in blood) can serve as surrogate biomarkers of exposure.
DNA adducts in WBCs reflect exposure over a relatively long period, and are indicative of the
individual's metabolic and DNA repair capability, both of which are genetically influenced.
Although protein adducts are not thought to be mechanistic intermediates in PAH-initiated
carcinogenesis, they offer several advantages as biomarkers. For example, hemoglobin and
albumin are much more abundant than DNA; their adducts are not subject to removal by
enzymatic repair, and they can integrate exposures over the protein lifespan of days to weeks.
In humans, PAH albumin and hemoglobin adducts have often been used to investigate
internal dosimetry of direct tobacco and environmental tobacco smoke (ETS) exposures.
Elevated benzo[a]pyrene adducts are often reported in smokers versus nonsmokers and in
children exposed to ETS from their smoking mothers (Hecht, 2004; Philips, 2002).
Benzo[a]pyrene-protein adduct levels in smokers are typically increased twofold compared with
nonsmokers (Scherer et al., 2000; Sherson et al., 1990).
4.1.2.2.1. DNA adduct measures in blood
Major advances in methodology have greatly increased the sensitivity and specificity
(Himmelstein et al., 2009; Jarabek et al., 2009) of DNA adducts measurements. The various
32 33
techniques utilized include immunoassays and immunohistochemistry, [ P]- and [ P]-
postlabelling of modified nucleotides paired with thin layer or high performance liquid
chromatography (TLC and HPLC, respectively), fluorescence and phosphorescence
spectroscopy, electrochemical detection, and mass spectrometry (Arlt et al., 2007; Poirier et al.,
32
2000). [ P]-postlabelling assays and immunological methods are the most commonly used. The
immunoassays employ an antiserum generated against benzo[a]pyrene-DNA adducts. In
addition to DNA adducts formed by benzo[a]pyrene, the antibody cross reacts with DNA adducts
formed by other carcinogenic polycyclic aromatic hydrocarbons (PAHs) (Weston, et al, 1989)
but has no affinity for DNA or PAHs alone. DNA adducts detected in non-laboratory samples
32
using this technique are therefore referred to as "PAH-DNA" adducts. The [ P]-postlabelling
method is highly sensitive in detecting bulky DNA adducts, including but not limited to
benzo[a]pyrene- and PAH-DNA adducts, enabling quantitation down to 1 DNA adduct/109
normal nucleotides, and achieving high resolution of individual adducts particularly when
combined with HPLC. Unless specific chromatographic standards are available to assist in
identification, adducts detected using this technique are referred to in general as "bulky DNA
adducts". Accelerator mass spectrometry is a relatively new method for adduct determination
that has demonstrated even greater sensitivity. Most studies do not identify the specific DNA
adduct structures, but instead report total "PAH-DNA," "bulky-DNA," "hydrophobic-DNA," or
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
36
37
38
"aromatic-DNA" adducts (Alexandrov et al., 2002). Further investigations are needed to
examine correlations between DNA adduct levels determined by different methods, and
correlations between DNA adducts and urinary PAH metabolites.
benzo[a]pyrene-DNA and PAH-DNA adducts have been measured in WBCs in several
PAH-exposed human groups and found to be associated with, and predictive of, elevated cancer
risks (Vineis and Perera, 2007; Pavanello et al., 1999). Coke oven workers experience especially
high exposures to PAHs and demonstrate increased lung cancer rates. Pavanello and coworkers
(2005) studied a group of 67 highly exposed coke oven workers for genetic factors that can
modulate individual responses to carcinogenic PAHs. Levels of BPDE-DNA adducts in
mononuclear WBCs (lymphocyte plus monocyte fraction) were associated with workplace PAH
exposure as indicated by urinary 1-OH-Py excretion. The authors concluded that the elevated
levels of BPDE-DNA adducts reflected both exposure and individual variation in expression of
genes involved with glutathione conjugation activity and DNA excision repair capacity.
In a similar study, Pavanello et al. (2006) screened 585 Caucasian municipal workers
(52% males, 20-62 years old) from northeast Italy for BPDE-DNA adduct formation in
peripheral lymphocytes. Forty-two percent of the participants had elevated anti-BPDE-DNA
8 8
adduct levels, defined as >0.5 adducts/10 nucleotides (mean, 1.28 ± 2.80 adducts/10
nucleotides). Comparison of adduct levels with questionnaire responses indicated that smoking,
frequent consumption of PAH-rich meals (>52 times/year vs. <52 times/year), and long time
periods spent outdoors (>4 hours/day vs. <4 hours/day) were risk factors as all increased BPDE-
DNA adduct levels significantly. Exposure to indoor combustion sources (use of fireplace, coal-
or wood stove >5 times/year) significantly increased the frequency of subjects positive for
BPDE-DNA adducts. Exposure to heavy traffic did not alter lymphocyte BPDE-DNA adduct
levels. Smoking and high-PAH diets were associated with increased BPDE-DNA adduct levels.
In nonsmokers, high-PAH diets and extended time spent indoors associated with increased
BPDE-DNA adduct formation, while in smokers, personal cigarette smoking was the only factor
that was positively correlated with adduct levels. These results demonstrate the potential utility
of BPDE-DNA adduct measurement as a biomarker of PAH exposure not only in heavily PAH-
exposed occupations and tobacco smokers, but also in the general population, including
nonsmokers.
Several studies investigated the level of DNA adducts in the WBCs of foundry workers
(Perera et al., 1988) or coke oven workers (Arnould et al., 2000; Assennato et al., 1993a, b) using
an immunoassay. The antibody's cross-reactivity with adducts originating from PAHs other than
32
benzo[a]pyrene generally results in higher reported adduct levels than the [ P]-postlabelling
method which can isolate benzo[a]pyrene-DNA adducts. All three studies involved several
levels of potential benzo[a]pyrene exposure as assessed by environmental sampling: 0.02, 5, 25,
and 45 (J,g/m3 (Arnould et al., 2000); <0.05, 0.05-0.2, and >0.2 (J,g/m3 (Perera et al., 1988); and
0.03-12.6 (J,g/m3 (Assennato et al., 1993a, b). Adduct levels ranged from 0.05 to 7 fmol PAH-
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
DNA per jag DNA, and variation within exposure groups was high; however, all three studies
reported statistically significant correlations between exposure and adduct levels. However,
there are a large number of studies, many of them conducted by the same laboratories cited here,
that did not report positive correlations between benzo[a]pyrene exposure and DNA adduct
formation in people exposed occupationally or to cigarette smoke (e.g., van Delft et al., 2001;
Arnould et al., 1999; Pan et al., 1998; Peluso et al., 1998; Lewtas et al., 1997; Assennato et al.,
1993a, b; Kriek et al., 1993; Paleologo et al., 1992; Herbert et al., 1990). These studies illustrate
that attempting to correlate DNA adducts with benzo[a]pyrene exposure by a single route of
exposure (e.g., inhalation), or by occupation alone, may produce highly variable or misleading
results.
Perera et al. (2005a) measured BPDE-DNA adduct levels in maternal and umbilical cord
blood obtained following normal delivery from 329 nonsmoking pregnant women exposed to
emissions from fires during the 4 weeks following the collapse of the World Trade Center
(WTC) building in New York City on 09/11/2001. BPDE-DNA adduct levels were highest in
study participants who lived within 1 mile of the WTC, with inverse correlation between cord
blood levels and distance from WTC. For the group of participants that resided within 1 mile of
o
the WTC, maternal levels were 0.30 ±0.16 adducts/10 nucleotides, umbilical cord levels were
0.28 ± 0.08; for the group employed within 1 mile of WTC, the corresponding values were 0.25
±0.11 and 0.24 ± 0.12, and for the unexposed referent group, the values were 0.22 ± 0.10 and
0.23 ± 0.10. Differences between referent and exposed subjects were marginally significant (0.1
>p> 0.05); however, the trend for adducts in maternal blood decreasing with increasing distance
from the WTC was statistically significant (p = 0.02) as was the finding that the percentage of
participants with detectable BPDE-DNA adducts increased significantly (trendp = 0.05) with
increasing time and proximity to WTC-from 52.6% (reference group) to 66.7% (employed
group) to 80.0%) (resident group).
Tuntawiroon et al. (2007) evaluated airborne PAH concentrations and internal biomarker
levels in 115 Thai school boys (8-12 years old) attending schools adjacent to high-density traffic
areas in Bangkok (high exposure group), compared with 69 boys (9-13 years old) attending
schools located in a provincial area (low exposure group). Ambient air concentrations (roadside
and school areas) and personal breathing zone air concentrations of 10 particulate PAHs were
measured in Bangkok and in the provincial location. Peripheral blood samples were collected
32
from the school boys for [ P]-postlabeling determination of bulky DNA adducts in lymphocytes.
Based on toxic equivalency factors (TEFs, Nisbet and LaGoy, 1992) for PAH, benzo[a]pyrene
equivalent exposures from personal breathing zone measurements in Bangkok children were
about 3.5-fold greater than in rural provincial children (p < 0.001). Interestingly, although bulky
DNA adducts were increased fivefold in Bangkok children compared with provincial children {p
< 0.001), adduct levels were negatively correlated with total PAH and benzo[a]pyrene equivalent
exposures.
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
4.1.2.2.2. Adduct measures in reproductive tissues
Zenzes et al. (1998) used immunostaining to detect PAH-DNA adducts in ovarian
granulosa-lutein cells from women undergoing procedures for reproductive assistance. The 32
women in the sample were separated by smoking status and consisted of 14 active smokers (1-20
cigarettes per day), 7 passive smokers (nonsmoker but husband smoked), and 11 non-smokers
with a non-smoking partner. The collected cells were fixed and stained with the anti-BP-DNA
antibody which is known to cross-react with several carcinogenic PAH-DNA adducts. In this
assay, darker nuclear staining results from a greater concentration of PAH-DNA adducts and is
assigned a higher score. The proportion of nuclei exhibiting respective intensities of staining
were combined to obtain an overall intensity score for the sample. The observed staining
intensity was related to smoking status, with a mean ± SE of 1.91 ± 0.10 among active smokers,
1.21 ± 0.20 among passive smokers, and 0.62 ± 0.17 among nonsmokers (p < 0.0001). The
authors concluded that smoking-related seminal DNA adducts could be a potential source of
transmissible zygotic DNA damage.
In another study using the same immunostaining assay, Zenzes et al. (1999a) investigated
the occurrence of PAH-DNA adducts in semen in relation to tobacco use. The study included 23
men (11 smokers and 12 nonsmokers), mean age 37 years, recruited through couples attending
an in vitro fertilization - embryo transfer clinic in Toronto. Among smokers, the mean amount
smoked was 20.6 cigarettes per day. The PAH-DNA adduct staining intensity was higher in the
sperm samples from smokers (mean ± SE, 1.73 ± 0.09) when compared with nonsmokers (0.93 ±
0.10) (p < 0.0001); 5.7% and 30.9% of the sperm from smokers and nonsmokers, respectively,
exhibited negative staining and 21.2% and 3.8%, respectively, exhibited strong staining. The
authors concluded that smoking-related DNA adducts in semen could potentially be a source of
zygotic DNA damage.
A related study examined the presence of PAH-DNA adducts in 112 blastomere cells
from 22 pre-implantation embryos available through an in vitro fertilization - embryo transfer
clinic (Zenzes et al., 1999b). The donated embryos were grouped with respect to the maternal
and paternal smoking status (n = 8 both parents smoked, n = 12 father smoked but mother did not
smoke, and n=7 neither parent smoked). Five of the embryos (4 from the father-only smoking
and 1 from a non-smoking couple) were very fragmented and were not used in the analysis of
PAH-DNA staining intensity. Among the smoking parents, women smoked less per day than
men (mean ± SE, 13.0 ± 3.0 and 16.8 ± 1.5 cigarettes per day for women and men, respectively).
The intensity score was higher for embryos with at least one parent who smoked (1.40 ± 0.28)
compared with embryos from nonsmoking parents (0.38 ± 0.14) (p = 0.015), but there was little
difference between embryos with two compared with one smoking parent (1.34 ± 0.30 and 1.48
± 0.55 for both smokers and one smoker, respectively). Similar results were seen using the
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
36
37
38
proportion of embryos exhibiting any staining rather than a mean staining measure, and with the
proportion of blastomere cells exhibiting staining. Intensity score was correlated with the
amount smoked by the father (p = 0.020) but not by the mother. Analysis of the presence of
PAH-DNA adducts in sperm cells also revealed an increased proportion and staining intensity in
samples from smokers compared with nonsmokers (p < 0.0001 for categorical analysis of
negative, weak, moderate and strong staining intensity). The authors noted that the demonstrated
presence of these adducts in the embryos, reflecting most strongly a paternal origin, may affect
the viability of the pregnancy.
4.1.2.3. Benzo[a]pyrene-Induced Cytogenetic Damage
Many studies measure cytogenetic damage as biomarkers of early biological effects
which also reflect exposure to genotoxic chemicals. Standard cytogenetic end points include
chromosomal aberration (CA), sister chromatid exchange (SCE), micronucleus (MN) formation,
hypoxanthine guanine phosphoribosyl transferase (hprt) mutation frequency, and GPA mutation
frequency (Gyorffy et al., 2008). These biomarkers are often incorporated in multi-endpoint
studies with other biomarkers of exposure. Because they indicate related but different endpoints,
there is often a lack of correlation between the different categories of biomarkers.
32
Merlo et al. (1997) evaluated DNA adduct formation (measured by [ P]-postlabelling)
and micronuclei in WBCs of 94 traffic policemen vs. 52 residents from the metropolitan area of
Genoa, Italy. All study subjects wore personal air samplers for 5 hours of one work shift, and
levels of benzo[a]pyrene and other PAHs. Policemen were exposed to 4.55 ng
3 3
benzo[a]pyrene/m air, compared with urban residents who were exposed to 0.15 ng/m . DNA
adduct levels in policemen were 35% higher than in urban residents (p = 0.007), but micronuclei
in urban residents were 20% higher than in policemen (p = 0.02). Linear regressions of DNA
adducts and MN incidence, respectively, vs. benzo[a]pyrene exposure levels did not reveal
significant correlations.
Perera and coworkers assessed DNA damage in Finnish iron foundry workers in two
separate studies and using three methodologies. Based on results from personal sampling and
stationary monitoring in both studies, three levels of benzo[a]pyrene air concentrations were
3 3 3
defined: low (<5 ng/m benzo[a]pyrene), medium (5-12 ng/m ), and high (>12 ng/m ). (Perera
et al., 1994, 1993). In the first study, involving 48 workers, several biomarkers were analyzed
for dose-response and interindividual variability (Perera et al., 1993). PAH-DNA adducts were
determined in WBCs using an immunoassay as described in Section 4.1.2.2.1 and enzyme-linked
immunosorbent assay (ELISA) with fluorescence detection. Mutations at the hprt locus were
also measured in WBC DNA. The latter assay is based on the fact that each cell contains only
one copy of the hprt gene, which is located on the X-chromosome. While male cells have only
one X-chromosome, female cells inactivate one of the two X-chromosomes at random. The gene
is highly sensitive to mutations such that in the event of a crucial mutation in the gene, enzyme
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
35
36
37
38
activity disappears completely from the cell. In addition, mutations at the GPA gene locus were
measured in red blood cells (RBCs). The GPA mutation frequency was not correlated with
either benzo[a]pyrene exposure or PAH-DNA adduct formation. However, both PAH-DNA
adduct levels and hprt mutation frequency increased with increasing benzo[a]pyrene exposure.
In addition, there was a highly significant correlation between incidence of hprt mutations and
PAH-DNA adduct levels (p = 0.004).
In a second study, Perera et al. (1994) surveyed 64 iron foundry workers with
assessments conducted in 2 successive years; 24 of the workers provided blood samples in both
years. Exposure to benzo[a]pyrene, collected by personal and area sampling in the 1st year of
"3
the study, ranged from <5 to 60 ng/m and was estimated to have decreased by 40% in the 2nd
year. The levels of PAH-DNA adducts were roughly 50% lower in the 2nd year, presumably
reflecting decreased exposure. The longer-lived hprt mutations were not as strongly influenced
by the decreasing exposure to benzo[a]pyrene. Study subjects who did not have detectable levels
of DNA adducts were excluded from the study. As in the previous study, a strong correlation
between DNA adduct levels and incidence of hprt mutations was observed (Perera et al., 1993).
Kalina et al. (1998) studied several cytogenetic markers in 64 coke oven workers and
34 controls employed at other locations within the same plant. Airborne benzo[a]pyrene and
seven other carcinogenic PAHs were collected by personal air samplers, which showed ambient
"3
benzo[a]pyrene concentrations ranging widely from 0.002 to 50 (J,g/m in coke oven workers and
from 0.002 to 0.063 (J,g/m3 in controls. CAs, SCEs, high-frequency cells (HFCs), and SCE
heterogeneity index (SCE-H) were all significantly increased with benzo[a]pyrene exposure.
Except for increases in HFCs, no effect of smoking was observed. Consistent with studies of
PAH-DNA adduct formation, reduced cytogenetic response at high exposure levels produced a
nonlinear dose-response relationship. The authors also evaluated the potential influence of
polymorphisms in enzymes involved in the metabolism of benzo[a]pyrene. Glutathione S-
transferase Ml (GSTM-1) and N-acetyl transferase (NAT)-2 polymorphisms were studied and
no evidence of the two gene polymorphisms having any influence on the incidence of
cytogenetic damage was found.
Motykiewicz et al. (1998) conducted a similar study of genotoxicity associated with
benzo[a]pyrene exposure in 67 female residents of a highly polluted industrial urban area of
Upper Silesia, Poland, and compared the results to those obtained from 72 female residents of
another urban but less polluted area in the same province of Poland. Urinary mutagenicity and
1-OH-Py levels, PAH-DNA adducts in oral mucosa cells (detected by immunoperoxidase
staining), SCE, HFC, CA, bleomycin sensitivity, and GSTM-1 and CYP1A1 polymorphisms in
blood lymphocytes were investigated. High volume air samplers and gas chromatography were
used to quantify ambient benzo[a]pyrene levels which during the summer were 3.7 ng/m in the
3 3
polluted area and 0.6 ng/m in the control area. During winter, levels rose to 43.4 and 7.2 ng/m
in the two areas, respectively. The cytogenetic biomarkers (CA and SCE/HFC), urinary
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
mutagenicity, and urinary 1-OH-Py excretion were significantly increased in females from the
polluted area, and differences appeared to be more pronounced during winter time. PAH-DNA
adduct levels were significantly increased in the study population, when compared to the
controls, only in the winter season. No difference in sensitivity to bleomycin-induced
lymphocyte chromatid breaks was seen between the two populations. As with the study by
Kalina et al. (1998), genetic polymorphisms assumed to affect the metabolic transformation of
benzo[a]pyrene were not associated with any difference in the incidence of DNA damage.
In a study of Thai school boys in urban (Bangkok) and rural areas, bulky (including but
not limited to BPDE-type) DNA adduct levels were measured in lymphocytes along with DNA
single strand breaks (SSBs), using the comet assay, and DNA repair capacity (Tuntawiroon et al.,
2007). Ambient air and personal breathing zone measurements indicated that Bangkok school
children experienced significantly higher exposures to benzo[a]pyrene and total PAHs. A
significantly higher level of SSBs (tail length 1.93 ± 0.09 vs. 1.28 ± 0.12 [j,m, +51%; p < 0.001)
was observed in Bangkok school children when compared with rural children, and this parameter
was significantly associated with DNA adduct levels. A significantly reduced DNA repair
capacity (0.45 ± 0.01 vs. 0.26 ± 0.01 y-radiation-induced deletions per metaphase, -42%;
p < 0.001) was also observed in the city school children, again significantly associated with
DNA adduct levels. It was not evident why higher environmental PAH exposure would be
associated with lowered DNA repair capacity. However, because the personal breathing zone
PAH levels and DNA adduct levels were not associated with each other, it is conceivable that the
city school children had a priori lower DNA repair capacities that contributed significantly to the
high adduct levels. The authors considered genetic differences between the two study
populations as a possible reason for this observation.
4.1.3. Epidemiologic Findings in Humans
The association between human cancer and contact with PAH-containing substances,
such as soot, coal tar, and pitch, has been widely recognized since the early 1900s (Bostrom et
al., 2002). Although numerous epidemiology studies establish an unequivocal association
between PAH exposure and human cancer, defining the causative role for benzo[a]pyrene and
other specific PAHs remains a challenge. In essentially all reported studies, either the
benzo[a]pyrene exposure and/or internal dose are not known, or the benzo[a]pyrene carcinogenic
effect cannot be distinguished from the effects of other PAH and non-PAH carcinogens.
Nevertheless, three types of investigations provide support for the involvement of
benzo[a]pyrene in some human cancers: molecular epidemiology studies; population- and
hospital-based case-control studies; and occupational cohort studies. In some cohort studies,
benzo[a]pyrene exposure concentrations were measured and thus provide a means to link
exposure intensity with observed cancer rates. In case-control studies, by their nature,
benzo[a]pyrene and total PAH doses can only be estimated.
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
4.1.3.1. Molecular Epidemiology and Case-Control Cancer Studies
Defective DNA repair capacity leading to genomic instability and, ultimately, increased
cancer risk is well documented (Wu et al., 2007, 2005). Moreover, sensitivity to mutagen-
induced DNA damage is highly heritable and thus represents an important factor that determines
individual cancer susceptibility. Based on studies comparing monozygotic and dizygotic twins,
the genetic contribution to BPDE mutagenic sensitivity was estimated to be 48.0% (Wu et al.,
2007). BPDE has been used as an etiologically relevant mutagen in case-control studies to
examine the association between elevated lung and bladder cancer risk and individual sensitivity
to BPDE-induced DNA damage. Mutagen sensitivity is determined by quantifying chromatid
breaks or DNA adducts in phytohemagglutinin-stimulated peripheral blood lymphocytes as an
indirect measure of DNA repair capacity.
In a hospital-based case-control study involving 221 lung cancer cases and 229 healthy
controls, DNA adducts were measured in stimulated peripheral blood lymphocytes after
incubation with BPDE in vitro (Li et al., 2001). Lung cells from cancer cases showed consistent
statistically significant elevations in induced BPDE-DNA adducts, compared with controls,
regardless of subgroup by age, sex, ethnicity, smoking history, weight loss, or family history of
cancer. The BPDE-induced DNA adduct levels, when grouped by quartile using the levels in
controls as cutoff points, were significantly dose-related with lung cancer risk (odds ratios [ORs]
1.11, 1.62, and 3.23; trend test,/? < 0.001). In a related hospital-based case-control study
involving 155 lung cancer patients and 153 healthy controls, stimulated peripheral blood
lymphocytes were exposed to BPDE in vitro (Wu et al., 2005). DNA damage/repair was
evaluated using the comet assay, and impacts on cell cycle checkpoints measured using a
fluorescence-activated cell-sorting method. The lung cancer cases exhibited significantly higher
levels of BPDE-induced DNA damage than the controls (p < 0.001), with lung cancer risk
positively associated with increasing levels of DNA damage when grouped in quartiles (trend
test,/? < 0.001). In addition, lung cancer patients demonstrated significantly shorter cell cycle
delays in response to BPDE exposure, which correlated with increased DNA damage.
Sensitivity to BPDE-induced DNA damage in bladder cancer patients supports the results
observed in lung cancer cases. In a hospital-based case-control study involving 203 bladder
cancer patients and 198 healthy controls, BPDE-induced DNA damage was specifically
evaluated at the chromosome 9p21 locus in stimulated peripheral blood lymphocytes (Gu et al.,
2008). Deletions of 9p21, which includes critical components of cell cycle control pathways, are
associated with a variety of cancers. After adjusting for age, sex, ethnicity, and smoking status,
individuals with high BPDE-induced damage at 9p21 were significantly associated with
increased bladder cancer risk (OR 5.28; 95% confidence interval [CI] 3.26-8.59).
Categorization of patients into tertiles for BPDE sensitivity relative to controls demonstrated a
dose-related association between BPDE-induced 9p21 damage and bladder cancer risk.
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
Collectively, the results of molecular epidemiology studies with lung and bladder cancer patients
indicate that individuals with a defective ability to repair BPDE-DNA adducts are at increased
risk for cancer and, moreover, that specific genes linked to tumorigenesis pathways may be
molecular targets for benzo[a]pyrene and other carcinogens.
Due to the importance of the diet as a benzo[a]pyrene exposure source, several
population- and hospital-based case-control studies have investigated the implied association
between dietary intake of benzo[a]pyrene and risk for several tumor types. In a study involving
193 pancreatic cancer cases and 674 controls (Anderson et al., 2005), another involving
626 pancreatic cancer cases and 530 controls (Li et al., 2007), and a third involving
146 colorectal adenoma cases and 228 controls (Sinha et al., 2005), dietary intake of
benzo[a]pyrene was estimated using food frequency questionnaires. In all studies, the primary
focus was on estimated intake of benzo[a]pyrene (and other carcinogens) derived from cooked
meat. Overall, cases when compared with controls had higher intakes of benzo[a]pyrene and
other food carcinogens, leading to the conclusion that benzo[a]pyrene plays a role in the etiology
of these tumors in humans. In a supportive follow-up case-control study of colorectal adenomas,
increased leukocyte PAH-DNA adducts were measured in cases when compared with controls,
using a method that recognizes BPDE and several other PAHs bound to DNA (Gunter et al.,
2007).
4.1.3.2. Cohort Cancer Studies
Epidemiologic studies of workers in PAH-related occupations indicate increased human
cancer risks associated with iron and steel production, roofing, carbon black production, and
exposure to diesel exhaust (Bosetti et al., 2007). Exposure to benzo[a]pyrene is only one of
numerous contributors to the cancer risk from complex PAH-containing mixtures that occur in
the workplace. Although some occupational cohort studies report measured or estimated
inhalation exposure concentrations for benzo[a]pyrene, none report biomarkers of internal
benzo[a]pyrene dose in study subjects (reviewed in Bosetti et al., 2007; Armstrong et al., 2004).
Several of these cohort studies (summarized below) demonstrate a positive exposure-response
relationship with cumulative PAH exposure using benzo[a]pyrene—or a proxy such as benzene-
soluble matter (BSM) that can be converted to benzo[a]pyrene—as an indicator substance.
These studies provide insight and support for the causative role of benzo[a]pyrene in human
cancer.
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
4.1.3.2.1. Cancer incidence in aluminum and electrode production plants.
Exposure to benzo[a]pyrene and BSM in aluminum smelter workers is strongly
associated with bladder cancer and weakly associated with lung cancer (Boffetta et al., 1997;
Tremblay et al., 1995; Armstrong et al., 1994; Gibbs, 1985; Theriault et al., 1984). In an
analysis of pooled data from nine cohorts of aluminum production workers, 688 respiratory tract
cancer cases were observed versus 674.1 expected (pooled RR 1.03; CI 0.96-1.11) (Bosetti et al.,
2007). A total of 196 bladder cancer cases were observed in eight of the cohorts, compared with
155.7 expected (pooled RR 1.29; CI 1.12-1.49). Based on estimated airborne benzo[a]pyrene
exposures from a meta-analysis of eight cohort studies, the predicted lung cancer relative risk
"3
(RR) per 100 (J,g/m -years of cumulative benzo[a]pyrene exposure was 1.16 (95% CI 1.05-1.28)
(Armstrong et al., 2004).
Spinelli et al. (2006) reported a 14-year update to a previously published historical cohort
study (Spinelli et al., 1991) of Canadian aluminum reduction plant workers. The results
confirmed and extended the findings from the earlier epidemiology study. The study surveyed a
total of 6,423 workers with >3 years of employment at an aluminum reduction plant in British
Columbia, Canada, between the years 1954 and 1997, and evaluated all types of cancers. The
focus was on cumulative exposure to coal tar pitch volatiles, measured as BSM and as
benzo[a]pyrene. Benzo[a]pyrene exposure categories were determined from the range of
predicted exposures over time from statistical exposure models. There were 662 cancer cases, of
which approximately 98% had confirmed diagnoses. The overall cancer mortality rate
(standardized mortality ratio [SMR] 0.97; CI 0.87-1.08) and cancer incidence rate (standardized
incidence ratio [SIR] 1.00; CI 0.92-1.08) were not different from that of the British Columbia
general population. However, this study identified significantly increased incidence rates for
cancers of the bladder (SIR 1.80; CI 1.45-2.21) and the stomach (SIR 1.46; CI 1.01-2.04). The
lung cancer incidence rate was only slightly higher than expected (SIR = 1.10; CI 0.93-1.30).
Significant dose-response associations with cumulative benzo[a]pyrene exposure were seen for
bladder cancer (p trend <0.001), stomach cancer (p trend <0.05), lung cancer (p trend <0.001),
non-Hodgkin lymphoma (p trend <0.001), and kidney cancer (p trend <0.01), although the
overall incidence rates for the latter three cancer types were not significantly elevated versus the
general population. Similar cancer risk results were obtained using BSM as the exposure
measure; the cumulative benzo[a]pyrene and BSM exposures were highly correlated (r = 0.94).
In several occupational cohort studies of workers in Norwegian aluminum production
plants, personal and stationary airborne PAH measurements were performed.
In a study covering 11,103 workers and 272,554 person x years of PAH exposure, cancer
incidence was evaluated in six Norwegian aluminum smelters (Romundstad et al., 2000a, b).
"3
Reported estimates of PAH exposure concentrations reached a maximum of 3,400 (J,g/m PAH
(680 (J,g/m benzo[a]pyrene). The overall number of cancers observed in this study did not differ
significantly from control values (SIR 1.03; CI 1.0-1.1). The data from this study showed
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
37
significantly increased incidences for cancer of the bladder (SIR 1.3; CI 1.1-1.5), and elevated,
but not significant, SIRs for larynx (SIR 1.3; CI 0.8-1.9), thyroid (SIR 1.4; CI 0.7-2.5), and
multiple myeloma (SIR 1.4; CI 0.9-1.9). Incidence rates for bladder, lung, pancreas, and kidney
cancer (the latter three with SIRs close to unity) were subjected to a cumulative exposure-
response analysis. The incidence rate for bladder cancer showed a trend with increasing
cumulative exposure and with increasing lag times (up to 30 years) at the highest exposure level.
The incidence of both lung and bladder cancers was greatly increased in smokers. The authors
reported that using local county rates rather than national cancer incidence rates as controls
increased the SIR for lung cancer (SIR 1.4; CI 1.2-1.6) to a statistically significant level.
4.1.3.2.2. Cancer incidence in coke oven, coal gasification, and iron and steel foundry
workers.
An increased risk of death from lung and bladder cancer is reported in some studies
involving coke oven, coal gasification, and iron and steel foundry workers (Bostrom et al., 2002;
Boffetta et al., 1997). An especially consistent risk of lung lung cancer across occupations is
noted when cumulative exposure is taken into consideration (e.g. RR of 1.16 per 100 unity-yrs
for aluminum smeleter workers, 1.17 for coke oven workers, and 1.15 for coal gasificiation
workers). In an analysis of pooled data from 10 cohorts of coke production workers, 762 lung
cancer cases were observed versus 512.1 expected (pooled RR 1.58; CI 1.47-1.69) (Bosetti et al.,
2007). Significant variations in risk estimates among the studies were reported, particularly in
the large cohorts (RRs of 1.1, 1.2, 2.0, and 2.6). There was no evidence for increased bladder
cancer risk in the coke production workers. Based on estimated airborne benzo[a]pyrene
exposures from a meta-analysis of 10 cohort studies, the predicted lung cancer RR per
-3
100 (J,g/m -years of cumulative benzo[a]pyrene exposure was 1.17 (95% CI 1.12-1.22)
(Armstrong et al., 2004).
A meta-analysis of data from five cohorts of gasification workers reported 251 deaths
from respiratory tract cancer, compared with 104.7 expected (pooled RR 2.58; 95% CI 2.28-
2.92) (Bosetti et al., 2007). Pooled data from three of the cohorts indicated 18 deaths from
urinary tract cancers, versus 6.0 expected (pooled RR 3.27; 95% CI 2.06-5.19). Based on
estimated airborne benzo[a]pyrene exposures from a meta-analysis of four gas worker cohort
-3
studies, the predicted lung cancer RR per 100 (J,g/m -years of cumulative benzo[a]pyrene
exposure was 1.15 (95% CI 1.11-1.20) (Armstrong et al., 2004).
Increased risks were reported in iron and steel foundry workers for cancers of the
respiratory tract, bladder and kidney. In an analysis of pooled data from 10 cohorts,
1,004 respiratory tract cancer cases were observed versus 726.0 expected (pooled RR 1.40; CI
1.31-1.49) (Bosetti et al., 2007). A total of 99 bladder cancer cases were observed in seven of
the cohorts, compared with 83.0 expected (pooled RR 1.29; CI 1.06-1.57). For kidney cancer,
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
37
38
40 cases were observed compared with 31.0 expected based on four studies (pooled RR 1.30;
95% CI 0.95-1.77).
Xu et al. (1996) conducted a nested case-control study, surveying the cancer incidence
among 196,993 active or retired workers from the Anshan Chinese iron and steel production
complex. A large number of historical benzo[a]pyrene measurements (1956-1995) were
available. The study included 610 cases of lung cancer and 292 cases of stomach cancer, with
959 matched controls from the workforce. After adjusting for nonoccupational risk factors such
as smoking and diet, significantly elevated risks for lung cancer and stomach cancer were
identified for subjects employed for 15 or more years, with ORs varying among job categories.
For either type of cancer, highest risks were seen among coke oven workers: lung cancer, OR =
3.4 (CI 1.4-8.5); stomach cancer, OR = 5.4 (CI 1.8-16.0).
There were significant trends for long-term cumulative benzo[a]pyrene exposure vs. lung
cancer (p = 0.004) or stomach cancer (p = 0.016) incidence. For cumulative total
benzo[a]pyrene exposures of < 0.84, 0.85-1.96, 1.97-3.2 and >3.2 the ORs for lung cancer were
1.1 (CI 0.8-1.7), 1.6 (CI 1.2-2.3), 1.6 (1.1-2.3) and 1.8 (CI 1.2-2.5). For cumulative total
benzo[a]pyrene exposures of < 0.84, 0.85-1.96, 1.97-3.2 and > 3.2 the ORs for stomach cancer
were 0.9 (CI 0.5-1.5), 1.7 (CI 1.1-2.6), 1.3 (0.8-2.1) and 1.7 (CI 1.1-2.7). However, the
investigators noted that additional workplace air contaminants were measured, which might have
influenced the outcome. Of these, asbestos, silica, quartz, and iron oxide-containing dusts may
have been confounders. For lung cancers, cumulative exposures to total dust and silica dust both
showed significant dose-response trends (p = 0.001 and 0.007, respectively), while for stomach
cancer, only cumulative total dust exposure showed a marginally significant trend (p = 0.061).
For cumulative total dust exposures of < 69, 69-279, 280-882 and > 883 mg/m3 the ORs for lung
cancer were 1.4 (CI 1.2-1.9), 1.2 (CI 1.0-2.19), 1.4 (CI 1.0-2.0) and 1.9 (CI 1.3-2.5),
respectively. For cumulative silica dust exposures of < 3.7, 3.7-10.39, 10.4-27.71 and > 27.72
mg/m3 the ORs for lung cancer were 1.7 (CI 1.2-2.4), 1.5 (CI 1.0-2.1), 1.5 (CI 1.0-2.1) and 1.8
(CI 1.2-2.5), respectively. For cumulative total dust exposures of < 69, 69-279, 280-882 and >
883 mg/m3 ORs for stomach cancer were 1.3 (CI 0.8-2.1), 14 (CI 0.9-2.2), 12 (CI 0.8-1.9) and
1.6 (CI 1.1-2.5), respectively.
Exposure-response data from studies of coke oven workers in the United States have
often been used to derive quantitative risk estimates for PAH mixtures, and for benzo[a]pyrene
as an indicator substance (Bostrom et al., 2002). However, there are numerous studies of coke
oven worker cohorts that do not provide estimates of benzo[a]pyrene exposure. An overview of
the results of these and other studies can be obtained from the review of Boffetta et al. (1997).
4.1.3.2.3. Cancer incidence in asphalt workers and roofers.
These groups encompass different types of work (asphalt paving vs. roofing) and also
different types of historical exposure that have changed from using PAH-rich coal tar pitch to the
80
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
use of bitumen or asphalt, both of which are rather low in PAHs due to their source (crude oil
refinery) and a special purification process. Increased risks for lung cancer were reported in
large cohorts of asphalt workers and roofers; evidence for increased bladder cancer risk is weak
(Burstyn et al., 2007; Partanen and Boffetta, 1994; Chiazze et al., 1991; Hansen, 1991, 1989;
Hammond et al., 1976). In an analysis of pooled data from two cohorts of asphalt workers, 822
lung cancer cases were observed versus 730.7 expected (pooled RR 1.14; CI 1.07-1.22) (Bosetti
et al., 2007). In two cohorts of roofers, analysis of pooled data indicated that 138 lung cancer
cases were observed, compared with 91.9 expected (pooled RR 1.51; CI 1.28-1.78) (Bosetti et
al., 2007).
4.1.3.3. Noncattcer Disease Caused by Benzo[a]pyrene
Because accumulating evidence indicates that PAH exposure is a risk factor for ischemic
heart disease (IHD), Burstyn et al. (2005) investigated 418 cases of fatal IHD in a cohort of
12,367 asphalt paving workers exposed to low-level PAH from bitumen and coal tar. The
follow-up started in 1953 and ended in 2000, with an average exposure of 17 ± 9 years
(minimum: one work season), resulting in 193,889 person-years of observation. Previous
analyses of this cohort indicated no association between PAH exposure and excess mortality
from cancer or by all causes. Quantitative estimates of exposure to benzo[a]pyrene were
obtained for paving operations on the basis of previously available personal exposure
measurements from workers in the asphalt industry (but not necessarily from cohort members).
Exposures were calculated as average (0-68 [reference], 68-105, 106-146, 147-272, and 273+
ng/m3) and cumulative (0-189 [reference], 189-501, 502-931, 932-2,012, and 2,013+ ng/m3-
years), respectively.
Cumulative and average exposure indices for benzo[a]pyrene were positively associated
with mortality from IHD; the highest RR coincided with an average exposure to benzo[a]pyrene
"3
of 273 ng/m or higher (RR = 1.64; CI = 1.13-2.28). A similar risk was observed for the highest
"3
cumulative benzo[a]pyrene exposure group (>2,013 ng/m -years) (RR = 1.58; CI = 0.98-2.55).
Length of employment had no influence on this result. A dose-response was evident for IHD,
but not for other types of cardiovascular disease. The RR remained elevated even with
adjustment for smoking as a confounder; the RR was 1.24 under the extreme assumption of 0%
never smokers, 30% former smokers, and 70% current smokers in the highest-exposed group.
The authors discussed the possibility of bias because some of their study subjects might have
been exposed to other IHD-causing factors that were not controlled for in their study, or might
have been misclassified.
An occupational study of Canadian aluminum smelter workers investigated the effect of
benzo[a]pyrene exposure on cardiopulmonary mortality (Friesen et al., 2010). Adjusted internal
comparisons for smoking were conducted using Cox regression for male subjects (n = 6,423).
Ischemic heart disease (IHD) was associated with cumulative benzo[a]pyrene exposure with a
81
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
hazard ratio of 1.62 (95% CI 1.06-2.46) in the highest benzo[a]pyrene exposure category. For
active employment, the hazard ratio for IHD was 2.39 (95% CI 0.95-6.05) in the highest
cumulative benzo[a]pyrene exposure category.
Other studies have reported potential prenatal effects and associated birth outcomes
induced by inhalation exposure to PAHs, including benzo[a]pyrene. Perera et al. (2005a) studied
329 nonsmoking pregnant women (30 ± 5 years old) possibly exposed to PAHs from fires during
the 4 weeks after 09/11/2001. Maternal and umbilical cord blood levels of benzo[a]pyrene
(BPDE)-DNA adducts were highest in study participants who lived within 1 mile of the WTC,
with an inverse correlation between cord blood levels and distance from the WTC. Neither cord
blood adduct level nor ETS alone was positively correlated with adverse birth outcomes.
However, the interaction between ETS exposure and cord blood adducts was significantly
associated with reduced birth weight and head circumference. Among babies exposed to ETS in
utero, a doubling of cord blood benzo[a]pyrene-DNA adducts was associated with an 8%
decrease in birth weight (p = 0.03) and a 3% decrease in head circumference (p = 0.04).
Perera et al. (2005b) compared various exposures—ETS, nutrition, pesticides, material
hardship—with birth outcomes (length, head circumference, cognitive development). ETS
exposure and intake of PAH-rich foods by pregnant women were determined by questionnaire.
Levels of benzo[a]pyrene(BPDE)-DNA adducts were determined in umbilical cord blood
collected at delivery. The study population consisted of Dominican or African-American
nonsmoking pregnant women (n = 529; 24 ± 5 years old) free of diabetes, hypertension, HIV,
and drug or alcohol abuse. Benzo[a]pyrene adducts, ETS, and dietary PAHs were not
significantly correlated with each other. However, the interaction between benzo[a]pyrene-DNA
adducts and ETS exposure was significantly associated with reduced birth weights (-6.8%; p =
0.03) and reduced head circumference (-2.9%; p = 0.04).
Tang et al. (2006) measured benzo[a]pyrene(BPDE)-DNA adducts in maternal and
umbilical cord blood obtained at delivery from a cohort of 150 nonsmoking women and their
newborns in China. Exposure assessment was related to the seasonal operation of a local, coal-
fired power plant; however, airborne PAH concentrations were not measured. Dietary PAH
intake was not included as a covariate because it did not significantly contribute to the final
models, but ETS, sex, and maternal height and weight were considered as covariates. DNA
adduct levels were compared to several birth outcomes and physical development parameters,
such as gestational age at birth; infant sex, birth weight, length, head circumference, and
malformations; maternal height and pregnancy weight total weight gain; complications of
pregnancy and delivery; and medications used during pregnancy.
High cord blood adduct levels were significantly associated with reduced infant/child
weight at 18 months (P = -0.048,p = 0.03), 24 months (P = -0.041,p = 0.027), and 30 months of
age (P = -0.040, p = 0.049); decreased birth head circumference was marginally associated with
DNA adduct levels (P = -0.011 ,p = 0.057). Maternal adduct levels were correlated neither with
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
35
36
37
38
cord blood adduct levels nor with fetal and child growth. Among female infants, cord blood
adduct levels were significantly associated with smaller birth head circumference (p = 0.022) and
with lower weight at 18 months (p = 0.014), 24 months (p = 0.012), and 30 months of age (p =
0.033), and with decreased body length at 18 months of age (p = 0.033). Among male infants,
the corresponding associations were also inverse but statistically not significant.
Considerable evidence of a deleterious effect of smoking on male and female fertility has
accumulated from epidemiological studies of time to pregnancy, ovulatory disorders, semen
quality, and spontaneous abortion (reviewed in Waylen et al., 2009; Cooper and Moley, 2008;
Soares and Melo, 2008). In addition, the effect of smoking, particularly during the time of the
perimenopausal transition, on acceleration of ovarian senescence (menopause) has also been
established (Midgette and Baron, 1990). More limited data is available pertaining specifically to
measures of benzo[a]pyrene and reproductive outcomes.
Neal et al. (2008, 2007) examined levels of benzo[a]pyrene and other PAHs in follicular
fluid and serum sample from 36 women undergoing in vitro fertilization at a clinic in Toronto,
and compared the successful conception rate in relation to benzo[a]pyrene levels. The women
were classified by smoking status, with 19 who were current cigarette smokers, 7 with passive or
sidestream smoke exposure (i.e., non-smoker with a partner who smoked) and 10 non-smoked
exposed. An early follicular phase blood sample and follicular fluid sample from the follicle at
the time of ovum retrieval were collected and analyzed for the presence of benzo[a]pyrene,
acenapthelene, phenanthrene, pyrene and chrysene using GC/MS (detection limit 5 pg/ml). The
frequency of non-dectable levels of serum benzo[a]pyrene was highest in the non-smoking group
(60.0%, 14.3%, and 21.0% below detection limit in non-smoking, sidestream smoke, and active
smoking groups, respectively). A similar pattern was seen with follicular fluid benzo[a]pyrene
(30.0%), 14.3%), and 10.5% below detection limit limit in non-smoking, sidestream smoke, and
active smoking groups, respectively). In the analyses comparing mean values across groups, an
assigned value of 0 was used for non-detectable samples. Follicular fluid benzo[a]pyrene levels
were higher in the active smoking group (mean ± SE, 1.32 ± 0.68 ng/ml) than in the sidestream
(0.05 ± 0.01 ng/ml) or non-smoking (0.03 ± 0.01 ng/ml) groups (p = 0.04). The between-group
differences in serum benzo[a]pyrene levels were not statistically significant (0.22 ±0.15, 0.98 ±
0.56, and 0.40 ± 0.13 ng/ml in non-smoking, sidestream smoke, and active smoking groups,
respectively), and there were no differences in relation to smoking status. Among active
smokers, the number of cigarettes smoked per day was strongly correlated with follicular fluid
benzo[a]pyrene levels (r= 0.7, p<0 .01). Follicular fluid benzo[a]pyrene levels were
significantly higher among the women who did not conceive (1.79 ng/ml ± 0.86) compared with
women who did get pregnant (mean approximately, 0.10 ng/ml, as estimated from graph) (p <
0.001), but serum levels of benzo[a]pyrene were not associated with successful conception.
A small case-control study conducted between August 2005 and February 2006 in
Lucknow city (Uttar Pradesh), India examined PAH concentrations in placental tissues (Singh et
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
al., 2008) in relation to risk of preterm birth. The study included 29 cases (delivery between 28
and < 36 weeks gestation) and 31 term delivery controls. Demographic data smoking history,
reproductive history and other information were collected by interview, and a 10 g sample of
placental tissue was collected form all participants. Concentration of specific PAHs in placental
tissue was determined using HPLC. In addition to benzo[a]pyrene, the PAHs assayed were
naphthalene, acenapththylene, phenanthrene, fluorene, anthracene, benzo(a)anthracene,
fluoranthene, pyrene, beno(k)fluoranthene, benzo(b)fluoranthene, benzo(g,h,i)perylene, and
dibenzo(a,h)anthracene. PAH exposure in this population was from environmental sources and
from cooking. The age of study participants ranged from 20 to 35 years. There was little
difference in birthweight between cases and controls (mean 2.77 and 2.75 in the case and control
groups, respectively). Placental benzo[a]pyrene levels were lower than the levels of the other
PAHs detected (mean 8.83 ppb in controls for benzo[a]pyrene compared with 25-30 ppb for
anthracene, beno(k)fluoranthene, benzo(b)fluoranthene, and dibenzo(a,h)anthracene, 59 ppb for
acenaphthylene, and 200 - 380 naphthalene, phenanthrene, fluoranthene, and pyrene; non-
detectable levels of fluorine, benzo(a)anthracene, and benzo(g,h,i)perylene were found). There
was little difference in benzo[a]pyrene levels between cases (mean ± SE 13.85 ± 7.06 ppb) and
controls (8.83 ± 5.84 ppb), but elevated levels of fluoranthene (325.91 ± 45.14 and 208.6 ±
21.93 ppb in cases and controls, p < 0.05) and benzo(b)fluoranthene (61.91 ± 12.43 and 23.84 ±
7.01 ppb in cases and controls, p < 0.05) were seen.
Wu et al. (2010) conducted a study of benzo[a]pyrene-DNA adduct levels in relation to
risk of fetal death in Tianjin, China. This case-control study included women who experienced a
missed abortion before 14 weeks gestational age, that is a fetal death that remained in utero and
so required surgical intervention. Cases were matched by age and gravidity to controls (women
undergoing induced abortion due to an unplanned or unwanted pregnancy). The study excluded
women who smoked, women with chronic disease and pregnancy complications, and women
with occupational exposures to PAHs. Residency within Tianjin for at least one year was also an
eligibility criterion. The participation rate was high: 81 of 84 eligible cases participanted and 81
of 89 eligible controls participated. Data pertaining to demographic characteristics, reproductive
history, and factors relating to potential PAH exposure were collected using a structured
interview, and samples from the aborted tissue were obtained. In two of the four hospitals used
in the study, blood samples from the women (n=51 cases and 51 controls) were also collected.
The presence of benzo[a]pyrene-BPDE adducts was assessed in the blood and tissue samples
using HPLC. There was no correlation between blood and aborted tissue levels of
benzo[a]pyrene adducts (r = -0.12 for the 102 blood-tissue pairs, r = -0.02 for the 51 case paris
and r = -0.21 for the 51 control pairs). (The authors noted that there was little difference
between women with and without blood samples in terms of the interview- based measures
collected or in terms of the DNA-adduct levels in aborted tissue.) benzo[a]pyrene-adduct levels
were similar but slightly lower in the aborted tissue of cases compared with controls (mean ± SD
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
4.8 ± 6.0 in cases and 6.0 ± 7.4 in controls, p = 0.29). In the blood samples, however,
benzo[a]pyrene-adduct levels were higher in cases (6.0 ± 4.7 and 2.7 ± 2.2 in cases and controls,
respectively, p < 0.001). In logistic regression analyses using a continuous adduct measure, the
OR was 1.35 (95% CI 1.11 - 1.64) per adduct/108 nucleotide. These results adjusted for
education and household income, but were very similar to the unadjusted results. Categorizing
exposure at the median value resulted in an adjusted OR of 4.27 (95% CI 1.41 - 12.99) in the
high compared with low benzo[a]pyrene-adduct group. There was no relation between
benzo[a]pyrene-adduct levels in the aborted tissue and missed abortion in the logistic regression
analyses using either the continuous (adjusted OR 0.97, 95% CI 0.93 - 1.02) or dichotomous
exposure measure (adjusted OR 0.76, 95% CI 0.37 - 1.54). Associations between missed
abortion and several interview-based measures of potential PAH exposure were also seen:
adjusted OR 3.07 (95% CI 1.31 - 7.16) for traffic congestion near residence, 3.52 (95% CI 1.44
- 8.57) for commuting by walking, 3.78 (95% CI 1.11 - 12.87) for routinely cooked during
pregnancy, and 3.21 (95% CI 0.98 - 10.48) for industrial site or stack near residence, but there
was no association with other types of commuting (e.g., by bike, car, or bus).
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS—ORAL, INHALATION, AND DERMAL
4.2.1. Oral
4.2.1.1. Subchronic Studies
De Jong et al. (1999) 35-day rat study
De Jong et al. (1999) treated male Wistar rats (eight/dose group) with benzo[a]pyrene
(98.6%) purity) dissolved in soybean oil by gavage 5 days/week for 35 days at doses of 0, 3, 10,
30, or 90 mg/kg-day (adjusted doses: 0, 2.14, 7.14, 21.4, and 64.3 mg/kg-day). At the end of the
exposure period, rats were necropsied, organ weights were determined, and major organs and
tissues were prepared for histological examination (adrenals, brain, bone marrow, colon, caecum,
jejunum, heart, kidney, liver, lung, lymph nodes, esophagus, pituitary, spleen, stomach, testis,
and thymus). Blood was collected for examination of hematological endpoints, but there was no
indication that serum biochemical parameters were analyzed. Immune parameters included
determinations of serum immunoglobulin levels (IgG, IgM, IgE, and IgA), relative spleen cell
distribution, and spontaneous cytotoxicity of spleen cell populations determined in a natural-
killer (NK) cell assay.
Body weight gain was decreased beginning at week 2 at the high dose of 90 mg/kg-day;
there was no effect at lower doses (De Jong et al., 1999). Hematology revealed a dose-related
decrease in RBC count, hemoglobin, and hematocrit at >10 mg/kg-day (Table 4-2). A minimal
but significant increase in mean cell volume and a decrease in mean cell hemoglobin
85
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
concentration were noted at 90 mg/kg-day, and may indicate dose-related toxicity for the RBCs
and/or RBC precursors in the bone marrow. A decrease in WBCs, attributed to a decrease in the
number of lymphocytes (approximately 50%) and eosinophils (approximately 90%), was
observed at 90 mg/kg-day; however, there was no effect on the number of neutrophils or
monocytes. A decrease in the cell number in the bone marrow observed in the 90 mg/kg-day
dose group was consistent with the observed decrease in the RBC and WBC counts at this dose
level. In the 90 mg/kg-day dose group, brain, heart, kidney, and lymph node weights were
decreased and liver weight was increased (Table 4-2). Decreases in heart weight at 3 mg/kg-day
and in kidney weight at 3 and 30 mg/kg-day were also observed, but these changes did not show
dose-dependent responses. Dose-related decreases in thymus weight were statistically
significant at >10 mg/kg-day (Table 4-2).
86
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-2. Exposure-related effects in male Wistar rats exposed to
benzo[a]pyrene by gavage 5 days/week for 5 weeks
Dose (mg/kg-d)
Effect
0
3
10
30
90
Hematologic effects
(mean ± SD; n = 7-8)
White blood cells (109/L)
14.96 ± 1.9
13.84 ± 3.0
13.69 ± 1.8a
13.58 ± 2.9a
8.53 ± l.la
Red blood cells (109/L)
8.7 ±0.2
8.6 ±0.2
8.3 ±0.2
7.8 ±0.4
7.1 ±0.4a
Hemoglobin (mmol/L)
10.5 ±0.2
10.4 ±0.3
9.8 ± 0.2a
9.5 ± 0.4a
8.6 ± 0.6a
Hematocrit (L/L)
0.5 ±0.01
0.5 ±0.01
0.47 ± 0.01a
0.46 ± 0.02a
0.43 ± 0.02a
Serum immunoglobulin levels
(mean ± SD; n = 7-8)
IgM
100 ± 13
87 ± 16
86 ±31
67 ± 16a
81 ±26
IgG
100 ±40
141±106
104 ±28
106 ± 19
99 ±29
IgA
100 ±28
73 ±29
78 ±67
72 ±22
39 ± 19a
IgE
100 ±65
50 ±20
228 ±351
145±176
75 ±55
Cellularity (mean ± SD; n = 7-8)
Spleen (cell number xio7)
59 ± 15
71 ± 14
59 ± 13
63 ± 10
41 ± 10a
Bone marrow (G/L)
31 ±7
36 ±5
31 ±8
27 ±8
19 ± 4a
Spleen cell distribution (%)
B cells
39± 4
36 ±2
34 ± 3a
32 ± 4a
23 ± 4a
T cells
40 ±9
48 ± 12
40 ±9
36 ±2
44 ±6
Th cells
23 ±7
26 ±7
24 ±5
22 ±4
26 ±4
Ts cells
24 ±5
26 ±6
24 ±7
19 ± 2
27 ±5
Body (g) and organ (mg) weights
(means; n = 7-8)
Body weight
305
282a
300
293
250a
Brain
1,858
1,864
1,859
1,784
l,743a
Heart
1,030
934a
1,000
967
863a
Kidney
1,986
1,761a
1,899
l,790a
l,626a
Liver
10,565
9,567
11,250
11,118
12,107a
Thymus
517 ±47
472 ± 90
438 ± 64a
388 ± 71a
198 ± 65a
Spleen
551
590
538
596
505
Mandibular lymph nodes
152
123
160
141
00
Mesenteric lymph nodes
165
148
130a
158
107a
Popliteal lymph nodes
19
18
19
17
10a
Thymus cortex surface area
(% of total surface area of thymus;
mean ± SD; n = 6-8)
77.9 ±3.8
74.4 ±2.2
79.2 ±5.9
75.8 ±4.0
68.9 ± 5.2a
"Significantly (p < 0.05) different from control mean. For body weight and organ weight means, SDs were only
reported for thymus weights.
Source: De Jong et al. (1999).
1
2 Statistically significant reductions were also observed in the relative cortex surface area
3 of the thymus and thymic medullar weight at 90 mg/kg-day, but there was no difference in cell
4 proliferation between treated and control animals using the proliferating cell nuclear antigen
5 (PCNA) technique. Changes in the following immune parameters were noted: dose-related and
6 statistically significant decrease in the relative number of B cells in the spleen at 10 (13%), 30
7 (18%), and 90 mg/kg-day (41%); significant decreases in absolute number of cells harvested in
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
36
37
38
the spleen (31%), in the number of B cells in the spleen (61%), and NK cell activity in the spleen
(E:T ratio was 40.9 ± 28.4% that of the controls) at 90 mg/kg-day; and a decrease in serum IgM
(33%>) and IgA (61%>) in rats treated with 30 and 90 mg/kg-day, respectively. The decrease in
the spleen cell count was attributed by the study authors to the decreased B cells and suggested a
possible selective toxicity of benzo[a]pyrene to B cell precursors in the bone marrow. The study
authors considered the decrease in IgA and IgM to be due to impaired production of antibodies,
suggesting a role of thymus toxicity in the decreased (T-cell dependent) antibody production. In
addition to the effects on the thymus and spleen, histopathologic examination revealed treatment-
related lesions only in the liver and forestomach at the two highest dose levels, but the incidence
data for these lesions were not reported by De Jong et al. (1999). Increased incidence for
forestomach basal cell hyperplasia (p < 0.05 by Fisher's exact test) was reported at 30 and
90 mg/kg-day, and increased incidence for oval cell hyperplasia in the liver was reported at
90 mg/kg-day (p < 0.01, Fisher's exact test). The results indicate that 3 mg/kg-day was a no-
observed-adverse-effect level (NOAEL) for effects on hematological parameters (decrease in
RBC count, hemoglobin, and hematocrit) and immune parameters (decreased thymus weight and
percent of B cells in the spleen) noted in Wistar rats at 10 mg/kg-day (the lowest-observed-
adverse-effect level [LOAEL]) and above. Lesions of the liver (oval cell hyperplasia) and
forestomach (basal cell hyperplasia) occurred at doses >30 mg/kg-day.
Knuckles et al. (2001) 90-day rat study
Knuckles et al. (2001) exposed male and female F344 rats (20/sex/dose group) to
benzo[a]pyrene (98%> purity) at doses of 0, 5, 50, or 100 mg/kg-day in the diet for 90 days. Food
consumption and body weight were monitored, and the concentration of benzo[a]pyrene in the
food was adjusted every 3-4 days to maintain the target dose. The authors indicated that actual
intake of benzo[a]pyrene by the rats was within 10%> of the calculated intake, and the nominal
doses were not corrected to actual doses. Hematology and serum chemistry parameters were
evaluated. Urinalysis was also performed. Animals were examined for gross pathology, and
histopathology was performed on selected organs (stomach, liver, kidney, testes, and ovaries).
Statistically significant decreases in RBC counts and hematocrit level (decreases as much as 10
and 12%o, respectively) were observed in males at doses >50 mg/kg-day and in females at
100 mg/kg-day. A maximum 12%> decrease (statistically significant) in hemoglobin level was
noted in both sexes at 100 mg/kg-day. Blood chemistry analysis showed a significant increase in
blood urea nitrogen (BUN) only in high-dose (100 mg/kg-day) males. Histopathology
examination revealed an apparent increase in the incidence of abnormal tubular casts in the
kidney in males at 5 mg/kg-day (40%>), 50 mg/kg-day (80%>) and 100 mg/kg-day (100%>),
compared to 10%> in the controls. Only 10%> of the females showed significant kidney tubular
changes at the two high dose levels compared to zero animals in the female control group. The
casts were described as molds of distal nephron lumen and were considered by the study authors
88
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
to be indicative of renal dysfunction. From this study, male F344 rats appeared to be affected
more severely by benzo[a]pyrene treatment than the female rats. However the statistical
significance of the kidney lesions are unclear. Several reporting gaps and inconsistencies
regarding the reporting of kidney abnormalities in Knuckles et al (2001) make interpretation of
the results difficult. Results of histopathological kidney abnormalities (characterized primarily
as kidney casts) were presented graphically and the data were not presented numerically in this
report. No indication was given in the graph that any groups are statistically different than
controls, though visual examination of the magnitude of response and error bars appears to
indicate a four fold increase in kidney casts in males compared to the control group (40
compared to 10%). The figure legend reports the data as "percentage incidence of abnormal
kidney tissues" and reports values are mean plus or minus standard deviation. However, text
under the materials and methods section states that for histopathological data, Fisher's Exact Test
was used. This would involve the pairwise comparison of incidence and not means. There are
additional internal inconsistencies in the data presented. Data appear to indicate that incidences
for males are as follows: control: 10%, 5mg/kg-day: 40%, 50 mg/kg-day: 80% and 100 mg/kg-
day: 100%, however, these incidences are inconsistent with the size of the study groups which
were reported as 6-8 animals per group. The study authors were contacted but did not respond to
EPA's request for clarification of study design and/or results. Due to issues of data reporting, a
LOAEL could not be established for the increased incidence of kidney lesions. Based on the
statistically significant hematological effects including decreases in RBC counts, hematocrit, and
BUN, the NOAEL in males was 5 mg/kg-day and the LOAEL was 50 mg/kg-day, based on in
F344 rats. No exposure-related histological lesions were identified in the stomach, liver, testes,
or ovaries in this study.
Kroese et al. (2001) 5-week rat study
In a range-finding study, Wistar (specific pathogen-free [SPF] Riv:TOX) rats
(10/sex/dose group) were administered benzo[a]pyrene (97.7% purity) dissolved in soybean oil
by gavage at dose levels of 0, 1.5, 5, 15, or 50 mg/kg body weight, 5 days/week for 5 weeks
(Kroese et al., 2001). Behavior, clinical symptoms, body weight, and food and water
consumption were monitored. None of the animals died during the treatment period. Animals
were sacrificed 24 hours after the last dose. Urine and blood were collected for standard
urinalysis and hematology and clinical chemistry evaluation. Liver enzyme induction was
monitored based on EROD activity in plasma. Animals were subjected to macroscopic
examination, and organ weights were recorded. The esophagus, stomach, duodenum, liver,
kidneys, spleen, thymus, lung, and mammary gland (females only) from the highest-dose and
control animals were evaluated for histopathology. Intermediate dose-groups were examined if
abnormalities were observed in the higher-dose groups.
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
A significant, but not dose-dependent, increase in food consumption in males at
>1.5 mg/kg-day and a decrease in females at >5 mg/kg-day was observed (Kroese et al., 2001).
Water consumption was statistically significantly altered in males only, a decrease at 1.5, 5, and
15 mg/kg-day and an increase at 50 mg/kg-day. Organ weights of lung, spleen, kidneys,
adrenals, and ovaries were not affected by treatment. There was a dose-related, statistically
significant decrease in thymus weight in males at 15 and 20 mg/kg-day (decreased by 28 and
33%, respectively) and a significant decrease in females at 50 mg/kg-day (decreased by 17%)
(Table 4-3). In both sexes, liver weight was statistically significantly increased only at
50 mg/kg-day by about 18% (Table 4-3).
Table 4-3. Exposure-related effects in Wistar rats exposed to
benzo[a]pyrene by gavage 5 days/week for 5 weeks
Organ
Dose (mg/kg)
0
1.5
5
15
50
Liver weight (g; mean ± SD)
Males
Females
6.10 ±0.26
4.28 ± 0.11
6.19 ± 0.19
4.40 ±0.73
6.13 ± 0.10
4.37 ±0.11
6.30 ± 0.14
4.67 ±0.17
7.20 ± 0.18a
5.03 ± 0.15a
Thymus weight (mg; mean ± SD)
Males
Females
471 ± 19
326 ± 12
434 ± 20
367 ± 23
418 ±26
351 ±25
342 ± 20a
317 ± 30
317 ± 21a
271 ± 16a
Basal cell hyperplasia of the
forestomach
(incidence with slight severity)
Males
Females
1/10
0/10
1/10
1/10
4/10
1/10
3/10
3/10a
7/10
7/10a
aSignificantly (p < 0.05) different from control mean; n = 10/sex/group.
Source: Kroese et al. (2001).
Hematological evaluation revealed only statistically nonsignificant, small dose-related
decreases in hemoglobin in both sexes, and RBC counts in males. Clinical chemistry analysis
showed a small, but statistically significant, increase in creatinine levels in males only at
1.5 mg/kg-day, but this effect was not dose-dependent. A dose-dependent induction of liver
microsomal EROD activity was observed, with a five-fold induction at 1.5 mg/kg-day compared
to controls, reaching 36-fold in males at 50 mg/kg-day; the fold induction in females at the top
dose was less than in males. At necropsy, significant, dose-dependent macroscopic findings
were not observed.
Histopathology examination revealed a statistically significant increase in basal cell
hyperplasia in the forestomach of females at doses >15 mg/kg-day (Kroese et al., 2001). The
induction of liver microsomal EROD was not accompanied by any adverse histopathologic
findings in the liver at the highest dose, 50 mg/kg-day, so the livers from intermediate-dose
90
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
groups were, therefore, not examined. An increased incidence of brown pigmentation of red
pulp (hemosiderin) in the thymus was observed in treated animals of both sexes. However, this
tissue was not examined in intermediate-dose groups. This range-finding 5-week study
identified a NOAEL of 5 mg/kg-day and a LOAEL of 15 mg/kg-day, based on decreased thymus
weight and forestomach hyperplasia in Wistar rats.
Kroese et al. (2001) 90-day rat study
Kroese et al. (2001) exposed Wistar (Riv:TOX) rats (10/sex/dose group) to
benzo[a]pyrene (98.6% purity, dissolved in soybean oil) by gavage at 0, 3, 10, or 30 mg/kg body
weight, 5 days/week for 90 days. The rats were examined daily for behavior and clinical
symptoms and by palpation. Food and water consumption, body weights, morbidity, and
mortality were monitored. At the end of the exposure period, rats were subjected to macroscopic
examination and organ weights were recorded. Blood was collected for hematology and serum
chemistry evaluation and urine was collected for urinalysis. All gross abnormalities, particularly
masses and lesions suspected of being tumors were also evaluated. The liver, stomach,
esophagus, thymus, lung, spleen, and mesenteric lymph node were examined histopathologically.
In addition, cell proliferation in forestomach epithelium was measured as the prevalence of S-
phase epithelial cells displaying bromodeoxyuridine (BrdU) incorporation.
There were no obvious effects on behavior of the animals, and no difference was
observed in survival or food consumption between exposed animals and controls (Kroese et al.
2001). Higher water consumption and slightly lower body weights than the controls were
observed in males but not females at the high dose of 30 mg/kg-day. Hematological
investigations showed only nonsignificant, small dose-related decreases in RBC count and
hemoglobin level in both sexes. Clinical chemistry evaluation did not show any treatment-
related group differences or dose-response relationships for alanine aminotransferase (ALT),
serum aspartate transaminase (AST), lactate dehydrogenase (LDH), or creatinine, but a small
dose-related decrease in y-glutamyl transferase (GGT) activity was observed in males only.
Urinalysis revealed an increase in urine volume in males at 30 mg/kg-day, which was not dose
related. At the highest dose, both sexes showed increased levels of urinary creatinine and a dose-
related increase in urinary protein. However, no further investigation was conducted to
determine the underlying mechanisms for these changes. At necropsy, reddish to brown/gray
discoloration of the mandibular lymph nodes was consistently noted in most rats; occasional
discoloration was also observed in other regional lymph nodes (axillary). Statistically significant
increases in liver weight were observed at 10 and 30 mg/kg-day in males only (15 and 29% ) and
a decrease in thymus weight in both sexes only at 30 mg/kg-day (17 and 33% decrease in
females and males, respectively, compared with controls) (Table 4-4). At 10 mg/kg-day, thymus
weight in males was decreased by 15%, but the decrease did not reach statistical significance.
91
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Table 4-4. Means ± SDa for liver and thymus weights in Wistar rats exposed
to benzo[a]pyrene by gavage 5 days/week for 90 days
Dose (mg/kg-d)
Organ
0
3
10
30
Liver weight (g)
Males
7.49 ±0.97
8.00 ±0.85
8.62 ± 1.30b
9.67 ± 1.17b
Females
5.54 ±0.70
5.42 ±0.76
5.76 ±0.71
6.48 ± 0.78b
Thymus weight (mg)
Males
380 ± 60
380 ±110
330 ±60
270 ± 40b
Females
320 ± 60
310 ± 50
300 ± 40
230 ± 30b
a Reported as SE, but judged to be SD (and confirmed by study authors).
bSignificantly (p < 0.05) different from control mean; student t-test (unpaired, two-tailed); n = 10/sex/group.
Source: Kroese et al. (2001).
Histopathologic examination revealed what was characterized by Kroese et al. (2001) as
basal cell disturbance in the epithelium of the forestomach in males (p < 0.05) and females
(p < 0.01) at 30 mg/kg-day. The basal cell disturbance was characterized by increased number of
basal cells, mitotic figures, and remnants of necrotic cells; occasionally early nodule
development; infiltration by inflammatory cells (mainly histiocytes); and capillary hyperemia,
often in combination with the previous changes (Kroese et al., 2001). Incidences for these
lesions (also described as "slight basal cell hyperplasia") in the 0, 3, 10, and 30-mg/kg-day
groups were 0/10, 2/10, 3/10, and 7/10 in female rats and 2/10, 0/10, 6/10, and 7/10 for male rats.
Nodular hyperplasia was noted in one animal of either sex at 30 mg/kg-day. A significant
(p < 0.05) increase in proliferation of forestomach epithelial cells was detected at doses
>10 mg/kg-day by morphometric of analysis of nuclei with BrdU incorporation. The mean
numbers of BrdU-staining nuclei per unit surface area of the underlying lamina muscularis
mucosa were increased by about two- and three- to four-fold at 10 and 30 mg/kg-day,
respectively, compared with controls. A reduction of thymus weight and increase in the
incidence of thymus atrophy (the report described the atrophy as slight, but did not specify the
full severity scale used in the pathology examination) was observed in males only at 30 mg/kg-
day (p< 0.01 compared with controls). Incidences for thymus atrophy for the control through
high-dose groups were 0/10, 0/10, 0/10, and 3/10 for females and 0/10, 2/10, 1/10, and 6/10 for
males. No significant differences were observed in the lungs of control and treated animals. In
the esophagus, degeneration and regeneration of muscle fibers and focal inflammation of the
muscular wall were judged to be a result of the gavage dosing rather than of benzo[a]pyrene
treatment.
The target organs of benzo[a]pyrene toxicity in this 90-day dietary study of Wistar rats
were the forestomach, thymus, and liver. The LOAEL for forestomach hyperplasia, decreased
92
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
thymus weight, thymus atrophy, and increased liver weight was 30 mg/kg-day and the NOAEL
was 10 mg/kg-day.
4.2.1.2. Chronic Studies and Cancer Bioassays
Kroese et al. (2001) 2-year rat study
Kroese et al. (2001) exposed Wistar (Riv:TOX) rats (52/sex/dose group) to
benzo[a]pyrene (98.6% purity) in soybean oil by gavage at nominal doses of 0, 3, 10, or
30 mg/kg-day, 5 days/week, for 104 weeks. Mean achieved dose levels were 0, 2.9, 9.6, and
29 mg/kg-day. Additional rats (6/sex/group) were sacrificed after 4 and 5 months of exposure
for analysis of DNA adduct formation in blood and major organs and tissues. The rats were
6 weeks old at the start of exposure. The rats were examined daily for behavior and clinical
symptoms and by palpation. Food and water consumption, body weights, morbidity, and
mortality were monitored during the study. Complete necropsy was performed on all animals
that died during the course of the study, were found moribund, or at terminal sacrifice (organ
weight measurement was not mentioned in the report by Kroese et al., 2001). The organs and
tissues collected and prepared for microscopic examination included: brain, pituitary, heart,
thyroid, salivary glands, lungs, stomach, oesophagus, duodenum, jejunum, ileum, caecum, colon,
rectum, thymus, kidneys, urinary bladder, spleen, lymph nodes, liver pancreas, adrenals, sciatic
nerve, nasal cavity, femur, skin including mammary tissue, ovaries/uterus, and testis/accessory
sex glands. Some of these tissues were examined only when gross abnormalities were detected.
All gross abnormalities, particularly masses and lesions that appeared to be tumors, were also
examined.
At 104 weeks, survival in the control group was 65% (males) and 50% (females),
whereas mortality in the 30 mg/kg-day dose group was 100% after about week 70. At 80 weeks,
survival percentages were about 90, 85 and 75% in female rats in the 0, 3, and 10 mg/kg-day
groups, respectively; in males, respective survival percentages were about 95, 90, and 85% at
80 weeks. Survival of 50% of animals occurred at 104, 104, about 90, and 60 weeks for control
through high-dose females; for males, the respective times associated with 65% survival were
104, 104, 104, and about 60 weeks. The high mortality rate in high-dose rats was attributed to
liver or forestomach tumor development, not to noncancer systemic effects. After 20 weeks,
body weight was decreased (compared with controls by >10%) in 30-mg/kg-day males, but not
in females. This decrease was accompanied by a decrease in food consumption. Body weights
and food consumption were not adversely affected in the other dose groups compared to
controls. In males, there was a dose-dependent increase in water consumption starting at week
13, but benzo[a]pyrene treatment had no significant effects on water consumption in females.
Tumors were detected at significantly elevated incidences at several tissue sites in female
and male rats at doses >10 and >3 mg/kg-day, respectively (Table 4-4; Kroese et al., 2001). The
93
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
tissue sites with the highest incidences of tumors were the liver (hepatocellular adenoma and
carcinoma) and forestomach (squamous cell papilloma and carcinoma) in both sexes (Table 4-4).
The first liver tumors were detected in week 35 in high-dose male rats. Liver tumors were
described as complex, with a considerable proportion (59/150 tumors) metastasizing to the lungs.
At the highest dose level, 95% of rats with liver tumors had malignant carcinomas (95/100;
Table 4-4). Forestomach tumors were associated with the basal cell proliferation observed
(without diffuse hyperplasia) in the forestomach of rats in the preliminary range-finding and
90-day exposure studies described previously in Section 4.2.1. At the highest dose level, 59% of
rats with forestomach tumors had malignant carcinomas (60/102; Table 4-4). Other tissue sites
with distinctly elevated incidences of tumors in the 30 mg/kg-day dose group included: the oral
cavity (papilloma and squamous cell carcinoma [SCC]) in both sexes, and the jejunum
(adenocarcinoma), kidney (cortical adenoma) and skin (basal cell adenoma and carcinoma) in
male rats (Table 4-4). In addition, auditory canal tumors (carcinoma or squamous cell papilloma
originating from pilo-sebaceous units including the Zymbal's gland) were also detected in both
sexes at 30 mg/kg-day, but auditory canal tissue was not histologically examined in the lower
dose groups and the controls (Table 4-4). Gross examination revealed auditory canal tumors
only in the high-dose group.
94
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-5. Incidences of exposure-related neoplasms in Wistar rats treated
by gavage with benzo[a]pyrene, 5 days/week, for 104 weeks
Dose (mg/kg-d)
0
3
10
30b
Site
Females"
Oral cavity
Papilloma
0/19
0/21
0/9
9/31°
see
1/19
0/21
0/9
9/31°
Basal cell adenoma
0/19
0/21
1/9
4/31
Sebaceous cell carcinoma
0/19
0/21
0/9
1/31
Oesophagus
Sarcoma undifferentiated
0/52
0/52
2/52
0/52
Rhabdomy o sarcoma
0/52
1/52
4/52
0/52
Fibrosarcoma
0/52
0/52
3/52
0/52
Forestomach
Squamous cell papilloma
1/52
3/51
20/510
25/52°
sec
0/52
3/51
10/51°
25/52°
Liver
Hepatocellular adenoma
0/52
2/52
7/52°
1/52
Hepatocellular carcinoma
0/52
0/52
32/52°
50/52°
Cholangiocarcinoma
0/52
0/52
1/52
0/52
Anaplastic carcinoma
0/52
0/52
1/52
0/52
Auditory canal
Benign tumor
0/0
0/0
0/0
1/20
Squamous cell papilloma
0/0
0/1
0/0
1/20
Carcinoma
0/0
0/1
0/0
13/20°
Males3
Oral cavity
Papilloma
0/24
0/24
2/37
10/38°
Squamouse cell carcinoma
1/24
0/24
5/37
11/38°
Basal cell adenoma
0/24
0/24
0/37
2/38
Sebaceous cell carcinoma
0/24
0/24
0/37
2/38
Forestomach
Squamous cell papilloma
0/52
7/52°
18/52°
17/52°
see
0/52
1/52
25/52°
35/52°
Jejunum
Adenocarcinoma
0/51
0/50
1/51
8/49°
Liver
Hepatocellular adenoma
0/52
3/52
15/52°
4/52
Hepatocellular carcinoma
0/52
1/52
23/52°
45/52°
Cholangiocarcinoma
0/52
0/52
0/52
1/52
Kidney
Cortical adenoma
0/52
0/52
7/52°
8/52°
Adenocarcinoma
0/52
0/52
2/52
0/52
Urothelial carcinoma
0/52
0/52
0/52
3/52
Auditory canal
Benign
0/1
0/0
1/7
0/33
Squamous cell papilloma
0/1
0/0
0/7
4/33
Carcinoma
0/1
0/0
2/7
19/33°
Sebaceous cell adenoma
0/1
0/0
0/7
1/33
95 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Table 4-5. Incidences of exposure-related neoplasms in Wistar rats treated
by gavage with benzo[a]pyrene, 5 days/week, for 104 weeks
Dose (mg/kg-d)
0
3
10
30b
Skin and mammary
Basal cell adenoma
2/52
0/52
1/52
10/51°
Basal cell carcinoma
1/52
1/52
0/52
4/51
see
0/52
1/52
1/52
5/51
Keratoacanthoma
1/52
0/52
1/52
4/51
T richoepithelioma
0/52
1/52
2/52
8/51°
Fibrosarcoma
0/52
3/52
5/52
0/51
Fibrous histiocytoma (malignant)
0/52
0/52
1/52
1/52
"Incidences are for number of rats with tumors compared with number of tissues examined histologically. Auditory
canal and oral cavity tissues were only examined histologically when abnormalities were observed upon
macroscopic examination.
bThis group had significantly decreased survival.
"Statistically significant difference (p < 0.01), Fisher's exact test; analysis of auditory canal tumor incidence was
based on assumption of n = 52 and no tumors in the controls.
Source: Kroese et al. (2001).
Kroese et al. (2001) did not systematically investigate nonneoplastic lesions detected in
rats sacrificed during the 2-year study, because the focus was to identify and quantitate tumor
occurrence. However, incidences were reported for nonneoplastic lesions in tissues or organs in
which tumors were detected (i.e., oral cavity, oesophagus, forestomach, jejunum, liver, kidney,
skin, mammary, and auditory canal). The reported nonneoplastic lesions associated with
exposure were the forestomach basal cell hyperplasia and clear cell foci of cellular alteration in
the liver. Incidences for forestomach basal cell hyperplasia in the control through high-dose
groups were: 1/52, 8/51, 13/51, and 2/52 for females and 2/50, 8/52, 8/52, and 0/52 in males.
Incidences for hepatic clear cell foci of cellular alteration were 22/52, 33/52, 4/52, and 2/52 for
females and 8/52, 22/52, 1/52, and 1/52 for males. These results indicate that the lowest dose
group, 3 mg/kg-day, was a LOAEL for increased incidence of forestomach hyperplasia and
hepatic histological changes in male and female Wistar rats exposed by gavage to
benzo[a]pyrene for up to 104 weeks. The lack of an increase in incidence of these nonneoplastic
lesions in the forestomach and liver at the intermediate and high doses (compared with controls)
may be associated with increased incidences of forestomach and liver tumors at these dose levels
(see Table 4-4).
As an adjunct study to the 2-year gavage study with Wistar rats, Kroese et al. (2001)
sacrificed additional rats (6/sex/group) after 4 and 5 months of exposure (0, 1,3, 10, or
30 mg/kg-day) for analysis of DNA adduct formation in WBCs and major organs and tissues.
Additional rats (6/sex/time period) were exposed to 0.1 mg/kg-day benzo[a]pyrene for 4 and
32
5 months for analysis of DNA adduct formation. Using the [ P]-postlabeling technique, five
96
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
benzo[a]pyrene-DNA adducts were identified in all of the examined tissues at 4 months (WBCs,
liver, kidney, heart, lung, skin, forestomach, glandular stomach, brain). Only one of these
adducts (adduct 2) was identified based on co-chromatography with a standard. This adduct,
identified as dG-N -BPDE, was the predominant adduct in all organs of female rats exposed to
10 mg/kg-day, except the liver and kidney, in which another adduct (unidentified adduct 4) was
predominant. Levels of total adducts (number of benzo[a]pyrene-DNA adducts per 1010
nucleotides) in examined tissues (from the single 10 mg/kg-day female rat) showed the following
order: liver > heart > kidney > lung > skin > forestomach ~ WBCs > brain. Mean values for
female levels of total benzo[a]pyrene-DNA adducts (number per 1010 nucleotides) in four organs
showed the same order, regardless of exposure group: liver > lung > forestomach ~ WBCs;
comparable data for males were not reported). Mean total benzo[a]pyrene-DNA adduct levels in
livers increased in both sexes from about 100 adducts per 1010 nucleotides at 0.1 mg/kg-day to
about 70,000 adducts per 1010 nucleotides at 30 mg/kg-day. In summary, these results suggest
that total benzo[a]pyrene-DNA adduct levels in tissues at 4 months were not independently
associated with the carcinogenic responses noted after 2 years of exposure to benzo[a]pyrene.
The liver showed the highest total DNA adduct levels and a carcinogenic response, but total
DNA adduct levels in heart, kidney, and lung (in which no carcinogenic responses were
detected) were higher than levels in forestomach and skin (in which carcinogenic responses were
detected).
Brune et al. (1981) 2-year rat study
Groups of Sprague-Dawley rats (32/sex/dose) were fed diets delivering a daily dose of
0.15 mg benzo[a]pyrene/kg body weight every 9th day or 5 times/week (Brune et al., 1981).
Other groups (32/ sex/dose) were given gavage doses of 0.15 mg benzo[a]pyrene (in aqueous
1.5% caffeine solution)/kg every 9th day, every 3rd day, or 5 times/week. The study included an
untreated control group (to compare with the dietary exposed groups) and a gavage vehicle
control group (each with 32 rats/sex). Rats were treated until moribundity or death occurred,
with average annual doses are reported in Table 4-2 (mg/kg-year, calculated by Brune et al.
[1981]). The following tissues were prepared for histopathological examination: tongue, larynx,
lung, heart, trachea, esophagus, stomach, small intestine, colon, rectum, spleen, liver, urinary
bladder, kidney, adrenal gland, and any tissues showing tumors or other gross changes. Survival
was similar among the groups, with the exception that the highest gavage-exposure group
showed a decreased median time of survival (Table 4-5). Increased incidences of portal-of-entry
tumors (forestomach, esophagus, and larynx) were observed in all of the gavage-exposed groups
and in the highest dietary exposure group (Table 4-5). Following dietary administration, all
observed tumors were papillomas. Following gavage administration, two malignant forestomach
tumors were found (one each in the mid- and high-dose groups) and the remaining tumors were
benign. The data in Table 4-5 show that the carcinogenic response to benzo[a]pyrene was
97
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
stronger with the gavage protocol compared with dietary exposure, and that no distinct difference
in response was apparent between the sexes. Tumors at distant sites (mammary gland, kidney,
pancreas, lung, urinary bladder, testes, hematopoietic, and soft tissue) were not considered
treatment-related as they were also observed at similar rates in the control group (data not
provided). The study report did not address noncancer systemic effects.
Table 4-6. Incidences of alimentary tract tumors in Sprague-Dawley rats
chronically exposed to benzo[a]pyrene in the diet or by gavage in caffeine
solution
Average annual
dose (mg/kg-yr)
Estimated average
daily dose"
(mg/kg-d)
Forestomach tumorsb
Total alimentary tract
tumors0 (larynx,
esophagus, forestomach)
Median
survival time
(wks)
benzo[a]pyrene by gavage in 1.5% caffeine solution
0
0
3/64 (4.7%)
6/64 (9.4%)
102
6
0.016
12/64 (18.8%)d
13/64 (20.3%)
112
18
0.049
26/64 (40.1%)e
26/64 (40.6%)
113
39
0.107
14/64 (21,9%)e
14/64 (21.9%)
87
benzo[a]pyrene in diet
0
0
2/64 (3.1%)
3/64 (4.7%)
129
6
0.016
1/64(1.6%)
3/64 (4.7%)
128
39
0.107
9/64 (14.1%)d
10/64(15.6%)
131
"Average annual dose divided by 365 d.
bNo sex-specific forestomach tumor incidence data were reported by Brune et al. (1981).
°Sex-specific incidences for total alimentary tract tumors were reported as follows:
Gavage (control - high dose): Male: 6/32, 7/32, 15/32, 8/32
Female: 0/32, 6/32, 11/32, 6/32
Diet (control - high dose): Male: 3/32, 3/32, 8/32
Female: 0/32, 0/32, 2/32
dSignificantly (p < 0.1) different from control using a modified %2test that accounted for group differences in
survival time.
"Significantly (p < 0.05) different from control using a modified %2 test that accounted for group differences in
survival time.
Source: Brune et al. (1981).
Beland and Culp (1998; Culp et al., 1998) 2-year mouse study
In the other modern cancer bioassay with benzo[a]pyrene, female B6C3Fi mice (48/dose
group) were administered benzo[a]pyrene (98.5% purity) at concentrations of 0 (acetone
vehicle), 5, 25, or 100 ppm in the diet for 2 years (Beland and Culp, 1998; Culp et al., 1998).
This study was designed to compare the carcinogenicity of coal tar mixtures with that of
benzo[a]pyrene and included groups of mice fed diets containing one of several concentrations
of two coal tar mixtures. Benzo[a]pyrene was dissolved in acetone before mixing with the feed.
Control mice received only acetone-treated feed. Female mice were chosen because they have a
lower background incidence of lung tumors than male B6C3Fi mice. Culp et al. (1998) reported
98
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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 the average daily intakes of benzo[a]pyrene in the 25- and 100-ppm groups were 104 and
430 [j,g/day, but did not report intakes for the 5-ppm group. Based on the assumption that daily
benzo[a]pyrene intake at 5 ppm was one-fifth of the 25-ppm intake (about 21 (j.g/day), average
daily doses for the three benzo[a]pyrene groups are estimated at 0.7, 3.3, and 16.5 mg/kg-day.
Estimated doses were calculated using time-weighted average (TWA) body weights of 0.032 kg
for the control, 5- and 25-ppm groups and 0.026 kg for the 100-ppm group (estimated from
graphically presented data). Food consumption, body weights, morbidity, and mortality were
monitored at intervals, and lung, kidneys, and liver were weighed at sacrifice. Necropsy was
performed on all mice that died during the experiment or survived to the end of the study period.
Limited histopathologic examinations (liver, lung, small intestine, stomach, tongue, esophagus)
were performed on all control and high-dose mice and on all mice that died during the
experimental period, regardless of treatment group. In addition, all gross lesions found in mice
of the low- and mid- dose groups were examined histopathologically.
None of the mice administered 100 ppm benzo[a]pyrene survived to the end of the study,
and morbidity/mortality was 100% by week 78. Decreased survival was also observed at 25 ppm
with only 27% survival at 104 weeks, compared with 56 and 60%, in the 5-ppm and control
groups, respectively. In the mid- and high-dose group, 60% of mice were alive at about 90 and
60 weeks, respectively. Early deaths in exposed mice were attributed to tumor formation rather
than other causes of systemic toxicity. Food consumption was not statistically different in
benzo[a]pyrene-exposed and control mice. Body weights of mice fed 100 ppm were similar to
those of the other treated and control groups up to week 46, and after approximately 52 weeks,
body weights were reduced in 100-ppm mice compared with controls. Body weights for the 5-
and 25-ppm groups were similar to controls throughout the treatment period. Compared with the
control group, no differences in liver, kidney, or lung weights were evident in any of the treated
groups (other organ weights were not measured).
Papillomas and/or carcinomas of the forestomach, esophagus, tongue, and larynx at
elevated incidences occurred in groups of mice exposed to 25 or 100 ppm, but no exposure-
related tumors occurred in the liver or lung (Table 4-6; Beland and Culp [1998]; Culp et al.
[1998]). The forestomach was the most sensitive tissue, and demonstrated the highest tumor
incidence among the examined tissues and was the only tissue with an elevated incidence of
tumors at 25 ppm (Table 4-6). In addition, most of the forestomach tumors in the exposed
groups were carcinomas, as 1, 31, and 45 mice had forestomach carcinomas in the 5-, 25-, and
100-ppm groups respectively. Nonneoplastic lesions were also found in the forestomach at
significantly (p < 0.05) elevated incidences: hyperplasia at >5 ppm and hyperkeratosis at >25
ppm (Table 4-6). The esophagus was the only other examined tissue showing elevated incidence
of a nonneoplastic lesion (basal cell hyperplasia, see Table 4-6). Tumors (papillomas and
carcinonas) were also significantly elevated in the esophagus and tongue at 100 ppm (Table 4-6).
Esophogeal carcinomas were detected in 1 mouse at 25 ppm and in 11 mice at 100 ppm. Tongue
99
DRAFT - DO NOT CITE OR QUOTE
-------
1 carcinomas were detected in seven 100-ppm mice; the remaining tongue tumors were
2 papillomas. Although incidences of tumors of the larynx were not significantly elevated in any
3 of the exposed groups, a significant dose-related trend was apparent (Table 4-6).
4
Table 4-7. Incidence of nonneoplastic and neoplastic lesions in female
B6C3Fi mice fed benzo[a]pyrene in the diet for up to 2 years
Incidence (%)
benzo[a]pyrene concentration (ppm) in diet"
0
5
25
100
Average daily doses (mg/kg-d)
Tissue and lesion
0
0.7
3.3
16.5
Liver (hepatocellular adenoma)
2/48
7/48
5/47
0/45
(2)
(15)
(11)
(0)
Lung (alveolar/bronchiolar adenoma and/or carcinoma)
5/48
0/48
4/45
0/48
(10)
(0)
(9)
(0)
Forestomach (papilloma and/or carcinoma)b
1/48
3/47
36/46b
46/47b
(2)
(6)
(78)
(98)
Forestomach (hyperplasia)13
13/48
23/47a
33/46b
37/47b
(27)
(49)
(72)
(79)
Forestomach (hyperkeratosis)15
13/48
22/47
33/46b
38/47b
(27)
(47)
(72)
(81)
Esophagus (papilloma and/or carcinoma)b
0/48
0/48
2/45
27/46b
(0)
(0)
(0)
(59)
Esophagus (basal cell hyperplasia)13
1/48
0/48
5/45
30/46b
(2)
(0)
(11)
(65)
Tongue (papilloma and/or carcinoma)b
0/49
0/48
2/46
23/48b
(0)
(0)
(4)
(48)
Larynx (papilloma and/or carcinoma)b
0/35
0/35
3/34
5/38
(0)
(0)
(9)
(13)
"Significant (p < 0.05) dose-related trend calculated for incidences of these lesions.
bSignificantly different from control incidence (p < 0.05); using a modified Bonferonni procedure for multiple
comparisons to the same control.
Source: Beland and Culp (1998); Culp et al. (1998).
5
6 Neal andRigdon (1967; Rigdon andNeal 1969, 1966) mouse study
7 Neal and Rigdon (1967) fed BAP (purity not reported) at concentrations of 0, 1, 10, 20,
8 30, 40, 45, 50, 100 and 250 ppm in the diets of male and female CFW-Swiss mice.
9 Corresponding doses (in mg/kg-day) were calculated1 as 0, 0.2, 1.8, 3.6, 5.3, 7.1, 8, 8.9, 17.8,
10 44.4 mg/kg-day. The age of the mice ranged from 17-180 days old and the treatment time from
11 1-197 days; the size of the treated groups ranged from 9 to 73. There were 289 mice (number of
1 Calculation: mg/kg-day = (ppm in feed x kg food/day)/kg body weight. Reference food consumption rates of
0.0062 kg/day (males) and 0.0056 kg/day (females) and reference body weights of 0.0356 kg (males) and 0.0305
(females) were used (U.S. EPA, 1988) and resulting doses were averaged between males and females.
100
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
mice/sex not stated) in the control group. No forestomach tumors were reported in the 0-, 0.2-
and 1.8 mg/kg-day dose groups. The incidence of forestomach tumors in the 20-, 30-, 40-, 45-,
50-, 100- and 250-ppm dose groups (3.6, 5.3, 7.1, 8, 8.9, 17.8, 44.4 mg/kg-day) were 1/23, 0/37,
1/40, 4/40, 23/34, 19/23 and 66/73, respectively.
Other oral exposure cancer bioassays in mice
Numerous other oral exposure cancer bioassays in mice have limitations which restrict
their usefulness for characterizing dose-response relationships between chronic-duration oral
exposure to benzo[a]pyrene and noncancer effects or cancer, but collectively, they provide strong
evidence that oral exposure to benzo[a]pyrene can cause portal-of-entry site tumors (see Table 4-
8 for references).
101
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Tumor incidence in oral exposure rodent cancer bioassays with limitations for describing dose-response
relationships for lifetime exposure to benzo[a]pyrene
Species/strain
Exposure
Results
Comments
Reference
Mouse/Hal CR
Groups of nine mice (9 wks old) were fed
benzo[a]pyrene in the diet (0, 0.2, or
0.3 mg/g diet) for 12 wks and sacrificed.
Estimated doses were 0, 27.3, or
41 mg/kg-d.
Incidence with forestomach tumors:
Control 0/9
Low 6/9
High 9/9
Less than lifetime exposure duration;
glandular stomach, lung, and livers from
control and exposed mice showed no
tumors.
Triolo et al., 1977
Mouse/HalCR
Groups of 12-20 mice (10 wks old) were
fed benzo[a]pyrene in the diet (0.1,0.3,
or 1.0 mg/g diet) for 12-20 wks.
Estimated doses were 14.3, 42.0, or
192 mg/kg-d.
Incidence with forestomach tumors:
Low 11/20 (18 wks)
Mid 13/19 (20 wks)
High 12/12 (12 wks)
Less than lifetime exposure duration; only
stomachs were examined for tumors;
tumors found only in forestomach.
Wattenberg, 1972
Mouse/CD-I
20 female mice (9 wks old) were given
1 mg benzo[a]pyrene by gavage 2
times/wk for 4 wks and observed for
19 wks. Estimated dose is 33 mg/kg-d,
using an average body weight of 0.030 kg
from reported data.
Incidence with forestomach tumors:
Exposed 17/20 (85%)
Controls 0/24
Less than lifetime exposure duration; only
stomach were examined for tumors; tumors
found only in forestomach.
El-Bayoumy, 1985
Mouse/BALB
25 mice (8 wks old) were given 0.5 mg
benzo[a]pyrene 2 times/wk for 15 wks.
5/25 mice had squamous carcinomas of
the forestomach; tumors were detected 28-
65 wks after treatment.
Less than lifetime exposure duration; the
following details were not reported:
inclusion of controls, methods for detecting
tumors, and body weight data.
Biancifiori et al.,
1967
Mouse/C3H
19 mice (about 3 mo old) were given
0.3 mL of 0.5%benzo[a]pyrene in
polyethylene glycol-400 by gavage,
once/d for 3 d.
By 30 wks, 7/10 mice had papillomas; no
carcinomas were evident
Less than lifetime exposure duration.
Berenblum and
Haran, 1955
Mouse/albino
Groups of 17-18 mice were given single
doses of benzo[a]pyrene and allowed to
survive until terminal sacrifice at 569 d.
Incidence of mice (that survived at least to
60 d) with forestomach papillomas:
Dose (|ig) Incidence
Experiment 1 Experiment 2
Control 0/17 0/18
12.5 3/17 2/18
50 0/17 1/17
200 8/17 a
Less than lifetime exposure duration; GI
tract examined for tumors with hand lens;
body weight data not reported.
Field and Roe, 1965
102
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Tumor incidence in oral exposure rodent cancer bioassays with limitations for describing dose-response
relationships for lifetime exposure to benzo[a]pyrene
Species/strain
Exposure
Results
Comments
Reference
Mouse/Hal CR
20 mice (9 wks old) were given
benzo[a]pyrene in the diet (0.3 mg
benzo[a]pyrene/g diet) for 6 wks and
sacrificed after 20 wks in the study.
8/20 exposed mice had forestomach
tumors.
Less than lifetime exposure duration; only
stomachs were examined for tumors;
tumors found only in forestomach; no
nonexposed controls were mentioned.
Wattenberg, 1974
Mouse/A/HeJ
12 female mice (9 wks old) were given
standard diet for 25 d, and 3 mg
benzo[a]pyrene by gastric intubation on d
7 and 21 of the study. Mice were killed
at 31 wks of age and examined for lung
tumors.
12/12 exposed mice had lung tumors
Less than lifetime exposure duration; only
lungs examined for tumors; no nonexposed
controls were mentioned.
Wattenberg, 1974
Mouse/white
Groups of 16-30 mice were given
benzo[a]pyrene in Methylene glycol
(0.001-10 mg) wkly for 10 wks and
observed until 19 mo.
Tumors in stomach antrum
Dose (mg) - Carcinoma - Papilloma
0.001 0/16 0/16
0.01- 0/26 2/26
0.1 0/24 5/24
1.0 11/30 12/30
10 16/27 7/27
Less than lifetime exposure duration.
Fedorenko et al.,
1967,as cited in U.S.
EPA, 1991a
Mouse/albino
Groups of about 160 female mice (70 d
of age; strain unknown) were given 0 or
8 mg benzo[a]pyrene mixed in the diet
over a period of 14 mo.
Gastric tumors were observed at the
following incidence:
Control 0/158
8 mg benzo[a]pyrene total 13/160
Close to lifetime exposure duration; daily
dose levels and methods of detecting
tumors were not clearly reported.
Chouroulinkov et al.,
1967
Mouse/A/J
Groups 40 female mice (8 wks old) were
given 0 or 0.25 mg benzo[a]pyrene (in
2% emulphor) by gavage 3 times/wk for
8 wks. Mice were killed at 9 mo of age
and examined for lung or forestomach
tumors.
Incidence for mice surviving at 9 mo of
age:
Lung tumors
Control 11/38
Exposed 22/36
Forestomach tumors
Control 0/38
Exposed 33/36
Less than lifetime duration of exposure;
only lungs and GI tract were examined for
tumors.
Robinson etal., 1987
Mouse/Swiss
albino
Groups of mice (9-14 wks old) were
given single doses of 0 or 0.05 mg
benzo[a]pyrene in polyethylene glycol-
400 by gavage. Surviving mice were
killed at 18 mo of age and examined for
macroscopic tumors.
Forestomach tumor incidence:
Dose (|ig) - Carcinoma - Papilloma
0 0/65 2/65
50 1/61 20/61
Less than lifetime duration of exposure;
exposure-related tumors only found in
forestomach.
Roe et al., 1970
103
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Tumor incidence in oral exposure rodent cancer bioassays with limitations for describing dose-response
relationships for lifetime exposure to benzo[a]pyrene
Species/strain
Exposure
Results
Comments
Reference
Mouse/ICR
Groups of 20 or 24 mice (71 d old) were
given 1.5 mgbenzo[a]pyrene by gavage
2 times/wk for 4 wks; terminal sacrifice
was at 211 d of age. Estimated dose was
about 50 mg benzo[a]pyrene/kg, using an
average body weight of 0.03 kg during
exposure from reported data.
Incidence of mice with forestomach
neoplasms.
Experiment 1 23/24
Experiment 2 19/20
Less than lifetime duration of exposure;
only stomachs were examined for tumors;
tumors found only in forestomach;
nonexposed controls were not mentioned.
Benjamin etal., 1988
Mouse/CFW
Groups of mice (mixed sex) were fed
benzo[a]pyrene in the diet (dissolved in
benzene and mixed with diet) at 0, 1, 10,
20, 30, 40, 45, 50, 100, or 250 ppm in the
diet.
ppm
Exposure
Forestomach tumor
Less than lifetime exposure duration; no
(d)
incidence
vehicle control group; animals ranged from
1
110
0/25
three wks to 6 mo old at the start of dosing;
10
110
0/24
only alimentary tract was examined for
20
110
1/23
tumors; (see also Rigdon and Neal, 1969,
30
110
0/37
1967, 1966).
40
110
1/40
45
110
4/40
50
152
24/34
100
110
19/23
250
118
66/73
Neil and Rigdon,
1967
Rat/ Sprague-
Dawley
Groups of Sprague-Dawley rats
(32/sex/dose) were fed diets delivering a
daily dose of 0.15 mgbenzo[a]pyrene/kg
body weight every 9th day or
5 times/week (Brune et al., 1981). Other
groups (32/ sex/dose) were given gavage
doses of 0.15 mg benzo[a]pyrene (in
aqueous 1.5% caffeine solution)/kg every
9 day, every 3rd day, or 5 times/week.
Dose larynx, esopogus, and forestomach
Doses are annual averages. Non-standard
(gavage)
tumors
treatment protocol involved animals being
0
6/64
treated for 5 days a week or fewer;
0.016
13/64
relatively high control incidence compared
0.049
26/64
to other gavage studies;
0.107
14/64
(diet)
0
3/64
0.016
3/64
0.107
10/64
Brune et al., 1981
Mouse/A/J
Groups of female mice were fed
benzo[a]pyrene in the diet at 0, 16, or 98
ppm for 260 d. Average intakes of
benzo[a]pyrene were 0, 40.6, and
256.6 |ig/mousc/d. Estimated doses are
0, 1.6, and 9.9 mg/kg-d using a chronic
reference body weight value of 0.026 kg
(U.S. EPA, 1988).
Incidence of mice surviving to 260 d:
Lung tumors
Control 4/21
16 ppm 9/25
98 ppm 14/27
Forestomach tumors
Control 0/21
16 ppm 5/25
98 ppm 27/27
Close to lifetime exposure duration; A/J
strain of mice particularly sensitive to
chemically induced cancer; only lungs and
stomachs were examined for tumors.
Weyandetal., 1995
104
DRAFT - DO NOT CITE OR QUOTE
-------
not evaluated
105
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
4.2.2. Inhalation
4.2.2.1. Short-term and Subchronic Studies
Wolff et al. (1989) exposed groups of 40 male and 40 female F344/Crl rats, via nose
"3
only, to 7.5 mg benzo[a]pyrene/m for 2 hours/day, 5 days/week for 4 weeks (corresponding to a
TWA of 0.45 mg/m ). Rats were 10-11 weeks old at the beginning of the experiment.
Benzo[a]pyrene (>98% pure) aerosols were formed by heating and then condensing the
vaporized benzo[a]pyrene. The particle MMAD was 0.21 [j,m. Subgroups of these animals
(six/sex/dose) were exposed for 4 days or 6 months after the end of the 4-week exposure,
respectively, to radiolabeled aluminosilicate particles. Lung injury was assessed by analyzing
clearance of radiolabeled aluminosilicate particles and via histopathologic evaluations. Body
and lung weights, measured in subgroups from 1 day to 12 months after the exposure did not
differ between controls and treated animals. Radiolabeled particle clearance did not differ
between the control and treated groups, and there were no significant lung lesions. This study
identified a NOAEL for lung effects of 0.45 mg/m -day for a short-term exposure.
4.2.2.2. Chronic Studies and Cancer Bioassays
Thyssen et al. (1981) conducted an inhalation study in which male Syrian golden
hamsters were exposed to benzo[a]pyrene for their natural lifetime. Groups of 20-30 male
Syrian golden hamsters (8 weeks old) were exposed by nose-only inhalation to NaCl aerosols
"3
(controls; 240 jag NaCl/m ) or benzo[a]pyrene condensed onto NaCl aerosols at three nominal
-3
concentrations of 2, 10, or 50 mg benzo[a]pyrene/m for 3-4.5 hours/day, 5 days/week for 1-
41 weeks, followed by 3 hours/day, 7 days/week for the remainder of study (until hamsters died
or became moribund). Thyssen et al. (1981) reported average measured benzo[a]pyrene
"3
concentrations to be 0, 2.2, 9.5, or 46.5 mg/m . More than 99% of the particles were between 0.2
and 0.5 [j.m in diameter, and over 80% had diameters between 0.2 and 0.3 [j,m. The particle
analysis of the aerosols was not reported to modern standards (MMAD and geometric SD were
not reported). Each group initially consisted of 24 hamsters; final group sizes were larger as
animals dying during the first 12 months of the study were replaced.
Survival was similar in the control, low-dose, and mid-dose groups, but was significantly
decreased in the high-dose group. Average survival times in the control, low-, mid-, and high-
dose groups were 96.4 ± 27.6, 95.2 ± 29.1, 96.4 ± 27.8, and 59.5 ± 15.2 weeks, respectively.
After the 60th week, body weights decreased and mortality increased steeply in the highest dose
group. Histologic examination of organs (a complete list of organs examined histologically was
not reported by Thyssen et al. [1981]) revealed a dose-related increase in tumors in the upper
respiratory tract, including the nasal cavity, pharynx, larynx, and trachea and in the digestive
tract in the mid- and high-dose groups (Table 4-8). A statistical analysis was not included in the
Thyssen et al. (1981) report. No lung tumors were observed. Squamous cell tumors in the
esophagus and forestomach were also observed in the high-dose group, presumably as a
106
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
consequence of mucociliary particle clearance. Tumors were detected in other sites, but none of
these appeared to be related to exposure. The results indicated that the pharynx and larynx,
including the epiglottis, were the main cancer targets (Table 4-8).
Table 4-9. Incidence of respiratory and upper digestive tract tumors in male
hamsters treated for life with benzo[a]pyrene by inhalation
Reported benzo[a]pyrene concentration (mg/m3)
0a
2b
10
50
Tumor site
Tumor incidence (latency in wksc)
Nasal cavity
0
0
3/26 (116 ± 1.5)
1/25 (79)
Larynx
0
0
8/26(107.1 ± 15.5)
13/25 (67.6 ± 12.1)
Trachea
0
0
1/26(115)
3/25 (63.3 ±33.3)
Lung
0
0
0
0
Pharynx
0
0
6/26 (97.2 ± 16.9)
14/25 (67.5 ± 12.2)
Esophagus
0
0
0
2/25 (70, 79)
Forestomach
0
0
1/26(119)
1/25 (72)
"Effective number of animals in control group: n = 27.
bEffective number of animals in 2 mg/m3 dose group: n = 27.
°Mean ± SD.
Source: Thyssen et al. (1981).
Under contract to the U.S. EPA, Clement Associates (1990) obtained the individual
animal data (including individual animal pathology reports, time-to-death data, and exposure
chamber monitoring data) collected by Thyssen et al. (1981). Re-analysis of the original data
revealed several errors and omissions in the published report. The actual exposure protocol was
as follows: 4.5 hours/day 5 days/week on weeks 1-12, 3 hours/day 5 days/week on weeks 13-
29, 3.7 hours/day 5 days/week on week 30, 3 hours/day 5 days/week on weeks 31-41, and
3 hours/day 7 days/week for the reminder of the experiment. In addition, actual exposure
concentrations varied widely from week to week. Because different animals were started at
different times, each individual animal had an exposure history somewhat different than others in
the same exposure group. In order to deal with this problem, Clement Associates (1990) used
the orignial individual animal data to calculate average continuous lifetime exposures for each
individual hamster. Group averages of individual average continuous lifetime exposure
-3
concentrations were 0, 0.25, 1.01, and 4.29 mg/m for the control through high-exposure groups.
For this assessment, the individual animal pathology reports prepared by Thyssen et al.
(1981) and obtained by Clement Associates (1990) were examined to independently assess the
numbers of hamsters with tumors in the larynx, pharynx, and nose in each group. Table 4-9
presents the number of animals with tumors in the larynx and pharynx and the numbers of
animals in each exposure group. Numbers of animals with either laryngeal or pharyngeal tumors
107
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
are also noted in Table 4-10, since these two types of tumors arise in close anatomical proximity
from similar cell types. Examination of the individual animal pathology reports also showed that
all of the nasal, forestomach, esophageal, and tracheal tumors occurred in animals that also had
either laryngeal or pharyngeal tumors, except for two animals in the mid-dose group that
displayed nasal tumors (one malignant and one benign) without displaying tumors in the pharynx
or larynx.
Table 4-10. Number of animals with pharynx and larynx tumors in male
hamsters exposed by inhalation to benzo[a]pyrene for life
Average
continuous
benzo[a]pyrene
concentration"
(mg/m3)
Number of
hamsters in
groupb
Larynxb
Pharynxb
Larynx or pharynx,
combined0
Malignant
All
Malignant
All
Malignant
All
Control
27
0
0
0
0
0
0
0.25
27
0
0
0
0
0
0
1.01
26
8
11
7
9
11
16
4.29
34
9
12
17
18
17
18
aAs calculated by Clement Associates (1990) from air monitoring data collected by Thyssen and colleagues.
bAs counted from information in Table D-l in Appendix D, which was obtained from examination of individual
animal pathology reports prepared by Thyssen and colleagues and obtained by Clement Associates.
°As counted from information in Table D-l in Appendix D. Nasal, forestomach, esophageal, and tracheal tumors
occurred in hamsters that also had tumors in the larynx or pharynx, except for two animals in the mid-dose group
that displayed nasal tumors (one malignant and one benign) without displaying tumors in the pharynx or larynx.
Several studies have investigated the carcinogenicity of benzo[a]pyrene in hamsters
exposed by intratracheal instillation. Single-dose studies verified that benzo[a]pyrene is
tumorigenic, but do not provide data useful characterizing dose-response relationships because of
their design (Kobayashi, 1975; Reznik-Schuller and Mohr, 1974; Henry et al., 1973; Mohr, 1971;
Saffiotti et al., 1968; Gross et al., 1965; Herrold and Dunham, 1962). One multiple-dose study,
which utilized very low doses (0.005, 0.02, and 0.04 mg, once every 2 weeks), failed to find any
tumorigenic response (Kunstler, 1983). Tumorigenic responses (mostly in the respiratory tract)
were found at higher dosage levels (0.25-2 mg benzo[a]pyrene once per week for 30-52 weeks)
in four multiple-dose studies (Feron and Kruysse, 1978; Ketkar et al., 1978; Feron et al., 1973;
Saffiotti et al., 1972). These studies identify the respiratory tract as a cancer target with exposure
to benzo[a]pyrene by intratracheal instillation and provide supporting evidence for the
carcinogenicity of benzo[a]pyrene at portal-of-entry sites.
4.2.3. Dermal Exposure
4.2.3.1. Skin-Tumor Initiation-Promotion Assays
108
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Results from numerous studies indicate that acute dermal exposure to benzo[a]pyrene
induces skin tumors in mice when followed by repeated exposure to a potent tumor promoter
(Weyand et al., 1992; Cavalieri et al., 1991, 1981; Rice et al., 1985; El-Bayoumy et al., 1982;
LaVoie et al., 1982; Raveh et al., 1982; Slaga et al., 1980, 1978; Wood et al., 1980; Hoffmann et
al., 1972). The typical exposure protocol in these studies involved the application of a single
dose of benzo[a]pyrene (typically >20 nmol per mouse) to dorsal skin of mice followed by
repeated exposure to a potent tumor promoter, such as 12-O-tetradecanoylphorbol-13-acetate
(TP A).
4.2.3.2. Carcinogenicity Dermal Bioassays
Poel (1959)
Poel (1959) applied benzo[a]pyrene in toluene to shaved interscapular skin of groups of
13-56 male C57L mice at doses of 0, 0.15, 0.38, 0.75, 3.8, 19, 94, 188, 376, or 752 jag,
3 times/week for up to 103 weeks or until the appearance of a tumor by gross examination
(3 times weekly). Some organs (not further specified) and interscapular skin in sacrificed mice
were examined histologically. With increasing dose level, the incidence of mice with skin
tumors increased and the time of tumor appearance decreased (see Table 4-10). Doses >3.8 jag
were associated with 100% mortality after increasingly shorter exposure periods, none greater
than 44 weeks. Poel (1959) did not mention the appearance of exposure-related tumors in tissues
other than interscapular skin.
Table 4-11. Skin tumor incidence and time of appearance in male C57L
mice dermally exposed to benzo[a]pyrene for up to 103 weeks
Dose (jig)a
Incidence of mice with
gross skin tumors
Time of first tumor
appearance (wks)
Incidence of mice
with epidermoid
carcinomab
Length of exposure
period(wks)
0 (Toluene)
0/33 (0%)
-
0/33 (0%)
92
0.15
5/55 (9%)
42-44°
0/55 (0%)
98
0.38
11/55 (20%)
24
2/55 (4%)
103
0.75
7/56(13%)
36
4/56 (7%)
94
3.8
41/49 (84%)
21-25
32/49 (65%)
82
19
38/38 (100%)
11-21
37/38 (97%)
25-44°
94
35/35 (100%)
8-19
35/35 (100%)
22-43
188
12/14 (86%)
9-18
10/14(71%)
20-35
376
14/14 (100%)
4-15
12/14 (86%)
19-35
752
13/13 (100%)
5-13
13/13 (100%)
19-30
indicated doses were applied to interscapular skin 3 times/wk for up to 103 wks or until time of appearance of a
grossly detected skin tumor.
bCarcinomas were histologically confirmed.
109
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
°Ranges reflect differing information in Tables 4 and 6 of Poel (1959).
Source: Poel (1959).
Poel (1960) applied benzo[a]pyrene in a toluene vehicle to shaved interscapular skin of
groups of 14-25 male SWR, C3HeB, or A/He mice 3 times/week at doses of 0, 0.15, 0.38, 0.75,
3.8, 19.0, 94.0, or 470 jag benzo[a]pyrene per application, until mice died or a skin tumor was
observed. Time ranges for tumor observations were provided, but not times of death for mice
without tumors, so it was not possible to evaluate differential mortality among all dose groups or
the length of exposure for mice without tumors. With increasing dose level, the incidence of
mice with skin tumors increased and the time of tumor appearance decreased (Table 4-11). The
lowest dose level did not induce an increased incidence of mice with skin tumors in any strain,
but strain differences in susceptibility were evident at higher dose levels. SWR and C3HeB mice
showed skin tumors at doses >0.38 jag benzo[a]pyrene, whereas AH/e mice showed tumors at
doses >19 [j,g benzo[a]pyrene (Table 4-11). Except for metastases of the skin tumors to lymph
nodes and lung, Poel (1960) did not mention the appearance of exposure-related tumors in
tissues other than interscapular skin.
Table 4-12. Skin tumor incidence and time of appearance in male SWR,
C3HeB, and A/He mice dermally exposed to benzo[a]pyrene for life or until
a skin tumor was detected
Dose (jtg)a
SWR mice
C3HeB mice
A/He mice
Tumor
incidenceb
Time of first
tumor
appearance
(weeks)
Tumor
incidenceb
Time of first
tumor
appearance
(weeks)
Tumor
incidenceb
Time of fist
tumor
appearance
(weeks)
0 (Toluene)
0/20 (0%)
0/17 (0%)
—
0/17 (0%)
—
0.15
0/25 (0%)
—
0/19 (0%)
—
0/18(0%)
—
0.38
2/22 (9%)
55-55
3/17 (18%)
81-93
0/19(0%)
—
0.75
15/18(83%)
25-72
4/17 (24%)
51-93
0/17 (0%)
—
3.8
12/17 (70%)
25-51
11/18(61%)
(35-73
0/17 (0%)
—
19.0
16/16 (100%)
12-28
17/17 (100%)
13-32
21/23 (91%)
21-40
94.0
16/17 (94%)
9-17
18/18(100%)
10-22
11/16(69%)
14-31
470.0
14/14 (100%)
5-11
17/17 (100%)
4-19
17/17 (100%)
4-21
"Indicated doses were applied 3times/week for life or until a skin tumor was detected. Mice were 10-14 wks old at
initial exposure.
incidence of mice exposed 10 or more wks with a skin tumor.
Source: Poel (1960).
110
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Roe et al. (1970) treated groups of 50 female Swiss mice with 0 (acetone vehicle), 0.1,
0.3, 1, 3, or 9 jag benzo[a]pyrene applied to the shaved dorsal skin 3 times/week for up to
93 weeks; all surviving mice were killed and examined for tumors during the following 3 weeks.
The dorsal skin of an additional control group was shaved periodically but was not treated with
the vehicle. Mice were examined every 2 weeks for the development of skin tumors at the site of
application. Histologic examinations included: (1) all skin tumors thought to be possibly
malignant; (2) lesions of other tissues thought to be neoplastic; and (3) limited nonneoplastic
lesions in other tissues. As shown in Table 4-13, markedly elevated incidences of mice with skin
tumors were only found in the two highest dose groups (3 or 9 jag ), compared with no skin
tumors in the control groups. Malignant skin tumors (defined as tumors with invasion or
penetration of the pannicuius carnosus muscle) were detected in 4/41 and 31/40 mice in the
3- and 9-[ig groups, respectively, surviving to at least 300 days. Malignant lymphomas were
detected in all groups, but the numbers of cases were not elevated compared with expected
numbers after adjustment for survival differences. Lung tumors were likewise detected in
control and exposed groups at incidences that were not statistically different.
Table 4-13. Tumor incidence in female Swiss mice dermally exposed to
benzo[a]pyrene for up to 93 weeks
Dose (jug)1'
Cumulative number of mice with skin
tumor/survivors
Skin tumor
incidenceb
Malignant
Lymphoma
incidence0
Lung
Tumor
incidence0
200 d
300 d
400 d
500 d
600 d
700 d
No treatment
0/48
0/43
0/40
0/31
0/21
0/0
0/43 (0%)
19/44 (43%)
12/41 (29%)
Acetone
0/49
0/47
0/45
0/37
0/23
0/0
0/47 (0%)
12/47 (26%)
10/46 (22%)
0.1
0/45
1/42
1/35
1/31
1/22
1/0
1/42 (2%)
11/43 (26%)
10/40 (25%)
0.3
0/46
0/42
0/37
0/30
0/19
0/0
0/42 (0%)
10/43 (23%)
13/43 (30%)
1
0/48
0/43
0/37
1/30
1/18
1/0
1/43 (2%)
16/44 (36%)
15/43 (35%)
3
0/47
0/41
1/37
7/35
8/24
8/0
8/41 (20%)
23/42 (55%)
12/40 (30%)
9
0/46
4/40
21/32
28/21
33/8
34/0
34/46 (74%)
9/40 (23%)
5/40 (13%)
"Doses were applied 3 times/wk for up to 93 wks to shaved dorsal skin.
bNumerator: number of mice detected with a skin tumor. Denominator: number of mice surviving to 300 d for all
groups except the highest dose group. For the highest dose group (in which skin tumors were first detected between
200 and 300 d), the number of mice surviving to 200 d was used as the denominator.
0 Numerator : number of mice detected with specified tumor. Denominator: number of mice surviving to 300 d
unless a tumor was detected earlier, in which case the number dying before 300 d without a tumor was subtracted
from the number of animals reported to have been examined.
Source: Roe et al. (1970).
Schmidt et al. (1973) dermally administered benzo[a]pyrene in acetone to female NMRI
mice (100/group) and female Swiss mice. Benzo[a]pyrene was applied to the shaved dorsal skin
twice weekly with doses of 0, 0.05, 0.2, 0.8, or 2 jag until spontaneous death occurred or until an
advanced carcinoma was observed. Skin carcinomas were identified by the presence of crater-
111
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
shaped ulcerations, infiltrative growth and the beginning of physical wasting (i.e., cachexia).
Necropsy was performed for all animals, and histopathological examination of the dermal site of
application and any other tissues with gross abnormalities was conducted. Skin tumors were
observed at the two highest doses in both strains of female mice (see Table 4-14), with induction
periods of 53.0 and 75.8 weeks for the 0.8 and 2.0 jag NMRI mice and 57.8 and 60.7 weeks for
the Swiss mice, respectively. The authors indicated that the latency period for tumor formation
was highly variable and significant differences among exposure groups could not be identified,
but no further timing information was available, including overall survival. Carcinoma was the
primary tumor type seen after lifetime application of benzo[a]pyrene to mouse skin.
Table 4-14. Skin tumor incidence in female NMRI and Swiss mice dermally
exposed to benzo[a]pyrene
Dose (jtg)a b
Skin tumor incidence (all types)
Incidence of papilloma
Incidence of carcinoma
Female NMRI mice
0 (Acetone)
0/100 (0%)
0/100 (0%)
0/100 (0%)
0.05
0/100 (0%)
0/100 (0%)
0/100 (0%)
0.2
0/100 (0%)
0/100 (0%)
0/100 (0%)
0.8
2/100 (2%)
0/100 (0%)
2/100 (2%)
2
30/100 (30%)
2/100 (2%)
28/100 (28%)
Female Swiss mice
0 (Acetone)
0/80 (0%)
0/80 (0%)
0/80 (0%)
0.05
0/80 (0%)
0/80 (0%)
0/80 (0%)
0.2
0/80 (0%)
0/80 (0%)
0/80 (0%)
0.8
5/80 (6%)
0/80 (0%)
5/80 (6%)
2
45/80 (56%)
3/80 (4%)
42/80 (52%)
"Mice were exposed until natural death or until they developed a carcinoma a
indicated doses were applied 2 times/wk to shaved skin of the back.
Source: Schmidt et al. (1973).
the site of application.
Schmahl et al. (1977) applied benzo[a]pyrene 2 times/week to the shaved dorsal skin of
female NMRI mice (100/group) at doses of 0, 1, 1.7, or 3 jag in 20 [jL acetone. The authors
reported that animals were observed until natural death or until they developed a carcinoma at
the site of application. The effective numbers of animals at risk was about 80% of the nominal
group sizes, which the authors attributed to autolyis; no information was provided concerning
when tumors appeared in the relevant groups, how long treatment lasted in each group, or any
times of death. Necropsy was performed on all mice and the skin of the back, as well as any
organs that exhibited macroscopic changes, were examined histopathologically. The incidence
of all types of skin tumors was increased in a dose related manner compared to controls (see
Table 4-15). Carcinoma was the primary tumor type observed following chronic dermal
112
DRAFT - DO NOT CITE OR QUOTE
-------
1 exposure to benzo[a]pyrene, and skin papillomas occurred infrequently. Dermal sarcoma was not
2 observed.
3
113
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Table 4-15. Skin tumor incidence in female NMRI mice dermally exposed to
benzo[a]pyrene
Dose (jig)a'b
Skin tumor incidence
(all types)
Incidence of papilloma
Incidence of carcinoma
0
1/81 (1%)
0/81 (0%)
0/81 (0%)
1
11/77 (14%)
1/77 (1%)
10/77 (13%)
1.7
25/88 (28%)
0/88 (0%)
25/88 (28%)
3
45/81 (56%)
2/81 (3%)
43/81 (53%)
"Mice were exposed until natural death or until they developed a carcinoma at the site of application,
indicated doses were applied 2 times/wk to shaved skin of the back.
Source: Schmahl et al. (1977).
Habs et al. (1980) applied benzo[a]pyrene to the shaved interscapular skin of female
NMRI mice (40/group) at doses of 0, 1.7, 2.8 or 4.6 jag in 20 [xL acetone twice weekly, from
10 weeks of age until natural death or gross observation of infiltrative tumor growth. Latency of
tumors, either as time of first appearance or average time of appearance of tumors, was not
reported. Necropsy was performed on all animals, and the dorsal skin, as well as any organs
showing gross alterations at autopsy, was prepared for histopathological examination. Age-
standardized mortality rates, using the total population of the experiment as the standard
population, were used to adjust tumor incidence findings in the study. Benzo[a]pyrene
application was associated with a statistically significant increase in the incidence of skin tumors
at each dose level (see Table 4-17).
Table 4-16. Skin tumor incidence in female NMRI mice dermally exposed to
benzo[a]pyrene
Dose (jig)a'b
Skin tumor incidence
Age-standardized tumor incidence0
0 (acetone)
0/35 (0%)
0%
1.7
8/34 (24%)
24.8%
2.8
24/35 (68%)
89.3%
4.6
22/36 (61%)
91.7%
aMice were exposed until natural death or until they developed a carcinoma at the site of application,
indicated doses were applied 2 times/wk to shaved skin of the back.
°Mortality data of the total study population were used to derive the age-standardized tumor incidence.
Source: Habs et al. (1980).
Grimmer et al. (1983) and Grimmer et al. (1984) applied benzo[a]pyrene (in 0.1 mL of a
1:3 solution of acetone:dimethyl sulfoxide [DMSO]) to the interscapular skin of female CFLP
mice (65-80/group) 2 times/week for 104 weeks. Doses were 0, 3.9, 7.7 and 15.4 jag in the 1983
114
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
experiment, and 0, 3.4, 6.7, and 13.5 jag in the 1984 experiment. Mice were observed until
spontaneous death, unless an advanced tumor was observed or if animals were found moribund.
Survival information was not provided; incidences reflect the number of animals placed on
study. Necropsy was performed on all mice. Histopathological examination of the skin and any
other organ showing gross abnormalities was performed. Chronic dermal exposure to
benzo[a]pyrene produced a dose-related increase in skin tumor incidence and a decrease in tumor
latency (see Table 4-17). Carcinoma was the primary tumor type observed and a dose-response
relationship was evident for carcinoma formation and incidence of all types of skin tumors.
Table 4-17. Skin tumor incidence and time of appearance in female CFLP
mice dermally exposed to benzo[a]pyrene for 104 weeks
Dose (jtg)a
Skin tumor incidence
(all types)
Incidence of
papilloma
Incidence of
carcinoma
Tumor appearance in
weeks
Grimmer et al. (1983)
0 (1:3 Solution of
acetone :DMSO)
0/80 (0%)
0/80 (0%)
0/80 (0%)
—
3.9
22/65 (34%)
7/65 (11%)
15/65 (23%)
74.6 ± 16.78b
7.7
39/64 (61%)
5/64 (8%)
34/64 (53%)
60.9 ± 13.90
15.4
56/64 (88%)
2/64 (3%)
54/64 (84%)
44.1 ±7.66
Grimmer et al. (1984).
0 (1:3 Solution of
acetone :DMSO)
0/65 (0%)
0/65 (0%)
0/65 (0%)
—
3.4
43/64 (67%)
6/64 (9%)
37/64 (58%)
61 (53-65)°
6.7
53/65 (82%)
8/65 (12%)
45/65 (69%)
47 (43-50)
13.5
57/65 (88%)
4/65 (6%)
53/65 (82%)
35 (32-36)
"Indicated doses were applied twice/week to shaved skin of the back.
b Mean ± SD.
0 Median with 95% confidence interval.
Habs et al. (1984) applied benzo[a]pyrene (in 0.01 mL acetone) to the shaved
interscapular skin of female NMRI mice at doses of 0, 2 or 4 jag, 2 times/week for life. Animals
were observed twice daily until spontaneous death, unless an invasive tumor was observed. All
animals were necropsied and histopathological examination was performed on the dorsal skin
and any other organ with gross abnormalities. Chronic dermal exposure to benzo[a]pyrene did
not affect body weight gain, but appeared to reduce survival at the highest dose with mean
survival times of 691, 648, and 528 days for the 0, 2, and 4 (J,g/day groups, respectively. The
total length of exposure for each group was not reported, but can be inferred from the survival
data. Latency also was not reported. Benzo[a]pyrene application resulted in a dose-related
increase the incidence of total skin tumors and skin carcinomas (see Table 4-18). Hematopoietic
tumors (at 6/20, 3/20, and 3/20) and lung adenomas (at 2/20, 1/20, and 0/20) were observed in
115
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
the controls and in the benzo[a]pyrene treatment groups, but did not appear to be treatment
related according to the study authors.
Table 4-18. Skin tumor incidence in female NMRI mice dermally exposed to
benzo[a]pyrene for life
Dose (jig)a'b
Skin tumor incidence
(all types)
Incidence of
papilloma
Incidence of
carcinoma
Mean survival time,
days (95%
confidence interval)
0 (Acetone)
0/20 (0%)
0/20 (0%)
0/20 (0%)
691 (600-763)
2
9/20 (45%)
2/20 (10%)
7/20 (35%)
648 (440-729)
4
17/20 (85%)
0/20 (0%)
17/20 (85%)
528 (480-555)
aMice were exposed until natural death or until they developed an invasive tumor at the site of application,
indicated doses were applied 2 times/wk to shaved interscapular skin.
Source: Habs et al. (1984).
Higginbotham et al. (1993)
Groups of 23-27 female Ah-receptor-responsive Swiss mice were treated on a shaved
area of dorsal skin with 0, 1, 4, or 8 nmol (0, 0.25, 1, or 2 (j,g/treatment) benzo[a]pyrene (>99%
pure) in acetone 2 times weekly for 40 weeks (Higginbotham et al., 1993). Surviving animals
were sacrificed 8 weeks later. Complete necropsies were performed, and tissues from the treated
area, lung, liver, kidney, spleen, urinary bladder, ovary, and uterus were harvested for
histopathologic examination. Histopathologic examination was performed on tissues from the
treated area, lungs, liver, kidneys, spleen, urinary bladder, uterus, and ovaries, as well as any
other grossly abnormal tissue. Lung adenomas occurred in each group (1/27, 2/24, 1/23, 1/23),
and other tumors were noted in isolated mice (i.e., malignant lymphoma (spleen) in one low-dose
and one mid-dose mouse; malignant lymphoma with middle organ involvement in one high-dose
mouse, and hemangioma (liver) in one mid-dose mouse) and were not considered dose related.
In addition, benzo[a]pyrene showed no skin tumors under the conditions of this bioassay.
Sivaketal. (1997)
Sivak et al. (1997) designed a study to compare the carcinogenicity of condensed asphalt
fumes (including benzo[a]pyrene and other PAHs) with several doses of benzo[a]pyrene alone.
For the purposes of this assessment, the exposure groups exposed to PAH mixtures are not
discussed. Groups of 30 male C3H /HeJ mice were treated dermally twice/week to 0, 0.0001,
0.001, or 0.01% (0, 0.05, 0.5, or 5 jag) benzo[a]pyrene in a 50 |il volume of
cyclohexanone/acetone (1:1) for 104 weeks beginning at 8 weeks of age. All mice were
necropsied, and skin samples from all as well as any grossly observed lesions were subjected to
histopathological examination. The incidence of skin tumors and mean survival times for each
116
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
group are shown in Table 4-19. All high dose mice died before the final sacrifice. The extent of
deaths prior to one year in each group was not provided, so that the reported incidence may
underestimate the tumor rate of animals exposed long enough to develop tumors. However, the
crude skin tumor rates show an increasing trend in incidence.
Table 4-19. Skin tumor incidence in male C3H /HeJ mice dermally exposed
to benzo[a]pyrene for 24 months
Skin tumor incidence
No. died before final
Mean survival time,
Dose (ng)a
(all types)
sacrifice
days
0 cyclohexanone/acetone (1:1)
0/30 (0%)
19
607
0.05
0/30 (0%)
15
630
0.5
5/30 (20%)
15
666
5.0
27/30 (90%)
30
449
"Indicated doses were applied twice/week to shaved dorsal skin.
Source: Sivak et al. (1997).
Albert et al. (1991)
To examine dose-response relationships and the time course of benzo[a]pyrene-induced
skin damage, DNA adduct formation, and tumor formation, groups of 43-85 female Harlan mice
were treated dermally with 0, 16, 32, or 64 jag of benzo[a]pyrene in 50 [xL of acetone once per
week for 29 weeks (Albert et al., 1991). Interscapular skin of each mouse was clipped 3 days
before the first application and every 2 weeks thereafter. Additional groups of mice were treated
for 9 weeks with 0, 8, 16, 32, or 64 jag radiolabeled benzo[a]pyrene to determine benzo[a]pyrene
diolepoxide-DNA (BPDE-DNA) adduct formation in the epidermis at several time points (1, 2,
4, and 9 weeks). Tumor formation was monitored only in the skin.
No tumors were present in vehicle-treated or untreated control mice. In exposed groups,
incidences of mice with skin tumors were not reported, but time-course data for cumulative
number of tumors per mouse, corrected for deaths from nontumor causes, were reported.
Tumors began appearing after 12-14 weeks of exposure for the mid- and high-dose groups and
at 18 weeks for the low-dose group. At study termination (35 weeks after start of exposure), the
mean number of tumors per mouse was approximately one per mouse in the low- and mid-dose
groups and eight per mouse in the high-dose group; indicating that most, if not all, mice in each
exposure group developed skin tumors and that the tumorigenic response was greatest in the
highest dose group. The majority of tumors were initially benign, with an average time of
8 weeks for progression from benign papillomas to malignant carcinomas. Epidermal damage
occurred in a dose-related manner (more severe in the high-dose group than in the low- and mid-
dose groups) and included statistically significant increases (compared with controls) in: [ H]-
thymidine labeling and mitotic indices; incidence of pyknotic and dark cells (signs of apoptosis);
117
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
and epidermal thickness. Only a minor expansion of the epidermal cell population was observed.
In the high-dose group, indices of epidermal damage increased to a plateau by 2 weeks of
exposure. The early time course of epidermal damage indices was not described in the low- and
mid-dose groups, since data for these endpoints were only collected at 20, 24, and 30 weeks of
exposure. An increased level of BPDE-DNA adducts, compared with controls, was apparent in
all exposed groups after 4 weeks of exposure in the following order: 64>32>16>8 [j,g/week. The
time-course data indicate that benzo[a]pyrene-induced increases in epidermal damage indices
and BPDE-DNA adducts preceded the appearance of skin tumors.
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL, INHALATION, AND
DERMAL
As discussed in Section 4.1.4.3, several studies of human cohorts have examined possible
associations between lower body weights or head circumference in newborns or infants with
benzo[a]pyrene-DNA adducts levels in cord blood and exposure to ETS (Tang et al., 2006;
Perera et al., 2005a, b). Available studies of reproductive or developmental endpoints in
animals exposed orally or by inhalation to benzo[a]pyrene are reviewed as follows in this
section. No studies that evaluated reproductive or developmental effects following dermal
exposure were identified.
4.3.1. Oral
Mohamed et al. (2010) investigated multi-generational effects in male mice following
exposure of six-week old C57BL/6 mice (10/group) to 0 (corn oil), 1, or 10 mg/kg-day
benzo[a]pyrene for 6 weeks by daily gavage. Following final treatment, male mice were allowed
to stabilize for one week prior to being mated with two untreated female mice to produce an F1
generation. Male mice were sacrificed one week after mating. F1 males were also mated with
untreated female mice as were F2 males. The mice of the Fl, F2, and F3 generations were not
exposed to benzo[a]pyrene. The F0, Fl, F2 and F3 mice were all sacrificed at the same age (14
weeks) and endpoints including testis histology, sperm count, sperm motility, and in vitro sperm
penetration (of hamster oocytes) were evaluated. These endpoints were analyzed statistically
using ANOVA and Tukey's honest significance test and results were reported graphically as
means +/- SD.
Testicular atrophy was observed in the benzo[a]pyrene treatment groups, but was not
statistically different than controls. Statistically significant reductions were observed in
epididymal sperm counts of F0 and Fl generations treated with the high or low dose of
benzo[a]pyrene. For F0 and Fl generations, epididymal sperm counts were reduced
approximately 50% and 70%, respectively, in the low and high dose groups. Additionally, sperm
motility was statistically significantly decreased in the high dose in the F0 and Fl generations.
Sperm parameters of the F3 generation were not statistically different from controls. An in vitro
118
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
sperm penetration assay revealed statistically significantly reduced fertilization in F0 and F1
generations of the low and high dose groups. However, the value of this in vitro test is limited as
it bypasses essential components of the intact animal system (US EPA, 1996). Based on
decreased epididymal sperm counts of F0 and F1 generations, a LOAEL of 1 mg/kg-day was
established from this study (no NOAEL was identified).
Xu et al., (2010) treated female Sprague-Dawley rats (6/group) to 0 (corn oil only), 5, or
10 mg/kg-day benzo[a]pyrene by gavage every other day for a duration of 60 days. This resulted
in time weighted average doses of 0, 2.5, and 5 mg/kg-day over the study period of 60 days.
Endpoints examined included ovary weight, estrous cycle, 17B-estradiol blood level, and ovarian
follicle populations (including primordial, primary, secondary, atretic, and corpora leutea).
Animals were observed daily for any clinical signs of toxicity and following sacrifice, gross
pathological examinations were made and any findings were recorded. All animals survived to
necropsy. A difference in clinical signs was not observed for the treated groups and body
weights were not statistically different in treated animals (though they appear to be depressed 6%
at the high dose). Absolute ovary weight was statistically significantly reduced in the both the
low and high dose groups, 11 and 15% respectively (see Table 4-20). Animals treated with the
high dose were noted to have a statistically significantly prolonged duration of the estrous cycle
and non-estrus phase compared to controls. Animals in the high dose group also had statistically
significantly depressed levels of estradiol (by approximately 25%) and decreased numbers of
primordial follicles (by approximately 20%). This study also indicated a strong apoptotic
response of ovarian granulosa cells as visualized through TUNEL labeling, however, the
strongest response was seen at the low dose; decreased apoptosis was also observed at the high
dose. Based on decreased ovary weight following 60 day oral exposure to benzo[a]pyrene, a
LOAEL of 2.5 mg/kg-day was established from this study (no NOAEL was identified).
Table 4-20. Means ± SD for ovary weight in female SD-rats
Dose (mg/kg-d)a
0
2.5
5
Ovary weight (g)
0.160 ±0.0146
0.143 ± 0.0098b
0.136 ±0.0098b
Body weight (g)
261.67 ± 12.0
249.17± 11.2
247.25 ± 11.2
a TWA doses over the 60 day study period
b Statistically different from controls (p < 0.05) using one-way ANOVA
Source: Xu et al. (2010).
119
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Zheng et al., (2010) treated male Sprague-Dawley rats to 0 (corn oil only), 1, or 5 mg/kg-
day benzo[a]pyrene by daily gavage for a duration of 30 (8/group) or 90 days (8/group). At
necropsy, the left testis of each animal was collected and weighed. Testes testosterone
concentrations were determined by radioimmunassay and results were expressed as ng/g testis
and reported graphically. Testicular testosterone was statistically significantly decreased in the
high dose group approximately 15% following 90 days of exposure. The low dose group also
appeared to have a similar average depression of testosterone levels; however, the change did not
reach statistical significance. Testosterone levels measured in animals sacrificed following 30
days of benzo[a]pyrene exposure were not statistically different than controls. Based on
decreased testicular testosterone levels following 90 day oral exposure to benzo[a]pyrene, a
LOAEL of 5 mg/kg-day and a NOAEL of 1 mg/kg-day were identified.
McCallister et al., (2008) administered 0 or 300 ug/kg benzo[a]pyrene by oral gavage in
peanut oil to pregnant Long Evans rats (n= 5 or 6) on GDs 14 to 17. At this exposure level, no
significant changes were see in number of pups per litter, pup growth, or liver to body weight
ratios in control compared to benzo[a]pyrene exposed offspring. Treatment related differences
in brain to body weight ratios were observed only on PND 15 and PND 30. Decreases in
cerebrocortical mRNA expression of the glutamatergic NMDA receptor subunit was
significantly reduced (50%) in treated offspring compared to controls. In addition, in utero
exposed offpring exhibited decreased evoked cortical neuronal activity in the barrel field cortex
when tested at PN 90-120.
Rigdon and Neal (1965) administered diets containing 1,000 ppm benzo[a]pyrene to
pregnant mice (nine/group) on GDs 10-21 or 5-21. The pups were reported as appearing
generally normal at birth, but cannibalism was elevated in the exposed groups. These results
contrast with an earlier study (Rigdon and Rennels, 1964) in which rats (strain not specified)
were fed diets containing benzo[a]pyrene at 1,000 ppm for approximately 28 days prior to
mating and during gestation. In the earlier study, five of eight treated females mated with
untreated males became pregnant, but only one delivered live young. The treated dam that
delivered had two live and two stillborn pups; one dead pup was grossly malformed. In the
remaining treated females, vaginal bleeding was observed on GDs 23 or 24. In the inverse
experimental design, three of six controls mated to benzo[a]pyrene-treated males became
pregnant and delivered live young. Visceral and skeletal examinations of the pups were not
conducted. These studies were limited by the small numbers of animals, minimal evaluation of
the pups, lack of details on days of treatment (food consumption, weight gain), and the
occurrence of cannibalism.
Reproductive effects of in utero exposure via oral route
MacKenzie and Angevine (1981) conducted a two-generation reproductive and
developmental toxicity study for benzo[a]pyrene in CD-I mice. Benzo[a]pyrene was
120
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
administered by gavage in 0.2 mL of corn oil to groups of 30 or 60 pregnant (the F0 generation)
mice at doses of 0, 10, 40, or 160 mg/kg-day on GDs 7-16 only. Therefore, unlike the standard
two-generation study, F1 animals were exposed only in utero. F1 offspring were evaluated for
postnatal development and reproductive function as follows. F1 pups (four/sex when possible)
were allowed to remain with their mothers until weaning on PND 20. Crossover mating studies
were then conducted. Beginning at 7 weeks of age, each F1 male mouse (n = 20-45/group) was
allowed to mate with two untreated virgin females for 5-day periods for 25 days (for a total
exposure of 10 untreated females/Fl male), after which time the males were separated from the
females. Fourteen days after separation from the males, (i.e., on days 14-19 of gestation), the
females were sacrificed and the numbers of implants, fetuses, and resorptions were recorded.
The F2 fetuses were then examined for gross abnormalities. Similarly, each F1 female mouse (n
= 20-55/group), beginning at 6 weeks of age, was paired with an untreated male for a period of 6
months. Males were replaced if the females failed to produce a litter during the first 30-day
period. All F2 young were examined for gross abnormalities on day 1 of life and their weights
were recorded on day 4 of age. This F2 group was sacrificed on day 20 postpartum, while the
F1 female was left with a male until the conclusion of the study. At 6 weeks of age, gonads of
groups of 10 male and 10 female F1 mice exposed to 0, 10, or 40 mg/kg-day benzo[a]pyrene in
utero were subjected to gross pathology and histologic examinations.
No maternal toxicity was observed. The number of F0 females with viable litters at
parturition at the highest dose was statistically significantly reduced by about 35% (Table 4-21),
but progeny were normal by gross observation. Parturition rates of the low- and mid-dose
groups were unaffected by treatment, and litter sizes of all treated groups were similar to the
control group throughout lactation. However, body weights of the F1 pups in the mid-dose and
high-dose groups were statistically significantly decreased on PND 20, by 7 and 13%,
respectively, and in all treated pups on PND 42, 6, 6, and 10% for the low, mid, and high dose,
respectively (Table 4-21). The number of F1 pups surviving to PNDs 20 and 42 was
significantly reduced at the high dose (p < 0.01), by 8 and 16%, respectively. When F1 males
were bred to untreated females and F1 females were mated with untreated males, a marked dose-
related decrease in fertility of > 30% was observed in both sexes, starting at the lowest exposure.
There were no treatment-associated gross abnormalities or differences in body weights in the F2
offspring.
Table 4-21. Reproductive effects in male and female CD-I F1 mice exposed
in utero to benzo[a]pyrene
Effect
Dose (mg/kg-d)a
0
10
40
160
121
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
F0 mice with viable litters at parturition
46/60 (77%)
21/30 (70%)
44/60 (73%)
13/30 (43%)b
Mean ± SEM pup weight (g) at PND 20
11.2 ± 0.1
11.6 ± 0.1
10.4 ± 0.1b
9.7 ± 0.2b
Mean ± SEM pup weight (g) at PND 42
29.9 ±0.2
28.2 ± 0.3b
28.0 ± 0.2b
26.8 ± 0.4b
F1 male fertility index0
80.4
52.0b
4.7b
0.0b
F1 female fertility indexd
100.0
65.7b
0.0b
0.0b
aPregnant F0 mice were administered daily doses of benzo[a]pyrene in corn oil on GDs 7-16.
bSignificantly (p < 0.05) different from control by unspecified tests.
beginning at 7 wks of age, each F1 male mouse (20-45/group) was exposed to 10 untreated females over a period
of 25 d. Index = (females pregnant/females exposed to males) x 100.
dBeginning at 6 wks of age, each F1 female mouse (20-55/group) was cohabitated with an untreated male for a
period of 6 mo.
Source: MacKenzie and Angevine (1981).
Exposure to benzo[a]pyrene caused a marked dose-related decrease in the size of the
gonads. In F1 males, testes weights were statistically significantly reduced. Testes from animals
exposed in utero to 10 and 40 mg/kg-day weighed approximately 60 and 18%, respectively, of
the weight of testes from the control animals (no F2 offspring were produced in the high dose
group). This was confirmed by histopathologic observation of atrophic seminiferous tubules in
the 40 mg/kg-day group that were smaller than those of controls and were empty except for a
basal layer of cells. The number of interstitial cells in the testes was also increased in this group.
Males from the 10 mg/kg-day group showed limited testicular damage; although all exhibited
evidence of tubular injury, each animal had some seminiferous tubules that displayed active
spermatogenesis. Ovarian tissue was absent or reduced in F1 females such that organ weights
were not possible to obtain. Examination of available tissue in these females revealed
hypoplastic ovaries with few follicles and corpora lutea (10 mg/kg-day) or with no evidence of
folliculogenesis (40 mg/kg-day). Ovarian tissue was not examined in highest-dose females.
The LOAEL in this study was 10 mg/kg-day, based on decreases in mean pup weight
(<5%) at PND 42 of F1 offspring of dams treated with 10, 40, or 160 mg/kg-day benzo[a]pyrene,
marked decreases in the reproductive capacity (as measured by fertility index) of both male and
female F1 offspring exposed at all three treatment levels of benzo[a]pyrene (by approximately
30% in males and females), decreased litter size (by about 20%) in offspring of F1 dams, and
also the dramatic decrease in size and alteration in anatomy of the gonads of both male and
female F1 mice exposed to 10 and 40 mg/kg-day benzo[a]pyrene in utero. A NOAEL was not
identified.
In another reproductive and developmental toxicity study, benzo[a]pyrene was
administered by gavage in corn oil to nine female NMRI mice at a dose of 10 mg/kg-day on GDs
7-16; a group of nine controls received corn oil (Kristensen et al., 1995). Body weights were
monitored. F0 females were kept with their offspring until after weaning (21 days after
delivery). At 6 weeks of age, one F1 female from each litter (n = 9) was caged with an untreated
male. The F2 offspring were inspected for gross deformities at birth, weight and sex were
recorded 2 days after birth, and the pups were sacrificed. The F1 females were sacrificed after 6
122
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
months of continuous breeding. The effects of benzo[a]pyrene treatment on fertility, ovary
weights, follicles, and corpora lutea were evaluated. F0 females showed no signs of general
toxicity, and there was no effect on their fertility. F1 females had statistically significantly lower
median numbers of offspring, number of litters, and litter sizes and a statistically significantly
greater median number of days between litters as compared with the controls (Table 4-22). At
necropsy, the F1 females from treated F0 females had statistically significantly reduced ovary
weights; histologic examination of the ovaries revealed decreased numbers of small, medium, or
large follicles and corpora lutea (Table 4-22). Only one dose group was used in this study, with
decreased F1 female fertility observed following in utero exposure at the LOAEL of 10 mg/kg-
day; no NOAEL was identified.
Table 4-22. Effect of prenatal exposure to benzo[a]pyrene on indices of
reproductive performance in F1 female NMRI mice
Endpoint (median with range in parentheses)
Control"
benzo[a]pyrene exposeda(10
mg/kg-d)
Number of F2 offspring
92 (26-121)
22b (0-86)
Number of F2 litters
8 (3-8)
3b (0-8)
F2 litter size (number of pups per litter)
11.5 (6-15)
8b (3-11)
Number of d between F2 litters
20.5 (20-21)
21b (20-23)
F1 ovary weight (mg)
13 (13-20)
9b (7-13)
Number of small follicles
44 (1-137)
0b (0-68)
Number of medium follicles
9 (5-25)
0b (0-57)
Number of large follicles
14 (6-23)
0b (0-19)
Number of corpora lutea
16 (6-35)
0b (0-14)
aGroups of nine female NMRI F0 mice were administered 0 or 10 mg benzo[a]pyrene/kg by gavage in corn oil on
GDs 7-16. One F1 female from each litter was continuously bred with an untreated male for 6 mo.
bSignificantly (p < 0.05) different from control group by Wilcoxon rank sum test or Kruskall-Wallis two-tailed test.
Source: Kristensen et al. (1995).
Reproductive effects in adults and repeated oral exposure
Rigdon and Neal (1965) conducted a series of experiments to assess the reproductive
effects of orally administered benzo[a]pyrene to Ah-responsive white Swiss mice. Female
animals (number not stated) were administered benzo[a]pyrene at 250, 500, or 1,000 ppm in the
feed before or during a 5-day mating period. Based on the initial body weight, the doses can be
estimated as 32, 56, and 122 mg/kg-day, respectively. No effect on fertility was observed at any
treatment dose, even when animals were fed 1,000 ppm benzo[a]pyrene for 20 days prior to
mating, but interpretation of this finding was marred by large variability in numbers of pregnant
females and litter sizes for both treated and control mice. In separate experiments, the fertility of
five male mice/group was not affected by exposure to 1,000 ppm in food for up to 30 days prior
to mating with untreated females. Histologic examinations showed that male mice fed 500 ppm
benzo[a]pyrene for 30 days had spermatozoa present in their testes; further details were not
123
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
provided. The only treatment-related effect was a lack of weight gain related to feed
unpalatabilty. While this study suggests that premating exposure of male or female mice to
doses up to 122 mg/kg-day for 20 days may not affect fertility, the sample sizes were too small
and study designs were too inconsistent to provide reliable NOAELs and LOAELs for
reproductive/devel opmental toxi city.
In an earlier study (Rigdon and Rennels, 1964) rats (strain not specified) were fed diets
containing benzo[a]pyrene at 1,000 ppm for approximately 28 days prior to mating and during
gestation. In this study, five of eight treated females mated with untreated males became
pregnant, but only one delivered live young. The treated dam that delivered had two live and
two stillborn pups; one dead pup was grossly malformed. In the remaining treated females,
vaginal bleeding was observed on GDs 23 or 24. In the inverse experimental design, three of six
controls mated to benzo[a]pyrene-treated males became pregnant and delivered live young.
Visceral and skeletal examinations of the pups were not conducted. These studies are
insufficiently reported and of insufficient design (e.g., inadequate numbers of animals for
statistical analysis) to provide reliable NOAELs or LOAELs for reproductive effects from
repeated oral exposure to benzo[a]pyrene.
Immunosuppression effects and in utero exposure via oral route
No studies were found that examined immune system endpoints following in utero
exposure via the oral route. The abstract of a report by Holladay and Smith (1994) referred to
gavage dosing in a study of immune endpoints in fetuses of pregnant mice exposed to
benzo[a]pyrene on GDs 13-17, but the methods section of the report described an i.p. injection
procedure in detail (see Section 4.4.2).
4.3.2. Inhalation
124
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Reproductive toxicity and in utero exposure via inhalation
Archibong et al. (2002) evaluated the effect of exposure to inhaled benzo[a]pyrene on
fetal survival and luteal maintenance in timed-pregnant F344 rats. Prior to exposure on GD 8,
laparotomy was performed to determine the number of implantation sites, and confirmed
pregnant rats were divided into three groups, consisting of rats that had four to six, seven to nine,
or more than nine conceptuses in utero. Rats in these groups were then assigned randomly to the
treatment groups or control groups to ensure a similar distribution of litter sizes. Animals
(10/group) were exposed to benzo[a]pyrene:CB aerosols at concentrations of 25, 75, or
100 (J,g/m via nose-only inhalation, 4 hours/day on GDs 11-20. Control animals were either
sham-exposed to CB or remained entirely unexposed. Results of particle size analysis of
generated aerosols were reported by several other reports from this laboratory (Inyang et al.,
2003; Ramesh et al., 2001a; Hood et al., 2000). Aerosols showed a trimodal distribution with
averages of 95% cumulative mass with diameters <15.85 [j,m; 89% <10 [j,m; 55% <2.5 [j,m; and
38%) <1 [j,m (Inyang et al., 2003). Ramesh et al. (2001a) reported that the (MMAD ± GSD) for
the 55%) mass fraction with diameters <2.5 [j,m was 1.7 ± 0.085. Progesterone, estradiol-17p,
and prolactin concentrations were determined in plasma collected on GDs 15 and 17. Fetal
survival was calculated as the total number of pups divided by the number of all implantation
sites determined on GD 8. Individual pup weights and crown-rump length per litter per
treatment were determined on PND 4 (PND 0 = day of parturition).
Archibong et al. (2002) reported that exposure of rats to benzo[a]pyrene caused
biologically and statistically significant (p < 0.05) reductions in fetal survival compared with the
two control groups; fetal survival rates were 78.3, 38.0, and 33.8%> per litter at 25, 75, and
"3
100 (J,g/m , respectively, and 96.1% with CB or 98.8%> per litter in untreated controls (see Table
4-23). Consequently, the number of pups per litter was also decreased in a concentration-
3 3
dependent manner. The decrease was ~50%> at 75 (j,g/m and ~65%> at 100 (j,g/m , compared with
sham-exposed and unexposed control groups. No effects on hormone levels were observed on
GDs 15 or 17 at the low-dose. Biologically significant decreases in mean pup weights
"3
(expressed as g per litter) of >5%> were observed at doses >75 (j,g/m (14 and 16%> decreases at 75
and 100 (J,g/m , respectively,/? < 0.05). Exposure to benzo[a]pyrene did not affect crown-rump
length (see Table 4-22).
Table 4-23. Pregnancy outcomes in female F344 rats treated with
benzo[a]pyrene on GDs 11-21 by inhalation
Parameter3
Administered concentration of benzo[a]pyrene (|ig/m3)
0 (unexposed
control)
0
(carbon black)
25
75
100
Implantation sites
8.6 ±0.2
8.8 ±0.1
8.8 ±0.5
9.0 ±0.2
8.8 ± 0.1
Pups per litter
8.5 ±0.2
8.7 ±0.2
7.4 ± 0.5b
4.2 ± 0. lb
3.0 ± 0.2b
125
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Survival (litter %)
98.9 ± 1.1
96.7 ± 1.7
78.3 ± 4.1b
38.0 ± 2.1b
33.8 ± 1.3b
Pup weight (g/litter)
10.6 ±0.1
8.8 ±0.1
10.5 ±0.2
9.1 ±0.2b
8.9 ± 0.1b
Crown-rump length
(mm/litter)
29.4 ±0.6
29.3 ±0.5
28.0 ±0.6
27.3 ±0.7
27.9 ±0.7
aValues presented as means ± SEM.
bSignificantly different from controls at p < 0.05 by one-tailed post-hoc t-testing following ANOVA.
Source: Archibong et al. (2002).
benzo[a]pyrene exposure at 75 (j,g/m caused a statistically significant decrease in plasma
progesterone, estradiol, and prolactin on GD 17; these levels were not determined in the rats
exposed to 100 (j,g/m (Archibong et al., 2002). Plasma prolactin is an indirect measure of the
activity of decidual luteotropin, a prolactin-like hormone whose activity is necessary for luteal
maintenance during pregnancy in rats. Control levels of prolactin increased from GDs 15 to 17,
"3
but this increase did not occur in the rats exposed to 75 (J,g/m . Although the progesterone
"3
concentration at 75 (J,g/m was significantly lower than in controls on GD 17, the authors thought
that the circulating levels should have been sufficient to maintain pregnancy; thus, the increased
loss of fetuses was thought to be caused by the lower prolactin levels rather than progesterone
deficiency. The reduced circulating levels of progesterone and estradiol-17(3 among
benzo[a]pyrene-treated rats were thought to be a result of limited decidual luteotropic support for
the corpora lutea. The authors proposed the following mechanism for the effects of
benzo[a]pyrene on fertility: benzo[a]pyrene or its metabolites decreased prolactin and decidual
luteotropin levels, compromising the luteotropic support for the corpora lutea and thereby
decreasing the plasma levels of progesterone and estradiol-17(3. The low estradiol-17(3 may
decrease uterine levels of progesterone receptors, thereby resulting in fetal mortality. Based on
biologically and statistically significant decreases in pups/litter and percent fetal survival/per
"3
litter, the LOAEL was 25 (J,g/m ; no NOAEL was identified.
Neurotoxicity and in utero exposure via inhalation
To evaluate the effects of benzo[a]pyrene on the developing central nervous system,
Wormley et al. (2004) exposed timed-pregnant F344 rats (10/group) to benzo[a]pyrene:CB
"3
aerosols by nose-only inhalation on GDs 11-21 for 4 hours/day at a concentration of 100 (J,g/m .
Results of particle size analysis of genenerated aerosols were reported by other reports from this
"3
laboratory (Ramesh et al., 2001a; Hood et al., 2000). Particle size analysis of a 100-[j,g/m
aerosol showed a trimodal distribution with averages of 95% cumulative mass with diameters
<15.85 [j,m; 90% <10 [j,m; 67.5% <2.5 [j,m; and 66.2% <1 [j,m; the MMAD ± GSD for the latter
fraction was 0.4 ± 0.02 [j,m (Hood et al., 2000). Dams were maintained to term and pups were
weaned on PND 30. Benzo[a]pyrene reduced the number of live pups to one-third of control
values without affecting the number of implantation sites. During PNDs 60-70 electrical
stimulation and evoked field potential responses were recorded via electrodes implanted into the
126
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
brains of the animals. Direct stimulation of perforant paths in the entorhinal region revealed a
diminution in long-term potentiation of population spikes across the perforant path-granular cell
synapses in the dentate gyrus of the hippocampus of F1 generation benzo[a]pyrene-exposed
animals; responses in exposed offspring were about 25% weaker than in control offspring.
Additionally, NMDA receptor subunit 1 protein (important for synaptic functioning) was down-
regulated in the hippocampus of benzo[a]pyrene exposed F1 pups. The authors interpreted their
results as suggesting that gestational exposure to benzo[a]pyrene inhalation attenuates the
capacity for long-term potentiation (a cellular correlate of learning and memory) in the F1
generation.
In a later study by this same group of investigators, Wu et al. (2003) evaluated the
generation of benzo[a]pyrene metabolites in F1 generation pups, as well as the developmental
profile for AhR and mRNA. In this study, confirmed pregnant F344 rats were exposed to
-3
benzo[a]pyrene:CB aerosols at 25, 75, or 100 (J,g/m via nose-only inhalation, 4 hours/day, for
10 days (GDs 11-21). Control animals were exposed to CB (sham) to control for inert carrier
effects or they remained untreated. Each benzo[a]pyrene concentration had its own set of
controls (CB and untreated). Two randomly selected pups were sacrificed on each of PND 0, 3,
5, 10, 15, 20, and 30. Body, brain, and liver weights were recorded. Benzo[a]pyrene metabolites
were analyzed in the cerebral cortex, hippocampus, liver, and plasma. A dose-related increase in
plasma and cortex benzo[a]pyrene metabolite concentrations in pups was observed.
Dihydrodiols (4,5-; 7,8-; 9,10-) dominated the metabolite distribution profile up to PND 15 and
the hydroxy (3-OH-benzo[a]pyrene; 9-OH-benzo[a]pyrene) metabolites after PND 15 at
100 (j,g/m (the only exposure concentration reported). Results indicated a dose-related decrease
in the ratio of the total number of pups born per litter to the total number of implantation sites per
3 3
litter. The number of resorptions at 75 and 100 (J,g/m , but not at 25 (J,g/m , was statistically
significantly increased compared with controls.
Adult male reproductive effects and repeated inhalation exposure
Inyang et al. (2003) evaluated the effect of sub-acute exposure to inhaled benzo[a]pyrene
on testicular steroidogenesis and epididymal function in rats. Male F344 rats (10/group),
"3
13 weeks of age, were exposed to benzo[a]pyrene:CB aerosols at 25, 75, or 100 (j,g/m via nose-
only inhalation, 4 hours/day for 10 days. Control animals were either exposed to CB (sham) to
control for exposure to the inert carrier, or they remained untreated. Each benzo[a]pyrene
concentration had its own set of controls (CB and untreated). Aerosols showed a trimodal
distribution with averages of 95% cumulative mass <15.85 [j,m; 89% <10 [j,m; 55% <2.5 [j,m; and
38%) <1 [j,m (Inyang et al., 2003); an earlier report from this laboratory indicated that the 55%
mass fraction had a MMAD ± GSD of 1.7 ± 0.085 (Ramesh et al., 2001a). Blood samples were
collected at 0, 24, 48, and 72 hours after cessation of exposure to assess the effect of
benzo[a]pyrene on systemic concentrations of testosterone and luteinizing hormone (LH),
127
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
hormones that regulate testosterone synthesis. Reproductive endpoints such as testis weight and
motility and density of stored (epididymal) sperm were evaluated.
Regardless of the exposure concentration, inhaled benzo[a]pyrene did not affect testis
weight or the density of stored sperm compared with controls. However, inhaled benzo[a]pyrene
caused a concentration-dependent reduction in the progressive motility of stored sperm.
"3
Progressive motility was similar at 75 and 100 |ig/m , but these values were significantly lower
(p < 0.05) than in any other group. The reduction in sperm motility post-cessation of exposure
was thought to be the result of benzo[a]pyrene limiting epididymal function. Benzo[a]pyrene
exposure to 75 (j,g/m caused a decrease in circulating concentrations of testosterone compared
with controls from the time of cessation of exposure (time 0) to 48 hours post-termination of
exposure (p < 0.05). However, the decrease was followed by a compensatory increase in
"3
testosterone concentration at 72 hours post-cessation of exposure. Exposure to 75 (j,g/m caused
a nonsignificant increase in plasma LH concentrations at the end of exposure compared with
controls, which increased further and turned significant (p < 0.05) for the remaining time of the
study period. The decreased plasma concentration of testosterone, accompanied by an increased
plasma LH level, was thought to indicate that benzo[a]pyrene did not have a direct effect on LH.
The authors also noted that the decreased circulating testosterone may have been secondary to
induction of liver CYP450 enzymes by benzo[a]pyrene. The authors concluded that subacute
exposure to benzo[a]pyrene contributed to impaired testicular endocrine function that ultimately
"3
impaired epididymal function. Based on this study, the NOAEL was 25 (J,g/m and the LOAEL
was 75 (J,g/m , based on a statistically significant reduction in the progressive motility of stored
"3
sperm and impairment of testicular function with 10 days of exposure at 75 (J,g/m .
In a follow up study with longer exposure duration, adult male F344 rats (10 per group)
-3
were exposed to benzo[a]pyrene:CB aerosols at 75 (J,g/m via nose-only inhalation, 4 hours/day
for 60 days (Archibong et al., 2008; Ramesh et al., 2008). Rats in the control group were
subjected to the nose-only restraint, but were not exposed to the CB carrier. Blood samples were
collected at 0, 24, 48, and 72 hours after exposure terminated, and the animals sacrificed for
tissue analyses following the last blood sampling. Data were analyzed statistically for
benzo[a]pyrene effects on weekly body weights, total plasma testosterone and LH
concentrations, testis weights, density of stored spermatozoa, sperm morphological forms and
motility, benzo[a]pyrene metabolite concentrations and AHH activity, and morphometric
assessments of testicular histologies. Relative to controls, the results indicated 34% reduced
testis weight (p < 0.025), reduced daily sperm production (p < 0.025) and reduced intratesticular
testosterone concentrations (p < 0.025). Plasma testosterone concentrations were reduced to
about one-third of the level in controls on the last day of exposure (day 60) and at 24, 48, and 72
hours later (p < 0.05). However, plasma LH concentrations in benzo[a]pyrene exposed rats were
elevated throughout the blood sampling time periods compared with controls (p < 0.05). In
testis, lung, and liver, AHH activity, and benzo[a]pyrene-7,8-dihydrodiol (precursor to the DNA-
128
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
reactive BPDE) and benzo[a]pyrene-3,6-dione metabolites were significantly (p < 0.05) elevated
relative to controls. Progressive motility and mean density of stored spermatozoa were
significantly reduced (p < 0.05). Weekly body weight gains were not affected by benzo[a]pyrene
exposure. These results indicate that 60-day exposure of adult male rats to benzo[a]pyrene:CB
-3
aerosols at 75 (j,g/m produced decreased testis weight; decreased intratesticular and plasma
testosterone concentrations; and decreased sperm production, motility, and density.
4.4. OTHER DURATION OR ENDPOINT-SPECIFIC STUDIES
4.4.1. Acute Neurological Studies
Saunders et al. (2001) administered benzo[a]pyrene (>97% pure in peanut oil) to F344
rats (10/sex/dose group) via a single gavage dose of 0, 12.5, 25, 50, 100, or 200 mg/kg. Separate
groups of animals were used for motor activity assessment and functional observational battery.
Motor activity (horizontal, vertical, total distance, and stereotypic activity) was measured over
2-hour intervals during the nocturnal phase (12 hours) for 5 consecutive days after treatment
(day 1). The functional observational battery, consisting of 29 tests assessing autonomic,
neuromuscular, CNS excitability, CNS activity, sensorimotor, and physiological activity, was
administered before dosing and at 2, 4, 6, 12, 24, 48, 72, and 96 hours after dosing. Body weight
was measured after the functional observational battery.
In both sexes of rat, body weight gain was significantly reduced on day 4 and/or 5 at
doses >25 mg/kg (Saunders et al., 2001). Body weight gains were comparable to controls in the
25 and 50 mg/kg groups by day 6. Higher doses of benzo[a]pyrene resulted in prolonged
reductions in body weight gain, with reductions of 21-26% in both sexes on day 9 after
treatment. Weight gain had returned to control levels after 2 weeks postdosing. At doses of
50 mg/kg and higher, all measures of motor activity were significantly depressed in both sexes
beginning 2 hours after dosing and persisting through the 12-hour post-dosing measurement. At
the 25 mg/kg dose, significant changes in motor activity were not observed until 4-6 hours after
treatment. When motor activity was measured over 24-hour intervals, significant depression of
motor activity (all measures) was observed at all doses on day 1 post-treatment and at >50 mg/kg
on day 2. The results of the functional observational battery showed significant effects on
neuromuscular endpoints (decreased mobility and grip strength, abnormal gait, loss of righting
reflex), autonomic endpoints (increased defecation and urination), and sensorimotor endpoints
(decreased response to sound, touch, and pain) at all doses and in both sexes. Effects on most
parameters peaked at 6 hours post-dosing, with return to control levels by 72 hours post-dosing.
The severity of effects on the FOB tests was greater in males than in females. This study
identified a LOAEL of 25 mg/kg benzo[a]pyrene for acute neurotoxicity; the NOAEL is
12.5 mg/kg.
In a study with nearly identical design, Saunders et al. (2002) treated male F344 rats
(10/dose) with single gavage doses of 0, 25, 50, 100, or 200 mg/kg benzo[a]pyrene (>97% pure,
129
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
in peanut oil). As in the study by Saunders et al. (2001), separate groups of animals were used
for motor activity assessment and functional observational battery (using the same endpoints
measured at the same time intervals); in addition, a third group of animals was treated at the
same doses and used for measurement of benzo[a]pyrene and its metabolites in plasma and brain
tissue. At 2, 4, 6, 12, 24, 48, 72, and 96 hours after dosing, rats in the metabolism groups were
sacrificed for analysis of benzo[a]pyrene and its metabolites in plasma and brain tissue. Dose-
and time-dependent effects on locomotor activity were observed in all treated groups. In
measurements conducted over the 12 hours on day 1 after treatment, significant (p < 0.05 relative
to vehicle controls) reductions in total distance traveled (considered by the authors to represent
the most accurate measure of locomotor activity) were observed at doses >50 mg/kg beginning at
2-4 hours post dosing and persisting up to the 12-hour time point. When assessed on a daily
basis over the 5 posttreatment days, total distance traveled was significantly lower than controls
in all treated groups on day 1 and in groups exposed to >50 mg/kg on day 2. Significant (p <
0.05) dose-dependent effects on neuromuscular, sensorimotor, and autonomic parameters were
observed with benzo[a]pyrene treatment. Significant increases in the severity of abnormal gait
and impaired sound and tail pinch responses, as well as decreased forelimb grip strength,
occurred in all dose groups; the severity of effects peaked at 4 or 6 hours after treatment.
Increased severity of landing foot splay was observed at doses >50 mg/kg, also peaking in
severity at 6 hours after treatment. Effects on this endpoint persisted from 2 to 24 hours post
treatment in the two high dose groups. Autonomic effects, consisting of increased frequency of
urination and defecation, occurred at doses >50 mg/kg; in the two high dose groups, these effects
began 4 hours after dosing and persisted through 24 hours. The onset and duration of the effects
observed in this study corresponded well with the plasma and brain tissue concentrations of
benzo[a]pyrene and its metabolites. In particular, levels of benzo[a]pyrene metabolites in brain
tissue peaked between 2 and 6 hours after dosing, and plasma levels peaked at 6 hours after
dosing, in all treated groups. By 72 hours after dosing, metabolite levels had returned to
baseline. Unmetabolized benzo[a]pyrene was detected in brain tissue only in the two highest
dose groups; levels in both brain tissue and plasma peaked between 2 and 6 hours after dosing.
Analysis of specific metabolite levels over time showed that the diol metabolites comprised a
larger percentage of the total metabolites at earlier time points (up to 12 hours after dosing),
while the hydroxyl metabolites predominated at later times. The authors postulated that the
observed toxic effects were associated with the production of benzo[a]pyrene diol metabolites
and/or related generation of ROS rather than an effect of the parent compound or hydroxyl
metabolites. The LOAEL in this study was 25 mg/kg based on suppression of locomotor activity
and evidence of impairment in the functional observational battery.
In a follow-up study, Saunders et al. (2006) attempted to correlate neurobehavioral
changes with levels of benzo[a]pyrene metabolites, antioxidant enzyme levels, and measures of
lipid peroxidation in selected brain regions. Single oral doses of 0, 25, 50, 100, or 200 mg/kg
130
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
benzo[a]pyrene (97% pure, in peanut oil) were administered by gavage to groups of 10 male
F344 rats. Motor activity, measured as total distance traveled over a 2-hour interval during the
nocturnal phase, was assessed at 0, 2, 4, 6, 12, 24, 48, 72, and 96 hours post treatment.
Additional groups of animals were exposed to the same doses and sacrificed for collection of
blood and specific brain tissues (hippocampus, striatum) at the same time points; benzo[a]pyrene
and its metabolites were measured by reverse phase HPLC in blood and brain tissues. In
addition, the activities of superoxide dismutase, catalase, glutathione peroxidase, and levels of
malondialdehyde in striatum and hippocampus were determined at 6 and 96 hours after
treatment. The authors reported that motor activity was significantly (p < 0.01) suppressed as
early as 2 hours after treatment and remained suppressed through the 48-hour time point in the
groups exposed to 50 and 200 mg/kg benzo[a]pyrene; however, the data were not reported, and it
was not clear whether the 100 mg/kg dose group was affected. The authors indicated that the
maximum suppression occurred at 12 hours, when motor activity was 72% lower than controls;
however, the affected dose group(s) was not reported. Data were reported graphically for the
25 mg/kg group only; based on the graph, a significant (p < 0.05) suppression of motor activity
occurred by 4 hours after treatment and persisted through the 12-hour measurement. Activity
had returned to control levels by 72 or 96 hours in all dose groups. As in the study reported by
Saunders et al. (2002), the levels of benzo[a]pyrene metabolites in the plasma and brain
correlated with the onset and duration of behavioral effects; metabolite concentrations peaked
between 2 and 6 hours after treatment, when suppression of motor activity occurred. In addition,
the toxification/detoxification ratio of benzo[a]pyrene metabolites (measured as the ratio of 7,8-
dihydrodiol 9,10-epoxide to 3[OH]benzo[a]pyrene) in plasma, cortex, cerebellum, hippocampus,
and striatum was higher (between 2 and 7) for the first 6 hours, indicating higher levels of the
more toxic epoxide metabolite. From 24 to 96 hours after treatment, the ratio was <1, indicating
that the hydroxyl form predominated. Measurement of antioxidant enzyme levels in the striatum
and hippocampus at 6 and 96 hours after dosing showed significant dose-related decreases in the
activities of superoxide dismutase and glutathione peroxidase, but enhanced catalase activity and
increased lipid peroxidation products in the striatum and hippocampus. The authors suggested
that benzo[a]pyrene-induced acute neurobehavioral effects may be associated with oxidative
stress resulting from generation of ROS and inhibition of brain antioxidants.
Evidence of neurotoxicity has also been reported in studies of acute exposure to
benzo[a]pyrene administered via parenteral routes. As part of a study of the neurotoxicity of
motorcycle exhaust, Liu et al. (2002) administered benzo[a]pyrene dissolved in corn oil via i.p.
injection to ICR mice (sex not specified; 4-6/group) at doses of 0, 50, or 100 mg/kg-day for
3 consecutive days. One day after the last treatment, motor nerve conduction velocity was
measured in the tails. The mice were then sacrificed for removal of the sciatic nerve, which was
assayed for Na /K+-ATPase activity. Although the methods section indicated that
benzo[a]pyrene-treated animals were tested for rotarod performance, results of this evaluation
131
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
were not reported. Data on motor nerve conduction velocity and Na+/K+-ATPase activity were
presented graphically. Exposure to benzo[a]pyrene at resulted in significant (p < 0.05)
depression of motor nerve conduction velocity (about 25 and 40% decrease from control for 50
and 100 mg/kg-day doses, respectively, based on visual inspection data presented graphically),
and decreased the Na+/K+-ATPase activities of sciatic nerves (about 20 and 30% decreases from
control, respectively).
Grova et al. (2007) evaluated the effects on short-term exposure to benzo[a]pyrene on
learning, memory, locomotor activity, and motor coordination. Groups of 10 female Balb/c mice
were treated with i.p. injections of benzo[a]pyrene (>97% pure, in vegetable oil) at doses of 0,
0.02, 0.2, 2, 20, or 200 mg/kg-day for 10 consecutive days. At the end of exposure, locomotor
activity was measured in open field activity, and motor coordination was assessed. Memory and
learning were evaluated using the Y maze (measures spontaneous alternation behavior) and the
Morris water maze (measures learning as escape latency in repeated trials). Finally, the animals
were sacrificed for removal of the brain; expression of the NMDA R1 receptor (involved in
cognitive function) subunit gene was measured in eight brain regions (cerebral trunk, cerebellum,
mesencephalum, hippocampus, hypothalamus, thalamus, frontal cortex, and temporal cortex) by
quantitative real-time reverse transcription PCR assay. In contrast to oral studies of
benzo[a]pyrene exposure (Saunders et al., 2006, 2002, 2001), injection of benzo[a]pyrene did not
affect locomotor activity at any dose in this study. At the lowest doses (0.02 and 0.2 mg/kg-
day), benzo[a]pyrene exposure resulted in reductions in the percentage of spontaneous
alternation in the Y maze. However, at the higher doses (>2 mg/kg-day), there was no difference
from controls. The authors attributed this finding to increased activity and arousal at the higher
doses, postulated to result from an anxiolytic effect of benzo[a]pyrene. In the 5th trial of the
water maze, all benzo[a]pyrene groups showed impairment; the escape latency was significantly
higher than controls. In contrast, the higher doses of benzo[a]pyrene resulted in significantly
reduced latency during the first trial. No differences from control were observed in the 2nd, 3rd,
and 4th trials, or 1 day after the last dose. Benzo[a]pyrene exposure resulted in modulation of
NMDA-R1 subunit gene expression; expression was significantly (p < 0.05) increased in the
cerebellum, mesencephalus, and hippocampus, but decreased in the frontal cortex and cerebral
trunk. Effects on gene expression in the cerebellum, frontal cortex, and hippocampus occurred at
all doses, while gene expression in the cerebral trunk was affected only at doses of >0.2 mg/kg-
day, and expression in the mesencephalus occurred only at doses of >2.0 mg/kg-day.
In a follow-up study, Grova et al. (2008) evaluated the effects of exposure to
benzo[a]pyrene on anxiety-related behaviors (performance in elevated-plus maze and hole-board
apparatus). The animals, group sizes, and doses were the same as in Grova et al. (2007), but
exposure occurred over 11 days. Behavioral tests were administered 30 minutes after the final
dose, and the animals were sacrificed 1 hour after the tests. Benzo[a]pyrene exposure at 20 and
200 mg/kg-day resulted in reduction in anxiety-related behavior as measured by the increased
132
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
number of head dippings in the hole-board apparatus. In the elevated-plus maze test, only the
highest dose resulted in a significant reduction in anxiety (as measured by higher percentage of
open arm entries and time spent in open arms).
4.4.2. Immunological Studies
Immunological effects (e.g., decreased thymus weight, decreased number of B cells in
spleen) have been reported in Wistar rats repeatedly exposed to benzo[a]pyrene doses >10-
15 mg/kg-day in standard oral toxicity studies (Kroese et al., 2001; De Jong et al., 1999).
Diminished immune responses elicited by the dermal sensitizer, 2,4-dinitrochlorobenzene
(DNCB) have been observed in C56BL/6 mice orally exposed to 13 mg/kg-day benzo[a]pyrene,
3 times/week for 4 weeks (van den Berg et al., 2005). No studies were located examining
immune system endpoints following inhalation exposure of animals to benzo[a]pyrene. Results
from studies of immune system endpoints in mice following i.p., subcutaneous (s.c.), or
intratracheal instillation exposure are consistent with immune suppression at dose levels
generally >40 mg/kg-day. The available animal studies identify immune suppression as a
potential hazard of repeated oral exposure to benzo[a]pyrene at doses >10-15 mg/kg-day.
4.4.2.1. Oral Exposure Immunological Studies
As discussed in Section 4.2.1.1, dose-related decreases in thymus weight and relative
number of B cells in the spleen were observed in male Wistar rats administered gavage doses
>10 mg/kg-day for 35 days (De Jong et al., 1999). At higher doses (>30 mg/kg-day), serum IgM
and IgA levels were decreased. At the highest dose tested (90 mg/kg-day), the relative cortex
surface area of thymus and thymic medullar weight were significantly reduced; NK cell activity
in the spleen was also reduced at this dose. No effects on the immune system were observed at
3 mg/kg-day (De Jong et al., 1999). In two additional subchronic gavage studies, thymus weight
was decreased in a dose-related manner in male Wistar rats exposed to doses >15 mg/kg-day
(5 days/week) for at least 5 weeks and in females exposed to doses of >30 mg/kg-day
(5 days/week) for 90 days (Kroese et al., 2001). No other immune system parameters were
assessed in these studies. In an adaptation of the sensitization-specific murine local lymph node
assay for use in testing immune function, van den Berg et al. (2005) tested several
immunomodulating compounds, including benzo[a]pyrene, for effects on the T-cell-dependent
immune response induced by the contact sensitizer, DNCB. Groups of eight male and eight
female C56BL/6 mice were given gavage doses of 13 mg/kg-day benzo[a]pyrene (purity not
reported) 3 times/week for 4 weeks, followed by sensitization with DNCB (0, 0.33, 0.66, or 1%
solutions in acetone: corn oil) applied topically to the backs of both ears for 3 consecutive days.
Three days after the last DNCB treatment, the lymph nodes under the application area were
excised, weighed, and homogenized for preparation of cell suspensions. The lymph node cell
"3
suspensions were tested for cell proliferation capacity via measurement of [ H]-thymidine
133
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
incorporation. Releases of the cytokines interferon (IFN)-y and interleukin (IL)-4 following
concanavalin A stimulation were assayed by ELIS A. At the highest concentration of the
"3
sensitizer, benzo[a]pyrene treatment reduced [ H]-thymidine incorporation into lymphocytes by
approximately 30% (based on visual inspection of data presented graphically; p = 0.008)
compared with untreated controls. In this treatment group, benzo[a]pyrene also reduced the
release of IFN-y (approximately 75% less than controls based on graphical data, not significant)
and IL-4 (approximately 60% less than controls based on graphical data, p < 0.001). The results
indicated that benzo[a]pyrene modulates the immune response elicited by the sensitizer DNCB.
4.4.2.2. Inhalation Exposure Immunological Studies
No studies were located that examined immune system endpoints following inhalation
exposure of animals to benzo[a]pyrene.
4.4.2.3. Other Exposure Route Immunological Studies
A number of studies have shown suppression of both humoral and cell-mediated immune
responses in mice following i.p., s.c., or intratracheal administration. Dose-related decreases in
spleen and serum IgM levels after challenge by sheep red blood cells (SRBC) were reported in
rats (10, 40 mg/kg-day) and mice (5, 20, 40 mg/kg-day) following s.c. injection of
benzo[a]pyrene for 14 days (Temple et al. 1993). Reduced spleen cell response to SRBC and
lipopolysaccharides were observed in B6C3Fi mice exposed to doses >40 mg/kg-day
benzo[a]pyrene by i.p. or s.c. injection for 4-14 days (Lyte and Bick, 1985; Dean et al., 1983;
Munson and White, 1983) or by intratracheal instillation for 7 days (Schnizlein et al., 1987).
B6C3Fi mice exhibited dose-dependent decreased resistance to Streptococcus pneumonia or
Herpes simplex type 2 following s.c. injection of 5, 20, or 40 mg/kg benzo[a]pyrene for 14 days
(Munson et al., 1985). Galvan et al. (2006) reported that single i.p. injections of mice with
50 mg/kg benzo[a]pyrene caused decreased pro/pre B-lymphocytes and neutrophils in bone
marrow, without affecting numbers of immature and mature B-lymphocytes or GR-1+ myeloid
cells. Several i.p. injection studies reported immune suppression effects in mice exposed to
benzo[a]pyrene in utero at doses ranging from 50 to 150 mg/kg; effects included decreased
spleen or thymus weights, suppression of antibody forming cells in response to sheep red blood
cells, decreased spleen, thymic, or bone marrow cellularity, and disrupted T-cell development
(Rodriguez et al. 1999; Holladay and Smith, 1995; Lummus and Henningsen, 1995; Holladay
and Smith, 1994; Urso and Johnson 1988; Urso et al., 1988; Urso and Gengozian, 1984, 1982,
1980).
In contrast to the studies that have shown decrements in immune response,
benzo[a]pyrene may also induce sensitization responses. Epicutaneous application of
benzo[a]pyrene (100 jag benzo[a]pyrene to C3H/HeN mice followed by ear challenge with 20 jag
134
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
benzo[a]pyrene 5 days later) produced a contact hypersensitivity (a significant ear swelling)
response (Klemme et al., 1987).
4.4.2.4. Other Exposure Route Developmental Immunotoxicy
While there are no oral or inhalation studies of benzo[a]pyrene on the developing
immune system, several i.p. injection studies indicate that this is an area of concern for both cell-
mediated and humoral immune ontogeny. In terms of cell-mediated effects, Urso and Gengozian
(1984) reported severe suppression of the mixed lymphocyte response and moderate suppression
of the graft-versus-host response in mice exposed in utero to 150 mg/kg from GD 11 to 17. Both
effects persisted until 18 mo of age. Holladay and Smith (1994) found that mice exposed to 0,
50, 100, or 150 mg/kg from GD 13 to 17 exhibited severe fetal thymic atrophy when examined
on GD 18. In the same study, expression data of cell surface markers (e.g. CD4, CD8) indicate
that benzo[a]pyrene may inhibit and/or delay thymocyte maturation, possibly contributing to the
observed thymic atrophy. Several other studies also show decreased thymocyte numbers and
disrupted T cell maturation after in utero exposure to benzo[a]pyrene (Rodriguez et al., 1999;
Holladay and Smith, 1995; Lummus and Henningsen, 1995; Urso et al., 1992; Urso and Johnson,
1987).
In addition to direct thymus effects, Holladay and Smith (1994) reported a large reduction
in total cellularity in the fetal liver, which is the primary hematopoietic organ during gestation
and a major source of thymocyte precursors beginning around GD 10-11 in mice (Pennit and
Vaddeur, 1989; Landreth and Dodson, 2005). This was accompanied by decreased expression of
terminal deoxynucleotidyl transferase (TdT), an intracellular marker known to be present in
cortical thymocyte progenitors in the fetal liver (Fine et al., 1990; Silverstone et al., 1976). This
data suggests that benzo[a]pyrene also disrupts liver hematopoiesis during gestation and may
interfere with prolymphoid seeding of the thymus, possibly contributing to thymic atrophy and
cell-mediated immunosuppression. Rodriguez et al. (1999) assessed downstream affects of T cell
development by showing that CD4+ T-cells were reduced in the spleen of 1-week old mice
following in utero benzo[a]pyrene exposure.
There is also some evidence of humoral immune disruption by benzo[a]pyrene during
fetal life. In a series of related studies, mice exposed to benzo[a]pyrene during mid (GD 11 to
13) or late (GD 16-18) gestation or both (GD 11 to 17) exhibited severe suppression of the
plaque-forming cell response to sheep red blood cells from 1 wk up to 18 mo after birth (Urso
and Gengozian, 1984, 1982, 1980). In their analysis of fetal liver cells, Holladay and Smith
(1994) reported large decreases in expression of TdT and CD45R cellular markers, both of which
are present on pre-B lymphocytes.
4.4.3. Cancer Bioassays (Other Routes of Exposure)
135
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Cancer bioassays following i.p. injection of mice with benzo[a]pyrene have consistently
found cancer responses. Newborn mouse bioassays involving postnatal injections of
benzo[a]pyrene (generally in the dose range of 0.5-1 |imol/mouse on PNDs 1, 8, and 15)
consistently found increases in liver or lung tumors, either increases in incidence of animals with
tumors or increased numbers of tumors per animal (LaVoie et al., 1994, 1987; Busby et al., 1989,
1984; Weyand and LaVoie, 1988; Wislocki et al., 1986; Buening et al., 1978; Kapitulnik et al.,
1978). Likewise, i.p. injection of pregnant mice with benzo[a]pyrene (100-150 mg/kg) during
gestation induced increased incidences of offspring with liver or lung tumors, compared with
controls (Urso and Gengozian, 1984; Bulay and Wattenberg, 1971). A/J adult mice given single
i.p. injections of benzo[a]pyrene showed a dose-related increase in the number of lung tumors
per mouse with doses ranging from about 5 to 200 mg/kg (Mass et al., 1993).
Tumorigenic responses to s.c. administered benzo[a]pyrene have been observed mostly at
the site of injection in studies with mice (Nikonova, 1977; Pfeiffer, 1977; Homburger et al.,
1972; Roe and Walters, 1967; Grant and Roe, 1963; Steiner, 1955; Rask-Nielson, 1950; Pfeiffer
and Allen, 1948; Bryan and Shimkin, 1943; Barry et al., 1935).
Positive cancer responses from other routes of exposure have included: (1) mammary
tumors in rats with intramammilary administration (Cavalieri et al., 1991, 1988a, b, c);
(2) cervical tumors in mice with intravaginal application (Naslund et al., 1987); (3) injection site
sarcomas with intramuscular injection (Sugiyama, 1973); (4) respiratory tract tumors in hamsters
with intratracheal instillation (Henry et al., 1973); and (5) tracheal epithelial tumors in rats with
intratracheal implantation (Topping et al., 1981, Nettesheim et al., 1977).
4.4.4. Atherogenesis Studies
Cigarette smoking (see Ramos and Moorthy, 2005; Miller and Ramos, 2001; Thirman et
al., 1994, for review) and, to a more limited degree, occupational exposure to PAH mixtures
(Burstyn et al., 2005) have been identified as risk factors associated with the development of
atherosclerotic vascular disease and increased risk for cardiovascular mortality. Based on results
from in vivo and in vitro animal studies, reactive metabolites of PAHs, including
benzo[a]pyrene, are thought to play a role in the progression of atherosclerosis leading to
hardening and thickening of the arteries (see Ramos and Moorthey, 2005; Miller and Ramos,
2001 for review). For example, in vivo exposure of Sprague-Dawley rats to 10 mg/kg
benzo[a]pyrene i.p. injections (once/week for 8 weeks) induced aortic wall lesions related to
atherosclerosis including loss of endothelial integrity and increase of smooth muscle cell mass
(Zhang and Ramos, 1997). The molecular mechanisms responsible for PAH-induced vascular
injury and the development of atherosclerosis are not well established, but current hypotheses
include roles for cell proliferative responses to injury of endothelial cells from reactive
metabolites (including ROS) and genomic alterations in smooth muscle cells from reactive
136
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
metabolites leading to transformed vasculature cells and eventual plaque formation (Ramos and
Moorthy, 2005).
Although many studies have been conducted in animal systems to study the mechanisms
by which PAHs may participate in the initiation and promotion of atherosclerosis, no studies
were located that examined relationships between levels of exposure to benzo[a]pyrene (via
environmentally relevant routes) and the development of aortic wall lesions related to
atherosclerosis, with the exception of a series of experiments involving repeated exposure of
Apolipoprotein E knock out (ApoE-/-) mice to oral doses of 5 mg/kg benzo[a]pyrene (Knaapen
et al., 2007; Curfs et al., 2005, 2004; Godschalk et al., 2003). ApoE-/- mice develop
spontaneous atherosclerosis, which is thought to be due to enhanced oxidative stress from the
lack of ApoE (Godschalk et al., 2003).
Treatment of male ApoE-/- mice with gavage doses of 5 mg/kg-day benzo[a]pyrene for
4 days produced increased levels of lipid peroxidation-derived DNA modifications (etheno-DNA
adducts) and BPDE-DNA adducts in aorta, compared with unexposed ApoE-/- controls
(Godschalk et al., 2003). Repeated exposure of male ApoE-/- mice to 5 mg/kg once a week for
12 or 24 weeks did not cause enhancement of the initiation of plaques in the aortic arch
(compared with unexposed ApoE-/- controls), but caused larger plaques with increased plaque
layering and number of lipid cores, and increased plaque content of T-lymphocytes, compared
with unexposed ApoE-/- controls (Curfs et al., 2004). In another study, gavage exposure of male
ApoE-/- mice with 5 mg/kg benzo[a]pyrene or 5 mg/kg benzo[e]pyrene (BeP) once per week for
24 weeks similarly increased plaque size and T-lymphocyte content (Curfs et al., 2005). In
addition, exposure to benzo[a]pyrene, and to a lesser extent BeP, was associated with increased
transforming growth factor beta (TGFpi) protein levels in plaque macrophages; TGFpi is
thought to play a role in the migration of T-lymphocytes. No exposure-related differences were
noted in the location or number of plaques, oxidative DNA damage (assessed by immunostaining
for 8-hydroxydeoxyguanosine, 8-OHdG), or apoptosis in the plaques (Curfs et al., 2005). As
expected, the lungs of benzo[a]pyrene-exposed mice showed several benzo[a]pyrene-DNA
adducts, which were not detectable in the lungs of BeP-exposed or control ApoE-/- mice (Curfs
et al., 2005). In another study, increased expression of monocyte-chemoattractant protein-1
(MCP-1) was found in aortic tissue from male ApoE-/- mice exposed to 5 mg/kg benzo[a]pyrene
once per week by gavage for 2 weeks; this protein is thought to recruit monocytes into
atherosclerotic lesions (Knaapen et al., 2007).
In summary, the results of the studies with ApoE-/- mice indicate that repeated oral
exposure to 5 mg/kg gavage doses of benzo[a]pyrene enhance the progression of (but do not
initiate) atherosclerosis through a general local inflammatory process. The involvement of PAH-
DNA adducts was not evident in these studies, as indicated by observations that BeP, which does
not cause DNA adducts, elicited similar plaque responses in ApoE-/- mice as benzo[a]pyrene.
Although these results demonstrate that repeated oral exposure to 5 mg/kg benzo[a]pyrene can
137
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
enhance atherosclerosis in animals, the altered genetic disposition of ApoE-/- mice limits their
usefulness in describing human-relevant dose-response relationships for oral exposure to
benzo[a]pyrene and atherosclerosis.
4.4.5. Reproductive Studies (Other Routes of Exposure)
Mattison et al. (1980) examined the response to i.p. exposure to a single dose of
benzo[a]pyrene in female DBa/2N mice (n=15 per dose group); effects on fertility, primordial
oocyte destruction, and response to pregnant mare's serum gonadotropins were evaluated. In the
10 week breeding study, dose groups included the vehicle control (corn oil) and 10, 100, 200 and
500 mg benzo[a]pyrene/kg. Complete infertility was seen in the 200 and 500 mg/day groups,
with decreased fertility seen in the 10 and 100 mg/kg dose groups, too. The total number of pups
born was 137, 91, 28, 0 and 0 and the mean number of pups per mouse per week was 0.91, 0.61,
0.20, 0.0, 0.0 in the 0, 10, 100, 200 and 500 mg/kg dose groups, respectively (p < 0.05 for
comparison of 0 to 10 mg/kg groups and 20 to 100 mg/kg groups). In a parallel study using a
single i.p. dose administered 21 days before sacrifice, the percent of primordial oocytes
destroyed (compared with controls) was 0, 18, 19, 56, 88, 100% for doses of 0, 5, 10, 50, 100,
and 200 mg benzo[a]pyrene/kg. The differences at doses > 50 mg/kg were statistically
significant (p < 0.05) compared with controls. The results from these studies were used to
calculated an ED50 (i.e., dose producing a reduction in fertility or number of oocytes) of 25.5
mg/kg for fertility reduction and 24.5 mg/kg for primordial oocyte destruction. There was no
effect of benzo[a]pyrene exposure on ovary weight or response to pregnant mare's serum
gonadotropin, indicating that the effect of exposure did not involve ovulation inhibition.
Another acute exposure study examined the effect of benzo[a]pyrene exposure on
ovulatory response (as determined by number of corpora lutea) in female C57BL/6N mice
(Swartz and Mattison, 1985). Benzo[a]pyrene was given as a control dose (corn oil vehicle), and
1, 5, 50, 100, 500 mg/kg i.p., 20 animlals were included per dose group and 5 were sacrificed at
weekly intervals. Ovaries were removed and serial sections were examined for histological
changes and counts of corpora lutea. There was a 35% mortality rate in the 500 mg/kg group,
but no evidence of treatment-related mortality in the other groups. Mean number of corpora
lutera in controls varied between 5.5 and 10.0 for the samples taken at 1, 2, 3, and 4 weeks post-
dose administration, with no time-related trend of increasing or decreasing number. The 1 mg/kg
dose group exhibited no decrease in number of corpora lutea compared with controls at any time
period (mean count varying between 6.2 and 7.2). The number of corpora lutea was decreased (p
< 0.05) at all doses > 5.0 mg/kg at 1 week post-administration (mean 0.0, 2.0, 0.0, 0.0 and 0.0 for
5.0, 10, 50, 100 and 500 mg/kg compared with 6.8 and 10.0 in the control and 1 mg/kg groups,
respectively); at 2 weeks decreases were seen at > 50 mg/kg and at weeks 3 and 4 a decrease was
seen only at doses >100 mg/kg. Thus in addition to the destruction of primordial follicles seen
138
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
in Mattison et al. (1980), this study also demonstrated an inhibition of ovulation by
benzo[a]pyrene that was time- and dose-related.
Miller et al. (1992) used the protocol described by Swartz and Mattison (1985) to further
examine the ovarian effects of acute benzo[a]pyrene exposure. As in the previous experiment,
doses of 1, 5, 10, 50, 100 and 500 mg/kg i.p. (with corn oil vehicle control) were administered to
female C57BL/6N mice (20 per group; 5 sacrificed at 1, 2, 3, and 4 weeks post-dose
administration). In addition to counts of corpora lutea, total ovarian volume and total and
individual corpora lutea volume were measured. The results pertaining to dose- and time-
dependent descreases in corpora lutea matched the results seen in Swartz and Mattison, 1985),
and similar trends were seen in the total ovarian volume and total corpora luteal volume
measures. However, volume of individual corpora lutea was increased in the treated animals
compared with controls at weeks 2, 3 and 4 post-treatment. The authors note that the recovery
seen in the effect on corpora lutea number by 2 weeks post-treatment at the lower doses may
reflect an effect specifically on antral follicles, whereas the longer recover period at higher doses
indicates an additional effect on growing follicles.
Borman et al. (2000) compared the ovarian effects of benzo[a]pyrene, two other PAHs
(9,10-dimethylbenzanthracene and 3-methylcholanthrene in female B6C3Fi mice and Fischer
344 rats); the ovotoxic 4-vinylcyclohexene (VHC)and its diepoxide metabolite (4-
vinylcyclohexene diexpoxide, VHD) were also included to allow calculation of an "ovotoxic
index." as positive controls. Doses of 0.0, 0.0075, 0.015, 0.075, 0.15, 0.75, 3.5, 7.5 and 15
mg/kg in the mouse, and an additional dose of 60 mg/kg for the rat were administered i.p.
(sesame oil vehicle) daily for 15 days (6-7 animals per treatment group). The size of the ovaries
and the number of primordial, primary and secondary (containing an oocyte) follicles was
determined was determined after sacrifice (4 hours after the last dose administration). The
ovotoxic index was defined as the lowest dose that resulted in a 50% loss (ED50) of primordial
follicles. For benzo[a]pyrene in mice, the ED50 was 3 mg/kg (0.012 mmol/kg) for primordial
follicles, a 50% loss in primary follicles was seen at 7.5 mg/kg (0.03 mmol/kg), but this level of
loss of secondary follicles was not seen even at the highest dose used. The ED50 for primordial
follicle loss was 0.02 mg/kg for dimethylbenzanthracene and 0.045 mg/kg for 3-
methylcholanthrene. In rats, the 60 mg/kg (0.24 mmol/kg) dose of benzo[a]pyrene resulted in a
50% loss of primary and secondary follicles, but a much smaller decrease in primordial follicles
was seen (approximately 75% of control counts in the 15 and 60 mg/kg groups).
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MOA
benzo[a]pyrene is a complete carcinogen in that it both initiates and promotes tumor
formation. Several mechanistic processes have been associated with benzo[a]pyrene
carcinogenicity, including oxidative metabolism, which gives rise to reactive intermediates (see
Section 3.3), and formation of DNA adducts, both of which can lead to genotoxicity and
139
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
mutations in specific cancer-related genes. The ability of benzo[a]pyrene to function as a tumor
promoter may be related to cytotoxicity, AhR affinity, and upregulation of genes related to
biotransformation, growth, and differentiation.
The following sections discuss mechanistic evidence for possible key events in the MOA
for cancer (a topic that is further discussed in Section 4.7.3). Information regarding MOAs for
noncancer effects noted above is discussed in Section 4.6.3.
4.5.1. Genotoxicity
The ability of benzo[a]pyrene to cause mutations and other forms of DNA damage in
both in vivo and in vitro studies is well documented (see Tables 4-24, 4-25 and 4-26). With
metabolic activation (inclusion of S9) benzo[a]pyrene is consistently mutagenic in the
prokaryotic Salmonella/Ames and E. coli assays (Table 4-24). A rare exception was observed, in
which benzo[a]pyrene did not induce mitotic recombination in eukaryotic S. cerevisiae
regardless of the presence of S9. In mammalian in vitro studies, benzo[a]pyrene is consistently
mutagenic, clastogenic and induces cell transformation both with and without metabolic
activation (Table 4-25). Cytogenetic damage in the form of chromosomal aberrations,
micronuclei, sister chromatid exchanges and aneuploidy are commonplace following
benzo[a]pyrene exposure as are DNA adduct formation, single strand breaks, and induction of
DNA repair and unscheduled DNA synthesis. The in vitro mammalian cell assays were
conducted in various test systems, including human cell lines.
In in vivo studies, benzo[a]pyrene consistently tested positive in multiple species and
strains and under various test conditions in the following assays: cell transformation,
chromosomal aberrations, DNA adducts, DNA strand breaks, micronuclei formation, gene
mutations (H-ras, K-ras, p53, /acZ, Hprt), sister chromatid exchanges, sperm abnormality, and
unscheduled synthesis. Negative results were nominally interspersed throughout the in vivo
mammalian assays, except for consistently negative results observed for unscheduled DNA
synthesis.
In human in vivo studies, exposures were to mixed PAHs through cigarette smoke or
occupational exposure. In a subset of these studies, benzo[a]pyrene-specific DNA adducts have
been detected, and it has been demonstrated qualitatively that benzo[a]pyrene metabolites
damage DNA in exposed humans (see Table 4-26).
140
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-24. In vitro genotoxicity studies of benzofalpyrene in non-mammallian cells
Result
Reference
+S9
- S9
Endpoint/Test System: prokaryotic cells
Forward mutation
S. typhimurium TM677
+
-
Rastetter et al. (1982)
S. typhimurium TM677
+
ND
Babson et al. (1986)
Reverse mutation
S. typhimurium TA1537; TA1538
+
-
Ames et al. (1973)
S. typhimurium TA1535
-
-
Ames et al. (1973)
S. typhimurium TA98; TA1538
+
ND
Ames et al. (1975)
S. typhimurium TA1537; TA1538
+
-
Glatt et al. (1975)
S. typhimurium TA 1535
-
-
Glatt et al. (1975)
S. typhimurium TA98; TA100; TA1538
+
ND
McCann et al. (1975)
S. typhimurium TA 1535
-
ND
McCann et al. (1975)
S. typhimurium TA1538
+
ND
Egert and Greim (1976)
S. typhimurium TA1537
+
ND
Oeschetal. (1976)
S. typhimurium TA1538, TA98
+
-
Wood et al. (1976)
S. typhimurium TA98; TA100; TA1537
+
-
Epleretal. (1977)
S. typhimurium TA1535
-
-
Epleretal. (1977)
S. typhimurium TA98; TA100
+
-
Obermeier and Frohberg (1977)
S. typhimurium TA100
+
ND
Tang and Friedman (1977)
S. typhimurium TA98
+
-
Pitts et al. (1978)
S. typhimurium TA100
+
ND
Bruce and Heddle (1979)
S. typhimurium TA98, TA100
+
ND
LaVoie et al. (1979)
S. typhimurium TA1538
+
-
Rosenkranz and Poirier (1979)
S. typhimurium TA98, TA100
+
-
Simmon etal. (1979a)
S. typhimurium TA98
+
ND
Hermann (1981)
S. typhimurium TA98, TA100
+
ND
Alfheim and Randahl (1984)
S. typhimurium TA100
+
ND
Norpothetal. (1984)
S. typhimurium TA98, TA100, TA 1538
ND
-
Glatt et al. (1985)
S. typhimurium TA97, TA98, TA100
+
-
Sakai et al. (1985)
141 DRAFT - DO NOT CITE OR QUOTE
-------
S. typhimurium TA100
+
-
Carver etal. (1986)
S. typhimurium TA98
+
-
Alzieu et al. (1987)
S. typhimurium TA97, TA98, TA100, TA1537
+
-
Glatt et al. (1987)
S. typhimurium TA97, TA98, TA100
+
ND
Marino (1987)
S. typhimurium TA100
+
ND
Pahlman and Pelkonen (1987)
S. typhimurium TA 98, TA100
+
-
Prasanna et al. (1987)
S. typhimurium TA98
+
ND
Ampy et al. (1988)
S. typhimurium TA98, TA100
+
ND
Bos et al. (1988)
S. typhimurium TA98
+
ND
Lee and Lin (1988)
S. typhimurium TA100
+
ND
Phillipson and Ioannides (1989)
S. typhimurium TA98
+
ND
Antignac etal. (1990)
S. typhimurium TA98
-
ND
Gao et al. (1991)
S. typhimurium TA98
+
ND
Balansky et al. (1994)
S. typhimurium TA100
-
ND
Balansky et al. (1994)
DNA damage
E. co//'/pol A
+
-
Rosenkranz and Poirier (1979)
E. co/z'/differential killing test
+
-
Tweats (1981)
E. coli WP2-WP100/rec-assay
+
ND
Mamber et al. (1983)
E. coli!SOS chromotest Pq37
+
-
Mersch-Sundermann et al. (1992)
Endpoint/Test System: Non-mammalian eukaryotes
Mitotic recombination
S. cerevisiae D4-RDII
ND
-
Siebert et al. (1981)
S. cerevisiae D3
-
-
Simmon (1979b)
142 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
Table 4- 25. In vitro genotoxicity studies of benzofalpyrene in mammalian cells
Assay!Test System
Result
Reference
+S9
- S9
Forward Mutation
Human AHH-1 lymphoblastoid cells
ND
+
Danheiser et al. (1989)
Human lymphoblast (AHH-1) cells (Hprt)
ND
+
Crespi et al. (1985)
Human lymphoblastoid (AHH-1) cell line
ND
+
Chenetal. (1996)
Human fibroblast (MRC5CV1) cell line (Hprt)
—
ND
Hanelt et al. (1997)
Human lymphoblast (TK) cells
ND
+
Barfknecht et al. (1982)
Human lymphoblast (TK6) cells
+
ND
Crespi et al. (1985)
Human embryonic epithelial (EUE) cells
ND
+
Rocchi et al. (1980)
Mouse L5178Y/HGPRT
+
—
Clive et al. (1979)
Mouse lymphoma (L5178Y/TK+/-) cells
+
—
Clive et al. (1979)
Mouse lymphoma (L5178Y/TK+/-) cells
+
ND
Amacheretal. (1980);
Amacher and Turner (1980)
Mouse lymphoma (L5178Y/TK+/-) cells
+
—
Amacher and Paillet (1983)
Mouse lymphoma (L5178Y/TK+/-) cells
+
ND
Arce et al. (1987)
Human HSC172 lung fibroblasts
+
—
Gupta and Goldstein (1981)
Human Q3-wp normal lung keratinocytes
+
ND
Allen-Hofmann and Rheinwald (1984)
Human SCC-13Y lung keratinocytes
ND
+
Allen-Hofmann and Rheinwald (1984)
Chinese hamster ovary (CHO) cells (aprt)
+
ND
Yang et al. (1999)
Chinese hamster ovary cells (5 marker loci)
+
+
Gupta and Singh (1982)
Chinese hamster V79 cells (Co-cultured with irradiated HepG2 cells)
+
ND
Diamond et al. (1980)
Chinese hamster V79 lung epithelial cells
+
ND
Huberman (1976)
Chinese hamster V79 lung epithelial cells
+
ND
Arce et al. (1987)
Chinese hamster V79 lung epithelial cells
+
ND
O'Donovan (1990)
Rat/Fischer, embryo cells/OuaR
ND
+
Mishra et al. (1978)
DNA damage
DNA adducts
143
DRAFT - DO NOT CITE OR QUOTE
-------
Human fibroblast (MRC5CV1) cell line
+
ND
Haneltetal. (1997)
Human peripheral blood lymphocytes
ND
+
Wienke et al. (1990)
Hamster tracheal cells
ND
+
Roggeband et al. (1994)
Chinese hamster V79 lung epithelial cells
+
ND
Arce et al. (1987)
Virus transformed Syrian hamster embryo and mouse C3H10T1/2 cells
ND
+
Arce et al. (1987)
Mouse lymphoma (L5178Y/TK+/-) cells
+
ND
Arce et al. (1987)
Rat tracheal cells
ND
+
Roggeband et al. (1994)
DNA damage/single strand breaks
Human fibroblast (MRC5CV1) cell line
+
ND
Haneltetal. (1997)
Human hepatoma (HepG2) cell line
ND
+
Tarantini et al. (2009)
Human prostrate carcinoma (DU145) cell line
ND
+
Nwagbara (2007)
Mouse embryo fibroblast (C3H/10T1/2 CL 8) cells
ND
+
Lubet et al. (1983)
Rat C18 trachea epithelial cells
ND
+
Cosma and Marchok, 1988;
Cosmaetal. (1988)
Rat lymphocytes
ND
+
Gao et al. (1991)
Unscheduled DNA synthesis
HeLa cells
+
ND
Martin etal. (1978)
Human fibroblasts
+
ND
Agrelo and Amos (1981)
Human fibroblasts
+
—
Robinson and Mitchell (1981)
Human HepG2
[+]
Valentin-Severin et al. (2004)
Hamster Primary embryo cells
ND
+
Casto et al. (1976)
Hamster tracheal cells
ND
+
Roggeband et al. (1994)
Rat Hepatocytes
[+]
Michalopoulos et al. (1978)
Rat tracheal cells
ND
—
Roggeband et al. (1994)
DNA repair
Human mammary epithelial cells
ND
+
Leadonetal. (1988)
Human skin fibroblasts
ND
+
Milo etal. (1978)
Baby hamster kidney (BHK21/cl3) cells
[+]
Feldmanetal. (1978)
144 DRAFT - DO NOT CITE OR QUOTE
-------
secondary mouse embryo fibroblasts (C57BL/6) and human lymphocytes
[+]
Shinohara and Cerutti (1977)
Rat/F344 hepatocytes
ND
+
Williams et al. (1982)
Cytogenetic damage
Chromosomal aberrations
Human blood cells
ND
+
Salama et al. (2001)
Human WI38 fibroblasts
+
—
Weinstein et al. (1977)
Chinese hamster CHL
+
—
Matsuoka et al. (1979)
Chinese hamster V79-4 lung epithelial cells
—
—
Popescu et al. (1977)
Mouse lymphoma (L5178Y/TK+/-) cells
+
ND
Arce et al. (1987)
Rat Liver RL1 cells
+
ND
Dean (1981)
Micronuclei
Human AHH-1 lymphoblastoid cells
ND
+
Crofton-Sleigh et al. (1993)
Human HepG2 liver cells
ND
+
Wu et al. (2003b)
Human lymphoblastoid (TK) cells
ND
+
Fowler etal. (2010)
Human MCL-5 lymphoblastoid cells
ND
+
Crofton-Sleigh et al. (1993)
Human peripheral blood lymphocytes
+
ND
Lo Jacono et al. (1992)
Chinese hamster V79 cells
ND
+
Whitwell et al. (2010)
Chinese hamster V79-MZ cells
ND
+
Matsuoka et al. (1999)
Sister chromatid exchanges
Human C-HC-4 and C-HC-20 hepatoma cells
ND
+
Abe et al. (1983a,b)
Human diploid fibroblast (TIG-II) cell line
+
+
Huh etal. (1982)
Human fibroblasts
ND
+
Juhletal. (1978)
Human blood cells
ND
+
Salama et al. (2001)
Human peripheral blood lymphocytes
ND
+
Rudiger et al. (1976)
Human peripheral blood lymphocytes
ND
+
Craig-Holmes and Shaw (1977)
Human peripheral blood lymphocytes
ND
+
Schoenwald et al. (1977)
Human peripheral blood lymphocytes
ND
+
Wienke et al. (1990)
Human peripheral blood lymphocytes
ND
+
Wienke et al. (1990)
145 DRAFT - DO NOT CITE OR QUOTE
-------
Human peripheral blood lymphocytes
+
—
Tohdaetal. (1980)
Human peripheral blood lymphocytes
+
ND
Lo Jacono et al. (1992)
Chinese hamster Don-6 cells
ND
+
Abe et al. (1983a,b)
Chinese hamster V79 lung epithelial cells
+
—
Popescu et al. (1977)
Chinese hamster V79 lung epithelial cells
+
ND
Mane et al. (1990)
Chinese hamster V79 lung epithelial cells
+
ND
Wojciechowski et al. (1981)
Chinese hamster V79 lung epithelial cells
+
ND
Arce et al. (1987)
Chinese hamster V79 lung epithelial cells
ND
+
Kulkaetal. (1993)
Chinese hamster ovary (CHO) cells
+
—
de Raat (1979)
Chinese hamster ovary (CHO) cells
+
—
HusgafVel-Pursiainen et al. (1986)
Chinese hamster ovary (CHO) cells
ND
+
Wolff and Takehisa (1977)
Chinese hamster ovary (CHO) cells
ND
+
Pal et al. (1978)
Hamster Chi cells
ND
+
Shimizu et al. (1984)
Rabbit peripheral blood lymphocytes
ND
+
Takehisa and Wolff (1978)
Rat ascites hepatoma AH66-B
ND
+
Abe et al. (1983a,b)
Rat esophageal tumor R1
ND
+
Abe et al. (1983a,b)
Rat hepatocyte (immortalized) cell lines (NRL cl-B, NRL cl-C and ARL)
+
ND
Kulkaetal. (1993)
Rat hepatoma (Reuber H4-II-E) cells
ND
+
Deanetal. (1983)
Rat liver cell line ARL 18
ND
+
Tong et al. (1981)
Rat pleural mesothelial cells
ND
+
Achard et al. (1987)
Aneuploidy
Chinese hamster V79-MZ cells
ND
+
Matsuoka et al. (1998)
Cell transformation
Human BEAS-2B lung cells
+
van Agen et al. (1997)
Human breast epithelial (MCF-10F, MCF-7, T24) cell lines
ND
+
Calafetal. (1993)
Baby hamster kidney (BHK21/cl3) cells
+
ND
Greb et al. (1980)
Golden hamster embyro cells
+
ND
Mager et al. (1977)
Syrian hamster embryo (SHE) cells
ND
+
DiPaolo et al. (1969, 1971)
Syrian hamster embryo cells
ND
+
Dunkel et al. (1981)
146 DRAFT - DO NOT CITE OR QUOTE
-------
Syrian hamster embryo cells
+
LeBoeuf et al. (1996)
Syrian hamster embyro (SHE) cells/focus assay
ND
+
Castoetal. (1977)
Fetal Syrain hamster lung (FSHL) cells
ND
+
Emuraetal. (1980, 1987)
Virus infected rat embryo RLV/RE and RAT cells; mouse embryo AKR/Me cells; Syrain
hamster embryo cells
+
Heidelberger et al. (1983)
Virus transformed Syrian hamster embryo and mouse C3H10T1/2 cells
ND
+
Arce et al. (1987)
Mouse C3H/10T1/2 embryo fibroblasts
+
Nesnowetal. (2002, 1997)
Mouse embryo fibroblast (C3H/10T1/2 CL 8) cells
ND
+
Peterson et al. (1981)
Mouse embryo fibroblast (C3H/10T1/2 CL 8) cells
ND
+
Lubet et al. (1983)
Mouse SHE cells; BALB/c-3t3 cells; C3H/10T1/2 cells; prostate cells
+
Heidelberger et al. (1983)
Mouse BALB/c-3T3 cells
ND
+
Dunkel et al. (1981)
Mouse BALB/c-3T3 cells
+
Matthews (1993)
Mouse BALB/c-3T3 clone A31-1-1
ND
+
Little and Vetroys (1988)
Rat embyro cells/SA7 virus transformation
ND
+
DiPaolo and Casto (1976)
Rat/Fischer, embryo cells (leukemia virus transformed)
ND
+
Dunkel et al. (1981)
Rat/Fischer, embryo cells/OuaR
ND
+
Mishra et al. (1978)
Key: [+] = S9 status not given; "+" = positive; = negative; ND = not determined.
1
2
147
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-26. In vivo genotoxicity studies of benzo|alpyrene
Endpoint
Test System
Test Conditions
Results
Dose
Comment
Reference
Mutations
Human/Blood T
lymphocytes (smokers
and nonsmokers); HPRT
locus mutation assay
T-cells of lung cancer patients
(smokers and nonsmokers from
lung cancer patients and population
controls with known smoking
status) analyzed for HPRT locus
mutations.
+
Smokers and
nonsmokers
Splicing mutations, base-
pair substitutions, frameshift
and deletion mutations
observed. Smokers and
nonsmokers had GC~>TA
transversions (13% and 6%,
respectively) and GC~>AT
transitions (24% and 35%,
respectively) in HPRT gene
consistent with in vitro
mutagenicity of
benzo[a]pyrene
Hackman et al.
(2000)
Mutations
Mouse/strains :T-stock,
(SECxC57BL)Fi,
(C3Hxl01)Fi,
(C3HxC57BL)Fi for
females; (101xC3H)F,
or (C3Hxl01)Fi for
males; dominant-lethal
mutation assay
12-wk old males dosed with
benzo[a]pyrene i.p. and mated 3.5-
6.5 days post-treatment with 12-wk
old females from different stocks;
sacrificed on days 12-15 after
vaginal plug was observed;
females kept in a 5 hr-dark phase
to synchronize ovulation 5 wks
before the start of the expt.;
fertilized eggs collected from 9-11
hrs after mating and first-cleavage
metaphase chromosomes prepared
20 hrs after mating
+
500 mg/kg
b.w.
The % of dominant lethal
mutations were in the order
of T-stock= (C3Hxl01)F! >
(SECxC57BL)Fi>
(C3HxC57BL)F i
Generoso et al.
(1979)
Mutations, GC
Mice/strains: Male
stocks: (101xC3H)Fi;
Female stocks (A):
(101xC3H)Fl, (B):
(C3Hxl01)Fl, (C):
(C3HxC57BL)F 1,
(D): (SECxC57BL)F 1,
(E):T-stock females;
dominant lethal
mutations
In dominant lethal assay (DLA),
12-wk-old males dosed i.p. with
benzo[a]pyrene and mated with 10-
12 wk-old (#1) stock A females; or
(#2) stock B females on the day of
dosing; or with (#3a) with stocks
B, C and D females 3.5-7.5 days
post-dosing, or with (#3b) with
stocks B, C, D and E females 3.5-
6.5 days post-dosing. Control
group mated at time corresponding
to 1.5-4.5 days post-treatment in
positive
for
DLA;
negative
for HT
500 mg/kg
b.w.
Dominant lethal effects
were observed in early to
middle (4.5-5.5 and 6.5-7.5
days post-treatment,
respectively) spermatozoa
and in preleptotene
spermatocytes (32.5-33.5
and 34.5-35.5 days post-
treatment). In the HTA, no
significant differences
observed between treated
and control progeny.
Generoso et al.
(1982)
148
DRAFT - DO NOT CITE OR QUOTE
-------
the test groups.
Mutations, GC
Mice/strains: Male
stocks: (101xC3H)Fi;
Female stocks (A):
(101xC3H)Fl, (B):
(C3Hxl01)Fl, (C):
(C3HxC57BL)F 1,
(D): (SECxC57BL)F 1,
(E):T-stock females;
heritable translocations
For heritable translocation assay
(HTA), males were mated with
stocks B and D females 3.5-7.7
days post-benzo[a]pyrene
treatment and male progeny
screened for translocation
heterozygosity.
-
500 mg/kg
b.w.
No significant differences
observed between treated
and control progeny.
Generoso et al.
(1982)
Mutations, GC/
spot test
Mouse/C57BL female x
T-strain male; somatic
mutation assay
Mice mated for a 5-day period;
10V4 days post-appearance of
vaginal plug, females injected i.p.
withbenzo[a]pyrene or vehicle;
offspring (pups) scored for
survival, morphology and presence
of white near-midline ventral spots
(WMVS) and recessive spots (RS).
+
100 and 500
mg/kg b.w.
Induced coat color mosaics
represent genetic changes
(e.g. point mutations) in
somatic cells. WMVS and
RS represent melanocyte
cell killing and
mutagenicity, respectively.
Benzo[a]pyrene caused high
incidence of RS but did not
correlated with WMVS.
Russell
(1977)
Mutations
Mouse//acZ transgenic
(Muta™ Mouse)
benzo[a]pyrene given orally in
corn oil for 5 consecutive days;
sacrificed 14 days after last dosing;
Eleven organs analyzed for lacZ
MF
+
125
mg/kg/day
Highest MF observed in
colon followed by ileum >
forestomach > bone marrow
= spleen > glandular
stomach > liver =
lung>kidney = heart
Hakura et al.
(1998)
Mutations
Mouse/C57BL/6J Dlb-1
congenic; Dlb-1 locus
assay
Animals dosed i) i.p. with vehicle
orbenzo[a]pyrene 2, 4, or 6 doses
at 96 hr intervals; or ii) single dose
of benzo[a]pyrene given i.p. or p.o.
alone or 96 hours following a
single i.p. dosing with 10 |ig/kg
TCDD
+
40 mg/kg
b.w.
benzo[a]pyrene caused a
dose-dependent increase in
mutant frequency; i.p. route
showed higher mutant
frequency than p.o. route;
induction of mutations were
associated with Ah-
responsiveness.
Brooks et al.
(1999)
149
DRAFT - DO NOT CITE OR QUOTE
-------
Mutations (Hprt
locus)
Mice/C57BL/6 (lacz
negative and A7J1 and
XPAT lymphocytes
Gavage in corn oil 3 times/wk for
0, 1, 5, 9, 13 wks; sacrificed 7 wks
after last treatment
+
13 mg/kg
Mutation sensitivity:
XPA>XPA+,+
Bol et al.
(1998)
Mutations
Mouse/Cockayne
syndrome-deficient (Csb~
); heterozygous (Csb+/~)
and wild type controls
(Csb+/+); Hprt mutation
frequency assay
Csb-'VlacZ+/- and Csb+/7lacZ+/-
mice were dosed i.p. with
benzo[a]pyrene thrice a wk for 5,
9, or 13 wks; For Hprt MF analysis
mice were sacrificed 3 wks after
last treatment; spleenocytes
collected; For lac Z MF analysis
mice were sacrificed 3 days after
last treatment and liver, lung and
spleen collected.
+
13 mg/kg
lac Z MF detected in all
tissues but no differences
between WT and Csh"
mice; Hprt mutations
significantly higher in Csh"
mice than control mice.
BPDE-dGuo adducts in
Hprt gene are preferentially
removed in WT mice than
Csh" mice.
Wijnhoven et al.
(2000)
Mutations
Mouse/B6C3Fl,
forestomach H -ras, K-
ras & p53 mutations
benzo[a]pyrene given in feed in a
2-year chronic feeding study;
+
5, 25, 100
ppm
68%K-ras (codons 12,13),
10%H-ras (codon 13), 10%
p53 mutations; all G~>T
transversions
Culp et al.
(2000)
Mutations
Mous dlacZ/galE
(Muta™ Mouse); Skin
painting study
Mice topically treated with a single
dose or in five divided doses daily;
sacrificed 7 or 21 days after the
single or final treatment; DNA
from skin, liver and lung analyzed
for mutations.
+sk or -
Li,Lu
1.25 or 2.5
mg/kg (25 or
50 |j.g/mouse)
Skin showed significant
dose- and time-dependent
increase in mutation
frequency; liver and lung
showed no mutations; MF
for single or multiple-dose
regimens were similar.
Dean et al.
(1998)
Mutations/ spot
test
Mouse/T-strain
benzo[a]pyrene given to pregnant
mice by gavage in 0.5 ml corn oil
on GDs 5-10
+
10 mg/mouse
(5x2 mg)
Davidson and
Dawson (1976)
Mutations (Hprt
locus)
Mouse, 129/Ola (Wild
type); splenic T
lymphocytes
Single i.p. injection followed by
sacrifice 7 wks post-treatment
+
0, 50, 100,
200, 400
mg/kg
dose-dependent increase in
HprtMF
Bol et al.
(1998)
Mutations
Mouse, A/J, male
Single i.p. injection followed by
sacrifice 28 days post-treatment
+
0, 0.05, 0.5, 5
50 mg/kg
Dose-dependent increase in
lung tissue K-ras codon 12
G-->T mutation frequency
Meng et al.
(2010)
Mutations/ gene
Mouse/CD-I; skin
papillomas (Ha-ras
mutations)
Female mice were initiated
topically with a single dose of
benzo[a]pyrene and 1 wk after
+
600
nmol/mouse
About 90% of papillomas
contained Ha-ras mutations,
all of them being
Colapietro et al.
(1993)
150 DRAFT - DO NOT CITE OR QUOTE
-------
initiation promoted twice weekly
with 5 nmol TPA for 14 wks. One
month after stopping TPA
application, papillomas collected
and DNA from 10 individual
papillomas were analyzed for Ha-
ras mutations by PCR and direct
sequencing.
transversions at codons 12
(20% GGA~>GTA), 13
(50% GGC~>GTC), and 61
(20% CAA~> CTA).
Mutation///?
vivo-in vitro
Rat, Wistar
Single dose by gavage; urine and
feces collected 0-24, 24-48 and 48-
72 hrs post-treatment; urine and
extracts of feces tested in S.
typhimurium TA100 strain with or
without S9 mix and (3-
glucuronidase
+
0, 1,5, 10,
100 mg/kg
Fecal extracts and urine
showed mutagenicity at and
above 1 and 10 mg/kg b.w.
Benzo [a]pyrene,
respectively. Highest
mutagenic activity observed
for 0-24 hrs post-treatment
for feces and 24-48 hrs post-
treatment for urine with (3-
glucuronidase ± S9 mix.
Willems et al
(1991)
Mutations, GC
/gene
D. melanogaster/ sex-
linked recessive lethal
test
Base males exposed to
benzo[a]pyrene were mated with
virgin females of Berlin K or mei-
9U strains;
±
10 mM
Data inconclusive due to
low fertility rates of mei-9u
females.
Vogel et al.
(1983)
Mutations, GC
/gene
D. melanogaster/ sex-
linked recessive lethal
test
Adult Berlin males treated orally
with benzo [a]pyrene
+
5 or 7.5 mM
Low mutagenic activity
Vogel et al.
(1983)
Mutations, GC
/gene
D. melanogaster/Berlin-
K and Oregon-K strains;
sex-linked recessive
lethal test
benzo[a]pyrene dissolved in
special fat and injected into the
abdomen of flies.
-
2 and 5 mM
Negative at both doses
Zijlstra and
Vogel
(1984)
Mutations, GC
/gene
D. melanogaster/ sex-
linked recessive lethal
test
Male Berlin K larvae treated with
benzo [a]pyrene for 9-11 days
+
0.1-4 mM
Threefold enhancement in
lethals in treated versus
controls
Vogel et al.
(1983)
151
DRAFT - DO NOT CITE OR QUOTE
-------
Mutations, GC
/gene
D. melanogaster/Canton-
S (WT) males, FM6
(homozygous for an X
chromosome) females;
sex-linked recessive
lethal test
Adult male flies were fed on filters
soaked inbenzo[a]pyrene for 48 or
72 hrs; Treated and control males
mated with FM6 females, males
transferred to new groups of
females at intervals of 3, 2, 2, and
3 days; four broods obtained; a
group of 100 daughters of each
male were mated again; scored for
% lethal
-
250, 500 ppm
Authors report incomplete
dissolution of
benzo[a]pyrene in DMSO as
a possible cause of negative
result.
Valencia and
Houtchens
(1981)
Mutation/gene
D. melanogaster;
somatic mutation - eye
color mosaicism
50 females and 20 females were
mated in a culture bottle for 48 hrs
allowing females to oviposit;
adults then discarded and the eggs
allowed to hatch; larvae fed on
benzo[a]pyrene deposited on food
surface and the emerging adult
males scored for mosaicly colored
eye sectors;
+
1, 2, or 3 mM
benzo[a]pyrene was
effective as a mutagen; no
dose-response observed
Fahmy and
Fahmy (1980)
DNA adducts
Human/white blood cells
Workers were exposed for 6-8
hrs/day for at least 4-6 months
before blood collection; leukocyte
DNA isolated, digested and
benzo[a]pyrene tetrols analyzed by
HPLC with fluorescent detection
(HPLC-FD). Low, medium, and
high exposure groups correspond
to < 0.15, 0.15 to 4, > 4 mg/m3 of
benzo[a]pyrene, respectively.
+
<0.15,0.15
to 4, > 4
|j,g/m3 of
benzo[a]pyre
ne
PAH exposure, CYP1A1
status and smoking
significantly affected DNA
adduct levels, i.e.
CYPlAl(*l/*2 or *24/*2a)
>CYP1A1*1/*1;
occupational >
environmental exposure;
smokers > nonsmokers;
adducts increased with dose
and duration of smoking
Rojas et al.
(2000)
DNA adducts
Human/white blood cells
Coke oven workers were exposed
to PAHs and benzo[a]pyrene-WBC
DNA analyzed by HPLC-FD for
BPDE-DNA adducts
+
0.14 |J.g/m3
BPDE-DNA adducts
detectable; no significant
difference between smokers
and nonsmokers; no
correlation with air
benzo[a]pyrene levels and
adduct levels
Mensing et al.
(2005)
152
DRAFT - DO NOT CITE OR QUOTE
-------
DNA adducts
and Mutations
Mouse/C57BL/6 lacZ
transgenic
Mice dosed with single i.p.
injection of benzo[a]pyrene in
DMSO; sacrificed 1, 3, 5, 7, 14,
21, and 28 days post-treatment;
spleen, lung, liver, kidney and
brain collected, DNA isolated and
analyzed for mutations in lacz
reporter gene in E. coli and adducts
by 32P-postlabeling assay.
+
50 mg/kg
b.w.
BPDE-dG adduct levels
peaked between 5 and 7
days post-treatment,
followed by gradual decline;
rate of removal highest in
lung, liver and spleen and
lowest in kidney and brain;
mutant frequencies peaked
between 7 and 14 days in
lung, spleen, liver and
kidney; brain was not
significant at any time point.
Boerrigter
(1999)
DNA adducts
Mice/
(Ahr+/+, Ahr+/~
Gavage; sacrificed 24 hr post-
treatment
+
100 mg/kg
b.w.
No induction of CYP in
Ahr . but all alleles positive
for adduct formation
Sagredo et al
2006
DNA adducts
Mice/C57BL/6J
Cyplal(+/-) and
Cyplal(-/-)
Single i.p. injection; sacrificed 24
hrs post-treatment; liver DNA
analyzed by 32P-postlabeling assay
+
500 mg/kg
b.w.
BPDE-DNA adduct levels
4-fold higher in Cyplal(-/-)
mice than Cyplal(+/-) mice
Uno et al.
(2001)
DNA adducts
Mouse/B6C3Fl
benzo[a]pyrene fed in diet for 4
(100 ppm) or for 1, 2, 8, 16 and 32
wks (5 ppm); sacrificed and liver,
lungs, forestomach, small intestine
collected; DNA analyzed by 32P-
postlabeling assay
+
5 ppm (32
wks) and 100
ppm (4 wks)
Linear dose-response in 4-
wk study; the 5 ppm groups
showed a plateau after 4
wks of feeding
Culp et al.
(2000)
DNA adducts
Mouse/Balb/c;
Single i.p. injection; sacrificed 12
hrs post-injection; liver and
forestomach collected; DNA
binding of [3H]benzo[a]pyrene
analyzed by scintillation counting.
+
140 |LtCi/100
gb.w.
Liver DNA had 3-fold
higher binding of
benzo[a]pyrene than that of
forestomach
Gangar et al.
(2006)
DNA adducts
Mice/BALB/cAnN
(BALB), CBA/JN
(CBA); 32P-postlabeling
assay
Animals dosed i.p. with or without
24 hr pretreatment with TCDD
+
50 and 200
mg/kg
Adduct levels similar in
both strains dosed with
benzo[a]pyrene alone.
TCDD pre-treatment had a
greater suppressive effect on
Wu et al.
(2008)
153 DRAFT - DO NOT CITE OR QUOTE
-------
adduct formation in B ALB
relative to CBA mice at low
dose but resulted in no
significant difference in
adduct levels at high dose.
DNA adducts
Mice/BALB/c, skin;
Four doses of benzo[a]pyrene
topically applied to the shaved
backs of animals at 0, 6, 30 and 54
hrs; sacrificed 1 day after last
treatment; DNA analyzed by 32P-
postlabeling assay
+
4x 1.2
Mmol/animal
Five adducts spots detected
Reddy et al.
(1984)
DNA adducts
Mice/Swiss, epidermal
and dermal skin
Single topical application on
shaved backs; sacrificed 1,3, and 7
days post-treatment; epidermal and
dermal cells separated; DNA
isolated, digested with DNAsel
and estimated DNA binding;
adducts separated by HPLC.
+
250 nmol in
150 Ml
acetone
Both cells positive for
benzo[a]pyrene adducts;
epidermis>dermis; adducts
persisted up to 7 days with a
gradual decline in levels
Oueslati et al.
(1992)
DNA adducts
Rats/CD, PBLs, lungs
and liver
Single i.p. injection; sacrificed 3
days post-treatment; DNA
analyzed by Nuclease Pl-
endhanced 32P-postlabeling assay.
+
2.5
mg/animal
BPDE-dG as major adducts
and several minor adducts
detected in all tissues
Ross et al.
(1991)
DNA adducts
Rats/Sprague-Dawley,
liver
Single i.p. injection followed by
sacrifice at 4 hours post-treatment;
liver DNA isolated and analyzed
by 32P-postlabeling assay.
+
100 mg/kg
b.w.
Two adduct spots detected
Reddy et al.
(1984)
DNA adducts
Rat/Lewis; lung and
liver
Animals received a single oral
dose of benzo[a]pyrene in
tricaprylin; sacrificed 1, 2, 4, 11,
and 21 days post-dosing; analyzed
liver and lung DNA for BP-DNA
adducts by 32P-postlabeling assay
and urine for 8-oxodG adducts by
HPLC-ECD.
+
10 mg/kg
BPDE-dG levels peaked 2
days after exposure in both
tissues, higher in lungs than
liver at all time points,
decline faster in liver than
lung; Increased 8-oxodG
levels in urine and
decreased levels in liver and
lung.
Briede et al.
(2004)
154
DRAFT - DO NOT CITE OR QUOTE
-------
DNA adducts
Rats/F344; 32P-
postlabeling assay
benzo[a]pyrene given in the diet
for 30, 60, or 90 days; animals
sacrificed and liver and lung
isolated and DNA extracted and
analyzed for adducts.
+
0, 5, 50, 100
mg/kg
Adduct levels linear at low
and intermediate doses,
nonlinear at high dose;
Ramesh and
Knuckels (2006)
DNA adducts
Rats/Wistar; liver and
PBL adducts
Single dose by gavage; sacrificed
24 hrs post-dosing; PBL and liver
DNA analyzed by 32P-postlabeling
for BP-DNA adducts
+
0, 10, 100
mg/kg b.w.
At 100 mg dose total adduct
levels in PBL were twofold
higher than the levels in
liver; adduct profiles
differed between PBL and
liver
Willems et al
(1991)
DNA strand
breaks
Rats/Sprague-Dawley;
Comet assay
instilled intratracheally with (i)
single dose of benzo[a]pyrene in
aqueous suspension; sacrificed at
3, 24, 48 hrs post-treatment;
alveolar macrophages, lung cells,
lymphocytes, hepatocytes collected
(ii) dose-response study and
sacrificed at 24 hours post-
treatment; lungs collected;
Controls received normal saline
instillation; All cells analyzed by
comet assay.
+
Expt#l: 3 mg
of
benzo[a]pyre
ne; Expt#2:
dose-response
study with
0.75, 1.5, 3
mg
benzo[a]pyre
ne
All time points showed
significant increase in SSB
(Expt#l); A dose-response
in SSB observed (Expt#2)
Garry et al.
(2003a,b)
DNA strand
breaks
Aquatic organisms:Carp
(Cyprinus carpio),
rainbow trout
(Oncorhynchus mykiss),
and clams (Spisula
sachalinensis); Comet
assay
All organisms acclimatized in
tanks for 2 days, water changed
every 24 hrs; exposed to
benzo[a]pyrene in DMSO in a
tank; one third volume of tank
contents changed every 12 hrs;
organisms sacrificed at 24, 48, 72,
and 96 hrs post-treatment; cell
suspensions prepared from liver
(carp and trout) or digestive gland
(clam) for comet assay
+
0.05,0.25,
0.5 and 1
ppm
Significant dose-response
for strand breaks observed;
carp and trout liver showed
highest response at 48 hrs
and clam digestive gland
showed time-dependent
increase at highest conc.
Kim and Hyun
(2006)
DNA strand
breaks
Rat, Brown Norway
UDS determined after 5 and 18 hrs
of a single i.g. dosing
-
62.5 mg/kg
negative at both time points
Mullaart et al.
(1989)
155 DRAFT - DO NOT CITE OR QUOTE
-------
Unscheduled
DNA synthesis
Rats/F344;
Single i.p. injection of
benzo[a]pyrene or DMSO;
sacrificed at 2 or 12 hrs post-
exposure; liver isolated, hepatocyte
cultures were setup and incubated
with 10 mCi/ml 3H-thymidine for 4
hrs; washed and autoradiography
performed
-
100 mg/kg
b.w.
benzo[a]pyrene was
negative at both time points
Mirsalis et al.
(1982)
Unscheduled
DNA synthesis
Mouse/HO S :HR-1
hairless; skin
Single topical application on two
spots on the backs after stripping
stratum corneum with adhesive
tape to enhance penetration;
sacrificed 24 hr post-treatment,
skin isolated [3H]thymidine;
cultured in; epidermal UDS
measured
+
0, 0.25, 0.5
and 1% (w/v)
in acetone
UDS index showed a dose-
dependent increase up to
0.5%benzo[a]pyrene dose
and then plateaued
Mori et al.
(1999)
Unscheduled
DNA synthesis
Rat/Brown Norway;
liver
Single intragastric injection;
sacrificed at 5 and 18 hours post-
injection
-
62.5 mg/kg
b.w.
benzo[a]pyrene was
negative at both time points
Mullaart et al.
(1989)
Unscheduled
DNA synthesis
Mouse/(C3Hf x 101 )F,
hybrid, germ cells
i.p. injection of benzo[a]pyrene;
[3H]Thymidine injection later
-
0.3 mL
Concentration not specified
Sega
(1979)
Unscheduled
DNA synthesis
Mouse, early spermatid
i.p. injection
-
250-500
mg/kg b.w.
Reviewed by Sotomayor
and Sega (2000)
Sega
(1982)
Chromosomal
aberrations
Hamster/Chinese bone
marrow
Single, i.p. injection of
benzo[a]pyrene dissolved in
tricapryline; animals sacrificed 24
hours post-exposure
+
25, 50, 100
mg/kg b.w.
benzo[a]pyrene induced
CAs at 50 mg/kg/bw only,
with negative responses at
the low and high dose
Bayer
(1979)
Chromosomal
aberrations
Mice/C57 (high AHH
inducible) and DBA
(low AHH inducible)
strains; 11-day old
embryos; adult bone
marrows
Study used 4 matings
(femalexmale): C57xC57;
DBAxDBA; C57xDBA;
DBAxC57; Pregnant mice treated
orally on GDI 1 with
benzo[a]pyrene; sacrificed 15 hrs
post-treatment; material liver, bone
+
150 mg/kg
Levels of CAs: hybrid
embryos > homozygous
DBA embryos >
homozygous C57 embryos;
Tissue AHH activity: C57
mothers and their embryos >
DBA females and their
Adler et al.
(1989)
156 DRAFT - DO NOT CITE OR QUOTE
-------
marrow and placenta and embryos
collected; male mice dosed
similarly and bone marrows
collected; individual embryo cell
suspensions and bone marrow
preparations scored for CAs.
Tissue AHH activity measured.
homozygous embryos. No
quantitative correlation
between BP-induced CAs
and AHH inducibility. No
differences in bone marrow
mitotic index of males of
different strains between
control and treatment
groups.
Chromosomal
aberrations
Mouse/lC3Fl hybrid
(101/ElxC31xEl)Fi;
CAs in bone marrow
Single dose by gavage; sacrificed
30 hrs of post-dosing; bone
marrow from femur isolated and
analyzed for CAs
+
63 mg/kg
Significant increase in CAs
in benzo [a]pyrene-treated
animals compared to
controls.
Adler and
Ingwersen
(1989)
Chromosomal
aberrations
RatsAVistar; PBLs
Single dose by gavage; sacrificed
6, 24 and 48 hrs post-treatment;
blood from abdominal aorta
collected, whole blood cultures set
up, CAs scored in 100 first-
division PBLs per animal
-
0, 10, 100,
200 mg/kg
b.w.
No difference between
control and treatment
groups at any dose or at any
sampling time observed
Willems et al
(1991)
Micronuclei
Hamster, Chinese, bone
marrow
Single, i.p. injection of
benzo[a]pyrene dissolved in
tricaprylin; animals sacrificed 30
hours post-exposure
-
100, 300, 500
mg/kg b.w.
Bayer
(1979)
Micronuclei
Mice/B6C3F1 (hybrid);
I.p. injection; several doses given
to calculate LD50
+
232 mg/kg
(LD50/7);
259 mg/kg
(LD50/4)
Study conducted to
determine the toxicity of
benzo [a]pyrene (LD50)
Salamone et al.
(1981)
Micronuclei
Mouse/CD-I andBDFl;
bone marrow
Dosed orally once, twice or thrice
at 24 hr intervals; sacrificed 24 hrs
after last treatment
+
250, 500,
1000, 2000
mg/kg b.w.
significant increase at all
doses; no dose-response;
double dosing at 500 mg/kg
dose gave best response
Shimada et al.
(1990)
Micronuclei
Mouse/CD-I & BDF1,
peripheral blood
reticulocytes
Given single i.p injection; tail
blood collected at 24 hr intervals
from 0 to 72 hrs
+
62.5, 125,
250, 500
mg/kg b.w.
maximum response seen at
48 hrs post-treatment
Shimada et al.
(1992)
Micronuclei
Rat/Sprague-Dawley,
peripheral blood
reticulocytes
Given single i.p injection; tail
blood collected at 24 hr intervals
from 0 to 96 hrs
+
62.5, 125,
250, 500,
1000 mg/kg
b.w.
maximum response seen at
72 hrs post-treatment
Shimada et al.
(1992)
157 DRAFT - DO NOT CITE OR QUOTE
-------
Micronuclei
Mouse/ICR [Hsd:
(ICR)Br]
benzo[a]pyrene was heated in olive
oil and given orally as a single
dose; males, females and pregnant
mothers used; pregnant mice dosed
on GDs 16-17 and sacrificed on
GDs 17-18; micronuclei evaluated
in adult bone marrow and fetal
liver
+
150 mg/kg
b.w.
All groups significantly
higher than controls for
MN; Fetal liver more
sensitive than any other
group
Harper et al.
(1989)
Micronuclei
Mouse/Swiss albino;
bone marrow
Given orally in corn oil; sacrificed
24 hr post-exposure
+
75 mg/kg
b.w.
Koraktar et al.
(1993)
Micronuclei
Mouse/Swiss; bone
marrow PCE
Given by gavage and sacrificed 36
hrs post-treatment
+
75 mg/kg
b.w.
Rao and Nandan
(1990)
Micronuclei
Mice/CD-I and MS/Ae
strains
i.p. and p.o. administration
+
62.5, 125,
250, 500
mg/kg
good dose response by both
routes, strains; i.p. better
than p.o.; MS/Ae strain
more sensitive than CD-I
strain
Awogi and Sato
(1989)
Micronuclei
Mouse/BDFl, bone
marrow
Male and female mice aged 12-15
wks given single i.p. injection of
benzo[a]pyrene or corn oil;
sacrificed 24, 48, and 72 hrs post-
treatment; bone marrow smears
prepared, stained with May-
Grunwald-Giemsa technique and
scored for MN PCEs.
+
0, 25, 50, 60
mg/kg b.w.
Positive at all doses, time
points and sexes tested.
Dose-dependent increase in
MN observed in both sexes;
males responded better than
females; highest positive
response observed at 72 hrs
post-injection
Balansky et al.
(1994)
Micronuclei
Mouse, HRA/Skh
hairless, keratinocytes
Single topical application
+
0.5, 5, 50,
100, 500
mg/mouse
He and Baker
(1991)
Micronuclei
Mouse/HO S :HR-1,
hairless; skin
micronuclei
Topical application once daily for
3 days; sacrificed 24 hrs after last
treatment
+
0.4, 1,2,4
mg
Nishikawa et al.
(2005)
Micronuclei
Mice/HR-1 hairless, skin
(benzo[a]pyrene with
slight radiation)
+
Exposure to sunlight
simulator to evaluate
photogenotoxicity and
chemical exposure
Hara et al.
(2007)
Micronuclei
Rat, Sprague-Dawley,
pulmonary alveolar
macrophages
i.t. instillation, once/day for 3 days
+
25 mg/kg
b.w.
De Flora et al.
(1991)
158 DRAFT - DO NOT CITE OR QUOTE
-------
Micronuclei
Rat, Sprague-Dawley,
bone marrow cells
i.t. instillation, once/day for 3 days
-
25 mg/kg
b.w.
De Flora et al.
(1991)
Micronuclei
Fish (Carp, rainbow
trout, clams); Blood and
hemolymph
+
0.05, 0.25,
0.5 and 1
ppm
Kim and Hyun
(2006)
Sister
chromatid
exchanges
Hamster/Chinese; SCEs
in bone marrow
8-12 wk-old animals dosed with
two i.p. injections of
benzo[a]pyrene given 24 hrs apart;
animals sacrificed 24 hrs after last
treatment, bone marrow from
femur isolated and metaphases
analyzed.
+
450 mg/kg
b.w.
Significant increase in
metaphase SCEs in
benzo [a]pyrene-treated
animals compared to
vehicle-treated controls.
Roszinsky-
Kocher et al.
(1979)
Sister
chromatid
exchanges
Hamster/Chinese,
Animals implanted s.c. with
bromodeoxyuridine (BrdU) tablet;
2 hrs later given phorone (125 or
250 mg/kg) i.p.; another 2 hrs later
dosed i.p. withbenzo[a]pyrene; 24
hrs post-BrdU dosing, animals
injected with colchicine 10 mg/kg
b.w., sacrificed 2 hrs later; bone
marrow from femur prepared for
SCE assay
+
50 or 100
mg/kg b.w.
SCEs increased with low
dose of phorone
significantly.
Bayer et al.
(1981)
Sister
chromatid
exchanges
Hamster/Syrian, fetal
liver
i.p. injection to pregnant animals
on GDs 11, 13 or 15; fetal liver
SCEs were analyzed
+
50 and 125
mg/kg b.w.
Produced doubling of SCE
frequency
Pereira et al.
(1982)
Sister
chromatid
exchanges
Hamster/Chinese, bone
marrow
NA
+
2.5, 25, 40,
50, 75, 100
mg/kd b.w.
Frequency of SCEs
increased >40 mg/kg b.w.
Bayer
(1979)
Sister
chromatid
exchanges
Mouse/DBA/2 &
C57BL/6, bone marrow
cells
Two intragastric injections given;
mice implanted with BrdU tablets,
sacrificed on day 5, SCE estimated
+
10 or 100
mg/kg b.w.
SCEs and BP-DNA adducts
in the order of C57B1/6
(AHH-inducible) < DBA/2
(AHH-noninducible)
Wielgosz et al.
(1991)
Sister
chromatid
exchanges
Mouse/DBA/2 &
C57BL/6, spleenic
lymphocytes
Two intragastric injections given;
mice killed on 5th day and cells
cultured for 48 hrs with BrdU.
+
10 or 100
mg/kg b.w.
SCEs and BP-DNA adducts
in the order of C57B1/6
(AHH-inducible) < DBA/2
(AHH-noninducible)
Wielgosz et al.
(1991)
159 DRAFT - DO NOT CITE OR QUOTE
-------
Sister
chromatid
exchanges
Rats/Wistar; PBLs
Single dose by gavage; sacrificed
6, 24 and 48 hrs post-treatement;
blood from abdominal aorta
collected, whole blood cultures set
up, SCEs scored in 50 second-
division metaphases in PBLs per
animal
+
0, 10, 100,
200 mg/kg
Linear dose-response at any
sampling time, however,
significant at the highest
dose only; no interaction
between dose and sampling
time
Willems et al
(1991)
Cell
transformation
Hamsters/LVG:LAK
strain (virus free);
Transplacental host-
mediated assay
Pregnant animals dosed i.p. with
benzo[a]pyrene on GD 10;
sacrificed on GD 13, fetal cell
cultures prepared, 10x10s
cells/plate; 5 days post-culture
trypsinized; subcultured every 4-6
days thereafter and scored for
plating efficiency and
transformation.
+
3 mg/100 g
b.w.
Quarles et al.
(1979)
8-oxodG, 8-oxodeoxyguanosine; AHH, aryl hydrocarbon hydroxylase; benzo[a]pyrene, benzo[a]pyrene; BPDE, benzo[a]pyrene diol epoxide; BrdU,
bromodeoxyuridine; CAs, chromosomal aberrations; CSB, Cockayne syndrome; CYP, cytochrome P450; DLA, dominant lethal assay; DMSO,
dimethylsulfoxide; ECD, electrochemical detection; FD, fluorescence detection; FM6, First Multiple No. 6 is an X chromosome with a complex of inversions (to
suppress cross-over) and visible markers such as yellow body, white eyes and narrow eyes. GC, germline cell; GD, gestational day; HPLC, high performance
liquid chromatography; HPRT, hypoxanthine-guanine phosphoribosyl transferase; HTA, heritable translocation assay; i.p., intraperitoneal; i.t., intratracheal; Li,
Liver; Lu, Lung; MF, mutation frequency; PBL, peripheral blood lymphocytes; PCE, polychromatic erythrocytes; SCEs, sister chromatid exchanges; SFS,
synchronous fluorescence spectrometry; Sk Skin TCDD, 2,3,7,8-tetrachlorodibenzodioxin; TP A, 12-tetradecanoyl-O-phorbol acetate; UDS, unscheduled DNA
synthesis; USERIA, ultrasensitive enzyme radioimmunoassay; WT, wild-type; XP, xeroderma pigmentosum;
160
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
4.5.2. Metabolic pathways
Diol epoxide pathway
benzo[a]pyrene diol epoxide metabolites interact preferentially with the exocyclic amino
groups of deoxyguanine and deoxyadenine (Geacintov et al., 1997; Jerina et al., 1991). Adducts
may give rise to mutations unless these adducts are removed by DNA repair processes prior to
replication. The stereochemical nature of the diol epoxide metabolite (i.e., anti- vs. syn-diol
epoxides) affects the number and type of adducts and mutation that occurs (Geacintov et al.,
1997). Transversion mutations (e.g., GC—>TA or AT—>TA) are the most common type of
mutation found in mammalian cells following diol epoxide exposure (Bostrom et al., 2002).
Strong evidence for the association between benzo[a]pyrene activation by the diol
epoxide pathway and key DNA-reactive and mutational events associated with tumor initiation
comes from the following observations: (1) (+)-anti-BPDE is very reactive with guanine
residues in DNA (Koreeda et al., 1978; Jeffrey et al., 1976); (2) (+)-anti-BPDE is more potent
than benzo[a]pyrene, benzo[a]pyrene phenols, and benzo[a]pyrene diols in mutagenicity assays
in bacterial and mammalian cells (Malaveille et al., 1977; Newbold and Brookes, 1976);
(3) When administered by ip injection to newborn mice, (+)-anti-BPDE is more potent than
benzo[a]pyrene phenols and benzo[a]pyrene diols and much more potent than benzo[a]pyrene
itself in lung tumorigenicity assays (Chang et al., 1987; Buening et al., 1978; Kapitulnik et al.,
1978); (4) (+)-anti-BPDE treatment resulted in ras gene codon 12 G—>T point mutations, the
activation of the H-ras-1 proto-oncogene and transformation of NIH/3T3 cells (Marshall et al.,
1984); (5) (+)-anti-BPDE forms DNA adducts at specific "hotspots" in the p53 tumor suppressor
gene that are commonly mutated in lung and other cancer patients (Denissenko et al., 1996;
Puisieux et al., 1991); (6) lung tumors from nonsmoking patients who were chronically exposed
to smoky coal emissions contained mutated p53 and showed a spectrum of mutations consistent
with (+)-anti-BPDE-associated mutations in the K-ras oncogene (DeMarini et al., 2001); (7)
elevated blood BPDE-DNA adducts have been observed in coke oven workers and chimney
sweeps, occupations associated with increased risks of cancer from PAH-containing complex
mixtures (Pavanello et al., 1999); (8) the spectrum of mutation in the K-ras, H-ras, and p53 genes
in forestomach tumors of mice fed benzo[a]pyrene in the diet for 2 years was consistent with (+)-
anti-BPDE DNA reactions (Culp et al., 2000); (9) K-ras mutations found in lung tumors from
A/J mice given single i.p. injections of benzo[a]pyrene showed several guanine mutations at
codon 12, which are indicative of (+)-anti-BPDE DNA adduct formation (Ross and Nesnow,
1999; Nesnow et al., 1998a, b, 1996, 1995; Mass et al., 1993); and (10) the major DNA adduct
formed in a murine embryonic fibroblast line transfected with human p53 DNA and exposed to
luM benzo[a]pyrene for 96 hours was (+)-anti-BPDE-DNA. The concomitant spectrum of p53
mutations in the latter study had features similar to those found in human lung cancer:
161
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
predominance of G—>T mutations, strand bias of transversions, and mutation hot spots at codons
157 to 158 (Liu et al., 2005).
As pointed out by Penning et al. (1999), the association between BPDE-DNA adducts
and tumors from benzo[a]pyrene exposure is not entirely specific given that dihydrodiol and diol
epoxides of benzo[a]pyrene are less potent tumorigenic agents in mouse skin than the parent
material (Slaga et al., 1977; Chouroulinkov et al., 1976) and oxidative damage to DNA has been
observed in rats treated with benzo[a]pyrene (Kim and Lee, 1997) and human mammary
epithelial cells exposed to benzo[a]pyrene (Leadon et al., 1988). In addition, although BPDE-
DNA adduct levels in forestomach tissue were linearly related to the amount of benzo[a]pyrene
consumed by mice in a 28-day study (Culp et al., 2000, 1998, 1996a; Culp and Beland, 1994),
levels of BPDE-DNA adduct in lung and liver tissue (which did not develop tumors with 2 years
of exposure to benzo[a]pyrene in the diet) were similar at 28 days to those in forestomach tumors
(Goldstein et al., 1998). These observations suggest that BPDE-adduct levels alone are not the
only path to benzo[a]pyrene-induced tumors and provide indirect evidence for the other
mutagenic pathways to benzo[a]pyrene tumor initiation.
Radical cation pathway
Radical cation formation involves a one-electron oxidation by CYP or peroxidase
enzymes (i.e., horseradish peroxidase, prostaglandin H synthetase) that produces electrophilic
radical cation intermediates (Cavalieri and Rogan, 1995, 1992). Radical cations can be further
metabolized to phenols and quinones (Cavalieri et al., 1988d, e), or they can form unstable
adducts with DNA that ultimately result in depurination (Cavalieri et al., 2005, 1993; Rogan et
al., 1993). The predominant depurinating adducts occur at the N-3 and N-7 positions of adenine
and the C-8 and N-7 positions of guanine (Cavalieri and Rogan, 1995).
Abasic sites resulting from base depurination undergo error-prone excision repair and can
induce mutations such as those found in the H-ras oncogene in mouse skin (Chakravarti et al.,
2000). One pathway to the formation of depurinating DNA adducts involves the formation of
DNA-reactive radical cations from benzo[a]pyrene via CYP peroxidases (Cavalieri and Rogan,
1995). In mouse skin exposed to 200 nmol benzo[a]pyrene for 4 hours, a mix of stable and
depurinating DNA adducts was found: (+)-anti-BPDE DNA and depurinating adducts accounted
for 22 and 74% of the identified adducts, respectively (Rogan et al., 1993). When mouse skin
was exposed to either the benzo[a]pyrene-7,8-diol or BPDE, only stable BPDE-DNA adducts
were found (Rogan et al., 1993). In mouse skin tumors induced by benzo[a]pyrene and
promoted by TP A, 7/13 examined tumors had H-ras oncogene mutations attributed to apurinic
sites generated by loss of N-7 and C-8 guanine adducts, and 2/13 tumors had H-ras mutations
attributed to loss of N-7 adenine adducts (Chakravarti et al., 1995). Results from the Rogan et al.
(1993) and Chakravarti (1995) studies provide strong in vivo evidence for the importance of both
162
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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 diol epoxide and the radical cation pathways in the activation of benzo[a]pyrene to initiate
mouse skin tumors, possibly by inducing mutations in critical genes.
o-Quinone/ROS pathway
The o-Quinone metabolites of PAHs are formed by enzymatic dehydrogenation of
dihydrodiols (Bolton et al., 2000; Penning et al., 1999; Harvey, 1996; ATSDR, 1995). DHH
enzymes are members of the a-keto reductase gene superfamily. o-Quinone metabolites are
potent cytotoxins, are weakly mutagenic, and are capable of producing a broad spectrum of DNA
damage. These metabolites can interact directly with DNA as well as result in the production of
ROS (i.e., hydroxyl and superoxide radicals) that may produce further cytotoxicity and DNA
damage. The DNA damage caused by o-quinones may include the formation of stable adducts
(Balu et al., 2004), N-7 depurinating adducts (McCoull et al., 1999), oxidative base damage (i.e.,
8-oxo-2'-deoxyguanosine or 8-oxo-dG) (Park et al., 2006a), and strand scission (Flowers et al.,
1997). The ROS generated by the o-quinone metabolites of benzo[a]pyrene and other PAHs
have been shown to induce mutation in the p53 tumor suppressor gene using an in vitro yeast
reporter gene assay (Park et al., 2008; Shen et al., 2006; Yu et al., 2002).
The o-quinone/ROS pathway also can produce depurinated DNA adducts from
benzo[a]pyrene metabolites (Jiang et al., 2007; 2005). In this pathway, and in the presence of
NAD(P)+, AKR oxidizes benzo[a]pyrene-7,8-diol to a ketol, which subsequently forms
benzo[a]pyrene-7-8-dione. This and other PAH o-quinones react with DNA to form unstable,
depurinating DNA adducts. In the presence of cellular reducing equivalents, o-quinones can also
activate redox cycles which produce DNA-ROS (Penning et al., 1996). DNA damage in in vitro
systems following exposure to benzo[a]pyrene-7,8-dione or other o-quinone PAH derivatives
occurs through the AKR pathway and can involve the formation of stable DNA adducts (Balu et
al., 2004), N-7 depurinated DNA adducts (McCoull et al., 1999), DNA damage from ROS
(8-oxo-dG) (Park et al., 2006a) and strand scission (Flowers et al., 1997, 1996).
Benzo[a]pyrene-7,8-dione and other PAH o-quinones have been shown to induce mutations in
the p53 tumor suppressor gene using an in vitro yeast reporter gene assay (Park et al., 2008; Shen
et al., 2006; Yu et al., 2002). When the yeast were exposed to varying concentration of
benzo[a]pyrene-7,8-dione or (+)-anti-BPDE, levels of 8-oxo-dG or (+)-anti-BPDE-DNA,
adducts were linearly related to p53 mutagenic frequencies with similar slopes, suggesting that
these two types of DNA lesions were equipotent in producing p53 mutations in this system (Park
et al., 2008). When the p53 mutations were sorted into dominant and recessive mutants, the
dominant mutations clustered to p53 mutation hotspots observed in human lung cancer tissue
(Park, 2008). The combined results provide strong in vitro evidence for the potential for the o-
Quinone/ROSpathway to produce several DNA-damaging products from benzo[a]pyrene (e.g.,
benzo[a]pyrene-7,8-dione and ROS) that lead to p53 mutations associated with human lung
cancer. In support of the operation of this pathway, and the other bioactivation pathways, in
163
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
humans, Jiang et al. (2007) used liquid chromatograpy-mass spectrometry (LC-MS) to provide
evidence for the formation of radical cations, diol epoxides, and o-quinones in cultured human
"3
lung H358 cells following exposure to 4 |iM [ H]-benzo[a]pyrene.
4.5.3. Mechanistic Studies- Mutagenesis and Tumor Initiation
Oncogene/tumor suppressor gene mutations (in vivo)
DeMarini et al. (2001) demonstrated mutations in the p53 tumor suppressor gene and the
K-ras oncogene in the lung tumors of nonsmokers, whose tumors were associated with exposure
to smoky coal. Lung tumors were obtained from 24 nonsmoking women from China (age 30-
63, mean age 48.5 ± 8.8 years) who used smoky coal in their homes without chimneys.
Bronchioloalveolar adenocarcinoma and acinar adenocarcinoma were observed in 54 and 46% of
the women studied, respectively. The observed mutations in lung tumors were primarily G—>T
transversions at either K-ras or p53. Mutation hotspots in the lung tumors examined
corresponded with hot spots for PAH adducts (codon 154), cigarette smoke associated mutations
(codon 249), and both of these events together (codon 273). The mutation spectrum was
described as unique and consistent with exposure to PAHs in the absence of cigarette smoke.
Mutations in the K-ras, H-ras, and p53 genes were assessed in forestomach tumors
(n = 31) of mice fed benzo[a]pyrene in the diet (0, 5, 25, or 100 ppm) for 2 years (Culp et al.,
2000). Forestomach tumors had K-ras mutations (68% of tumors) that were G—>T or C
transversions in codon 12 or 13. H-ras (codon 13) and p53 mutations characterized as G—>T or
C transversions were also each found in 10% of forestomach tumors.
K-ras mutations were observed in A/J mouse lung tumors (Nesnow et al., 1998a, b, 1996,
1995; Mass et al., 1993). Benzo[a]pyrene was administered to male A/J mice (20/group) as a
single i.p. injection (0, 10, 50, 100, or 200 mg/kg in tricaprylin) and the presence of lung
adenomas were evaluated 8 months following injection. The number of lung adenomas/mouse
was significantly greater than control (p < 0.05) for benzo[a]pyrene doses >50 mg/kg. Lung
tumor DNA was isolated and DNA sequence analysis of K-ras mutations was performed for 19
separate lung tumors. The DNA sequence analysis demonstrated several guanine mutations at
codon 12, including GGT—>TGT (56% of tumors), GGT—>GTT (25% of tumors), and
GGT—>GAT (19%) of tumors).
H-ras mutations were studied in skin papillomas of SENCAR mice resulting from dermal
initiation by benzo[a]pyrene or benzo[a]pyrene-7,8-dihydrodiol (400 nmol) followed by TP A
promotion (Chakravarti et al., 2000, 1995). PCR amplification of the H-ras gene and sequencing
revealed that codon 13 (GGC to GTC) and codon 61 (CAA to CTA) mutations in papillomas
corresponded to the relative levels of depurinating adducts of guanine and adenine, despite the
formation of significant amounts of stable DNA adducts.
164
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
DNA adducts in target tissues detectedfollowing chemical exposure
DNA adducts of benzo[a]pyrene have been measured in target tissues of humans exposed
to PAH mixtures and experimental animals exposed to benzo[a]pyrene. Phillips et al. (2002)
provided a review of smoking-related DNA adducts detected in human respiratory tract tissues.
BPDE-DNA adducts were detected in the lung, parenchyma of cigarette smokers with lung
cancer (Godschalk et al., 2002; Bartsch et al., 1999; Alexandrov et al., 1992). DNA was isolated
from the normal tissue (i.e., noncancerous) of the lung which was obtained during surgery. A
study using lung samples obtained on autopsy revealed that the average level of BPDE-DNA
o
adducts was higher in smokers (4.46 ± 5.76 per 10 bases) compared to ex-smokers (4.04 ± 2.37
8 8
per 10 bases) and nonsmokers (1.76 ± 1.69 per 10 bases) (Lodovici et al., 1998).
BPDE-DNA adducts were measured in skin biopsies of eczema patients treated with coal
tar preparations (Godschalk et al., 2001). Godschalk et al. (1998a) performed a study examining
the level of DNA adducts in biopsies of treated skin and in WBCs in psoriasis patients being
treated with coal tar. Urinary 1-OH-Py levels were serially monitored in all subjects. Skin
biopsies were taken before and after five treatments, at which point the average number of DNA
o
adduct levels increased from 2.9 to 63.3 per 10 nucleotides. Total WBC DNA adducts increased
8 8
from 0.33 to 0.89 per 10 after five treatments, and then doubled to 1.59 per 10 when sampled
1 week later. There was an increase in 1-OH-Py levels from 0.75 to 186 [j,g/L after one treatment
and to 266 [j,g/L after five treatments. One week later, mean 1-OH-Py levels were reduced to
2.4 (J,g/L. Adduct levels in the skin increased over 20-fold with five treatments, while WBC
adduct levels approximately doubled over the same period.
DNA adduct levels were examined in the forestomach of groups of female B6C3Fi mice
fed benzo[a]pyrene in the diet at concentrations of 5, 25, or 100 ppm for 28 days (Culp et al.,
2000, 1998, 1996a, b; Culp and Beland, 1994). [32P]-postlabeling of forestomach DNA of
benzo[a]pyrene-treated mice revealed one major adduct characterized as dG-N -BPDE. There
was a linear relationship between the amount of benzo[a]pyrene consumed and the concentration
of dG-N -BPDE in the forestomach of mice. For benzo[a]pyrene, forestomach tumor incidence
increased sharply with adduct concentrations between 50 and 140 fmol/mg DNA and in coal-tar
fed mice. Tumor incidence increased sharply with dG-N -BPDE adduct levels between 20 and
60 fmol/mg DNA. The same levels of adduct were present in lung and liver of benzo[a]pyrene-
treated mice, although only the forestomach exhibited benzo[a]pyrene-induced tumors
(Goldstein et al., 1998). The presence of adducts in tumor-free tissue suggests that DNA adduct
levels alone are not necessarily predictors of tumor outcome.
DNA adducts were identified and quantified in experiments using the A/J mouse lung
model which results in lung adenomas in male A/J mice 8 months following a single i.p.
injection (Nesnow et al., 1998a, b, 1996, 1995; Ross et al., 1995). Benzo[a]pyrene was
administered to male A/J mice (20-25/group) as a single i.p. injection (0, 20, 50, or 100 mg/kg in
165
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
tricaprylin) and DNA was isolated from lung tissues at several time points between 1 and 21
days following injection. The primary DNA adduct identified in mouse lung tumors was a
benzo[a]pyrene bay region diol epoxide adduct of guanine (7R, 8S, 9S-trihydroxy-10R-[2N-2'-
deoxyguanosyl]-7, 8, 9, 10-tetrahydro-benzo[a]pyrene). Two minor adducts were also observed
to result from the metabolism of 9-hydroxy-benzo[a]pyrene and trans-7,8-dihydroxy-7,8-
32
dihydro-benzo[a]pyrene. Quantitative analysis of DNA adducts by [ P]-postlabeling illustrated
the importance of measuring DNA adduct levels over time. Total DNA adducts accumulated
rapidly between 3 and 9 days after exposure followed by a gradual decrease. A time-integrated
DNA adduct level (TIDAL) was linearly related to the administered dose of benzo[a]pyrene
(Ross et al., 1995).
Following i.p injection, benzo[a]pyrene also induced DNA adducts in the lungs, liver,
and peripheral blood lymphocytes of rats (Ross et al., 1991, 1990), the liver of Lewis rats
(Godschalk et al., 1998b), and lungs of BALB/c mice (Van Schooten et al., 1991). Qian et al.
(1998) treated male CD rats with intratracheal instillation of fume condensates of roofing
asphalts and evaluated adducts in lung cells and WBCs. Adducts were seen in the lungs but not
in WBCs, leading the authors to conclude that WBCs may not be a suitable surrogate for lung
cells. The adducts were not characterized, so the benzo[a]pyrene-specificity of the results cannot
be evaluated. Formation of DNA adducts from benzo[a]pyrene metabolites also has been
observed in the lung and liver of male Sprague-Dawley rats after intratracheal administration of
benzo[a]pyrene (De Flora et al., 1991; Weyand and Bevan, 1987).
DNA adducts have been reported in the lung and skin of dermally treated SENCAR mice
(Mukhtar et al., 1986), in the epidermis of Swiss mice (Oueslati et al., 1992), and in the skin of
an unspecified strain of mice (Ingram et al., 2000). When Talaska et al. (1996) compared the
dose-duration-response of benzo[a]pyrene-induced adducts in the skin, lung, and liver of Hsd
(ICR) BR mice treated dermally with 10, 25, or 50 nmol benzo[a]pyrene, accumulation of
adducts was found to be linear with dose in the skin and lung. In skin painting studies with
female SENCAR mice and various PAHs, Melendez-Colon et al. (1999) found that carcinogenic
potency correlated with DNA adduct levels in epidermal DNA rather than in the formation of
apurinic sites. Alexandrov and Rojas-Moreno (1990) found DNA adducts in epidermal
keratinocytes and dermal fibroblasts of Swiss mice treated dermally with benzo[a]pyrene but not
in similarly treated Wistar rats. BPDE-DNA adducts were measured in the lung, stomach, and
skin of male Lewis rats (15/group) following a single exposure to 10 mg/kg benzo[a]pyrene via
the intratracheal, gavage, and dermal routes, respectively (Godschalk et al., 2000).
4.5.3. Tumor Promotion and Progression
benzo[a]pyrene has been shown to promote the growth of previously initiated cells,
resulting in the formation of tumors in the skin (see Section 4.2.3.1). The tumor promotion
166
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
properties of benzo[a]pyrene may be due to a compensatory response to cytotoxicity or via an
AhR-mediated effect on cell growth and differentiation.
Cytotoxicity and inflammatory response
The cytotoxicity of benzo[a]pyrene metabolites may contribute to tumor promotion via
inflammatory responses leading to cell proliferation (Burdick et al., 2003). Benzo[a]pyrene is
metabolized to o-quinones, which are cytotoxic, and can generate ROS (Bolton et al., 2000;
Penning, 1999). Benzo[a]pyrene o-quinones reduce the viability and survival of rat and human
hepatoma cells (Flowers-Geary et al., 1996, 1993). Cytotoxicity was also induced by
benzo[a]pyrene and BPDE in a human prostate carcinoma cell line (Nwagbara et al., 2007).
Inflammatory responses to cytotoxicity may contribute to the tumor promotion process. For
example, benzo[a]pyrene quinones (1,6-, 3,6-, and 6,12-benzo[a]pyrene-quinone) generated ROS
and increased cell proliferation by enhancing the epidermal growth factor receptor
(EFGR)pathway in cultured breast epithelial cells (Burdick et al., 2003).
Several studies have demonstrated that exposure to benzo[a]pyrene increases the
production of inflammatory cytokines which may contribute to cancer progression. Gar<;on et al.
(2001a, b) exposed Sprague-Dawley rats by inhalation to benzo[a]pyrene with or without ferrous
oxide (Fe2C>3) particles. They found that benzo[a]pyrene alone or in combination with Fe2C>3
particles elicited mRNA and protein synthesis of the inflammatory cytokine, IL-1. Tamaki et al.
(2004) also demonstrated a benzo[a]pyrene-induced increase in IL-1 expression in a human
fibroblast-like synoviocyte cell line (MH7A). Benzo[a]pyrene increases the expression of the
mRNA for CCL1, an inflammatory chemokine, in human macrophages (N'Diaye et al., 2006).
The benzo[a]pyrene-induced increase in CCL1 mRNA was inhibited by the potent AhR
antagonist 3 '-methoxy-4'-nitroflavone.
AhR-mediated effects
The promotional effects of benzo[a]pyrene may also be related to AhR affinity and the
upregulation of genes related to biotransformation (i.e., induction of CYP1 Al), growth, and
differentiation (Bostrom et al., 2002). Figure 4-1 illustrates the function of the AhR and depicts
the genes regulated by this receptor as belonging to two major functional groups (i.e., induction
of metabolism or regulation cell differentiation and proliferation). PAHs bind to the cytosolic
AhR in complex with heat shock protein 90 (Hsp90). The ligand-bound receptor is then
transported to nucleus in complex with the ARNT. The AhR complex interacts with the Ah
responsive elements (AHRE) of the DNA to increase the transcription of proteins associated with
induction of metabolism and regulation of cell differentiation and proliferation.
167
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
PAH
AHR
Hsp90
Hsp90
ARNT
ARNT
AHRE
Enhanced
specific
mRNA
production
ARNT
nucleus
Increased
synthesis of
PAH metabolizing
enzymes
Increased
synthesis of
proteins that
regulate cell
differentiation and
proliferation
AHREdna = Ah-responsive elements of DNA; Hsp90 = heat shock protein 90
Source: Okey et al. (1994).
Figure 4-1. Interaction of PAHs with the AhR.
Binding to the AhR induces enzymes that increase the formation of reactive metabolites,
resulting in DNA binding and, eventually, tumor initiation. In addition, with persistent exposure,
the ligand-activated AhR triggers epithelial hyperplasia, which provides the second step leading
from tumor initiation to promotion and progression (Nebert et al., 1993). Ma and Lu (2007)
reviewed several studies of benzo[a]pyrene toxicity and turnorigenicity in mouse strains with
high and low affinity AhRs. Disparities were observed in the tumor pattern and toxicity of Ah-
responsive (+/+ and +/-) and Ah-nonresponsive (-/-) mice. Ah-responsive mice were more
susceptible to toxicity and tumorigenicity in proximal target tissues such as the liver, lung, and
skin. For example, Shimizu et al. (2000) reported that AhR knock out mice (-/-), treated with
benzo[a]pyrene by s.c. injection or dermal painting, did not develop skin cancers at the treatment
site, while AhR-responsive (+/+) or heterozygous (+/-) mice developed tumors within 18-
25 weeks after treatment. Benzo[a]pyrene treatment increased CYP1 Al expression in the skin
and liver of AhR-positive mice (+/- or +/+), but CYP1 Al expression was not altered by
benzo[a]pyrene treatment in AhR knock out mice (-/-). Talaska et al. (2006) also showed that
benzo[a]pyrene adduct levels in skin were reduced by 50% in CYP1A2 knock out mice and by
168
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
90% in AhR knock out mice compared with WT C57B16/J mice following a single dermal
application of 33 mg/kg benzo[a]pyrene for 24 hours. Ma and Lu (2007) further noted that Ah-
nonresponsive mice were at greater risk of toxicity and tumorigenicity in remote organs, distant
from the site of exposure (i.e., bone marrow). As an example, Uno et al. (2006) showed that
benzo[a]pyrene (125 mg/kg/-day, p.o. for 18 days) caused marked wasting, immunosuppression,
and bone marrow hypocellularity in CYP1 Al knock out mice, but not in WT mice.
Some studies have demonstrated the formation of DNA adducts in the liver of AhR
knock out mice following i.p. or oral exposure to benzo[a]pyrene (Sagredo et al., 2006; Uno et
al., 2006; Kondraganti et al., 2003). These findings suggest that there may be alternative (i.e.,
non-AhR mediated) mechanisms of benzo[a]pyrene activation in the mouse liver. Sagredo et al.
(2006) studied the relationship between the AhR genotype and CYP metabolism in different
organs of the mouse. AhR+/+, , and mice were treated once with 100 mg/kg benzo[a]pyrene
by gavage. CYP1A1, CYP1B1, and AhR expression was evaluated in the lung, liver, spleen,
kidney, heart, and blood, via RT-PCR, 24 hours after treatment. CYP1 Al RNA was increased in
the lung and liver and CYP IB 1 RNA was increased in the lung following benzo[a]pyrene
treatment in AhR+/+ and +/~ mice (generally higher in heterozygotes). Benzo[a]pyrene treatment
did not induce CYP1A1 or CYP1B1 enzymes in AhR 7 mice. The expression of CYP1A1 RNA,
as standardized to (3-actin expression, was generally about 40 times that of CYP1B1. The
concentration of benzo[a]pyrene metabolites and the levels of DNA and protein adducts were
increased in mice lacking the AhR, suggesting that there may be an AhR-independent pathway
for benzo[a]pyrene metabolism and activation. The high levels of benzo[a]pyrene DNA adducts
in organs other than the liver of AhR 1 mice may be the result of slow detoxification of
benzo[a]pyrene in the liver, allowing high concentrations of the parent compound to reach
distant tissues.
Uno et al. (2006) also demonstrated a paradoxical increase in liver DNA adducts in AhR
ko mice following oral exposure to benzo[a]pyrene. WT C57BL/6 mice and several knock out
mouse strains (CYP1A2 1 and CYP1B1 single ko, CYPlAl/lBl and CYP1A2/1B1
double ko) were studied. Benzo[a]pyrene was administered in the feed at 1.25, 12.5, or 125
mg/kg for 18 days (this dose is well tolerated by WT C57BL/6 mice for 1 year, but lethal within
30 days to the CYP1 Al 1 mice). Steady-state blood levels of benzo[a]pyrene, reached within 5
days of treatment, were -25 times higher in CYP 1A1 and -75 times higher in CYPlAl/lBl^
than in WT mice, while clearance was similar to WT mice in the other knock out mouse strains.
32
DNA adduct levels, measured by [ P]-postlabeling in liver, spleen, and bone marrow, were
highest in the CYP1 Al 1 mice at the two higher doses, and in the CYP1A1/1B1 mice at the
mid dose only. Adduct patterns, as revealed by 2-dimensional chromatography, differed
substantially between organs in the various knock out types.
Dertinger et al. (2001, 2000) demonstrated that AhR signaling may play a role in
cytogenetic damage caused by benzo[a]pyrene. The in vivo formation of micronuclei in
169
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
peripheral blood reticulocytes of C57B1/6J mice induced by a single i.p. injection of
benzo[a]pyrene (150 mg/kg) was eliminated by prior treatment with the potent AhR antagonist
3'-methoxy-4'-nitroflavone. This antagonist also protected AhR null allele mice from
benzo[a]pyrene-induced increases in micronuclei formation, suggesting that 3'-methoxy-4'-
nitroflavone may also act through a mechanism independent of the AhR (Dertinger et al., 2000).
Several in vitro studies have suggested that the AhR plays a role in the disruption of cell
cycle control, possibly leading to cell proliferation and tumor promotion following exposure to
benzo[a]pyrene (Andrysik et al., 2007; Chung et al., 2007; Chen et al., 2003). Chung et al.
(2007) showed that benzo[a]pyrene-induced cytotoxicity and apoptosis in mouse hepatoma
(Hepalclc7) cells occurred through a p53 and caspace-dependent process requiring the AhR.
An accumulation of cells in the S-phase of the cell cycle (i.e., DNA synthesis and replication)
was also observed, suggesting that this process may be related to cell proliferation. Chen et al.
(2003) also demonstrated the importance of the AhR in benzo[a]pyrene-7,8-dihydrodiol- and
BPDE-induced apoptosis in human HepG2 cells. Both the dihydrodiol and BPDE affected Bcl2
(a member of a family of apoptosis suppressors) and activated caspase and p38 mitogen-
activated protein (MAP) kinases, both enzymes that promote apoptosis. When the experiments
were conducted in a cell line that does not contain ARNT (see Figure 4-1), the dihydrodiol was
not able to initiate apoptotic event sequences, indicating that activation to BPDE by CYP1 Al
was required. BPDE did not induce apoptosis-related events in a p38-defective cell line,
illustrating the importance of MAP kinases in this process. In rat liver epithelial cells (WB-F344
cells), in vitro exposure to benzo[a]pyrene resulted in apoptosis, a decrease in cell number, an
increase in the percentage of cells in S-phase (comparable to a proliferating population of WB-
F334 cells), and increased expression of cell cycle proteins (e.g., cyclin A) (Andrysik et al.,
2007). Benzo[a]pyrene-induced apoptosis was attenuated in cells transfected with a dominant-
negative mutation of the AhR.
Inhibition of gap junctional intercellular communication
Gap junctions are channels between cells that allow substances of a molecular weight up
to roughly 1 kDa to pass from one cell to the other. This process of metabolic cooperation is
crucial for differentiation, proliferation, apoptosis, and cell death and consequently for the two
epigenetic steps of tumor formation, promotion, and progression. Chronic exposure to many
toxicants results in down-regulation of gap junctions. For tumor promoters, such as TPA or
TCDD, inhibition of intercellular communication is correlated with their promoting potency
(Sharovskaya et al., 2006; Yamasaki, 1990).
Blaha et al. (2002) surveyed the potency of 35 PAHs, including benzo[a]pyrene, to
inhibit gap junctional intercellular communication (GJIC). The scrape loading/dye transfer assay
was employed using a rat liver epithelial cell line that was incubated in vitro for 15, 30, or 60
minutes with 50 [xM benzo[a]pyrene. After incubation, cells were washed, and then a line was
170
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
scraped through the cells with a surgical blade. Cells were exposed to the fluorescent dye lucifer
yellow for 4 minutes and then fixed with formalin. Spread of the dye from the scrape line into
cells remote from the scrape was estimated under a fluorescence microscope. Benzo[a]pyrene
reduced spread of the dye after 30 minutes of exposure (approximately 50% of control).
Recovery of GJIC was observed 60 minutes after exposure.
Sharovskaya et al. (2006) studied the effects of carcinogenic and noncarcinogenic PAHs
on GJIC in HepG2 cells. Individual carcinogenic PAHs inhibited GJIC in a temporary fashion
(70-100% within 24 hours), but removal of the PAH from culture reversed the effect.
Noncarcinogenic PAHs had very little effect on GJIC. Benzo[a]pyrene at 20 |iM inhibited GJIC
completely within 24 hours, while its noncarcinogenic homolog, BeP, produced <20% inhibition.
The effect was not AhR-dependent, because benzo[a]pyrene inhibited GJIC in HepG2 cells to
the same extent as in hepatoma G27 cells, which express neither CYP1 Al nor AhR. The authors
concluded that the effects of benzo[a]pyrene and BeP on GJIC were direct (i.e., not caused by
metabolites).
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS
4.6.1. Oral
Numerous epidemiological studies are available which investigate associations between
PAH (including benzo[a]pyrene) dietary intake and cancer incidence; however, no studies were
found which evaluate the contribution of benzo[a]pyrene through dietary exposure in humans
and noncancer health effects. Several studies in animal models are available evaluating the
sensitive noncancer effects following subchronic or chronic exposure to benzo[a]pyrene. The
types of effects observed following oral exposure were predominantly effects in the reproductive
and immune systems. Additionally, some minor hematological effects and kidney and
forestomach effects were observed. Studies that identified NOAELs and LOAELS for
noncancer effects in animals repeatedly exposed to benzo[a]pyrene by the oral route are
summarized in Table 4-27.
171
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-27. NOAELS and LOAELs for noncancer effects in animals repeatedly exposed to benzo[a]pyrene by the oral
route
NOAEL
LOAEL
Species/sex
Dose
Duration
mg/kg-d
Response at LOAEL
Comments
Reference
Wistar rat/
male and female
0, 3, 10, or
30 mg/kg-d,
gavage, 5 d/wk
2 yrs
ND
3
t Forestomach hyperplasia, t liver clear
cell foci of alteration
Hematological and organ weight
variables were not measured at
terminal sacrifice.
No exposure-related changes in
noncancer histology in oral cavity,
esophagus, forestomach, jejunum,
liver, kidney, skin, mammary gland,
or auditory canal (noncancer lesions
were only detected in tissues that
developed tumors),
t Forestomach tumors in males at
>3 mg/kg-d.
t Forestomach, liver, and kidney
(males only) tumors at 10 and
30 mg/kg-d.
Kroese et al.,
2001
B6C3Fi
mouse/female
only
Estimated
doses: 0, 0.7,
3.3, or
16.5 mg/kg-d in
diet
2 yrs
ND
0.7
3.3
0.7
3.3
16.5
t Forestomach hyperplasia
t Forestomach hyperkeratosis
t Esophagus (basal cell hyperplasia)
No exposure-related changes in
weight or histology of liver, kidney,
or lung weight (other organs not
measured).
Hematologic variables were not
examined.
t Forestomach tumors at 3.3 and
16.5 mg/kg-d;
t Esophagus and tongue tumors at
16.5 mg/kg-d.
Beland and
Culp, 1998;
Culp et al.,
1998
Wistar rat/
male and female
0, 3, 10, or
30 mg/kg-d,
5 d/wk
90 d
10
3
30
10
I Thymus weight, t liver weight,
t forestomach hyperplasia, t slight thymic
atrophy
t Forestomach epithelial cell proliferation
index (BrdU incorporation)
No exposure-related changes in
hematological variables or histology
of lung, spleen, or lymph node.
Kroese et al.,
2001
F344 rat/
male and female
0, 5, 50, or
100 mg/kg-d in
diet
90 d
ND
5
50
5
50
100
t renal tubular casts in males
| RBCs and hematocrit in males
i RBCs and hematocrit in females,
I hemoglobin in both sexes, flivenbody
weight ratio in males
No exposure-related changes in
other organ weights measured
(stomach, testes, ovaries), or in
histology of stomach, liver, testes or
ovaries (other tissues were not
examined).
Knuckles et al.,
2001
172
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-27. NOAELS and LOAELs for noncancer effects in animals repeatedly exposed to benzo[a]pyrene by the oral
route
NOAEL
LOAEL
Species/sex
Dose
Duration
mg/kg-d
Response at LOAEL
Comments
Reference
SD male rats
0, 1 or 5 mg/kg-
day by gavage
90 d
1
5
i testicular testosterone
Testosterone levels measured in
animals sacrificed after 30 d, were
not statistically different than
controls
Zheng et al.,
2010
SD female rats
0,2.5 or 5
mg/kg-day,
gavage1
60 d
ND
2.5
2.5
5
i ovary weight
i estrogen and primordial follicles; altered
estrous cyclicity
Xu et al., 2010
C57BL/6 male
mice
0, 1, or 10
mg/kg-day
exposure to FO
generation
42 d
ND
1
1
10
i epididymal sperm count in F0 and F1
generations
i sperm motility in F0 mice
Mohamed et
al., 2010
Wistar rat/male
only
0, 3, 10, 30, or
90 mg/kg-d,
gavage, 5 d/wk
35 d
3
10
30
10
30
90
i RBCs, hemoglobin, and hematocrit,
i thymus weight, j percent B cells in
spleen
t Forestomach hyperplasia, j serum IgM
and IgA
t Liver oval cell hyperplasia
No exposure-related histological
changes in adrenals, brain, bone
marrow, colon, caecum, jejunum,
heart, kidney, lung, lymph nodes,
esophagus, pituitary, spleen,
stomach, testis, or thymus.
De Jong et al.,
1999
Wistar rat/
male and female
0, 1.5, 5, 15, or
50 mg/kg-d,
gavage, 5 d/wk
35 d
5
15
15
50
I Thymus weight, t forestomach
hyperplasia
t Liver weight
No exposure-related changes in
hematological variables, weights of
kidney, spleen, lung, adrenals or
ovaries, or histology of liver,
kidney, spleen, thymus, lung, or
mammary gland.
Kroese et al.,
2001
CD-I mouse/ FO
female;
F1 male and
female
0,10, 40, or
160 mg/kg-d,
gavage
GDs 7-16
ofFO
pregnancy
40
10
ND
ND
160
40
10
10
I Number of F0 females with viable litters
| F1 body weight at PND 20
| F1 body weight at PND 42
| F1 male and Flfemale fertility index
Beginning at 6-7 wks of age, each
F1 male mouse (20-45/group) was
exposed to 10 untreated females
over a period of 25 d. Beginning at
6 wks of age, each F1 female mouse
(20-55/group) was cohabitated with
an untreated male for a period of
6 mo.
MacKenzie and
Angevine,
1981
173
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-27. NOAELS and LOAELs for noncancer effects in animals repeatedly exposed to benzo[a]pyrene by the oral
route
Species/sex
Dose
Duration
NOAEL
LOAEL
Response at LOAEL
Comments
Reference
mg/kg-d
NMRI mouse/
F0 female;
F1 female
0 or 10 mg/kg-d
GDs 7-16
ofFO
pregnancy
ND
10
| F1 female fertility (j number of F2 litters
and F2 litter size; j ovary weight, and j
numbers of small, medium, or large
follicles and corpora lutea)
Exposed F0 females showed no
gross signs of toxicity and no
effects on fertility. One F1 female
from each litter was continuously
bred with an untreated male for
6 mo.
Kristensen et
al., 1995
ND = not determined
1 Time weighted average dose ; animals treated by gavage, every other day to 0, 5, or 10 mg/kg-day
1
2
174
DRAFT - DO NOT CITE OR QUOTE
-------
35
30
25
20
15
10
NOAEL O LOAEL
9
o
9
9
o
o
o
9
9
o
o
FS hyp; rat;
FS hyp;
\1/ thymus
\1/ thymus
\1/ B-cells in
\1/ ovary wt;
\1/ follicles
\1/ immune
\1/ F1 body
si F1
(1)
mouse; (2)
wt; rat; (1)
wt; rat; (3)
spleen; rat;
rat; (4)
and
testosterone;
epididymal
response;
weight and
fertility;
(3)
estrogen; rat;
rats; (5)
sperm count;
mouse; (7)
fertility;
mouse; (9)
(4)
mice; (6)
mouse; (8)
2 years
35-90 day
Gestational (days 7-16)
(1) Kroese et al., 2001; (2) Beland and Culp, 1998; Culp et al., 1998; (3) De Jong et al., 1999; (4) Xu et al., 2010; (5) Zheng et al., 2010; (6) Moliamed et al..
2010; (7) van den Berg et al.. 2005; (8) MacKenzie and Angevine, 1981; (9) Kristensen et al.. 1995; f= increased; |= decreased; FS = forestomach; hyp =
hyperplasia; wt = weight.
Figure 4-2. NOAELs and LOAELs for selected noncancer effects from repeated oral exposure to
benzo[a]pyrene.
175
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
The two oral chronic-duration studies identify forestomach hyperplasia in rats (gavage
exposure) and mice (dietary exposure) as a sensitive effect—a LOAEL of 3 mg/kg-day for
forestomach hyperplasia in male and female Wistar rats (Kroese et al., 2001) and a LOAEL of
0.7 mg/kg-day for forestomach hyperplasia in female B6C3Fi mice (Beland and Culp, 1998;
Culp et al., 1998). In both rats and mice, an increasing incidence of animals with forestomach
tumors with increasing dose was also observed.
Several immune related effects have been observed in animals treated subchronicly with
benzo[a]pyrene, including decreased thymus weight, decreased % of B cells in the spleen,
decreased RBCs, and decreased serum immunoglobulins. LOAELs for decreased thymus
weights were 10 and 15 mg/kg-day in two different studies of Wistar rats exposed by gavage for
35 days (Kroese et al., 2001; De Jong et al., 1999) and 30 mg/kg-day for Wistar rats exposed to
benzo[a]pyrene in the diet for 90 days (Kroese et al., 2001). Thymus weights were not measured
in the the available chronic studies. Decreased thymus weight was accompanied by a decreased
percentage of B cells in spleen and decreased serum IgM and IgA in one 35-day Wistar rat study
(De Jong et al., 1999) and increased incidence of slight thymic atrophy in the 90-day Wistar rat
study (Kroese et al., 2001). Thymus atrophy, but no histological thymus lesions, was noted in
the other 35-day study with Wistar rats (Kroese et al., 2001). Other support for immune effects
as a potential effect from repeated oral exposure to benzo[a]pyrene is shown by decreased
immune responses in lymph nodes to the dermal sensitizer, DNCB, in C56BL/6 mice given 13
mg/kg-day (LOAEL) 3 times/week for 4 weeks (van den Berg et al., 2005; see Section 4.4.2).
Effects on RBC counts were also observed across the rat subchronic duration studies
(Table 4-25). LOAELs for decreased RBCs were 10 mg/kg-day in Wistar rats exposed by
gavage for 35 days (De Jong et al., 1999) and 50 mg/kg-day in male F344 rats exposed in the
diet for 90 days (Knuckles et al., 2001), but no significant exposure-related changes in RBC
counts were observed in Wistar rats in another 35 day study at doses up to 50 mg/kg-day (Kroese
et al., 2001) or at 30 mg/kg-day in Wistar rats exposed in the diet for 90 days (Kroese et al.,
2001; see Table 4-26). When observed, the magnitudes of the decreases in RBC, hemoglobin, or
hematocrit were generally small: about 18% at 90 mg/kg-day and <10% at lower doses in Wistar
rats (De Jong et al., 1999) and about 10% in F344 rats (Knuckles et al., 2001). Hematologic
variables were not measured at the terminal sacrifices in the chronic duration studies in rats
(Kroese et al., 2001) or mice (Beland and Culp, 1998; Culp et al., 1998).
Kidney effects characterized as increased incidence of renal tubular casts in male F344
rats were observed in a study by Knuckles et al 2001). The most sensitive effect observed in this
study was an increase in abnormal tubular casts in the kidney in males at 5 mg/kg-day (40%),
50 mg/kg-day (80%) and 100 mg/kg-day (100%), compared to 10% in the controls. In females,
176
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
only 10% showed significant kidney tubular changes at the two high dose levels compared to
zero incidence in controls.
Reproductive and developmental effects following gestational exposure to
benzo[a]pyrene have been observed in animal models. Decreased male reproductive endpoints
including decreased testicular testosterone, decreased epididymal sperm count, and decreased
sperm motility have been observed in rodents treated subchronically (Mohamed et al., 2010;
Zheng et al., 2010). In addition, female reproductive endpoints including decreased ovary
weight, decreased estrogen, decreased primordial follicles, and estrus cyclicity have been
observed in female rats treated for 60 days (Xu et al., 2010). Impaired reproductive performance
in F1 mouse offspring (male and female) has been observed following exposure of F0 mice to
oral doses as low as 10 mg/kg-day during GDs 7-16 (Kristensen et al., 1995; MacKenzie and
Angevine, 1981). Effects observed included decreased ovary weight in F1 females and reduced
fertility as reflected by decreased mean number of F2 litters. F1 females had statistically
significantly lower median numbers of offspring, number of litters, and litter sizes and a
statistically significantly greater median number of days between litters as compared with the
controls (Kristensen et al., 1995). Another study of gestationally treated dams (GD 7-16)
identified statistically significant decrements in fertility, pup weight, and reproductive organ
weights and histology (MacKenzie and Angevine, 1981). These mouse developmental/
reproductive toxicity studies observed effects at the lowest dose tested (10 mg/kg-day).
Reductions in motor activity, decreased grip strength, and decreased response to sound, touch,
and pain were observed in F344 rats following administration of single gavage doses of
>25 mg/kg (Saunders et al., 2006, 2002, 2001; see Section 4.4.1), but similar evaluations of
neurological endpoints following repeated oral exposure of animals to benzo[a]pyrene were not
located.
Studies with ApoE-/- mice, which spontaneously develop atherosclerosis, show that
repeated oral exposure to 5 mg/kg gavage doses of benzo[a]pyrene enhances the progression of
atherosclerosis through a general local inflammatory process (Knaapen et al., 2007; Curfs et al.,
2005, 2004; Godschalk et al., 2003; see Section 4.4.4); however, available data are inadequate to
assess oral exposure dose-response relationships for benzo[a]pyrene-induced atherosclerosis in
normal test animals.
4.6.2. Inhalation
Several epidemiological studies have associated increased occupational exposure of
benzo[a]pyrene with cardiac endpoints, specifically ischemic heart disease (Friesen et al., 2010;
Burstyn et al., 2005). Other studies have reported potential prenatal effects, birth outcomes, and
decreased fertility associated with increased exposure to benzo[a]pyrene. Decreased head
177
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
circumference, decreased birth weight, and decreased postnatal weight have been reported (Tang
et al., 2006; Perera et al., 2005a,b;) in addition to increased risk of early fetal death (Wu et al.,
2010). Furthermore, elevated levels of benzo[a]pyrene in follicular fluid have been associated
with reduced fertility (Neal et al., 2008).
In addition to epidemiological studies, several repeated-exposure inhalation toxicity studies in
animals exist for benzo[a]pyrene (Archibong et al., 2002, 2008; Ramesh et al., 2008; Inyang et
al., 2003; Wormley et al., 2004). A lifetime-exposure carcinogenicity study of Syrian golden
hamsters exposed to benzo[a]pyrene condensed onto NaCl aerosols at nominal concentrations of
-3
2, 10, or 50 mg/m is available (Thyssen et al., 1981); however, noncancer effects were not
evaluated.
Although no standard developmental toxicity studies are available for benzo[a]pyrene via
the inhalation route, decreased fetal survival and number of pups per litter were observed
following exposure of pregnant F344 rats to aerosols of benzo[a]pyrene and CB at
"3
concentrations >25 (j,g/m on GDs 11-20 (Archibong et al., 2002). Decreased levels of plasma
progesterone, estradiol, and prolactin were observed on GD 17 in dams exposed to 75 (j,g/m , but
"3
not in those exposed to 25 (j,g/m (Archibong et al., 2002). Other rat studies from the same
-3
laboratory have associated inhalation exposures to 100 (j,g/m benzo[a]pyrene:CB aerosols
during gestation with changes in electrophysiological variables in the hippocampus (Wormley et
al., 2004). Duration-dependent effects on male reproductive endpoints including increased
luteinizing hormone, decreased circulating and intratesticular concentrations of testosterone,
decreased testis weight, and decreased sperm motility have also been observed following
"3
exposure of adult male F344 rats to benzo[a]pyrene:CB aerosols at 75 (j,g/m for 10 or 60 days
(Archibong et al., 2008; Ramesh et al., 2008; Inyang et al., 2003).
178
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-28. NOAELS and LOAELs for noncancer effects in animals repeatedly exposed to benzo[a]pyrene by the
inhalation route
NOAEL
LOAEL
Species/sex
Dose
Duration
fig/m3
Response at LOAEL
Comments
Reference
F344 rats
0, 25, 75, 100
Hg/m3
4hr/d
GD 11-20
ND
25
25
75
i pups/litter, litter survival (%)
t resorptions
i plasma progesterone, estradiol, and
prolactin
carbon black used as carrier particle
Archibong et
al., 2002; Wu
et al 2003
F344 rats
0, 75 ng/m3
4hr/d
60 d
ND
75
t luteinizing hormone
i testosterone
i decreased testis weight
i sperm motility
carbon black used as carrier particle
in treatment group; controls not
exposed to carbon black
Archibong et
al., 2008
F344 rats
0, 100 |ig/m3
4hr/d
GD
ND
100
i pups/litter
electrophysiological changes in the
hippocampus
Wormley et al.,
2004
179
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
4.6.3. Dermal
Though numerous chronic cancer bioassays exist for benzo[a]pyrene by the dermal route,
noncancer effects were not reported in these studies, nor are studies available evaluating
noncancer effects in humans exposed dermally to benzo[a]pyrene.
4.6.4. Mode-of-Action Information
4.6.4.1. Forestomach Lesions from Oral Exposure
The development of forestomach hyperplasia in mice and rats from subchronic or chronic
oral exposures (by gavage and diet) to benzo[a]pyrene is reasonably expected to involve a cell
proliferative response to cytotoxicity from reactive benzo[a]pyrene metabolic intermediates,
based on the extensive findings from research on the bioactivation of benzo[a]pyrene and
carcinogenicity (see reviews on the bioactivation of benzo[a]pyrene by Xu et al., 2009; Jiang et
al., 2007, 2005; Xue and Warshawsky, 2005; Penning et al., 1999; Harvey 1996; Cavalieri and
Rogan, 1995). Reactive intermediates that can react with cellular macromolecules and
potentially lead to cytotoxicity include BPDE, benzo[a]pyrene radical cations, benzo[a]pyrene o-
quinones, and ROS. Reactive benzo[a]pyrene metabolites are also well known to reduce the
viability and survival of cultured cells involving mechanisms related to stimulation of apoptosis
(Andrysik et al., 2007; Chung et al., 2007; Nwagbara et al., 2007; Chen et al., 2003; Jyonouchi et
al., 1999; Flowers-Geary et al., 1996, 1993). Molecular details of cell proliferative responses to
cytotoxicity or apoptosis from benzo[a]pyrene metabolites are poorly understood, but Burdick et
al. (2006, 2003) provided evidence that benzo[a]pyrene o-quinones could inhibit apoptosis and
increase cell proliferation in a model human mammary epithelial cell system (MCF10A) via
activation of the epidermal growth factor receptor (EGFR) by ROS. The relationship of these
findings to development of benzo[a]pyrene-induced forestomach hyperplasia is unknown.
4.6.4.2. Immune System Effects
Decreased thymus weight, decreased number of B cells in spleen, and immune
suppression have been observed following oral, i.p., s.c., or intratracheal instillation exposure to
benzo[a]pyrene. DeJong et al. (1999) and Kroese et al. (2001) found decreased thymus weight
due to benzo[a]pyrene at oral doses >10 mg/kg body weight. In addition, several studies report
thymus effects at higher doses and/or by other routes of exposure (e.g. Rodriguez et al., 1999;
Holladay and Smith, 1994). Reduced thymus size or weight have been noted to be among the
first indicators of immunotoxicity (Schuurman et al. 1992; Luster et al. 1988), and correlate well
with adverse histopathologic effects and the presence of lesions in the thymic cortex (Germolec
et al. 2004a, 2004b; Wachsmuth 1983).
180
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Interpretation of decreased thymus weight as an adverse effect is supported by general
immunology literature as well as chemical-specific data. The thymic cortex is known to be a
major site of thymocyte proliferation and selection for maturation, and impairment can lead to
cell-mediated immune suppression (Kuper 2002, 1992; De Waal et al. 1997). Reduced thymus
weight is often attributed to decreased thymocyte proliferation or increased thymocyte apoptosis
in the thymic cortex (Kamath et al., 1997; Vandebriel et al., 1999). DeJong et al. (1999) reported
a decrease in estimated thymic cortex weight at 10, 30, and 90 mg/kg benzo[a]pyrene, and
reduced medulla weight at 90 mg/kg, but did not calculate the ratio between the two. This
suggests that both parts of the thymus were affected by benzo[a]pyrene, but that the cortex may
be more sensitive. De Jong also used immunohistochemistry data to show that cell proliferation
was not affected, suggesting that thymic atrophy may be due to increased rates of thymocyte
apoptosis. However, a study using the murine LLNA showed that proliferation activity
decreased after a single 13 mg/ml oral dose of benzo[a]pyrene (van den Berg et al., 2005).
In addition to thymus effects, decreased B cell percentages in the spleen were observed at
10, 30, and 90 mg/kg-day benzo[a]pyrene in a dose response pattern. The absolute number of B
cells, however, was not significantly lower than control animals until 90 mg-kg/day. There is
also supportive evidence of humoral immune suppression at higher doses, as well as evidence of
overall toxicity of the bone marrow. This theory is supported by the decrease in IgA and IgM
(incating T-cell dependant effects), and the dose-related toxicity of RBCs observed by Dejong et
al. (1999).
The MOA by which benzo[a]pyrene produces immune system effects is not understood,
but several in vitro studies have been conducted to investigate potential contributing
mechanisms. Benzo[a]pyrene induced myelotoxicity in human cord blood cells (Carfi et al.,
2007) and mouse bone marrow cultures (Legraverend et al., 1983), suppressed mouse B cell
lymphopoiesis (Hardin et al., 1992), and inhibited mitogen-induced proliferative responses of
mouse spleen cell cultures (Lee and Urso, 2007). Benzo[a]pyrene inhibition of the proliferative
responses of spleen cells to a mitogen was diminished by the presence of the AhR antagonist and
CYP inhibitor, a-NF, indicating the potential importance of benzo[a]pyrene metabolites in the
immune suppression effect (Lee and Urso, 2007). Similarly, the CYP1A1 inhibitor 1-(1-
propynyl)pyrene blocked B-cell growth inhibition by benzo[a]pyrene, but not through the
metabolite BPDE (Allan et al. 2006).
Carfi et al. (2007) described a series of in vitro assays designed to assess cytotoxicity,
myelotoxicity, cytokine release, and mitogen responsiveness in rat, mouse, and human cells (i.e.,
peripheral lymphocytes and cord blood cells, and spleen cells from rats and mice only). The
cytotoxicity half maximal inhibitory concentration (IC50) value for benzo[a]pyrene was >200 |iM
in human, rat, and mouse cells. Benzo[a]pyrene produced myelotoxicity as evaluated by a dose-
related decrease in colony scoring for the colony forming unit-granulocyte macrophage (CFU-
GM) assay using human cord blood cells. Benzo[a]pyrene reduced the release of specific
181
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
cytokines from HL following phytohemagglutinin (PHA), gamma-interferon (y-INF), and
lipopolysaccharide (LPS) tumor necrosis factor (TNF-a) stimulation. Mitogen responsiveness in
rat and mouse spleen cells following stimulation with LPS or PHA was decreased by exposure to
benzo[a]pyrene. T-lymphocyte proliferation induced by anti-CD3 antibody was also inhibited by
benzo[a]pyrene in HL, but was not affected in mouse spleen cells at the highest concentration
(160 (jM).
Myelotoxicity was also observed in mouse bone marrow cultures exposed to
benzo[a]pyrene as evidenced by decreased cell survival (Legraverend et al., 1983). The findings
in bone marrow cultures from Ah-responsive (C57BL/6) and Ah-nonresponsive (DBA/2) mice
suggest that AhR affinity may play a role in benzo[a]pyrene-induced myelotoxicity.
Benzo[a]pyrene is more toxic to bone marrow cells from C57BL/6 mice in vitro compared to
cells cultured from DBA/2 mice.
Hardin et al. (1992) treated cultured bone marrow cells from DBA/2 and C57BL/6 mice
4 8
with benzo[a]pyrene at concentrations between 10" and 10" M. Benzo[a]pyrene suppressed B
cell lymphopoiesis in a dose-dependent manner even at the lowest concentration used. Bone
marrow cells from Ah-nonresponsive DBA/2 mice were less sensitive to the immunosuppressive
action of benzo[a]pyrene, compared with Ah-responsive C57BL/6 mice. The AhR antagonist
and CYP450 inhibitor a-NF prevented benzo[a]pyrene-induced inhibition of B cell
lymphopoiesis from C57BL/6 mice in a concentration-dependent fashion. Benzo[a]pyrene also
induced apoptosis in cultured bone marrow cells.
Spleen cell cultures derived from C3H/HeJ and CBY/D2 mice were exposed to benzo[a]pyrene
and assessed for T-lymphocyte proliferation in response to mitogenic or antigenic stimulation
(Lee and Urso, 2007). Benzo[a]pyrene (10 [xM) produced an 80% decrease in the allogeneic
mixed lymphocyte response (MLR) assay, which is a measure of the proliferative response to
antigenic stimulation. Benzo[a]pyrene (0.1, 1, and 10 |iM) also produced a dose-dependent
inhibition of the proliferative response to the mitogen Concanavalin A (Con A). This inhibition
did not occur in spleen cells treated with the AhR antagonist and CYP450 inhibitor, a-NF.
BPDE-DNA adducts were detected in CH3 spleen cells cultured with 10 [jM benzo[a]pyrene in
the presence of Con A.
4.6.4.3. Developmental and Reproductive Toxicity Effects
Developmental and reproductive toxicity effects have been associated with oral and
inhalation exposure to benzo[a]pyrene. Impaired fertility, with associated lesions in ovarian
(decreased follicles) and testicular (atrophic seminiferous tubules) tissues, has been observed in
male and female F1 offspring following exposure of F0 female mice to 10 mg/kg-day
benzo[a]pyrene on GDs 7-16; a decrease in the number of F0 females with viable litters was
observed at a higher dose level of 160 mg/kg-day (Kristensen et al., 1995; MacKenzie and
Angevine, 1981). Inhalation exposure of pregnant rats to benzo[a]pyrene:CB aerosols during
182
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
gestation has also been associated with decreased fetal survival and number of pups per litter
(Archibong et al., 2002), decreased levels of plasma progesterone, estradiol, and prolactin
(Archibong et al., 2002), changes in electrophysiological variables in the hippocampus
(Wormley et al., 2004) and decreased cortical neuron activity (McCallister et al., 2008).
Inhalation exposure of adult male rats to benzo[a]pyrene:CB aerosols (75 (j,g/m ) for 10 or 60
days caused decreased circulating and intratesticular concentrations of testosterone, decreased
testis weight, and decreased sperm motility (Archibong et al., 2008; Ramesh et al., 2008; Inyang
et al., 2003).). Acute i.p. exposure studies in adult female DBa/2N or C57BL/6N mice
demonstrated a 50% destruction of primordial follicles with a single dose of 25 mg/kg (Mattison
et al., 1980) and a 15-day exposure of 3 mg/kg-day (Boorman et al., 2000). Other reproductive-
related effects in these studies include decreased fertility (number of pups) (Mattison et al., 1980)
and ovulatory inhibition, as indicated by decreased number of corpora lutea, one week after a
dose of 5 mg/kg and 3-4 weeks after a dose of 100 mg/kg (Miller et al., 1992; Swartz and
Mattison, 1985).
In vivo studies have suggested that the mechanism of decreased fertility in females and
decreased fetal survival may result from changes in circulating hormones (i.e., decreased
progesterone, estradiol-17P, and prolactin levels) responsible for maintaining the uterine
environment in a state that can support embryonic and fetal development (Archibong et al., 2002,
see Section 4.3.2). Several in vitro studies have demonstrated low affinity binding of
benzo[a]pyrene to the estrogen receptor and alteration of estrogen-dependent gene expression
(Liu et al., 2006; Van Lipzig et al., 2005; Vondracek et al., 2002; Fertuck et al., 2001; Charles et
al., 2000); however, the role of these changes in benzo[a]pyrene-induced reproductive toxicity is
unknown. Fertuck et al. (2001) showed in vitro effects of benzo[a]pyrene on estrogen-receptor-
mediated gene expression, but did not demonstrate estrogen-mediated uterotrophic effects
(increased uterine weight or lactoferrin mRNA expression) in ovarectomized C57BL/J6 or
DBA/2 mice following in vivo administration of benzo[a]pyrene (0.1-10 mg/kg-day, p.o., for 3
consecutive days).
The mechanism(s) by which benzo[a]pyrene or its metabolites impair the development of
follicles in the ovary have been the focus of study for more than 30 years (Mattison and
Thorgeirsson, 1977). AHH is found in the ovary, and inhibition of AHH activity reduces the
level of oocyte destruction seen with benzo[a]pyrene exposures in mice (Mattison and
Thorgeirsson, 1979). AHH activation is not sufficient to explain the variation across strains in
oocyte destruction, however (Mattison and Nightingale, 1980). Studies using intraovarian
injection of metabolites of benzo[a]pyrene indicate that the epoxide metabolite (+)-(7R,8S)-diol-
(9S,10R)-epoxide-2 is most strongly correlated with the oocyte counts in exposed mice
(Takizawa et al.,1984).
Most of the loss of oocytes that occurs in utero and throughout the reproductive lifespan
in mice and in women occurs through programmed cell death (apoptosis), which is regulated by
183
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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 protein Bax. Activation of Bax leads to increased oocyte death, and PAHs (including
benzo[a]pyrene) have been demonstrated to activate Bax gene transcription in mice (Matikainen
et al., 2002; 2001). However, an in vitro study of mouse ovarian cells obtained from 4-day old
pups did not show any evidence of increased markers of apoptosis with treatments of
benzo[a]pyrene concentrations of up to 1000 ng/ml for 6 and for 24 hours (Tuttle et al., 2009).
Other mechanisms may be more relevant for the ovulatory inhibition effects seen with
benzo[a]pyrene exposures. In an intro experiment, Neal et al. (2007) demonstrated a dose-
dependent decrease in FSH-stimulated rat follicle growth, with 158, 99, 75, 38, 30 and 38%
change in follicle area for benzo[a]pyrene exposure concentrations of 0, 1.5, 5.0, 15, 45 and 135
ng /ml (p < 0.05 for all differences compared with controls). The authors noted that the lowest
dose at which this effect was seen, 1.5 ng/ml, was similar to the mean concentration of
benzo[a]pyrene seen in follicular fluid samples from women who smoked.
Several in vitro studies have investigated the possible mechanisms for impaired
spermatogenesis by benzo[a]pyrene including enhancement of apoptosis of spermatogonia
(Revel et al., 2001), inhibition of spermatid meiosis (Georgellis et al., 1990), Sertoli cell
cytotoxicity (Raychoudhury and Kubinski, 2003), and altered androgen hormone regulation
(Inyang et al., 2003; Vinggaard et al., 2000).
Revel et al. (2001) reported dose-related increases in apoptosis of spermatogonia
harvested from the vas deferens of male BALB/c mice administered benzo[a]pyrene doses
ranging from 0.5 to 50 mg/kg via s.c. injection for 5 weeks. The competitive AhR inhibitor,
resveratrol (50 mg/kg, s.c.) given simultaneously with 5 mg/kg benzo[a]pyrene s.c. for 5 weeks,
suppressed both BPDE DNA adduct formation and apoptosis, suggesting a role for the AhR in
this benzo[a]pyrene-induced male reproductive toxicity (Revel et al., 2001).
Georgellis et al. (1990) reported that concentrations of 0.1 |iM benzo[a]pyrene incubated
in vitro with seminiferous tubule segments from Sprague-Dawley rats with microsome
preparations from the whole rat testes inhibited meiotic division of the spermatids and was
highly cytotoxic.
Raychoudhury and Kubinski (2003) isolated Sertoli cells from CD rats and incubated the
cells in culture with benzo[a]pyrene. Benzo[a]pyrene was cytotoxic to these cells at 50 and
100 [j,g/mL. Treatment of the cells for 24 hours with 10 [j,g/mL benzo[a]pyrene induced cell
killing through an apoptotic response (as measured by fluorescence labeling of apoptotic DNA
fragments).
Inyang et al. (2003) demonstrated that benzo[a]pyrene inhalation altered circulating
levels or cellular responsiveness to androgenic hormones such as testosterone (see Section 4.3.2).
Vinggaard et al., 2000) showed that benzo[a]pyrene antagonized the human androgen receptor
(hAR) in a sensitive reporter gene assay based on CHO cells transiently cotransfected with a
hAR vector and an MMTV-LUC vector (antiandrogen IC50 of 3.9 |iM),
184
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
4.7. EVALUATION OF CARCINOGENICITY
4.7.1. Summary of Overall Weight-of-Evidence
Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
benzo[a]pyrene is "carcinogenic to humans." This conclusion is based on evidence of
carcinogenicity in humans following exposure to different PAH mixtures containing
benzo[a]pyrene, extensive and consistent evidence of carcinogenicity in laboratory animals
exposed to benzo[a]pyrene via all routes of administration, and strong evidence that the
biological processes leading to benzo[a]pyrene carcinogenesis in laboratory animals are also
present in humans. Bioactivation of benzo[a]pyrene leads to the formation of DNA-reactive
metabolites which can produce mutations in key genes, such as the p53 tumor suppressor gene
and the ras oncogene, leading to tumor formation (see Section 4.7.3.).
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence
There is a large body of evidence for human carcinogenicity for several PAH mixtures
containing benzo[a]pyrene, such as soot, coal tars, coal-tar pitch, mineral oils, and shale oils
(IARC, 2010; Baan et al., 2009; Straif et al., 2005). There is also evidence of carcinogenicity in
occupations involving exposure to PAH mixtures containing benzo[a]pyrene, such as aluminum
production, chimney sweeping, coal gasification, coal-tar distillation, coke production, iron and
steel founding, and paving and roofing with coal tar pitch (IARC, 2010; Baan et al., 2009; Straif
et al., 2005). Increased cancer risks have been reported among other occupations involving
exposure to PAH mixtures such as carbon black and diesel exhaust (Bosetti et al., 2007; Straif et
al., 2005). Benzo[a]pyrene is also a notable constituent of tobacco smoke (IARC 2004). An
increasing number of studies report exposure biomarkers such as benzo[a]pyrene- or PAH-DNA
adducts in white blood cells, and several cohort studies (summarized in Section 4.1) demonstrate
a positive exposure-response relationship with cumulative PAH exposure using
benzo[a]pyrene—or a proxy such as BSM that can be converted to benzo[a]pyrene—as an
indicator substance. Because benzo[a]pyrene is only one of many PAHs that could contribute to
these observed increases in cancer, the epidemiologic studies provide credible but limited
support for a causative role of benzo[a]pyrene in human cancer.
In laboratory animals (i.e., rats, mice, and hamsters), exposure to benzo[a]pyrene via the
oral, inhalation, and dermal routes have been associated with carcinogenic responses both
systemically and at the site of administration. Chronic oral exposure to benzo[a]pyrene was
associated with forestomach and liver tumors in male and female Wistar rats (Kroese et al.,
2001), forestomach tumors in male and female Sprague-Dawley rats (Brune et al., 1981), and
forestomach, esophagus, tongue, and larynx tumors in female B6C3Fi mice (Beland and Culp,
1998; Culp et al., 1998). Auditory canal tumors were also observed in male and female Wistar
rats (Kroese et al., 2001). Repeated or short-term oral exposure to benzo[a]pyrene was
associated with forestomach tumors in more than 10 additional bioassays with several strains of
185
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
mice (see Table 4-4 in Section 4.2.1.2). Chronic inhalation exposure to benzo[a]pyrene was
associated with tumors in the larynx and pharynx of male Syrian golden hamsters exposed to
benzo[a]pyrene:NaCl aerosols (Thyssen et al., 1981). Intratracheal instillation of benzo[a]pyrene
was associated with respiratory tract tumors in more than 10 additional studies with hamsters
(see Section 4.2.2.2 for references). Chronic dermal application of benzo[a]pyrene (2-3
times/week) has been associated with mouse skin tumors in 12 bioassays (see Section 4.2.3.2 for
references). Skin tumors in rats, rabbits, and guinea pigs have also been associated with repeated
application of benzo[a]pyrene to skin in the absence of exogenous promoters (WHO, 1998;
ATSDR, 1995; IARC, 1983, 1973). When followed by repeated exposure to a potent tumor
promoter, acute dermal exposure to benzo[a]pyrene induced skin tumors in numerous studies of
mice, indicating that benzo[a]pyrene is a strong tumor-initiating agent in the mouse skin model
(see Section 4.2.3.1 for references).
Carcinogenic responses in animals exposed to benzo[a]pyrene by other routes of
administration include: (1) liver or lung tumors in newborn mice given acute postnatal i.p.
injections; (2) increased lung tumor multiplicity in A/J adult mice given single i.p. injections; (3)
injection site tumors in mice following s.c. injection; (4) injection site sarcomas in mice
following intramuscular injection; (5) mammary tumors in rats with intramammilary
administration; (6) cervical tumors in mice with intravaginal application; and (7) tracheal tumors
in rats with intratracheal implantation (see Section 4.4.3 for references).
Benzo[a]pyrene is classified as an alternant PAH, or a compound composed solely of
fused benzene rings. Nonalternant PAHs contain both benzene and five carbon rings. Among
alternant PAHs, important structural features related to enhanced mutagenicity and
carcinogenicity include the presence of at least four rings (Bostrom et al., 2002). The
carcinogenic activity of PAH compounds is influenced by specific structural features. Recently,
this knowledge has been exploited in an effort to derive quantitative structure activity
relationship (QSAR) methods to evaluate the relationship between specific PAH structural
features and mechanistic events related to carcinogenesis (Bruce et al., 2008; Vijayalakshmi et
al., 2008). Alternant PAHs having four or more benzene rings exhibit greater carcinogenic
potency than PAHs with two or three benzene rings (Bostrom et al., 2002). The carcinogenic
activity of PAHs is also related to the specific arrangement of the benzene rings. As a general
rule, PAHs with at least four rings and a classic bay- or fjord-region (formed entirely by benzene
rings) may be characterized as containing structural alerts for carcinogenesis. However, this
structural characterization is likely to be overly simplistic and other features may be important to
carcinogenesis.
As discussed in Section 4.5.1.1, several lines of evidence related to tumor initiation
following mutagenicity are available for benzo[a]pyrene including: (1) in vivo detection of
cancer-relevant oncogene/tumor suppressor gene mutations in target tissue; (2) in vivo detection
of DNA adducts in target tissue; (3) in vivo DNA adducts, gene mutations, cytogenetic damage,
186
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
and other measures of primary DNA damage in non-target tissues; and (4) in vitro DNA adduct
formation, mutations, cytogenetic damage, and primary DNA damage in cells from target and
nontarget tissues.
4.7.3. Mode-of-Action Information
4.7.3.1. Hypothesized MO A
The carcinogenicity of benzo[a]pyrene, the most studied and best characterized PAH, is
well documented (Xu et al., 2009; Jiang et al., 2007, 2005; Xue and Warshawsky, 2005; Ramesh
et al., 2004; Bostrom et al., 2002; Penning et al., 1999; WHO, 1998; Harvey, 1996; ATSDR,
1995; Cavalieri and Rogan, 1995; U.S. EPA, 1991b). EPA has concluded that benzo[a]pyrene
induces carcinogenicity via a mutagenic mode of action. Mutagenicity is a well-established
cause of carcinogenicity. This hypothesized mode of action is presumed to apply to all tumor
types and is relevant for all routes of exposure. The principal key events associated with the
mode of action for benzo[a]pyrene include: (1) bioactivation of benzo[a]pyrene to reactive
metabolites (2) direct DNA damage by the reactive metabolites, including the formation of DNA
adducts (3) formation and fixation of DNA mutations, particularly in tumor suppressor genes or
oncogenes and (4) clonal expansion of mutated cells. These events are depicted in Figure 4-3.
r
Key Events in the MO A for benzo[a]pyrene Carcinogenicity
Exposure Metabolism Initiation Promotion Progression
diol epoxide
Binding to AhR
Mutation
(Transversion)
K-Ras, H-Ras
and p53 targets
DNA Adducts
Upregulation of
genes related to
biotransformation,
growth, and
differentiation
radical cation
Mutation
(Depurination)
H-Ras target
Proliferation of
initiated cells
benzo[a n tctr
DNA Adducts
Neoplasm
Inflammatory
response
o-quinone and
ROS
Mutation
(Depurination)
Oxidative damage
and strand
DNA Adducts
and oxidative
base damage
Cytotoxicity
187
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure 4-3. Proposed principal pathways and key events in the
benzo[a]pyrene carcinogenic MOA.
4.7.3.2. Experimental Support for the Hypothesized MOA
Strength, consistency, specificity of association. There is an extensive database of in vitro and in
vivo studies demonstrating the genotoxicity and mutagenicity of benzo[a]pyrene following
metabolic activation (see Tables 4-24, 4-25 and 4-26). In vitro studies overwhelmingly support
the formation of DNA adducts, mutagenesis in bacteria, yeast and mammalian cells, several
measures of cytogenetic damage (CA, SCE, MN), and DNA damage. In vivo systems in animal
models are predominantly positive for somatic mutations following benzo[a]pyrene exposure.
Additionally, some evidence exists that benzo[a]pyrene can induce mutations in germ cells.
Benzo[a]pyrene is thought to be converted into reactive intermediates via three principal
metabolic pathways: (1) activation to a reactive diol epoxide via CYP1A1/1B1 and epoxide
hydrolase; (2) activation to a reactive radical cation via CYP peroxidases and (3) activation to a
reactive and redox active o-quinone metabolite via AKR1 Al and AKR1C1-1C4 (Xu et al., 2009;
Jiang et al., 2007, 2005; Xue and Warshawsky, 2005; Penning et al., 1999; Harvey 1996;
Cavalieri and Rogan, 1995). All three of these pathways (discussed in detail in Sections 3.3. and
4.5.2) lead to DNA damage including DNA adducts, depurination, and/or oxidative damage to
DNA. DNA damage, if not correctly repaired prior to replication, can subsequently give rise to
mutations.
Benzo[a]pyrene-DNA adducts, biomarkers of exposure and of effect, have been
extensively demonstrated with in vitro cell systems, in vivo animals studies, and in human target
tissues, including skin and lung (see Section 4.1.2.). Specifically, elevated BPDE-DNA adducts
have been observed in coke oven workers and chimney sweepers, occupations associated with
increased risks of cancer from complex PAH-containing mixtures (Pavanello et al., 1999).
BPDE-DNA adducts were also found to be elevated in the lungs of cigarette smokers with lung
cancer (Godschalk et al., 2002; Phillips et al., 2002; Bartsch et al., 1999; Alexandrov et al.,
1992). Multiple epidemiological studies have indicated that PAH exposed individuals who are
homozygous for a CYP1 Al polymorphism which increases the inducability of this enzyme (thus
increasing the production of reactive diol epoxide metabolites) have increased levels of
benzo[a]pyrene-DNA adducts (Bartsch et al., 2006; Aklillu et al., 2005; Alexandrov et al., 2002;
Perera and Weinstein, 2000). In addition, this population of individuals also has a greater risk of
certain tumors, including those of the lung.
Mutations in the K-ras, H-ras, and p53 genes were assessed in forestomach tumors of
mice fed benzo[a]pyrene in the diet for 2 years (Culp et al., 2000). Forestomach tumors had K-
ras mutations (68% of tumors) that were G—>T or C transversions in codon 12 or 13. H-ras
(codon 13) and p53 mutations characterized as G—>T or C transversions were also found.
188
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
K-ras mutations were observed in the A/J mouse lung tumor model following IP
treatment with benzo[a]pyrene, and were observed to be qualitatively different than K-ras
mutations in spontaneous lung tumors from control animals (Nesnow et al., 1996). Lung tumor
DNA was isolated and DNA sequence analysis of K-ras mutations was performed for 19 separate
lung tumors in the benzo[a]pyrene treated group and the control group. The DNA sequence
analysis demonstrated several guanine mutations at codon 12, which were different than the
spectrum of mutations found in untreated animals. Specifically, the frequency of GGT—>TGT
transversion mutations were significantly higher in the benzo[a]pyrene treated animals compared
to controls (56% vs. 0%) whereas GGT—>GAT transition mutations were the predominant
mutation in lung tumors from control animals (58% vs. 19% in benzo[a]pyrene group).
Some human data exist which correlate the frequency of PAH-DNA adducts with gene
mutations in highly PAH-exposed populations. In a study of iron foundry workers (a high PAH
exposure population which has been demonstrated to have an increased risk of lung cancer
[Bosetti et al., 2007]), biological samples from workers were analyzed for DNA adducts and
somatic gene mutations at the hprt locus (Perera et al. 1994, 1993). A strong correlation between
PAH-DNA adduct levels and incidence of hprt mutations was observed in individuals with
detectable levels of adducts, indicating that somatic mutations were increased in parallel with
PAH-DNA adducts in workers exposed to PAHs.
Data in humans from a study by Marini et al. (2001) indicate that the types of mutations
commonly found in response to benzo[a]pyrene exposure in in vitro and animal models are
similar to the spectrum of mutations in critical tumor suppressor genes and/or oncogenes in
populations highly exposed to PAHs. DeMarini et al. (2001) demonstrated mutations in the p53
tumor suppressor gene and the K-ras oncogene in lung tumors obtained from 24 nonsmoking
women from China, whose tumors were associated with exposure in their homes to smoky coal
from the use of stoves with no chimneys. The observed mutations in lung tumors were primarily
G—>T transversions at either K-ras or p53. Mutation hotspots in the lung tumors corresponded
with hot spots for PAH mutations (codon 154, codon 249, and codon 273).
Dose-response concordance and temporal relationship. The metabolism of benzo[a]pyrene to
reactive metabolites is a necessary event which precedes mutagenesis. Mutation assays of
benzo[a]pyrene in salmonella typhimurium are overwhelmingly positive with the inclusion of S9
metabolic liver fractions, but are negative without the addition of the S9 metabolic enzymes (see
Table 4-24).
In mice, a dose-response and temporal relationship has been demonstrated between the
formation of BPDE-DNA adducts and skin and forestomach tumors. In a study using mice
treated dermally with benzo[a]pyrene once or twice per week for 15 weeks, a linear dose-
response of benzo[a]pyrene-induced adducts in the skin, lung, and liver was observed (Talaska et
al. 1996). Another study examined the dose-response relationship and the time course of
189
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
benzo[a]pyrene-induced skin damage, DNA adduct formation, and tumor formation in female
mice. Mice were treated dermally with 0, 16, 32, or 64 jag of benzo[a]pyrene once per week for
29 weeks (Albert et al., 1991). Indices of skin damage and levels of BPDE-DNA adducts in skin
reached plateau levels in exposed groups by 2-4 weeks of exposure. With increasing dose level,
levels of BPDE-DNA adducts (fmol/[j,g DNA) initially increased in a linear manner and began to
plateau at doses >32 [j,g/week. Tumors began appearing after 12-14 weeks of exposure for the
mid- and high-dose groups and at 18 weeks for the low-dose group. At study termination
(35 weeks after start of exposure), the mean number of tumors per mouse was approximately one
per mouse in the low- and mid-dose groups and eight per mouse in the high-dose group. The
time-course data indicate that benzo[a]pyrene-induced increases in BPDE-DNA adducts
preceded the appearance of skin tumors, consistent with the formation of DNA adducts as a
precursor event in benzo[a]pyrene induced skin tumors.
Culp et al. (1996a) compared dose-response relationships for BPDE-DNA adducts and
tumors in female B6C3Fi mice exposed to benzo[a]pyrene in the diet at 0, 18.5, 90, or 350
[j,g/day for 28 days (to examine adducts) or 2 years (to examine tumors). The benzo[a]pyrene
dose-tumor response data showed a sharp increase in forestomach tumor incidence between the
18.5 (J,g/day group (6% incidence) and the 90 (J,g/day group (78% incidence). The BPDE-DNA
adduct levels in forestomach showed a relatively linear dose-response throughout the
benzo[a]pyrene dose range tested. The appearance of increased levels of BPDE-DNA adducts in
the target tissue at 28 days is temporally consistent with the contribution of these adducts to the
initiation of forestomach tumors. Furthermore, about 60% of the examined tumors had
mutations in the K-ras oncogene at codons 12 and 13, which were G—>T or G—>C transversions
indicative of BPDE reactions with DNA (Culp et al., 1996a).
Biological plausibility and coherence. A mutagenic MOA for benzo[a]pyrene is supported by a
large body of research over time with consistent evidence of benzo[a]pyrene activation to
reactive metabolites leading to DNA-damage and mutational events associated with tumor
initiation. Mutagenicity as a mode of action for carcinogenicity in humans is generally accepted
and is a biologically plausible mechanism for tumor induction. The formation of DNA adducts
and oncogene/tumor suppressor mutations in organs that also displayed an increase in tumor
incidence in rats and mice indicates coherence of these effects. Benzo[a]pyrene has been shown
to be mutagenic in vivo and in vitro, across species and tissue types.
4.7.3.2. Other Possible MO As
In addition to mutagenicity, other MO As which contribute to the carcinogenicity of
benzo[a]pyrene are possible, but are not as well studied. The tumor promotion properties of
benzo[a]pyrene may involve cell proliferative responses to cytotoxicity or apoptosis from
benzo[a]pyrene metabolites, AhR-mediated effects on cell growth and differentiation, or anti-
190
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
apoptotic signals elicited by metabolites (Burdick et al., 2006, 2003; Chen et al., 2003; Nebert et
al., 1993). Results from some studies indicate that exposure to benzo[a]pyrene or its metabolites
increases the production of inflammatory cytokines, such as IL-1, which may contribute to tumor
promotion (N'Diaye et al., 2006; Tamaki et al., 2004; Gar<;on et al., 2001a, b). Benzo[a]pyrene
also has been shown to inhibit GJIC, a characteristic associated with well-known tumor
promoters such as TPA (Sharovskaya et al., 2006; Blaha et al., 2002). In summary, though there
are limited data which support other processes that may contribute to the carcinogenicity of
benzo[a]pyrene (inflammation, cytotoxicity, anti-apoptotic signaling, etc.); the available
evidence indicates that the primary mode of action for benzo[a]pyrene involves DNA reactivity
and mutagenicity leading to carcinogenesis.
4.7.3.4. Conclusions About the Hypothesized Mode of Action
The MOA of mutagenicity of benzo[a]pyrene through reactive metabolites is extensively
supported by a large body of research. Mutations from DNA reactive benzo[a]pyrene
metabolites occur as early events in the carcinogenic process and are not believed to be acquired
following cytoxicity or regenerative proliferation. Several lines of evidence relating to
mutagenicity and tumor initiation are available for benzo[a]pyrene including: in vitro evidence
of DNA adducts, mutations, cytogenetic damage, and primary DNA damage; in vivo DNA
adducts, gene mutations, cytogenetic damage, and other measures of primary DNA damage;
detection of DNA adducts in target tissue in vivo; and detection of cancer-relevant
oncogene/tumor suppressor gene mutations in target tissue in vivo. Taken together, these data
provide support for a mutagenic MOA for benzo[a]pyrene-induced cancer.
Support for the hypothesized MOA in test animals
Benzo[a]pyrene induces gene mutations in a variety of in vivo and in vitro systems and
produces tumors in all animal species tested and all routes of exposure. Strong, consistent
evidence indicates that the postulated key events: the metabolism benzo[a]pyrene to a DNA-
reactive intermediates, the formation of DNA adducts, and the occurrence of subsequent
mutations in oncogenes and tumor suppressor genes occur in animal models.
Relevance of the hypothesized MOA to humans
Mutagenicity is a well-established cause of carcinogenicity. A substantial database of
information on benzo[a]pyrene indicates that the postulated key events: the metabolism of
benzo[a]pyrene to a DNA-reactive intermediates, the formation of DNA adducts, and the
formation of subsequent mutations in oncogenes and tumor suppressor genes all occur in human
tissues. The following lines of evidence from human studies provide support that the
hypothesized mutagenic MOA is relevant to humans: the activation of benzo[a]pyrene to DNA
reactive metabolites occurs in humans in qualitatively and quantitatively similar manner
191
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
compared to animals; DNA adducts specific to benzo[a]pyrene have been found in a wide variety
of human tissues and are elevated in populations exposed to high levels PAHs; increased DNA
mutations have been strongly associated with increasing levels of PAH-DNA adducts in workers
occupationally exposed to PAHs; and the spectra of benzo[a]pyrene induced mutations in p53
tumor suppressor genes and ras oncogenes observed in controlled in vitro cell systems and in
vivo animal studies are similar to the spectra of mutations in tumors from PAH-exposed humans.
Populations or life stages particularly susceptible to the hypothesized MO A
The mutagenic mode of action is considered relevant to all populations and lifestages.
The current understanding of biology of cancer indicates that mutagenic chemicals, such as
benzo[a]pyrene, are expected to exhibit a greater effect in early life versus later life exposure
(U.S. EPA, 2005b; Vesselinovitch et al., 1979). Although the developing fetus and infants may
have lower levels of some bioactivating enzymes than adults (e.g., CYP1A1/1B1), infants or
children are expected to be more susceptible to benzo[a]pyrene-induced cancer at certain tissue
sites. Newborn or infant mice developed liver and lung tumors more readily than young adult
mice following acute i.p. exposures to benzo[a]pyrene (Vesselinovitch et al., 1975; see Section
4.8.1). These results indicate that exposure to benzo[a]pyrene during early life stages presents
additional risk for cancer, compared with exposure during adulthood. The Supplemental
Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA,
2005b) recommends the application of age-dependent adjustment factors (ADAFs) for
carcinogens that act through a mutagenic mode of action. Given the weight of the available
evidence, benzo[a]pyrene acts through a mutagenic mode of carcinogenic action and the ADAFs
should be applied.
Population variability in metabolism and detoxification of benzo[a]pyrene, in addition to
DNA repair capability, may affect cancer risk. Polymorphic variations in the human population
in CYP1A1, CYP1B1, and other CYPs have been implicated as determinants of increased
individual lung cancer risk in some studies (Aklillu et al., 2005; Alexandrov et al., 2002; Perera
and Weinstein, 2000). The Phase II cytosolic GST, mediated by variants of the GSTM1 and
GSTT1 genes, prevents the formation of BPDE-DNA adducts. Some evidence suggests that
humans lacking a functional GST gene have higher BPDE-DNA adduct levels and thus are at
greater risk for cancer (Vineis et al., 2007a; Pavanello et al., 2004; Perera and Weinstein, 2000;
Alexandrov et al., 2002). In addition, acquired deficiencies or inherited gene polymorphisms
that affect the efficiency or fidelity of DNA repair may also influence individual susceptibility to
cancer from environmental mutagens (Matullo et al., 2003; Shen et al., 2003; Cheng et al., 2000;
Perera and Weinstein, 2000; Wei et al., 2000; Amos et al., 1999). In general, however, available
support for the role of single polymorphisms in significantly modulating human PAH cancer risk
is relatively weak or inconsistent. Combinations of metabolic polymorphisms, on the other hand,
192
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
may be critical determinants of cumulative DNA-damaging dose, and thus susceptibility to
cancer from benzo[a]pyrene exposure.
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
4.8.1. Possible Childhood Susceptibility
Increased childhood susceptibility to benzo[a]pyrene is supported by several lines of
evidence including epidemiological studies reporting associations between adverse birth
outcomes and developmental effects and internal biomarkers of exposure to benzo[a]pyrene,
presumably via exposure to complex PAH mixtures (Tang et al. 2008, 2006; Perera et al.,
2005a,b). The occurrence of BPDE-DNA in maternal and umbilical cord blood in conjunction
with exposure to ETS was associated with reduced birth weight and head circumference in
pregnant women living in the vicinity of fires from the 09/11/2001 disaster site in New York
City (Perera et al., 2005a). In other studies, elevated levels of BPDE-DNA adducts in umbilical
cord blood were associated with: (1) reduced birth weights or reduced head circumference in the
offspring of 529 Dominican or African-American nonsmoking women (Perera et al., 2005b); and
(2) decreased body weight at 18, 24, and 30 months and deficits in several areas of development
as assessed by the Gesell Developmental Schedules at 24 months in the offspring of nonsmoking
Chinese women living in the vicinity of a coal-fired power plant (Tang et al., 2008, 2006).
Developmental neurotoxicity
Studies in humans and experimental animals indicate that exposure to PAHs in general,
and benzo[a]pyrene in particular, may impact neurological development at relatively low
exposure levels. Observational studies in humans have suggested associations between
gestational exposure to PAHs and later measures of neurodevelopment (Perera et al., 2009; Tang
et al., 2008). An observational study of a Chinese population living in close proximity to a coal
fired power plant found increased levels of benzo[a]pyrene-DNA adducts in cord blood were
associated with decreased developmental quotients in offspring (Tang et al., 2008). In addition,
a study of pregnant women living or working near the World Trade Center site in NYC found
high PAH exposure during pregnancy was associated with a reduction in verbal and full scale IQ
of offspring at 5 years of age (Perera et al., 2009).
A study in pregnant rats exposed by inhalation showed a dose related increase in
benzo[a]pyrene metabolites in the cerebral cortex and hippocampus of pups, indicating the fetal
brain is exposed to benzo[a]pyrene and/or its metabolites following maternal inhalation exposure
(Wu et al., 2003). Another study which treated pregnant rat dams to benzo[a]pyrene by
inhalation found a decrease in long term potentiation (LTP) in the hippocampus of gestationally
treated pups compared with controls, indicating a possible effect on learning and memory in
benzo[a]pyrene exposed animals, though functional tests were not conducted (Wormley et al.,
2004). Another study by the same group treated rat dams by gavage with low levels of
193
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
benzo[a]pyrene (300 |ig/kg) on GD 14-17 and observed benzo[a]pyrene metabolites in the brains
of pups and diminished cortical neuronal activity to sensory input in exposed offspring compared
with controls (McCallister et al., 2008). In addition, a study of mice exposed to benzo[a]pyrene
lactationally observed statistically significant differences in performance in several neuromotor
and behavioral tests which indicated decreased righting reflex and disinhibition behavior
(Bouayed et al., 2009).
Reproductive effects
Epidemiological studies indicate that exposure to complex mixtures of PAHs, such as
through cigarette smoke, is associated with measures of decreased fertility in humans (El Nemr
et al., 1998; Neal et al., 2005) and that prenatal exposure to cigarette smoking is associated with
reduced fertility of women later in life (Weinberg et al., 1989). A case-control study in a
Chinese population has also indicated that women with elevated levels of benzo[a]pyrene-DNA
adducts in maternal blood were four times more likely to have experienced a missed abortion
(Wuetal., 2010).
Oral multigenerational studies of benzo[a]pyrene exposure in mice demonstrate effects
on fertility and the development of reproductive organs in male and female offspring exposed to
benzo[a]pyrene during development at levels in which no overt toxicity or depression in fertility
is seen in the parental animals (Mackenzie and Angevine 1981; Kristensen et al., 1995).
MacKenzie and Angevine (1981) exposed groups of mice to benzo[a]pyrene on GDs 7-
16 and reproductive outcomes of the offspring were investigated. Fertility of the F1 generation
was decreased in a dose dependant manner. At maturity, the fertility of these animals was tested.
The F1 male and F1 female fertility indices were significantly decreased in each exposure group.
These reductions in fertility indices were associated with decreased testes and ovary weight in
the F1 animals. Male offspring showed histological damage of the seminiferous tubules and
female offspring had hypoplastic ovaries with few follicles and corpora lutea. Similar results
were reported in a study in which female mice were exposed by gavage to benzo[a]pyrene on
GDs 7-16 (Kristensen et al., 1995). F1 females had decreased mean ovary weight and reduced
fertility. At necropsy, the F1 females had reduced ovary weights with decreased numbers of
small, medium, or large follicles and corpora lutea. Inhalation exposure of pregnant female rats
to benzo[a]pyrene:CB aerosols during gestation has also been associated with decreased fetal
survival and number of pups per litter associated with decreased levels of plasma progesterone,
estradiol, and prolactin (Archibong et al., 2002).
These reductions in fertility observed in animal models are supported by a large database
of animal studies in adult animals indicating that benzo[a]pyrene is ovotoxic with effects
including decreased ovary weight, decreased primordial follicles, and reduced fertility (Mattison
et al., 1980; Swartz and Mattison 1985; Miller et al., 1992; Borman et al, 2000).
194
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Developmental Immune effects
The severity and persistence of immune effects observed during in utero studies suggests
that immunotoxicity may be greater during gestation than adulthood (Dietert and Pieperbrink,
2006; Holladay and Smialowicz, 2000). Urso and Gengozian (1982) provide experimental
support demonstrating immunosuppression from benzo[a]pyrene exposure during gestation was
greater than for mice exposed after birth to a 25-fold higher dose. There is also substantial
general literature indicating that disruption of the immune system during certain critical periods
of development (e.g., initiation of hematopoiesis; migration of stem cells; expansion of
progenitor cells) may have significant and lasting impacts on lifetime immune function (e.g.
Burns-Naas et al., 2008; Dietert, 2008; Landreth et al., 2002; Dietert et al., 2000), as well as
more specific studies showing increased dose sensitivity and disease persistence from
developmental versus adult chemical exposure (reviewed in Luebke et al., 2006).
Thymus toxicity is a sensitive and specific effect of benzo[a]pyrene and has been
observed in both prenatal and adult exposure studies. The thymus serves as a major site of
thymocyte proliferation and selection for maturation, and impairment can lead to cell-mediated
immune suppression (Kuper 2002, 1992; De Waal et al., 1997). The thymus is believed to be
critical for T lymphocyte production during early life and not in adulthood (Hakim et al., 2005;
Schonland et al., 2003; Petrie et al., 2002; Mackall et al., 1995). Therefore, the decreases in
thymus weight observed in studies of adult animals exposed to benzo[a]pyrene suggest that
immunosuppression may be a heightened concern for individuals developmentally exposed to
benzo[a]pyrene.
Cancer
As mentioned above in section 4.7.3.4, investigations in young animals exposed to
benzo[a]pyrene provide evidence that early life exposure may present increased risk of cancer.
Comparisons of cancer responses in newborn (1 day old), infant (15 days old), and young adult
(42 days old) mice indicate that exposure to benzo[a]pyrene during early life stages can present
additional risk for cancer, compared with exposure during young adulthood (Vesselinovitch et
al., 1975), but studies designed to compare risks of cancer from early-life (including gestational
and pre-weaning) plus chronic adulthood exposures with risks from chronic adulthood exposure
alone were not located. Following i.p. injection of single doses of 75 or 150 mg/kg
benzo[a]pyrene to newborn (1 day old), infant (15 days old), or young adult (42 days old),
newborn and infant mice more readily developed tumors than young adult mice in the liver and
lung, the most predominant tissue sites of cancer development under these exposure conditions
(Vesselinovitch et al., 1975). The benzo[a]pyrene-exposed groups also displayed increased
incidences of stomach and lymphoreticular tumors, but the data indicated that these tumors
developed more readily with exposure at 42 or 15 days, compared with exposure on PND 1 .
195
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
4.8.2. Possible Gender Differences
Cheng et al. (2007) had conducted a study in which they found that lung tumor tissue
from nonsmoking females contained higher benzo[a]pyrene-DNA adduct levels than that from
nonsmoking males. The adduct levels were associated with CYP1 Al protein levels in the same
tissues. Female lung cancer tissue contains higher levels of DHH activity than male lung cancer
tissue. DHH is an enzyme that can divert benzo[a]pyrene-7,8-dihydrodiol into the quinone
pathway, thus preventing BPDE and DNA adduct formation. It is highly expressed in liver, but
only weakly in lung. Cheng et al. (2007) decided to investigate benzo[a]pyrene-DNA adduct
formation in several lung cancer cell lines to elucidate the roles of CYP1 Al and DHH in this
process. They found that DNA adduct levels were increased in cell lines that contain elevated
CYP1 Al and DHH isoform 1 activities. When DHH1 activity was blocked in these cells, DNA
adduct levels were increased. Benzo[a]pyrene-DNA adduct levels in 120 lung tumor samples
were associated with the protein levels of CYP1 Al, but not DHH1. Comparing tumor tissues
from both genders lung cancer patients they observed that a significantly higher percentage of
female lung tumors had measurable CYP1 Al levels, but were negative for DHH1, compared
with male tumors. The authors suggested that a gender difference in DHH1 activity was in part
responsible for the increased lung tumor incidence in females.
Chang et al. (2007) conducted a study also based in the increased incidence of lung
cancer in human females, but focused on the benzo[a]pyrene interaction with estrogen that result
in elevated COX-2 expression. COX-2 (aka PHS-2) can activate the procarcinogen
benzo[a]pyrene-7,8-dihydrodiol to BPDE. Human bronchial epithelial cells were treated with
benzo[a]pyrene and/or 17(3-estradiol. The combined, but not the individual treatments induced
COX-2 expression. The authors considered their findings as mechanistic evidence towards
understanding the gender difference in susceptibility towards benzo[a]pyrene.
Taioli et al. (2007) reviewed the evidence for a connection between MPO polymorphism
and lung cancer (a more detailed overview of this study is given in Section 4.8.3.3). MPO
converts benzo[a]pyrene metabolites into highly reactive epoxides and a known polymorphism
in its promoter region is said to afford some protection from lung cancer. Several genetic
variants of MPO are known, most of which result in deficiency of the enzyme. The MPO-G/G
genotype (WT) is said to be associated, among others, with acute promyelocytic leukemia,
aerodigestive tract cancer, coronary artery disease, early-onset multiple sclerosis, and an
increased incidence of Alzheimer disease. A multi-study analysis was conducted after several
epidemiologic studies had suggested an association between this gene polymorphism and lung
cancer incidence. The data were stratified for ethnicity, age, gender, and smoking status but
neither age nor gender showed any association for MPO polymorphism and lung cancer risk.
The authors hypothesized that age- and gender-related associations with MPO genotype and lung
cancer incidence, as had been observed in other studies, may be related to age- and gender-
dependent smoking habits rather than to the gene polymorphism itself.
196
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Mammary epithelium and tissues from the female genital tract have the ability to activate
benzo[a]pyrene. Morris and Seifter (1992) made a strong case for benzo[a]pyrene as a potential
causative agent in breast cancer, not only based on the anti-estrogenic action of benzo[a]pyrene
but also on its tendency to accumulate in adipose and hence breast tissue. Jeffy et al. (2002) also
pointed out that PAHs are risk factors for breast cancer. Because the populations in the studies
presented in Section 4.1.4 were predominantly, if not exclusively males, data for breast or
cervical cancers in relation to benzo[a]pyrene exposure are not available.
Gender differences in the response to benzo[a]pyrene have been demonstrated in
numerous animal studies (Knuckles et al., 2001; Kroese et al., 2001; Ramesh et al., 2001a, 2000;
Hood et al., 2000; Rodriguez et al., 1997; Weyand et al., 1994; Turusov et al., 1990; Weyand and
Bevan, 1987). The differences ranged from variations in feed intake, with or without effects on
body weight, to differences in disposition, all the way to different susceptibility towards
benzo[a]pyrene-induced cancers. In some studies, females and males were dosed differently; in
general, females were more resistant to benzo[a]pyrene toxicity than males. None of the studies
presented cogent explanations for the observed differences.
In the 2-year bioassay by Kroese et al. (2001), female rats had fewer tumors of the
forestomach and auditory canal, but more tumors of the liver, compared with males. There was a
rather striking negative dose response for pituitary tumors in females (eight to zero tumors from
control to highest dose), but not in males. Brune et al. (1981) used both sexes of animals in their
study, but did not report their findings for the sexes separately, allowing the conclusion that no
obvious sex differences were observed.
Soyka (1980) found that prenatal treatment of mice with benzo[a]pyrene affected
response to an enzyme-inducing challenge with 3-MC in a gender-specific way when the animals
were 3 months old. Female offspring of benzo[a]pyrene-treated mice had significantly elevated
hepatic microsomal aminopyrine demethylase activity, while CYP450 levels were significantly
lower in male offspring.
Sharma et al. (1997) specifically attempted to resolve the gender difference in cancer
susceptibility of CD-I mice. They focused on glutathione S-transferase % (GSTP) because it
detoxifies BPDE. They noted that constitutive expression of GSTP in the liver of the male CD-I
mouse was higher than in the female and that GSTP activity was much more inducible by the
antioxidant butylhydroxyanisole in the female than in the male mouse. They reported that
female mice were more susceptible to the carcinogenic effect of benzo[a]pyrene than males, but
only females could be partially protected from benzo[a]pyrene-induced carcinogenesis by the co-
administration of butylhydroxyanisole. This is an indication that Phase II enzymes may play a
role in the gender difference towards benzo[a]pyrene toxicity.
Martin et al. (2004) used a transgenic mouse model to address the question of gender
differences. They used p53 heterozygous Tg. AC (v-Ha-ras) mice, a strain possessing a
carcinogen-inducible ras oncogene, but only one functional p53 tumor suppressor gene. The
197
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
authors emphasized that these two mutations have been observed frequently in human tumors.
Female and male animals received 20 mg/kg benzo[a]pyrene by gavage in corn oil twice weekly
for 10 weeks. Eighteen weeks after termination of dosing, tissues were collected for histologic
processing. There were evident differences in gender response (percent of total, female
control/male control:female treated/male treated): mortality (60/60:70/40), thymus hyperplasia
(9/30:27/10), bladder papilloma (0/0:0/13), hepatocyte vacuolization (54/92:86/55),
hematopoietic cell proliferation in spleen (57/8:82/27), lymph node hyperplasia (0/0:0/33), and
malignant lymphoma (0/0:7/36). Neoplasia of the forestomach did not show a gender difference.
The focus of this study was evaluation of feed modifications in carcinogenesis studies (some of
the animals were given N-acetyl cysteine in the feed); the authors did not speculate on the
reasons for the gender differences. Wei et al. (2000) (see Section 4.1.4) observed that female
lung cancer patients displayed less DNA repair capacity than males.
benzo[a]pyrene has also been described as an anti-androgen (see Sections 4.4.1.3 and
4.5.3) in animal studies. The AhR mediates anti-estrogenic effects of its ligands (see Section
4.6.3.2). The findings in animals—anti-estrogenic and potential protection from breast cancer—
are at odds with postulates made for humans that benzo[a]pyrene may advance the development
of breast cancer (Jeffy et al., 2002; Morris and Seifter, 1992). Li et al. (1999) reported that 41%
of the samples of noncancerous breast tissue from breast cancer patients contained
benzo[a]pyrene-like DNA adducts, while no such adducts were detected in tissues obtained from
breast reduction surgery patients. Gender-specific expression of glycine N-m ethyl transferase
(GNMT) has been shown for mice (Section 4.5.2) and might help to explain gender-related
effects, including cancer formation, in this species. Applicability to human populations,
however, has not been addressed.
4.8.3. Genetic Polymorphisms
The metabolic formation and subsequent binding to critical positions in DNA of ultimate
carcinogenic forms of PAH are recognized as key mechanistic events in tumorigenesis.
Increased PAH exposure concentrations are associated with increased levels of DNA adducts and
other biomarkers in both target and surrogate tissues. Observed biomarker levels vary
considerably even among persons with apparently comparable exposures (Garte et al., 2007;
Perera and Weinstein, 2000). While laboratory variation and uncertain exposure estimates
contribute to the differences, inter-individual variation in PAH metabolism (activation and
detoxification) may be especially important. In particular, heritable metabolic gene (single
nucleotide polymorphisms [SNPs]) appear in some cases to influence individual susceptibility to
specific types of cancer, in addition to factors such as ethnicity, age, gender, nutrition, hormonal
and immune status, and preexisting health impairment.
In humans, benzo[a]pyrene is metabolized to the highly DNA-reactive BPDE by
microsomal CYP Phase I enzymes, mediated primarily by the CYP1A1 and CYP1B1 genes.
198
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Polymorphic variations in the human population that result in high inducibility of CYP1 Al,
CYP1B1, and other CYPs have been implicated as determinants of increased lung cancer risk in
some studies (Aklillu et al., 2005; Alexandrov et al., 2002; Perera and Weinstein, 2000). The
Phase II cytosolic GST, mediated by variants of the GSTM1 and GSTT1 genes, prevent the
formation of BPDE-DNA adducts. Some evidence suggests that humans lacking a functional
GST gene have higher PAH(BPDE)-DNA adduct levels and thus be at greater risk for cancer
(Vineis et al., 2007a; Pavanello et al., 2004; Alexandrov et al., 2002; Perera and Weinstein,
2000). Similarly, acquired deficiencies or inherited gene polymorphisms that affect the
efficiency or fidelity of DNA repair may also influence individual susceptibility to cancer
(Matullo et al., 2003; Shen et al., 2003; Cheng et al., 2000; Perera and Weinstein, 2000; Wei et
al., 2000; Amos et al., 1999). In general, however, support for the role of SNP in significantly
modulating PAH cancer risk is relatively weak (i.e., small increases in cancer risk) or
inconsistent, possibly due in part to small study size and the use of DNA-adduct detection
methods with low specificity (i.e., bulky DNA-adducts) and sensitivity. Combinations of
metabolic polymorphisms, on the other hand, are receiving increased attention as critical
determinants of cumulative DNA-damaging dose, and thus individual cancer risk.
Following a report from Japan that the GSTM1 null genotype combined with a mutated
CYP1 Al genotype was associated with increased lung cancer risk, Rojas and coworkers (2000)
measured specific BPDE-DNA adducts in leukocytes (HPLC with fluorometric detection) to
evaluate the impact of CYP1A1, GSTM1, and GSTT1 genotype combinations. The human
subjects were 89 PAH-exposed coke oven workers (smokers and nonsmokers) and 44 power
plant workers (all smokers) not occupationally exposed to PAH. Increased adduct levels were
significantly correlated with CYP1A1 polymorphism, occupational PAH exposure, and smoking.
Combinations of genotypes were observed to have a significant impact on BPDE-DNA adducts,
ranging from the absence of adducts in subjects with the active CYP1A1/GSTM1 genotypes to
the highest BPDE-DNA adduct level in the most susceptible combination of mutated CYP1 Al
with null GSTM1 genotype. The results provide mechanistic support for distinguishing high-
susceptibility benzo[a]pyrene-exposed subgroups, and for understanding their association with
increased cancer rates.
Pavanello et al. (2005) studied associations between xeroderma pigmentosum
(XP)-linked gene polymorphisms, the GSTM1 polymorphism, and bulky BPDE-type DNA
adduct formation in peripheral lymphocytes from 67 highly PAH-exposed male Polish coke oven
workers. The four XP genotypes studied impart low NER capacity, while the GSTM1 active or
null genotypes are associated with effective or ineffective removal of biologically active
benzo[a]pyrene metabolites via GSH conjugation. Workers were questioned for smoking habits,
charbroiled meat consumption, and other factors that might have affected their PAH exposure.
PAH exposure was assessed via urinary 1-OH-pyrene levels. There was a statistically significant
difference in the number of DNA adducts between the GSTM1 active and null carriers (3.37 ±
199
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
o
2.20 vs. 6.73 ± 6.61 adducts per 10 nucleotides, respectively). For the DNA repair gene
polymorphisms, there was an increase in DNA adduct numbers from homozygous WT carriers to
low-DNA-repair homozygous variant carriers. This difference was statistically significant for
homozygous carriers of the XPC-PAT and XPA-A23G variants, but not for the homozygous
XPD-Lys75 lGln and XPD-Asp312Asn variants. Individuals with a combination of DNA repair-
unfavorable XPC or XPA genotypes and the GSTMl-null variant generally fell into the highest
tertile of DNA adduct numbers. Smoking status and diet did not influence urinary 1-OH-Py or
BPDE-DNA adduct levels. These results support the conclusion that certain gene polymorphism
combinations affecting DNA repair or detoxification capacities may increase health risks
resulting from PAH exposure.
Porter et al. (2005) also evaluated the influence of XPA gene variants involved in NER
on BPDE-induced cytotoxicity. SV40-transformed human skin fibroblasts from an XP patient
with a nonsense XPA mutation were stably transfected with the WT XPA gene or either of two
rare XPA variants; the transfected genes could be overinduced with ponasterone A. The WT
XPA and both variants had greatly improved survival compared to XPA-free cells. Survival was
even more improved by ponasterone A induction in the variant, but not the WT cells. These
findings indicate that the polymorphic XPAs show greater ability to repair BPDE-induced DNA
damage, and thus may offer some protection from benzo[a]pyrene-induced genotoxicity, while
the nonsense mutation is likely to increase genotoxic risk.
200
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)
5.1.1. Choice of Principal Study and Critical Effect—with Rationale and Justification
There are limited data establishing associations between increased risk for noncancer
health effects in humans and exposure to benzo[a]pyrene. Several epidemiology studies have
reported associations between adverse birth outcomes including reduced birth weight, postnatal
body weight, and head circumference and internal biomarkers of exposure to benzo[a]pyrene
(BPDE-DNA adducts) via exposure to complex PAH mixtures (Tang et al., 2008, 2006; Perera et
al., 2005a, b). However, extrapolations from these studies are complicated by the concomitant
exposure to multiple PAHs and other components in the mixture. Thus, studies in humans were
not selected to serve as the basis of the RfD.
The subchronic and chronic oral exposure animal database includes a 2-year gavage
cancer bioassay with male and female Wistar rats (Kroese et al. 2001), a 2-year dietary cancer
bioassay with female B6C3Fi mice (Beland and Culp, 1998; Culp et al., 1998), a 90-day gavage
study with male and female Wistar rats (Kroese et al., 2001), a 90-day dietary study with male
and female F344 rats (Knuckles et al., 2001), and a 35-day study in male Wistar rats evaluating
immune endpoints (De Jong et al., 1999). Also available are five reproductive/developmental
toxicity studies in rodents examining reproductive endpoints in male Sprague-Dawley rats
(Zheng et al., 2010) and C57BL/6 mice (Mohamed et al., 2010), in offspring of treated CD-I and
NMRI female mice (Kristensen et al., 1995; MacKenzie and Angevine, 1981), and in female
Sprague-Dawley rats (Xu et al., 2010).
Kroese et al. (2001) exposed Wistar rats to benzo[a]pyrene in soybean oil by gavage at
doses of 0, 3, 10, or 30 mg/kg-day, 5 days/week, for 2 years. This study was primarily designed
as a cancer bioassay and did not evaluate other endpoints. An increase in the incidence of
animals with forestomach hyperplasia, compared with the control incidence, occurred at the low
and mid-dose but not the high-dose level; an dose-related, increased incidence of forestomach
tumors was observed at doses > 3 mg/kg-day. An increased incidence of hepatic clear cell foci
of cellular alteration was also observed at the 3 mg/kg-day, but not at the 10 or 30 mg/kg-day.
At the two highest exposure levels, elevated incidences of liver tumors were observed.
Female B6C3Fi mice were administered benzo[a]pyrene in the diet at average daily doses
of 0, 0.7, 3.3, and 16.5 mg/kg-day for 2 years (Beland and Culp, 1998; Culp et al., 1998). An
increase in the incidence of mice with forestomach hyperplasia, compared with the control
incidence, occurred at the lowest exposure level (23/47 at 0.7 mg/kg-day versus 13/48 in
controls). Similar to the rat bioassay (Kroese et al., 2001), forestomach hyperplasia was
observed with increasing incidence of animals with forestomach tumors (squamous cell
201
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
papillomas or carcinomas) with increasing dose. No other dose-related effects were reported in
this cancer bioassay.
A 90 day study also reported by Kroese et al. (2001) treated animals by gavage 5
days/week with 0, 3, 10, or 30 mg/kg benzo[a]pyrene in corn oil. The most sensitive effects
observed included increased liver weight and decreased thymus weight. Increases in liver weight
greater than 10% of controls were observed at 10 and 30 mg/kg-day in males only, and were
statistically significant. However, there were no statistically significant elevations in liver
enzymes screened in serum (ALT, AST, LDH, and GGT). The biological significance of
increased liver weight in males in the absence of elevated liver enzymes in the serum is unclear.
A decrease in thymus weight was observed in both sexes at 30 mg/kg-day, at 17 and 33% in
females and males, respectively, compared with controls. At 10 mg/kg-day, thymus weight in
males was decreased by 15% (not statistically significant). An increase in the incidence of
thymus atrophy was also observed in the 30 mg/kg-day males that showed a reduction of thymus
weight. Incidences for thymus atrophy (categorized in severity as slight) for the control through
high-dose groups were 0/10, 0/10, 0/10, and 3/10 for females and 0/10, 2/10, 1/10, and 6/10 for
males. The thymus is an organ involved in the maturation of immune cells especially early in
development. A change in thymus weight in the adult animal may be accompanied by alterations
of the immune system in functional assays, but the significance of thymus weight changes alone
is unknown.
Knuckles et al. (2001) exposed male and female F344 rats (6-8 per group) to
benzo[a]pyrene at doses of 0, 5, 50, or 100 mg/kg-day in the diet for 90 days. Statistically
significant decreases in RBC counts and hematocrit level (decreases as much as 10 and 12%,
respectively) were observed in males at doses >50 mg/kg-day and in females at 100 mg/kg-day.
The effect observed in this study at the lowest dose was an increase in abnormal tubular casts in
the kidney in males in which increases were observed at 5 mg/kg-day (40%), 50 mg/kg-day
(80%>) and 100 mg/kg-day (100%), compared to 10%> in the controls. In females, only 10%
showed significant kidney tubular changes at the two high dose levels compared to zero
incidence in controls. The incidences for kidney lesions were not provided; instead the data are
reported graphically as rounded percent incidences. Several reporting gaps in Knuckles et al
(2001) make interpretation of the results difficult. Specifically, the authors do not provide
statistical analysis of the renal endpoint nor do they provide the incidence data which would
allow for independent statistical analysis. The study author was contacted, but additional
clarification of the study data was not provided. Therefore, due to reporting gaps and resulting
reduced confidence, this study was not considered further in selecting the principal study.
De Jong et al. (1999) treated male Wistar rats (eight/dose group) with benzo[a]pyrene by
gavage 5 days/week for 35 days at doses of 0, 3, 10, 30, and 90 mg/kg-day. Hematological and
immunological changes were reported. Small, but statistically significant, dose-related decreases
in RBC count (5%) and associated measures (hemoglobin, and hematocrit) were observed at
202
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
>10 mg/kg-day. In addition, a dose-related and statistically significant decrease in the relative
number of B cells (13%) in the spleen was observed at >10 mg/kg-day compared to controls.
Dose-related decreases in thymus weight were statistically significant at >10 mg/kg-day.
Decreases in heart weight at 3 mg/kg-day and in kidney weight at 3 and 30 mg/kg-day were also
observed, but these changes did not show dose-dependent responses. At doses above 10 mg/kg-
day, significant decreases were observed in absolute number of cells harvested in the spleen, in
the number of B cells in the spleen, and in NK cell activity in the spleen, as well as a decrease in
serum IgM and IgA in rats.
Zheng et al., (2010) treated male Sprague-Dawley rats (8/group) to 0, 1, or 5 mg/kg-day
benzo[a]pyrene by daily corn oil gavage for a duration of 30 or 90 days. Testicular testosterone
was statistically significantly decreased in the high dose group (approximately 15%) following
90 days of exposure. The low dose group also appeared to have a similar average depression of
testosterone levels; however, the change did not reach statistical significance.
Mohamed et al. (2010) investigated multi-generational effects in male mice following
exposure of six-week old C57BL/6 mice (10/group) to 0 (corn oil), 1, or 10 mg/kg-day
benzo[a]pyrene for 6 weeks by daily corn oil gavage. Following final treatment, male mice were
mated with two untreated female mice to produce an F1 generation; F1 and F2 males were also
mated with untreated female mice. The mice of the Fl, F2, and F3 generations were not exposed
to benzo[a]pyrene. Statistically significant reductions of approximately 50% were observed in
epididymal sperm counts of F0 and Fl generations treated with the low dose of benzo[a]pyrene.
For F0 and Fl generations of the high dose group, epididymal sperm counts were reduced
approximately 70%. Means and variances were not reported but were presented graphically.
This study indicates that exposure to benzo[a]pyrene may have transgenerational effects on
sperm count. However, due to incomplete reporting, this study was not considered further for
selection as the principal study but was considered to be supportive of low dose male
reproductive effects following benzo[a]pyrene exposure.
MacKenzie and Angevine (1981) exposed groups of 30-60 female CD-I mice to 0, 10,
40, or 160 mg/kg-day benzo[a]pyrene on GDs 7-16. Crossover mating studies were then
conducted in which Fl offspring were mated continuously with untreated mice to determine
effects on fertility. Benzo[a]pyrene did not appear to be overtly toxic to mothers or offspring.
However, statistically significant decreased pup weight was observed at all dose levels at day 42.
At the lowest dose tested, pup weight was decreased 6% compared to control. At maturity, the
fertility of these animals was tested. The Fl male and Fl female fertility indices (i.e., percent of
mated animals that were pregnant) were significantly decreased in each exposure group as
follows (control through high-dose groups): Fl males: 80.4, 52.0, 4.7, and 0.0; Fl females 100,
65.7, 0.0, and 0.0. After six months on the breeding study, 34% of the gestationally treated
females in the 10 mg/kg-day dose group failed to produce any litters, and the the Fl females in
this dose group that did litter produced statistically significantly smaller litter sizes (19%
203
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
reduction in mean litter size). These reductions in fertility indices were associated with
decreased testes and ovary weight in the F1 animals. Testes weight was decreased 40 and 88%
at 10 and 40 mg/kg-day, respectively, and was associated with histologic evidence of injury to
the seminiferous tubules. In female F1 animals, severe reductions in ovarian tissues were
observed at all dose levels such that ovary weight measurements were difficult to obtain.
Examination of available tissue in these females revealed hypoplastic ovaries with few follicles
and corpora lutea (10 mg/kg-day) or with no evidence of folliculogenesis at the higher dose
(40 mg/kg-day).
Similar results were reported in a study in which groups of nine NMRIF0 female mice
were exposed by gavage to 0 or 10 mg/kg-day benzo[a]pyrene on GDs 7-16 (Kristensen et al.,
1995). F1 females were continuously bred with an untreated male for 6 months. F1 females had
decreased mean ovary weight (30% decreased, compared with controls) and reduced fertility as
reflected by decreased mean number of F2 litters (three compared with eight for control F1
females). F1 females had statistically significantly lower median numbers of offspring, number
of litters, and litter sizes and a statistically significantly greater median number of days between
litters as compared with the controls. At necropsy, the F1 females had statistically significantly
reduced ovary weight with histologic examination revealing decreased numbers of small,
medium, or large follicles and corpora lutea.
Xu et al., (2010) treated female Sprague-Dawley rats (6/group) to 0, 5, or 10 mg/kg-day
benzo[a]pyrene by corn oil gavage every other day for a duration of 60 days. This resulted in
time weighted average doses of 0, 2.5, and 5 mg/kg-day over the study period of 60 days.
Absolute ovary weight was statistically significantly reduced in the both the low and high
benzo[a]pyrene dose groups (11 and 15%, respectively; see Table 5-1). Animals in the high dose
group also had statistically significantly depressed levels of estradiol (by approximately 25%)
and decreased numbers of primordial follicles (by approximately 20%) compared to controls.
Statistically significantly altered estrus cyclicity was also evident in the high dose of
benzo[a]pyrene.
Table 5-1. Means ± SD for ovary weight in female SD-rats
Dose (mg/kg-d)a
0
2.5
5
204
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Ovary weight (g)
0.160 ±0.0146
0.143 ± 0.0098b
0.136 ±0.0098b
Body weight (g)
261.67 ± 12.0
249.17± 11.2
247.25 ± 11.2
a TWA doses over the 60 day study period
b Statistically different from controls (p < 0.05) using one-way ANOVA
Source: Xu et al. (2010).
5.1.2. Methods of Analysis
A number of noncancer effects observed following chronic or subchronic administration
of benzo[a]pyrene were modeled with U.S. EPA's Benchmark Dose (BMD) Modeling Software
(BMDS) where data were amenable. These endpoints included increased liver weight, decreased
thymus weight, decreased percent of splenic B-cells, increased forestomach hyperplasia, and
decreased ovary weight (Xu et al., 2010; Kroese et al., 2001; De Jong et al., 1999). Zheng et al.
(2010) did not provide enough information (i.e., incidences or means and variances) to allow for
BMD modeling. Data from other studies could not be modeled due study design utilizing only
one dose (Kristensen et al., 1995) or a highly elevated magnitude of response at the lowest dose
(MacKenzie and Angevine 1981) in which extrapolation down to a suitable benchmark response
would be unsupported by the available models.
In accordance with U.S. EP A's Benchmark Dose Technical Guidance Document (U.S.
EPA, 2000b), the BMD and the 95% lower confidence limit on the BMD (BMDL) were
estimated using a benchmark response (BMR) of 1 standard deviation (SD) from the control
mean for continous data or a BMR of 10% extra risk for dichotomous data in the absence of
information regarding what level of change is considered biologically significant, and also to
facilitate a consistent basis of comparison across endpoints and assessments. A summary of
modeling results for each endpoint is listed below in Table 5-2. Further details including the
output and graph for the best fit model can be found in Appendix B. In general, model fit was
assessed by a chi-square goodness-of-fit test (i.e., models with p < 0.1 failed to meet the
goodness-of-fit criterion) and the Akaike Information Criterion (AIC) value (i.e., a measure of
the deviance of the model fit that allows for comparison across models for a particular endpoint).
Of the models exhibiting adequate fit, the model yielding the lowest AIC value was selected as
the best-fit model. (U.S. EPA, 2000b).
For the forestomach hyperplasia endpoint, all data sets provided adequate descriptions of
the dose-response relationship from chronic oral exposure to benzo[a]pyrene, but at the highest
dose level for the rats (Kroese et al., 2001), the incidence of forestomach hyperplasia was not
increased relative to controls. It is possible that the forestomach hyperplasia observed following
benzo[a]pyrene exposure may be a precursor to the development of forestomach tumors, but
specific data supporting this conclusion are unavailable. Regardless, the male and female data
205
DRAFT - DO NOT CITE OR QUOTE
-------
1 sets in rats (Kroese et al., 2001) were modeled without the data from the highest dose group due
2 to the nonmontonic increase in response to increasing dose (Kroese et al., 2001).
3 Points of departure (PODs) for endpoints that were not amenable to BMD modeling were
4 identified using a NOAEL/LOAEL approach (Zheng et al., 2010; Kristensen et al., 1995;
5 MacKenzie and Angevine, 1981). A LOAEL of 5 mg/kg-day was identified for Zheng et al.
6 (2010) for significantly descreased testicular testosterone. A LOAEL of 10 mg/kg-day was
7 identified as the POD for Mackenzie and Angevine for decreased postnatal body weight and
8 decreased fertility of male and female mice treated during gestation. A POD based on the
9 LOAEL of 10 mg/kg/day was established from Kristensen et al. (1995) based on decreased
10 fertility and decreased ovary weight in female mice treated during gestation.
206
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-2. Summary of BMDs and BMDLs for modeled noncancer effects following oral exposure
Endpoint/data
Exposure
duration
BMR
Fitted model
Goodness-of-
fit />-value
AIC
BMD
(mg/kg-d)
BMDL
(mg/kg-d)
Reference
Increased Liver Weight
in Male Wistar Mice
90 d
10%
Linear (1° polynomial),
Power
0.58
49.51
8.11
5.8
Kroese et
al., 2001
Polynomial (2°)
0.74
50.53
4.53
2.29
Hill
0.82
50.48
4.1
1.24
Decreased Thymus
Weight in Male Wistar
Mice
90 d
1 SD
Linear, Polynomial
(2°), Power
(nonconstant variance)
0.23
380.71
16.40
11.30
Kroese et
al., 2001
Hill (nonconstant
variance)
NA
Decreased Thymus
Weight in Female
Wistar rats
90 d
1 SD
Linear
0.81
349.12
10.52
7.64
Kroese et
al., 2001
Hill
NA
Polynomial (2°)
0.77
350.80
13.29
7.77
Power
NA
Decreased Thymus
Weight in Male Wistar
Mice
35 d
1 SD
Linear, Polynomial (2°)
0.52
381.41
14.41
11.58
De Jong et
al., 1999
Hill
0.42
382.91
11.15
6.19
Power
NA
Decreased Splenic B-
cells in Male Wistar rats
35 d
1 SD
Linear, Polynomial (2°),
Power (constant
variance)
0.21
145.28
15.58
12.43
De Jong et
al., 1999
Hill (constant variance)
0.18
146.18
10.24
5.31
Increased Forestomach
Hyperplasia3 in Male
Wistar Rats
2 yrs
10%
Log-logistic
0.13
112.27
5.31
2.39
Kroese et
al., 2001
Gamma, Multistage,
Weibull
0.12
112.37
5.63
2.67
Logistic
0.09
112.93
7.25
4.35
LogProbit
0.06
113.88
8.36
4.52
Probit
0.10
112.87
7.09
4.13
Increased Forestomach
Hyperplasia3 in Female
2 yrs
10%
Log-logistic
0.32
117.04
2.15
1.35
Kroese et
al., 2001
Logistic
0.06
120.02
4.23
3.28
207 DRAFT - DO NOT CITE OR QUOTE
-------
Wistar Rats
LogProbit
0.02
121.13
3.91
2.57
Gamma, Multistage,
Weibull
0.24
117.42
2.40
1.59
Probit
0.06
119.74
3.99
3.06
Increased Forestomach
Hyperplasia in Female
B6C3Fi mice
2 yrs
10%
Log-logistic
0.21
193.3
0.329
0.115
Beland and
Culp, 1998
Logistic
0.06
194.7
0.757
0.545
LogProbit
0.29
192.1
0.670
0.448
Gamma, Multistage,
Weibull
0.42
191.3
0.421
0.295
Probit
0.03
196.6
0.946
0.711
Decreased Ovary
Weight in Female
Sprague-Dawley Rats
60 d
1 SD
Linear, Polynomial (1°)
0.39
-138.67
2.3
1.5
Xu et al.,
2010
Power
NA
a The best fit of each model considered is summarized. For continuous models (linear, polynomial, power, Hill), if an
adequate required including a variance model, only the results including modeled variance are summarized and the use of
nonconstant variance is indicated; otherwise constant variance was assumed. Details in Appendix B.
bData for the high-dose group were excluded from the modeled dataset due to a decreased incidence judged to be due to
competing effects masking the response.
NA = not applicable, model failed
208
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Several candidate principal studies (Kroese et al., 2001; De Jong et al., 1999; Zheng et
al., 2010; Kristensen et al., 1995; MacKenzie and Angevine, 1981) reported liver, thymus,
immune, and reproductive effects at higher doses relative to the ovary and forestomach effects
(reported by Kroese et al., 2001; Beland and Culp, 1998) and were considered less sensitive
measures of benzo[a]pyrene effects. Forestomach hyperplasia was not selected as the critical
effect, even though it was observed at lower doses compared with other effects, based on the
consideration that the reproductive and fertility effects, observed in animals and supported by
human data, appear to better characterize noncancer low dose effects of BaP. Specifically, the
Xu et al., (2010) study was chosen as the principal study and decreased ovarian weight as the
critical effect for the derivation of the RfD. This study identified biologically and statistically
significant decreases in ovary weight, estrogen, and primordial follicles, and altered estrus
cycling in treated animals. These reductions in female reproductive parameters are supported by
a large database of animal studies indicating that benzo[a]pyrene is ovotoxic with effects
including decreased ovary weight, decreased primordial follicles, and reduced fertility (Mattison
et al., 1980; MacKenzie and Angevine 1981; Swartz and Mattison 1985; Miller et al., 1992;
Kristensen et al., 1995; Borman et al, 2000). Additionally, studies indicate that exposure to
complex mixtures of PAHs, such as through cigarette smoke, is associated with measures of
decreased fertility in humans (El Nemr et al., 1998; Neal et al., 2005). Specific associations have
also been made between infertility and increased levels of benzo[a]pyrene in follicular fluid in
women undergoing in vitro fertilization (Neal et al., 2008).
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs)
Of the endpoints discussed in section 5.1.1., decreased ovary weight in female rats
(BMDLisd of 1.5 mg/kg-day) reported by Xu et al. (2010) was selected to serve as the critical
effect for the RfD. A total UF of 3000 was applied to the POD of 1.5 mg/kg-day to account for
several areas of uncertainty.
An UFa of 10 was applied to account for toxicokinetic and toxicodynamic differences
associated with extrapolation from animals to humans. The available data do not provide
quantitative information on the difference in susceptibility to benzo[a]pyrene between rats and
humans.
An UFh of 10 was applied to account for variability in susceptibility among members of
the human population (i.e., interindividual variability). Insufficient information is available to
quantitatively estimate variability in human susceptibility to benzo(a)pyrene.
An UFs of 10 was applied for the extrapolation of subchronic-to-chronic exposure
duration. The 60-day study by Xu et al. (2010) falls well short of a lifetime duration.
Therefore, it is unknown whether effects would be more severe or would be observed at lower
doses with a longer exposure duration.
209
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
An UFl of 1 was applied for LOAEL-to-NOAEL extrapolation because the current
approach is to address this factor as one of the considerations in selecting a BMR for BMD
modeling. In this case, a BMR of a 1 SD change from the control mean in ovary weight was
selected under an assumption that it represents a minimal biologically significant response level.
An UFd of 3 was applied to account for deficiencies in the benzo[a]pyrene toxicity
database. Limited observational studies in humans have suggested associations between
biomarkers of internal dose of benzo[a]pyrene and adverse birth outcomes (including reduced
birth weight, postnatal body weight, head circumference, and neurodevelopment) and decreased
fertility (Edwards et al., 2010; Neal et al., 2008; Tang et al., 2008, 2006; Perera et al., 2009;
2005a, b). However, the likely contribution of multiple exposure routes in these studies make
extrapolation to exposure concentrations uncertain. Several animal studies exist for
benzo[a]pyrene to inform noncancer effects, including subchronic oral toxicity studies in rats and
mice, and two developmental studies and several reproductive studies in mice and rats. The lack
of a standard multigenerational study (specifically, one which includes exposure from pre-mating
to lactation) is a data gap, especially considering benzo[a]pyrene has been shown to affect
fertility in adult male and female animals by multiple routes of exposure (Mohamed et al., 2010;
MacKenzie and Angevine 1981; Kristensen et al., 1995; Archibong et al., 2008; Borman et al.,
2000; Swartz and Mattison 1985). In addition, the lack of a study examining functional
neurological endpoints following in utero exposure is also a data gap considering the available
epidemiological evidence showing the association of in utero PAH exposure and indicators of
decreased neurological development (Edwards et al., 2010, Perera et al., 2009, Tang et al., 2008).
Therefore, an UF of 3 was applied to the POD for the lack of a standard multigenerational
reproductive toxicity study and a neurodevelopmental study.
The RfD for benzo[a]pyrene was calculated as follows:
RfD = BMDLisd - UF
= 1.5 mg/kg-day ^ 3000
= 0.0005 mg/kg-day
5.1.4. Previous RfD Assessment
No RfD was derived in the previous IRIS assessment.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
5.2.1. Choice of Principal Study and Critical Effect—with Rationale and Justification
The only chronic inhalation study available for benzo[a]pyrene was designed as a cancer
bioassay and did not report noncancer endpoints (Thyssen et al. 1981). However, several repeat
210
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
dose reproductive and developmental toxicity studies are available in which effects on fetal
survival and the male reproductive system have been observed.
Reproductive system variables were adversely impacted in male F344 rats (10/group)
"3
exposed to benzo[a]pyrene aerosols at 75 |ig/m for 60 days (Archibong et al., 2008; Ramesh et
al., 2008). The testes from exposed rats weighed 34% less than those from unexposed controls.
Treatment with benzo[a]pyrene was also associated with reductions (compared with control
values) in total tubular volume and weight (-20%) and tubular length (-40%); total weight and
volume of interstitium per paired testis (12%); and percentage of progressively motile stored
spermatozoa, stored sperm density, and percentage of morphologically normal sperm (-70-
80%>). Daily sperm production and levels of circulating and intratesticular testosterone were also
decreased in treated rats (by about 50, 70, and 80% respectively), compared with controls.
Luteinizing hormone levels were 50-60% higher in treated rats than controls 72 hours after
exposure.
In a developmental study, timed-pregnant F344 rats (10/group) were exposed to
benzo[a]pyrene on GDs 11-21 as a carbon black aerosol at 100 |ig/m to assess neurological
endpoints in offspring during PNDs 60-70 (Wormley et al., 2004). Pups of the F1 generation
were weaned on PND 30 and tested for long-term potentiation electrophysiological responses in
the hippocampus during PNDs 60-70. Although the number of implantation sites in treated rats
was within 1% of unexposed controls, the percentage of pups born relative to recorded
implantation sites in each dam (the birth index) was reduced by 65% in treated rats compared
with unexposed controls. In addition, protein levels of NMDA receptor subunit 1 were down-
regulated (by 18%) on PND 10 and 67% on PND 30) in the hippocampus of benzo[a]pyrene-
exposed F1 pups, and the magnitude of the long-term potentiation response across the perforant
path-granular cells in the hippocampus of F1 rats was consistently weaker than the response
observed for the controls (about 25% weaker), suggesting that exposure to benzo[a]pyrene via
the inhalation route attenuates the capacity for long-term potentiation in the F1 generation.
However, no functional tests to assess neurotoxicity were conducted in this study.
In another developmental toxicity study, timed-pregnant F344 rats (10/group) were
"3
exposed to benzo[a]pyrene aerosols at concentrations of 25, 75, or 100 |ig/m on GDs 11-20 and
evaluated for post-implantation fetal survival and hormone levels associated with pregnancy
(Archibong et al., 2002). The total number of implantation sites in treated rats was within 5% of
the values obtained for sham-exposed and unexposed controls. However, dose-dependent trends
were observed for decreased numbers of pups per litter and percent fetal survival per litter with
increasing benzo[a]pyrene concentrations. The number of pups/litter was decreased by
"3
approximately 14, 50, and 65% at 25, 75, and 100 |ig/m , respectively, compared with sham-
exposed and unexposed controls. Percent survival/was similarly reduced by about 20, 60, and
65%), respectively, at the same exposure concentrations. In addition, biologically significant
decreases in pup weights (presented as g/litter) were observed at concentrations >75 |ig/m (14
211
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
"3
and 16% decreases at 75 and 100 (j,g/m , respectively). Levels of plasma progesterone, estradiol-
"3
17p, and prolactin on GD 17 were decreased by about 12, 14, and 35%, respectively, at 25 |ig/m
and 17, 60, and 70%, respectively, at 75 |ig/m compared with respective controls.
The study by Archibong et al. (2002) was selected as the principal study as it observed
biologically significant effects at the lowest dose tested by the inhalation route. This study
indicates that the developing fetus is a sensitive target following inhalation exposure to
"3
benzo[a]pyrene. A LOAEL of 25 |ig/m was identified based on exposure to benzo[a]pyrene on
GDs 11-20 that caused biologically significant reductions in fetal survival and body weight
decreases in the surviving pups (see Table 5-3). The observed decrease in pup weight and fetal
survival were selected as critical effects as they are the most sensitive noncancer effects observed
following inhalation exposure to benzo[a]pyrene. Though only a few studies exist which
evaluate benzo[a]pyrene by the inhalation route, additional support for this endpoint can be
found from oral studies of benzo[a]pyrene. A developmental/reproductive study conducted via
the oral route in mice observed decreased survival of litters, decreased pup weight, and decreased
reproductive organ weight following in utero exposure to benzo[a]pyrene on GD 7-16
(MacKenzie and Angevine, 1981).
5.2.2. Methods of Analysis- Adjustment to a Human Equivalent Concentration (HEC)
By definition, the RfC is intended to apply to continuous lifetime exposures for humans
(U.S. EPA, 1994). EPA recommends that adjusted continuous exposures be used for inhalation
developmental toxicity studies as well as for studies of longer durations (U.S. EPA, 2002). The
LOAEL of 25 |ig/m based on decreased pup weight and fetal survival reported in the
developmental study by Archibong et al. (2002) was selected to serve as the POD.
Table 5-3. Pregnancy outcomes in female F344 rats treated with
benzo[a]pyrene on GDs 11-21 by inhalation
Parameter3
Administered concentration of benzo[a]pyrene (|ig/m3)
0 (unexposed
control)
0
(carbon black)
25
75
100
Implantation sites
8.6 ±0.2
8.8 ±0.1
8.8 ±0.5
9.0 ±0.2
8.8 ± 0.1
Pups per litter
8.5 ±0.2
8.7 ±0.2
7.4 ± 0.5b
4.2 ± 0. lb
3.0 ± 0.2b
Survival (litter %)
98.9 ± 1.1
96.7 ± 1.7
78.3 ± 4.1b
38.0 ± 2.1b
33.8 ± 1.3b
Pup weight (g/litter)
10.6 ±0.1
8.8 ±0.1
10.5 ±0.2
9.1 ±0.2b
8.9 ± 0.1b
Crown-rump length
(mm/litter)
29.4 ±0.6
29.3 ±0.5
28.0 ±0.6
27.3 ±0.7
27.9 ±0.7
aValues presented as means ± SEM.
bSignificantly different from controls at p < 0.05 by one-tailed post-hoc t-testing following ANOVA.
Source: Archibong et al. (2002).
212
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Data for decreased pup survival were not amenable to BMD modeling due to the pattern
of variability (heterogeneous variances) in the data set. Therefore, the LOAEL from this study
was used as the POD. The LOAEL from this study is based on a 4 hour exposure of pregnant
"3
rats to 25 |ig/m benzo[a]pyrene on GDs 11-20. This concentration was adjusted to account for
the discontinuous daily exposure as follows:
PODadj = POD x hours exposed per day/24 hours
= LOAEL x (4 hr/24 hr)
= 25 |ig/m3 x 4/24
= 4.2 |ig/m3
The human equivalent concentration (HEC) was calculated from the PODadj by
multiplying by a dosimetric adjustment factor (DAF), which, in this case, was the regional
deposited dose ratio (RDDRer) for extrarespiratory (i.e. systemic) effects as described in
Methods for Derivation of Inhalation Reference Concentrations and Application of Inhalation
Dosimetry (U.S. EPA, 1994b). The observed developmental effects are considered systemic in
nature (i.e., extrarespiratory) and the current normalizing factor for extrarespiratory effects of
particles is body weight. In the case of benzo[a]pyrene, the RDDRer was calculated as follows:
BW (V ) (F )
RDDRfr = hx,LjJaxI toWa
BW (V ) (F )
-LJ¥Ya vte/h tot /h
where:
BW = body weight (kg)
Ve = ventilation rate (L/min)
Ftot = total fractional deposition
The total fractional deposition (Ftot) includes particle deposition in the nasal-pharyngeal
region, the tracheobronchial region, and the pulmonary region. Ftot for both animals and
humans was calculated using the Multi-Path Particle Dosimetry model, a computational model
that can be used for estimating human and rat airway particle deposition and clearance (MPPD;
Version 2.0 © 2006, publicly available through the Hamner Institute). The Ftot was calculated
based on the average particle size of 1.7 ± 0.085 (mass median aerodynamic diameter ±
geometric standard deviation) as reported in the description of particle generation methods in
Ramesh et al. (2000). For the model runs, the Yeh-Schum 5-lobe model was used for the human
and the asymmetric multiple path model was used for the rat (see Appendix C for MPPD model
output). Both models were run under nasal breathing scenarios with the inhalability adjustment
selected. A geometric standard deviation (GSD) of 1 was used as the default by the model
because the reported GSD of 0.085 < 1.05.
213
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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 human parameters used in the model for calculating FTOt and in the subsequent
calculation of the PODhec were as follows: human - BW, 70 kg; Ve, 13.8 L/min; breathing
frequency, 16 per minute; tidal volume, 860 mL; FRC (functional residual capacity), 3300 mL;
and URT (upper respiratory tract) volume, 50 mL. Although the most sensitive population in the
principal study is the developing fetus, the adult rat dams were exposed. Thus, adult human
parameters were used in the calculation of the HEC to extrapolate from a pregnant rat to a
pregnant human. The parameters used for the rat were BW, 0.25 kg (based on the approximate
weight of a 100 day-old, female timed-pregnant Sprague-Dawley rat); Ve, 0.18 L/min; breathing
frequency, 102 per minute; tidal volume, 1.8 mL; FRC (functional residual capacity), 4 mL; and
URT (upper respiratory tract) volume, 4.42 mL. All other parameters were set to the default
value (see Appendix C).
Under these conditions, the MPPD model calculated FTot values of 0.621 for the human
and 0.181 for the rat. Using the above equation, the RDDRERwas calculated to be 1.06.
From this, the PODhec was calculated as follows:
PODhec = PODadj x RDDRer
PODhec = 4.2 |ig/m3 x 1.1
PODhec = 4.6 |ig/m3
5.2.3. RfC Derivation- Including Application of Uncertainty Factors (UFs)
The critical effect for the derivation of the RfC was identified as decreased fetal survival
and decreased pup weight associated with inhalation exposure to pregnant rats on GDs 11-20.
The LOAEL for decreased fetal survival was adjusted to a continuous human equivalent
concentration and used as the POD for the derivation of the RfC. A total UF of 1000 was
applied to the PODhec to account for four main areas of uncertainty:
A UFa of 3 was applied to account for uncertainties in extrapolating from rats to humans.
Application of a UF of 10 encompasses two areas of uncertainty: toxicokinetic and
toxicodynamic uncertainties. In this assessment, the toxicokinetic component is mostly
addressed by the determination of a HEC as described in the RfC methodology (U.S. EPA,
1994b). Therefore, a UF of 3 was applied to account for the remaining toxicodynamic
uncertainties in the extrapolation from rats and humans.
A UFh of 10 was applied to account for variability in susceptibility among members of
the human population (i.e., interindividual variability). Insufficient information is available to
quantitatively estimate variability in human susceptibility to benzo(a)pyrene.
A UFl of 10 was applied to account for the use of a LOAEL. A NOAEL was not
identified for decreased fetal survival observed by Archibong et al (2002). At the lowest dose,
benzo[a]pyrene treated dams gave birth to 15% fewer pups compared to dams treated with
214
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
vehicle alone (carbon black particles). Due to the lack of a NOAEL and the inability to model
the data set for decreased fetal survival, a UF of 10 was applied to extrapolate to a NOAEL.
A UFs of 1 was applied to account for extrapolation from subchronic to chronic exposure
because developmental toxicity resulting from a narrow period of exposure was used as the
critical effect. The developmental period is recognized as a susceptible life stage when exposure
during a time window of development is more relevant to the induction of developmental effects
than lifetime exposure (U.S. EPA, 1991a).
A UFd of 3 was applied to account for deficiencies in the benzo[a]pyrene toxicity
database. One developmental study exists for benzo[a]pyrene by the inhalation route. The
developmental study by Archibong et al., 2002, which was used as the basis for the RfC,
observed decreased fetal survival and decreased litter size following gestational treatment on
GDs 11-20. Limited observational studies in humans have suggested associations between
biomarkers of internal doses of benzo[a]pyrene and adverse birth outcomes (including reduced
birth weight and postnatal weight, decreased head circumference, and impaired
neurodevelopment) and decreased fertility (Neal et al., 2008; Tang et al., 2008, 2006; Perera et
al., 2005a, b). A multigenerational reproductive study examining these types of effects in
animals does not exist for the inhalation route. However, oral multigenerational studies indicate
that effects on fertility would be expected in male and female offspring exposed to
benzo[a]pyrene during development (Mackenzie and Angevine 1981; Kristensen et al., 1995). In
addition, the lack of a study examining functional neurological endpoints following in utero
exposure is also a data gap considering the available epidemiological evidence showing the
association of in utero PAH exposure and indicators of decreased neurological development
(Edwards et al., 2010, Perera et al., 2009, Tang et al., 2008). Therefore, a UF of 3 was applied
to the POD for the lack of a multigenerational reproductive toxicity study and a
neurodevelopmental study.
The RfC for benzo[a]pyrene was calculated as follows:
RfC = LOAELadj[hec] ^"UF
= 4.6 |ig/m3-day ^ 1000
= 4.6 x 10"3 |ig/m3-day or 5 x 10"6 mg/m3
5.2.4. Previous RfC Assessment
An RfC was not derived in the previous IRIS assessment.
215
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE AND INHALATION
REFERENCE CONCENTRATION
The following discussion identifies uncertainties associated with the RfD and RfC for
benzo[a]pyrene. To derive the RfD, the UF approach (U.S. EPA, 2000, 1994b) was applied to a
POD based on decreased ovary weight in female rats exposed to benzo[a]pyrene. To derive the
RfC, this same approach was applied to a POD from a developmental study for the effect of
decreased fetal survival. Uncertainty factors were applied to the POD to account for
extrapolating from an animal bioassay to human exposure, the likely existence of a diverse
population of varying susceptibilities, and for database deficiencies. These extrapolations are
carried out with default approaches given the lack of data to inform individual steps.
The database for benzo[a]pyrene contains limited human data. The observation of effects
associated with benzo[a]pyrene exposure in humans is complicated by several factors including
the existence of benzo[a]pyrene in the environment as one component of complex mixtures of
PAHs, exposure to benzo[a]pyrene by multiple routes of exposure, and the difficulty in obtaining
accurate exposure information. Data on the effects of benzo[a]pyrene alone are derived from a
large database of studies in animal models. The database for oral benzo[a]pyrene exposure
includes two chronic bioassays in rats and mice, two developmental studies in mice, and several
subchronic studies in rats.
Although the database is adequate for RfD derivation, there is uncertainty associated with
the database, because a NOAEL was not identified in the reproductive and developmental oral
toxicity studies, comprehensive two-generation reproductive/developmental toxicity studies are
not available, and immune system endpoints affected in the sub chronic-duration studies were not
evaluated in the chronic-duration toxicity studies. Additionally, the only available chronic
studies of oral exposure to benzo[a]pyrene focused primarily on neoplastic effects. These studies
identify forestomach hyperplasia as one of the more sensitive histological effects following
repeated oral exposure to benzo[a]pyrene. However, data from chronic cancer bioassays for
benzo[a]pyrene show no increase in this endpoint at the high dose in rats. An increased
incidence of forestomach tumors is observed at similar doses; suggesting that this effect may be
pre-neoplastic in nature.
The only chronic inhalation study of benzo[a]pyrene was designed as a lifetime
carcinogenicity study and did not examine noncancer endpoints (Thyssen et al., 1981). However
subchronic and short term inhalation studies are available which examine developmental and
reproductive endpoints in rats. Developmental studies by the inhalation route identified
biologically significant reductions in the number of pups/litter and percent fetal survival and
possible neurodevelopmental effects (e.g., diminished electrophysiological responses to stumuli
in the hippocampus) following gestational exposures. Additionally, a 60 day oral study in male
rats reported male reproductive effects (e.g., decreased testes weight and sperm production and
motility), but provides limited information to characterize dose-response relationships with
216
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
chronic exposure scenarios. One area of uncertainty pertains to the lack of information regarding
fertility in animals exposed gestationally to benzo[a]pyrene, especially in light of developmental
studies by the oral route indicating reduced fertility in the F1 generation and decreased
reproductive organ weights. The database also lacks a multigenerational reproductive study via
the inhalation route. Areas of uncertainty include the lack of chronic inhalation studies focusing
on noncancer effects, limited data on dose-response relationships for impaired male or female
fertility with gestational exposure or across several generations, and limited data on immune
system endpoints with chronic exposure to benzo[a]pyrene.
The toxicokinetic and toxicodynamic differences for benzo[a]pyrene between the animal
species in which the POD was derived and humans are unknown. PBPK models can be useful
for the evaluation of interspecies toxicokinetics; however, the benzo[a]pyrene database lacks an
adequate model that would inform potential differences. There is some evidence from the oral
toxicity data that mice may be more susceptible than rats to some benzo[a]pyrene effects (such
as ovotocity [Borman et al., 2000]), though the underlying mechanistic basis of this apparent
difference in not understood. Most importantly, it is unknown which animal species may be
more comparable to humans.
5.4. CANCER ASSESSMENT
As discussed in Section 4.7, benzo[a]pyrene is "carcinogenic to humans" based on
evidence of carcinogenicity in humans exposed to different PAH mixtures containing
benzo[a]pyrene, extensive and consistent evidence of carcinogenicity in laboratory animals
exposed to benzo[a]pyrene via several routes of administration, and extensive and consistent
evidence that the mode of action of carcinogenesis in laboratory animals also occurs in humans
exposed to PAH mixtures containing benzo[a]pyrene.
5.4.1. Oral Exposure—Oral Slope Factor
5.4.1.1. Choice of Study/Data—with Rationale and Justification—Oral Exposure
Numerous cancer bioassays exist which identify tumors, primarily of the alimentary tract,
following oral exposure in rodents (see Table 4-8 for references). These studies provide support
for the carcinogenic hazard for benzo[a]pyrene, however, are not suitable for dose-response
analysis due to limitations in study design, methods, and/or reporting. Specifically, several of
these studies 1) lack a vehicle control group 2) use only one benzo[a]pyrene dose group or 3) use
a single one-time exposure to benzo[a]pyrene (Benjamin et al., 1988; Robinson et al., 1987; El
Bayoumy, 1985; Wattenberg, 1974; Roe et al., 1970; Biancifiori et al. 1967; Chouroulinkov et
al., 1967; Field and Roe, 1970; Berenblum and Haran 1955). Of the controlled, multiple dose-
group, repeat-dosing studies that remain, most treated animals for less than a year, which is less
217
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
optimal for extrapolating to a lifetime exposure (Weyand et al., 1995; Triolo et al., 1977;
Fedorenko et al., 1967; Neal and Rigdon 1967).
Three 2-year oral bioassays remain which associate lifeftime benzo[a]pyrene exposure
with forestomach, liver, oral cavity, jejunum, kidney, auditory canal (Zymbal's gland) tumors,
and skin or mammary gland tumors in male and female Wistar rats (Kroese et al., 2001);
forestomach tumors in male and female Sprague-Dawley rats (Brune et al., 1981); and
forestomach, esophagus, tongue, and larynx tumors in female B6C3Fi mice (male mice were not
tested; Beland and Culp, 1998; Culp et al., 1998). Brune et al. (1981) dosed rats (32/sex/group)
with several concentration of benzo[a]pyrene dissolved a 1.5% caffeine solution, sometimes as
infrequently as once every 9th day, for up to two years and observed increased forestomach
tumors. This study was not selected for quantitation due to the non-standard treatment protocol.
The rat bioassay by Kroese et al. (2001) and the mouse bioassay by Beland and Culp (1998)
were conducted in accordance to Good Laboratory Practice (GLP) principles as established by
OECD. These studies included histological examinations for tumors in many different tissues,
contained three exposure levels and controls, contained adequate numbers of animals per dose
group (~50/sex/group), treated animals for two years or until death, and included detailed
reporting of methods and results (including individual animal data).
Therefore, the Kroese et al. (2001) and Beland and Culp s(1998) tudies were selected as
the best available studies for dose-response analysis and extrapolation to lifetime cancer risk
following oral exposure to benzo[a]pyrene.
5.4.1.2. Dose-response Data—Oral Exposure
Details of the rat (Kroese et al., 2001) and female mouse (Beland and Culp, 1998) study
designs are provided in Section 4.2.1.2. Dose-related, statistically significant increasing trends
in tumors were noted at the following sites:
Squamous cell carcinomas or papillomas of the forestomach or oral cavity in male and
female rats;
Squamous cell carcinomas or papillomas of the forestomach, tongue, larynx, or
esophagus in female mice;
Auditory canal carcinomas in male and female rats;
Kidney urothelial carcinomas in male rats;
Jejunum adenocarcinomas in female and male rats;
Hepatocellular adenomas or carcinomas in male and female rats;
Squamous cell carcinomas or basal cell tumors of the skin or mammary gland in male
rats.
These tumors were generally observed earlier during the study with increasing exposure
levels, and showed statistically significantly increasing trends in incidence with increasing
218
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
exposure level (Cochran-Armitage trend test, p<0.001). These data are summarized in Tables 5-
4 (male and female rats) and 5-5 (female mice). As recommended by the NTP (McConnell et al.,
1986), etiologically similar tumor types, i.e., benign and malignant tumors of the same cell type,
were combined for these tabulations when it was judged that the benign tumors could progress to
the malignant form, as outlined in the Cancer Guidelines (U.S. EPA, 2005a). In addition, when
one tumor type occurred across several functionally related tissues, as with squamous cell tumors
in the tongue, esophagus, larynx and forestomach, or adenocarcinomas of the jejunum or
duodenum, these incidences were also aggregated as counts of tumor-bearing animals.
In the rat study (Kroese et al., 2001), the oral cavity and auditory canal were examined
histologically only if a lesion or tumor was observed grossly at necropsy. Consequently, dose-
response analysis for these sites was not straightforward. Use of the number of tissues examined
histologically as the number at risk would tend to overestimate the incidence, because the
unexamined animals were much less likely to have a tumor. On the other hand, use of all
animals in a group as the number at risk would tend to underestimate if any of the unexamined
animals had tumors which could only be detected microscopically. The oral cavity squamous
cell tumors were combined with those in the forestomach because both are part of the alimentary
tract, recognizing that there was some potential for underestimating this cancer risk.
The auditory canal tumors from the rat study were not considered for dose-response
separately or combined with another site. First, very few tissues were examined in the control
and lower dose groups (see Table 4-4). Also, the tumors were not clearly related to any other
site or incidence type, as they were described as a mixture of squamous and sebaceous cells
derived from pilosebaceous units. The tumors found were observed mainly in the high dose
groups and were highly coincident with the oral cavity and forestomach tumors. That is, only
one mid-dose male with an auditory canal tumor did not also have a forestomach or oral cavity
squamous cell tumor. No low-dose male or female rats were found with auditory canal tumors.
While the investigators did not suggest that these tumors were metastases from other sites (in
which the auditory canal tumors could be reflections of other tumor types), it is difficult to
conclude that they are independent on a purely statistical basis without sufficient low-dose data.
Therefore dose-response analysis was not pursued for this site.
219
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-4. Incidence data for tumors in Wistar rats exposed to
benzo[a]pyrene by gavage, 5 days/week for 104 weeks
Tumor site/type"
Administered dose (mg/kg-d)
0
3
10
30
HED (mg/kg-d)b
Male rats
0
0.54
1.81
5.17
Forestomach or oral cavity: squamous cell papilloma
or carcinoma
0/51
8/51
45/49
52/52
Hepatocellular adenoma or carcinoma
0/51
4/50
38/49
49/50
Jejunum/duodenum: adenocarcinoma
0/50
0/48
1/48
9/39
Kidney: urothelial carcinoma
0/52
0/52
0/52
3/52
Skin or mammary gland:
Basal cell adenoma or carcinoma
Squamous cell carcinoma
2/52
0/51
1/50
1/50
1/49
1/49
13/40
6/40
Female rats
0
0.49
1.62
4.85
Forestomach or oral cavity: squamous cell papilloma
or carcinoma
1/52
6/50
30/47
50/51
Hepatocellular adenoma or carcinoma
0/52
1/50
39/47
51/51
Jejunum/duodenum: adenocarcinoma
0/50
0/46
0/45
4/42
aFor each tissue site, the numerator of the tumor incidence value is the number of animals bearing the specified
tumors. The denominators are the number of animals examined histologically, minus the number of animals who
died before the earlier of the first occurrence of the tumor type in each group or Week 52.
bHEDs for continuous exposure were calculated using the animal to human scaling factor for each dose group x the
administered dose x 5 d/7 d. Scaling factors used the form (TWA body weight/70)0 25, with the U.S. EPA (1988)
reference body weight for humans (70 kg), and the TWA body weight for each dose group. See Table D-4 for
more information.
Source: Kroese et al. (2001).
Table 5-5. Incidence data for tumors in female B6C3Fi mice exposed to
benzo[a]pyrene in the diet for 104 weeks
Tumor site/type"
Administered dose (mg/kg-d)b
0
0.7
3.3
16.5
HED (mg/kg-d)c
0
0.10
0.48
2.32
Forestomach, esophagus, tongue, larynx: squamous
cell papilloma or carcinoma
1/48
3/48
38/46
46/47
a The numerator of the tumor incidence value is the number of animals bearing any of the listed tumors (see Table
4-6 in Section 4.2.1.2). The denominators are the number of tissues examined histologically, minus the number of
animals who died before the earlier of the first occurrence of the tumor type in each group or Week 52.
bAdministered doses were calculated using TWA body weight for mice and reported food intakes.
°HEDs were calculated using the animal to human scaling factor for each dose group x the administered dose.
Scaling factors were calculated using U.S. EPA (1988) reference body weights for humans (70 kg), and the TWA
body weight for each dose group: (TWA body weight/70)025 x dose = HED. See Table D-5 for more information.
Source: Beland and Culp (1998).
5.4.1.3. Dose Adjustments and Extrapolation Method(s)—Oral Exposure
220
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
The EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend
that the method used to characterize and quantify cancer risk from a chemical is determined by
what is known about the mode of action 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 mode of action because of DNA reactivity, or if another mode of action
that is anticipated to be linear is applicable. In this assessment, EPA concluded that
benzo[a]pyrene causes cancer via a mutagenic MOA (as discussed in Section 4.7.3). Thus, a
linear approach to low-dose extrapolation was used.
The high-dose groups of both the rat and mouse studies were dead or moribund by week
79 for female mice, week 72 for female rats, and week 76 for male rats. Due to the occurrence of
multiple tumor types, earlier occurrence with increasing exposure, and early termination of the
high-dose group in each study, methods that can reflect the influence of competing risks and
intercurrent mortality on site-specific tumor incidence rates are preferred. EPA has generally
used a model which incorporates the time at which death-with-tumor occurred as well as the
dose; the multistage-Weibull model is multistage in dose and Weibull in time, and has the form:
P(d, t) = 1 - exp[-(q0 + qid + q2d2 + ... + qkJc) x (t ± t0f7,
where P(d, t) represents the lifetime risk (probability) of cancer at dose d (i.e., human equivalent
exposure in this case) and age t (in bioassay weeks); parameters qt > 0, for i = 0, 1, ..., k; t is the
time at which the tumor was observed; and z is a parameter 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 either when all tumors are
considered incidental or because of a lack of data to estimate the time reliably. The dose-
response analyses were conducted using the computer software program MultiStage-Weibull
(U.S. EPA, 2010), which is based on Weibull models drawn from Krewski et al. (1983).
Parameters were estimated using the method of maximum likelihood.
Two general characteristics of the observed tumor types were considered prior to
modeling; allowance for different, although unidentified modes of action, and allowance for
relative severity of tumor types. First, etiologically different tumor types were not combined
across sites prior to modeling (that is, overall counts of tumor-bearing animals were not
tabulated), in order to allow for the possibility that different tumor types could have different
dose-response relationships due to different underlying mechanisms or factors, such as latency.
Consequently, all of the tumor types listed separately in Tables 5-4 and 5-5 were also modeled
separately.
Additionally, the multistage-Weibull model can address relative severity of tumor types
by distinguishing between tumors as being either fatal or incidental to the death of an animal, in
order to adjust partially for competing risks. Incidental tumors are those tumors thought not to
221
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
have caused the death of an animal, while fatal tumors are thought to have resulted in animal
death. Cause of death information for most early animals deaths was provided by the
investigators of both of the bioassays. In the rat study, tumors of the forestomach or liver were
the principal cause of death for most animals dying or sacrificed (due to moribundity) before the
end of the study, while tumors of the forestomach were the most common cause of early deaths
in the mouse study.
Adjustments for approximating human equivalent slope factors applicable for continuous
exposure were applied prior to dose-response modeling. First, continuous daily exposure for the
gavage study in rats (Kroese et al, 2001) was estimated by multiplying each administered dose
by (5 days)/(7 days) = 0.71, under the assumption of equal cumulative exposure yielding
equivalent outcomes. Dosing was continuous in the mouse diet study (Beland and Culp, 1998),
so no continuous adjustment was necessary. Next, consistent with the Guidelines for
Carcinogen Risk Assessment (U.S. EPA, 2005a), an adjustment for cross-species scaling was
applied to address toxicological equivalence across species. Following EPA's cross-species
scaling methodology, the time-weighted daily average doses were converted to human equivalent
doses on the basis of (body weight)3'4 (U.S. EPA, 1992). This was accomplished by multiplying
0 25
administered doses by (animal body weight(kg)/70 kg) ' (U.S.EPA, 1992), where the animal
body weights were time-weighted averages from each group (see Tables D-4, D-5), and the U.S.
EPA (1988) reference body weight for humans is 70 kg. It was not necessary to adjust the
administered doses for lifetime equivalent exposure prior to modeling for the groups terminated
early, because the multistage-Weibull model characterizes the tumor incidence as a function of
time, from which it provides an extrapolation to lifetime exposure.
The multistage-Weibull model was applied to the datasets. Specific n-stage Weibull
models were selected for each tumor dataset based on the values of the log-likelihoods according
to the strategy used by EPA (U.S. EPA, 2002). 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. This method generally led to the same
conclusion as selecting the model fit with the lowest AIC. If a model with one more stage fitted
the low-dose data better than the most parsimonious model, then the model with one higher stage
was selected.
PODs for estimating low-dose risk were identified at doses at the lower end of the
observed data, generally corresponding to 10% extra risk, where extra risk is defined as [P(d) -
P(0)]/[1 - P(0)]. The lifetime oral cancer slope factor for humans is defined as the slope of the
line from the lower 95% bound on the exposure at the POD to the control response (slope factor
= 0.1/BMDLio). This slope, a 95% upper confidence limit (UCL) represents a plausible upper
bound on the true risk.
222
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
5.4.1.4. Oral Slope Factor Derivation
The PODs estimated for each tumor site are summarized in Table 5-5. Details of the
model selection process are provided in Table D-6. Using linear extrapolation from the
BMDLio, human equivalent oral slope factors were derived for each gender/tumor site
combination and are listed in Table 5-6.
Table 5-6. Human equivalent PODs and oral slope factors derived from
multistage-Weibull modeling of tumor incidence data at multiple tissue sites
in Wistar rats and B6C3Fi mice exposed to benzojalpyrene orally
'or 2 years
Species
Sex
Tumor
BMD10
(mg/kg-d)
BMDL10
(mg/kg-d)
Slope Factor"
(mg/kg-d)1
Rats
Male
Forestomach, oral cavity: squamous cell
tumors
0.453
0.281
0.4
Hepatocellular adenomas or carcinomas
0.651
0.449
0.2
Jejunum/duodenum adenocarcinomas
3.03
2.38
0.04
0.5b
Kidney: urothelial carcinomas
4.65
2.50
0.04
Skin, mammary:
Basal cell tumors
Squamous cell tumors
2.86
2.64
2.35
1.77
0.04
0.06
Female
Forestomach, oral cavity: squamous cell
tumors
0.539
0.328
0.3
0.3b
Hepatocellular adenomas or carcinomas
0.575
0.507
0.2
Jejunum/duodenum adenocarcinomas
3.43
1.95
0.05
Mice
Female
Forestomach, esophagus, tongue, larynx:
squamous cell tumors
0.127
0.071
1
aHuman equivalent slope factor = 0.1/BMDLi0hed; see Appendix D for details of modeling results.
b Estimates of risk of incurring at least one of the tumor types listed.
Oral slope factors derived from rat bioassay data varied by gender and tumor site
(Table 5-6). Values ranged from 0.04 per mg/kg-day, based on kidney tumors in males, to
0.4 per mg/kg-day, based on alimentary tract tumors in males. Slope factors based on liver
tumors in male and female rats (approximately 0.2 per mg/kg-day) were only slightly lower than
slope factors based on alimentary tract tumors. The oral slope factor for female mice was
highest, at 1 per mg/kg-day for alimentary tract tumors (Table 5-6), approximately fourfold
higher than the oral slope factor derived from the alimentary tract tumors in male rats.
Although the time-to-tumor modeling helps account for competing risks associated with
decreased survival times and other tumors, considering the tumor sites individually still does not
convey the total amount of risk potentially arising from the sensitivity of multiple sites—that is,
the risk of developing any combination of the increased tumor types, not just the risk of
developing all simultaneously. One approach suggested in the Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005a) would be to estimate cancer risk from tumor-bearing animals.
EPA traditionally used this approach until the National Resource Council (NRC) document
Science and Judgment (NRC, 1994) made a case that this approach would tend to underestimate
overall risk when tumor types occur in a statistically independent manner. In addition,
223
DRAFT - DO NOT CITE OR QUOTE
-------
1 application of one model to a composite data set does not accommodate biologically relevant
2 information that may vary across sites or may only be available for a subset of sites. For
3 instance, the time courses of the multiple tumor types evaluated varied, as is suggested by the
4 variation in estimates of z (see Table 5-6), from 1.5 (e.g., male rat skin or mammary gland basal
5 cell tumors), indicating relatively little effect of age on tumor incidence, to 3.7 (e.g., male mouse
6 alimentary tract tumors), indicating a more rapidly increasing response with increasing age (in
7 addition to exposure level). The result of fitting a model with parameters which can reflect
8 underlying mechanisms, such as z in the multistage-Weibull model, would be difficult to
9 interpret with composite data (i.e., counts of tumor-bearing animals). A simpler model, such as
10 the multistage model, could be used for the composite data but relevant biological information
11 would then be ignored.
12 Following the recommendations of the NRC (1994) regarding combining risk estimates,
13 statistical methods which can accommodate the underlying distribution of slope factors are
14 optimal, such as through maximum likelihood estimation or through bootstrapping or Bayesian
15 analysis. However, these methods have not yet been extended to models such as the multistage-
16 Weibull model. A method involving the assumption that the variability in the slope factors could
17 be characterized by a normal distribution is detailed below (U.S. EPA, 2010). Using the results
18 in female rats to illustrate, the overall risk estimate involved the following steps:
19
20 1)
21
22
23
24
25
26
27
28
29 2)
30
31
32
33
34
35
36 3)
37
38
39
40 4)
41
42
43
224 DRAFT - DO NOT CITE OR QUOTE
It was assumed that the tumor groupings modeled above were statistically
independent—that is, that the occurrence of a liver tumor was not dependent
upon 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) 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), in order to reach the region of each
estimated dose-response function where the slope was reasonably constant and
upper bound estimation was still numerically stable. For these data, a 10" risk
was generally the lowest risk necessary. The oral slope factor for each site was
then estimated by R/BMDLr, as for the estimates for each tumor site above.
The maximum likelihood estimates (MLE) of unit potency (that is, risk per unit
of exposure) estimated by R/BMDr, were summed across the alimentary tract,
liver, and jejunum/duodenum in female rats.
An estimate of the 95% (one-sided) upper bound on the summed oral slope
factor 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 formula:
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
rearranged to:
95% UCL = MLE + 1.645 x s.d.,
s.d. = (UCL-MLE) / 1.645,
where 1.645 is the t-statistic corresponding to a one-sided 95% confidence
interval and >120 degrees of freedom, and the standard deviation (s.d.) is the
square root of the variance of the MLE. The variances (variance = s.d. ) 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 MLEs was calculated from
the expression above for the UCL, using the variance of the sum of the MLE to
1/2
obtain the relevant s.d (s.d. = variance ).
The resulting composite slope factor for all tumor types for male rats was 0.5 per mg/kg-
d, about 25% higher than the slope factor based on the most sensitive tumor site, oral cavity and
forestomach, while for female rats the composite slope factor was equivalent to that for the most
sensitive site (Table 5-6; see Table D-7 for details of the composite slope factor estimates).
The risk estimates from rats and mice spanned a nearly five-fold range. As there are no
data to support any one result as most relevant for extrapolating to humans, the most sensitive
result was used to derive the oral slope factor. The recommended slope factor for assessing
human cancer risk associated with chronic oral exposure to benzo[a]pyrene is 1 per mg/kg-day,
based on the alimentary tract tumor response in female B6C3Fi mice.
5.4.2. Inhalation Exposure—Inhalation Unit Risk
5.4.2.1. Choice of Study/Data—with Rationale and Justification—Inhalation Exposure
Inhalation exposure to benzo[a]pyrene was associated with nasal adenocarcinomas and
squamous cell tumors in the larynx, pharynx, trachea, esophagus, and forestomach, of male
Syrian golden hamsters exposed to benzo[a]pyrene:NaCl aerosols at concentrations of 10 or 50
"3
mg/m until natural death (up to 133 weeks) for 3-4.5 hours/day, 5-7 days/week (Thyssen et al.,
1981). Supportive evidence for the carcinogenicity of inhaled benzo[a]pyrene comes from 10
additional studies with hamsters exposed to benzo[a]pyrene via intratracheal instillation (see
Section 4.2.2.2 for references). However, the use of intratracheal dosing alters the deposition,
clearance, and retention of substances and therefore studies utilizing this exposure technique are
not as useful for the quantitative extrapolation of cancer risk from the inhalation of
benzo[a]pyrene in the environment (Driscoll et al., 2000).
The Thyssen et al. (1981) bioassay represents the only lifetime inhalation cancer bioassay
available for describing dose-response relationships for cancer from inhaled benzo[a]pyrene.
Limitations of the study include the following: (1) only male animals were included; (2) particle
analysis of aerosols was not reported [i.e., MMAD and geometric SD were not reported], and
(3) benzo[a]pyrene exposure occurred through the inhalation of hygroscopic particles
[benzo[a]pyrene was adsorbed onto NaCl aerosols] which may have a different deposition than
225
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
benzo[a]pyrene adsorbed onto non-hygroscopic particles in the environment. Strengths of the
study include exposure to hamsters for life, histological tumor examination of organs, use of
multiple exposure groups, including approximately 30 male hamsters per group, and the
availability of individual animal pathology reports with time of death and tumor detection data.
Although the study has a few limitations, the strengths of the study support use of the data to
derive an inhalation unit risk for benzo[a]pyrene.
5.4.2.2. Dose-response Data—Inhalation Exposure
Survival was decreased relative to control only in the high-dose exposure group; mean
"3
survival times in the 0, 2, and 10 mg/m concentration groups were 96.4, 95.2, and 96.4 weeks,
"3
respectively, and 59.5 weeks in the 50 mg/m group animals. Overall, tumors occurred earlier in
the highest benzo[a]pyrene exposure group than in the mid-exposure group. Increased
incidences of benign and malignant tumors of the larynx, trachea, pharynx, esophagus and
forestomach were seen with increasing exposure concentration. Benign tumors—papillomas,
polyps and papillary polyps—were considered by the study authors as early stages of the
squamous cell carcinomas in these tissues.
Nasal cavity tumors were also observed in the mid- and high-dose groups. Consideration
of early mortality (using the poly-3 approach; Bailer and Portier, 1988) suggested that an
increasing dose-response was consistent with these data. However trend testing was not
statistically significant (p=0.08), and the site was not considered further for unit risk derivation.
Table E-l in Appendix E summarizes the individual animal tumor data, noting the presence or
absence of a tumor in these tissues, whether or not the tissue was available for examination by
the pathologist, and the time of death. A summary of the incidence of these tumors is provided
in Table 5-7.
Table 5-7. Incidence of tumors in male hamsters exposed by inhalation to
benzo[a]pyrene for life
Average
continuous
benzo[a]pyrene
concentration"
(mg/m3)
Number
of
hamsters
in groupb
Larynx
Pharynx
Trachea
Esophagus
Forestomach
Any
Tumor0
Nasal
Cavity
Tumors
Control
27
0
0
0
0
0
0
0
0.25
27
0
0
0
0
0
0
0
1.01
26
11
9
2
0
1
18
4
4.29
34
12
18
3
2
1
18
1
aCalculated from air monitoring data.
bNumber of animals examined histologically, minus the number of animals who died before the earlier of the first
occurrence of the tumor type in each group or Week 52.
0 Includes any animal with squamous cell carcinoma of the larynx, pharynx, trachea, esophagus, or forestomach.
Source: Thyssen et al. (1981) and a reanalysis of this data by Clement Associates (1990). See Appendix E for
more detailed incidence data.
226
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
5.4.2.3. Dose Adjustments and Extrapolation Method(s)—Inhalation Exposure
A toxicokinetic model to assist in cross species scaling of benzo[a]pyrene inhalation
exposure was not available. In addition, default dosimetry adjustments utilized in the
benzo[a]pyrene RfC calculation could not be applied because aerosol particle distribution data
were not available for the hamster inhalation bioassay by Thyssen et al. (1981). The carrier
particle used in Thyssen et al. (1981) was sodium chloride, a soluble hygroscopic particle, and
the approaches presented in the RfC methodology guidelines (US EPA 1994b) were developed
for insoluble and nonhygroscopic particles.
The availability of the raw chamber air monitoring data and individual times on study
allowed the calculation of time-weighted average (TWA) continuous exposure rates for each
hamster. Group averages of individual TWA continuous exposure concentrations were 0, 0.25,
3 3
1.01, and 4.29 mg/m , respectively, for the 0, 2, 10, and 50 mg/m study concentrations.
A time-to-tumor dose-response model was fit to the time-weighted average exposure
concentrations and the individual animal occurrence data for tumors in the larynx, pharynx,
trachea, esophagus, and forestomach (Table E-l in Appendix E) using the computer software
program multistage-Weibull (U.S. EPA, 2010) as described in Section 5.4.1.3. The
investigators did not determine cause of death for any of the animals. Since in the available oral
bioassays the investigators considered these same tumors to be fatal at least some of the time,
bounding estimates for the Thyssen et al. data were developed by treating the tumors alternately
as either all incidental or all fatal. In either case, therefore, an estimate of to (the time between a
tumor first becoming observable and causing death) could not be estimated.
Because benzo[a]pyrene is expected to cause cancer via a mutagenic MO A, a linear
approach to low dose extrapolation from the BMCLio was used (U.S. EPA, 2005a).
5.4.2.4. Inhalation Unit Risk Derivation
Modeling results are provided in Appendix E. The BMC (0.28 mg/m3) and BMCL (0.20
"3
mg/m ) associated with an extra risk of 10% were calculated based on the occurrence of upper
respiratory and upper digestive tract tumors in male hamsters exposed to aerosols of
benzo[a]pyrene for 104 weeks using the multistage-Weibull model Using linear extrapolation
from the BMCLio of 0.20 mg/m3, an inhalation unit risk of 0.5 per mg/m3 or 5 x 10"4 per jug/m3
was calculated.
5.4.3. Dermal Exposure—Dermal Slope Factor
5.4.3.1. Choice of Study/Data—with Rationale and Justification—Dermal Exposure
Skin cancer in humans has been documented to result from occupational exposure to
complex mixtures of PAHs including benzo[a]pyrene such as coal tar, coal tar pitches, non-
refined mineral oils, shale oils and soot (IARC, 2010; Baan et al., 2009; Boffetta et al., 1997;
227
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
WHO, 1998; ATSDR, 1995), but no studies of human exposures to benzo[a]pyrene alone are
known to exist. In animal models, numerous dose response studies have demonstrated an
increased incidence of skin tumors with increasing dermal exposure to benzo[a]pyrene, in all
species tested (mice, rabbits, rats, and guinea pigs), though most benzo[a]pyrene chronic dermal
bioassays which provide quantitative information have been conducted in mice. In addition,
mice appear to be the most sensitive laboratory model of carcinogenesis following dermal
benzo[a]pyrene exposure. Therefore, this analysis focuses on chronic carcinogenicity bioassays
in several strains of mice demonstrating increasing incidence of benign and malignant skin
tumors, and earlier occurrence of tumors with increasing exposure, following repeated dermal
exposure to benzo[a]pyrene for the animals' lifetime. These studies involved 2- or 3-times/week
exposure protocols, at least two exposure levels plus controls, and included histopathological
examinations of the skin and other tissues (Sivak et al., 1997; Grimmer et al., 1984; 1983; Habs
et al., 1980; 1984; Schmahl et al., 1977; Schmidt et al., 1973; Roe et al., 1970; Poel, 1960; 1959).
These data sets are described in greater detail in Section 4.2.3.2.
Because of the availability of the lifetime studies listed above, other carcinogenicity
studies were not considered for this assessment. The other studies included: 1) early "skin
painting" studies of benzo[a]pyrene carcinogenicity in mouse skin which did not report sufficient
information to estimate the doses applied (e.g., Wynder and Hoffman 1959; Wynder et al, 1957);
2) initiation-promotion studies utilizing acute dosing of benzo[a]pyrene followed by repeated
exposure to a potent tumor promoter (sometimes benzo[a]pyrene at a lower dose than the
initiation step), because they are not as relevant for calculating risks from constant
benzo[a]pyrene exposure alone; 3) bioassays with one benzo[a]pyrene dose level or with only
dose levels inducing 90-100% incidence of mice with tumors, because they provide relatively
little information about the shape of the dose-response relationship (e.g., Wilson and Holland,
1988); 4) studies with shorter exposure and observation periods (i.e., less than one year; Levin et
al., 1977; Nesnow et al, 1983; Albert et al., 1991; Emmett et al, 1981; Higginbotham et al., 1993)
which are less relevant for characterizing lifetime risk; and 5) studies involving vehicles
expected to interact with or enhance benzo[a]pyrene carcinogenicity (e.g., Bingham and Falk,
1969) which precludes assessment of carcinogenic risks of benzo[a]pyrene alone.
5.4.3.2. Dose-response Data—Dermal Exposure
Several studies were considered for dose-response modeling for derivation of the dermal
slope factor for benzo[a]pyrene, reflecting a relatively large database. Study designs and the
extent of data reported varied across the studies, with no individual animal data available. All of
the studies identified in the previous section (Sivak et al., 1997; Grimmer et al., 1984; 1983;
Habs et al., 1980; 1984; Schmahl et al., 1977; Schmidt et al., 1973; Roe et al., 1970; Poel, 1960;
1959) were considered further in order to evaluate overall consistency of the available database.
These data sets are presented in Tables 5-8 through 5-12, and are grouped by study
228
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
characteristics such as mouse strain and vehicle in order to facilitate qualitative comparisons
where possible.
Each data set was examined for study design and both strengths and limitations that could
potentially impact the dose-response evaluation, including the potential of early mortality to
impact the number at risk for developing tumors, and the length of exposure. Although nearly all
studies reported lifetime exposures, this term most often indicated that exposure continued until
natural death, not a scheduled sacrifice at 104 weeks.
The studies by Poel (1959, 1960) were conducted in male mice and used toluene as the
vehicle (see Tables 4-11, 4-12, respectively). In addition to a control group, the 1959 study
included nine dose groups of one mouse strain (C57L) and the 1960 study included seven dose
groups of 3 other mouse strains. Both studies demonstrated high mortality and tumor incidence
at higher exposure levels. As noted in Table 4-11, all C57L mice in dose groups with >3.8
[j,g/application died by Week 44 of the study (Poel, 1959). Therefore, these five dose groups
were omitted prior to dose-response modeling because of the relatively large uncertainty in
characterizing cancer risk in relation to lifetime exposure. Four dose groups in addition to
control remained. Among these groups mice survived and were exposed until weeks 83-103.
According to the lifespan ranges provided, at least one mouse in each dose group died before the
first appearance of tumor, but insufficient information was available to determine how many;
consequently the incidence denominators were not adjusted. The dose-response data are
summarized in Table 5-7.
For the Poel (1960) studies, all tumors in the highest three dose groups for each of the
three mouse strains had occurred by Week 40 (see Table 4-12). While these observations
support concern for cancer risk, as noted above such results are relatively uncertain for
estimating lifetime cancer risk. In addition, there was no information indicating duration of
exposure for the mice without tumors; although exposure was for lifetime, it might have been as
short as for the mice with tumors. Overall, these datasets did not provide sufficient information
to estimate the extent of exposure associated with the observed tumor incidence. Consequently
the experiments reported by Poel (1960) were not used for dose-response modeling.
The studies listed in Table 5-9 all used acetone as the vehicle and either Swiss or NMRI,
female mice (Roe et al., 1970; Schmidt et al., 1973; Schmahl et al., 1977; Habs et al., 1980,
1984). Roe et al. (1970) applied benzo[a]pyrene dermally for 93 weeks or until natural death;
with the exception of the highest dose group, each group still had approximately 20 animals at 86
weeks (Table 4-14). The tumors were first observed in the lowest and highest dose groups
during the interval of weeks 29-43. Mice that died before week 29 were likely not at risk of
tumor development. However because tumor incidence and mortality were reported in 100-day
intervals, mice that had not been on study long enough to develop tumors were not easily
identifiable. Incidence denominators reflect the number of animals alive at Week 29, and thus
229
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
may tend to lead to underestimates of tumor risk if the number of animals at risk have been
overestimated.
Schmidt et al. (1973) did not report survival information, instead the authors provided
incidences based on the numbers of mice initially included in each dose group at the start of the
study. Overall latency was reported for the two high dose groups in each series, but these data
only describe the survival of mice with tumors (animals were removed from study when a tumor
appeared). It is not clear how long exposures lasted overall in each dose group, or whether some
mice may have died on study from other causes before tumors appeared. While it is possible that
no mice died during the study, all of the other studies considered here demonstrate mortality.
However, the data were modeled as reported, recognizing the possibility of underestimating risk
associated with incidences reported and lack of duration of exposure.
Schmahl et al. (1977) reported that reduced numbers of animals at risk (77-88 mice per
dose group compared with the initial group sizes of 100) resulted from varying rates of autolysis.
No other survival or latency information was provided, so all exposures were assumed to have
lasted for 104 weeks and were modeled as reported. Given the results of the other studies, it
seems possible that the numbers at risk in each group may be overestimated, which could lead to
an underestimate of lifetime risk.
Habs et al. (1980) reported age-standardized skin tumor incidence rates, indicating earlier
mortality in the two highest dose groups (2.8 and 4.6 ^/application). These rates were used to
estimate the number at risk in the dose-response modeling, by dividing the number of mice with
tumors by the age-standardized rates (see Table 5-9). Exposure lasted longer than 104 weeks in
the two lower exposure groups, at about 120 and 112 weeks, and until about 88 weeks in the
highest exposure group. Incidence in the two lower exposure groups may be higher than if the
exposure had lasted just 104 weeks. There was mortality in the first 52 weeks of exposure, about
10-15% in the three exposure groups, but because there was no information concerning when
tumors first appeared it is not possible to determine how much the early mortality may have
impacted the number of mice at risk in each group.
Habs et al. (1984) reported mean survival times (with 95% confidence intervals) for each
dose group. The confidence intervals supported the judgment that the control and lower dose
groups were treated for 104 weeks. The higher dose group (4 (j,g/application) was probably
treated for less than 104 weeks, because the upper 95% confidence limit for the mean survival
was approximately 79 weeks (Table 4-20). However, since it was not possible to estimate a
more realistic duration for this group, an estimate of 104 weeks was used.
Grimmer et al. (1983 and 1984), studied female CFLP mice, using acetone:DMSO (1:3)
as the vehicle (see Table 5-10). Mean or median latency times were reported (as well as
measures of variability), but no information concerning overall length of exposure or survival
was included in the results. The total of tumor-bearing mice and the reported percentages of
mice with any skin tumors was reported and varied at most one animal from the number of
230
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
animals initially placed on study. The decreasing latency and variability and increasing tumor
incidence with increasing benzo[a]pyrene exposure suggests that exposure probably did not last
for 104 weeks in at least the high dose group, but the available information did not provide
duration of exposure. The data reported were modeled under the assumption that at least some
animals in each group were treated and survived until Week 104.
The study listed in Table 5-11, Sivak et al. (1997) exposed male C3H /HeJ mice dermally
to benzo[a]pyrene in cyclohexanone/acetone (1:1) for 24 months, and reported mean survival
times for each group (see Table 4-21). All high dose mice died before the final sacrifice. From
the information provided it is apparent that the animals in the control and lower two dose groups
survived until study termination at Week 104. The study authors did not report how long
treatment in the highest dose group lasted, but estimation of the figure from the publication
suggest that exposure duration was 74 weeks. The tumor incidences and estimated duration of
exposure for each dose group are presented in Table 5-11.
231
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-8. Skin tumor incidence, benign or malignant, in C57L male mice
dermally exposed to benzo[a]pyrene
Average
First
Length
daily
appearance
of
Lifetime
Skin tumor
Mouse
dose
of tumor
Exposure
Average
incidence (all
Study
strain
Dose (jig)a
(Hg/d)
(weeks)
(weeks)
Daily Doseb
types)
Poel (1959)
C57L
0 (toluene)
0
—
92
0.00
0/33 (0%)
0.15
0.06
42
98
0.05
5/55 (9%)
0.38
0.16
24
103
0.16
11/55 (20%)
0.75
0.32
36
94
0.24
7/56(13%)
3.8
1.63
21-25
82
0.80
41/49 (84%)
"Doses were applied to interscapular skin 3 times/wk for up to 103 weeks or until time of appearance of a grossly
detected skin tumor. See Table 4-11 for data of five highest dose groups (19-752 (xg) in which all mice died by
Week 44. These groups were not considered for dose-response modeling.
bSee Section 5.4.3.3. for discussion of extrapolation to lifetime average daily doses.
232
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-9. Skin tumor incidence, benign or malignant in female Swiss or
NMRI mice dermally exposed to benzo[a]pyrene
Average
First
Lifetime
daily
appearance
Length of
average daily
Mouse
dose
of tumor
exposure
dose
Skin tumor
Study
strain
Dose (fig)
(Hg/d)
(weeks)
(weeks)
(|Ag/d)
incidence (all types)
Roe et al.
Swiss
0 (acetone)
0
—
93
0.00
0/49 (0%)
(mo)3-13
0.1
0.04
29-43
93
0.03
1/45 (2%)
0.3
0.13
—
93
0.09
0/46 (0%)
1
0.43
57-71
93
0.31
1/48 (2%)
3
1.29
43-57
93
0.92
8/47 (20%)
9
3.86
29-43
93
2.76
34/46 (74%)
Schmidt et al.
NMRI
0 (acetone)
0
—
104d
0
0/100 (0%)
(1973)°
0.05
0.01
—
104
0.01
0/100 (0%)
0.2
0.06
—
104
0.06
0/100 (0%)
0.8
0.23
53e
104
0.23
2/100 (2%)
2
0.57
76e
104
0.57
30/100 (30%)
Swiss
0 (acetone)
0
—
104
0
0/80 (0%)
0.05
0.01
—
104
0.01
0/80 (0%)
0.2
0.06
—
104
0.06
0/80 (0%)
0.8
0.23
00
104
0.23
5/80 (6%)
2
0.57
61e
104
0.57
45/80 (56%)
Schmahl et al.
NMRI
0 (acetone)
0
—
104
0
1/81 (1%)
(1977)°
1
0.29
NR
104
0.29
11/77 (14%)
1.7
0.49
NR
104
0.49
25/88 (28%)
3
0.86
NR
104
0.86
45/81 (56%)
Habs et al.
NMRI
0 (acetone)
0
—
128
0
0/35 (0%% )e
(1980)°
1.7
0.49
NR
120
0.49
8/34 (24.8%)
2.6
0.74
NR
112
0.74
24/27 (89.3%)
4.6
1.31
NR
88
0.80
22/24 91.7%)
Habs et al.
NMRI
0 (acetone)
0
—
104
0
0/20 (0%)
(1984)°
2
0.57
NR
104
0.57
9/20 (45%)
4
1.14
NR
104
1.14
17/20 (85%)
aDoses were applied 3 times/week for up to 93 wks to shaved dorsal skin.
^Numerator: number of mice detected with a skin tumor. Denominator: number of mice surviving to 29 weeks
(200 days).
°Doses were applied 2 times/week to shaved skin of the back. Mice were exposed until natural death or until they
developed a carcinoma at the site of application.
dExposure periods not reported were assumed to be 104 weeks; indicated in italics.
"Central tendency estimates; range or other variability measure not reported.
fThe percentages were reported by the authors as age-standardized tumor incidences, derived using mortality data
from the entire study population. The incidences reflect reported counts of tumor-bearing animals and
denominators estimated from the reported age-standardized rates.
NR=not reported.
233
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Table 5-10. Skin tumor incidence, benign or malignant, in female CFLP mice
dermally exposed to benzo[a]pyrene
Mean or
median time
Lifetime
Average
of tumor
Length of
average daily
Skin tumor
daily dose
appearance
exposure
dose
incidence (all
Study
Dose (jig)a
(|Ag/d)
(weeks)
(weeks)d
(Hg/d)
types)6
Grimmer et al.
0 (1:3 acetone:DMSO)
0
104
0
0/80 (0%)
(1983)
3.9
1.1
74.6 ± 16.8b
104
1.1
22/65 (34%)
7.7
2.2
60.9 ± 13.9
104
2.2
39/64 (61%)
15.4
4.4
44.1 ±7.7
104
4.4
56/64 (88%)
Grimmer et al.
0 (1:3 acetone:DMSO)
0
104
0
0/80 (0%)
(1984)
3.4
0.97
61 (53-65)°
104
0.97
43/64 (67%)
6.7
1.9
47 (43-50)
104
1.9
53/65 (82%)
13.5
3.9
35 (32-36)
104
3.9
57/65 (88%)
"Indicated doses were applied twice/week to shaved skin of the back for up to 104 weeks.
Vlean ± SD.
0 Median and 95% confidence limit.
dAssumed exposure period is indicated in italics.
"Incidence denominators were calculated from reported tumor-bearing animals and reported percentages.
Table 5-11. Skin tumor incidence, benign or malignant, in male C3H /HeJ
mice dermally exposed to benzo[a]pyrene(Sivak et al., 1997)
Average
First
Lifetime
daily
appearance
Length of
average
Skin tumor
dose
of tumor
exposure
daily dose
incidence (all
Dose (ng)a
(ng/d)
(weeks)
(weeks)b
(ng/d)
types)
0 (1:1 cyclohexanone/acetone)
0
104
0.0
0/30 (0%)
0.05
0.01
104
0.01
0/30 (0%)
0.5
0.14
NR
104
0.14
5/30 (17%)
5.0
1.4
-43
74
0.51
27/30 (90%)
"Indicated doses were applied twice/week to shaved dorsal skin.
bAssumed exposure period is indicated in italics.
NR=not reported.
5.4.3.3. Dose Adjustments and Extrapolation Method(s)—Dermal Exposure
As with the oral and inhalation benzo[a]pyrene carcinogenicity data (see sections 5.4.1.3
and 5.4.2.3), benzo[a]pyrene's dermal exposure carcinogenicity data were generally
characterized by earlier occurrence of tumors and increased mortality with increasing exposure
level. However, individual animal data were not available for any of the identified studies.
Therefore, time to tumor modeling was not possible. Each of the dermal data sets was modeled
using the multistage model, incorporating adjustments for early mortality, when data were
available.
234
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
First, for all studies, administered doses were converted to average daily doses using the
equation:
Average dose/day = (^/application) x (number of exposures/week 7 days/week).
Next, lifetime equivalent doses were estimated for study groups that were reported to end
"3
before 104 weeks by multiplying the relevant average daily doses by (Le/104) , where Le is the
length of exposure, based on observations that tumor incidence tends to increase with age (Doll,
1971). Note that exposure periods less than 52 weeks would lead to a relatively large adjustment
"3
[i.e., (52/104) = 0.125, or an eightfold lower dose than administered], reflecting considerable
uncertainty in lifetime equivalent dose estimates generated from relatively short studies.
The multistage-cancer model in the EPA BMDS (version 2.1) was fit to each data set in
Tables 5-8 through 5-11. The multistage model with the most parsimonious fit (fewest
parameters yielding an adequate fit) was selected to calculate the potential POD from each data
set (see Appendix F for details). Because the multistage model is preferred for cancer modeling,
the conventional a-level of 0.05 was used to judge goodness-of-fit. If there was no adequate fit
using the multistage-cancer model, then other dichotomous models were considered. If there
was still no adequate fit, high doses were dropped incrementally and the multistage-cancer model
was considered before attempting other models. BMDs and BMDLs associated with an extra
risk of 10% were calculated. Because benzo[a]pyrene is expected to cause cancer via a
mutagenic MO A, a linear approach to low dose extrapolation from the PODs (i.e., BMDLio) was
used (U.S. EPA, 2005a) for candidate dermal slope factors.
5.4.3.4. Dermal Slope Factor Derivation
Adequate model fits were found using the multistage model for all but one of the mouse
skin tumor incidence data sets in Tables 5-8 to 5-11, as described in Appendix F. In one case,
the data from Grimmer et al. (1984) could not be adequately fit by the multistage model initially,
and the other dichotomous models available in BMDS were considered. Due to the supralinear
shape of the dose-response data, only the log-logistic and dichotomous Hill models provided
adequate fits. Also due to the supralinear dose-response shape, the point of departure for slope
factor derivation was identified near the lowest response of-70%, in order to avoid excessive
extrapolation of the fitted model.
Dermal slope factors, calculated in units of risk per (|ig/day) using linear extrapolation
from the BMDLio values, ranged from 0.25 to 1.8 per |ig/day, a roughly 7-fold range (see Table
5-12). A number of differences among studies contribute to this range, including solvent choice,
sex and strain of mice studied, dose ranges and the level of detail reported. Mouse strains were
not repeated across sexes among these studies, so it cannot be established whether male mice are
235
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
generally more or less sensitive than female mice to benzo[a]pyrene dermal carcinogenicity, or
whether Swiss or NRMI mice are more or less sensitive than other strains. In addition, different
solvents were used in the various studies with varying strain and sex combinations tested. For
example, toluene was used in one male study only and all of the female studies used acetone.
Thus, any possible impact of the solvents used is not clear. The estimates derived from the two
studies in males were at the higher end of the range of slope factors derived, but the available
information is too limited to conclude that males are more sensitive than females. Also, as noted
earlier, incomplete mortality information in several of the female mouse studies (Schmidt et al.,
1973; Schmahl et al., 1977; Grimmer et al., 1983, 1984; and Habs et al., 1980, 1984) suggests
that the derived dermal slope factors may underestimate cancer risk.
The BMDLio of 0.066 |ig/animal-day, based on the tumor response in C3H/HeJ male
mice (Sivak et al., 1997), is recommended for developing a human dermal slope factor because it
is the lowest POD among studies with lower doses where intercurrent mortality was less likely to
impact the number at risk and represented a chronic duration of exposure.
Table 5-12. PODs derived from skin tumor incidence data in mice exposed to
benzo[a]pyrene by the dermal route of exposure"
Reference
Mouse strain
Solvent
BMD10
(jig/animal-d)
BMDL10
(jig/animal-d)
Male mice
Sivak et al., 1997
C3H/HeJ
Acetone/
cyclohexanone
0.12
0.066
Poel, 1959
C57L
Toluene
0.12
0.077
Female mice
Habs et al., 1984
NMRI
Acetone
0.078
0.056
Grimmer et al., 1984
CFLP
Acetone/DMSO
1.07b
0.48b
Schmahl et al., 1977
NMRI
Acetone
0.23
0.15
Schmidt et al., 1973
Swiss
Acetone
0.28
0.22
Grimmer et al., 1983
CFLP
Acetone/DMSO
0.24
0.21
Habs et al., 1980
NMRI
Acetone
0.29
0.22
Schmidt et al., 1973
NMRI
Acetone
0.33
0.29
Roe et al., 1970
Swiss
Acetone
0.69
0.39
aSee Appendix F for details of modeling results.
bBMR=70% for this dataset, in order to avoid excessive extrapolation via the fitted model.
5.4.3.5. Dermal Slope Factor Cross Species Scaling
Different methodologies have been established for interspecies scaling of points of
departure (PODs) used to derive oral slope factors and inhalation unit risks. Cross-species
adjustment of oral doses is based on allometric scaling using the three-fourths power of body
weight. This adjustment accounts for more rapid distribution, metabolism, and clearance in
small animals (US EPA 2005). Cross-species extrapolation of inhalation exposures is based on
standard dosimetry models that consider factors such as solubility, reactivity, and persistence
236
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
(US EPA 1994). However, no established methodology exists to adjust for interspecies
differences in dermal toxicity at the point of contact. Because there is no established
methodology for cross-species extrapolation of dermal toxicity, several alternative approaches
were evaluated (see Appendix H). Among the alternative described in Appendix H, cross-
species adjustment based on allometric scaling using body weight to the 3/4 power was selected.
Under this approach, rodents and humans exposed to the same daily dose of a carcinogen,
adjusted for BW3 4, would be expected to have equal lifetime risks of cancer.
The PODm derived from the mouse study by Sivak et al., (1997) is adjusted to a human
equivalent dose (HED) as follows:
POD HED(|ig/day) = PODM fag/day) x (BW„/BWVI)34
PODhed (|ig/day) = 0.066 |ig/day x (70 kg / 0.035 kg)3/4
= 19.7 |ig/day
The resulting PODhed is used to calculate the dermal slope factor for benzo[a]pyrene:
DSF = 0.1/PODhed
DSF = 0.1/(19.7 |ig/day) = 0.005 (jig/day)1
Note that the DSF should only be used with lifetime human exposures < 20 |ig/day, the
human equivalent of the bioassay POD, because above this level the dose-response relationship
may not be proportional to mass of the compound applied.
Several assumptions are made in the use of this scaling method. First, it is assumed that
the toxicokinetic processes in the skin will scale with interspecies differences in whole body
toxicokinetics. Secondly, it is assumed that the risk at low doses of benzo[a]pyrene is linear;
however, one study indicates that at high doses of benzo[a]pyrene, carcinogenic potency is
related to mass applied per unit skin and not to total mass (Davies 1967). However, this may be
due to promotional effects, such as inflammation, that are observed at high doses of
benzo[a]pyrene.
This slope factor has been developed for a local effect and it is not intended to estimate
systemic risk of cancer following dermal absorption of benzo[a]pyrene into the systemic
circulation. Although some information suggests that benzo[a]pyrene metabolites can enter
systemic circulation following dermal exposure (Godschalk et al 1998), lifetime skin cancer
bioassays which have included pathological examination of other organs, have not found
237
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
elevated incidences of tumors at distal sites (Poel 1959; Roe et al., 1970; Schmidt et al., 1973;
Schmahl et al., 1977; Habs et al., 1980; Higginbotham et al.,1993). In addition, benzo[a]pyrene
tends to bind to targets within the skin rather than enter the plasma receptor fluid (a surrogate
measure of systemic absorption) in in vitro human skin experiments. These data are consistent
with benzo[a]pyrene's metabolism to reactive metabolites within the viable layers of the skin
(Wester et al., 1990). Some studies indicate that the fraction of benzo[a]pyrene left within the
viable layers of the skin is a large portion of the applied dose (Moody et al., 2007; 1995). Taken
together, these data support the conclusion that the risk of skin cancer following dermal exposure
likely outweighs cancer risks at distal organs.
5.4.4. Application of Age-Dependent Adjustment Factors
Based on sufficient support in laboratory animals and relevance to humans (see Section
4.7.3) benzo[a]pyrene is determined to be carcinogenic by a mutagenic MOA. According to the
Supplemental Guidance for Assessing Susceptibility from Early Life Exposure to Carcinogens
("Supplemental Guidance") (U.S. EPA, 2005b), individuals exposed during early life to
carcinogens with a mutagenic MOA are assumed to have increased risk for cancer. The oral
"3
slope factor of 1.4 per mg/kg-day, inhalation unit risk of 0.5 per mg/m , and dermal slope factor
of 0.0051 per |ig/day for benzo[a]pyrene, calculated from data applicable to adult exposures, do
not reflect presumed early life susceptibility to this chemical. Though some chemical specific
data exist for benzo[a]pyrene which demonstrate increased early life susceptibility to cancer
(Vesselinovitch et al. 1984), these data were not considered sufficient to develop separate risk
estimates for childhood exposure, as they used acute, i.p. exposures (U.S. EPA, 2005b). In the
absence of adequate chemical-specific data to evaluate differences in age-specific susceptibility,
the Supplemental Guidance (U.S. EPA, 2005b) recommends that age-dependent adjustment
factors (ADAFs) be applied in estimating cancer risk.
The Supplemental Guidance (U.S. EPA, 2005b) establishes ADAFs for three specific age
groups. These ADAFs and their corresponding age groupings are: 10 for individuals exposed
<2 years, 3 for exposed individuals 2 to <16 years, and 1 for exposed individuals >16 years. The
10- and 3-fold adjustments are combined with age specific exposure estimates when estimating
cancer risks from early life (<16 years age) exposures to benzo[a]pyrene To illustrate the use of
the ADAFs established in the Supplemental Guidance (U.S. EPA, 2005b), sample calculations
are presented for three exposure duration scenarios, including full lifetime, assuming a constant
benzo[a]pyrene exposure of 0.001 mg/kg-day (Table 5-13).
Calculations for the application of ADAFs to oral exposures are presented in Table 5-13;
calculations for exposures by the inhalation and dermal routes follow the same procedure (Table
5-13 and 5-14). Exposure duration scenarios include full lifetime exposure (assuming a 70-year
lifespan), and two 30-year exposures at ages 0-30 and ages 20-50. Table 5-13 lists the four
factors (ADAFs, cancer risk estimate, assumed exposure, and duration adjustment) that are
238
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
needed to calculate the partial cancer risk based on the early age-specific group. The cancer risk
for each age group is the product of the four factors in columns 2-5. Therefore, the cancer risk
following daily benzo[a]pyrene oral exposure in the age group 0 to <2 years is the product of the
values in columns 2-5 or 10 x lx 0.001 x 2/70 = 4 x 10"4. The cancer risk for specific exposure
duration scenarios that are listed in the last column are added together to get the total risk. Thus,
a 70-year (lifetime) risk estimate for continuous exposure to 0.001 mg/kg-day benzo[a]pyrene is
2 x 10" , which is adjusted for early-life susceptibility and assumes a 70-year lifetime and
constant exposure across age groups.
Table 5.13. Application of ADAFs to benzo[a]pyrene cancer risk following a
lifetime (70-year) oral exposure
Age
Group
ADAF
Unit risk
Exposure Concentration
Duration
adjustment
Cancer Risk
for Specific
Exposure
Duration
Scenarios
(per mg/kg-day)
(mg/kg-day)
0-<2 yrs
10
1
0.001
2 yrs/70 yrs
0.0003
2—<16 yrs
3
1
0.001
14 yrs/70 yrs
0.0006
>16 yrs
1
1
0.001
54 yrs/70 yrs
0.0007
Total Risk
0.002
In calculating the cancer risk for a 30-year constant exposure to benzo[a]pyrene at an
exposure level of 0.001 mg/kg-day from ages 0-30, the duration adjustments would be 2/70,
14/70, and 14/70, and the partial risks for the three age groups would be 3 x 10"4, 6 x 10"4, and
2 x 10"4, which would result in a total risk estimate of 1 x 10"3.
In calculating the cancer risk for a 30-year constant exposure to benzo[a]pyrene at an
exposure level of 0.001 mg/kg-day from ages 20-50, the duration adjustments would be 0/70,
0/70, and 30/70. The partial risks for the three groups are 0, 0, and 4 x 10"4, which would result
in a total risk estimate of 4 x 10"4.
Consistent with the approaches for the oral route of exposure, the ADAFs should also be
applied when assessing cancer risks for subpopulations with early life exposures to
benzo[a]pyrene via the inhalation and dermal routes are presented in Tables 5-14 and 5-15.
Table 5-14. Application of ADAFs to benzo[a]pyrene cancer risk following a
lifetime (70-year) inhalation exposure
Unit risk
Exposure Concentration
Duration
Cancer Risk
Age
Group
ADAF
(per fig/m3)
(figlm3)
adjustment
for Specific
Exposure
Duration
Scenarios
239
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
0-<2 yrs
10
5x 10-4
1
2 yrs/70 yrs
0.0001
2—<16 yrs
3
5x 10-4
1
14 yrs/70 yrs
0.0003
>16 yrs
1
5x 10-4
1
54 yrs/70 yrs
0.0004
Total Risk
0.0008
Table 5-15. Application of ADAFs to benzo[a]pyrene cancer risk following a
lifetime (70-year) dermal exposure
Age
Group
ADAF
Unit risk
Exposure Concentration
Duration
adjustment
Cancer Risk
for Specific
Exposure
Duration
Scenarios
(per fig/day)
(fig/day)
0-<2 yrs
10
0.005
0.001
2 yrs/70 yrs
1 x 10"6
2—<16 yrs
3
0.005
0.001
14 yrs/70 yrs
3 x 10-6
>16 yrs
1
0.005
0.001
54 yrs/70 yrs
4x 10-6
Total Risk
8 x 10-5
5.4.5. Uncertainties in Cancer Risk Values
5.4.5.1. Oral Slope Factor
Uncertainty in the recommended oral slope factor is reflected in the range of slope factors
among tumors sites and species; the lowest and highest slope factors listed in Table 5-6 show
about a 35-fold difference. While the highest risk estimates were derived from the incidence
data for forestomach tumors in both rats and mice, the oral slope factor based on the mouse
forestomach data was about threefold higher than the oral slope factor based on male rat data
(Table 5-6). These comparisons show that the selection of target organ, animal species, and
dosimetric extrapolation can impact the oral cancer risk estimate. However, all of the activation
pathways implicated in benzo[a]pyrene carcinogenicity have been observed in human tissues and
associations have been made between the spectra of mutations in tumor tissues from
benzo[a]pyrene-exposed animals and humans exposed to complex PAH mixtures containing
benzo[a]pyrene (see Section 4.7.3).
5.4.5.2. Inhalation Unit Risk
Only one animal cancer bioassay by the inhalation route is available which describes the
dose-response relationship for respiratory tract tumors with chronic inhalation exposure to
benzo[a]pyrene (Thyssen et al., 1981). Although corroborative information on dose-response
relationships in other animal species is lacking, the findings for upper respiratory tract tumors are
consistent with findings in other hamster studies with intratracheal administration of
benzo[a]pyrene. This study is adequate for dose-response analysis and derivation of an
inhalation unit risk estimate, but some associated uncertainty includes the inability to apply U.S.
240
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
EPA (1994b) dosimetry approaches to extrapolate inhaled doses from animals to humans, due to
the use of a soluble hygroscopic carrier particle (NaCl) for the delivery of benzo[a]pyrene. One
likely consequence of the use of hygroscopic carrier particles would be the growth of
benzo[a]pyrene-NaCl particles in the humid environment of the respiratory tract resulting in
increased particle diameter and resulting changes in particle deposition, specifically, increased
impaction in the upper respiratory tract (Xu and Yu 1985; Ferron 1994; Asgharian 2004;
Varghese and Gangamma 2009). Exposure to benzo[a]pyrene in the environment predominantly
occurs via non-soluble, non-hygroscopic particles. The potential impact of differences in carrier
particle on the magnitude of the inhalation unit risk is unknown.
5.4.5.3 Dermal Slope Factor
Uncertainty in the recommended dermal slope factor is partly reflected in the range of
slope factors derived from the modeled mouse skin tumor data sets: the lowest and highest
dermal slope factors listed in Table 5-12 show a 7-fold difference (0.25-1.8 ug/day) in
magnitude. There is some indication that the recommended dermal slope factor may
underestimate cancer risk, due to inadequate data to take the observed decreasing tumor latency
with increasing exposure level into account. Reliance on studies with the lowest exposure levels
where early mortality due to benzo[a]pyrene exposure was low and exposures continued for
approximately 104 weeks may minimize this source of uncertainty.
Human dermal exposure to benzo[a]pyrene in the environment likely occurs
predominantly through soil contact. The available mouse dermal bioassays of benzo[a]pyrene
relied on delivery of benzo[a]pyrene to the skin in a solvent solution (typically acetone or
toluene). The use of a volatile solvent likely results in a larger dose of benzo[a]pyrene available
for uptake into the skin (compared to soil). Reliance on these studies may overestimate the risk
of skin tumors from benzo[a]pyrene contact through soil; however, cancer bioassays delivering
benzo[a]pyrene through a soil matrix are not available.
There is uncertainty in extrapolating from the intermittent exposures in the mouse assays
to daily exposure scenarios. This assessment makes the assumption that risk is proportional to
total cumulative exposure. The extent to which this assumption under- or overestimates risk is
unknown.
The available data were not useful to determine which animal species may be the best
surrogate for human dermal response to benzo[a]pyrene. In extrapolation of the animal dermal
information to humans the inherent assumption is that equal area of skin from a mouse or human
would have equal probability of developing a tumor upon benzo[a]pyrene exposure.
Qualitatively, the toxicokinetics and toxicodynamics in mouse and human skin appear to be
similar (Knafla et al., 2010; Bickers et al., 1984). Specifically, all of the activation pathways
implicated in benzo[a]pyrene carcinogenicity have been observed in mouse and human skin and
associations have been made between the spectra of mutations in tumor tissues from
241
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
benzo[a]pyrene-exposed animals and humans exposed to complex PAH mixtures containing
benzo[a]pyrene (see Section 4.7.3).
This dermal slope factor for benzo[a]pyrene is based on skin cancer and it is not
developed to represent systemic cancer risk from dermal exposure. It is unclear whether dermal
exposure to benzo[a]pyrene would result in elevated risk of systemic tumors. Some studies in
humans suggest that although the skin may be responsible for a "first pass" metabolic effect,
benzo[a]pyrene-specific adduct levels have been detected in WBC following dermal exposure to
benzo[a]pyrene, indicating that dermally applied benzo[a]pyrene enters systemic circulation
(Godschalk et al., 1998). Although none of the lifetime dermal bioassays in mice, which
included macroscopic examination of internal organs, reported an elevation of systemic tumors
in benzo[a]pyrene-treated mice compared to controls (Poel 1959; Roe et al., 1970; Schmidt et al.,
1973; Schmahl et al., 1977; Habs et al., 1980; Higginbotham et al.,1993), most of these studies
attempted to remove animals with grossly observed tumors of the skin from the study before the
death of the animal, possibly minimizing the development of more distant tumors with longer
latency. The risk of benzo[a]pyrene-induced point of contact tumors in the skin likely competes
with systemic risk of tumors. Currently, the potential contribution of dermally absorbed
benzo[a]pyrene to systemic cancer risk is unclear.
5.5.5. Previous Cancer Assessment
The previous cancer assessment for benzo[a]pyrene was posted on the IRIS database in
1987. At that time, benzo[a]pyrene was classified as a probable human carcinogen (Group B2)
based on inadequate data in humans and sufficient data in animals via several routes of exposure.
An oral slope factor was derived from the geometric mean of four slope factor estimates based
on studies in Sprague-Dawley rats (Brune et al., 1981) and CFW-Swiss mice (Neal and Rigdon,
1967). Brune et al. (1981) administered 0.15 mg/kg benzo[a]pyrene in the diet every 9th day or
5 days/week in a 1.5% caffeine solution until rats were moribund or dead. A single slope factor
estimate of 11.7 per mg/kg-day, based on a linearized multistage model applied to the combined
incidence of forestomach, esophageal, and laryngeal tumors, was derived. In the Neal and
Rigdon (1967) bioassay, mice administered benzo[a]pyrene in the diet at concentrations ranging
from 1 to 250 ppm for up to 197 days developed significantly increased incidences of
forestomach tumors. This study utilized mixed sex dose groups with mice from 3 weeks to 6
months old at the start of dosing. This study did not include concurrent controls. This
necessitated the use of historical controls (from SWR/J mice) for the incidence of forestomach
tumors from a study by (Rabstein et al., 1973). Three modeling procedures were used to derive
risk estimates from these data. For one risk estimate, Clement Associates (1990) fit a two-stage
response model, based on exposure-dependent changes in both transition rates and growth rates
of preneoplastic cells, to derive a value of 5.9 per mg/kg-day. In a U.S. EPA report (1991b), a
value of 9.0 per mg/kg-day, derived by linear extrapolation from the 10% response point to the
242
DRAFT - DO NOT CITE OR QUOTE
-------
1 background of an empirically fitted dose-response curve, was identified. Finally, using a
2 Weibull-type model to reflect less-than-lifetime exposure to benzo[a]pyrene, the same U.S. EPA
3 report (1991b) derived an upper-bound slope factor estimate of 4.5 per mg/kg-day. Since the
4 variance for the four slope factor estimates was low, and in order to consider all of the available
5 data, the geometric mean of these four estimates, 7.3 per mg/kg-day, was recommended as the
6 oral slope factor.
7 An inhalation unit risk and dermal slope factor were not previously available on IRIS.
8
9
243
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
6. MAJOR CONCLUSIONS IN Till CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE
6.1. HUMAN HAZARD POTENTIAL
benzo[a]pyrene is a five-ring nonsubstituted PAH, which is produced through natural and
anthropogenic processes involving the incomplete combustion or pyrrolysis of carbon-containing
materials. Benzo[a]pyrene exists in the environment in complex mixtures, which may consist of
numerous PAHs, including heterocyclic and nonheterocyclic forms as well as aza arenes and
nitro-substituted PAHs. The magnitude of human exposure to benzo[a]pyrene and other PAHs
depends on several factors related to lifestyle (e.g., diet, tobacco smoking), occupation, and
living conditions (e.g., urban versus rural setting, domestic heating, and cooking methods).
There are limited data establishing associations between increased risk for noncancer
health effects in humans and exposure to benzo[a]pyrene. Several epidemiology studies have
reported associations between adverse birth outcomes including reduced birth weight, postnatal
body weight, and head circumference with internal biomarkers of exposure to benzo[a]pyrene
(BPDE-DNA adducts) via exposure to complex PAH mixtures (Tang et al., 2008, 2006; Perera et
al., 2005a, b). However, extrapolations from these studies are complicated by the concomitant
exposure to multiple PAHs and other components in the mixture.
There is evidence of human carcinogenicity for several PAH mixtures containing
benzo[a]pyrene, such as soot, coal tars, coal-tar pitch, mineral oils, and shale oils (IARC, 2010;
Baan et al., 2009; Straif et al., 2005). There is also evidence of carcinogenicity in occupations
involving exposure to PAH mixtures containing benzo[a]pyrene, such as aluminum production,
chimney sweeping, coal gasification, coal-tar distillation, coke production, iron and steel
founding, and paving and roofing with coal tar pitch (IARC, 2010; Baan et al., 2009; Straif et al.,
2005). Benzo[a]pyrene is also a notable constituent of tobacco smoke (IARC 2004). An
increasing number of studies report exposure biomarkers such as benzo[a]pyrene- or PAH-DNA
adducts, and several cohort studies demonstrate a positive exposure-response relationship with
cumulative PAH exposure using benzo[a]pyrene as an indicator substance. Because
benzo[a]pyrene is only one of many PAHs that could contribute to these observed increases in
cancer, the epidemiologic studies provide credible but limited support for a causative role of
benzo[a]pyrene in human cancer. Studies in multiple species of laboratory animals indicate that
benzo[a]pyrene is carcinogenic by all routes of exposure.
6.2. DOSE RESPONSE
6.2.1. RfD
Limited human data are available to inform noncancer health effects following chronic
oral exposure to benzo[a]pyrene. Animal studies reporting effects of benzo[a]pyrene include
244
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
several chronic cancer bioassays (which report limited noncancer endpoints; Kroese et al. 2001;
Beland and Culp, 1998; Culp et al., 1998), several subchronic studies (Knuckles et al., 2001; De
Jong et al., 1999), developmental toxicity studies (Kristensen et al., 1995; MacKenzie and
Angevine, 1981) and several reproductive toxicity studies (Mohamed et al., 2010; Xu et al.,
2010; Zheng et al., 2010).
In consideration of the available studies reporting low-dose effects of chronic and
subchronic oral exposure to benzo(a)pyrene in animals, the Xu et al., (2010) study was chosen as
the principal study. This study identified biologically and statistically significant decrements in
ovary weight, number of primordial follicles, estrogen levels, and estrus cyclicity. These
reductions in female reproductive endpoints observed in rats are supported by a large database of
animal studies indicating that benzo[a]pyrene, administered by multiple routes of exposure, is
ovotoxic with effects including decreased ovary weight, decreased primordial follicles, and
reduced fertility (Mattison et al., 1980; MacKenzie and Angevine, 1981; Swartz and Mattison
1985; Miller et al., 1992; Kristensen et al.,1995; Borman et al, 2000; Archibong et al., 2002).
Additionally, studies indicate that exposure to complex mixtures of PAHs, such as through
cigarette smoke, is associated with measures of decreased fertility in humans (El Nemr et al.,
1998; Neal et al., 2005). Specific associations have also been made between infertility and
increased levels of benzo[a]pyrene in follicular fluid in women undergoing in vitro fertilization
(Neal et al., 2008).
The RfD of 0.0005 mg/kg-day (0.5 |ig/kg-day) was derived using a BMDLisd of
1.5 mg/kg-day for reduced ovary weight in SD rats exposed to benzo[a]pyrene via gavage for 60
days (Xu et al., 2010). To derive the RfD, this POD was divided by a total UF of 3000 (factors
of 10 for animal-to-human extrapolation, human interindividual variability in susceptibility, and
subchronic to chronic extrapolation, and 3 for database deficiencies). The default animal-to-
human extrapolation and human variability factors were applied because of the lack of
quantitative information to assess toxicokinetic or toxicodynamic differences between animals
and humans and the range of susceptibilities in human populations. A subchronic to chronic
extrapolation factor was applied because the POD was chosen from a study with a less than
lifetime exposure duration. In addition, a database uncertainty factor of 3 was applied to account
for deficiencies in the benzo[a]pyrene toxicity database, primarily the lack of a standard
multigenerational reproductive study and the lack of a neurodevelopmental study.
The overall confidence in the RfD is low-to-medium. Confidence in the principal study
(Xu et al., 2010) is medium. The design, conduct, and reporting of this subchronic toxicity study
were adequate; however, the number of dose groups and number of animals per group were low.
Confidence in the database is low-to-medium primarily due to the lack of a multigeneration
reproductive toxicity study (with exposure from pre-mating to sexual maturity) and the lack of a
neurodevelopmental study. Reflecting medium confidence in the principal study and low-to-
medium confidence in the database, confidence in the RfD is low-to-medium.
245
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
6.2.2. RfC
The only chronic inhalation study available for benzo[a]pyrene was designed as a cancer
bioassay and did not report noncancer endpoints (Thyssen et al. 1981). However, several
repeated dose reproductive and developmental toxicity studies are available in which effects on
fetal survival and the male reproductive system have been observed following inhalation
exposure.
Archibong et al. (2002) was selected as the principal study as it observed biologically
significant effects in F344 rats at the lowest dose tested by the inhalation route. This study
indicates that the developing fetus is a sensitive target following inhalation exposure to
"3
benzo[a]pyrene. Exposure to benzo[a]pyrene at 25 |ig/m on GDs 11-20 caused biologically
significant reductions in fetal survival and body weight decreases in the surviving pups. The
observed decrease in pup weight and fetal survival were selected as critical effects as they are the
most sensitive noncancer effects observed following inhalation exposure to benzo[a]pyrene.
Additional support for this endpoint can be found from an oral study of benzo[a]pyrene in mice
which observed decreased survival of litters, decreased pup weight, and decreased reproductive
organ weight following in utero exposure to benzo[a]pyrene on GD 7-16 (MacKenzie and
Angevine, 1981). Though only a few studies exist that evaluate benzo[a]pyrene by the
inhalation route, the oral studies support the reproductive and developmental effects observed in
the available inhalation studies.
The RfC of 5 x 10"6 mg/m3-day was derived using a LOAELadj[hec] of 4.6 |ig/m3-day for
decrease in pup weight and fetal survival in F344 rats exposed to benzo[a]pyrene aerosols on
GDs 11-20 (Archibong et al., 2002). To derive the RfC, the POD was divided by a total UF of
1000 (factors of 3 for animal-to-human extrapolation, 10 for human interindividual variability in
susceptibility and LOAEL-to-NOAEL extrapolation, and 3 for database deficiencies). The
default animal-to-human extrapolation and human variability factors were applied because of the
lack of quantitative information to assess toxicodynamic differences between animals and
humans, whereas the toxicokinetic component is addressed by the determination of a HEC as
described in the RfC methodology (U.S. EPA, 1994b). The default human variability factor was
applied because of the lack of information regarding the range of susceptibilities in human
populations. A LOAEL-to-NOAEL extrapolation factor was applied because a NOAEL was not
identified for decreased fetal survival observed by Archibong et al (2002). In addition, a
database uncertainty factor of 3 was applied to account for deficiencies in the benzo[a]pyrene
toxicity database, primarily the lack of a standard multigenerational reproductive study.
The overall confidence in the RfC is low-to-medium. Confidence in the principal study
(Archibong et al., 2002) is medium. The conduct, and reporting of this developmental dietary
study were adequate, however, a NOAEL was not identified. Confidence in the database is low-
to-medium due to the lack of a multigeneration reproductive toxicity study, the lack of studies on
246
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
immune endpoints, and the lack of information regarding subchronic and chronic inhalation
exposure. Reflecting medium confidence in the principal study and low-to-medium confidence
in the database, confidence in the RfC is low-to-medium.
6.2.3. Cancer
Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
benzo[a]pyrene is "carcinogenic to humans" based on evidence of carcinogenicity in humans
exposed to different PAH mixtures containing benzo[a]pyrene, extensive and consistent evidence
of carcinogenicity in laboratory animals exposed to benzo[a]pyrene via several routes of
administration, and strong evidence that mechanisms of carcinogenesis in laboratory animals
also occur in humans exposed to PAH mixtures containing benzo[a]pyrene. Strong evidence
links the metabolism of benzo[a]pyrene to DNA-reactive agents with key mutational events in
genes that can lead to tumor initiation. Specifically, the metabolic activation of benzo[a]pyrene
occurs in human tissues, and associations have been made between spectra of mutations in the
p53 tumor suppressor gene or ras oncogenes induced by benzo[a]pyrene metabolites and the
spectra of mutations in these genes in tumor tissue from benzo[a]pyrene-exposed animals and
humans.
Several lines of evidence relating to mutagenicity and tumor initiation are available for
benzo[a]pyrene including: in vitro evidence of DNA adducts, mutations, cytogenetic damage,
and primary DNA damage; in vivo DNA adducts, gene mutations, cytogenetic damage, and other
measures of primary DNA damage; detection of DNA adducts in target tissue in vivo; and
detection of cancer-relevant oncogene/tumor suppressor gene mutations in target tissue in vivo.
Taken together, these data provide support for a mutagenic MOA for benzo[a]pyrene-induced
cancer. Because benzo[a]pyrene is expected to cause cancer via a mutagenic MOA, a linear
approach to low-dose extrapolation was used in the derivation of the cancer risk estimates.
In the absence of appropriate benzo[a]pyrene-specific data to adjust cancer risk values for
early life exposure, ADAFs combined with age-specific exposure estimates should be applied to
the cancer risk values (oral slope factor, inhalation unit risk, and dermal slope factor) when
assessing cancer risks for individuals exposed during early life periods, as per U.S. EPA (2005b)
Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens.
6.2.3.1. Cancer—Oral
Lifetime oral exposure to benzo[a]pyrene has been associated with forestomach, liver,
oral cavity, jejunum or duodenum, and auditory canal tumors in male and female Wistar rats
(Kroese et al., 2001), with forestomach tumors in male and female Sprague-Dawley rats (Brune
et al., 1981), and with forestomach, esophagus, tongue, and larynx tumors in female B6C3F1
mice (male mice were not tested); (Beland and Culp, 1998; Culp et al., 1998). Less-than-lifetime
oral exposure to benzo[a]pyrene is also associated with forestomach tumors in more than 10
247
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
additional bioassays with several strains of mice (see Section 4.2.1.2.). Both the rat bioassay by
Kroese et al. (2001) and the mouse bioassay by Beland and Culp (1998) included histological
examinations for tumors, three exposure levels and controls, and 50 animals per dose group. The
chronic studies by Kroese et al. and Beland and Culp studies were therefore selected for dose-
response analysis.
EPA used the multistage-Weibull model in the derivation of the oral slope factor because
it incorporates the time at which death-with-tumor occurred and can account for differences in
mortality observed between the exposure groups in the rat bioassay. Using linear extrapolation
from the BMDLio, human equivalent oral slope factors were derived for each gender/tumor site
combination (slope factor = 0.1/BMDLio). The oral slope factor of 1 per mg/kg-day is based on
the tumor response in the alimentary tract (forestomach, esophagus, tongue, larynx) of female
B6C3Fi mice exposed to benzo[a]pyrene in the diet for 2 years (Beland and Culp, 1998). The
slope factor was derived by linear extrapolation from a human equivalent BMDLio of 0.07
mg/kg-day for forestomach, esophagus, tongue, and larynx papillomas or carcinomas. The
recommended slope factor was selected as the factor with the highest value among a range of
slope factors derived from tumor responses at several sites in the 2-year male and female Wistar
rat bioassay by Kroese et al. (2001) and the 2-year female B6C3Fi mouse bioassay by Beland
and Culp (1998).
6.2.3.2. Cancer—Inhalation
Inhalation exposure to benzo[a]pyrene was associated with squamous cell neoplasia in
the larynx, pharynx, trachea, esophagus, and forestomach, of male Syrian golden hamsters
exposed to benzo[a]pyrene condensed onto NaCl particles (Thyssen et al., 1981). Supportive
evidence for the carcinogenicity of inhaled benzo[a]pyrene comes from 10 additional studies
with hamsters exposed to benzo[a]pyrene via intratracheal instillation (see Section 4.2.2.2 for
references). The Thyssen et al. (1981) bioassay represents the best available data for describing
dose-response relationships for cancer from inhaled benzo[a]pyrene.
A time-to-tumor dose-response model was fit to the time-weighted average exposure
concentrations and the individual animal occurrence data for tumors in the larynx, pharynx,
trachea, esophagus, and forestomach. The inhalation unit risk of 5 x 10"4 per jug/m3 was
"3
calculated by linear extrapolation (slope factor = 0.1/BMDLio) from a BMDLio of 0.20 mg/m
for the occurrence of upper respiratory and upper digestive tract tumors in male hamsters
chronically exposed by inhalation to benzo[a]pyrene (Thyssen et al., 1981).
6.2.3.3. Cancer—Dermal
Skin cancer in humans has been documented to result from occupational exposure to
complex mixtures of PAHs including benzo[a]pyrene such as coal tar, coal tar pitches, non-
refined mineral oils, shale oils and soot (Boffetta et al., 1997; WHO, 1998; ATSDR, 1995). No
248
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
human studies of exposure to benzo[a]pyrene alone are known to exist. In animal models,
numerous dose response studies have demonstrated the increased incidence of skin tumors with
increasing dermal exposure of benzo[a]pyrene, in all species tested (mice, rabbits, rats, and
guinea pigs), though most benzo[a]pyrene chronic dermal bioassays have been conducted in
mice. This analysis focuses on chronic carcinogenicity bioassays in several strains of mice
demonstrating increasing incidence of benign and malignant skin tumors following repeated
dermal exposure to benzo[a]pyrene for the animals' lifetime.
As with the oral and inhalation benzo[a]pyrene carcinogenicity data (see Sections 5.4.1.3
and 5.4.2.3), benzo[a]pyrene's dermal exposure carcinogenicity data were generally
characterized by earlier occurrence of tumors with increasing exposure and increased mortality
with increasing exposure level. Each of the dermal data sets was modeled using the multistage
model, incorporating adjustments for early mortality, when data were available, prior to
modeling.
The POD of 0.0066 |ig/day, based on the tumor response in C3H/HeJ male mice (Sivak
et al., 1997), is recommended for developing a human dermal slope factor because it is the
highest POD among studies with low observed tumor response to benzo[a]pyrene (20%).
Following the modeling, this POD from the Sivak et al. (1997) dataset in male C3H/HeJ mice
was adjusted by allometric scaling. The dermal slope factor of 0.005 per jig/day was calculated
by linear extrapolation (slope factor = 0. 1/BMDLio-hed) from the human equivalent POD (19.7
|ig/day) for the occurrence of skin tumors in male mice chronically exposed dermally to
benzo[a]pyrene (Sivak et al., 1997).
This dermal slope factor has been calculated based on the risk of skin tumors in mice
following dermal exposure to benzo[a]pyrene. As this slope factor has been developed for a
local effect, it is not intended to estimate systemic risk of cancer following dermal absorption of
benzo[a]pyrene into the systemic circulation.
249
DRAFT - DO NOT CITE OR QUOTE
-------
7. REFERENCES
Agarwal, R; Medrano, EE; Khan, IU; et al. (1991) Metabolism of benzo[a]pyrene by human melanocytes in culture.
Carcinogenesis 12(10): 1963—1966.
Aklillu, E; Ovrebo, S; Botnen, IV; et al. (2005) Characterization of common CYP1B1 variants with different
capacity for benzo[a]pyrene-7,8-dihydrodiol epoxide formation from benzo[a]pyrene. Cancer Res 65(12):5105—
5111.
Albert, RE; Miller, ML; Cody, T; et al. (1991) Benzo[a]pyrene-induced skin damage and tumor promotion in the
mouse. Carcinogenesis 12(7): 1273-1280.
Alexandrie, A-K; Warholm, M; Carstensen, U; et al. (2000) CYP1A1 and GSTM1 polymorphisms affect urinary 1-
hyroxypyrene levels after PAH exposure. Carcinogenesis 21(4):699-676.
Alexandrie, AK; Nyberg, F; Warholm, M; et al. (2004) Influence of CYP1A1, GSTM1, GSTT1, and NQOl
genotypes and cumulative smoking dose on lung cancer risk in a Swedish population. Cancer Epidemiol
Biomarkers Prev 13(6):908-914.
Alexandrov, K; Rojas-Moreno, M. (1990) In vivo DNA adduct formation by benzo(a)pyrene in mouse and rat
epidermal and dermal fibroblasts after topical application of an initiating dose of benzo(a)pyrene. Arch
Geschwulstforsch 60(5):329-340.
Alexandrov, K; Rojas, M; Goldberg, M; et al. (1990) A new sensitive fluorometric assay for the metabolism of (±)-
7,8-dihydroxy-7,8-dihydrobenzo[a]pyrene by human hair follicles. Carcinogenesis 11:2157-2161.
Alexandrov, K; Rojas, M; Geneste, O; et al. (1992) An improved fluorometric assay for dosimetry of
benzo(a)pyrene diol-epoxide-DNA adducts in smokers' lung: comparisons with total bulky adducts and aryl
hydrocarbon hydroxylase activity. Cancer Res 52(22):6248-6253.
Alexandrov, K; Cascorbi, I; Rojas, M; et al. (2002) CYP1 Al and GSTM1 genotypes affect benzo[a]pyrene DNA
adducts in smokers' lung: comparison with aromatic/hydrophobic adduct formation. Carcinogenesis 23(12): 1969-
1977.
Amos, CI; Xu, W; Spitz, MR. (1999) Is there a genetic basis for lung cancer susceptibility? Recent Results Cancer
Res 151:3-12.
Anderson, KE; Kadlubar, FF; Kulldorff, M; et al. (2005) Dietary intake of heterocyclic amines and benzo(a)pyrene:
associations with pancreatic cancer. Cancer Epidemiol Biomarkers Prev 14(9):2261-2265.
Andersson, K; Levin, JO; Nilsson, CA. (1983) Sampling and analysis of particulate and gaseous polycyclic aromatic
hydrocarbons from coal tar sources in the working environment. Chemosphere 12(2): 197-207.
Andrysik, Z; Vondracek, J; Machala, M; et al. (2007) The aryl hydrocarbon receptor-dependent deregulation of cell
cycle control induced by polycyclic aromatic hydrocarbons in rat liver epithelial cells. Mutat Res 615(1 -2):87—97.
Archibong, AE; Inyang, F; Ramesh, A; et al. (2002) Alteration of pregnancy related hormones and fetal survival in
F-344 rats exposed by inhalation to benzo(a)pyrene. Reprod Toxicol 16(6):801-808.
Archibong, AE; Ramesh, A; Niaz, MS; et al. (2008) Effects of benzo(a)pyrene on intra-testicular function in F-344
rats. Int J Environ Res Public Health 5(l):32-40.
Ariese, F; Verkaik, M; Hoornweg, GP; et al. (1994) Trace analysis of 3-hydroxy benzo[a]pyrene in urine for the
biomonitoring of human exposure to polycyclic aromatic hydrocarbons. J Anal Toxicol 18(4): 195-204.
250
DRAFT - DO NOT CITE OR QUOTE
-------
Arlt, VM; Frei, E; Schmeiser, HH. (2007) ECNIS-sponsored workshop on biomarkers of exposure and cancer risk:
DNA damage and DNA adduct detection and 6th GUM-32P-postlabelling workshop, German Cancer Research
Center, Heidelberg, Germany, 29-30 September 2006. Mutagenesis 22(l):83-88.
Armstrong, B; Tremblay, C; Baris, D; et al. (1994) Lung cancer mortality and polynuclear aromatic hydrocarbons: a
case-cohort study of aluminum production workers in Arvida, Quebec, Canada. Am J Epidemiol 139(3):250-262.
Armstrong, B; Hutchinson, E; Unwin, J; et al. (2004) Lung cancer risk after exposure to polycyclic aromatic
hydrocarbons: a review and meta-analysis. Environ Health Perspect 112(9):970-978.
Arnould, JP; Pfohl-Leszkowicz, A; Bach, V; et al. (1999) Biological monitoring exposure of workers from plant
producing carbon electrodes: quantification of benzo[a]pyrene DNA-adducts in leukocytes, by a 32P-postlabelling
method and an immunoassay. Hum Exp Toxicol 18(5):314—321.
Arnould, JP; Kubiak, R; Belowski, J; et al. (2000) Detection of benzo[a]pyrene-DNA adducts in leukocytes of coke
oven workers. Pathol Biol (Paris) 48(6):548-553.
Assennato, G; Ferri, GM; Foa, V; et al. (1993a) Correlation between PAH airborne concentration and PAH-DNA
adducts levels in coke-oven workers. Int Arch Occup Environ Health 65(1 Suppl):S143-S145.
Assennato, G; Ferri, GM; Tockman, MS; et al. (1993b) Biomarkers of carcinogen exposure and cancer risk in a coke
plant. Environ Health Perspect 99:237-239.
ATSDR (Agency for Toxic Substances and Disease Registry). (1995) Toxicological profile for polycyclic aromatic
hydrocarbons. Atlanta, GA: Agency for Toxic Substances and Disease Registry.
Autrup, H; Seremet, T. (1986) Excretion of benzo[a]pyrene-Gua adduct in the urine of benzo[a]pyrene-treated rats.
Chem Biol Interact 60(2):217-226.
Autrup, H; Harris, CC; Stoner, GD; et al. (1978) Metabolism of [3H]benzo[a]pyrene by cultured human bronchus
and cultured human pulmonary alveolar macrophages. Lab Invest 38(3):217-224.
Autrup, H; Wefald, FC; Jeffrey, AM; et al. (1980) Metabolism of benzo[a]pyrene by cultured tracheobronchial
tissues from mice, rats, hamsters, bovines and humans. Int J Cancer 25(2):293-300.
Balu, N; Padgett, WT; Lambert, GR; et al. (2004) Identification and characterization of novel stable deoxy guano sine
and deoxyadenosine adducts of benzo[a]pyrene-7,8-quinone from reactions at physiological pH. Chem Res Toxicol
17(6):827-838.
Bao, H; Vepakomma, M; Sarkar, MA. (2002) Benzo(a)pyrene exposure induces CYP1 Al activity and expression in
human endometrial cells. J Steroid Biochem Mol Biol 81 (1):37—45.
Barry, G; Cook, JW; Haslewood, AD; et al. (1935) The production of cancer by pure hydrocarbons-Part III. Proc R
Soc London 117:318-351.
Bartsch, H; Rojas, M; Nair, U; et al. (1999) Genetic cancer susceptibility and DNA adducts: studies in smokers,
tobacco chewers, and coke oven workers. Cancer Detect Prev 23(6):445-453.
Becher, G; Bjorseth, A. (1983) Determination of exposure to polycyclic aromatic hydrocarbons by analysis of
human urine. Cancer Lett 17(3):301—311.
Beland, F; Culp, S. (1998) Chronic bioassay of two composite samples from selected manufactured gas plant waste
sites. Jefferson, AK: Research, NCfT; Technical Report 6722.02 (unpublished).
Benjamin, H; Storkson, J; Talias, PG; et al. (1988) Reduction of benzo[a]pyrene-induced forestomach neoplasms in
mice given nitrite and dietary soy sauce. Food Chem Toxicol 26(8):671-678.
251
DRAFT - DO NOT CITE OR QUOTE
-------
Bentsen-Farmen, RK; Botnen, I; Eilertsen, E; et al. (1999a) The effect of CYP1A1 induction on the formation of
benzo(a)pyrene adducts in liver and lung DNA and plasma albumin in rats exposed to benzo(a)pyrene: Adduct
quantitation by immunoassay and an HPLC method. Biomarkers 4(l):37-47.
Bentsen-Farmen, RK; Botnen, IV; Noto, H; et al. (1999b) Detection of polycyclic aromatic hydrocarbon metabolites
by high-pressure liquid chromatography after purification on immunoaffinity columns in urine from occupationally
exposed workers. Int Arch Occup Environ Health 72(3): 161-168.
Berenblum, I; Haran, N. (1955) The influence of croton oil and of polyethylene glycol-400 on carcinogenesis in the
forestomach of the mouse. Cancer Res 15(8):510—516.
Bevan, DR; Sadler, VM. (1992) Quinol diglucuronides are predominant conjugated metabolites found in bile of rats
following intratracheal instillation of benzo[a]pyrene. Carcinogenesis 13(3):403-407.
Bevan, DR; Ulman, MR. (1991) Examination of factors that may influence disposition of benzo[a]pyrene in vivo:
vehicles and asbestos. Cancer Lett 57(2): 173-179.
Bevan, DR; Weyand, EH. (1988) Compartmental analysis of the disposition of benzo[a]pyrene in rats.
Carcinogenesis 9(11):2027-2032.
Biancifiori, C; Caschera, F; Giornelli-Santulli, F; et al. (1967) The action of oestrone and four chemical carcinogens
in intact and ovariectomized BALB/c/Cb/Se mice. Br J Cancer 21:452^159.
Bingham, E; Falk, HL. (1969) Environmental carcinogens. The modifying effect of cocarcinogens on the threshold
response. Arch Environ Health 19(6):779-783.
Bjerseth, A; Bjorseth, 0; Fjeldstad, PE. (1978a) Polycyclic aromatic hydrocarbons in the work atmosphere. I.
Determination in an aluminum reduction plant. Scand J Work Environ Health 4(3):212-223.
Bjerseth, A; Bjorseth, 0; Fjeldstad, PE. (1978b) Polycyclic aromatic hydrocarbons in the work atmosphere. II.
Determination in a coke plant. Scand J Work Environ Health 4(3):224-236.
Blaha, L; Kapplova, P; Vondracek, J; et al. (2002) Inhibition of gap-junctional intercellular communication by
environmentally occurring polycyclic aromatic hydrocarbons. Toxicol Sci 65(1 ):43—51.
Boerrigter, ME. (1999) Treatment of lacZ plasmid-based transgenic mice with benzo[a]pyrene: measurement of
DNA adduct levels, mutant frequencies, and mutant spectra. Environ Mol Mutagen 34(2-3): 140-147.
Boffetta, P; Jourenkova, N; Gustavsson, P. (1997) Cancer risk from occupational and environmental exposure to
polycyclic aromatic hydrocarbons. Cancer Causes Control 8(3):444-472.
Bol, SA; van Steeg, H; Jansen, JG; et al. (1998) Elevated frequencies of benzo(a)pyrene-induced Hprt mutations in
internal tissue of XPA-deficient mice. Cancer Res 58(13):2850-2856.
Bolton, JL; Trush, MA; Penning, TM; et al. (2000) Role of quinones in toxicology. Chem Res Toxicol 13(3): 135—
160.
Bompart, G; Clamens, S. (1990) In vitro and in vivo benzo(a)pyrene metabolism in rat liver and lung microsomes:
effect of sodium selenite. J Toxicol Clin Exp 10(1):3—14.
Bond, JA; Harkema, JR; Russell, VI. (1988) Regional distribution of xenobiotic metabolizing enzymes in
respiratory airways of dogs. Drug Metab Dispos 16(1): 116-124.
Borman, SM; Christian, PJ; Sipes, IG; Hoyer, PB. (2000) Ovotoxicity in female Fischer rats and B6 mice induced
by low-dose exposure to three polycyclic aromatic hydrocarbons: comparison through calculation of an ovotoxic
index. Toxicol App Pharmacol 167;191-198.
Bosetti, C; Boffetta, P; La Vecchia, C. (2007) Occupational exposures to polycyclic aromatic hydrocarbons, and
respiratory and urinary tract cancers: a quantitative review to 2005. Ann Oncol 18:431^46.
252
DRAFT - DO NOT CITE OR QUOTE
-------
Bostrom, CE; Gerde, P; Hanberg, A; et al. (2002) Cancer risk assessment, indicators, and guidelines for polycyclic
aromatic hydrocarbons in the ambient air. Environ Health Perspect 110(Supp 3):451-488.
Bowman, ED; Rothman, N; Hackl, C; et al. (1997) Interindividual variation in the levels of certain urinary
polycyclic aromatic hydrocarbon metabolites following medicinal exposure to coal tar ointment. Biomarkers 2:321—
327.
Brooks, RA; Gooderham, NJ; Edwards, RJ; et al. (1999) The mutagenicity of benzo[a]pyrene in mouse small
intestine. Carcinogenesis 20(1): 109-114.
Bruce, ED; Austenrieth, RL; Burghardt, RC; et al. (2008) Using quantitative structure-activity relationships (QSAR)
to predict toxic endpoints for polycyclic aromatic hydrocarbons (PAH). J Toxicol Environ Health Part A 71:1073—
1084.
Brune, H; Deutsch-Wenzel, RP; Habs, M; et al. (1981) Investigation of the tumorigenic response to benzo(a)pyrene
in aqueous caffeine solution applied orally to Sprague-Dawley rats. J Cancer Res Clin Oncol 102(2): 153—157.
Bryan, WR; Shimkin, MB. (1943) Quantitative analysis of dose-response data obtained with three carcinogenic
hydrocarbons in strain C3H male mice. J Natl Cancer Inst 3:503-531.
Buckley, TJ; Waldman, JM; Dhara, R; et al. (1995) An assessment of a urinary biomarker for total human
environmental exposure to benzo[a]pyrene. Int Arch Occup Environ Health 67(4):257-266.
Buening, MK; Wislocki, PG; Levin, W; et al. (1978) Tumorigenicity of the optical enantiomers of the
diastereomeric benzo[a]pyrene 7,8-diol-9,10-epoxides in newborn mice: exceptional activity of (+)-7beta,8alpha-
dihydroxy-9alpha,10alpha-epoxy-7,8,9,10-tetrahydrobenzo[a Jpyrene. Proc Natl Acad Sci USA 75(11):5358—5361.
Bulay, OM; Wattenberg, LW. (1971) Carcinogenic effects of polycyclic hydrocarbon carcinogen administration to
mice during pregnancy on the progeny. J Natl Cancer Inst 46(2):397-402.
Burdick, AD; Davis, JW; Liu, KJ; et al. (2003) Benzo(a)pyrene quinones increase cell proliferation, generate
reactive oxygen species, and transactivate the epidermal growth factor receptor in breast epithelial cells. Cancer Res
63(22):7825-7833.
Burdick, AD; Ivnitski-Steele, ID; Lauer, FT; et al. (2006) PYK2 mediates anti-apoptotic AKT signaling in response
to benzo[a]pyrene diol epoxide in mammary epithelial cells. Carcinogenesis 27(11 ):2331—2340.
Burns-Naas, LA; Hastings, KL; Ladies, GS; et al. (2008) What's so special about the developing immune system?
Int J Toxicol 27(2):223-254.
Burstyn, I; Kromhout, H; Partanen, T; et al. (2005) Polycyclic aromatic hydrocarbons and fatal ischemic heart
disease. Epidemiology 16(6):744-750.
Burstyn, I; Kromhout, H; Johansen, C; et al. (2007) Bladder cancer incidence and exposure to polycyclic aromatic
hydrocarbons among asphalt pavers. Occup Environ Med 64(8):520-526.
Busby, WF. Jr.; Goldman, ME; Newberne, PM; et al. (1984) Tumorigenicity of fluoranthene in a newborn mouse
lung adenoma bioassay. Carcinogenesis 5(10):1311-1316.
Busby, WF, Jr.; Stevens, EK; Martin, CN; et al. (1989) Comparative lung tumorigenicity of parent and mononitro-
polynuclear aromatic hydrocarbons in the BLU:Ha newborn mouse assay. Toxicol Appl Pharmacol 99(3):555-563.
Byczkowski, JZ; Gessner, T. (1989) Effects of inhibition of NADPH: cytochrome P-450 reductase on
benzo(a)pyrene metabolism in mouse liver microsomes. Int J Biochem 21(5):525-529.
Byczkowski, JZ; Kulkarni, AP. (1990) Lipid peroxidation-coupled co-oxygenation of benzo(a)pyrene and
benzo(a)pyrene-7,8-dihydrodiol in human term placental microsomes. Placenta 11(1): 17-26.
253
DRAFT - DO NOT CITE OR QUOTE
-------
Byczkowski, JZ; Kulkarni, AP. (1992) Vanadium redox cycling, lipid peroxidation and co-oxygenation of
benzo(a)pyrene-7,8-dihydrodiol. Biochim Biophys Acta 1125(2): 134—141.
Carfi, M; Gennari, A; Malerba, I; et al. (2007) In vitro tests to evaluate immunotoxicity: a preliminary study.
Toxicology 229(1-2): 11-22.
Cavalieri, EL; Rogan, EG. (1992) The approach to understanding aromatic hydrocarbon carcinogenesis. The central
role of radical cations in metabolic activation. Pharmacol Ther 55(2): 183-199.
Cavalieri, EL; Rogan, EG. (1995) Central role of radical cations in metabolic activation of poly cyclic aromatic
hydrocarbons. Xenobiotica 25(7):677-688.
Cavalieri, E; Rogan, E; Toth, B; et al. (1981) Carcinogenicity of the environmental pollutants cyclopenteno-
[cd]pyrene and cyclopentano[cd]pyrene in mouse skin. Carcinogenesis 2(4):277-281.
Cavalieri, E; Rogan, E; Cremonesi, P; et al. (1988a) Tumorigenicity of 6-halogenated derivatives of benzo[a]pyrene
in mouse skin and rat mammary gland. J Cancer Res Clin Oncol 114:10-15.
Cavalieri, E; Rogan, E; Sinha, D. (1988b) Carcinogenicity of aromatic hydrocarbons directly applied to rat
mammary gland. J Cancer Res Clin Oncol 114:3-9.
Cavalieri, E; Rogan, E; Higginbotham, S; et al. (1988c) Tumor-initiating activity in mouse skin and carcinogenicity
in rat mammary gland of fluorinated derivatives of benzo[a]pyrene and 3-methylcholanthrene. J Cancer Res Clin
Oncol 114(1): 16-22.
Cavalieri, EL; Devanesan, PD; Rogan, EG. (1988d) Radical cations in the horseradish peroxidase and prostaglandin
H synthase mediated metabolism and binding of benzo[a]pyrene to deoxyribonucleic acid. Biochem Pharmacol
37(11):2183—2187.
Cavalieri, EL; Rogan, EG; Cremonesi, P; et al. (1988e) Radical cations as precursors in the metabolic formation of
quinones from benzo[a]pyrene and 6-fluorobenzo[a]pyrene. Fluoro substitution as a probe for one-electron oxidation
in aromatic substrates. Biochem Pharmacol 37(11):2173—2182.
Cavalieri, EL; Rogan, EG; Devanesan, PD; et al. (1990) Binding of benzo[a]pyrene to DNA by cytochrome P-450
catalyzed one-electron oxidation in rat liver microsomes and nuclei. Biochemistry 29(20):4820-4827.
Cavalieri, EL; Higginbotham, S; RamaKrishna, NV; et al. (1991) Comparative dose-response tumorigenicity studies
of dibenzo[alpha,l]pyrene versus 7,12-dimethylbenz[alpha]anthracene, benzo[alpha]pyrene and two
dibenzo[alpha,l]pyrene dihydrodiols in mouse skin and rat mammary gland. Carcinogenesis 12(10): 1939—1944.
Cavalieri, EL; Rogan, EG; Ramakrishna, NVS; et al. (1993) Mechanisms of benzo[a]pyrene and
7,12-dimethylbenz[a]-anthracene activation: qualitative aspects of the stable and depurination DNA adducts
obtained from radical cations and diol epoxides. In: Garrigues, P; Lamotte, M; eds. Polycyclic aromatic compounds.
Synthesis, properties, analytical measurements, occurrence and biological effects. Langhorne, PA: Gordon and
Breach Science Publishers; pp. 725-732.
Cavalieri, EL; Rogan, EG; Li, KM; et al. (2005) Identification and quantification of the depurinating DNA adducts
formed in mouse skin treated with dibenzo[a,l]pyrene (DB[a,l]P) or its metabolites and in rat mammary gland
treated with DB[a,l]P. Chem Res Toxicol 18(6):976-983.
Cavret, S; Laurent, C; Feidt, C; et al. (2003) Intestinal absorption of 14C from 14C-phenanthrene,
14C-benzo[a]pyrene and 14C-tetrachlorodibenzo-para-dioxin: approaches with the Caco-2 cell line and with portal
absorption measurements in growing pigs. Reprod, Nutr, Dev 43(2): 145-154.
Cenni, A; Sciarra, G; Sartorelli, P; et al. (1993) Environmental and biological monitoring of polycyclic aromatic
hydrocarbons (PAHs) in coke plants and other workplaces. Med Lav 84(5):379-386.
254
DRAFT - DO NOT CITE OR QUOTE
-------
Chakravarti, D; Pelling, JC; Cavalieri, EL; et al. (1995) Relating aromatic hydrocarbon-induced DNA adducts and
c-H-ras mutations in mouse skin papillomas: the role of apurinic sites. Proc Natl Acad Sci U S A 92(22): 10422-
10426.
Chakravarti, D; Mailander, PC; Cavalieri, EL; et al. (2000) Evidence that error-prone DNA repair converts
dibenzo[a,l]pyrene-induced depurinating lesions into mutations: formation, clonal proliferation and regression of
initiated cells carrying H-ras oncogene mutations in early preneoplasia. Mutat Res 456(1-2): 17—32.
Chanda, S; Erexson, G; Frost, D; et al. (2007) 26-Week dermal oncogenicity study evaluating pure trans-capsaicin
in Tg.AC hemizygous mice. Int J Toxicol 26:123-133.
Chang, RL; Wood, AW; Conney, AH; et al. (1987) Role of diaxial versus diequatorial hydroxyl groups in the
tumorigenic activity of a benzo[a]pyrene bay-region diol epoxide. Proc Natl Acad Sci U S A 84(23):8633-8636.
Chang, LW; Chang, YC; Ho, CC; et al. (2007) Increase of carcinogenic risk via enhancement of cyclooxygenase-2
expression and hydroxyestradiol accumulation in human lung cells as a result of interaction between benzo[a]pyrene
and 17-beta estradiol. Carcinogenesis 28(7): 1606-1612.
Charles, GD; Bartels, MJ; Zacharewski, TR; et al. (2000) Activity of benzo[a]pyrene and its hydroxylated
metabolites in an estrogen receptor-alpha reporter gene assay. Toxicol Sci 55(2):320-326.
Chen, S; Nguyen, N; Tamura, K; et al. (2003) The role of the Ah receptor and p38 in benzo[a]pyrene-
7,8-dihydrodiol and benzo[a]pyrene-7,8-dihydrodiol-9,10-epoxide-induced apoptosis. J Biol Chem 278(21):19526-
19533.
Cheng, L; Spitz, MR; Hong, WK; et al. (2000) Reduced expression levels of nucleotide excision repair genes in lung
cancer: a case-control analysis. Carcinogenesis 21(8): 1527-1530.
Cheng, YW; Tsai, HJ; Wu, JY; et al. (2007) A possible role for dihydrodiol dehydrogenase in the formation of
benzo[a]pyrene-DNA adducts in lung cancer cells and tumor tissues. Environ Mol Mutagen 48(1): 14-21.
Chiazze, L, Jr.; Watkins, DK; Amsel, J. (1991) Asphalt and risk of cancer in man. Br J Ind Med 48(8):538-542.
Chipman, JK; Bhave, NA; Hirom, PC; et al. (1982) Metabolism and excretion of benzo[a]pyrene in the rabbit.
Xenobiotica 12(6):397-404.
Choi, JY; Lee, KM; Cho, SH; et al. (2003) CYP2E1 and NQOl genotypes, smoking and bladder cancer.
Pharmacogenetics 13(6):349—355.
Chouroulinkov, I; Gentil, A; Guerin, M. (1967) Study of the carcinogenic activity of 9,10-dimethyl-benzanthracene
and of 3,4-benzopyrene given orally. Bull Cancer 54(l):67-78.
Chouroulinkov, I; Gentil, A; Grover, PL; et al. (1976) Tumour-initiating activities on mouse skin of dihydrodiols
derived from benzo[a]pyrene. Br J Cancer 34(5):523-532.
Chung, JY; Kim, JY; Kim, WR; et al. (2007) Abundance of aryl hydrocarbon receptor potentiates benzo[a]pyrene-
induced apoptosis in Hepalclc7 cells via CYP1 Al activation. Toxicology 235(l-2):62-72.
Clement Associates. (1990) Development of a dose-response model for inhaled B[a]P. Fairfax, VA: Clement
Associates.
Cogliano, VJ; Baan, RA; Straif, K; et al. (2008) Use of mechanistic data in IARC evaluations. Environ Mol
Mutagen 49(2): 100-109.
Cohen, GM; Haws, SM; Moore, BP; et al. (1976) Benzo(a)pyren-3-yl hydrogen sulphate, a major ethyl acetate-
extractable metabolite of benzo(a)pyrene in human, hamster and rat lung cultures. Biochem Pharmacol
25(23):2561-2570.
255
DRAFT - DO NOT CITE OR QUOTE
-------
Conney, AH; Chang, RL; Jerina, DM; et al. (1994) Studies on the metabolism of benzo[a]pyrene and dose-
dependent differences in the mutagenic profile of its ultimate carcinogenic metabolite. Drug Metab Rev
26(1-2): 125-163.
Cooper, AR; Moley, KH. (2008) Maternal tobacco use and its preimplantation effects on fertility: more reasons to
stop smoking. Semin Reprod Med. 26:204-212.
Creasia, DA; Poggenburg, JK, Jr.; Nettesheim, P. (1976) Elution of benzo[alpha]pyrene from carbon particles in the
respiratory tract of mice. J Toxicol Environ Health l(6):967-975.
Cresteil, T. (1998) Onset of xenobiotic metabolism in children: toxicological implications. Food Addit Contam
15(Suppl):45-51.
Crofton-Sleigh, C; Doherty, A; Ellard, S; et al. (1993) Micronucleus assays using cytochalasin-blocked MCL-5
cells, a proprietary human cell line expressing five human cytochromes P-450 and microsomal epoxide hydrolase.
Mutagenesis 8(4):363-372.
Culp, SJ; Beland, FA. (1994) Comparison of DNA adduct formation in mice fed coal tar or benzo[a]pyrene.
Carcinogenesis 15(2):247-252.
Culp, SJ; Gaylor, D; Sheldon, W; et al. (1996a) DNA adduct measurements in relation to small intestine and
forestomach tumor incidence during the chronic feeding of coal tar or benzo(a)pyrene to mice. Polycyclic Aromat
Compd 11:161-168.
Culp, SJ; Gaylor, DW; Sheldon, WG; et al. (1996b) Relationship between DNA adduct levels and tumor incidence
in mice fed coal tar or benzo[a]pyrene for two years. Proc Am Assoc Cancer Res 37:101.
Culp, SJ; Gaylor, DW; Sheldon, WG; et al. (1998) A comparison of the tumors induced by coal tar and
benzo[a]pyrene in a 2-year bioassay. Carcinogenesis 19(1): 117-124.
Culp, SJ; Warbritton, AR; Smith, BA; et al. (2000) DNA adduct measurements, cell proliferation and tumor
mutation induction in relation to tumor formation in B6C3F1 mice fed coal tar or benzo[a]pyrene. Carcinogenesis
21(7): 1433—1440.
Curfs, DM; Lutgens, E; Gijbels, MJ; et al. (2004) Chronic exposure to the carcinogenic compound benzo[a]pyrene
induces larger and phenotypically different atherosclerotic plaques in ApoE-knockout mice. Am J Pathol
164(1): 101—108.
Curfs, DM; Knaapen, AM; Pachen, DM; et al. (2005) Polycyclic aromatic hydrocarbons induce an inflammatory
atherosclerotic plaque phenotype irrespective of their DNA binding properties. FASEB J 19(10): 1290—1292.
Dahl, AR; Coslett, DS; Bond, JA; et al. (1985) Metabolism of benzo[a]pyrene on the nasal mucosa of Syrian
hamsters: comparison to metabolism by other extrahepatic tissues and possible role of nasally produced metabolites
in carcinogenesis. J Natl Cancer Inst 75(1): 135-139.
Davidson, GE; Dawson, GW. (1976) Chemically induced presumed somatic mutations in the mouse. Mutat Res
38(2): 151—154.
Davidson, GE; Dawson, GW. (1977) The induction of somatic mutations in mouse embryos by benzo(a)pyrene.
Arch Toxicol 38(1-2):99-l03.
Davies RF. (1969) Tumorigenicity of very fresh tobacco smoke condensate to mouse skin. Br J Cancer. 1969
23(4):858-60.
Dean, JH; Luster, MI; Boorman, GA; et al. (1983) Selective immunosuppression resulting from exposure to the
carcinogenic congener of benzopyrene in B6C3F1 mice. Clin Exp Immunol 52(l):199-206.
Dean, SW; Coates, A; Brooks, TM; et al. (1998) Benzo[a]pyrene site of contact mutagenicity in skin of Muta
mouse. Mutagenesis 13(5):515—518.
256
DRAFT - DO NOT CITE OR QUOTE
-------
De Flora, S; D'Agostini, F; Izzotti, A; et al. (1991) Prevention by N-acetylcysteine of benzo[a]pyrene clastogenicity
and DNA adducts in rats. Mutat Res 250(1-2): 87-93.
De Jong, WH; Kroese, ED; Vos, JG; et al. (1999) Detection of immunotoxicity of benzo[a]pyrene in a subacute
toxicity study after oral exposure in rats. Toxicol Sci 50:214-220.
DeMarini, DM; Landi, S; Tian, D; et al. (2001) Lung tumor KRAS and TP53 mutations in nonsmokers reflect
exposure to PAH-rich coal combustion emissions. Cancer Res 61(18):6679-6681.
Denissenko, MF; Pao, A; Tang, M; et al. (1996) Preferential formation of benzo[a]pyrene adducts at lung cancer
mutational hotspots in P53. Science 274:430-432.
Dennison, JE. (2005) Alkene and aromatic compounds. In: Reddy, MB; Yang, RSH; Clewell, HJ; et al.; eds.
Physiologically based pharmacokinetic modeling. Hoboken, NJ: John Wiley & Son, Inc.; pp. 103-104.
Dertinger, SD; Lantum, HB; Silverstone, AE; et al. (2000) Effect of 3'-methoxy-4'-nitroflavone on benzo[a]pyrene
toxicity. Aryl hydrocarbon receptor-dependent and -independent mechanisms. Biochem Pharmacol 60(2): 189-196.
Dertinger, SD; Nazarenko, DA; Silverstone, AE; et al. (2001) Aryl hydrocarbon receptor signaling plays a
significant role in mediating benzo[a]pyrene- and cigarette smoke condensate-induced cytogenetic damage in vivo.
Carcinogenesis 22(1): 171—177.
Deutsch, J; Vatsis, KP; Coon, MJ; et al. (1979) Catalytic activity and stereoselectivity of purified forms of rabbit
liver microsomal cytochrome P-450 in the oxygenation of the (-) and (+) enantiomers of trans-7,8-dihydroxy-
7,8-dihydrobenzo[alpha]pyrene. Mol Pharmacol 16(3): 1011-1018.
De Waal, EJ; Schuurman, HJ; Van Loveren, H; et al. (1997) Differential effects of 2,3,7,8-tetrachlorodibenzo-p-
dioxin, bis(tri-n-butyltin) oxide and cyclosporine on thymus histophysiology. Crit Rev Toxicol 27(4):381-430.
Diamond, L; Kruszewski, F; Aden, DP; et al. (1980) Metabolic activation of benzo[a]pyrene by a human hepatoma
cell line. Carcinogenesis 1 (10):871—875.
Dietert, RR. (2008) Developmental immunotoxicology (DIT): Windows of vulnerability, immune dysfunction and
safety assessment. J Immunotoxicol 5(4):401-412.
Dietert, RR; Etzel, RA; Chen, D; et al. (2000) Workshop to identify critical windows of exposure for children's
health: Immune and respiratory systems work group summary. Environ Health Perspect 108(Suppl 3):483-490.Divi,
RL; Beland, FA; Fu, PP; et al. (2002) Highly sensitive chemiluminescence immunoassay for benzo[a]pyrene-DNA
adducts: validation by comparison with other methods, and use in human biomonitoring. Carcinogenesis
23(12):2043-2049.
Doll, R. (1971) The age distribution of cancer: implication for models of carcinogenesis. J R Statistical Society
A134: 133-166.
Donovan, PJ; Smith, GT; Nardone, R. (2004) The mutagenic effects of 7,12-dimethylbenz[a]anthacene,
3-methylcholanthrene and benzo[a]pyrene to the developing Syrian hamster fetus measured by an in vivo/in vitro
mutation assay. Mutation Research 554(1-2): 111-120.
Dreij, K; Sundberg, K; Johansson, AS; et al. (2002) Catalytic activities of human alpha class glutathione transferases
toward carcinogenic dibenzo[a,l]pyrene diol epoxides. Chem Res Toxicol 15(6):825—831.
Eaton, DL; Bammler, TK. (1999) Concise review of the glutathione S-transferases and their significance to
toxicology. Toxicol Sci 49(2):156-164.
Edwards, SC; Jedrychowski, W; Butscher, M; et al. (2010). Prenatal exposure to airborne polycyclic aromatic
hydrocarbons and children's intelligence at 5 years of age in a prospective cohort study in Poland. Environ Health
Perspect. 118(9): 1326-31.
257
DRAFT - DO NOT CITE OR QUOTE
-------
El-Bayoumy, K. (1985) Effects of organoselenium compounds on induction of mouse forestomach tumors by
benzo(a)pyrene. Cancer Res 45:3631-3635.
El-Bayoumy, K; Hecht, SS; Hoffmann, D. (1982) Comparative tumor initiating activity on mouse skin of
6-nitrobenzo[a]pyrene, 6-nitrochrysene, 3-nitroperylene, 1-nitropyrene and their parent hydrocarbons. Cancer Lett
16(3):333—337.
Emmett, EA; Bingham, EM; Barkley, W. (1981) A carcinogenic bioassay of certain roofing materials. Am J Ind
Med 2(l):59-64.
Ewing, P; Blomgren, B; Ryrfeldt, A; et al. (2006) Increasing exposure levels cause an abrupt change in the
absorption and metabolism of acutely inhaled benzo(a)pyrene in the isolated, ventilated, and perfused lung of the rat.
Toxicol Sci 91(2):332-340.
Fang, JL; Lazarus, P. (2004) Correlation between the UDP-glucuronosyltransferase (UGT1 Al) TATAA box
polymorphism and carcinogen detoxification phenotype: significantly decreased glucuronidating activity against
benzo(a)pyrene-7,8-dihydrodiol(-) in liver microsomes from subjects with the UGT1 Al *28 variant. Cancer
Epidemiol Biomarkers Prev 13(1): 102—109.
Fedorenko, Z; Yansheva, N. (1967) Experimental reproduction of tumors of the antral part of the stomach in mice
by administration of various dose of 3,4-benzpyrene. Hyg Sanit 32(5): 168-173.
Feron, VJ; Kruysse, A. (1978) Effects of exposure to furfural vapour in hamsters simultaneously treated with
benzo[alpha] pyrene or diethylnitrosamine. Toxicology 11 (2): 127—144.
Feron, VJ; de Jong, D; Emmelot, P. (1973) Letter: Dose-response correlation for the induction of respiratory-tract
tumours in Syrian golden hamsters by intratracheal instillations of benzo(a)pyrene. Eur J Cancer 9(5):387-390.
Fertuck, KC; Matthews, JB; Zacharewski, TR. (2001) Hydroxylated benzo[a]pyrene metabolites are responsible for
in vitro estrogen receptor-mediated gene expression induced by benzo[a]pyrene, but do not elicit uterotrophic effects
in vivo. Toxicol Sci 59(2):231-240.
Field, E; Roe, F. (1965) Tumor promotion in the forestomach epithelium of mice by oral administration of citrus
oils. J Natl Cancer Inst 35(5):775-784.
Fine, JS; Silverstone, AE; Gasiewicz, TA. (1990) Impairment of prothymocyte activity by 2,3,7,8-
tetrachlorodibenzo-p-dioxin. J Immunol 144(4): 1169-1176.
Flowers, L; Bleczinski, WF; Burczynski, ME; et al. (1996) Disposition and biological activity of benzo[a]pyrene-
7,8-dione. A genotoxic metabolite generated by dihydrodiol dehydrogenase. Biochemistry 35(42): 13664-13672.
Flowers, L; Ohnishi, ST; Penning, TM. (1997) DNA strand scission by polycyclic aromatic hydrocarbon o-
quinones: role of reactive oxygen species, Cu(II)/Cu(I) redox cycling, and o-semiquinone anion radicals.
Biochemistry 36(28):8640-8648.
Flowers-Geary, L; Harvey, RG; Penning, TM. (1993) Cytotoxicity of polycyclic aromatic hydrocarbon o-quinones
in rat and human hepatoma cells. Chem Res Toxicol 6(3):252-260.
Flowers-Geary, L; Bleczinki, W; Harvey, RG; et al. (1996) Cytotoxicity and mutagenicity of polycyclic aromatic
hydrocarbon ortho-quinones produced by dihydrodiol dehydrogenase. Chem Biol Interact 99(l-3):55-72.
Fontana, RJ; Lown, KS; Paine, MF; et al. (1999) Effects of a chargrilled meat diet on expression of CYP3A,
CYP1 A, and P-glycoprotein levels in healthy volunteers. Gastroenterology 117(1):89—98.
Forster, K; Preuss, R; Rossbach, B; et al. (2008) 3-Hydroxybenzo[a]pyrene in the urine of workers with
occupational exposure to polycyclic aromatic hydrocarbons in different industries. Occup Environ Med 65(4):224-
229.
258
DRAFT - DO NOT CITE OR QUOTE
-------
Foth, H; Kahl, R; Kahl, GF. (1988) Pharmacokinetics of low doses of benzo[a]pyrene in the rat. Food Chem
Toxicol 26(1 ):45—51.
Fowler, P; Whitwell, J; Jeffrey, L; et al. (2010) Cadmium chloride, benzo[a]pyrene and cyclophosphamide tested in
the in vitro mammalian cell micronucleus test (MNvit) in the human lymphoblastoid cell line TK6 at Covance
laboratories, Harrogate UK in support of OECD draft Test Guideline 487. Mutat Res.
Frederick, CB; Bush, ML; Lomax, LG; et al. (1998) Application of a hybrid computational fluid dynamics and
physiologically based inhalation model for interspecies dosimetry extrapolation of acidic vapors in the upper
airways. Toxicol Appl Pharmacol 152(1):211-231.
Friesen, MC; Demers, PA; Spinelli, JJ; et al. (2010) Chronic and acute effects of coal tar pitch exposure and
cardiopulmonary mortality among aluminum smelter workers. Am J Epidemiol. 172(7):790-9.
Fu, PP; Harvey, RG; Beland, FA. (1978) Molecular orbital theoretical prediction of the isomeric products formed
from reactions of arene oxides and related metabolites of polycyclic aromatic hydrocarbons. Tetrahedron 34(7):866.
Galvan, N; Page, TJ; Czuprynski, CJ; et al. (2006) Benzo(a)pyrene and 7,12-dimethylbenz(a)anthrecene
differentially affect bone marrow cells of the lymphoid and myeloid lineages. Toxicol Appl Pharmacol 213(2): 105—
116.
Gangar, SC; Sandhir, R; Rai, DV; et al. (2006) Preventive effects of Azadirachta indica on benzo(a)pyrene-DNA
adduct formation in murine forestomach and hepatic tissues. Phytother Res 20:889-895.
Garcon, G; Gosset, P; Garry, S; et al. (2001a) Pulmonary induction of proinflammatory mediators following the rat
exposure to benzo(a)pyrene-coated onto Fe203 particles. Toxicol Lett 121(2): 107-117.
Gargon, G; Zerimech, F; Hannothiaux, M; et al. (2001b) Antioxidant defense disruption by polycyclic aromatic
hydrocarbons-coated onto Fe(2)0(3) particles in human lung cells (A549). Toxicology 166(3): 129—137.
Garte, S; Taioli, E; Raimondi, S; et al. (2007) Effects of metabolic genotypes on intermediary biomarkers in subjects
exposed to PAHS: results from the EXP AH study. Mutat Res 620(l-2):7-15.
Geacintov, NE; Cosman, M; Hingerty, BE; et al. (1997) NMR solution structures of stereoisometric covalent
polycyclic aromatic carcinogen-DNA adduct: principles, patterns, and diversity. Chem Res Toxicol 10(2): 111-146.
Gelboin, HV. (1980) Benzo[alpha]pyrene metabolism, activation and carcinogenesis: role and regulation of mixed-
function oxidases and related enzymes. Physiol Rev 60(4): 1107-1166.
Georgellis, A; Toppari, J; Veromaa, T; et al. (1990) Inhibition of meiotic divisions of rat spermatocytes in vitro by
polycyclic aromatic hydrocarbons. Mutat Res 231(2): 125—135.
Gerde, P; Scholander, P. (1989) An experimental study of the penetration of polycyclic aromatic hydrocarbons
through a model of the bronchial lining layer. Environ Res 48(2):287-295.
Gerde, P; Muggenburg, BA; Hoover, MD; et al. (1993a) Disposition of polycyclic aromatic hydrocarbons in the
respiratory tract of the beagle dog. I. The alveolar region. Toxicol Appl Pharmacol 121 (2): 313—318.
Gerde, P; Muggenburg, BA; Sabourin, PJ; et al. (1993b) Disposition of polycyclic aromatic hydrocarbons in the
respiratory tract of the beagle dog. II. The conducting airways. Toxicol Appl Pharmacol 121 (2):319—327.
Gerde, P; Muggenburg, BA; Henderson, RF. (1993c) Disposition of polycyclic aromatic hydrocarbons in the
respiratory tract of the beagle dog. III. Mechanisms of the dosimetry. Toxicol Appl Pharmacol 121(2):328—334.
Gerde, P; Muggenburg, BA; Lundborg, M; et al. (2001) The rapid alveolar absorption of diesel soot-adsorbed
benzo[a]pyrene: bioavailability, metabolism and dosimetry of an inhaled particle-borne carcinogen. Carcinogenesis
22(5):741-749.
259
DRAFT - DO NOT CITE OR QUOTE
-------
Germolec, DR; Kashon, M; Nyska, A; et al. (2004) The accuracy of extended histopathology to detect immunotoxic
chemicals. Toxicol Sci 82(2):504-514.
Germolec, DR; Nyska, A; Kashon, M; et al. (2004) Extended histopathology in immunotoxicity testing:
Interlaboratory validation studies. Toxicol Sci 78(1): 107-115.
Gibbs, GW. (1985) Mortality of aluminum reduction plant workers, 1950 through 1977. J Occup Med 27(10):761-
770.
Ginsberg, GL; Atherholt, TB. (1989) Transport of DNA-adducting metabolites in mouse serum following
benzo[a]pyrene administration. Carcinogenesis 10(4):673-679.
Godschalk, RW; Ostertag, JU; Moonen, EJ; et al. (1998a) Aromatic DNA adducts in human white blood cells and
skin after dermal application of coal tar. Cancer Epidemiol Biomarkers Prev 7(9):767-773.
Godschalk, RW; Maas, LM; Kleinjans, JC; et al. (1998b) Influences of DNA isolation and RNA contamination on
carcinogen-DNA adduct analysis by 32P-postlabeling. Environ Mol Mutagen 32(4):344-350.
Godschalk, RW; Moonen, EJ; Schilderman, PA; et al. (2000) Exposure-route-dependent DNA adduct formation by
polycyclic aromatic hydrocarbons. Carcinogenesis 21 (1 ):87—92.
Godschalk, RW; Ostertag, JU; Zandsteeg, AM; et al. (2001) Impact of GSTM1 on aromatic-DNA adducts and p53
accumulation in human skin and lymphocytes. Pharmacogenetics 11 (6):537—543.
Godschalk, R; Nair, J; Van Schooten, FJ; et al. (2002) Comparison of multiple DNA adduct types in tumor adjacent
human lung tissue: effect of cigarette smoking. Carcinogenesis 23(12):2081-2086.
Godschalk, R; Curfs, D; Bartsch, H; et al. (2003) Benzo[a]pyrene enhances lipid peroxidation induced DNA damage
in aorta of apolipoprotein E knockout mice. Free Radic Res 37(12): 1299—1305.
Gold, LS; Zeiger, E. (1997) Handbook of carcinogenicity potency and genotoxicity databases (CPDB). Boca Raton,
FL: CRC Press.
Goldstein, LS; Weyand, EH; Safe, S; et al. (1998) Tumors and DNA adducts in mice exposed to benzo[a]pyrene and
coal tars: implications for risk assessment. Environ Health Perspect 106 (Suppl 6): 1325—1330.
Goshman, LM and Heidelberger, C. (1967). Binding of tritium-labeled polycyclic hydrocarbons to dna of mouse
skin. Cancer Research 27: 1678-1688.
Gozukara, EM; Belvedere, G; Robinson, RC; et al. (1981) The effect of epoxide hydratase on benzo[a]pyrene diol
epoxide hydrolysis and binding to DNA and mixed-function oxidase proteins. Mol Pharmacol 19(1): 153-161.
Granberg, AL; Brunstrom, B; Brandt, I. (2000) Cytochrome P450-dependent binding of 7,12-
dimethylbenz [a] anthracene (DMBA) and benzo[a]pyrene (B[a]P) in murine heart, lung, and liver endothelial cells.
Arch Toxicol 74(10):593-601.
Grant, G; Roe, FJ. (1963) The effect of phenanthrene on tumour induction by 3,4-benzopyrene administered to
newly born mice. Br J Cancer 17:261-265.
Grimmer, G; Brune, H; Deutsch-Wenzel, R; et al. (1983) On the contribution of polycyclic aromatic hydrocarbons
to the carcinogenic impact of automobile exhaust condensate evaluated by local application onto mouse skin.
Cancer Lett 21(1):105-113.
Grimmer, G; Brune, H; Deutsch-Wenzel, R; et al. (1984) Contribution of polycyclic aromatic hydrocarbons to the
carcinogenic impact of gasoline engine exhaust condensate evaluated by implantation into the lungs of rats. J Natl
Cancer Inst 72(3):733-739.
260
DRAFT - DO NOT CITE OR QUOTE
-------
Grimmer, G; Dettbarn, G; Jacob, J. (1993) Biomonitoring of polycyclic aromatic hydrocarbons in highly exposed
coke plant workers by measurement of urinary phenanthrene and pyrene metabolites (phenols and dihydrodiols). Int
Arch Occup Environ Health 65(3): 189-199.
Grimmer, G; Dettbam, G; Naujack, K; et al. (1994) Relationship between inhaled PAH and urinary excretion of
phenanthrene, pyrene and benzo(a)pyrene metabolites in coke plant workers. Polycyclic Aromat Compd 5:269-277.
Gross, P; Tolker, E; Babyak, MA; et al. (1965) Experimental lung cancer in hamsters. Repetitive intratracheal
applications of two carcinogenic hydrocarbons. Arch Environ Health 11:59-65.
Grova, N; Valley, A; Turner, JD; et al. (2007) Modulation of behavior and NMDA-R1 gene mRNA expression in
adult female mice after sub-acute administration of benzo(a)pyrene. Neurotoxicology 28(3):630-636.
Grova, N; Schroeder, H; Farinelle, S; et al. (2008) Sub-acute administration of benzo[a]pyrene (B[a]P) reduces
anxiety-related behaviour in adult mice and modulates regional expression of N-methyl-D-aspartate (NMDA)
receptors genes in relevant brain regions. Chemosphere 73(1 Suppl):S295-302.
Grover, PL. (1986) Pathways involved in the metabolism and activation of polycyclic hydrocarbons. Xenobiotica
16(10-11 ):915—931.
Gu, J; Horikawa, Y; Chen, M; et al. (2008) Benzo(a)pyrene diol epoxide-induced chromosome 9p21 aberrations are
associated with increased risk of bladder cancer. Cancer Epidemiol Biomarkers Prev 17(9):2445-2450.
Guillemette, C. (2003) Pharmacogenomics of human UDP-glucuronosyltransferase enzymes. Pharmacogenomics J
3(3): 136—158.
Gundel, J; Schaller, KH; Angerer, J. (2000) Occupational exposure to polycyclic aromatic hydrocarbons in a
fireproof stone producing plant: biological monitoring of 1-hydroxypyrene, 1-, 2-, 3- and 4-hydroxyphenanthrene,
3-hydroxybenz(a)anthracene and 3-hydroxybenzo(a)pyrene. Int Arch Occup Environ Health 73(4):270-274.
Gunter, MJ; Divi, RL; Kulldorff, M; et al. (2007) Leukocyte polycyclic aromatic hydrocarbon-DNA adduct
formation and colorectal adenoma. Carcinogenesis 28(7): 1426-1429.
Gyorffy, E; Anna, L; Kovacs, K; et al. (2008) Correlation between biomarkers of human exposure to genotoxins
with focus on carcinogen-DNA adducts. Mutagenesis 23(1): 1-18.
Habs, M; Schmahl, D; Misfeld, J. (1980) Local carcinogenicity os some environmentally relevant polycyclic
aromatic hydrocarbons after lifelong topical application to mouse skin. Arch Geschwulstforsch 50(3):266-274.
Habs, M; Jahn, SA; Schmahl, D. (1984) Carcinogenic activity of condensate from coloquint seeds (Citrullus
colocynthis) after chronic epicutaneous administration to mice. J Cancer Res Clin Oncol 108(1): 154-156.
Hakim, FT; Memon, SA; Cepeda, R; et al. (2005) Age-dependent incidence, time course, and consequences of
thymic renewal in adults. J Clin Invest 115(4):930-939.
Hakura, A; Tsutsui, Y; Sonoda, J; et al. (1998) Comparison between in vivo mutagenicity and carcinogenicity in
multiple organs by benzo[a]pyrene in the lacZ transgenic mouse (Muta Mouse). Mutat Res 398(1-2): 123-130.
Hakura, A; Shimada, H; Nakajima, M; et al. (2005) Salmonella/human S9 mutagenicity test: a collaborative study
with 58 compounds. Mutagenesis 20(3):217-228.
Hall, M; Grover, PL. (1988) Stereoselective aspects of the metabolic activation of benzo[a]pyrene by human skin in
vitro. Chem Biol Interact 64(3):281-296.
Hammond, EC; Selikoff, I J; Lawther, PL; et al. (1976) Inhalation of benzpyrene and cancer in man. Ann N Y Acad
Sci 271:116-124.
261
DRAFT - DO NOT CITE OR QUOTE
-------
Hanelt, S; Helbig, R; Hartmann, A; et al. (1997) A comparative investigation of DNA adducts, DNA strand breaks
and gene mutations induced by benzo[a]pyrene and (+/-)-anti-benzo[a]pyrene-7,8-diol 9,10-oxide in cultured human
cells. MutatRes 390:179-188.
Hansen, ES. (1989) Cancer incidence in an occupational cohort exposed to bitumen fumes. Scand J Work Environ
Health 15(2): 101-105.
Hansen, ES. (1991) Mortality of mastic asphalt workers. Scand J Work Environ Health 17(l):20-24.
Hara, T; Nishikawa, T; Sui, H; et al. (2007) In vivo photochemical skin micronucleus test using a sunlight simulator:
detection of 8-methoxypsoralen and benzo[a]pyrene in hairless mice. Mutat Res 631:1 -8.
Hardin, JA; Hinoshita, F; Sherr, DH. (1992) Mechanisms by which benzo[a]pyrene, an environmental carcinogen,
suppresses B cell lymphopoiesis. Toxicol Appl Pharmacol 117(2): 155—164.
Harvey, RG. (1996) Mechanisms of carcinogenesis of polycyclic aromatic compounds. Polycyclic Aromat Compd
9:1-23.
Hayes, JD; Pulford, DJ. (1995) The glutathione S-transferase supergene family: regulation of GST and the
contribution of the isoenzymes to cancer chemoprotection and drug resistance. Crit Rev Biochem Mol Biol
30(6):445-600.
Hayes, JD; Strange, RC. (2000) Glutathione S-transferase polymorphisms and their biological consequences.
Pharmacology 61 (3): 154—166.
Hecht, SS. (2002) Human urinary carcinogen metabolites: biomarkers for investigating tobacco and cancer.
Carcinogenesis 23(6):907-922.
Hecht, SS. (2004) Carcinogen derived biomarkers: applications in studies of human exposure to secondhand tobacco
smoke. Tob Control 13 (Suppl l):48-56.
Hecht, SS; Grabowski, W; Groth, K. (1979) Analysis of faeces for benzo[a]pyrene after consumption of charcoal-
broiled beef by rats and humans. Food Cosmet Toxicol 17(3):223-227.
Hecht, SS; Chen, M; Yagi, H; et al. (2003) r-l,t-2,3,c-4-Tetrahydroxy-l,2,3,4-tetrahydrophenanthrene in human
urine: a potential biomarker for assessing polycyclic aromatic hydrocarbon metabolic activation. Cancer Epidemiol
Biomarkers Prev 12(12): 1501-1508.
Henry, MC; Port, CD; Bates, RR; et al. (1973) Respiratory tract tumors in hamsters induced by benzo(a)pyrene.
Cancer Res 33(7): 1585-1592.
Herbert, R; Marcus, M; Wolff, MS; et al. (1990) A pilot study of detection of DNA adducts in white blood cells of
roofers by 32P-postlabelling. IARC Sci Publ (104):205-214.
Herrold, KM; Dunham, LJ. (1962) Induction of carcinoma and papilloma of the Syrian hamster by intratracheal
instillation of benzo[a]-pyrene. J Natl Cancer Inst 28:467-491.
Higginbotham, S; RamaKrishna, NV; Johansson, SL; et al. (1993) Tumor-initiating activity and carcinogenicity of
dibenzo[a,l]pyrene versus 7,12-dimethylbenz[a]anthracene and benzo[a]pyrene at low doses in mouse skin.
Carcinogenesis 14(5):875-878.
Himmelstein, MW; Boogaard, PJ; Cadet, J; et al. (2009). Creating context for the use of DNA adduct data in cancer
risk assessment: II. Overview of methods of identification and quantitation of DNA damage. Crit Rev Toxicol.
39(8):679-94.
Hirom, PC; Chipman, JK; Millburn, P; et al. (1983) Enterohepatic circulation of the aromatic hydrocarbons
benzo[a]pyrene and naphthalene. In: Extrahepatic drug metabolism and chemical carcinogenesis. Amsterdam:
Elsvier Science Publishers; pp. 275-281.
262
DRAFT - DO NOT CITE OR QUOTE
-------
Hoffmann, D; Rathkamp, G; Nesnow, S; et al. (1972) Fluoranthenes: quantitative determination in cigarette smoke,
formation by pyrolysis, and tumor-initiating activity. J Natl Cancer Inst 49(4): 1165-1175.
Holladay, SD; Smith, BJ. (1994) Fetal hematopoietic alterations after maternal exposure to benzo[a]pyrene: a
cytometric evaluation. J Toxicol Environ Health 42(3):259-273.
Holladay, SD; Smith, BJ. (1995) Benzo[a]pyrene-induced alterations in total immune cell number and cell-surface
antigen expression in the thymus, spleen and bone marrow of b6c3fl mice. Vet Hum Toxicol 37(2):99-104.
Holladay, SD; Smialowicz, RJ. (2000) Development of the murine and human immune system: Differential effects
of immunotoxicants depend on time of exposure. Environ Health Perspect 108(Suppl 3):463-473.
Homburger, F; Hsueh, SS; Kerr, CS; et al. (1972) Inherited susceptibility of inbred strains of Syrian hamsters to
induction of subcutaneous sarcomas and mammary and gastrointestinal carcinomas by subcutaneous and gastric
administration of polynuclear hydrocarbons. Cancer Res 32(2):360-366.
Horton, AW; Bingham EL; Graf Burton, MJ; et al. (1965) Carcinogenesis of the skin. III. The contribution of
elemental sulfur and of organic sulfur compounds. Cancer Res 25:17591763.
Howard, P; Meylan, W. (1997) Handbook of physical properties of organic chemicals. Boca Raton, FL: Lewis
Publishers, Inc.
Hu, Z; Wells, PG. (2004) Human interindividual variation in lymphocyte UDP-glucuronosyltransferases as a
determinant of in vitro benzo[a]pyrene covalent binding and cytotoxicity. Toxicol Sci 78(1):32^10.
IARC (International Agency for Research on Cancer). (1973) IARC monographs on the evaluation of carcinogenic
risks to humans. Vol. 3. Certain polycyclic aromatic hydrocarbons and heterocyclic compounds. Lyon, France:
International Agency for Research on Cancer
IARC (International Agency for Research on Cancer). (1983) IARC monographs on the evaluation of carcinogenic
risks to humans. Vol. 32. Polynuclear aromatic hydrocarbons, Part 1, chemical environmental and experimental
data. Lyon, France: International Agency for Research on Cancer.
IARC (International Agency for Research on Cancer). (1984) Polynuclear aromatic compounds. Part 3, Industrial
exposures in aluminium production, coal gasification, coke production, and iron and steel founding. IARC
monographs on the evaluation of carcinogenic risks to humans. Volume 34. Lyon, France: International Agency for
Research on Cancer.
IARC (International Agency for Research on Cancer). (1985) Polynuclear aromatic compounds. Part 4, bitumens
coal-tars and derived products shale-oils and soots. IARC International Agency for Research on Cancer Monographs
on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. Volume 35. Lyon, France: International
Agency for Research on Cancer.
IARC (International Agency for Research on Cancer). (1987a) IARC monographs on the evaluation of carcinogenic
risks to humans. Overall evaluation of carcinogenicity: an updating of IARC Monographs volumes 1 to 42.
Supplement 7. Lyon, France: International Agency for Research on Cancer.
IARC (International Agency for Research on Cancer). (1987b) Tobacco smoking. IARC monographs on the
evaluation of carcinogenic risks to humans. Volume 30. Lyon, France: International Agency for Research on
Cancer.
IARC (International Agency for Research on Cancer). (1998) IARC Monographs on the Evaluation of Carcinogenic
Risks to Humans: preliminary classification. Available online at http://monographs.iarc.fr/ENG/Meetings/92-
pahs.pdf.
Ingram, AJ; Phillips, JC; Davies, S. (2000) DNA adducts produced by oils, oil fractions and polycyclic aromatic
hydrocarbons in relation to repair processes and skin carcinogenesis. J Appl Toxicol 20(3): 165-174.
263
DRAFT - DO NOT CITE OR QUOTE
-------
Inyang, F; Ramesh, A; Kopsombut, P; et al. (2003) Disruption of testicular steroidogenesis and epididymal function
by inhaled benzo(a)pyrene. Reprod Toxicol 17(5):527-537.
Izzotti, A; Balansky, R; Scatolini, L; et al. (1991) [Research on adducts of benzo(a)pyrene and DNA in different
cells using synchronous fluorescence spectrophotometry], Bollettino - Societa Italiana Biologia Sperimentale
67(9):869-874.
Izzotti, A; Bagnasco, M; D'Agostini, F; et al. (1992) Evaluation of BPDE-DNA adducts in lung cells. Perspectives
in human biomonitoring and chemoprevention of cancer. Mutat Res 27(2): 191.
Jarabek, AM; Pottenger, LH; Andrews, LS; et al. (2009); Creating context for the use of DNA adduct data in cancer
risk assessment: I. Data organization. Crit Rev Toxicol. 39(8):659-78.
Jeffrey, AM; Jennette, KW; Blobstein, SH; et al. (1976) Letter: Benzo[a]pyrene-nucleic acid derivative found in
vivo: structure of a benzo[a]pyrenetetrahydrodiol epoxide-guanosine adduct. J Am Chem Soc 98(18):5714-5715.
Jeffy, BD; Chirnomas, RB; Romagnolo, DF. (2002) Epigenetics of breast cancer: polycyclic aromatic hydrocarbons
as risk factors. Environ Mol Mutagen 39(2-3):235-244.
Jerina, DM; Chadha, A; Cheh, AM; et al. (1991) Covalent bonding of bay-region diol epoxides to nucleic acids.
Adv Exp Med Biol 283:533-553.
Jiang, H; Shen, YM; Quinn, AM; et al. (2005) Competing roles of cytochrome P450 1A1/1B1 and aldo-keto
reductase 1 Al in the metabolic activation of (+/-)-7,8-dihydroxy-7,8-dihydro-benzo[a]pyrene in human
bronchoalveolar cell extracts. Chem Res Toxicol 18(2):365-374.
Jiang, H; Gelhaus, SL; Mangal, D; et al. (2007) Metabolism of benzo[a]pyrene in human bronchoalveolar H358
cells using liquid chromatography-mass spectrometry. Chem Res Toxicol 20(9): 1331-1341.
Johnson, D; Bruice, TC. (1975) Nucleophilic catalysis of the aromatization of an arene oxide. The reaction of
trimethylamine with 4-carbo-tert-butoxybenzene. J Am Chem Soc 97:6901-6903.
Johnson, M; Pullen, L; Cogliano, J. (1997) An approach for calculating risk from dermal exposure to
benzo[a]pyrene based on point of contact effects. Unpublished review draft report.
Joseph, P; Jaiswal, AK. (1998) NAD(P)H:quinone oxidoreductase 1 reduces the mutagenicity of DNA caused by
NADPH:P450 reductase-activated metabolites of benzo(a)pyrene quinones. Br J Cancer 77(5):709-719.
Jyonouchi, H; Sun, S; Iijima, K; et al. (1999) Effects of anti-7,8-dihydroxy-9,10-epoxy- 7,8,9,10-
tetrahydrobenzo[a]pyrene on human small airway epithelial cells and the protective effects of myo-inositol.
Carcinogenesis 20(1): 139—145.
Kalina, I; Brezani, P; Gajdosova, D; et al. (1998) Cytogenetic monitoring in coke oven workers. Mutat Res
417(1):9—17.
Kamath, AB; Xu, H; Nagarkatti, PS; et al. (1997) Evidence for the induction of apoptosis in thymocytes by 2,3,7,8-
tetrachlorodibenzo-p-dioxin in vivo. Toxicol Appl Pharmacol 142(2):367-377.
Kapitulnik, J; Wislocki, PG; Levin, W; et al. (1978) Tumorigenicity studies with diol-epoxides of benzo[a]pyrene
which indicate that (+/-)-trans-7beta,8alpha-dihydroxy-9alpha,10alpha-epoxy-7,8,9,10-tetrahydr obenzo[a]pyrene is
an ultimate carcinogen in newborn mice. Cancer Res 38:354-358.
Kao, J; Hall, J; Shugart, LR; et al. (1984) An in vitro approach to studying cutaneous metabolism and disposition of
topically applied xenobiotics. Toxicol Appl Pharmacol 75(2):289-298.
Kao, J; Patterson, FK; Hall, J. (1985) Skin penetration and metabolism of topically applied chemicals in six
mammalian species, including man: an in vitro study with benzo[a]pyrene and testosterone. Toxicol Appl
Pharmacol 81(3 Pt 1):502-516.
Kao, J; Hall, J; Helman, G. (1988) In vitro percutaneous absorption in mouse skin: influence of skin appendages.
Toxicol Appl Pharmacol 94(1):93—103.
264
DRAFT - DO NOT CITE OR QUOTE
-------
Kapitulnik, J; Wislocki, PG; Levin, W; et al. (1978) Tumorigenieity studies with diol-epoxides of benzo(a)pyrene
which indicate that (+/-)-trans-7beta,8alpha-dihydroxy-9alpha,10alpha-epoxy-7,8,9,10-tetrahydr obenzo(a)pyrene is
an ultimate carcinogen in newborn mice. Cancer Res 38(2):354-358.
Kawamura, Y; Kamata, E; Ogawa, Y; et al. (1988) The effect of various foods on the intestinal absorption of benzo-
a-pyrene in rats. J Food Hyg Soc Jpn 29(1):21—25.
Keller, GM; Christou, M; Pottenger, LH; et al. (1987) Product inhibition of benzo[a]pyrene metabolism in
uninduced rat liver microsomes: effect of diol epoxide formation. Chem Biol Interact 61(2): 159-175.
Ketkar, M; Reznik, G; Schneider, P; et al. (1978) Investigations on the carcinogenic burden by air pollution in man.
Intratracheal instillation studies with benzo(a)pyrene in bovine serum albumin in Syrian hamsters. Cancer Lett
4(4):235-239.
Kiefer, F; Cumpelik, 0; Wiebel, FJ. (1988) Metabolism and cytotoxicity of benzo(a)pyrene in the human lung
tumour cell line NCI-H322. Xenobiotica 18(6):747-755.
Kim, KB; Lee, BM. (1997) Oxidative stress to DNA, protein, and antioxidant enzymes (superoxide dismutase and
catalase) in rats treated with benzo(a)pyrene. Cancer Lett 113(1 -2):205—212.
Kim, HS; Kwack, SJ; Lee, BM. (2000) Lipid peroxidation, antioxidant enzymes, and benzo[a]pyrene-quinones in
the blood of rats treated with benzo[a]pyrene. Chem Biol Interact 127(2): 139—150.
Kim, IY; Hyun, CK. (2006) Comparative evaluation of the alkaline comet assay with the micronucleus test for
genotoxicity monitoring using aquatic organisms. Ecotoxicol Environ Saf 64:288-297.
Kleiner, HE; Vulimiri, SV; Hatten, WB; et al. (2004) Role of cytochrome p4501 family members in the metabolic
activation of polycyclic aromatic hydrocarbons in mouse epidermis. Chem Res Toxicol 17(12): 1667-1674.
Klemme, JC; Mukhtar, H; Elmets, CA. (1987) Induction of contact hypersensitivity to dimethylbenz(a)anthracene
and benzo(a)pyrene in C3H/HeN mice. Cancer Res 47(22):6074-6078.
Knaapen, AM; Curfs, DM; Pachen, DM; et al. (2007) The environmental carcinogen benzo[a]pyrene induces
expression of monocyte-chemoattractant protein-1 in vascular tissue: a possible role in atherogenesis. Mutat Res
621 (31):41.
Knafla, A; Phillips, KA; Brecher, RW; et al. (2006) Development of a dermal cancer slope factor for
benzo[a]pyrene. Reg Toxicol Pharmacol 45:159-168.
Knox, EG. (2005) Childhood cancers and atmospheric carcinogens. J Epidemiol Community Health 59(2): 101-105.
Knuckles, ME; Inyang, F; Ramesh, A. (2001) Acute and subchronic oral toxicities of benzo[a]pyrene in F-344 rats.
Toxicol Sci 61(2):382-388.
Kobayashi, N. (1975) Production of respiratory tract tumors in hamsters by benzo(a)pyrene. Gann 66(3):311—315.
Kondraganti, SR; Fernandez-Salguero, P; Gonzalez, FJ; et al. (2003) Polycyclic aromatic hydrocarbon-inducible
DNA adducts: evidence by 32P-postlabeling and use of knockout mice for Ah receptor-independent mechanisms of
metabolic activation in vivo. Int J Cancer 103(1):5—11.
Koreeda, M; Moore, PD; Wislocki, PG; et al. (1978) Binding of benzo[a]pyrene 7,8-diol-9,10-epoxides to DNA,
RNA, and protein of mouse skin occurs with high stereoselectivity. Science 199(4330):778-781.
Kriek, E; Van Schooten, FJ; Hillebrand, MJ; et al. (1993) DNA adducts as a measure of lung cancer risk in humans
exposed to polycyclic aromatic hydrocarbons. Environ Health Perspect 99:71-75.
Kristensen, P; Eilertsen, E; Einarsdottir, E; et al. (1995) Fertility in mice after prenatal exposure to benzo[a]pyrene
and inorganic lead. Environ Health Perspect 103(6):588-590.
265
DRAFT - DO NOT CITE OR QUOTE
-------
Kroese, ED; Muller, JJA; Mohn, GR; et al. (2001) Tumorigenic effects in Wistar rats orally administered
benzo(a)pyrene for two years: Implications for human cancer risks associated with oral exposure to polycyclic
aromatic hydrocarbons. Bilthoven, Netherlands: RIVM 658603.
Kunstler, K. (1983) Failure to induce tumors by intratracheal instillation of automobile exhaust condensate and
fractions thereof in Syrian golden hamsters. Cancer Lett 18(1): 105—108.
Kuper, C; Beems, R; Bruijntjes, J; et al. (1992) Normal development, growth, and aging of the thymus. In: Mohr, U;
Dungworth, DL; Capen, CC; eds. Pathobiology of the aging rat, Vol. 1. Washington, DC: ILSI Press; pp. 25-48.
Kuper, CF; De Heer, E; Van Loveren, H; et al. (2002) Immune system. In: Hascheck-Hock, WM; Rousseaux, CG;
Wallig, MA; eds. Handbook of Toxicologic Pathology, Vol. 2. New York: Academic Press; pp. 585-644.
LaGoy, PK; Quirk, TC. (1994) Establishing generic remediation goals for the polycyclic aromatic hydrocarbons:
critical issues. Environ Health Perspect 102:348-352.
Landreth, K; Dodson, S. (2005) Development of the rodent immune system. In: Holladay, SD; ed. Developmental
Immunotoxicology. Boca Raton, FL: CRC Press; pp. 3-19.
Landreth, KS. (2002) Critical windows in development of the rodent immune system. Human and Experimental
Toxicology 21(9-10):493-498.
LaVoie, EJ; Amin, S; Hecht, SS; et al. (1982) Tumour initiating activity of dihydrodiols of benzo[b]fluoranthene,
benzo[j]fluoranthene, and benzo[k]fluoranthene. Carcinogenesis 3(l):49-52.
LaVoie, EJ; Braley, J; Rice, JE; et al. (1987) Tumorigenic activity of non-alternant polynuclear aromatic
hydrocarbons in newborn mice. Cancer Lett 34(1): 15-20.
LaVoie, EJ; Cai, ZW; Meschter, CL; et al. (1994) Tumorigenic activity of fluoranthene, 2-methylfluoranthene and
3-methylfluoranthene in newborn CD-I mice. Carcinogenesis 15(10):2131—2135.
Lazarova, M; Labaj, J; Eckl, P; et al. (2006) Comparative evaluation of DNA damage by genotoxicants in primary
rat cells applying the comet assay. Toxicol Lett 164(l):54-62.
Leadon, SA; Stampfer, MR; Bartley, J. (1988) Production of oxidative DNA damage during the metabolic activation
of benzo[a]pyrene in human mammary epithelial cells correlates with cell killing. Proc Natl Acad Sci U S A
85:4365-4368.
Lee, ME; Urso, P. (2007) Suppression of T lymphocyte proliferation to antigenic and mitogenic stimuli by
Benzo(alpha)pyrene and 2-aminofluorene metabolites. Immunopharmacol Immunotoxicol 29(3-4):425^138.
Legraverend, C; Harrison, DE; Ruscetti, FW; et al. (1983) Bone marrow toxicity induced by oral benzo[a]pyrene:
protection resides at the level of the intestine and liver. Toxicol Appl Pharmacol 70(3):390-401.
Le Marchand, L; Sivaraman, L; Pierce, L; et al. (1998) Associations of CYP1A1, GSTM1, and CYP2E1
polymorphisms with lung cancer suggest cell type specificities to tobacco carcinogens. Cancer Res 58(21):4858-
4863.
Leng, SG; Zheng, YX; Pan, ZF; et al. (2004) [A study on the inherited susceptibility of chromosomal damage in
peripheral blood lymphocytes among coke oven workers]. Zhonghua Yu Fang Yi Xue Za Zhi 38(2):94-98.
Leung, HW; Henderson, RF; Bond, JA; et al. (1988) Studies on the ability of rat lung and liver microsomes to
facilitate transfer and metabolism of benzo [a]pyrene from diesel particles. Toxicology 51 (1): 1 -9.
Levin, W; Wood, AW; Yagi, H; et al. (1976) Carcinogenicity of benzo[a]pyrene 4,5-, 7,8-, and 9,10-oxides on
mouse skin. Proc Natl Acad Sci USA 73(l):243-247.
266
DRAFT - DO NOT CITE OR QUOTE
-------
Levin, W; Wood, AW; Wislocki, PG; et al. (1977) Carcinogenicity of benzo-ring derivatives of benzo(a)pyrene on
mouse skin. Cancer Res 37(9):3356-3361.
Levin, W; Wood, A; Chang, R; et al. (1982) Oxidative metabolism of polycyclic aromatic hydrocarbons to ultimate
carcinogens. Drug Metab Rev 13(4):555-580.
Lewtas, J; Walsh, D; Williams, R; et al. (1997) Air pollution exposure-DNA adduct dosimetry in humans and
rodents: evidence for non-linearity at high doses. Mutat Res 378(1 -2):51—63.
Li, D; Zhang, W; Sahin, AA; et al. (1999) DNA adducts in normal tissue adjacent to breast cancer: a review. Cancer
Detect Prev 23(6):454-462.
Li, D; Firozi, PF; Wang, LE; et al. (2001) Sensitivity to DNA damage induced by benzo(a)pyrene diol epoxide and
risk of lung cancer: a case-control analysis. Cancer Res 61 (4): 1445—1450.
Li, D; Day, RS; Bondy, ML; et al. (2007) Dietary mutagen exposure and risk of pancreatic cancer. Cancer
Epidemiol Biomarkers Prev 16(4):655-661.
Li, P; Li, X; Stagnitti, F; (2009) Studies on the sources of benzo[a]pyrene in grain and aboveground tissues of rice
plants. J Hazard Mater. 162(1 ):463-8.
Liao, KH; Dobrev, ID; Dennison, JE, Jr.; et al. (2002) Application of biologically based computer modeling to
simple or complex mixtures. Environ Health Perspect 110 (Suppl 6): 957-963.
Likhachev, AJ; Beniashvili, DS; Bykov, VJ; et al. (1992) Relevance of quantitation of benzo(a)pyrene metabolites
in animal excretes to evaluate individual human cancer risk. Prog Clin Biol Res 374:435-452.
Lin, P; Hsueh, YM; Ko, JL; et al. (2003) Analysis of NQOl, GSTP1, and MnSOD genetic polymorphisms on lung
cancer risk in Taiwan. Lung Cancer 40(2): 123-129.
Liu, SH; Wang, JH; Chuu, JJ; et al. (2002) Alterations of motor nerve functions in animals exposed to motorcycle
exhaust. J Toxicol Environ Health A 65:803-812.
Liu, Z; Muehlbauer, KR; Schmeiser, HH; et al. (2005) p53 mutations in benzo(a)pyrene-exposed human p53 knock-
in murine fibroblasts correlate with p53 mutations in human lung tumors. Cancer Res 65(7):2583-2587.
Liu, S; Abdelrahim, M; Khan, S; et al. (2006) Aryl hydrocarbon receptor agonists directly activate estrogen receptor
alpha in MCF-7 breast cancer cells. Biol Chem 387(9): 1209-1213.
Lodovici, M; Akpan, V; Giovannini, L; et al. (1998) Benzo[a]pyrene diol-epoxide DNA adducts and levels of
polycyclic aromatic hydrocarbons in autoptic samples from human lungs. Chem Biol Interact 116(3): 199-212.
Luebke, RW; Chen, DH; Dietert, R; et al. (2006) The comparative immunotoxicity of five selected compounds
following developmental or adult exposure. J Toxicol Environ Health B Crit Rev 9(1): 1-26.
Lummus, ZL; Henningsen, G. (1995) Modulation of t-cell ontogeny by transplacental benzo(a)pyrene. Int J
Immunopharmacol 17(4):339-350.
Luster, MI; Munson, AE; Thomas, PT; et al. (1988) Development of a testing battery to assess chemical-induced
immunotoxicity: National toxicology program's guidelines for immunotoxicity evaluation in mice. Fundam Appl
Toxicol 10(1):2-19.
Lyte, M; Bick, PH. (1985) Differential immunotoxic effects of the environmental chemical benzo[a]pyrene in young
and aged mice. Mech Ageing Dev 30(3):333—341.
Lyte, M; Blanton, RH; Myers, MJ; et al. (1987) Effect of in vivo administration of the carcinogen benzo(a)pyrene
on interleukin-2 and interleukin-3 production. Int J Immunopharmacol 9(3):307-312.
Ma, Q; Lu, AY. (2007) CYP1A induction and human risk assessment: an evolving tale of in vitro and in vivo
studies. Drug Metab Dispos 35(7): 1009-1016.
267
DRAFT - DO NOT CITE OR QUOTE
-------
Mackall, CL; Fleisher, TA; Brown, MR; et al. (1995) Age, thymopoiesis, and cd4+ t-lymphocyte regeneration after
intensive chemotherapy. NEngl J Med 332(3): 143-149.
MacKenzie, KM; Angevine, DM. (1981) Infertility in mice exposed in utero to benzo(a)pyrene. Biol Reprod
24(1): 183—191.
MacNicoll, AD; Grover, PL; Sims, P. (1980) The metabolism of a series of polycyclic hydrocarbons by mouse skin
maintained in short-term organ culture. Chem Biol Interact 29(2):169-188.
Malaveille, C; Kuroki, T; Sims, P; et al. (1977) Mutagenicity of isomeric diol-epoxides of benzo[a]pyrene and
benz[a]anthracene in S. typhimurium TA98 and TA100 and in V79 Chinese hamster cells. Mutat Res 44(3):313—
326.
Mallet, WG; Mosebrook, DR; Trush, MA. (1991) Activation of (+-)-trans-7,8-dihydroxy-
7,8-dihydrobenzo[a]pyrene to diolepoxides by human polymorphonuclear leukocytes or myeloperoxidase.
Carcinogenesis 12(3):521—524.
Mancini,R; Romano, G; Sgambato, A; et al. (1999). Polycyclic aromatic hydrocarbon-DNA adducts in cervical
smears of smokers and nonsmokers. Gynecol Oncol. 75:68-71.
Marnett, LJ. (1990) Prostaglandin synthase-mediated metabolism of carcinogens and a potential role for peroxyl
radicals as reactive intermediates. Environ Health Perspect 88:5-12.
Marshall, CJ; Vousden, KH; Phillips, DH. (1984) Activation of c-Ha-ras-1 proto-oncogene by in vitro modification
with a chemical carcinogen, benzo(a)pyrene diol-epoxide. Nature 310(5978):586-589.
Martin, FL; Cole, KJ; Williams, JA; et al. (2000) Activation of genotoxins to DNA-damaging species in exfoliated
breast milk cells. Mutat Res 470(2): 115—124.
Martin, FL; Cole, KJ; Muir, GH; et al. (2002) Primary cultures of prostate cells and their ability to activate
carcinogens. Prostate Cancer Prostatic Dis 5(2):96-104.
Martin, KR; Jokinen, MP; Honeycutt, HP; et al. (2004) Tumor profile of novel p53 heterozygous Tg.AC (v-Ha-ras)
bitransgenic mice treated with benzo(a)pyrene and fed dietary N-acetyl-L-cysteine (NAC). Toxicol Sci 81(2):293-
301.
Mass, MJ; Jeffers, AJ; Ross, JA; et al. (1993) Ki-ras oncogene mutations in tumors and DNA adducts formed by
benz[j]aceanthrylene and benzo[a]pyrene in the lungs of strain A/J mice. Mol Carcinog 8(3): 186-192.
Matkainen,T; Perez, GI; Jurisicova, A; et al. (2001) Aromatic hydrocarbon receptor-driven BAX gene expression is
required for premature ovarian failure caused by biohazardous environmental chemicals. Nature Genetics 28:355-
360.
Matkainen,T; Moriyama, T; Morita, Y; et al. (2002) Ligand activation of the aromatic hydrocarbon receptor
transcription factor drives Bax-dependent apoptosis in developing fetal ovarian germ cells. Endocrinology 143;615-
620.
Matsuoka, A; Matsuura, K; Sakamoto, H; et al. (1998) Spindle disturbances induced by benzo[a]pyrene and
7,12-dimethylbenz[a]anthracene in a Chinese hamster cell line (V79-MZ) and the stability of the numerical
chromosome aberrations that follow. Mutat Res 419(1-3): 1-12.
Matsuoka, A; Matsuura, K; Sakamoto, H; et al. (1999) A proposal for a simple way to distinguish aneugens from
clastogens in the in vitro micronucleus test. Mutagenesis 14(4):385-389.
Mattison, DR; Thorgeirsson, SS. (1977) Genetic differences in mouse ovarian metabolism of benzo[a]pyrene and
oocyte toxicity. Biochem Pharmacol 26:909-912.
268
DRAFT - DO NOT CITE OR QUOTE
-------
Mattison, DR; Thorgeirsson, SS. (1979). Ovarian aryl hydrocarbon hydroxylase activity and primordial ooctye
toxicity of polcyclic aromatic hydrocarbons in mice. Cancer Res 39;3471-3475.
Mattison, DR; Nightingale, MR. (1980) The biochemical and genetic characteristics of murine ovarian aryl
hydrocarbon (benzo[a]pyrene) hydroxylase activity and its relationship to primordial oocyte destruction by
polycyclic aromatic hydrocarbons. Toxic App Pharmacol 56:399-408.
Mattison, DR; White, NB; Nightingale, MR (1980). The effect of benzo(a)pyrene on fertility, primordial oocyte
number, and ovarian response to pregnant mare' s serum gonadotropin. Ped Pharmacol 1:143-151.
Matullo, G; Peluso, M; Polidoro, S; et al. (2003) Combination of DNA repair gene single nucleotide polymorphisms
and increased levels of DNA adducts in a population-based study. Cancer Epidemiol Biomarkers Prev 12(7):674-
677.
McCabe, DP; Flynn, EJ. (1990) Deposition of low dose benzo(a)pyrene into fetal tissue: influence of protein
binding. Teratology 41(1 ):85-95.
McCallister MM; Maguire M; Ramesh A; et al., (2008) Prenatal exposure to benzo(a)pyrene impairs later-life
cortical neuronal function. Neurotoxicology 29(5):846-54.
McCarver, DG; Hines, RN. (2002) The ontogeny of human drug-metabolizing enzymes: phase II conjugation
enzymes and regulatory mechanisms. J Pharmacol Exp Ther 300(2):361-366.
McCoull, KD; Rindgen, D; Blair, IA; et al. (1999) Synthesis and characterization of polycyclic aromatic
hydrocarbon o-quinone depurinating N7-guanine adducts. Chem Res Toxicol 12(3):237-246.
Melendez-Colon, VJ; Luch, A; Seidel, A; et al. (1999) Cancer initiation by polycyclic aromatic hydrocarbons results
from formation of stable DNA adducts rather than apurinic sites. Carcinogenesis 20(10): 1885—1891.
Melikian, AA; Bagheri, K; Hecht, SS; et al. (1989) Metabolism of benzo[a]pyrene and 7 beta,8 alpha-dihydroxy-9
alpha, 10 alpha-epoxy-7,8,9,10-tetrahydrobenzo[a pyrene in lung and liver of newborn mice. Chem Biol Interact
69(2-3):245-257.
Melikian, AA; Sun, P; Prokopczyk, B; et al. (1999) Identification of benzo[a]pyrene metabolites in cervical mucus
and DNA adducts in cervical tissues in humans by gas chromatography-mass spectrometry. Cancer Lett
146(2): 127-134.
Meng, F; Knapp, GW; Green, T; et al. (2010) K-Ras mutant fraction in A/J mouse lung increases as a function of
benzo[a]pyrene dose. Environ Mol Mutagen. 51(2): 146-55.
Mensing, T; Marczynski, B; Engelhardt, B; et al. (2005) DNA adduct formation of benzo[a]pyrene in white blood
cells of workers exposed to polycyclic aromatic hydrocarbons. Int J Hyg Environ Health 208:173-178.
Merk, HF; Mukhtar, H; Kaufmann, I; et al. (1987) Human hair follicle benzo[a]pyrene and benzo[a]pyrene 7,8-diol
metabolism: effect of exposure to a coal tar-containing shampoo. J Invest Dermatol 88(1):71—76.
Merlo, F; Bolognesi, C; Peluso, M; et al. (1997) Airborne levels of polycyclic aromatic hydrocarbons: 32P-
postlabeling DNA adducts and micronuclei in white blood cells from traffic police workers and urban residents. J
Environ Pathol Toxicol Oncol 16(2-3): 157—162.
Midgette, AS; Baron, JA. (1990) Cigarette smoking and the risk of natural menopause. Epidemiology; 1:474-479.
Miller, MM; Plowchalk, DR; Weitzman, GA et al. (1992) The effect of benzo(a)pyrene on murine ovarian and
corpora lutea volumes. Am J OBstet Gynecol 166:1535-1541.
Miller, KP; Ramos, KS. (2001) Impact of cellular metabolism on the biological effects of benzo[a]pyrene and
related hydrocarbons. Drug Metab Rev 3 3 (1): 1 -3 5.
269
DRAFT - DO NOT CITE OR QUOTE
-------
Mirvish, SS; Ghadirian, P; Wallcave, L; et al. (1981) Effect of diet on fecal excretion and gastrointestinal tract
distribution of unmetabolized benzo(a)pyrene and 3- methylcholanthrene when these compounds are administered
orally to hamsters. Cancer Res 41 (6): 2289-2293.
Mohamed, el-SA; Song, WH; Oh, SA; et al.(2010) The transgenerational impact of benzo(a)pyrene on murine male
fertility. Hum Reprod. 2010 25(10):2427-33.
Mohr, U. (1971) Diethylnitrosamine-induced carcinogenesis. Diaplacental effect. Fortschr Med 89(6):251-253
[German],
Moir, D. (1999) Physiological modeling of benzo(a)pyrene pharmacokinetics in the rat. In: Salem, H; Katz, SA;
eds. Toxicity assessment alternatives: methods, issues, opportunities. Totowa, NJ: Humana Press Inc.; pp. 79-95.
Moir, D; Viau, A; Chu, I; et al. (1998) Pharmacokinetics of benzo[a]pyrene in the rat. J Toxicol Environ Health A
53(7):507-530.
Monteith, DK; Novotny, A; Michalopoulos, G; et al. (1987) Metabolism of benzo[a]pyrene in primary cultures of
human hepatocytes: dose-response over a four-log range. Carcinogenesis 8(7):983-988.
Moore, BP; Hicks, RM; Knowles, MA; et al. (1982) Metabolism and binding of benzo(a)pyrene and 2-
acetylaminofluorene by short-term organ cultures of human and rat bladder. Cancer Res 42(2):642-648.
Moore, LE; Wiencke, JK; Bates, MN; et al. (2004) Investigation of genetic polymorphisms and smoking in a
bladder cancer case-control study in Argentina. Cancer Lett 211(2): 199-207.
Moorthy, B; Miller, KP; Jiang, W; et al. (2003) Role of cytochrome P4501B1 in benzo[a]pyrene bioactivation to
DNA-binding metabolites in mouse vascular smooth muscle cells: evidence from 32P-postlabeling for formation of
3-hydroxybenzo[a]pyrene and benzo[a]pyrene-3,6-quinone as major proximate genotoxic intermediates. J
Pharmacol Exp Ther 305(1 ):394-401.
Morris, JJ; Seifter, E. (1992) The role of aromatic hydrocarbons in the genesis of breast cancer. Med Hypotheses
38(3): 177—184.
Morse, MA; Carlson, GP. (1985) Distribution and macromolecular binding of benzo[a]pyrene in SENCAR and
BALB/c mice following topical and oral administration. J Toxicol Environ Health 16(2):263-276.
Motykiewicz, G; Michalska, J; Pendzich, J; et al. (1998) A molecular epidemiology study in women from Upper
Silesia, Poland. Toxicol Lett 96-97:195-202.
Mukhtar, H; Asokan, P; Das, M; et al. (1986) Benzo(a)pyrene diol epoxide-I-DNA adduct formation in the
epidermis and lung of SENCAR mice following topical application of crude coal tar. Cancer Lett 33(3):287-294.
Munson, AE; White, KLJ. (1983) Effects of subchronic fourteen day exposure to benzopyrene in B6C3F1 male and
female mice on specific immunological parameters. American Petroleum Institute. Submitted under TSCA Section
FYI; NTIS No. OTSAX-1182-0220.
Munson, AE; White, KL; Lysy, HH. (1985) Effects of subchronic fourteen day exposure to benzopyrene in B6C3F1
female mice on host resistance. Submitted under TSCA Section FYI; EPA Document No. FYI-AX-0485-0220,
Fiche No. OTS0000220-2.
Naslund, I; Rubio, CA; Auer, GU. (1987) Nuclear DNA changes during pathogenesis of squamous carcinoma of the
cervix in 3,4-benzopyrene-treated mice. Anal Quant Cytol Histol 9(5):411-418.
N'Diaye, M; Le, FE; Lagadic-Gossmann, D; et al. (2006) Aryl hydrocarbon receptor- and calcium-dependent
induction of the chemokine CCL1 by the environmental contaminant benzo[a]pyrene. J Biol Chem 281(29): 19906-
19915.
270
DRAFT - DO NOT CITE OR QUOTE
-------
Neal, J; Rigdon, H. (1967) Gastric tumors in mice fed benzo[a]pyrene: a quantitative study. Tex Rep Biol Med
25(4):553-557.
Neal, MS; Zhu, J; Holloway, AC; Foster, WG. (2007) Follicle growth is inhibited by benzo-[a]-pyrene, at
concentrations representative of human exposure, in an isolated rat follicle culture assay. Human Repro 22;961-967.
Neal, MS; Zhu, J; Foster, WG. (2008) Quantification of benzo[a]pyrene and other PAHs in the serum and follicular
fluid of smokers versus non-smokers. Repro Toxicol 2008:100-106.
Nebert, DW; Puga, A; Vasiliou, V. (1993) Role of the Ah receptor and the dioxin-inducible [Ah] gene battery in
toxicity, cancer, and signal transduction. Ann N Y Acad Sci 685:624-640.
Nesnow, S; Triplett, LL; Slaga, TJ. (1983) Mouse skin tumor initiation-promotion and complete carcinogenesis
bioassays: mechanisms and biological activities of emission samples. Environ Health Perspect 47:255-268.
Nesnow, S; Gold, A; Sangaiah, R; et al. (1984) Mouse skin tumor-initiating activity of benz[e]aceanthrylene and
benz[l]aceanthrylene in SENCAR mice. Cancer Lett 22(3):263-268.
Nesnow, S; Mass, MJ; Ross, JA; et al. (1998a) Lung tumorigenic interactions in strain A/J mice of five
environmental polycyclic aromatic hydrocarbons. Environ Health Perspect 106 Suppl 6:1337-1346.
Nesnow, S; Ross, JA; Mass, MJ; et al. (1998b) Mechanistic relationships between DNA adducts, oncogene
mutations, and lung tumorigenesis in strain A mice. Exp Lung Res 24(4): 3 95-405.
Nesnow, S; Ross, JA; Stoner, GD; et al. (1995) Mechanistic linkage between DNA adducts, mutations in oncogenes
and tumorigenesis of carcinogenic environmental polycyclic aromatic hydrocarbons in strain A/J mice. Toxicology
105(2-3):403-413.
Nesnow, S; Ross, JA; Stoner, GD; et al. (1996) Tumorigenesis of carcinogenic environmental polycyclic aromatic
hydrocarbons in strain A/J mice: linkage to DNA adducts and mutations in oncogenes. Polycyclic Aromat Compd
10:259-266.
Nettesheim, P; Griesemer, RA; Martin, DH; et al. (1977) Induction of preneoplastic and neoplastic lesions in grafted
rat tracheas continuously exposed to benzo(a)pyrene. Cancer Res 37(5): 1272-1278.
Neubert, D; Tapken, S. (1988) Transfer of benzo(a)pyrene into mouse embryos and fetuses. Arch Toxicol 62(2-
3):236-239.
Newbold, RF; Brookes, P. (1976) Exceptional mutagenicity of a benzo(a)pyrene diol epoxide in cultured
mammalian cells. Nature 261(5555):52-54.
Ng, KM; Chu, I; Bronaugh, RL; et al. (1992) Percutaneous absorption and metabolism of pyrene, benzo[a]pyrene,
and di(2-ethylhexyl) phthalate: comparison of in vitro and in vivo results in the hairless guinea pig. Toxicol Appl
Pharmacol 115(2):216-223.
Nikonova, TV. (1977) [Transplacental effect of benz(a)pyrene and pyrene]. Biulleten Eksperimentalnoi Biologii i
Meditsiny 84(7): 88-91.
Nikoyan, A; De, MM; Sari-Minodier, I; et al. (2007) Evaluation of a battery of Salmonella typhimurium tester
strains for biomonitoring of mutagenic polycyclic aromatic hydrocarbons, nitroarenes and aromatic amines.
Mutation Research 626(1 -2):88—101.
Niwa, Y ; Terashima, T; Sumi, H. (2003). Topical application of the immunosuppressant tacrolimus accelerates
carcinogenesis in mouse skin. Br J Dermatol. 149(5):960-7.
Nordling, MM; Nygren, J; Bergman, J; et al. (2002) Toxicological characterization of a novel in vivo
benzo[a]pyrene metabolite, 7-oxo-benz[d]anthracene-3,4-dicarboxylic acid anhydride. Chem Res Toxicol
15(10): 1274—1280.
271
DRAFT - DO NOT CITE OR QUOTE
-------
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press.
NTP (National Toxicology Program). (2005) 11th Report on carcinogens. Public Health Service, U.S. Department of
Health and Human Services, Research Triangle Park, NC.
Nwagbara, 0; Darling-Reed, SF; Tucker, A; et al. (2007) Induction of cell death, DNA strand breaks, and cell cycle
arrest in DU145 human prostate carcinoma cell line by benzo[a]pyrene and benzo[a]pyrene-7,8-diol-9,10-epoxide.
Int J Environ Res Public Health 4(1): 10-14.
Obana, H; Hori, S; Kashimoto, T; et al. (1981) Polycyclic aromatic hydrocarbons in human fat and liver. Bull
Environ Contam Toxicol 27(l):23-27.
Oesch, F. (1980) Microsomal epoxide hydrolase. In: Enzymatic basis of detoxification. Academic Press; pp. 277-
290.
Okey, AB; Riddick, DS; Harper, PA. (1994) Molecular biology of the aromatic hydrocarbon (dioxin) receptor.
Trends Pharmacol Sci 15(7):226-232.
O'Neil, MJ; Smith, A; Heckelman, PE; et al. (2001) The Merck index: An encyclopedia of chemicals, drugs, and
biologicals. 13th ed. Whitehouse Station, NJ: Merck & Co., Inc. .
O'Neill, IK; Povey, AC; Bingham, S; et al. (1990) Systematic modulation by human diet levels of dietary fibre and
beef on metabolism and disposition of benzo[a]pyrene in the gastrointestinal tract of Fischer F344 rats.
Carcinogenesis 11 (4):609—616.
O'Neill, IK; Goldberg, MT; el Ghissassi, F; et al. (1991) Dietary fibre, fat and beef modulation of colonic nuclear
aberrations and microcapsule-trapped gastrointestinal metabolites of benzo[a]pyrene-treated C57/B6 mice
consuming human diets. Carcinogenesis 12(2): 175-180.
Oueslati, R; Alexandrov, K; Chouikha, M; et al. (1992) Formation and persistence of DNA adducts in epidermal and
dermal mouse skin exposed to benzo(a)pyrene in vivo. In Vivo 6(2):231-235.
Paleologo, M; Van Schooten, FJ; Pavanello, S; et al. (1992) Detection of benzo[a]pyrene-diol-epoxide-DNA adducts
in white blood cells of psoriatic patients treated with coal tar. Mutat Res 281(1): 11—16.
Pan, G; Hanaoka, T; Yamano, Y; et al. (1998) A study of multiple biomarkers in coke oven workers--a cross-
sectional study in China. Carcinogenesis 19(11): 1963-1968.
Park, SJ; Zhao, H; Spitz, MR; et al. (2003) An association between NQOl genetic polymorphism and risk of bladder
cancer. Mutat Res 536(1—2): 131—137.
Park, JH; Troxel, AB; Harvey, RG; et al. (2006a) Polycyclic aromatic hydrocarbon (PAH) o-quinones produced by
the aldo-keto-reductases (AKRs) generate abasic sites, oxidized pyrimidines, and 8-oxo-dGuo via reactive oxygen
species. Chem Res Toxicol 19(5):719—728.
Park, JH; Gelhaus, S; Vedantam, S; et al. (2008) The pattern of p53 mutations caused by PAH o-quinones is driven
by 8-oxo-dGuo formation while the spectrum of mutations is determined by biological selection for dominance.
Chem Res Toxicol 21(5): 1039-1049.
Parman, T; Wells, PG. (2002) Embryonic prostaglandin H synthase-2 (PHS-2) expression and benzo[a]pyrene
teratogenicity in PHS-2 knockout mice. FASEB J 16(9): 1001-1009.
Partanen, T; Boffetta, P. (1994) Cancer risk in asphalt workers and roofers: review and meta-analysis of
epidemiologic studies. Am J Ind Med 26(6):721-740.
272
DRAFT - DO NOT CITE OR QUOTE
-------
Pastorelli, R; Guanci, M; Cerri, A; et al. (1998) Impact of inherited polymorphisms in glutathione S-transferase Ml,
microsomal epoxide hydrolase, cytochrome P450 enzymes on DNA, and blood protein adducts of benzo(a)pyrene-
diolepoxide. Cancer Epidemiol Biomarkers Prev 7(8):703-709.
Pataki, J; Huggins, C. (1969) Molecular site of substituents of benz(a)anthracene related to carcinogenicity. Cancer
Res 29(3):506-509.
Pavanello, S; Favretto, D; Brugnone, F; et al. (1999) HPLC/fluorescence determination of anti-BPDE-DNA adducts
in mononuclear white blood cells from PAH-exposed humans. Carcinogenesis 20(3):431-435.
Pavanello, S; Siwinska, E; Mielzynska, D; et al. (2004) GSTM1 null genotype as a risk factor for anti-BPDE-DNA
adduct formation in mononuclear white blood cells of coke-oven workers. Mutat Res 558(1-2):53-62.
Pavanello, S; Pulliero, A; Siwinska, E; et al. (2005) Reduced nucleotide excision repair and GSTMl-null genotypes
influence anti-B[a]PDE-DNA adduct levels in mononuclear white blood cells of highly PAH-exposed coke oven
workers. Carcinogenesis 26(1): 169-175.
Pavanello, S; Pulliero, A; Saia, BO; et al. (2006) Determinants of anti-benzo[a]pyrene diol epoxide-DNA adduct
formation in lymphomonocytes of the general population. Mutat Res 611(1 -2): 54—63.
Peluso, M; Merlo, F; Munnia, A; et al. (1998) 32P-postlabeling detection of aromatic adducts in the white blood cell
DNA of nonsmoking police officers. Cancer Epidemiol Biomarkers Prev 7(1):3—11.
Penning, TM; Ohnishi, ST; Ohnishi, T; et al. (1996) Generation of reactive oxygen species during the enzymatic
oxidation of polycyclic aromatic hydrocarbon trans-dihydrodiols catalyzed by dihydrodiol dehydrogenase. Chem
Res Toxicol 9(l):84-92.
Penning, TM; Burczynski, ME; Hung, CF; et al. (1999) Dihydrodiol dehydrogenases and polycyclic aromatic
hydrocarbon activation: generation of reactive and redox active o-quinones. Chem Res Toxicol 12(1): 1-18.
Pereira, MA; McMillan, L; Kaur, P; et al. (1982) Effect of benzo[a]pyrene on sister-chromatid exchange in fetal
hamster liver exposed in utero. Mutat Res 105:343-347.
Perera, FP; Hemminki, K; Young, TL; et al. (1988) Detection of polycyclic aromatic hydrocarbon-DNA adducts in
white blood cells of foundry workers. Cancer Res 48(8):2288-2291.
Perera, FP; Tang, DL; O'Neill, JP; et al. (1993) HPRT and glycophorin A mutations in foundry workers: relationship
to PAH exposure and to PAH-DNA adducts. Carcinogenesis 14(5): 969-973.
Perera, FP; Dickey, C; Santella, R; et al. (1994) Carcinogen-DNA adducts and gene mutation in foundry workers
with low-level exposure to polycyclic aromatic hydrocarbons. Carcinogenesis 15(12):2905-2910.
Perera, FP; Weinstein, IB. (2000) Molecular epidemiology: recent advances and future directions. Carcinogenesis
21(3):517—524.
Perera, FP; Tang, D; Rauh, V; et al. (2005a) Relationships among polycyclic aromatic hydrocarbon-DNA adducts,
proximity to the World Trade Center, and effects on fetal growth. Environ Health Perspect 113(8): 1062-1067.
Perera, FP; Rauh, V; Whyatt, RM; et al. (2005b) A summary of recent findings on birth outcomes and
developmental effects of prenatal ETS, PAH, and pesticide exposures. Neurotoxicology 26(4):573-587.
Perera FP, Li Z, Whyatt R, Hoepner L, et al., (2009) Prenatal airborne polycyclic aromatic hydrocarbon exposure
and child IQ at age 5 years. Pediatrics. 124(2):el95-202.
Persson, E; Larsson, P; Tjalve, H. (2002) Cellular activation and neuronal transport of intranasally instilled
benzo(a)pyrene in the olfactory system of rats. Toxicol Lett 133(2—3):211-219.
Petrie, HT. (2002) Role of thymic organ structure and stromal composition in steady-state postnatal t-cell
production. Immunol Rev 189:8-19.
273
DRAFT - DO NOT CITE OR QUOTE
-------
Petridou-Fischer, J; Whaley, SL; Dahl, AR. (1988) In vivo metabolism of nasally instilled benzo[a]pyrene in dogs
and monkeys. Toxicology 48(1 ):31—40.
Petry, T; Schmid, P; Schlatter, C. (1996) The use of toxic equivalency factors in assessing occupational and
environmental health risk associated with exposure to airborne mixtures of polycyclic aromatic hydrocarbons
(PAHs). Chemosphere 32(4):639-648.
Pfeiffer, EH. (1977) Oncogenic interaction of carcinogenic and non-carcinogenic polycyclic aromatic hydrocarbons
in mice. IARC Sci Publ 16:69-77.
Pfeiffer, CA; Allen, E. (1948) Attempts to produce cancer in Rhesus monkeys with carcinogenic hydrocarbons and
estrogens. Cancer Res 8(3):97-127.
Phillips, DH. (2002) Smoking-related DNA and protein adducts in human tissues. Carcinogenesis 23(12): 1979—
2004.
Poel, WE. (1959) Effect of carcinogenic dosage and duration of exposure on skin-tumor induction in mice. J Natl
Cancer Inst 22(1): 19-43.
Poel, WE. (1960) Skin as a test site for the bioassay of carcinogens and carcinogen precursors. J Natl Cancer Inst
Monogr 10:611-631.
Poirier, MC; Santella, RM; Weston, A. (2000) Carcinogen macromolecular adducts and their measurement.
Carcinogenesis 21 (3):353—359.
Porter, PC; Mellon, I; States, JC. (2005) XP-A cells complemented with Arg228Gln and Val234Leu polymorphic
XPA alleles repair BPDE-induced DNA damage better than cells complemented with the wild type allele. DNA
Repair (Amst) 4(3):341-349.
Potter, D; Booth, ED; Brandt, HC; et al. (1999) Studies on the dermal and systemic bioavailability of polycyclic
aromatic compounds in high viscosity oil products. Arch Toxicol 73(3): 129-140.
Puisieux, A; Lim, S; Groopman, J; et al. (1991) Selective targeting of p53 gene mutational hotspots in human
cancers by etiologically defined carcinogens. Cancer Res 51 (22):6185—6189.
Qian, HW; Ong, T; Nath, J; et al. (1998) Induction of DNA adducts in vivo in rat lung cells by fume condensates of
roofing asphalt. Teratogenesis, Carcinogenesis, and Mutagenesis 18(3): 131—140.
Rahman, A; Barrowman, JA; Rahimtula, A. (1986) The influence of bile on the bioavailability of polynuclear
aromatic hydrocarbons from the rat intestine. Can J Physiol Pharmacol 64(9): 1214-1218.
Ramesh, A; Inyang, F; Hood, DB; et al. (2000) Aryl hydrocarbon hydroxylase activity in F-344 rats subchronically
exposed to benzo(a)pyrene and fluoranthene through diet. JBiochemMol Toxicol 14(3): 155—161.
Ramesh, A; Greenwood, M; Inyang, F; et al. (2001a) Toxicokinetics of inhaled benzo[a]pyrene: plasma and lung
bioavailability. Inhal Toxicol 13(6):533—555.
Ramesh, A; Inyang, F; Hood, DB; et al. (2001b) Metabolism, bioavailability, and toxicokinetics of
benzo(alpha)pyrene in F-344 rats following oral administration. Exp Toxicol Pathol 53(4):275-290.
Ramesh, A; Greenwood, M; Inyang, F; et al. (2003) Aryl hydrocarbon hydroxylase (AHH) and benzo(a)pyrene
metabolite in F-344 rats subchronically exposed to inhaled benzo[a]pyrene. Toxicologist 72(S-1):325.
Ramesh, A; Walker, SA; Hood, DB; et al. (2004) Bioavailability and risk assessment of orally ingested polycyclic
aromatic hydrocarbons. Int J Toxicol 23(5):301—333.
Ramesh, A; Knuckles, ME. (2006) Dose-dependent benzo(a)pyrene [B(a)P]-DNA adduct levels and persistence in
F-344 rats following subchronic dietary exposure to B(a)P. Cancer Lett 240:268-278.
274
DRAFT - DO NOT CITE OR QUOTE
-------
Ramesh, A; Inyang, F; Lunstra, DD; et al. (2008) Alteration of fertility endpoints in adult male F-344 rats by
subchronic exposure to inhaled benzo(a)pyrene. Exp Toxicol Pathol 60(4-5):269-280.
Ramos, KS; Moorthy, B. (2005) Bioactivation of polycyclic aromatic hydrocarbon carcinogens within the vascular
wall: Implications for human atherogenesis. Drug Metab Rev 37(4):595-610.
Rask-Nielson, R. (1950) The susceptibility of the thymus, lung, subcutaneous and mammary tissues in strain street
mice to direct application of small doses of four different carcinogenesis hydrocarbons. Br J Cancer 4:108-116.
Raveh, D; Slaga, TJ; Huberman, E. (1982) Cell-mediated mutagenesis and tumor-initiating activity of the ubiquitous
polycyclic hydrocarbon, cyclopenta[c,d]pyrene. Carcinogenesis 3(7):763-766.
Raychoudhury, SS; Kubinski, D. (2003) Polycyclic aromatic hydrocarbon-induced cytotoxicity in cultured rat
Sertoli cells involves differential apoptotic response. Environ Health Perspect 111 (1 ):33—38.
Reardon, KF; Liao, KH; Yang, RS. (2005) Development of a hybrid reaction network physiologically based
pharmacokinetic model of benzo(a)pyrene and its metabolites. Toxicol Sci 84(1 -S): 177.
Reddy, MV; Gupta, RC; Randerath, E; et al. (1984) 32P-postlabeling test for covalent DNA binding of chemicals in
vivo: application to a variety of aromatic carcinogens and methylating agents. Carcinogenesis 5:231-243.
Renzik-Schuller, H; Mohr, U. (1974) Investigations on the carcinogenic burden by air pollution in man. IX. Early
pathological alterations of the bronchial epithelium in Syrian golden hamsters after intratracheal instillation of benzo
(a) pyrene. 1. Morphological studies from semi-thin sections. Zentralbl Bakteriol [Orig B] 159(5-6):493-502.
Revel, A; Raanani, H; Younglai, E; et al. (2001) Resveratrol, a natural aryl hydrocarbon receptor antagonist,
protects sperm from DNA damage and apoptosis caused by benzo(a)pyrene. Reprod Toxicol 15(5):479-486.
Rice, JE; Makowski, GS; Hosted, TJ, Jr.; et al. (1985) Methylene-bridged bay region chrysene and phenanthrene
derivatives and their keto-analogs: mutagenicity in Salmonella typhimurium and tumor-initiating activity on mouse
skin. Cancer Lett 27(2): 199-206.
Rigdon, RH; Neal, J. (1965) Effects of feeding benzo(a)pyrene on fertility, embryos, and young mice. J Natl Cancer
Inst 34:297-305.
Rigdon, RH; Neal, J. (1966) Gastric carcinomas and pulmonary adenomas in mice fed benzo (a) pyrene. Tex Rep
Biol Med 24(2): 195-207.
Rigdon, RH; Neal, J. (1969) Relationship of leukemia to lung and stomach tumors in mice fed benzo (a) pyrene.
Proc Soc Exp Biol Med 130(1): 146-148.
Rigdon, RH; Rennels, EG. (1964) Effect of feeding benzpyrene on reproduction in the rat. Experientia 20(4):224-
226.
Robertson, IG; Guthenberg, C; Mannervik, B; et al. (1986) Differences in stereoselectivity and catalytic efficiency
of three human glutathione transferases in the conjugation of glutathione with 7 beta,8 alpha-dihydroxy-9 alpha, 10
alpha-oxy-7,8,9,10-tetrahydrobenzo(a)pyrene. Cancer Res 46(5):2220-2224.
Robinson, DJ; de Bruijn, HS; van der Veen, N; et al. (1998). Fluorescence photobleaching of ala-induced
protoporphyrin ix during photodynamic therapy of normal hairless mouse skin: the effect of light dose and
lrradiance and the resulting biological effect. Photochemistry and Photobiology 67(1): 140-1 49.
Robinson, M; Laurie, RD; Bull, RJ; et al. (1987) Carcinogenic effects in A/J mice of particulate of a coal tar paint
used in potable water systems. Cancer Lett 34(l):49-54.
Rodriguez, LV; Dunsford, HA; Steinberg, M; et al. (1997) Carcinogenicity of benzo[ajpyrene and manufactured gas
plant residues in infant mice. Carcinogenesis 18(1): 127-135.
275
DRAFT - DO NOT CITE OR QUOTE
-------
Rodriguez, JW; Kirlin, WG; Wirsiy, YG; et al. (1999) Maternal exposure to benzo[a]pyrene alters development of t
lymphocytes in offspring. Immunopharmacol Immunotoxicol 21(2):379-396.
Roe, FJ; Waters, MA. (1967) Induction of hepatoma in mice by carcinogens of the polycyclic hydrocarbon type.
Nature 214(5085):299-300.
Roe, FJ; Peto, R; Kearns, F; et al. (1970) The mechanism of carcinogenesis by the neutral fraction of cigarette
smoke condensate. Br J Cancer 24(4):788-806.
Rogan, EG; Devanesan, PD; RamaKrishna, NV; et al. (1993) Identification and quantitation of benzo[a]pyrene-
DNA adducts formed in mouse skin. Chem Res Toxicol 6(3):356-363.
Rogan, EG; RamaKrishna, NV; Higginbotham, S; et al. (1990) Identification and quantitation of 7-(benzo[a]pyren-
6-yl)guanine in the urine and feces of rats treated with benzo[a]pyrene. Chem Res Toxicol 3(5):441^144.
Rojas, M; Alexandrov, K; Cascorbi, I; et al. (1998) High benzo[a]pyrene diol-epoxide DNA adduct levels in lung
and blood cells from individuals with combined CYP1A1 MspI/Msp-GSTMl*0/*0 genotypes. Pharmacogenetics
8(2): 109-118.
Rojas, M; Cascorbi, I; Alexandrov, K; et al. (2000) Modulation of benzo[a]pyrene diolepoxide-DNA adduct levels
in human white blood cells by CYP1A1, GSTM1 and GSTT1 polymorphism. Carcinogenesis 21(1):35—41.
Rojas, M; Godschalk, R; Alexandrov, K; et al. (2001) Myeloperoxidase--463A variant reduces benzo[a]pyrene diol
epoxide DNA adducts in skin of coal tar treated patients. Carcinogenesis 22(7): 1015-1018.
Romundstad, P; Haldorsen, T; Andersen, A. (2000a) Lung and bladder cancer among workers in a Norwegian
aluminium reduction plant. Occup Environ Med 57(7):495^99.
Romundstad, P; Andersen, A; Haldorsen, T. (2000b) Cancer incidence among workers in six Norwegian aluminum
plants. Scand J Work Environ Health 26(6):461-469.
Ross, JA; Nesnow, S. (1999) Polycyclic aromatic hydrocarbons: correlations between DNA adducts and ras
oncogene mutations. Mutat Res 424(1-2): 155-166.
Ross, J; Nelson, G; Kligerman, A; et al. (1990) Formation and persistence of novel benzo(a)pyrene adducts in rat
lung, liver, and peripheral blood lymphocyte DNA. Cancer Res 50(16):5088-5094.
Ross, J; Nelson, G; Erexson, G; et al. (1991) DNA adducts in rat lung, liver and peripheral blood lymphocytes
produced by i.p. administration of benzo[a]pyrene metabolites and derivatives. Carcinogenesis 12(10): 1953-1955.
Ross, JA; Nelson, GB; Wilson, KH; et al. (1995) Adenomas induced by polycyclic aromatic hydrocarbons in strain
A/J mouse lung correlate with time-integrated DNA adduct levels. Cancer Res 55(5): 1039-1044.
Rossella, F; Campo, L; Pavanello, S; et al. (2009) Urinary polycyclic aromatic hydrocarbons and monohydroxy
metabolites as biomarkers of exposure in coke-oven workers. Occup Environ Med Epub ahead of print 15 Feb 2009
Roth, RA; Vinegar, A. (1990) Action by the lungs on circulating xenobiotic agents, with a case study of
physiologically based pharmacokinetic modeling of benzo(a)pyrene disposition. Pharmacol Ther 48(2): 143—155.
Russell, LB. (1977) Validation of the in vivo somatic mutation method in the mouse as a prescreen for germinal
point mutations. Arch Toxicol 38(l-2):75-85.
Saffiotti, U; Cefis, F; Kolb, LH. (1968) A method for the experimental induction of bronchogenic carcinoma.
Cancer Res 28(1): 104-124.
Saffiotti, U; Montesano, R; Sellakumar, AR; et al. (1972) Respiratory tract carcinogenesis induced in hamsters by
different dose levels of benzo-(a)pyrene and ferric oxide. J Natl Cancer Inst 49(4): 1199-1204.
Sagredo, C; Ovrebo, S; Haugen, A; et al. (2006) Quantitative analysis of benzo[a]pyrene biotransformation and
adduct formation in Ahr knockout mice. Toxicol Lett 167(3): 173-182.
276
DRAFT - DO NOT CITE OR QUOTE
-------
Salama, SA; Sierra-Torres, CH; Oh, HY; et al. (2001) Variant metabolizing gene alleles determine the genotoxicity
of benzo[a]pyrene. Environ Mol Mutagen 37:17-26.
Salmon, AG; Brown, JP; Collins, JF; et al. (2003) Evaluation of alternate dose metrics for benzo(a)pyrene inhalation
cancer potency using a nasal-PBPK hamster model. Toxicologist 72:85.
Santella, RM; Nunes, MG; Blaskovic, R; et al. (1994) Quantitation of polycyclic aromatic hydrocarbons, 1-
hydroxypyrene, and mutagenicity in urine of coal tar-treated psoriasis patients and untreated volunteers. Cancer
Epidemiol Biomarkers Prev 3(2): 137—140.
Saunders, CR; Shockley, DC; Knuckles, ME. (2001) Behavioral effects induced by acute exposure to
benzo[a]pyrene in F-344 rats. Neurotox Res 3:557-579.
Saunders, CR; Ramesh, A; Shickley, DC. (2002) Modulation of neurotoxic behavior of F-344 rats by temporal
disposition of benzo[a]pyrene. Toxicol Lett 129:33-45.
Saunders, CR; Das, SK; Ramesh, A; et al. (2006) Benzo(a)pyrene-induced acute neurotoxicity in the F-344 rat: role
of oxidative stress. J Appl Toxicol 26(5):427^138.
Scherer, G; Frank, S; Riedel, K; et al. (2000) Biomonitoring of exposure to polycyclic aromatic hydrocarbons of
nonoccupationally exposed persons. Cancer Epidemiol Biomarkers Prev 9(4):373-380.
Schlede, E; Kuntzman, R; Haber, S; et al. (1970) Effect of enzyme induction on the metabolism and tissue
distribution of benzo(alpha)pyrene. Cancer Res 30(12):2893-2897.
Schmahl, D; Schmidt, KG; Habs, M. (1977) Syncarcinogenic action of polycyclic hydrocarbons in automobile
exhaust gas condensates. IARC Sci Publ (16):53—59.
Schmidt, KG; Schmael, D; Misfield, J. (1973) Investigations on the carcinogenic burden in man. VI. Experimental
investigation to determine a dose-response relationship and to estimate a threshold dose of benzo[a]pyrene in the
skin of two different mouse strains. Zbl Bakt Hyg I Abs B 158:62-68.
Schnizlein, CT; Munson, AE; Rhoades, RA. (1987) Immunomodulation of local and systemic immunity after
subchronic pulmonary exposure of mice to benzo(a)pyrene. Int J Immunopharmacol 9(1 ):99—106.
Schoket, B; Doty, WA; Vincze, I; et al. (1993) Increased sensitivity for determination of polycyclic aromatic
hydrocarbon-DNA adducts in human DNA samples by dissociation-enhanced lanthanide fluoroimmunoassay
(DELFIA). Cancer Epidemiol Biomarkers Prev 2(4):349-353.
Schonland, SO; Zimmer, JK; Lopez-Benitez, CM; et al. (2003) Homeostatic control of t-cell generation in neonates.
Blood 102(4): 1428-1434.
Schuurman, HJ; Van Loveren, H; Rozing, J; et al. (1992) Chemicals trophic for the thymus: Risk for
immunodeficiency and autoimmunity. Int J Immunopharmacol 14(3):369-375.
Schwarz, D; Kisselev, P; Cascorbi, I; et al. (2001) Differential metabolism of benzo[a]pyrene and benzo[a]pyrene-
7,8-dihydrodiol by human CYP1 Al variants. Carcinogenesis 22(3):453-459.
Sega, GA. (1979) Unscheduled DNA synthesis (DNA repair) in the germ cells of male mice-its role in the study of
mammalian mutagenesis. Genetics 92:s49-58.
Sharma, R; Haque, AK; Awasthi, S; et al. (1997) Differential carcinogenicity of benzo[a]pyrene in male and female
CD-I mouse lung. J Toxicol Environ Health 52(l):45-62.
Sharovskaya, J; Kobliakova, I; Solomatina, N; et al. (2006) Effect of some carcinogenic and non-carcinogenic
polycyclic aromatic hydrocarbons on gap junction intercellular communication in hepatoma cell cultures. Eur J Cell
Biol 85(5):387-397.
277
DRAFT - DO NOT CITE OR QUOTE
-------
Shen, H; Spitz, MR; Qiao, Y; et al. (2003) Smoking, DNA repair capacity and risk of nonsmall cell lung cancer. Int
J Cancer 107(l):84-88.
Shen, YM; Troxel, AB; Vedantam, S; et al. (2006) Comparison of p53 mutations induced by PAH o-quinones with
those caused by anti-benzo[a]pyrene diol epoxide in vitro: role of reactive oxygen and biological selection. Chem
Res Toxicol 19(11): 1441-1450.
Shendrikova, A; Aleksandrov, VA. (1974) Comparative penetration of polycyclic hydrocarbons through the rat
placenta into the fetus. Bull Exp Biol Med 77:169-171.
Sherson, D; Sabro, P; Sigsgaard, T; et al. (1990) Biological monitoring of foundry workers exposed to polycyclic
aromatic hydrocarbons. Br J Ind Med 47(7):448-453.
Shimizu, Y; Nakatsuru, Y; Ichinose, M; et al. (2000) Benzo[a]pyrene carcinogenicity is lost in mice lacking the aryl
hydrocarbon receptor. Proc Natl Acad Sci USA 97(2):779-782.
Shukla, Y, Yadav, A, Arora, A. (2002). Carcinogenic and cocarcinogenic potential of cypermethrin on mouse skin.
Cancer Letters 182:33-41.
Shultz, MA; Morin, D; Chang, AM; et al. (2001) Metabolic capabilities of CYP2F2 with various pulmonary
toxicants and its relative abundance in mouse lung subcompartments. J Pharmacol Exp Ther 296(2):510-519.
Shvedova, A; Kisin, ER; Murray, AR; et al. (2002). Antioxidant Balance and Free Radical Generation in Vitamin
E . Chem. Res. Toxicol., 2002, 15 (11), pp 1451-1459.
Silverstone, AE; Cantor, H; Goldstein, G; et al. (1976) Terminal deoxynucleotidyl transferase is found in
prothymocytes. J Exp Med 144(2):543-548.
Singh, VK; Singh, J; Anand, M; et al. (2008) Comparison of polycyclic aromatic hydrocarbon levels in placental
tissues of Indian women with full- and preterm deliveries. Int J Hyg Environ Health 211:63 9-647.
Sinha, R; Kulldorff, M; Gunter, MJ; et al. (2005) Dietary benzo[a]pyrene intake and risk of colorectal adenoma.
Cancer Epidemiol Biomarkers Prev 14(8):2030-2034.
Sivak A, Niemeier R, Lynch D, et al., (1997). Skin carcinogenicity of condensed asphalt roofing fumes and their
fractions following dermal application to mice. Cancer Lett. 117(1): 113-23.
Slaga, TJ; Bracken, WM; Viaje, A; et al. (1977) Comparison of the tumor-initiating activities of benzo(a)pyrene
arene oxides and diol-epoxides. Cancer Res 37(11):4130—4133.
Slaga, TJ; Huberman, E; Selkirk, JK; et al. (1978) Carcinogenicity and mutagenicity of benz(a)anthracene diols and
diol-epoxides. Cancer Res 38(6): 1699-1704.
Slaga, TJ; Iyer, RP; Lyga, W; et al. (1980) Comparison of the skin tumor-initiating activities of dihydrodiols, diol-
epoxides, and methylated derivatives of various polycyclic aromatic hydrocarbons. In: Bjorseth, A; Dennis, AJ; eds.
Polynuclear aromatic hydrocarbons: chemistry and biological effects. Columbus, OH: Battelle Press; pp. 753-769.
Soares, SR; Melo, MA. 2008. Cigarette smoking and reproductive function. Curr Opin Obstet Gynecol. 20:281-91.
Soyka, LF. (1980) Hepatic drug metabolizing enzyme activity and tumorigenesis in mice following perinatal
exposure to benzo(a)pyrene. Pediatr Pharmacol 1(1):85—96.
Spinelli, JJ; Band, PR; Svirchev, LM; et al. (1991) Mortality and cancer incidence in aluminum reduction plant
workers. J Occup Med 33(11): 1150—1155.
Spinelli, JJ; Demers, PA; Le, ND; et al. (2006) Cancer risk in aluminum reduction plant workers (Canada). Cancer
Causes Control 17(7):939-948.
278
DRAFT - DO NOT CITE OR QUOTE
-------
Stansbury, KH; Noll, DM; Groopman, JD; et al. (2000) Enzyme-mediated dialdehyde formation: an alternative
pathway for benzo[a]pyrene 7,8-dihydrodiol bioactivation. Chem Res Toxicol 13(11): 1174—1180.
Stavric, B; Klassen, R. (1994) Dietary effects on the uptake of benzo[a]pyrene. Food Chem Toxicol 32(8):727-734.
Steiner, PE. (1955) Carcinogenicity of multiple chemicals simultaneously administered. Cancer Res 15(9):632-635.
Straif, K; Baan, R; Grosse, Y; et al. (2005) Carcinogenicity of polycyclic aromatic hydrocarbons. Lancet Oncology
6:931-932.
Sugiyama, T. (1973) Chromosomal aberrations and carcinogenesis by various benz(a)anthracene derivatives. Gann
64(6):637-639.
Sun, JD; Wolff, RK; Kanapilly, GM. (1982) Deposition, retention, and biological fate of inhaled benzo(a)pyrene
adsorbed onto ultrafine particles and as a pure aerosol. Toxicol Appl Pharmacol 65(2):231-244.
Surh, YJ; Tannenbaum, SR. (1995) Sulfotransferase-mediated activation of 7,8,9,10-tetrahydro-7-ol, 7,8-
dihydrodiol, and 7,8,9,10-tetraol derivatives of benzo[a]pyrene. Chem Res Toxicol 8(5):693-698.
Swartz, WJ; Mattison, DR.(1985) Benzo(a)pyrene inhibits ovulation in C57BL/6N mice. Anat Rec 212:268-276.
Taioli, E; Sram, RJ; Binkova, B; et al. (2007) Biomarkers of exposure to carcinogenic PAHs and their relationship
with environmental factors. Mutat Res 620(1 -2): 16-21.
Takizawa, K; Yagi, H; Jerina, DM; Mattison, DR. Murine strain differences in ovotoxicity following intraovarian
injection with benzo(a)pyrene, (+)-(7R,8S),-oxide, (-)-(7R,*R)-dihydrodiol, or (+)-(7R,8S)-diol-(9S,10R)-epoxide-
2. Cancer Res 44:2571-2576.
Talaska, G; Jaeger, M; Reilman, R; et al. (1996) Chronic, topical exposure to benzo[a]pyrene induces relatively high
steady-state levels of DNA adducts in target tissues and alters kinetics of adduct loss. Proc Natl Acad Sci USA
93(15):7789-7793.
Talaska, G; Ginsburg, D; LaDow, K; et al. (2006) Impact of Cypla2 or Ahr gene knockout in mice: implications for
biomonitoring studies. Toxicol Lett 162(2-3):246-249.
Tamaki, A; Hayashi, H; Nakajima, H; et al. (2004) Polycyclic aromatic hydrocarbon increases mRNA level for
interleukin 1 beta in human fibroblast-like synoviocyte line via aryl hydrocarbon receptor. Biol Pharm Bull
27(3):407-410.
Tang, D; Li, TY; Liu, JJ; et al. (2006) PAH-DNA adducts in cord blood and fetal and child development in a
Chinese cohort. Environ Health Perspect 114(8): 1297—1300.
Tang, D; Li, TY; Liu, JJ, et al. (2008) Effects of prenatal exposure to coal-burning pollutants on children's
development in China. Environ Health Perspect 116:674-679.
Tarantini, A; Maitre, A; Lefebvre, E; et al. (2009) Relative contribution of DNA strand breaks and DNA adducts to
the genotoxicity of benzo[a]pyrene as a pure compound and in complex mixtures. Mutat Res 671:67-75.
Terracini, B; Shubik, P; Dellaporta, A. (1960). A Study of Skin Carcinogenesis in the Mouse with Single
Applications of 9,10-Dimethyl-l,2-benzanthracene at Different Dosages. Cancer Research Vol. 20:1538-1542.
Theriault, G; Tremblay, C; Cordier, S; et al. (1984) Bladder cancer in the aluminium industry. Lancet 1(8383):947-
950.
Thirman, MJ; Albrecht, JH; Krueger, MA; et al. (1994) Induction of cytochrome CYPIA1 and formation of toxic
metabolites of benzo[a]pyrene by rat aorta: a possible role in atherogenesis. Proc Natl Acad Sci U S A
91 (12):5397—5401.
279
DRAFT - DO NOT CITE OR QUOTE
-------
Thyssen, J; Althoff, J; Kimmerle, G; et al. (1981) Inhalation studies with benzo[a]pyrene in Syrian golden hamsters.
J Natl Cancer Inst 66(3):575-577.
Topping, DC; Martin, DH; Nettesheim, P. (1981) Determination of cocarcinogenic activity of benzo[e]pyrene for
respiratory tract mucosa. Cancer Lett 11 (4):315—321.
Tremblay, C; Armstrong, B; Theriault, G; et al. (1995) Estimation of risk of developing bladder cancer among
workers exposed to coal tar pitch volatiles in the primary aluminum industry. Am J Ind Med 27(3):335-348.
Temple, L; Kawabata, TT; Munson, AE; et al. (1993) Comparison of ELISA and plaque-forming cell assays for
measuring the humoral immune response to SRBC in rats and mice treated with benzo[a]pyrene or
cyclophosphamide. Fundam Appl Toxicol 21(4):412-419.
Triolo, AJ; Aponte, GE; Herr, DL. (1977) Induction of aryl hydrocarbon hydroxylase and forestomach tumors by
benzo(a)pyrene. Cancer Res 37:3018-3021.
Tsuruda, L; Hou, Y; Penning, TM. (2001) Stable expression of rat dihydrodiol dehydrogenase (AKR1C9) in human
breast MCF-7 cells results in the formation of PAH-o-quinones and enzyme mediated cell death. Chem Res Toxicol
14(7):856-862.
Tuntawiroon, J; Mahidol, C; Navasumrit, P; et al. (2007) Increased health risk in Bangkok children exposed to
polycyclic aromatic hydrocarbons from traffic-related sources. Carcinogenesis 28(4):816-822.
Turusov, VS; Nikonova, TV; Parfenov, Y. (1990) Increased multiplicity of lung adenomas in five generations of
mice treated with benz(a)pyrene when pregnant. Cancer Lett 55(3):227-231.
Tuttle, AM; Stampfli, M; Foster, WG. (2009) Cigarette smoke causes follicle loss i nmice ovaries at concentrations
representative of human exposure. Human Repro 24:1452-1459.
Uno, S; Dalton, TP; Shertzer, HG; et al. (2001) Benzo[a]pyrene-induced toxicity: paradoxical protection in
Cyplal(-/-) knockout mice having increased hepatic benzo[a]pyrene-DNA adduct levels. Biochem Biophys Res
Commun 289:1049-1056.
Uno, S; Dalton, TP; Dragin, N; et al. (2006) Oral benzo[a]pyrene in Cypl knockout mouse lines: CYP1 Al
important in detoxication, CYP1B1 metabolism required for immune damage independent of total-body burden and
clearance rate. Mol Pharmacol 69(4): 1103-1114.
Urso, P; Gengozian, N. (1980) Depressed humoral immunity and increased tumor incidence in mice following in
utero exposure to benzo[alpha]pyrene. J Toxicol Environ Health 6(3):569-576.
Urso, P; Gengozian, N. (1982) Alterations in the humoral immune response and tumor frequencies in mice exposed
to benzo[a]pyrene and x-rays before or after birth. J Toxicol Environ Health 10(4-5):817-835.
Urso, P; Gengozian, N. (1984) Subnormal expression of cell-mediated and humoral immune responses in progeny
disposed toward a high incidence of tumors after in utero exposure to benzo[a]pyrene. J Toxicol Environ Health
14(4):569-584.
Urso, P; Johnson, RA. (1987) Early changes in t lymphocytes and subsets of mouse progeny defective as adults in
controlling growth of a syngeneic tumor after in utero insult with benzo(a)pyrene. Immunopharmacology 14(1): 1 -
10.
Urso, P; Johnson, RA. (1988) Quantitative and functional change in T cells of primiparous mice following injection
of benzo(a)pyrene at the second trimester of pregnancy. Immunopharmacol Immunotoxicol 10(2): 195—217.
Urso, P; Ryan, MC; Bennett, JS. (1988) Changes in peripheral blood cells in mice after injection with
benzo(a)pyrene during pregnancy. Immunopharmacol Immunotoxicol 10(2): 179-193.
280
DRAFT - DO NOT CITE OR QUOTE
-------
Urso, P; Zhang, W; Cobb, JR. (1992) Immunological consequences from exposure to benzo(a)pyrene during
pregnancy. Scand J Immunol Suppl 11:203-206.
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment of chemical
mixtures. Federal Register 51(185):34014-34025. Available online at http://www.epa.gov/iris/backgr-d.htm
(accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (1986b) Guidelines for mutagenicity risk assessment. Federal
Register 51(185):34006-34012. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15,
2009).
U.S. EPA (Environmental Protection Agency). (1988) Recommendations for and documentation of biological values
for use in risk assessment. Prepared by the Environmental Criteria and Assessment Office, Office of Health and
Environmental Assessment, Cincinnati, OH for the Office of Solid Waste and Emergency Response, Washington,
DC; EPA 600/6-87/008. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (1991a) Dose-response analysis of ingested benzo(a)pyrene (CAS no
50-32-8). Cincinnati, OH: U.S. Environmental Protection Agency. EPA600R92045.
U.S. EPA (Environmental Protection Agency). (1991b) Drinking water criteria document for polycyclic aromatic
hydrocarbons (PAHs). Cincinnati, OH: Agency, Environmental Protection Agency; PB92173459.
U.S. EPA (Environmental Protection Agency). (1994a) Interim policy for particle size and limit concentration issues
in inhalation toxicity studies. Federal Register 59(206):53799. Available online at http://www.epa.gov/iris/backgr-
d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (1994b) Methods for derivation of inhalation reference
concentrations and application of inhalation dosimetry. Office of Research and Development, Washington, DC;
EPA/600/8-90/066F. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (1995) Use of the benchmark dose approach in health risk
assessment. Risk Assessment Forum, Washington, DC; EPA/630/R-94/007. Available online at
http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=42601 (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (1996) Guidelines for reproductive toxicity risk assessment. Federal
Register 61(212):56274-56322. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15,
2009).
U.S. EPA (Environmental Protection Agency). (1998) Guidelines for neurotoxicity risk assessment. Federal
Register 63(93):26926-26954. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15,
2009).
U.S. EPA (Environmental Protection Agency). (2000a) Science policy council handbook: risk characterization.
Office of Science Policy, Office of Research and Development, Washington, DC; EPA 100-B-00-002. Available
online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2000b) Benchmark dose technical guidance document. External
review draft. Risk Assessment Forum, Washington, DC; EPA/630/R-00/001. Available online at
http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2000c) Supplementary guidance for conducting for health risk
assessment of chemical mixtures. Risk Assessment Forum, Washington, DC; EPA/630/R-00/002. Available online
at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2002) A review of the reference dose and reference concentration
processes. Risk Assessment Forum, Washington, DC; EPA/630/P-02/0002F. Available online at
http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
281
DRAFT - DO NOT CITE OR QUOTE
-------
U.S. EPA (Environmental Protection Agency). (2005a) Guidelines for carcinogen risk assessment. Risk Assessment
Forum, Washington, DC; EPA/630/P-03/001B. Available online at http://www.epa.gov/iris/backgr-d.htm (accessed
January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2005b) Supplemental guidance for assessing susceptibility from
early-life exposure to carcinogens. Risk Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available
online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2006a) Science policy council handbook: peer review. Third
edition. Office of Science Policy, Office of Research and Development, Washington, DC; EPA/100/B-06/002.
Available online at http://www.epa.gov/iris/backgr-d.htm (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2006b) A Framework for Assessing Health Risk of Environmental
Exposures to Children. National Center for Environmental Assessment, Washington, DC, EPA/600/R-05/093F.
Available online at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=158363 (accessed January 15, 2009).
U.S. EPA (Environmental Protection Agency). (2007) Framework for determining a mutagenic mode of action for
carcinogenicity: Using EPA's 2005 cancer guidelines and supplemental guidance for assessing susceptibility from
early-life exposure to carcinogens. External Review Draft. EPA 120/R-07/002-A. Available online at
http://www.epa.gov/osa/mmoaframework/pdfs/MMOA-ERD-FINAL-83007.pdf (accessed April 21, 2009).
Vainio, H; Uotila, P; Hartiala, J; et al. (1976) The fate of intratracheally installed benzo(a)pyrene in the isolated
perfused rat lung of both control and 20-methylcholanthrene pretreated rats. Res Commun Chem Pathol Pharmacol
13(2):259—271.
Van Agen, B; Maas, LM; Zwingmann, IH; et al. (1997) B[a]P-DNA adduct formation and induction of human
epithelial lung cell transformation. Environ Mol Mutagen 30(3):287-292.
van Delft, JH; Steenwinkel, MS; van Asten, JG; et al. (2001) Biological monitoring the exposure to polycyclic
aromatic hydrocarbons of coke oven workers in relation to smoking and genetic polymorphisms for GSTM1 and
GSTT1. Ann Occup Hyg 45(5):395-408.
van den Berg, FA; Baken, KA; Vermeulen, JP; et al. (2005) Use of the local lymph node assay in assessment of
immune function. Toxicology 211(1-2): 107-114.
van der Bijl, P; van Eyk, AD. (1999) Penetration of benzo[a]pyrene through human buccal and vaginal mucosa.
Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontics 87(4):452^155.
Vandebriel, RJ; Spiekstra, SW; Hudspith, BN; et al. (1999) In vitro exposure effects of cyclosporin a and bis(tri-n-
butyltin)oxide on lymphocyte proliferation, cytokine (receptor) mRNA expression, and cell surface marker
expression in rat thymocytes and splenocytes. Toxicology 135(l):49-66.
Van Lipzig, MMH; Vermeulen, NPE; Gusinu, R; et al. (2005) Formation of estrogenic metabolites of
benzo[a]pyrene and chrysene by cytochrome P450 activity and their combined and supra-maximal estrogenic
activity. Environ Toxicol Pharmacol 19(1 ):41—55.
van Rooij, JG; Bodelier-Bade, MM; Jongeneelen, FJ. (1993) Estimation of individual dermal and respiratory uptake
of polycyclic aromatic hydrocarbons in 12 coke oven workers. Br J Ind Med 50(7):623-632.
Van Schooten, FJ; Hillebrand, MJ; Scherer, E; et al. (1991) Immunocytochemical visualization of DNA adducts in
mouse tissues and human white blood cells following treatment with benzo[a]pyrene or its diol epoxide. A
quantitative approach. Carcinogenesis 12(3):427-433.
Verschueren, K. (2001) Handbook of environmental data on organic chemicals. New York, NY: John Wiley &
Sons, Inc; pp. 295-305.
Vesselinovitch, SD; Kyriazis, AP; Mihailovich, N. (1975) Conditions modifying development of tumors in mice at
various sites by benzo(a)pyrene. Cancer Res 35:2948-2953.
282
DRAFT - DO NOT CITE OR QUOTE
-------
Vesselinovitch, SD, Rao, KVN; Mihailovich, N. (1979) Neoplastic response of mouse tissues during perinatal age
periods and its significance in chemical carcinogenesis. Natl Cancer Inst Monogr 51: 239-250.
Vijayalakshmi, KP; Suresh, CH. (2008) Theoretical studies on the carcinogenicity of polycyclic aromatic
hydrocarbons. J Comput Chem 29:1108-1117.
Vineis, P; Perera, F. (2007) Molecular epidemiology and biomarkers in etiologic cancer research: the new in light of
the old. Cancer Epidemiol Biomarkers Prev 16(10): 1954-1965.
Vineis, P; Anttila, S; Benhamou, S; et al. (2007a) Evidence of gene gene interactions in lung carcinogenesis in a
large pooled analysis. Carcinogenesis 28(9):1902-1905.
Vineis, P; Veglia, F; Garte, S; et al. (2007b) Genetic susceptibility according to three metabolic pathways in cancers
of the lung and bladder and in myeloid leukemias in nonsmokers. Ann Oncol 18(7): 1230-1242.
Vinggaard, AM; Hnida, C; Larsen, JC. (2000) Environmental polycyclic aromatic hydrocarbons affect androgen
receptor activation in vitro. Toxicology 145(2—3): 173—183.
Vogel, EW; Zijlstra, JA; Blijleven, WG. (1983) Mutagenic activity of selected aromatic amines and polycyclic
hydrocarbons in Drosophila melanogaster. Mutat Res 107(1):53—77.
Vondracek, J; Kozubik, A; Machala, M. (2002) Modulation of estrogen receptor-dependent reporter construct
activation and GO/Gl-S-phase transition by polycyclic aromatic hydrocarbons in human breast carcinoma MCF-7
cells. Toxicol Sci 70(2): 193-201.
Wachsmuth, E. (1983) Evaluating immunopathological effects of new drugs. Immunotoxicology 237-250.
Waidyanatha, S; Zheng, Y; Rappaport, SM. (2003) Determination of polycyclic aromatic hydrocarbons in urine of
coke oven workers by headspace solid phase microextraction and gas chromatography-mass spectrometry. Chem
Biol Interact 145(2): 165-174.
Wang, JJ; Frazer, DG; Law, B; et al. (2003) Identification and quantification of urinary benzo[a]pyrene and its
metabolites from asphalt fume exposed mice by microflow LC coupled to hybrid quadrupole time-of-flight mass
spectrometry. Analyst 128(7):864-870.
Warshawsky, D; Barkley, W. (1987) Comparative carcinogenic potencies of 7H-dibenzo[c,g]carbazole,
dibenz[a,j]acridine and benzo[a]pyrene in mouse skin. Cancer Lett 37(3):337-344.
Watanabe, S; Kamiguchi, Y. (2001) Chromosome analysis of human spermatozoa following in vitro exposure to
cyclophosphamide, benzo(a)pyrene and N-nitrosodimethylamine in the presence of rat liver S9. Mutat Res 491(1—
2): 57-63.
Waylen, AL; Metwally, M; Jones, GL; Wilkinson, AJ; Ledger. WL. ( 2009) Effects of cigarette smoking upon
clinical outcomes of assisted reproduction: a meta-analysis. Hum Reprod Update. 15:31-44.
Wei, H; Tye, L; Bresnick, E; et al. (1990). Inhibitory effect of apigenin, a plant flavonoid, on epidermal ornithine
decarboxylase and skin tumor promotion in mice. Cancer Research 50:499-502.
Wei, Q; Cheng, L; Amos, CI; et al. (2000) Repair of tobacco carcinogen-induced DNA adducts and lung cancer risk:
a molecular epidemiologic study. J Natl Cancer Inst 92(21): 1764-1772.
Wei, C; Cacavale, RJ; Kehoe, JJ; et al. (2001) CYP1A2 is expressed along with CYP1 Al in the human lung.
Cancer Lett 164(l):25-32.
Wei, C; Caccavale, RJ; Weyand, EH; et al. (2002) Induction of CYP1 Al and CYP1A2 expressions by prototypic
and atypical inducers in the human lung. Cancer Lett 178(1):25—36.
Wester, RC; Maibach, HI; Bucks, DA; et al. (1990) Percutaneous absorption of [14CJDDT and
[14C]benzo[a]pyrene from soil. Fundam Appl Toxicol 15(3):510—516.
283
DRAFT - DO NOT CITE OR QUOTE
-------
Weyand, EH; Bevan, DR. (1986) Benzo(a)pyrene disposition and metabolism in rats following intratracheal
instillation. Cancer Res 46(11 ):5655—5661.
Weyand, EH; Bevan, DR. (1987) Species differences in disposition of benzo[a]pyrene. Drug Metab Dispos
15(4):442-448.
Weyand, EH; LaVoie, EJ. (1988) Comparison of PAH DNA adduct formation and tumor initiating activity in
newborn mice. Proc Am Assoc Cancer Res 29(10):98.
Weyand, EH; He, ZM; Ghodrati, F; et al. (1992) Effect of fluorine substitution on benzo[j]fluoranthene
genotoxicity. Chem Biol Interact 84(1):37—53.
Weyand, EH; Wu, Y; Patel, S; et al. (1994) Biochemical effects of manufactured gas plant residue following
ingestion by B6C3F1 mice. J Toxicol Environ Health 42(1):89-107.
Weyand, EH; Chen, YC; Wu, Y; et al. (1995) Differences in the tumorigenic activity of a pure hydrocarbon and a
complex mixture following ingestion: benzo[a]pyrene vs manufactured gas plant residue. Chem Res Toxicol
8(7):949-954.
Whitlock, JP, Jr. (1999) Induction of cytochrome P4501A1. Annu Rev Pharmacol Toxicol 39:103-125.
Whitwell, J; Fowler, P; Allars, S; et al. (2010) 5-Fluorouracil, colchicine, benzo[a]pyrene and cytosine arabinoside
tested in the in vitro mammalian cell micronucleus test (MNvit) in Chinese hamster V79 cells at Covance
Laboratories, Harrogate, UK in support of OECD draft Test Guideline 487. Mutat Res.
WHO (World Health Organizatino). (1998) Environmental health criteria. Vol. 202. Selected non-heterocyclic
polycyclic aromatic hydrocarbons. International Programme on Chemical Safety, Geneva, Switzerland.
Wiersma, DA; Roth, RA. (1983a) Total body clearance of circulating benzo(a)pyrene in conscious rats: effect of
pretreatment with 3-methylcholanthrene and the role of liver and lung. J Pharmacol Exp Ther 226(3):661-667.
Wiersma, DA; Roth, RA. (1983b) The prediction of benzo[a]pyrene clearance by rat liver and lung from enzyme
kinetic data. Mol Pharmacol 24(2):300-308.
Wijnhoven, SW; Kool, HJ; van Oostrom, CT; et al. (2000) The relationship between benzo[a]pyrene-induced
mutagenesis and carcinogenesis in repair-deficient Cockayne syndrome group B mice. Cancer Res 60(20): 5681-
5687.
Williams, JA; Martin, FL; Muir, GH; et al. (2000) Metabolic activation of carcinogens and expression of various
cytochromes P450 in human prostate tissue. Carcinogenesis 21 (9): 1683—1689.
Wilson, JS; Holland, LM. (1988) Periodic response difference in mouse epidermis chronically exposed to crude-oils
or benzo[a]pyrene: males vs. females. Toxicology 50(l):83-94.
Wislocki, PG; Bagan, ES; Lu, AY; et al. (1986) Tumorigenicity of nitrated derivatives of pyrene, benz[a]anthracene,
chrysene and benzo[a]pyrene in the newborn mouse assay. Carcinogenesis 7(8): 1317—1322.
Withey, JR; Shedden, J; Law, FC; et al. (1993) Distribution of benzo[a]pyrene in pregnant rats following inhalation
exposure and a comparison with similar data obtained with pyrene. J Appl Toxicol 13(3): 193-202.
Wolff, RK; Griffith, WC; Henderson, RF; et al. (1989) Effects of repeated inhalation exposures to 1-nitropyrene,
benzo[a]pyrene, Ga203 particles, and S02 alone and in combinations on particle clearance, bronchoalveolar lavage
fluid composition, and histopathology. J Toxicol Environ Health 27(1): 123-138.
Wood, AW; Levin, W; Chang, RL; et al. (1980) Mutagenicity and tumor-initiating activity of cyclopenta(c,d)pyrene
and structurally related compounds. Cancer Res 40(3):642-649.
284
DRAFT - DO NOT CITE OR QUOTE
-------
Wormhoudt, LW; Commandeur, JN; Vermeulen, NP. (1999) Genetic polymorphisms of human N-acetyltransferase,
cytochrome P450, glutathione-S-transferase, and epoxide hydrolase enzymes: relevance to xenobiotic metabolism
and toxicity. Crit Rev Toxicol 29(1):59-124.
Wormley, DD; Chirwa, S; Nayyar, T; et al. (2004) Inhaled benzo(a)pyrene impairs long-term potentiation in the F1
generation rat dentate gyrus. Cell Mol Biol 50(6):715-721.
Wu, J; Hou, H; Ritz, B; Chen, Y. (2010) Exposure to polycyclic aromatic hydrocarbons and missed abortion in early
pregnancy in a Chinese population. Sci Total Environ 408:2312-2318.
Wu, Y; Kim, SJ; Weyand, EH. (1994) Bioavailability of manufactured gas plant residue (MPG) components in mice
following ingestion: comparison of adulterated powder and gel diets. Polycyclic Aromatic Hydrocarbons 7:175-
182.
Wu, MT; Simpson, CD; Christiani, DC; et al. (2002) Relationship of exposure to coke-oven emissions and urinary
metabolites of benzo(a)pyrene and pyrene in coke-oven workers. Cancer Epidemiol Biomarkers Prev 11 (3):311—
314.
Wu, J; Ramesh, A; Nayyar, T; et al. (2003) Assessment of metabolites and AhR and CYP1A1 mRNA expression
subsequent to prenatal exposure to inhaled benzo(a)pyrene. Int J Dev Neurosci 21(6):333-346.
Wu, X; Roth, JA; Zhao, H; et al. (2005) Cell cycle checkpoints, DNA damage/repair, and lung cancer risk. Cancer
Res 65(l):349-357.
Wu, X; Gu, J; Spitz, MR. (2007) Mutagen sensitivity: a genetic predisposition factor for cancer. Cancer Res
67(8):3493-3495.
Wu, Q; Suzuki, JS; Zaha, H; et al. (2008) Differences in gene expression and benzo[a]pyrene-induced DNA adduct
formation in the liver of three strains of female mice with identical AhRb2 genotype treated with 2,3,7,8-
tetrachlorodibenzo-p-dioxin and/or benzo[a]pyrene. J Appl Toxicol 28:724-733.
Wynder, EL; Hoffmann, D. (1959) A study of tobacco carcinogenesis. VII. The role of higher polycyclic
hydrocarbons. Cancer 12:1079-1086.
Wynder, EL; Fritz, L; Furth, N. (1957) Effect of concentration of benzopyrene in skin carcinogenesis. J Natl Cancer
Inst 19(3):361—370.
Xu, Z; Brown, LM; Pan, GW; et al. (1996) Cancer risks among iron and steel workers in Anshan, China, Part II:
Case-control studies of lung and stomach cancer. Am J Ind Med 30(1 ):7—15.
Xu, D; Penning, TM; Blair, IA; et al. (2009) Synthesis of phenol and quinone metabolites of benzo[a]pyrene, a
carcinogenic component of tobacco smoke implicated in lung cancer. J Org Chem 74(2):597-604.
Xue, W; Warshawsky, D. (2005) Metabolic activation of polycyclic and heterocyclic aromatic hydrocarbons and
DNA damage: a review. Toxicol Appl Pharmacol 206(l):73-93.
Yamasaki, H. (1990) Gap junction intercellular communication and carcinogenesis. Carcinogenesis 11(7): 1051-
1058.
Yang, SK; Roller, PP; Fu, PP; et al. (1977) Evidence for a 2,3-epoxide as an intermediate in the microsomal
metabolism of benzo[a]pyrene to 3-hydroxybenzo[a]pyrene. Biochem Biophys Res Commun 77(4): 1176—1182.
Yang, JJ; Roy, TA; Krueger, AJ; et al. (1989) In vitro and in vivo percutaneous absorption of benzo[a]pyrene from
petroleum crude-fortified soil in the rat. Bull Environ Contam Toxicol 43(2):207-214.
Yang, H; Mazur-Melnyk, M; de Boer, JG; et al. (1999) A comparison of mutational specificity of mutations induced
by S9-activated B[a]P and benzo[a]pyrene-7,8-diol-9,10-epoxide at the endogenous aprt gene in CHO cells. Mutat
Res 423:23-32.
285
DRAFT - DO NOT CITE OR QUOTE
-------
Yu, D; Berlin, JA; Penning, TM; et al. (2002) Reactive oxygen species generated by PAH o-quinones cause change-
in-function mutations in p53. Chem Res Toxicol 15(6):832-842.
Zhao, J; Lahiri-Chatterjee, M; Y Sharma, Y. et al; (2000). Inhibitory effect of a flavonoid antioxidant silymarin on
benzoyl peroxide-induced tumor promotion, oxidative stress and inflammatory responses in SENCAR mouse skin.
Carcinogenesis 21(4): 811-816.
Zeilmaker, MJ; van Eijkeren, JCH. (1997) Modeling Ah-receptor dependent P450 induction I. Cellular model
definition and its corporation in a PBPK model of 2,3,7,8-TCDD. RIVM Report 604138 001. National Institute of
Public Health and the Environment, B, The Netherlands.
Zeilmaker, MJ; van Eijkeren, JCH; Kroese, ED. (1999a) PBPK simulated DNA adduct formation: relevance for the
risk assessment of benzo(a)pyrene. Bilthoven, The Netherlands.
Zeilmaker, MJ; van Eijkeren, JCH; Oiling, M. (1999b) A PBPK-model for B(a)P in the rat relating dose and liver
DNA-adduct level. Bilthoven, The Netherlands.
Zenzes, MT; Bielecki, R; Reed, T.E. (1999a) Detection of benzo[a]pyrene diol epoxide-DNA adducts in sperm of
men exposed to cigarette smoke, adducts in ovarian cells of women exposed to cigarette smoke. Fertil Steril
1999;72:330-335.
Zenzes, MT; Puy, LA; Bielecki, R; et al. (1999b) Detection of benzo[a]pyrene diol epoxide-DNA adducts in
embryos from smoking couples: evidence for transmission by spermatozoa. Mol Hum Reprod 5(2): 125—131.
Zenzes, MT; Puy, LA; Bielecki, R; et al. (1999b) Detection of benzo[a]pyrene diol epoxide-DNA adducts in
embryos from smoking couples: evidence for transmission by spermatozoa. Mol Hum Reprod 5(2): 125—131.
Zhang, Y; Ramos, KS. (1997) The induction of proliferative vascular smooth muscle cell phenotypes by
benzo[a]pyrene does not involve mutational activation of ras genes. Mutat Res 373(2):285-292.
Zhang, YJ; Hsu, TM; Santella, RM. (1995) Immunoperoxidase detection of polycyclic aromatic hydrocarbon-DNA
adducts in oral mucosa cells of smokers and nonsmokers. Cancer Epidemiol Biomarkers Prev 4:133-138.
Zhang, YJ; Weksler, BB; Wang, L; et al. (1998) Immunohistochemical detection of polycyclic aromatic
hydrocarbon-DNA damage in human blood vessels of smokers and non-smokers. Atherosclerosis 140:325-331.
Zheng, Z; Fang, JL; Lazarus, P. (2002) Glucuronidation: an important mechanism for detoxification of
benzo[a]pyrene metabolites in aerodigestive tract tissues. Drug Metab Dispos 30(4):397-403.
286
DRAFT - DO NOT CITE OR QUOTE
-------
1 APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
2 COMMENTS AND DISPOSITION
3
4
5 [THIS PAGE INTENTIONALLY LEFT BLANK]
287
DRAFT - DO NOT CITE OR QUOTE
-------
1 APPENDIX B. BENCHMARK DOSE MODELING RESULTS FOR NONCANCER
2
3 Increased liver weight (Kroese et al, 2001) male
4
5
6 Table B-l. Liver weight (±SD)a in male F344 rats administered
7 benzo[a]pyrene by gavage for 90 days
8
Organ
Dose (mg/kg-d)
0
3
10
30
Liver weight (g)
Males
7.49 ±0.97
8.00 ±0.85
8.62 ± 1.30b
9.67 ± 1.17b
a Reported as SE, but judged to be SD (and confirmed by study authors).
bSignificantly (p < 0.05) different from control mean; student t-test (unpaired, two-tailed); n = 10/sex/group.
9
10
11
12
13 Table B-2. BMD modeling results for increased liver weight in male rats,
14 with BMR=10%
15
Study
Endpoint
Model
AIC
Goodness-of-
fit p value
BMD
BMDL
Kroese et al.,
2001
Liver weight
Linear (1° polynomial),
Power
49.51
0.58
8.11
5.8
Polynomial (2°)a
50.53
0.74
4.53
2.29
Hill
50.48
0.82
4.1
1.24
16
17 a In order to consider apparent curvature in the dose-response data, the polynomial coefficients were allowed to be
18 negative; a satisfactory fit was achieved, with monotonically increasing predictions within the observed data range.
19 Since the AIC was higher than for the linear model, the 2-degree polynomial was not selected as the best fit.
20
288
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Linear Model with 0.95 Confidence Level
10.5
10
9.5
c
o
CL
<1)
q:
C
ro
<1)
8.5
7.5
6.5
Linear
10
dose
15
20
13:09 01/12 2010
Figure B-l. Fit of polynomial model to data on increased liver weight in
male Wistar rats—90 days.
Model output:
Polynomial Model. (Version: 2.13; Date: 04/08/2 008)
Input Data File: C:\USEPA\BMDS21\Data\linLiverwtKroeseLinearDefault.(d)
Gnuplot Plotting File: C:\USEPA\BMDS21\Data\linLiverwtKroeseLinearDefault.plt
Tue Jan 12 13:09:58 2010
BMDS Model Run
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 +
Dependent variable = Liver_wt
Independent variable = Dose
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1.18058
289
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
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
77
78
79
80
81
82
83
84
rho =
beta_0 =
beta 1 =
0
7.71695
0 . 0951703
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
alpha beta_0 beta_l
alpha 1 -3.6e-011 le-010
beta_0 -3 . 6e-011 1 -0.68
beta 1 le-010 -0.68 1
Parameter Estimates
Variable
alpha
beta_0
beta 1
Estimate
1. 09179
7.71695
0 . 0951703
95.0% Wald Confidence Interval
Std. Err. Lower Conf. Limit Upper Conf. Limit
0.244132 0.613302 1.57028
0.224102 7.27772 8.15618
0.0197929 0.0563769 0.133964
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev
Scaled Res.
0
2.1
7.1
21.4
10
10
10
10
7.49
8.62
9.67
7.72
7 . 92
8.39
9.75
0 . 97
0.85
1.3
1.17
1. 04
1. 04
1. 04
1. 04
-0.687
0 .252
0.688
-0 .253
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = SigmaA2
Likelihoods of Interest
Model Log(likelihood) # Param's AIC
A1 -21.212822 5 52.425644
A2 -20.156688 8 56.313377
A3 -21.212822 5 52.425644
fitted -21.756413 3 49.512826
R -30.879511 2 65.759022
Explanation of Tests
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Note:
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.)
Test
Tests of Interest
-2*log(Likelihood Ratio) Test df
p-value
290
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Test 1
Test 2
Test 3
Test 4
21.4456
2 .11227
2 .11227
1.08718
6
3
3
2
0 . 001525
0.5494
0.5494
0.5807
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect
0.1
Risk Type
Relative risk
Confidence level
0 . 95
BMD
8.10857
BMDL
5.80436
291
DRAFT - DO NOT CITE OR QUOTE
-------
1
2 Decreased thymus weight (males) Kroese et al., 2001
3
4
Table B-3. Means ± SDa for thymus weight in male Wistar rats exposed to
benzo[a]pyrene by gavage 5 days/week for 90 days
Organ
Dose (mg/kg-d)
0
3
10
30
Thymus weight (mg)
Males
380 ± 60
380±110
330 ±60
270 ± 40b
a Reported as SE, but judged to be SD (and confirmed by study authors).
bSignificantly (p < 0.05) different from control mean; student t-test (unpaired, two-tailed); n = 10/sex/group.
Source: Kroese et al. (2001).
Table B-4. Model predictions for decreased thymus weight in male Wistar rats—90
days
Model
Variance
/j-valuc"
Goodness-of-fit
/j-valuc
AIC
bmd1sd
(mg/kg-d)
BMDLisd
(mg/kg-d)
Constant variance
Linear
0.01
0.74
384.84
12.97
8.97
Nonconstant variance
Hilf
NA
Linear, Polynomial (2-degree),
Power0
0.30
0.23
380.71
16.40
11.30
NA = not applicable, model failed;
292
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Linear Model with 0.95 Confidence Level
Linear
450
400
350
300
250
BMDL
BMD
0
5
10
15
20
dose
15:33 10/15 2009
BMDs and BMDLs indicated are associated with a change of 1 SD from the control, and are in
units of mg/kg-day.
Source: Kroese et al. (2001).
Figure B-2. Fit of linear model (nonconstant variance) to data on decreased
thymus weight in male Wistar rats—90 days.
Linear (nonconstant variance)
Polynomial Model. (Version: 2.13; Date: 04/08/2008)
Input Data File:
C:\USEPA\IRIS\benzo[a] pyrene\RfD\Kroese2 001\90day\thymusweight\male\durationadjusted\2
Linkrolin.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese2 001\90day\thymusweight\male\durationadjusted\2
Linkrolin.pit
Thu Oct 15 15:33:37 2009
BMDS Model Run
The form of the response function is:
Y [dose] = beta_0 + beta_l*dose + beta_2*dose*2 + ...
293 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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
Dependent variable = mean
Independent variable = dose
The polynomial coefficients are restricted to be negative
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha = 8.56121
rho = 0
beta_0 = 380.763
beta 1 = -5.3285
Asymptotic Correlation Matrix of Parameter Estimates
lalpha rho beta_0 beta_l
lalpha 1 -1 0.048 -0.061
rho -1 1 -0.04 8 0.061
beta_0 0.048 -0.048 1 -0.84
beta 1 -0.061 0.061 -0.84 1
Parameter Estimates
Interval
Variable
Limit
lalpha
0 .288754
rho
7.94967
beta_0
411.351
beta_l
3 . 17249
Estimate
-18.8293
4.66515
378.954
-5.14219
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
9.75429 -37.9473
1.67581 1.38062
16.5291 346.558
1.00497 -7.11189
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0 10 380 379 60 84.3 0.0392
2.1 10 380 368 110 78.8 0.475
7.1 10 330 342 60 66.6 -0.591
21.4 10 270 269 40 37.9 0.0908
Model Descriptions for likelihoods calculated
294
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
Model A1: Yij = Mu(i) + e(ij]
Var{e(ij)} = Sigma*2
Model A2 : Yij = Mu(i) + e(ij]
Var{e(ij)} = Sigma(i)*2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model Log(likelihood) # Param's AIC
A1 -189.116991 5 388.233982
A2 -183.673279 8 383.346558
A3 -184.883626 6 381.767253
fitted -186.353541 4 380.707081
R -196.353362 2 396.706723
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 25.3602 6 0.0002928
Test 2 10.8874 3 0.01235
Test 3 2.42069 2 0.2981
Test 4 2.93983 2 0.2299
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
295
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
BMD
BMDL
16 .4008
11 .2965
296
DRAFT - DO NOT CITE OR QUOTE
-------
Decreased thymus weight females (Kroese et al., 2001)
Table B-5. Means ± SDa for thymus weight in female Wistar rats exposed to
benzo[a]pyrene by gavage 5 days/week for 90 days
Organ
Dose (mg/kg-d)
0
3
10
30
Thymus weight (mg)
Females
320 ± 60
310 ± 50
300 ± 40
230 ± 30b
a Reported as SE, but judged to be SD (and confirmed by study authors).
bSignificantly (p < 0.05) different from control mean; student t-test (unpaired, two-tailed); n = 10/sex/group.
Source: Kroese et al. (2001).
Table B-6. Model predictions for decreased thymus weight in female Wistar
rats—90 days
Model (constant variance)
Variance
/j-valuc"
Means
/j-valuc1'
AIC
bmd1sd
(mg/kg-d)
BMDLisd
(mg/kg-d)
Hillb
NA
Linear0
0.17
0.81
349.12
10.52
7.64
Polynomial (2-degree)°'d
0.17
0.77
350.80
13.29
7.77
Powerb
NA
aValues <0.10 fail to meet conventional goodness-of-fit criteria.
bPower restricted to >1.
Coefficients restricted to be negative.
dLowest degree polynomial with an adequate fit is reported.
BMD/BMC = maximum likelihood estimate of the dose/concentration associated with the selected BMR; NA = not
applicable; model failed to generate
Source: Kroese et al. (2001).
297
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Linear Model with 0.95 Confidence Level
Linear
360
340
320
300
280
260
240
220
200
BMDL
BMD
0
5
10
15
20
dose
16:27 10/15 2009
BMDs and BMDLs indicated are associated with a change of 1 SD from the control, and are in
units of mg/kg-day.
Source: Kroese et al. (2001).
Figure B-3. Fit of linear model (constant variance) to data on decreased
thymus weight in female Wistar rats—90 days.
Polynomial Model. (Version: 2.13; Date: 04/08/2008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese200l\90day\thymusweight\female\durationadjusted
\2Linkrolin.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese200l\90day\thymusweight\female\durationadjusted
\2Linkrolin.pit
Thu Oct 15 16:27:44 2009
BMDS Model Run
The form of the response function is:
Y [dose] = beta_0 + beta_l*dose + beta_2*dose*2 + ...
Dependent variable = mean
298 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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
Independent variable = dose
rho is set to 0
The polynomial coefficients are restricted to be negative
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1
rho = 0 Specified
beta_0 = 322.144
beta 1 = -4.2018
Asymptotic Correlation Matrix of Parameter Estimates
*** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix
alpha beta_0 beta_l
alpha 1 2.4e-008 -2.3e-008
beta_0 2.4e-008 1 -0.68
beta 1 -2.3e-008 -0.68 1
Parameter Estimates
Interval
Variable
Limit
alpha
2811.69
beta_0
340.73
beta_l
2 . 56026
Estimate
1954.92
322 .144
-4.2018
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
437.134 1098.16
9.48287 303.558
0.837537 -5.84334
Table of Data and Estimated Values of Interest
Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
10
320
322
60
44 . 2
-0 .153
2 .1
10
310
313
50
44 . 2
-0.237
7 . 1
10
300
292
40
44 . 2
0 . 55
21. 4
10
230
232
30
44 . 2
-0.159
Model Descriptions for likelihoods calculated
299
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
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
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model Log(likelihood) # Param's AIC
A1 -171.357252 5 352.714504
A2 -168.857234 8 353.714467
A3 -171.357252 5 352.714504
fitted -171.562118 3 349.124237
R -181.324151 2 366.648303
Explanation of Tests
Test 1:
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 24.9338 6 0.0003512
Test 2 5.00004 3 0.1718
Test 3 5.00004 3 0.1718
Test 4 0.409733 2 0.8148
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
300
DRAFT - DO NOT CITE OR QUOTE
-------
BMD
BMDL
301 DRAFT - DO NOT CITE OR QUOTE
10 . 5228
7.64037
-------
1 Decreased thymus weight males (DeJong et al., 1999)
2
3
Table B-7. Means ± SD for thymus weight in male Wistar rats exposed to
benzo[a]pyrene by gavage 5 days/week for 35 days
4
Organ
Dose (mg/kg-d)
0
3
10
30
90
Thymus weight (mg)
(means; n = 7-8)
517 ±47
472 ± 90
438 ± 64a
388 ± 71a
198 ± 65a
5
6
Table B-8. Model predictions for decreased thymus weight in male Wistar
rats—35 days
Model (constant variance)
Variance
/>-valucb
Means
/>-valucb
AIC
BMD1sd
(mg/kg-d)
BMDLisd
(mg/kg-d)
Hillc
0.50
0.42
382.91
11.15
6.19
Linear1*, Polynomial (2-dcgrcc)'1''1
0.50
0.52
381.41
14.41
11.58
Powerd
NA
aNumber of animals was reported to be 7-8 per dose group, and was not specified for each group; forBMD
modeling purposes, n = 8 was used.
bValues <0.10 fail to meet conventional goodness-of-fit criteria.
°Power restricted to >1.
Coefficients restricted to be negative.
eLowest degree polynomial with an adequate fit is reported.
BMD/BMC = maximum likelihood estimate of the dose/concentration associated with the selected BMR; NA = not
applicable; model failed to generate
Source: De Jong et al. (1999).
7
8
302
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Linear Model with 0.95 Confidence Level
CD
CO
c
o
Q.
CO
CD
cr
c
m
CD
550
500
450
400
350
300
250
200
150
Linear
BMDL
BMD
0 10 20 30 40 50 60
dose
04:31 10/19 2009
BMDs and BMDLs indicated are associated with a change of 1 SD from the control, and are in
units of mg/kg-day.
Source: De Jong et al. (1999).
Figure B-4. Fit of linear model (constant variance) to data on decreased
thymus weight in male Wistar rats—35 days.
Model Output:
Polynomial Model. (Version: 2.13; Date: 04/08/2008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\dej ongl99 9\3 5day\thymusweightmale\durationadj usted\2L
indejlin.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\dej ongl99 9\3 5day\thymusweightmale\durationadj usted\2L
indejlin.pit
Mon Oct 19 04:31:24 2009
BMDS Model Run
The form of the response function is:
Y [dose] = beta_0 + beta_l*dose + beta_2*dose*2 + ...
303 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Dependent variable = mean
Independent variable = dose
rho is set to 0
The polynomial coefficients are restricted to be negative
A constant variance model is fit
Total number of dose groups = 5
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
the user.
Default Initial Parameter Values
alpha = 1
rho = 0 Specified
beta_0 = 489.769
beta 1 = -4.5927
Asymptotic Correlation Matrix of Parameter Estimates
*** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix
alpha beta_0 beta_l
alpha 1 -3 . le-009 -3.2e-009
beta_0 -3.le-009 1 -0.62
beta 1 -3.2e-009 -0.62 1
Parameter Estimates
95.0% Wald Confidence
Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf.
Limit
alpha 4382.3 979.911 2461.71
6302.89
beta_0 489.769 13.3751 463.555
515.984
beta_l -4.5927 0.438716 -5.45257
3 . 73283
Table
of
Data and Estimated Values
of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std Dev
Scaled R<
0 f
3
517
490
47
66.2
1. 16
2.1 f
3
472
480
90
66.2
-0.347
7.1 f
3
438
457
64
66.2
-0.819
21.4 f
3
388
391
71
66.2
-0 .149
64 . 3 f
3
198
194
65
66.2
0 .151
304
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log(likelihood)
-186 .580733
-184 .896632
-186 .580733
-187 .706565
-214.086904
# Param's
6
10
6
3
2
AIC
385 .161466
389 .793264
385 .161466
381 .413130
432.173809
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (A1 vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test 1
Test 2
Test 3
Test 4
58.3805
3.3682
3.3682
2 .25166
<.0001
0.4982
0.4982
0.5218
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
305
DRAFT - DO NOT CITE OR QUOTE
-------
1 Risk Type = Estimated standard deviations from the control mean
2
3 Confidence level = 0.95
4
5 BMD = 14.4139
6
7
8 BMDL = 11.577
9
306
DRAFT - DO NOT CITE OR QUOTE
-------
1 Decreased Splenic B cells
2
3
Table B-9. Exposure-related effects in male Wistar rats exposed to
benzo[a]pyrene by gavage 5 days/week for 5 weeks
Effect
Dose (mg/kg-d)
0
3
10
30
90
Spleen cell distribution (%)
B cells
39± 4
36 ±2
34 ± 3a
32 ± 4a
23 ± 4a
T cells
40 ±9
48 ± 12
40 ±9
36 ±2
44 ±6
Th cells
23 ±7
26 ±7
24 ±5
22 ±4
26 ±4
Ts cells
24 ±5
26 ±6
24 ±7
19 ± 2
27 ±5
4 "Significantly (p < 0.05) different from control mean.
5
6 Source: De Jong et al. (1999).
7
8
9
Table B-10. Model predictions for decreased spleen B-cells in male Wistar
rats—35 days
Model
Variance
/j-valuc"
Means
/j-valuc"
AIC
BMD1sd
(mg/kg-d)
BMDLisd
(mg/kg-d)
Constant variance
Hillb
0.30
0.18
146.18
10.24
5.31
Linear0, Polynomial (2-degree)c'd,
Powerb
0.30
0.21
145.28
15.58
12.43
aValues <0.10 fail to meet conventional goodness-of-fit criteria.
bPower restricted to >1.
Coefficients restricted to be negative.
dLowest degree polynomial with an adequate fit is reported.
BMD/BMC = maximum likelihood estimate of the dose/concentration associated with the selected BMR; NA = not
applicable; model failed to generate these values
Source: De Jong et al. (1999).
307
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Linear Model with 0.95 Confidence Level
Linear
40
35
30
25
20
BMDL BMD
0
10
20
30
40
50
60
dose
05:06 10/19 2009
BMDs and BMDLs indicated are associated with a change of 1 SD from the control, and are in
units of mg/kg-day.
Source: De Jong et al. (1999).
Figure B-5. Fit of linear model to data on decreased spleen B-cells in male
Wistar rats—35 days.
Model Output:
Polynomial Model. (Version: 2.13; Date: 04/08/2008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\dej ongl99 9\3 5day\spleenBcell\durationadjusted\2Lindej
lin.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\dej ongl99 9\3 5day\spleenBcell\durationadjusted\2Lindej
lin.pit
Mon Oct 19 05:06:33 2009
BMDS Model Run
The form of the response function is:
Y [dose] = beta_0 + beta_l*dose + beta_2*dose*2 + ...
308
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
Dependent variable = mean
Independent variable = dose
rho is set to 0
The polynomial coefficients are restricted to be negative
A constant variance model is fit
Total number of dose groups = 5
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 12.2
rho = 0 Specified
beta_0 = 37.0148
beta 1 = -0.222068
Asymptotic Correlation Matrix of Parameter Estimates
*** The model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix
alpha beta_0 beta_l
alpha 1 -2 . 4e-009 3.8e-009
beta_0 -2.4e-009 1 -0.62
beta 1 3.8e-009 -0.62 1
Parameter Estimates
Interval
Variable
Limit
alpha
17 .2084
beta_0
38 .3846
beta_l
0.177138
Estimate
11.9647
37 . 0148
-0 . 222068
95.0% Wald Confidence
Std. Err. Lower Conf. Limit Upper Conf.
2.6754 6.72106
0.698873 35.6451
0.0229237 -0.266997
Table
of
Data and Estimated Values
of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std Dev
Scaled R<
0 f
3
39
37
4
3 .46
1. 62
2.1 f
3
36
36 . 5
2
3 .46
-0.449
7.1 f
3
34
35.4
3
3.46
CO
21.4 f
3
32
32 .3
4
3.46
-0 .215
64 . 3 f
3
23
22 . 7
4
3 .46
0 .216
309
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
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
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2 : Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model Log(likelihood) # Param's AIC
A1 -67.358091 6 146.716182
A2 -64.934513 10 149.869025
A3 -67.358091 6 146.716182
fitted -69.639287 3 145.278575
R -93.795081 2 191.590163
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (A1 vs A2)
Test 3: Are variances adequately modeled? (A2 vs.
Test 4: Does the Model for the Mean Fit? (A3 vs.
(Note: When rho=0 the results of Test 3 and Test
A3)
fitted)
2 will be the
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 57.7211 8 <.0001
Test 2 4.84716 4 0.3033
Test 3 4.84716 4 0.3033
Test 4 4.56239 3 0.2068
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
310 DRAFT - DO NOT CITE OR QUOTE
-------
Confidence level
BMD
BMDL
0 .95
15 .5764
12 .4286
311
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Forestomach hyperplasia
All available dichotomous models in the EPA BMDS (version 2.1) were fit to the
incidence data shown in Table B-9 for forestomach hyperplasia in rats and mice orally exposed
to benzo[a]pyrene for 2 years (Kroese et al., 2001; Beland and Culp, 1998). In accordance with
U.S. EPA (2000) guidance, BMDs and BMDLs associated with an extra risk of 10% are
calculated for all models.
Adequate model fit was judged by three criteria: goodness-of-fitp-walue (p> 0.1), visual
inspection of the dose-response curve, and scaled residual at the data point (except the control)
closest to the predefined benchmark response (BMR). Among all of the models providing
adequate fit to the data, the lowest BMDL is selected as the POD when the difference between
the BMDLs estimated from these models are more than threefold; otherwise, the BMDL from
the model with the lowest Akaike's Information Criterion (AIC) is chosen. If an adequate fit to
the data was not achieved using the protocol above for the full dataset, doses were dropped
(starting with the highest dose) until an adequate fit was achieved.
Table B-ll. Dose-response data for forestomach hyperplasia in Wistar rats
and B6C3Fi rats orally exposed to benzo[a]pyrene for 2 years
Species, sex (reference)
Administered dose (mg/kg-day)
0
3
10
30
Duration-adjusted dose (x 5/7)
0
2.1
7.14
21.4
Wistar rat, female (Kroese et al., 2001)
1/52
8/51
13/51
2/52
Wistar rat, male (Kroese et al., 2001)
2/50
8/52
8/52
0/52
Administered dose (mg/kg-day)
0
0.7
3.3
16.5
B6C3Fi mice, female (Beland and Culp, 1998)
13/48
23/47
33/46
46/47
All data sets provided adequate descriptions of the dose-response relationship for
forestomach hyperplasia from chronic oral exposure to benzo[a]pyrene, but at the highest dose
level in rats, the incidence of hyperplasia was not increased. It is possible that the forestomach
hyperplasia observed following benzo[a]pyrene exposure may be a precursor to the development
of forestomach tumors, but specific data supporting this conclusion are unavailable. Regardless,
the male and female data sets in rats (Kroese et al., 2001) were modeled without the data from
the highest dose group due to the nonmontonic increase in response to increasing dose (Kroese et
al., 2001).
312
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Table B-12. Summary of BMDs and BMDLs from the best fitting model
forestomach hyperplasia—oral exposure
Endpoint/data
Strain/species
Exposure
duration
Best fitting
model
BMD
(mg/kg-d)
BMDL
(mg/kg-d)
Reference
Forestomach
hyperplasia
(highest dose
excluded)3
Wistar rat (male)
2 yrs
Log-logistic
5.31
2.39
Kroese et al., 2001
Forestomach
hyperplasia
(highest dose
excluded)3
Wistar rat
(female)
2 yrs
Log-logistic
2.15
1.35
Kroese et al., 2001
Forestomach
hyperplasia
B6C3Fi mouse
(female)
2 yrs
Log-logistic
0.33
0.12
Beland and Culp,
1998
aData for the high-dose group were excluded in the modeled datset due to the absence of an increase in incidence at
the high dose.
Forestomach hyperplasia-Male Wistar Rats, 2 yrs (Kroese etal., 2001)
Table B-13. Model predictions for forestomach hyperplasia in male Wistar
rats in a 2-year study
Model
Degrees of
freedom
x2
X2 Goodness-
of-fit /j-valuc
AIC
BMD10
(mg/kg-d)
BMDL10
(mg/kg-d)
Highest dose excludedb
Gamma0, Multistage"1,
Weibulf
1
2.39
0.12
112.37
5.63
2.67
Logistic
1
2.83
0.09
112.93
7.25
4.35
LogLogistic
1
2.28
0.13
112.27
5.31
2.39
LogProbit
1
3.64
0.06
113.88
8.36
4.52
Probit
1
2.78
0.10
112.87
7.09
4.13
aValues <0.10 fail to meet conventional goodness-of-fit criteria.
bData for the high-dose group were excluded in the modeled datset due to the absence of an increase in incidence at
the high dose; likely related to the statistically significantly increased incidence of forestomach tumors in these
animals.
°Power restricted to >1.
dBetas restricted to >0; lowest degree polynomial with an adequate fit is reported (1-degree polynomial).
Source: Kroese et al. (2001).
313
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
0.3
0.25
0.2
0.15
0.1
0.05
0
BMDL
BMD
0
2
3
4
5
6
7
dose
08:57 10/14 2009
Figure B-6. Fit of log logistic model to data on forestomach hyperplasia in
male Wistar rats in a 2-year study.
Model output:
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese2 001\chronic\forestomachhyperplasia\male\durati
onadjusted\3Logkrolog.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese2 001\chronic\forestomachhyperplasia\male\durati
onadjusted\3Logkrolog.pit
Tue Oct 13 21:42:15 2009
BMDS Model Run
The form of the probability function is:
P [response] = background+(1-background)/[1 + EXP(-intercept-slope*Log(dose))]
Dependent variable = incidence
Independent variable = dose
Slope parameter is restricted as slope >= 1
314
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
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
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.04
intercept = -3.27842
slope = 1
the user,
background
intercept
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
background intercept
1 -0.7
-0.7 1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
0 . 0623861
-3 . 86644
1
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(1i ke1i hood)
-53 . 0468
-54 .1335
-55.5429
# Param's
3
2
1
Deviance Test d.f.
2 . 17337
4 . 99229
P-value
0 .1404
0.0824
AIC :
112 .267
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.0624 3.119 2.000 50 -0.654
2.1000 0.1019 5.297 8.000 52 1.239
7.1400 0.1843 9.584 8.000 52 -0.566
Chi*2 = 2.28 d.f. = 1 P-value = 0.1306
315
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 5.308
BMDL = 2.38692
Forestomach hyperplasia- Female Wistar Rats, 2 yrs (Kroese et al., 2001)
Table B-14. Model predictions for forestomach hyperplasia in female
Wistar rats in a 2-year study
Model
Degrees of
freedom
x2
X2 Goodness-
of-fit />-valuc
AIC
BMD10
(mg/kg-d)
BMDL10
(mg/kg-d)
Highest dose excludedb
Logistic
1
3.68
0.06
120.02
4.23
3.28
LogLogistic
1
0.98
0.32
117.04
2.15
1.35
LogProbit
1
5.09
0.02
121.13
3.91
2.57
Gamma0, Multistage11,
Weibulf
1
1.40
0.24
117.42
2.40
1.59
Probit
1
3.47
0.06
119.74
3.99
3.06
aValues <0.10 fail to meet conventional goodness-of-fit criteria.
''Data for the high-dose group were excluded in the modeled datset due to the absence of an increase in incidence at
the high dose; likely related to the statistically significantly increased incidence of forestomach tumors in these
animals.
°Power restricted to >1.
dBetas restricted to >0; lowest degree polynomial with an adequate fit is reported (1-degree polynomial).
BMD = maximum likelihood estimate of the dose/concentration associated with the selected BMR
Source: Kroese et al. (2001).
316
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Log-Logistic Model with 0.95 Confidence Level
0)
§
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
Log-Log istic
BMDL
BMD
08:59 10/14 2009
Figure B-7. Fit of log logistic model to data on forestomach hyperplasia in
female Wistar rats in a 2-year study.
Model output:
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese2001\chronic\forestomachhyperplasia\female\dura
tionadjusted\3Logkrolog.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese2001\chronic\forestomachhyperplasia\female\dura
tionadjusted\3Logkrolog.pit
Wed Oct 14 08:59:23 2009
BMDS Model Run
The form of the probability function is:
P [response] = background+(1-background)/[1 + EXP(-intercept-slope*Log(dose))]
Dependent variable = incidence
Independent variable = dose
Slope parameter is restricted as slope >= 1
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
317 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
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
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.0192308
intercept = -2.7983
slope = 1
the user,
background
intercept
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -slope
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
background intercept
1 -0.46
-0.46 1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
0 . 0238694
-2.96044
1
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
* - Indicates that this value is not calculated.
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(1i ke1i hood)
-56 . 048
-56 .5181
-63 .1579
# Param's Deviance Test d.f. P-value
3
2 0.940072 1 0.3323
1 14.2198 2 0.000817
AIC :
117.036
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0.0000
2 . 1000
7.1400
0 . 0239
0.1196
0.2874
1. 241
6 . 101
14.658
1. 000
8 . 000
13.000
52
51
51
-0 .219
0 . 819
-0.513
Chi*2 = 0.98
d.f. = 1
P-value = 0.3216
Benchmark Dose Computation
Specified effect = 0.1
318
DRAFT - DO NOT CITE OR QUOTE
-------
1 Risk Type
2
3 Confidence level
4
5 BMD
6
7 BMDL
8
9
10 Forestomach hyperplasia- female mice, 2 yrs (Beland and Culp, 1998)
Table B-15. Model predictions for forestomach hyperplasia in female
B6C3F1 mice in a 2-year study
Model
Degrees of
freedom
x2
X2 Goodness-
of-fit />-valuc
AIC
BMD10
(mg/kg-d)
BMDL10
(mg/kg-d)
Logistic
2
5.71
0.06
194.7
0.757
0.545
LogLogistic
2
1.55
0.21
193.3
0.329
0.115
LogProbit
2
2.49
0.29
192.1
0.670
0.448
Multistage0, Weibullb,
Gammab
2
1.74
0.42
191.3
0.421
0.295
Probit
2
7.04
0.03
196.6
0.946
0.711
"Values <0.10 fail to meet conventional goodness-of-fit criteria.
bPower restricted to >1.
°Betas restricted to >0; lowest degree polynomial with an adequate fit is reported (1-degree polynomial).
BMD = maximum likelihood estimate of the dose/concentration associated with the selected BMR
Source: Beland and Culp (1998).
Extra risk
0 . 95
2 . 14515
1. 34776
319
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
Figure B-8. Fit of log logistic model to data on forestomach hyperplasia in
female B6C3Fi mice in a 2-year study.
Model output:
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File:
C:\Usepa\BMDS2l\Data\lnl_benzo[a]pyrene_BelandCulp_mice_4s_hyperplasia_generic_dich_10
¦ (d)
Gnuplot Plotting File:
C:\Usepa\BMDS2l\Data\lnl_benzo[a]pyrene_BelandCulp_mice_4s_hyperplasia_generic_dich_10
. pit
BMDS Model Run
The form of the probability function is:
P [response] = background+(1-background)/[1 + EXP(-intercept-slope*Log(dose))]
Dependent variable = NumAff
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
320
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Default Initial Parameter Values
background = 0.270833
intercept = -0.637972
slope = 1.38091
Asymptotic Correlation Matrix of Parameter Estimates
background intercept
background 1 -0.66
intercept -0.66 1
slope 0.46 -0.8
slope
0.46
-0.8
1
Parameter Estimates
Interval
Variable
Limit
background
intercept
slope
Estimate
0 .286381
-0.789676
1.26641
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(1i ke1i hood)
-92 .8312
-93.6497
-125.58
# Param's
4
3
1
Deviance Test d.f.
1. 63686
65 .4982
P-value
0.2008
<.0001
AIC :
193.299
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
0.2864
13.746
13 . 000
48
-0.238
0.7000
0.4464
20.979
23 . 000
47
0 .593
3.3000
0 .7667
35.270
33 . 000
46
-0.791
16.5000
0.9575
45.005
46.000
47
0 . 720
Chi*2 = 1.55 d.f. = 1 P-value = 0.2127
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.329083
321
DRAFT - DO NOT CITE OR QUOTE
-------
BMDL = 0.115446
322 DRAFT - DO NOT CITE OR QUOTE
-------
1 Decreased ovary weight-female rats, 60 days (Xu et al., 2010)
2
3
Table B-16. Means ± SDa for ovary weight in female SD-rats
Organ
Dose (mg/kg-d)a
0
2.5
5
Ovary weight (mg)
0.160 ±0.0146
0.143 ± 0.0098b
0.136 ±0.0098b
a TWA doses over the 60 day study period
b Statistically different (p < 0.05) from controls using one-way ANOVA
Table B-17. Model predictions for decreased ovary weight in female SD-
rats—60 days
Model
Goodness-of-fit
/j-valuc
AIC
bmd1sd
(mg/kg-d)
BMDLisd
(mg/kg-d)
Power
N/A
Linear, Polynomial (1°)
0.39
-138.67
2.27
1.49
NA = not applicable, model failed;
4
5
323
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
0.17
0.16
o
CL
a. 0.15
c
ro
0
0.14
0.13
0
dose
16:03 12/14 2010
Figure B-9. Fit of linear/polynomial (1°) model to data on decreased ovary
weight
Model Output:
Polynomial Model. (Version: 2.16; Date: 05/26/2010)
Input Data File:
C:/USEPA/BMDS212/Data/benzo[a]pyrene/Bap_AbsOvaryWeight/Xu2 010_AbsOvaryWeight_Linear_l
SD.(d)
Gnuplot Plotting File:
C:/USEPA/BMDS212/Data/benzo[a]pyrene/Bap_AbsOvaryWeight/Xu2 010_AbsOvaryWeight_Linear_l
SD.pit
Tue Dec 14 13:51:32 2010
The form of the response function is:
Y [dose] = beta 0 + beta l*dose + beta 2*dose*2 + ...
Dependent variable = Mean
Independent variable = Dose
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Linear
BMDL
Linear Model with 0.95 Confidence Level
Total number of dose groups = 3
324
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =
rho =
beta_0 =
beta 1 =
0.000136
0
0.158333
-0.0048
Speci fied
the user.
alpha
beta_0
beta 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -rho
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha beta_0 beta_l
1 4e-010 -4.5e-010
4e-010 1 -0.77
-4.5e-010 -0.77 1
Parameter Estimates
Variable
alpha
beta_0
beta 1
Estimate
0.000118889
0.158333
-0.0048
Std. Err.
3.962 96e-0 05
0.00406354
0.00125904
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
4.12162e-005
0.150369
-0.00726768
0.000196562
0.166298
-0.00233232
Table of Data and Estimated Values of Interest
Dose
Obs Mean
Est Mean Obs Std Dev Est Std Dev Scaled Res.
0
2 . 5
5
0 . 16
0 .143
0 .136
0 . 158
0 . 146
0 . 134
0 . 0147
0.0098
0.0098
0 .0109
0 .0109
0 .0109
0.374
-0 . 749
0.374
Model Descriptions for likelihoods calculated
Model A1: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = Sigma*2
325
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Likelihoods of Interest
Model Log(likelihood) # Param's AIC
A1 72.766595 4 -137.533190
A2 73.468565 6 -134.937129
A3 72.766595 4 -137.533190
fitted 72.335891 3 -138.671782
R 67.008505 2 -130.017010
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (A1 vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1 12.9201 4 0.01167
Test 2 1.40394 2 0.4956
Test 3 1.40394 2 0.4956
Test 4 0.861408 1 0.3533
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 2.27159
BMDL = 1.49968
326 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
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 C. ADDITIONAL CALCULATIONS FOR THE RfC
Region: Entire Lung
Wed, 03/17/2010. 02:07:20 PM EDT
0.750 I-
0.600
0.450
5 0.300
0.150
0.0
0.621
0.127
0.045
Region
Species & Model Info:
Species/Geometry: Human Limited
FRC 'vblume: 3300.00 ml
Head Vblume: 50.00 ml
Breathing Route: nasal
Breathing Parameters:
Tidal Vblume: 860.00 ml
Breathing Frequency: 18.00 1 Anin
Inspiratory Fraction: 0.50
Pause Fraction: 0.00
Particle Properties:
Diameter: tutUtAD: 1.70 (jm
GSD: 1.00
Concentration: 4.20 |jgAn"3
Figure C-l. Human Fractional Deposition
Species = humanlimited
FRC = 3300.0
Head volume = 50.0
Density = 1.0
Number of particles calculated = single
Diameter = 1.70000000000000 02 iam MMAD
Inhalability = yes
GSD = 1.0
Breathing interval: One single breath
Concentration = 4.2
Breathing Frequency = 16.0
Tidal Volume = 860.0
Inspiratory Fraction = 0.5
Pause Fraction = 0.0
Breathing Route = nasal
Region: Entire Lung
Region: Entire Lung
Region Deposition Fraction
327
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
Head
0.449
3
TB
0.045
4
P
0.127
5
Total
0.621
6
328 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Wed, 03/17/2010, 02:15:27 FM EDT
Region: Entire Lung
0.250
0.200
0.150
o
S 0.100
S2.
a)
a
0.050
0.0
0.072
0.181
0.068
Head
Region
Figure C-2. Rat Fractional Deposition
Species = rat
FRC =4.0
Head volume = 0.42
Density = 1.0
Number of particles calculated = single
Diameter = 1.70000000000000 02 iam MMAD
Inhalability = yes
GSD = 1.0
Breathing interval: One single breath
Concentration = 4.2
Breathing Frequency = 102.0
Tidal Volume = 1.8
Inspiratory Fraction = 0.5
Pause Fraction = 0.0
Breathing Route = nasal
Region: Entire Lung
Region: Entire Lung
Region Deposition Fraction
Head 0.072
TB 0.041
P 0.068
Total 0.181
Species & Model Info:
Species/Geometry: Rat
FRC Vtilume: 4.00 ml
Head Vblume: 0.42 ml
Breathing Route: nasal
Breathing Parameters:
Tidal Vblume: 1.80 ml
Breathing Frequency: 102.00 1Anin
Inspiratory Fraction: 0.50
Pause Fraction: 0.00
Particle Properties:
Diameter: IvMAD: 1.70 |jm
GSD: 1.00
Concentration: 4.20 pgAn'^
329
DRAFT - DO NOT CITE OR QUOTE
-------
APPENDIX D. TIME-TO-TUMOR MODELING FOR THE ORAL SLOPE FACTOR
Table D-l. Tumor incidence data, with time to death with tumor; male rats
exposed by gavage to Benzo|alpyrene (Kroese et al., 2001)
Numbers of Animals with
Skin or Mammary
Gland
Duodenum
Kidney
Oral Cavity or
or Jejunum
Basal Cell
Squamous
Urothelial
Dose Group
Week of
Total
Forestomach Tumors
Liver Tumors
Tumors
Tumors
Cell Tumors
Carcinoma
(mg/kg-day)
Death
Examined
Incid."
Fatal"
Incid.
Fatal
Incid.
Incid.
Incid.
Incid.
44
1
0
0
0
0
0
1
0
0
80
1
0
0
0
0
0
0
0
0
82
1
0
0
0
0
0
0
0
0
84
1
0
0
0
0
0
0
0
0
89
1
0
0
0
0
0
0
0
0
90
3
0
0
0
0
0
0
0
0
91
1
0
0
0
0
0
0
0
0
92
1
0
0
0
0
0
0
0
0
93
1
0
0
0
0
0
0
0
0
0
94
1
0
0
0
0
0
0
0
0
95
2
0
0
0
0
0
0
0
0
96
2
0
0
0
0
0
0
0
0
97
1
0
0
0
0
0
0
0
0
98
1
0
0
0
0
0
0
0
0
100
3
0
0
0
0
0
1
0
0
104
1
0
0
0
0
0
0
0
0
105
1
0
0
0
0
0
0
0
0
108
7
0
0
0
0
0
0
0
0
109
22
0
0
0
0
0
0
0
0
29
1
0
0
0
0
0
0
0
0
40
1
1
0
0
0
0
0
0
0
74
1
0
0
0
0
0
0
0
0
76
1
0
0
0
0
0
0
0
0
79
1
0
0
0
0
0
0
0
0
82
1
0
0
0
0
0
0
0
0
T
92
2
0
0
0
0
0
0
0
0
J
93
1
0
0
0
0
0
0
0
0
94
1
0
0
0
0
0
0
0
0
95
2
0
0
0
0
0
0
0
0
98
1
0
0
0
0
0
0
0
0
107
10
4
0
1
0
0
0
0
0
108
15
2
0
3
0
0
1
1
0
109
14
1
0
0
0
0
0
0
0
39
1
0
0
0
0
0
0
0
0
47
2
0
0
0
0
0
0
0
0
63
1
1
0
0
0
0
0
0
0
68
2
2
0
0
0
0
0
0
0
69
1
1
0
0
0
0
0
0
0
77
1
0
0
1
0
0
0
0
0
80
1
0
0
1
0
0
0
0
0
81
1
1
0
0
0
1
0
0
0
84
1
1
0
0
1
0
0
0
0
10
86
1
0
0
1
0
0
0
0
0
90
1
1
0
0
0
0
0
0
0
95
3
3
0
2
0
0
0
0
0
97
1
1
0
0
1
0
0
0
0
100
1
1
0
1
0
0
0
0
0
102
1
1
0
1
0
0
0
0
0
103
1
1
0
1
0
0
0
0
0
104
3
3
0
3
0
0
0
0
0
107
12
12
0
11
0
0
0
1
0
108
11
11
0
11
0
0
1
0
0
109
6
5
0
3
0
0
0
0
0
330 DRAFT - DO NOT CITE OR QUOTE
-------
Table D-l. Tumor incidence data, with time to death with tumor; male rats
exposed by gavage to Benzo|alpyrene (Kroese et al., 2001)
Dose Group
(mg/kg-day)
Week of
Death
Total
Examined
Numbers of Animals with
Oral Cavity or
Forestomach Tumors
Liver Tumors
Duodenum
or Jejunum
Tumors
Skin or Mammary
Gland
Kidney
Urothelial
Carcinoma
Basal Cell
Tumors
Squamous
Cell Tumors
Incld."
Fatal"
Incid.
Fatal
Incid.
Incid.
Incid.
Incid.
32
1
1
0
0
0
0
0
0
0
35
1
1
0
1
0
0
0
0
0
37
1
1
0
0
0
0
0
0
0
44
1
0
1
1
0
0
0
0
0
45
2
0
2
0
0
0
0
0
47
1
1
0
1
0
0
0
0
0
48
1
1
0
1
0
0
0
0
0
49
1
1
0
1
0
0
0
0
0
50
1
1
0
1
0
0
0
0
0
51
1
1
0
1
0
1
0
0
0
30
52
4
3
1
3
1
0
1
1
0
53
1
1
0
1
0
0
1
0
0
56
2
1
1
1
1
0
0
0
0
58
2
2
0
2
0
0
1
0
0
59
2
2
0
2
0
0
0
0
0
60
2
1
1
1
1
1
0
0
0
61
3
2
1
1
2
1
0
0
0
62
5
5
0
0
4
3
0
0
0
63
5
5
0
4
1
1
2
1
2
64
2
2
0
1
1
0
0
0
1
65
3
2
1
1
2
0
3
2
0
66
1
1
0
0
1
0
0
0
0
67
3
1
2
2
1
1
1
1
0
68
1
1
0
1
0
0
0
0
0
70
2
2
0
1
1
1
1
0
0
71
1
1
0
1
0
0
1
1
0
73
1
0
1
1
0
0
1
0
0
76
1
1
0
0
1
0
1
0
0
a Incidental, denotes presence of tumors not known to have caused death of particular animals. "Fatal" denotes incidence of tumors reported by
the study investigators to have caused death of particular animals.
331
DRAFT - DO NOT CITE OR QUOTE
-------
Table D-2. Tumor incidence data, with time to death with tumor; female rats
Numbers of Animals with
Duodenum or
Week
Oral Cavity or Forestomach
Jejunum
Dose Group
Tumors
Liver Tumors
Tumors
of
(mg/kg-day)
Death
Total Examined
Incidental"
Fatal"
Incidental
Fatal
Incidental
64
1
0
0
0
0
0
69
1
0
0
0
0
0
75
1
0
0
0
0
0
0
104
1
0
0
0
0
0
106
4
0
0
0
0
0
107
7
0
0
0
0
0
108
7
0
0
0
0
0
109
30
1
0
0
0
0
8
1
0
0
0
0
0
47
1
0
0
0
0
0
52
1
0
0
0
0
0
60
1
0
0
0
0
0
65
1
0
0
0
0
0
76
1
0
0
0
0
0
77
1
0
0
0
0
0
83
0
0
0
0
0
3
85
1
0
0
0
0
0
86
1
0
0
0
0
0
88
1
0
0
0
0
0
93
0
0
0
0
0
94
1
0
0
0
0
0
97
1
1
0
0
0
0
107
6
2
0
1
0
0
108
9
2
0
0
0
0
109
21
1
0
0
0
0
42
1
0
0
0
0
0
43
1
0
0
0
0
0
44
1
0
0
0
0
0
45
1
0
0
0
0
0
48
1
0
0
0
0
0
55
1
0
0
1
0
0
59
1
0
0
0
0
0
75
1
0
0
1
0
0
76
0
0
1
0
0
77
80
1
0
1
0
0
0
1
0
0
0
0
81
1
1
0
0
1
0
82
1
1
0
1
0
0
83
1
0
1
0
0
85
1
0
1
1
0
10
86
1
1
0
0
1
0
87
88
89
1
1
1
1
0
0
0
1
1
0
0
1
1
0
0
0
91
1
0
0
0
1
0
95
1
0
0
0
0
0
96
1
0
0
0
0
0
98
2
0
1
1
0
99
3
0
1
2
0
102
1
1
0
0
1
0
104
1
1
0
1
0
0
105
2
1
0
1
1
0
106
1
1
0
0
1
0
107
5
5
0
5
0
0
108
7
7
0
7
0
0
109
4
2
0
2
0
0
332 DRAFT - DO NOT CITE OR QUOTE
-------
Table D-2. Tumor incidence data, with time to death with tumor; female rats
Dose Group
(mg/kg-day)
Week
of
Death
T otal Examined
Numbers of Animals with
Oral Cavity or Forestomach
Tumors
Liver Tumors
Duodenum or
Jejunum
Tumors
Incidental"
Fatal"
Incidental
Fatal
Incidental
26
1
0
0
0
0
0
44
4
4
0
3
1
0
47
3
3
0
2
1
0
48
1
1
0
0
1
0
54
1
0
0
1
0
0
55
3
3
0
1
2
0
56
2
2
0
0
2
0
57
2
2
0
2
0
0
58
4
3
1
0
4
0
59
2
1
1
0
2
0
30
60
1
0
1
1
0
0
61
2
2
0
0
2
0
62
2
2
0
1
1
0
63
3
3
0
0
3
0
64
5
5
0
0
5
3
66
3
3
0
0
3
0
67
2
1
1
0
2
0
68
1
1
0
0
1
0
69
4
3
1
1
3
1
71
4
3
1
1
3
0
72
2
1
1
0
2
0
' "Incidental" denotes presence of tumors not known to have caused death of particular animals. "Fatal" denotes incidence of tumors indicated by
the study investigators to have caused death of particular animals.
333
DRAFT - DO NOT CITE OR QUOTE
-------
Table D-3. Tumor incidence data, with time to
death with tumor; female mice exposed to
Benzojalpyrene via diet (Beland and Culp, 1998)
Dose
Number of Animals With Alimentary
Group
Week
Tract squamous cell tumors
(ppm in
of
Total
diet)
Death
Examined
Fatal3
Incidental
31
l
0
0
74
l
0
0
89
2
0
0
91
1
0
0
93
2
0
0
94
2
0
0
0
97
2
0
0
98
2
0
0
99
1
0
0
100
2
0
0
101
2
0
0
104
1
0
0
105
29
0
1
25
1
0
0
55
1
0
0
83
1
0
0
86
1
0
0
87
2
0
0
88
2
0
0
90
1
0
0
5
94
1
0
0
95
2
0
0
96
1
0
0
97
2
0
0
98
2
0
0
101
2
0
0
102
2
0
0
105
27
0
3
44
1
1
0
47
1
0
0
64
1
0
0
70
1
1
0
77
1
1
0
80
1
0
0
81
1
1
0
84
2
1
1
85
1
1
0
86
1
1
0
88
1
1
0
25
89
1
0
0
90
4
4
0
93
3
2
1
94
2
2
0
96
3
0
2
97
1
1
0
98
1
1
0
99
2
1
1
100
1
1
0
101
1
0
0
102
2
2
0
104
1
1
0
105
13
0
10
334
DRAFT - DO NOT CITE OR QUOTE
-------
Table D-3. Tumor incidence data, with time to
death with tumor; female mice exposed to
Benzojalpyrene via diet (Beland and Culp, 1998)
Dose
Number of Animals With Alimentary
Group
Week
Tract squamous cell tumors
(ppm in
of
Total
diet)
Death
Examined
Fatal3
Incidental
39
l
l
0
40
l
l
0
42
l
l
0
47
2
0
49
l
0
0
50
l
1
0
53
l
0
0
55
3
0
56
l
1
0
57
l
1
0
58
l
1
0
59
3
3
0
60
1
1
0
100
61
3
3
0
62
5
5
0
63
4
4
0
64
3
3
0
65
2
2
0
66
3
3
0
68
1
1
0
69
2
2
0
70
2
2
0
71
1
1
0
72
1
1
0
73
1
1
0
74
1
1
0
79
1
1
0
' "Incidental" denotes presence of tumors not known to have caused death of particular animals. "Fatal" denotes incidence of tumors indicated by
the study investigators to have caused death of particular animals.
335
DRAFT - DO NOT CITE OR QUOTE
-------
Table D-4. Derivation of HEDs to use for BMD
modeling of Wistar rat tumor incidence data from
Kroese et al. (2001)
benzo[a]pyrene dose
(mg/kg-d)
TWA body weight
(kg)
Interspecies
Scaling factor"
HEDb
(mg/kg-d)
Male
3
0.349
0.27
0.54
10
0.349
0.27
1.81
30
0.288
0.25
5.17
Female
3
0.222
0.24
0.49
10
0.222
0.24
1.62
30
0.222
0.24
4.85
a Scaling factors were calculated using U.S. EPA (1988) reference body weights for
humans (70 kg), and the TWA body weights for each dose group: rat-to-human =
(TWA body weight/70)025 = scaling factor.
b HED = administered dose x scaling factor.
Table D-5. Derivation of HEDs for BMD modeling of B6C3F1 female mouse
tumor incidence data from Beland and Culp (1998)
benzo[a]pyrene
dose in diet
(ppm)
Intake (jtg/d)
TWA body
weight average
(kg)
Administered
Dose" (mg/kg-d)
Scaling factorb
HEDC (mg/kg-d)
5
21
0.032
0.7
0.15
0.10
25
104
0.032
3.3
0.15
0.48
100
430
0.027
16.5
0.14
2.32
a Administered doses in mg/kg/day were calculated from dietary concentrations of benzo[a]pyrene using the TWA
body weight and reported food intakes for mice.
b Scaling factors were calculated using U.S. EPA (1988) reference body weights for humans (70 kg), and the TWA
body weights for each dose group: mouse-to-human = (TWA body weight/70)025 = scaling factor. HED =
administered dose x scaling factor.
0 HED = administered dose x scaling factor
336
DRAFT - DO NOT CITE OR QUOTE
-------
Table D-6. Summary of Model Selection and Modeling Results for best-fitting multistage-Weibull models, using
time-to-tumor data for rats (Kroese et al., 2001)
Selected model
Model
stages
Number
of
pa l am.
Responses at mg/kg-d levels'
parameter
estimates
Sex
EndPoints
LL"
r
AIC
BMDio
0
0.54
1.8
5.2
c
to
Model Selection Rationale
Males
Oral Cavity and
0
8
45
52
Forestomach:
1
-284.891
NR
4
577.8
0.104
0.0
21.1
39.8
38.7
2.2
44
Squam. Cell
2
-198.795
172.2
5
407.6
0.678
0.0
3.3
25.4
48.0
1.3
44
Tumors
3
-108.512
180.6
6
229.0
0.453
0.0
6.8
41.7
50.0
3.7
41
Lowest AIC, best fit to low dose data
Hepatocellular
0
4
38
49
Tumors
1
-179.664
NR
4
367.3
0.181
0
13.6
32.1
42
1
52
2
-145.749
67.8
5
301.5
0.472
0.0
6.5
37.2
49
2.3
48.4
3
-138.544
14.4
6
289.1
0.651
0.0
3.4
36.8
49.6
3.5
40.2
Lowest AIC, best fit to low dose data
Duodenum and
0
0
1
9
Jejunum
1
-31.781
NR
3
69.6
2.64
0.0
1.1
3.4
5.7
1
NR
Tumors
2
-28.941
5.7
4
65.9
3.04
0.0
0.2
1.8
8.2
1
NR
3
-28.439
1.0
5
66.9
3.03
0.0
0.0
1.0
9.0
1.8
NR
Best fit to data
Kidney:
0
0
0
3
Uroethelial
1
-12.956
NR
3
31.9
9.16
0.0
0.3
1
1.7
1
NR
Carcinoma
2
-11.837
2.2
4
31.7
5.71
0.0
0.1
0.5
2.5
1
NR
3
-11.398
0.9
5
32.8
4.65
0.0
0.0
0.3
2.7
1.7
NR
Best fit to data
Skin and
2
1
1
13
Mammary
1
-52.314
NR
3
110.6
1.88
1.0
2.5
5.5
8.3
1
NR
Gland: Basal
2
-48.570
7.5
4
105.1
2.58
1.1
1.3
3.4
11.4
1
NR
Cell Tumors
3
-47.362
2.4
5
104.7
2.86
1.2
1.2
2.3
12.5
1.4
NR
Lowest AIC, best fit to low dose data
Skin and
0
1
1
6
Mammary
1
-28.745
NR
3
63.5
3.36
0.0
0.9
2.7
4.5
1
NR
Lowest AIC, best fit to low dose data
Gland: Squam.
2
-28.145
1.2
4
64.3
2.75
0.0
0.4
2.4
5.2
1.9
NR
Cell Tumors
3
-27.652
1.0
5
65.3
2.64
0.0
0.4
2.1
5.5
3.0
NR
Females
Oral Cavity and
1
6
30
50
Forestomach:
1
-134.532
NR
4
277.1
0.245
1.0
10.1
23.7
35.1
1
58
Squam. Cell
2
-100.809
67.4
5
211.6
0.428
0.8
6.8
33.2
47.9
2.5
52
Tumors
3
-94.512
12.6
6
201.0
0.539
1.1
4.9
31.8
49.4
3.5
47
Lowest AIC, best fit to low dose data
Flepatocellular
0
1
39
51
Tumors
1
-293.771
NR
4
595.5
0.146
0.0
14.6
32.6
43.8
1
44
2
-382.470
177.4
5
774.9
0.370
0.0
8.1
38.5
50.1
2.2
44
3
-228.170
308.6
6
468.3
0.575
0.0
3.0
38.4
51.4
3.1
39
Lowest AIC, best fit to low dose data
Duodenum and
0
0
0
4
Jejunum
1
-15.948
NR
3
37.9
6.00
0.0
0.4
1.3
2.4
1
NR
Tumors
2
-14.518
2.9
4
37.0
4.33
0.0
0.0
0.7
3.3
1.1
NR
3
-13.878
1.3
5
37.8
3.43
0.0
0.0
0.4
3.6
2.3
NR
Best fit to low dose data
" LL=log-likelihood.
b X2 = chi-squared statistic for testing the difference between 2 model fits, from 2 x |(LLj - LLj) evaluated for i-j degrees of freedom. In all cases the difference was evaluated for consecutive numbers of
stages; i-j = 1, and ¦£ at a = 0.05 is 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.
337
DRAFT - DO NOT CITE OR QUOTE
-------
Table D-6. Summary of Model Selection and Modeling Results for best-fitting multistage-Weibull models, using
time-to-tumor data for rats (Kroese et al., 2001)
Selected model
Model
stages
Number
of
pa l am.
Responses at mg/kg-d levels'
parameter
estimates
Sex
EndPoints
LL"
AIC
BMDio
0
0.54
1.8
5.2
c
to
Model Selection Rationale
Males
Oral Cavity and
0
8
45
52
Forestomach:
1
-284.891
NR
4
577.8
0.104
0.0
21.1
39.8
38.7
2.2
44
Squam. Cell
2
-198.795
172.2
5
407.6
0.678
0.0
3.3
25.4
48.0
1.3
44
Tumors
3
-108.512
180.6
6
229.0
0.453
0.0
6.8
41.7
50.0
3.7
41
Lowest AIC, best fit to low dose data
Hepatocellular
0
4
38
49
Tumors
1
-179.664
NR
4
367.3
0.181
0
13.6
32.1
42
1
52
2
-145.749
67.8
5
301.5
0.472
0.0
6.5
37.2
49
2.3
48.4
3
-138.544
14.4
6
289.1
0.651
0.0
3.4
36.8
49.6
3.5
40.2
Lowest AIC, best fit to low dose data
Duodenum and
0
0
1
9
Jejunum
1
-31.781
NR
3
69.6
2.64
0.0
1.1
3.4
5.7
1
NR
Tumors
2
-28.941
5.7
4
65.9
3.04
0.0
0.2
1.8
8.2
1
NR
3
-28.439
1.0
5
66.9
3.03
0.0
0.0
1.0
9.0
1.8
NR
Best fit to data
Kidney:
0
0
0
3
Uroethelial
1
-12.956
NR
3
31.9
9.16
0.0
0.3
1
1.7
1
NR
Carcinoma
2
-11.837
2.2
4
31.7
5.71
0.0
0.1
0.5
2.5
1
NR
3
-11.398
0.9
5
32.8
4.65
0.0
0.0
0.3
2.7
1.7
NR
Best fit to data
Skin and
2
1
1
13
Mammary
1
-52.314
NR
3
110.6
1.88
1.0
2.5
5.5
8.3
1
NR
Gland: Basal
2
-48.570
7.5
4
105.1
2.58
1.1
1.3
3.4
11.4
1
NR
Cell Tumors
3
-47.362
2.4
5
104.7
2.86
1.2
1.2
2.3
12.5
1.4
NR
Lowest AIC, best fit to low dose data
Skin and
0
1
1
6
Mammary
1
-28.745
NR
3
63.5
3.36
0.0
0.9
2.7
4.5
1
NR
Lowest AIC, best fit to low dose data
Gland: Squam.
2
-28.145
1.2
4
64.3
2.75
0.0
0.4
2.4
5.2
1.9
NR
Cell Tumors
3
-27.652
1.0
5
65.3
2.64
0.0
0.4
2.1
5.5
3.0
NR
Females
Oral Cavity and
1
6
30
50
Forestomach:
1
-134.532
NR
4
277.1
0.245
1.0
10.1
23.7
35.1
1
58
Squam. Cell
2
-100.809
67.4
5
211.6
0.428
0.8
6.8
33.2
47.9
2.5
52
Tumors
3
-94.512
12.6
6
201.0
0.539
1.1
4.9
31.8
49.4
3.5
47
Lowest AIC, best fit to low dose data
Flepatocellular
0
1
39
51
Tumors
1
-293.771
NR
4
595.5
0.146
0.0
14.6
32.6
43.8
1
44
2
-382.470
177.4
5
774.9
0.370
0.0
8.1
38.5
50.1
2.2
44
3
-228.170
308.6
6
468.3
0.575
0.0
3.0
38.4
51.4
3.1
39
Lowest AIC, best fit to low dose data
Duodenum and
0
0
0
4
Jejunum
1
-15.948
NR
3
37.9
6.00
0.0
0.4
1.3
2.4
1
NR
Tumors
2
-14.518
2.9
4
37.0
4.33
0.0
0.0
0.7
3.3
1.1
NR
3
-13.878
1.3
5
37.8
3.43
0.0
0.0
0.4
3.6
2.3
NR
Best fit to low dose data
NR = not relevant.
338
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Male Rat (Kroese et al., 2001): Squamous Cell Papilloma or Carcinoma in Oral Cavity or
Forestomach
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: OralForstKroeseM3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 3.6
t_0 = 39.1111
beta_0 = 0
beta_l = 8 . 8911e-009
beta_2 = 1.60475e-031
beta 3 = 1.95818e-008
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -beta_0 -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c t_0 beta_l beta_3
c 1 -0.53 -0.93 -0.99
t_0 -0.53 1 0.47 0.57
beta_l -0.93 0.47 1 0.9
beta 3 -0.99 0.57 0.9 1
Variable
c
t_0
beta_0
beta_l
beta_2
beta 3
Estimate
3 .74559
41.4581
0
4 .37816e-009
0
1. 01904e-008
Parameter Estimates
Std. Err.
0 .447309
2 .14975
NA
1. 07528e-008
NA
1.94164e-008
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
2.86888 4.6223
37.2447 45.6716
-1.66 97e-008
-2.78651e-008
2 .54533e-008
4 .82458e-008
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param
Fitted Model -108.512 6
AIC
229 . 024
DOSE
Data Summary
CONTEXT
F I
U Total Expected Response
0
52
0
0
0
52
0 . 00
0.54
44
0
8
0
52
6 .77
1.8
7
0
45
0
52
41.69
5.2
0
9
43
0
52
49 . 97
Minimum observation time for F tumor context =
44
339
DRAFT - DO NOT CITE OR QUOTE
-------
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Confidence level = 0.9
Time = 104
Specified effect = 0.1 0.01 0.001
BMD = 0.453471 0.0633681 0.00636659
BMDL = 0.281044 0.0286649 0.00285563
BMDU = 0.612462 0.248377 > 0.0509326
Incidental Risk: OralForstKroeseM3
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose = 0.00 Dose = 0.54
~ 20 40 60 80 100 ~ 20 40 60 80 100
Time
Time
340
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
84
Male Rat (Kroese et al., 2001): Hepatocellular Adenoma or Carcinoma
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: LiverKroeseM3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 3.6
t_0 = 34.6667
beta_0 = 0
beta_l = 2 . 73535e-009
beta_2 = 8.116e-028
beta 3 = 1.43532e-008
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -beta_0 -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c t_0 beta_l beta_3
c 1 -0.84 -0.88 -1
t_0 -0.84 1 0.71 0.86
beta_l -0.88 0.71 1 0.86
beta 3 -1 0.86 0.86 1
Variable
c
t_0
beta_0
beta_l
beta_2
beta 3
Estimate
3 .49582
40 .2211
0
4 .43906e-009
0
2 .35065e-008
Parameter Estimates
Std. Err.
0.629257
5.65421
NA
1.76 051e-0 08
NA
6 .4799 9e-008
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
2.26249 4.72914
29.1391 51.3032
-3 . 00664e-008
-1. 0349 9e-007
3 .89445e-008
1.50512e-007
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood)
# Param
AIC
Fitted
Model
-
138.544
6
289.088
Data
. Summary
CONTEXT
C
F
I
U
Total
Expected Resp<
DOSE
0
52
0
0
0
52
0 . 00
0.54
48
0
4
0
52
3 .38
1.8
14
2
36
0
52
36 .81
5.2
3
17
32
0
52
49.55
Minimum observation time for F tumor context =
52
341
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Confidence level = 0.9
Time = 104
Specified effect = 0.1 0.01 0.001
BMD = 0.6507 0.173556 0.0199908
BMDL = 0.44868 0.0530469 0.00530386
BMDU = 0.772467 0.352684 > 0.159927
Incidental Risk: Hepatocellular_Kroese_M3
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose = 0.00
Dose = 0.54
-Q
05
-Q
O
00
O
O
O
O
i i r
20 40 60 80 100
-Q
05
-Q
O
00
O
O
O
O
i i i i i r
0 20 40 60 80 100
Time
Time
Dose = 1.81
Dose = 5.17
-Q
05
_Q
O
00
O
O
O
O
OQDQO O.O
i 1 1—i—r
20 40 60 80 100
-Q
05
_Q
O
00
O
O
i 1 1 1—i—T
0 20 40 60 80 100
Time
Time
342
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Male Rat (Kroese et al., 2001): Duodenum or Jejunum Adenocarcinoma
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: DuoJejKroeseM3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
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 = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 1.63636
t_0 = 0 Specified
beta_0 = 4 . 31119e-027
beta_l = 2 . 96347e-025
beta_2 = 0
beta 3 = 1.76198e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0 -beta_l -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta_3
1 -1
-1 1
c
beta 3
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
1.77722
0
0
0
9 .82635e-007
Parameter Estimates
Std. Err.
2 . 03042
NA
NA
NA
8.29355e-006
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-2.20233 5.75677
-1.52 724e-0 05
1.72377e-005
NA
Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -28.4387 5 66.8773
DOSE
Data Summary
CONTEXT
F I
U Total Expected Response
0
52
0
0
0
52
0 . 00
0.54
52
0
0
0
52
0 . 03
1.8
51
0
1
0
52
1. 04
5.2
43
0
9
0
52
8 . 96
Benchmark Dose Computation
Risk Response = Incidental
343
DRAFT - DO NOT CITE OR QUOTE
-------
Risk Type = Extra
Specified effect = 0.1
Confidence level = 0.9
Time = 104
Specified effect = 0.1 0.01 0.001
BMD = 3.03291 1.38578 0.642252
BMDL = 2.37782 0.418285 0.0420835
BMDU = 3.87183 1.76166 0.811476
Incidental Risk: DuoJej_Kroese_M3
Dose = 0.00
Dose = 0.54
-Q
<0
-Q
O
LO
O
O
O _|
"T"
20
~i r"
40 60
~i r~
80 100
— LO
_Q t-
2 o
o
o
o
"T"
20
n—r
40 60
n—r~
80 100
Time
Time
Dose = 1.81 Dose = 5.17
Time
Time
344
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Male Rat (Kroese et al., 2001): Skin or Mammary Gland Basal Cell Tumors
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: SKinMamBasalKroeseM3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
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 = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 1.38462
t_0 = 0 Specified
beta_0 = 3.84298e-005
beta_l = 1. 06194e-028
beta_2 = 0
beta 3 = 6 .84718e-006
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_l -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta_0 beta_3
c 1-1-1
beta_0 -1 1 0.99
beta 3 -1 0.99 1
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
1.47227
2 . 54786e-005
0
0
4 .81611e-006
Parameter Estimates
Std. Err.
1.76686
0 . 000211261
NA
NA
3.4 9e-0 05
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-1.9907 4.93525
-0 . 0 0 03 88585 0 . 0 0 04 3 9542
-6 .35866e-005
7 .32188e-005
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param
AIC
Fitted
Model
-47.3623
5
104.725
Data Summary
CONTEXT
C
F I
U
Total
Expected Response
DOSE
0
50
0 2
0
52
1.18
0.54
51
0 1
0
52
1.22
1.8
51
0 1
0
52
2 .32
5.2
39
0 13
0
52
12 .54
345
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Confidence level = 0.9
Time = 104
Specified effect = 0.1 0.01 0.001
BMD = 2.86276 1.30804 0.606222
BMDL = 2.35118 0.415897 0.0424277
BMDU = 3.62258 1.69571 0.761447
Incidental Risk: Skin Mam Basal Kroese M3
Dose = 0.54
Dose = 1.81
05
-Q
o
00
o
o
o
o
0 20 40 60
Time
80
20 40 60
Dose = 5.17
05
O
00
O
O
O
o
I I I
20 40 60
Time
346
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
84
Male Rat (Kroese et al., 2001): Skin or Mammary Gland Squamous Cell Tumors
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: SKinMamSCCKroeseM3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
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 = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c 3
t_0 = 0 Specified
beta_0 = 0
beta_l = 1. 25256e-008
beta_2 = 1.25627e-030
beta 3 = 3 .34696e-009
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0 -beta_2
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 beta_3
c 1 -0.99 -1
beta_l -0.99 1 0.99
beta 3 -1 0.99 1
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
2 . 96213
0
1.50104e-008
0
3 . 9084e-009
Parameter Estimates
Std. Err.
2 .591
NA
1.86 972e-0 07
NA
4 .15374e-008
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-2.11613 8.04039
-3.51447e-007
-7.75033e-008
3 .81468e-007
8.53201e-008
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood)
# Param
AIC
Fitted
Model
-27.652
5
65.304
Data
Summary
CONTEXT
C
F
I
U
Total
Expected Resp<
DOSE
0
52
0
0
0
52
0 . 00
0.54
51
0
1
0
52
0 .42
1.8
51
0
1
0
52
2 .12
5.2
46
0
6
0
52
5.51
Benchmark Dose Computation
Risk Response = Incidental
347
DRAFT - DO NOT CITE OR QUOTE
-------
Risk Type = Extra
Confidence level = 0.9
Time = 104
Specified effect = 0.1 0.01 0.001
BMD = 2.6414 0.64109 0.070558
BMDL = 1.76931 0.211043 0.0210552
BMDU = 4.42145 2.03605 > 0.564463
Incidental Risk: OralForstKroeseM3
points show nonpararn. est. for Incidental (unfilled) and Fatal (filled)
Dose= 0.00 Dose= 0.54
CO
CD
CO
CD
CN
i=i
O
CD
i 1 1 1 r
20 40 60 80 100
Time
Dose = 1.81
CO
CD
CO
C=i
CD
cn
CD
CD
CD
20 40 60 30 100
Time
CO
CD
CO
CD
r-j
CD
CD
CD
i 1 1 1 r
20 40 60 80 100
Time
Dose = 5.17
CO
CD
CO
CD
CD
CN
C=i
O
CD
20 40 60 80 100
Time
Dose Response plot
348
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Male Rat (Kroese et al., 2001): Kidney Urothelial Carcinomas
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: KidneyUrothelialCarKroeseM3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
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 = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 1.63636
t_0 = 0 Specified
beta_0 = 3 . 78734e-027
beta_l = 1. 59278e-027
beta_2 = 2.718e-024
beta 3=4.96063e-007
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0 -beta_l -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta_3
1 -1
-1 1
c
beta 3
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
1.74897
0
0
0
3 .11107e-007
Parameter Estimates
Std. Err.
3 .79403
NA
NA
NA
4 . 90313e-006
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-5.68719 9.18512
-9.29885e-006
9.92107e-006
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood)
Fitted Model -11.3978
# Param
5
AIC
32 .7956
Data Summary
CONTEXT
C
F
I
U
Total
Expected
DOSE
0
52
0
0
0
52
0 . 00
0.54
52
0
0
0
52
0 . 01
1.8
52
0
0
0
52
0.29
5.2
49
0
3
0
52
2 .71
Benchmark Dose Computation
349
DRAFT - DO NOT CITE OR QUOTE
-------
Risk Response = Incidental
Risk Type = Extra
Confidence level = 0.9
Time = 104
Specified effect = 0.1 0.01 0.001
BMD = 4.64886 2.12413 0.984449
BMDL = 2.49972 0.734665 0.0748097
BMDU = 9.01023 3.49311 1.61892
Incidental Risk: Kidney_Kroese_M3
Dose = 0.00 Dose = 0.54
in
o
20 40 60 80 1 00
Time
in
o
350
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
Female Rat (Kroese et al., 2001): Oral Cavity or Forestomach, Squamous Cell Papilloma
or Carcinoma
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: OralForstKroeseF3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 3.6
t_0 = 45.1111
beta_0 = 1.11645e-009
beta_l = 4 . 85388e-009
beta_2 = 0
beta_3 = 1.95655e-008
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c t_0 beta_0 beta_l beta_3
c 1 -0.79 -0.92 -0.93 -1
t_0 -0.79 1 0.73 0.72 0.8
beta_0 -0.92 0.73 1 0.79 0.92
beta_l -0.93 0.72 0.79 1 0.91
beta 3 -1 0.8 0.92 0.91 1
Variable
c
t_0
beta_0
beta_l
beta_2
beta 3
Estimate
3 .52871
46 .553
1.53589e-009
7 .57004e-009
0
2 .53126e-008
Parameter Estimates
Std. Err.
0.701117
5 . 93306
5 .4052 3e-009
2 . 9647e-008
NA
7.664 04e-008
95.0% Wald Confidence Interval
Lower Conf. Limit
2 .15454
34 . 9244
-9 . 05817e-009
-5 . 0536 9e-008
-1.249e-007
Upper Conf. Limit
4 . 90287
58.1816
1.21299e-008
6 .5677e-008
1.75525e-007
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood)
# Param
AIC
Fitted
Model
-
94 .5119
6
201. 024
Data
. Summary
CONTEXT
C
F
I
U
Total
Expected Response
DOSE
0
51
0
1
0
52
1.14
0.49
46
0
6
0
52
4 . 90
1.6
22
0
30
0
52
31.81
4.6
2
7
43
0
52
49.43
351
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Minimum observation time for F tumor context =
58
Benchmark Dose Computation
Risk Response
Risk Type
Confidence level =
Time
Incidental
Extra
0.9
104
Specified effect = 0.1
BMD = 0.538801
BMDL = 0.328135
BMDU = 0.717127
0 .01
0 .0981283
0.0345104
0.325909
0 . 001
0 . 0100797
0 . 00344714
> 0.0806373
Incidental Risk: OralForstKroeseF3
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose = 0.00 Dose = 0.49
_Q
CO
_Q
O
CO
O
¦sf
d
o
d
80 100
1 1 1 T
0 20 40 60
_Q
CO
_Q
O
00
o
¦sf
o
i—i—i—i—i—r
0 20 40 60 80 100
Time
Time
Dose = 1.62
Dose = 4.58
_Q
CO
_Q
O
00
o
¦sf
o
i—i—i—i—i—r
0 20 40 60 80 100
_Q
CO
_Q
O
00
o
¦sf
o
o
o
20 40 60 80 100
Time
Time
352
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Female Rat (Kroese et al., 2001): Hepatocellular Adenoma or Carcinoma
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: LiverKroeseF3.(d)
Fri Apr 16 09:08:03 2010
Timer to Tumor Model, Liver Hepatocellular Tumors, Kroese et al, Female
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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 3.6
t_0 = 31.7778
beta_0 = 0
beta_l = 4 . 9104e-031
beta_2 = 5.45766e-030
beta 3 = 3 .44704e-008
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -beta_0 -beta_l -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c t_0 beta_3
1 -0.9 -1
-0.9 1 0.92
-1 0.92 1
c
t_0
beta 3
Parameter Estimates
95.0% Wald Confidence Interval
Variable
Estimate
Std. Err.
Lower Conf. Limit
Upper Conf. Limit
c
3 .11076
0.549208
2 . 03434
4 .18719
t_0
38.6965
5 .21028
28 .4846
48 . 9085
beta 0
0
NA
beta 1
0
NA
beta 2
0
NA
beta 3
2 . 94354e-007
7 .19418e-007
-1.1156 8e-0 06
1.7043 9e-006
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param
AIC
Fitted
Model
-228.17
6
468.34
Data
Summary
CONTEXT
C
F
I
U
Total
Expected Response
DOSE
0
52
0
0
0
52
0 . 00
0.49
51
0
1
0
52
3 . 02
1.6
13
12
27
0
52
38.36
4.6
1
38
13
0
52
51.36
353
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Minimum observation time for F tumor context = 44
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Confidence level = 0.9
Time = 104
Specified effect = 0.1 0.01 0.001
BMD = 0.575127 0.262783 0.12179
BMDL = 0.506633 0.134213 0.0152934
BMDU = 0.629806 0.287232 0.133064
Incidental Risk: Hepatocellular_Kroese_F3
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose = 0.00
Dose = 0.49
n
TO
n
2
CL
00
d
d
o
d
T
0
20
~l 1
40 60
n r~
80 100
n
TO
n
2
CL
00
o
o
o
o
T
0
20
"I 1-
40 60
~1 1—
80 100
Time
Time
Dose = 1.62
Dose = 4.58
n
TO
n
o
oo
o
o
o
o
n
ro
n
o
oo
o
o
o
o
0 20 40 60 80 100
T"
0
20
40 60
80 100
Time
Time
354
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Female Rat (Kroese et al., 2001): Duodenum or Jejunum Adenocarcinoma
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: DuoJejKroeseF3.(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 = CONTEXT
Independent variables = DOSE, TIME
Total number of observations = 2 08
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 = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
Default Initial Parameter Values
c = 2.25
t_0 = 0 Specified
beta_0 = 0
beta_l = 0
beta_2 = 0
beta 3 = 7.289e-008
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0 -beta_l -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta_3
1 -1
-1 1
c
beta 3
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
2 .32531
0
0
0
5.3220 9e-008
Parameter Estimates
Std. Err.
3 .58729
NA
NA
NA
7 . 98487e-007
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
-4.70565 9.35626
-1.51178e-0 06
1.61823e-006
NA - Indicates that this parameter has hit a bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -13.8784 5 37.7569
DOSE
Data Summary
CONTEXT
F I
U Total Expected Response
0
52
0
0
0
52
0 . 00
0.49
52
0
0
0
52
0 . 01
1.6
52
0
0
0
52
0 .44
4.6
48
0
4
0
52
3 .57
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
Confidence level = 0.9
355 DRAFT - DO NOT CITE OR QUOTE
-------
G\ Lh
Probability
o\
=\
3
a>
CD
II
ff>
ho
Probability
o
o
o
2
o
H
O
HH
H
ffl
o
o
c
o
H
w
=1
3
0
CD
II
¦Pk.
bi
oo
-p- u>
Probability
cn
fl)
=1
3
0
fl)
a
CD
m
Probability
D
c
o
=\
3
cd
-------
Table D-7. Summary of human equivalent overall cancer risk values, based on male and
Data set
Tumor Site
BMDooi
BMDLooi
Risk value® at
SD
SD2
Prop, of
total
variance
BMDooi
BMDLooi
Males
Oral cavity/forestomach
6.37E-03
2.86E-03
1.57E-01
3.50E-01
1.17E-01
1.38E-02
0.64
Liver
2.00E-02
5.30E-03
5.00E-02
1.89E-01
8.42E-02
7.09E-03
0.33
Duodenum/ j ejunum
6.42E-01
4.21E-02
1.56E-03
2.38E-02
1.35E-02
1.82E-04
0.01
Skin/mammary gland:
basal cell
6.06E-01
4.24E-02
1.65E-03
2.36E-02
1.33E-02
1.78E-04
0.01
Skin/mammary gland:
squam. cell
7.06E-02
2.11E-02
1.42E-02
4.75E-02
2.03E-02
4.10E-04
0.02
Kidney
9.84E-01
7.48E-02
1.02E-03
1.34E-02
7.51E-03
5.64E-05
0.00
Sum, risk values at BMDooi:
2.25E-01
Sum, SD2:
2.17E-02
Overall SDb:
1.47E-01
Upper bound on sum of risk estimates0:
4.68E-01
Females
Oral cavity/forestomach
3.45E-03
1.01E-02
2.90E-01
9.92E-02
1.16E-01
1.35E-02
0.91
Liver
1.53E-02
1.22E-01
6.54E-02
8.21E-03
3.48E-02
1.21E-03
0.08
Duodenum/ jejunum
5.85E-02
7.27E-01
1.71E-02
1.38E-03
9.56E-03
9.13E-05
0.01
Sum, risk values at BMDooi:
1.09E-01
Sum, SD2:
1.48E-02
Overall SD:
1.22E-01
Upper bound on sum of risk estimates0:
3.09E-01
" Risk value=0.001/BMDL0oi
b Overall SD = (Sum, SD2)05
0 Upper bound on the overall risk estimate :
Sum of BMDooi risk values + 1.645 x Overall SD.
Table D-8. Summary of model selection among multistage-Weibull models fit to
alimentary tract tumor data for female mice (Beland and Culp, 1998)
Model
stages
1
2
LL"
Number
of
X2 b param.
AIC
BMD1(
-340.271 NR
-309.620 61.3
-306.265 6.7
688.5 0.104
629.2 0.102
624.5
0.127
Responses (a} mg/kg-d levels'
0
0.1 0.48
2.3
1
0.6
0.7
0.9
3 38
14.6 34.1
4.5 33.2
3.2 30.8
46
36.3
41.8
41.9
Selected model
parameter
estimates
to
3.4
5.5
6.9
18
16
14
Model Selection
Rationale
Lowest AIC, best lit
to low dose data
" LL=log-likelihood.
b ¦£ = chi-squared statistic for testing the difference between 2 model fits, from 2 x |(LLj - LLj) evaluated for i-j degrees of freedom. In all
cases the difference was evaluated for consecutive numbers of stages; i-j = 1, and at a = 0.05 is 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.
NR = not relevant.
357
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
84
Female Mice (Beland and Culp, 1998): Alimentary Tract Squamous Cell Tumors
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: C:\mswl0-09\benzo[a]pyrene_FemaleSquamF3i.(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 = 191
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 64
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.49012e-008
User Inputs Initial Parameter Values
c 2
t_0 = 15
beta_0 = 1.6e-014
beta_l = 0
beta_2 = 5.5e-012
beta 3 = 4.4e-012
Asymptotic Correlation Matrix of Parameter Estimates
c
t_0
beta 0
beta 1 beta 2 beta 3
c
1
-0.78
-0 . 97
-0.42 -0.99 -0.99
ft
1
o
o
.78
1
0.76
0.39 0.74 0.84
beta 0 -0
. 97
0.76
1
0.33 0.97 0.96
beta 1 -0
.42
0.39
0.33
1 0.31 0.46
beta 2 -0
. 99
0.74
0 . 97
0.31 1 0.97
beta 3 -0
. 99
0.84
0 . 96
0.46 0.97 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable
Estimate
Std.
Err.
Lower Conf. Limit Upper Conf. Limit
c
6 . 92317
1.
33874
4.29929 9.54705
t_0
13 . 9429
4 .
96646
4.20881 23.677
beta 0
2.46916e-016
1.47619
e- 015
-2 . 64636e-015 3 .14 019e-015
beta 1
0
1.3 052 5e-014
-2 . 55825e-014 2 .55825e-014
beta 2
5.85452e-014
3.75144e-013
-6 .76 723e-013 7 . 93 813e-013
beta 3
9.76542e-014
5.62017e-013
-1. 00388e-012 1.19919e-012
Log(likelihood) #
Param
AIC
Fitted Model
306.265
6
624
.53
Data Summary
Class
C
F
I
U Total
Expected
Response
Dose
0 47
0
1
0 48
0 . 93
0.1 45
0
3
0 48
3.21
0.48 8
23
15
1 47
30.82
2.3 1
46
0
1 48
41. 91
Minimum observation time for F tumor context =
39
Benchmark Dose Computation
Risk Response = Incidental
358
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Risk Type = Extra
Specified effect = 0.1
Confidence level = 0.9
Time = 104
BMD = 0.126983
BMDL = 0.0706103
BMDU = 0.179419
Incidental Risk: BaP_FemaleSquamF3i
points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
Dose = 0.00
Dose= 0.10
CO
o
o
r-j
czi
oo
CD
(JD
o
o
C-i
O
CD
CD
Dose= 0.48
Dose = 2.32
Time
Time
359
DRAFT - DO NOT CITE OR QUOTE
-------
1 APPENDIX E. TIME-TO-TUMOR MODELING FOR THE INHALATION UNIT RISK
2
3
Table E-l
. Individual pathology and tumor occurrence data for male Syrian
hamsters exposed to benzo|alpyrene via inhalaf
ion for lifetime (Thyssen et al.,
1981)
Admin.
Exposure
Time
Cone.
on
Number
Nasal
(mg/m3)
Study
Examined
Larynx
Pharynx
Trachea
Esophagus
Forestomach
Cavity
17
1
0
oa
0
0
0
0
39
1
0
0
0
0
0
0
45
1
0
0
0
0
0
0
79
1
0
0
0
0
0
0
83
1
0
0
0
0
0
0
85
1
0
oa
0
0
0
0
86
1
0
0
0
0
0
0
88
0
0
0
0
0
0
89
0
0
0
0
0
0
90
1
0
0
0
0
0
0
101
1
0
0
0
0
0
0
102
1
0
0
0
0
0
0
0
103
1
0
0
0
0
0
0
106
1
0
0
0
0
0
0
108
1
0
0
0
0
0
0
109
1
0
0
0
0
0
0
112
1
0
0
0
0
0
0
115
1
0
0
0
0
0
0
116
1
0
oa
0
0
0
0
122
1
0
0
0
0
0
0
123
1
0
0
0
0
0
0
124
1
oa
0
0
0
0
0
125
1
0
0
0
0
0
0
127
1
0
oa
0
0
0
0
132
1
0
0
0
0
0
0
14
1
oa
oa
0
0
0
0
35
1
0
0
0
0
0
0
53
1
0
0
0
0
0
0
59
1
0
0
0
0
0
0
71
1
0
0
0
0
0
0
78
1
0
0
0
0
0
0
80
1
0
0
0
0
0
0
85
1
0
0
0
0
0
0
87
1
0
0
0
0
0
0
88
1
0
0
0
0
0
0
93
1
0
0
0
0
0
0
o
98
1
0
oa
0
0
0
0
L
99
1
0
0
0
0
0
0
102
1
0
0
0
0
0
0
103
1
0
0
0
0
0
0
108
1
0
0
0
0
0
0
111
1
0
0
0
0
0
0
113
1
0
0
0
0
0
0
114
1
0
0
0
0
0
0
115
1
0
0
0
0
0
0
116
1
0
0
0
0
0
0
117
1
0
0
0
0
0
0
120
1
0
0
0
0
0
0
122
2
oa
oa
0
0
0
0
133
2
0
0
0
0
0
0
360
DRAFT - DO NOT CITE OR QUOTE
-------
Table E-l
. Individual pathology and tumor occurrence data for male Syrian
hamsters exposed to benzo|alpyrene via inhalaf
ion for lifetime (Thyssen et al.,
1981)
Admin.
Exposure
Time
Cone.
on
Number
Nasal
(mg/m3)
Study
Examined
Larynx
Pharynx
Trachea
Esophagus
Forestomach
Cavity
31
1
0
0
0
0
0
0
32
1
0
0
0
0
0
0
52
1
0
0
0
0
0
0
67
1
0
0
0
0
0
0
73
1
0
0
0
0
0
0
76
0
2
0
0
0
0
80
1
1
0
0
0
0
0
85
1
0
0
0
0
0
0
94
1
1
0
0
0
0
0
100
1
0
0
0
0
0
0
10
102
1
0
1
0
0
0
0
105
1
1
1
0
0
0
0
111
1
0
1
0
0
0
0
113
1
0
1
0
0
0
0
114
1
1
1
0
0
0
0
115
1
1
oa
1
0
0
1
116
1
0
0
1
0
0
1
117
1
1
0
0
0
0
0
118
3
lb
0
0
1
1
122
1
1
0
0
0
0
0
124
1
1
1
0
0
0
0
125
1
0
0
0
0
0
1
20
1
oa
oa
oa
0
0
0
21
1
oa
oa
oa
0
0
0
25
oa
oa
oa
0
0
0
29
1
oa
oa
oa
0
0
0
30
1
oa
oa
oa
0
0
0
34
1
oa
oa
oa
0
0
0
36
oa
oa
oa
0
0
0
37
1
oa
oa
oa
0
0
0
40
r
r
r
0
0
0
41
1
0
0
0
0
0
0
43
1
0
0
0
0
0
0
47
1
l
l
0
0
0
0
50
48
1
0
l
0
0
0
0
51
1
0
oa
0
0
0
0
56
1
l
1
0
0
0
0
57
1
0
1
0
0
0
0
60
1
0
1
0
0
0
0
63
1
0
0
0
0
0
64
1
0
1
0
0
1
0
66
1
l
1
0
0
0
0
68
1
0
1
0
0
0
0
70
1
l
1
0
1
0
0
71
1
l
1
1
0
0
0
72
1
l
1
0
0
0
0
73
2
2
2
0
0
0
0
79
4
3
4
1
1
0
1
1
2 a Tissue was not examined for one animal of total examined.
3 b Tissue was not examined for two animals of total examined.
361
DRAFT - DO NOT CITE OR QUOTE
-------
Table E-2. Summary of model selection among multistage-Weibull models fit to tumor
data for male hamsters3 (Thyssen et al., 1981)
Model
stages
Number
of
LLb /;c param. AIC BMD10
Responses (a} mg/kg-d levels'1
Selected model
parameter
estimates
Model Selection
Rationale
0
0.25
1.01
4.29
c
to
Oral tract tumors: all tumors considered incidental to
cause of death
0
0
18
18
1
-26 NR 3 58 0.090
0.0
5.6
15.9
17.2
2.2
NR
2
-19.967 12.1 4 47.9 0.285
0.0
1.9
16.0
18.2
4.2
NR
Lowest AIC, best fit
to data; maximum
number of stages that
could be fit
Oral tract tumors: all tumors considered to be cause of
death
0
0
18
18
1
-160.646 NR 3 327.292 0.136
Not available
4.9
NR
2
-147.428 26.4 4 302.857 0.421
Not available
6.7
NR
3
-144.522 5 299.043 0.648
Not available
9.0
NR
Lowest AIC; best fit
to data (see graphs)
a All animals with missing tissues were omitted.
b LL=log-likelihood.
c ¦£ = chi-squared statistic for testing the difference between 2 model fits: f=2* (LLj - LLj) evaluated for i-j degrees of freedom (df). In
all cases the difference was evaluated for consecutive numbers of stages; i-j = 1, and ¦£ for 1 df at a = 0.05 is 3.84.
11 "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.
NR = not relevant.
362
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Output for Oral tract tumors: all tumors considered incidental to cause of death
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: C:\msw\benzo[a]pyrene-Thyssen_inc2st.(d)
The form of the probability function is:
P[response] = l-EXP{-(t - t_0)Ac *
(beta_0+beta_l*doseA1+beta_2 *doseA2)}
The parameter betas are restricted to be positive
Dependent variable = Class
Independent variables = Cone, Time
Total number of observations = 96
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 = 32
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 3.6
t_0 = 0 Specified
beta_0 = 1.18657e-031
beta_l = 1. 4 9e-03 0
beta 2 = 6.10362e-008
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_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_2
1 -1
-1 1
c
beta 2
Variable
c
beta_0
beta_l
beta 2
Estimate
4 .21938
0
0
4 . 00402e-009
Parameter Estimates
Std. Err.
0.840997
NA
NA
1.495e-008
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
2.57105 5.8677
-2.52 974e-008
3.33054e-008
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood)
# Param
AIC
Fitted
Model
-19.967
4
47 . 9339
Data
Summary
Class
C
F
I
U
Total
Expected Resp<
Cone
0
23
0
0
0
23
0 . 00
0.25
24
0
0
0
24
1. 92
1
8
0
18
0
26
16 . 04
4.3
5
0
18
0
23
18 .22
Benchmark Dose Computation
Risk Response = Incidental
Risk Type = Extra
363
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Specified effect =
Confidence level =
0.1
0.9
Time
104
BMD
BMDL
BMDU
0 .284958
0 .197807
0 .350247
Incidental Risk: BaP-Thyssen_inc2st
Dose = 0.00
Dose = 0.25
_Q
CO
_Q
O
00
O
O
O
O
.Q
CO
.Q
O
CO
O
O
O
O
0 20
60
100
0 20
60
100
Time
Time
Dose = 1.00
Dose = 4.29
_Q
CO
_Q
O
00
O
o
o
o
CO
O
00
O
o
o
o
Time
Time
364
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Output for Oral tract tumors: all tumors considered to be cause of death
Multistage Weibull Model. (Version: 1.6.1; Date: 11/24/2009)
Solutions are obtained using donlp2-intv, (c) by P. Spellucci
Input Data File: C:\msw\benzo[a]pyrene-Thyssen_allfatal_noU_3st.(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 = Cone, Time
Total number of observations = 96
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 = 32
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
c = 4.5
t_0 = 0 Specified
beta_0 = 0
beta_l = 1. 37501e-010
beta_2 = 2.84027e-010
beta 3 = 1.44668e-037
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -t_0 -beta_0 -beta_l -beta_2
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
c beta_3
1 -1
-1 1
c
beta 3
Variable
c
beta_0
beta_l
beta_2
beta 3
Estimate
8 . 95016
0
0
0
3 .43452e-019
Parameter Estimates
Std. Err.
0.896607
NA
NA
NA
1. 3 972 7e-018
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
7.19284 10.7075
-2.39515e-018
3 . 08205e-018
NA - Indicates that this parameter has hit a
bound implied by some inequality constraint
and thus has no standard error.
Log(likelihood) # Param AIC
Fitted Model -144.522 5 299.043
Data Summary
Class
C
F
I
U
Total
Cone
0
23
0
0
0
23
0.25
24
0
0
0
24
1
8
18
0
0
26
4.3
5
18
0
0
23
Minimum observation time for F tumor context =
40
365
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Benchmark Dose Computation
Risk Response = Fatal
Risk Type = Extra
Specified effect = 0.1
Confidence level = 0.9
Time = 104
BMD = 0.647659
BMDL = 0.461415
BMDU = 0.719325
Fatal Risk: BaP-Thyssen_allfatal_noll_3st
Dose = 0.00
Dose = 0.25
_Q
CC
_Q
O
CL
0~
O
^T
o
O
CD
0 20
_Q
03
_Q
O
qI
CO
o
o
o
o
0 20
Time
Time
Dose= 1.00
Dose= 4.29
_Q
03
_Q
O
0 20
^3
05
_Q
o
~I
0 20
Time
366
DRAFT - DO NOT CITE OR QUOTE
-------
1
APPENDIX F. BENCHMARK MODELING FOR THE DERMAL SLOPE FACTOR
Table F-l. Summary of model selection and modeling results for best-
fitting multistage models, for multiple data sets of skin tumors in mice
following dermal benzo[a]pyrene exposure
Data Set
(See Section 5.4.3.2.,
Tables 5-7 to 5-10)
Degree
of
Model
df
Goodness-
of-fit
/j-value
LLb
BMDio
Og/d)
BMDLio
Og/d)
Model selection rationale, with best
fitting model in boldface"
Figure
number
Poel (1959)
male C57L
2
3
4
2
2
2
0.027
0.053
0.068
-91.28
-90.43
-90.12
NR
1.7
0.58
NA
0.127
0.122
NA
0.078
0.077
(Inadequate fit for 1-2 stage models)
Most parsimonious fit
Improvement in fit not statistically
significant over 2-stage fit
F-l
Roe et al. (1970)
female Swiss
1
2-5
5
3
0.110
0.463
-64.56
-58.81
NR
11.5
0.299
0.689
0.233
0.394
Most parsimonious fit; significant
improvement over 1-stage fit
F-2
Schmidt et al. (1973)
female NMRI
1
2
3
4
4
4
0.008
0.609
0.999
-80.34
-72.68
-70.95
NR
15.3
3.5
0.256
0.329
0.381
0.194
0.287
0.326
Most parsimonious fit
Improvement in fit not statistically
significant over 2-stage fit
F-3
Schmidt et al. (1973)
female Swiss
1
2
3-4
4
4
3
<0.01
0.514
0.983
-87.99
-75.66
-73.66
NR
24.7
4.0
0.116
0.216
0.282
0.093
0.192
0.223
Most parsimonious fit; significant
improvement over 2-stage fit
F-4
Schmahl et al. (1977)
female NMRI
1
2
2
1
0.136
0.939
-147.20
-145.13
NR
4.14
0.140
0.233
0.117
0.149
Most parsimonious fit; significant
improvement over 1-stage fit
F-5
Habs et al. (1980)
female NMRI
2
3
3
3
0.009
0.207
-41.18
-37.34
NR
7.7
NA
0.294
NA
0.215
(Inadequate fit for 1-2 stage models)
Most parsimonious fit
F-6
Habs et al. (1984)
female NMRI
1
2
2
1
0.577
1.000
-22.78
-22.22
NR
1.1
0.078
0.171
0.056
0.060
Most parsimonious fit
F-7
Grimmer et al. (1983)
female CFLP
1
2-3
3
2
0.850
0.972
-108.94
-108.56
NR
0.76
0.245
0.292
0.208
0.213
Most parsimonious fit
Improvement in fit not statistically
significant over 1-stage fit
F-8
Grimmer et al.
(1984),
female CFLP
Multistage
1-3
3
0.003
205.3b
NR
NA
NA
Inadequate fit
F-9
Other: LogLogistic
Dich.-Hill
LogProbit
Gamma, Weibull
Logistic
Probit
2
1
3
2
2
0.919
1.000
0.047
<0.01
<0.01
195.8
197.7
200.2
250.5
255.4
NR
NR
NR
NR
NR
1.07
0.902
NA
NA
NA
0.479
0.533
NA
NA
NA
Best fit; slope parameter unrestricted
Slope parameter unrestricted
Inadequate fit
Same model as multistage (above)
Inadequate fit
Inadequate fit
F-10
Multistage,
high dose
dropped11
1-2
2
0.499
NR
NR
0.106
0.088
Best fit from multistage model
F-l 1
Sivak et al. (1997)
male CeH/HeJ
1
2-3
3
3
0.059
0.998
-27.92
-23.30
9.2
0.036
0.109
0.026
0.058
Most parsimonious fit
F-12
a Adequate fit: goodness-of-fit p>0.05, scaled residuals <2.0, good fit near BMR, lack of extreme curvature not reflected in the observed
data.
b LL=Log-likelihood; values for Grimmer et al. (1984) are AICs, in order to compare across models.
°X2= 2 x |(LLj - LLj)|, where i and j are consecutive numbers of stages. The test was evaluated for 1 degree of freedom (df). ¦£ for 1 df at
a = 0.05 is 3.84.
11 The preferred multistage model did not adequately fit the data for Grimmer et al. (1984), thus, the remaining suite of models were fit to
the data. The POD for Grimmer et al. (1984) was based on the LogLogistic model. For comparison purposes, the multistage model was it
fit to these data for Grimmer et al. (1984) with the highest dose dropped.
367
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-l. Fit of multistage model to skin tumors in C57L mice exposed
dermally to benzo[a]pyrene (Poel, 1959); graph and model output.
0.8
0.6
0.4
0.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
dose
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File: C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Poel_1959_MultiCanc3_0.1.(d)
Gnuplot Plotting File:
C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Poel_1959_MultiCanc3_0.1.pit
[add notes here]
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2-beta3*doseA3)]
The parameter betas are restricted to be positive
Dependent variable = NumAff
Independent variable = LADD
Total number of observations = 5
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0449589
Beta(1) = 0.490451
Beta(2) = 0
Beta(3) = 2.68146
Asymptotic Correlation Matrix of Parameter Estimates
368 DRAFT - DO NOT CITE OR QUOTE
Multistage Cancer
Linear extrapolation
Multistage Cancer Model with 0.95 Confidence Level
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
( *** The model parameter(s) -Beta(2)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background Beta(l) Beta(3)
Background 1 -0.87 0.74
Beta(1) -0.87 1 -0.92
Beta(3) 0.74 -0.92 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0.0176699 * * *
Beta(1) 0.79766 * * *
Beta(2) 0 * * *
Beta(3) 2.17146 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood) # Param'
-87.1835 5
-90.4265 3
-141.614 1
Deviance Test d.f.
6 .48606
108.86
P-value
0 . 03905
<.0001
AIC:
186.853
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
0 . 0177
0.583
0 . 000
33
-0.770
0.0500
0 . 0563
3 . 098
5 . 000
55
1.112
0.1600
0.1430
7.866
11. 000
55
1.207
0.2400
0 .2128
11.917
7 . 000
56
-1.605
0.8000
0.8293
40.635
41. 000
49
0.139
Chia2 = 5.88 d.f. = 2 P-value = 0.0528
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.126567
BMDL = 0.0777875
BMDU = 0.272961
Taken together, (0.0777875, 0.272961) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 1.28555
369
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-2. Fit of multistage model to skin tumors in female Swiss mice
exposed dermally to benzo[a]pyrene (Roe et al., 1970); graph and model
output.
Multistage Cancer Model with 0.95 Confidence Level
0.8
0.6
0.2
0
Multistage Cancer
Linear extrapolation
BMDL
BMD
0 0.5 1 1.5 2 2.5
dose
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File: C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Roe_1970_Setting.(d)
Gnuplot Plotting File: C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Roe_1970_Setting.pit
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2-beta3*doseA3-beta4*doseA4-beta5*doseA5)]
The parameter betas are restricted to be positive
Dependent variable = tumors
Independent variable = LADD
Total number of observations = 6
Total number of records with missing values = 0
Total number of parameters in model = 6
Total number of specified parameters = 0
Degree of polynomial = 5
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
370
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
Beta(1) = 0.0962491
Beta(2) = 0.141689
Beta(3) = 0
Beta(4) = 0
Beta(5) = 0
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Beta(3) -Beta(4) -Beta(5)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Background Beta(l) Beta(2)
Background 1 -0.57 0.45
Beta(1) -0.57 1 -0.94
Beta(2) 0.45 -0.94 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0.00584893 * * *
Beta(1) 0.0379152 * * *
Beta(2) 0.166839 * * *
Beta(3) 0 * * *
Beta(4) 0 * * *
Beta(5) 0 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-56 .1835
-57.5694
-118.948
Param's
6
3
1
Deviance Test d.f.
P-value
2.77176
125.529
0.4282
<.0001
AIC :
121.135
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
0.0058
0 .275
0 . 000
47
-0.526
0.0300
0.0071
0.321
1. 000
45
1.204
0.0900
0.0106
0 .444
0 . 000
42
-0.670
0.3100
0 . 0331
1.423
1. 000
43
-0.361
0.9200
0.1664
6 .821
8 . 000
41
0 .494
2.7600
0.7488
34 .444
34 . 000
46
-0.151
Chi^2 = 2.57 d.f. = 3 P-value = 0.4626
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.689131
BMDL = 0.393806
BMDU = 0.952365
Taken together, (0.393806, 0.952365) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 0.253932
371
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-3. Fit of multistage model to skin tumors in female NMRI mice
exposed dermally to benzo[a]pyrene (Schmidt et al., 1973); graph and model
output.
Multistage Cancer Model with 0.95 Confidence Level
0.4
Multistage Cancer
Linear extrapolation
BMDL
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmidtl973femaleNMRI\2MulSchMS_.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmidtl973femaleNMRI\2MulSchMS_.plt
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2)]
The parameter betas are restricted to be positive
Dependent variable = incidence
Independent variable = dose
Total number of observations = 5
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.4 9012e-008
372
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
**** We are sorry but Relative Function and Parameter Convergence ****
**** are currently unavailable in this model. Please keep checking ****
**** the web sight for model updates which will eventually ****
**** incorporate these convergence criterion. Default values used. ****
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 1.11271
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta (2)
Beta(2) 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0 * * *
Beta(1) 0 * * *
Beta(2) 0.970648 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-70.8903
-72.6831
-118.917
Param's
5
1
1
Deviance Test d.f.
P-value
3.58562
96.054
0.465
<.0001
AIC :
147.366
Goodness of Fit
Scaled
Dose
Est. Prob.
Expected
Observed
Size
Residue
0.0000
0.0000
0 . 000
0 . 000
100
0 . 000
0.0100
0.0001
0 . 010
0 . 000
100
-0 . 099
0.0600
0.0035
0.349
0 . 000
100
-0.592
0.2300
0.0501
5 . 005
2 . 000
100
-1.378
0.5700
0.2705
27 . 048
30 . 000
100
0.665
Chi^2 = 2.70 d.f. = 4 P-value = 0.6091
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.329464
BMDL = 0.286624
BMDU = 0.384046
Taken together, (0.286624, 0.384046) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 0.348889
373
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-4. Fit of multistage model to skin tumors in female Swiss mice
exposed dermally to benzo[a]pyrene (Schmidt et al., 1973); graph and model
output.
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmidtl973swissmice\3MulSchMS_.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmidtl973swissmice\3MulSchMS_.plt
BMDS Model Run
Multistage Cancer Model with 0.95 Confidence Level
0 0.1 0.2 0.3 0.4 0.5
dose
Multistage Cancer
Linear extrapolation
BMDL
BMD
iii
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2-beta3*doseA3)]
The parameter betas are restricted to be positive
Dependent variable = incidence
Independent variable = dose
Total number of observations = 5
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.4 9012e-008
**** We are sorry but Relative Function and Parameter Convergence ****
**** are currently unavailable in this model. Please keep checking ****
**** the web sight for model updates which will eventually ****
3 74 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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
77
78
79
80
81
82
83
**** incorporate these convergence criterion. Default values used. ****
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0.338951
Beta(3) = 3.8728
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(2) Beta(3)
Beta(2) 1 -0.99
Beta(3) -0.99 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0 * * *
Beta(1) 0 * * *
Beta(2) 0.108125 * * *
Beta(3) 4.31441 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-73 .5285
-73 .6628
-150.708
Param's
5
2
1
Deviance Test d.f.
P-value
0 .268637
154 .359
0 . 9658
< . 0001
AIC:
151.326
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000
0.0000
0 . 000
0 . 000
80
0 . 000
0.0100
0.0000
0 . 001
0 . 000
80
-0 . 035
0.0600
0.0013
0.106
0 . 000
80
-0.325
0.2300
0 . 0566
4 .524
5 . 000
80
0.230
0.5700
0.5657
45.260
45 . 000
80
-0 . 059
Chia2 = 0.16 d.f. = 3 P-value = 0.9833
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.282007
BMDL = 0.223401
BMDU = 0.309888
Taken together, (0.223401, 0.309888) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 0.447626
375
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-5. Fit of multistage model to skin tumors in female NMRI mice
exposed dermally to benzo[a]pyrene (Schmahl et al., 1977); graph and model
output.
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmahll977femaleNMRI\2MulschMS_.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmahll977femaleNMRI\2MulschMS_.plt
Multistage Cancer Model with 0.95 Confidence Level
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
dose
BMDL
11111111111111
Multistage Cancer
Linear extrapolation
BMD
111111111111111111111
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2) ]
The parameter betas are restricted to be positive
Dependent variable = incidence
Independent variable = dose
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.4 9012e-008
**** We are sorry but Relative Function and Parameter Convergence ****
**** are currently unavailable in this model. Please keep checking ****
**** the web sight for model updates which will eventually ****
376
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
**** incorporate these convergence criterion. Default values used. ****
Default Initial Parameter Values
Background = 0.0115034
Beta(1) = 0.284955
Beta(2) = 0.750235
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l) Beta(2)
Background 1 -0.67 0.47
Beta(1) -0.67 1 -0.94
Beta(2)
0.47
-0 . 94
Variable
Background
Beta(1)
Beta(2)
Parameter Estimates
Estimate
0 . 0123066
0 .274413
0.764244
Std. Err.
* - Indicates that this value is not calculated.
95.0% Wald Confidence Interval
Lower Conf. Limit Upper Conf. Limit
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(likelihood)
-145.127
-145.13
-184 .158
# Param's Deviance Test d.f. P-value
4
3 0.00579898 1 0.9393
1 78.0608 3 <.0001
AIC:
296 .261
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.0123 0.997 1.000 81 0.003
0.2900 0.1446 11.137 11.000 77 -0.045
0.4900 0.2813 24.756 25.000 88 0.058
0.8600 0.5567 45.096 45.000 81 -0.022
Chia2 = 0.01 d.f. = 1 P-value = 0.9393
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.232893
BMDL = 0.148895
BMDU = 0.320396
Taken together, (0.148895, 0.320396) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 0.671616
377
DRAFT - DO NOT CITE OR QUOTE
-------
Figure F-6. Fit of multistage model to skin tumors in female NMRI mice
exposed dermally to benzo[a]pyrene (Habs et al., 1980); graph and model
output.
Multistage Cancer Model with 0.95 Confidence Level
Multistage Cancer
1 - Linear extrapolation ¦
^ BMDL. | PMD ^^
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
dose
378
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
84
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008)
Input Data File: M:\_BMDS\msc_BAP_HABS1980_MultiCanc3_0.1.(d)
Gnuplot Plotting File: M:\_BMDS\msc_BAP_HABS1980_MultiCanc3_0.1.pit
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseA1-beta2 *doseA2-beta3 *doseA3)]
The parameter betas are restricted to be positive
Dependent variable = NumAff
Independent variable = LADD
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
Degree of polynomial = 3
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 4.23649
Beta(3) = 0
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l) -Beta(2)
have been estimated at a boundary point, or have been specified
by the user,
and do not appear in the correlation matrix )
Beta(3)
Beta(3) 1
Parameter Estimates
Interval
Variable
Conf. Limit
Background
Beta(1)
Beta(2)
Beta(3)
Estimate
0
0
0
4 .1289
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-34 .8527
-37.3373
-82 .5767
Param's
4
1
1
Deviance Test d.f.
P-value
4 . 96903
95 .4478
0.1741
<.0001
AIC:
76 .6745
379
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0 . 0000
0.4900
0.7400
0.8000
Chia2 =4.56
0 . 0000
0.3848
0.8123
0.8792
d.f. = 3
0.000 0.000 35 0.000
13.082 8.000 34 -1.791
21.933 24.000 27 1.019
21.102 22.000 24 0.563
P-value = 0.2067
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0 . 95
0 .294407
0 .215151
0.320955
Taken together, (0.215151, 0.320955) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
0 .46479
380
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-7. Fit of multistage model to skin tumors in female NMRI mice
exposed dermally to benzo[a]pyrene (Habs et al., 1984); graph and model
output.
Multistage Cancer Model with 0.95 Confidence Level
1
0.8
0
1
c
o
0.6
0.4
0.2
0
Multistage Cancer
Linear extrapolation
BMDL BMD
0 0.2 0.4 0.6 0.8 1
dose
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File: C:\Usepa\BMDS21\mscDax_Setting.(d)
Gnuplot Plotting File: C:\Usepa\BMDS21\mscDax_Setting.plt
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl) ]
The parameter betas are restricted to be positive
Dependent variable = tumors
Independent variable = LADD
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
3 81 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Default Initial Parameter Values
Background = 0
Beta(1) = 1.66414
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 )
Beta(1)
Beta(1) 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0 * * *
Beta(1) 1.35264 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-22 .217
-22.7878
-41. 0539
Param's
3
1
1
Deviance Test d.f.
P-value
1.14175
37.6739
0.565
<.0001
AIC:
47.5757
Dose
Goodness of Fit
Est._Prob. Expected Observed Size
Scaled
Residual
0.0000
0.5700
1.1400
ChiA2 = 1.10
0.0000
0.5375
0.7860
d.f. = 2
0.000 0.000 20 0.000
10.749 9.000 20 -0.784
15.721 17.000 20 0.697
P-value = 0.5765
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0 . 95
0 . 0778926
0 . 0558466
0.111853
Taken together, (0.0558466, 0.111853) is a 90
interval for the BMD
two-sided confidence
Multistage Cancer Slope Factor =
1.79062
382
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-8. Fit of multistage model to skin tumors in female CFLP mice
exposed dermally to benzo[a]pyrene (Grimmer et al, 1983); graph and model
output.
Multistage Cancer Model with 0.95 Confidence Level
1
0.8
0.6
0
1
c
o
0.4
0.2
0
Multistage Cancer
Linear extrapolation
BMDL BMD
0 0.5 1 1 .5 2 2.5 3 3.5 4 4.5
dose
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Grimmerl983CFLPmice\lMulGriMS_.(d)
Gnuplot Plotting File:
C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Grimmerl983CFLPmice\lMulGriMS_.plt
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl) ]
The parameter betas are restricted to be positive
Dependent variable = incidence
Independent variable = dose
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 250
Relative Function Convergence has been set to: 2.22045e-016
Parameter Convergence has been set to: 1.4 9012e-008
383 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
**** We are sorry but Relative Function and Parameter Convergence ****
**** are currently unavailable in this model. Please keep checking ****
**** the web sight for model updates which will eventually ****
**** incorporate these convergence criterion. Default values used. ****
Default Initial Parameter Values
Background = 0
Beta(1) = 0.478645
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 )
Beta(1)
Beta(1) 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0 * * *
Beta(1) 0.430366 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-108.532
-108.943
-186.434
Param's
4
1
1
Deviance Test d.f.
0.823537
155.805
P-value
0.8436
<.0001
AIC :
219.887
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.0000 0.000 0.000 80 -0.000
1.1100 0.3798 24.687 22.000 65 -0.687
2.2000 0.6120 39.169 39.000 64 -0.043
4.4000 0.8495 54.366 56.000 64 0.571
Chi^2 =0.80 d.f. = 3 P-value = 0.8496
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.244816
BMDL = 0.208269
BMDU = 0.289606
Taken together, (0.208269, 0.289606) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 0.480148
384
DRAFT - DO NOT CITE OR QUOTE
-------
Figure F-9. Fit of multistage model to skin tumors in female CFLP mice
exposed dermally to benzo[a]pyrene (Grimmer et al., 1984); graph and
model output.
Multistage Cancer Model with 0.95 Confidence Level
1
Multistage Cancer
Linear extrapolation
2
dose
385
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008)
Input Data File:
C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Grimmerl984_MultiCancl_0.1.(d)
Gnuplot Plotting File:
C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Grimmerl984_MultiCancl_0.1.pit
Wed Apr 27 17:11:28 2011
[add notes here]
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl)]
The parameter betas are restricted to be positive
Dependent variable = NumAff
Independent variable = LADD
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
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 = 0.311241
Beta(1) = 0.502556
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 )
Beta(1)
Beta(1) 1
Parameter Estimates
95.0% Wald Confidence
Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper
Conf. Limit
Background 0 * *
Beta(1) 0.796546 * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-95.8385
-101.643
-175.237
Param's
4
1
1
Deviance Test d.f.
11.61
158.797
P-value
0 . 008846
<.0001
AIC:
205.287
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size Residual
0.0000 0.0000 0.000 0.000 65 0.000
386 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
0.9700 0.5382 34.446 43.000 64
1.9100 0.7816 50.804 53.000 65
3.9000 0.9552 62.091 57.000 65
2 .145
0.659
-3.054
Chi ^2 = 14.36 d.f. = 3 P-value = 0.0025
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.132272
BMDL = 0.113427
BMDU = 0.154848
Taken together, (0.113427, 0.154848) is a 90 % two-sided confidence
interval for the BMD
Multistage Cancer Slope Factor = 0.881621
387 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-9. Fit of log-logistic model to skin tumors in female CFLP mice
exposed dermally to benzo[a]pyrene (Grimmer et al., 1984); graph and
model output.
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
1
0.8
0.6
0.4
0.2
0
BMDL
BMD
1.5
2
2.5
3
3.5
4
0
0.5
1
dose
Logistic Model. (Version: 2.12; Date: 05/16/2008)
Input Data File:
C:\Usepa\BMDS21\Data\lnl_benzo[a]pyrene_Grimmerl984_Grimmerl984_0.7Ou.(d)
Gnuplot Plotting File:
C:\Usepa\BMDS21\Data\lnl_benzo[a]pyrene_Grimmerl984_Grimmerl984_0.70u.pit
BMDS Model Run
The form of the probability function is:
P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))]
Dependent variable = NumAff
Independent variable = LADD
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0
intercept = 0.799142
slope = 0.894129
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,
388
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
intercept
slope
and do not appear in the correlation matrix )
intercept slope
1 -0.68
-0.68 1
Interval
Variable
Conf. Limit
background
intercept
slope
Parameter Estimates
Estimate
0.783559
0 . 922655
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood) # Param'
-95.8385 4
-95.9236 2
-175.237 1
Deviance Test d.f.
P-value
0.17031
158.797
0.9184
<.0001
AIC:
195.847
Dose
Est. Prob.
Goodness of Fit
Expected Observed Size
Scaled
Residual
0 . 0000
0 . 9700
1. 9100
3 . 9000
Chia2 = 0.17
0 . 0000
0.6804
0.7991
0.8849
d.f. = 2
0.000 0.000 65 0.000
43.543 43.000 64 -0.146
51.941 53.000 65 0.328
57.516 57.000 65 -0.200
P-value = 0.9190
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.7
Extra risk
0 . 95
1. 07152
0 .478669
389
DRAFT - DO NOT CITE OR QUOTE
-------
Figure F-ll. Fit of multistage model to skin tumors in female CFLP mice
exposed dermally to benzo[a]pyrene (Grimmer et al., 1984), highest dose
dropped; graph and model output.
Multistage Cancer Model with 0.95 Confidence Level
Multistage Cancer
Linear extrapolation
BMDL BMD
0 0.5 1 1.5 2
dose
390
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
77
78
79
80
81
82
83
84
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008)
Input Data File:
C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Grimmerl984_drophidose_Setting.(d)
Gnuplot Plotting File:
C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Grimmerl984_drophidose_Setting.pit
BMDS Model Run
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2)]
The parameter betas are restricted to be positive
Dependent variable = tumors
Independent variable = LADD
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0806622
Beta(1) = 0.88595
Beta(2) = 0
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(2)
have been estimated at a boundary point, or have been specified
by the user,
and do not appear in the correlation matrix )
Beta(1)
Beta(1) 1
Parameter Estimates
Interval
Variable
Conf. Limit
Background
Beta(1)
Beta(2)
Estimate
0 . 997118
0
Std. Err.
95.0% Wald Confidence
Lower Conf. Limit Upper
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)
-71.5928
-72.2756
-134 .46
# Param'
3
1
1
Deviance Test d.f.
P-value
1.36568
125.735
0.5052
<.0001
AIC:
146 .551
Goodness of Fit
Scaled
391
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Dose
Est. Prob.
Expected
Observed
Size
Residual
0 . 0000
0 . 9700
1. 9100
Chia2 = 1.3S
0 . 0000
0.6199
0.8511
d.f. = 2
0.000 0.000 65
39.671 43.000 64
55.322 53.000 65
P-value = 0.4992
0 . 000
0.857
-0.809
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0 . 95
0.105665
0 . 0881529
0.149328
Taken together, (0.0881529, 0.149328) is a 90
interval for the BMD
% two-sided confidence
Multistage Cancer Slope Factor =
1.13439
392
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Figure F-12. Fit of multistage model to skin tumors in female CFLP mice
exposed dermally to benzo[a]pyrene (Sivak et al., 1997); graph and model
output.
Multistage Cancer Model with 0.95 Confidence Level
0.8
0.6
0.4
0.2
0.1 0.2 0.3 0.4 0.5
dose
Multistage Cancer Model. (Version: 1.7; Date: 05/16/2 008)
Input Data File: C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Sivakl993_MultiCanc2_0.1.(d)
Gnuplot Plotting File:
C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Sivakl993_MultiCanc2_0.1.pit
[add notes here]
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*doseAl-beta2*doseA2) ]
The parameter betas are restricted to be positive
Dependent variable = NumAff
Independent variable = LADD
Total number of observations = 4
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0
Beta(1) = 0.0936505
Beta(2) = 8.67239
393 DRAFT - DO NOT CITE OR QUOTE
Multistage Cancer
Linear extrapolation
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(2)
Beta(2) 1
Parameter Estimates
95.0% Wald Confidence Interval
Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
Background 0 * * *
Beta(1) 0 * * *
Beta(2) 8.9375 * * *
* - Indicates that this value is not calculated.
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood) # Param'
-23.2693 4
-23.3009 1
-69.5898 1
Deviance Test d.f.
0 . 0631003
92 .641
P-value
0 . 995S
<.0001
AIC:
48.6018
Dose
Est. Prob.
Goodness of Fit
Expected Observed Size
Scaled
Residual
0.0000
0.0100
0.1400
0.5100
Chia2 =0.04
0.0000
0.0009
0.1607
0 . 9022
d.f. = 3
0.000 0.000 30 0.000
0.027 0.000 30 -0.164
4.821 5.000 30 0.089
27.065 27.000 30 -0.040
P-value = 0.9982
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
BMDU =
0.1
Extra risk
0 . 95
0.108575
0.058484
0.129641
Taken together, (0.058484, 0.129641) is a 90
interval for the BMD
two-sided confidence
Multistage Cancer Slope Factor =
1.70987
394
DRAFT - DO NOT CITE OR QUOTE
-------
1 APPENDIX G. Additional Information in Support of the Dermal Slope factor
2
3 Mouse Dermal Carcinogenesis Exposure Methods
4 Studies which macroscopically examined systemic organs for tumors include Roe et al.,
5 1970; Schmidt et al., 1973; Schmahl et al., 1977; Habs et al., 1980, 1984; Grimmer et al., 1983,
6 1984. The studies by Roe et al. 1970 and Habs et al. 1984 observed systemic tumors which the
7 authors did not consider to be treatment related. The other studies which conducted post mortem
8 macroscopic examinations of abnormal tissues, did not report any treatment related systemic
9 effects.
395
DRAFT - DO NOT CITE OR QUOTE
-------
Table G-l: Exposure methods for selected lifetime dermal exposure mouse cancer bioassays for benzo[a]pyrene-
induced skin tumors
Mouse
strain
sex
n
Applied dose of
benzo[a]pyrene,
times per week,
vehicle
Application method, volume
Application region
Duration
(weeks)
Comments
Reference
CeH/HeJ
male
30
0, 0.05, 0.5, or 5 (ig
2x/wk
cyclohexane/acetone (1:1)
not specified
0.050 ml
shaved dorsal skin
104
Sivak et al.,
1997
C57L
male
13-55
0,0.15,0.38,0.75,3.8, 19,
94, 188,376, or 752 (ig
3x/wk
toluene
calibrated needle dropper
0.0075 ml
(solvent let dry between drops to
limit spread)
shaved interscapular
skin
103
mice were 18-20 weeks of
age at the start of study
"principal organs"
examined
Poel, 1959
SWR
male
14-25
0,0.15,0.38,0.75,3.8,
19.0, 94.0, or 470 (ig
3x/wk
toluene
Calibrated needle pipette
0.0075 ml
Shaved interscapular
skin
lifetime1
exposed until tumor
development or death
Poel, 1960
C3HeB
male
14-25
0,0.15,0.38,0.75,3.8,
19.0, 94.0, or 470 (ig
benzo[a]pyrene
3x/wk
toluene
Calibrated needle pipette
0.0075ml
Shaved interscapular
skin
104
lifetime1
exposed until tumor
development or death
Poel, 1960
A/He
male
14-25
0,0.15,0.38,0.75,3.8,
19.0, 94.0, or 470 (ig
benzo[a]pyrene
3x/wk
toluene
Calibrated needle pipette
0.0075 ml
Shaved interscapular
skin
104
lifetime1
exposed until tumor
development or death
Poel, 1960
Swiss SPF
female
50
0, 0.1, 0.3, 1, 3, or 9 (ig
3x/wk
acetone
Calibrated pipette
0.25 ml
Entire shaved dorsal
area
93
systemic post-mortem
exam
Roe et al.,
1970
396 DRAFT - DO NOT CITE OR QUOTE
-------
NMRI
female
100
0, 0.05, 0.2, 0.8, or 2 (ig
2x/wk
acetone
Drop burette (dose delivered in
single drop; volume not given)
Shaved back
104
lifetime1
macroscopic examination
of internal organs
Schmidt et
al., 1973
Swiss SPF
female
80
0, 0.05, 0.2, 0.8, or 2 (ig
2x/wk
acetone
Drop burette (dose delivered in
single drop; volume not given)
Shaved back
104
lifetime1
macroscopic examination
of internal organs
Schmidt et
al., 1973
NMRI
female
100
0, 1, 1.7, or 3 (ig
2x/wk
acetone
Drop application with syringe
0.02 ml
Shaved back
104
lifetime1
macroscopic examination
of internal organs
Schmahl et
al., 1977
NMRI
female
65
0, 1.7, 2.8, and 4.6 ug
2x/wk
Drop application by calibrated
syringe,
0.02 ml
dorsal skin in
interscapular area
104
lifetime1
macroscopic examination
of internal organs
Habs et al.,
1980
CFLP
female
65
0, 3.9, 7.7 and 15.4 (ig
acetone/DMSO (1:3)
2x/wk
Drop
0.1 ml
Dorsal skin,
interscapular area
104
macroscopic examination
of internal organs
Grimmer et
al., 1983
CFLP
female
65
0, 3.4, 6.7, and 13.5 (ig
acetone:DMSO (1:3)
2x/wk
Drop
0.1 ml
Dorsal skin,
interscapular area
104
macroscopic examination
of internal organs
Grimmer et
al., 1984
NMRI
female
20
0, 2 or 4 (ig
acetone
2x/wk
drop application by calibrated
syringe
0.01 ml
dorsal skin in
interscapular area
lifetime1
macroscopic examination
of internal organs
Habs et al.,
1984
Treated until natural death or sacrifice following tumor formation
397
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
APPENDIX H. Alternative Approaches for Cross-Species Scaling of the Dermal Slope
Factor
Several publications which develop a dermal slope factor for benzo[a]pyrene are
available in the peer reviewed literature (Knafla et al., 2010; 2006; Hussain et al., 1998; LaGoy
and Quirk 1994; Sullivan et al., 1991). With the exception of the 2010 Knafla et al. publication,
none of these approaches applied quantitative adjustments to account for interspecies differences,
though the proposed slope factors were developed to account for human risk. Knafla et al.
(2010) qualitatively discuss processes which could affect the extrapolation between mice and
humans including skin metabolic activity adduct formation, stratum corneum thickness,
epidermal thickness, etc. Ultimately, the authors apply an adjustment based on the increased
epidermal thickness of human skin on the arms and hands compared to mouse interscapular
epidermal thickness. They hypothesize that the carcinogenic potential of benzo[a]pyrene may be
related to the thickness of the epidermal layer.
Because there is no established methodology for cross-species extrapolation of dermal
toxicity, several alternative approaches were evaluated. Each approach begins with the POD of
0.066 |ig/day that was based on a 10% extra risk for skin tumors in male mice (see Section
5.4.3). Based on the assumptions of each approach, a dermal slope factor for humans is
calculated. The discussion of these approaches uses the following abbreviations:
DSF = dermal slope factor
PODm = point of departure (for 10% extra risk) from mouse bioassay, in (ig/day
BWm= mouse body weight = 0.035 kg (assumed)
BWh = human body weight = 70 kg (assumed)
SAh = total human surface area = 19,000 cm2 (assumed)
SAm = total mouse surface area = 100 cm2 (assumed)
Approach 1. No interspecies adjustment to daily applied dose (POD) in mouse model
Under this approach, a given mass of benzo[a]pyrene, applied daily, would pose the same
risk in an animal or in humans, regardless of whether it is applied to a small surface area or to a
larger surface area at a proportionately lower concentration.
DSF = 0.1/PODm
DSF= 0.1/0.066 |ig/day = 2 (jig/day)"1
Assumptions: The same mass of benzo[a]pyrene, applied daily, would have same potency in
mice as in human skin regardless of treatment area.
Approach 2. Cross-species adjustment based on whole body surface-area scaling
Under this approach, animals and humans are assumed to have equal lifetime cancer risk
with equal average whole body exposures in loading units (|ig/cm2-day). As long as doses are
low enough that risk is proportional to the mass of applied compound, the daily dermal dose of
benzo[a]pyrene can be normalized over the total surface area.
POD (|ig/cm2-day) = PODM/sa (|ig/cm2-day) = PODM (|ig/day) / SAM (cm2)
POD = (0.066 |ig/day) / 100 cm2
398 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
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
= 0.00066 |ig/cm2-day
DSF = 0.1/(0.00066 |ig/cm2-day) ~ 152 (jig/cn^-day)"1
Assumptions: Mouse and human slope factors are equipotent if total dermal dose is averaged
over equal fractions of the entire surface area. Tumor potency of benzo[a]pyrene is assumed to
be related to overall dose and not dose per unit area. For example, someone exposed to 0.01
|ig/day on 10 cm2 would be assumed to have the same potential to form a skin tumor as someone
treated with 0.01 |ig/day over 10,000 cm2.
Approach 3. Cross-species adjustment based on body weight
Under this approach, a given mass of benzo[a]pyrene is normalized relative to the body
weight of the animal or human. This approach has been used for oral doses for noncancer
effects.
PODm/ BWm= 0.066 |ig/0.035 kg-day =1.9 |ig/kg-day
DSF = 0.1/1.9 |ig/kg-day = 0.05 (jig/kg-day)"1
Assumptions: The potency of point of contact skin tumors is related to bodyweight and humans
and mice would have an equal likelihood of developing skin tumors based on a dermal dose per
kg basis.
Issues: Skin cancer following benzo[a]pyrene exposure is a local effect and not likely dependent
on body weight.
Approach 4. Cross-species adjustment based on allometric scaling using body weight to
the 3/4 power
Under this approach, rodents and humans exposed to the same daily dose of a carcinogen,
adjusted for BW3 4, would be expected to have equal lifetime risks of cancer. That is, a lifetime
dose expressed as |ig/kg3/4-day would lead to an equal risk in rodents and humans. This scaling
reflects the empirically observed phenomena of more rapid distribution, metabolism, and
clearance in smaller animals. The metabolism of benzo[a]pyrene to reactive intermediates is a
critical step in the carcinogenicity of benzo[a]pyrene, and this metabolism occurs in the skin.
POD (|ig/day) = PODM (ng/day) x (BWH/ BWM)3/4
POD (|ig/day) = 0.066 |ig/day x (70 kg / 0.035 kg)3/4
= 19.7 |ig/day
DSF = 0.1/(19.7 |ig/day) ~ 0.005 (jig/day)1
Assumptions: Risk at low doses of benzo[a]pyrene is dependent on absolute dermal dose and not
dose per unit of skin, meaning a higher exposure concentration of benzo[a]pyrene contacting a
smaller area of exposed skin could carry the same risk of skin tumors as a lower exposure
concentration of benzo[a]pyrene that contacts a larger area of skin.
399
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Issues: It is unclear if scaling of doses based on bodyweight ratios will correspond to differences
in metabolic processes in the skin of mice and humans.
Synthesis of the alternative approaches to cross-species scaling
A comparison of the above approaches is provided in Table H-l below. The lifetime risk
from a nominal human dermal exposure to benzo[a]pyrene over a 5% area of exposed skin
(approximately 950 cm2), estimated at 1 x 10 "4 |ig/day*, is calculated for each of the approaches
in order to judge whether the method yields risk estimates that are unrealistically high.
Other potential interspecies adjustments
The above discussion presents several mathematical approaches that result from varying
assumptions about what is the relevant dose metric for determining equivalence across species.
Biological information (that is not presently comprehensive or detailed enough to develop robust
models) that could be used in future biologically based models for cross-species extrapolation
include:
a. Quantitative information on interspecies differences in partitioning from exposure
medium to the skin and absorption through the skin
b. Thickness of the stratum corneum between anatomical sites and between species
c. Thickness of epidermal layer
d. Skin permeability
e. Metabolic activity of skin
f. Formation of DNA adducts in skin
400
DRAFT - DO NOT CITE OR QUOTE
-------
1
2 Table H-l. Alternative approaches to cross-species scaling
Approach
Assumptions
Dose metric
DSF
Risk at nominal
exposure
(0.0001 Hg/day)*
1. Mass-per-
day scaling
Eaual mass ocr dav (us /d). if aoolied to eaual areas of skin (cm2). will affect similar
numbers of cells across species. Cancer risk is proportional to the area (cm2) exposed if
the loading rate (|ig /cm2-d) is the same. This approach assumes that risk is proportional
to dose expressed as mass per day. This approach implies that any combination of loading
rate (|ig /cm2-day) and skin area exposed (cm2) that have the same product when
multiplied, will result in the same risk.
Hg/day
1 per ng/day
2 x 10-4
2. Surface-
area scaling
Equal mass per dav (|ia /d). if applied to equal fractions of total skin surface (cm2) will
have similar cancer risks. That is, whole-body lifetime exposure [e.g., 5%-of-the-body
lifetime exposure] at the same loading rate (|ig /cm2-d) gives similar cancer risks across
species. This approach assumes that risk is proportional to dose expressed as mass per
area per day. This approach implies that risk does not increase with area exposed as long
as dose per area remains constant.
|ig/cm2-day
152 per
|ig/cm2-day
8 x 10"7
3. Body-
weight
scaling
The skin is an organ with thickness and volume; benzo[a]pyrene is distributed within this
volume of skin. Cancer risk is proportional to the concentration of benzo[a]pyrene in the
exposed volume of skin. Equal mass per day (|ig /d), if distributed within equal fractions
of total body skin will have similar cancer risks. That is, whole-body lifetime exposure
[e.g., 5%-of-the-body lifetime exposure] at the same loading rate (|ig /cm2-d) gives similar
cancer risks across species. This approach assumes that risk is proportional to dose
expressed as mass per kg body weight per day. This approach implies that any
combination of dose (|ig /day) and body weight (kg) that have the same result when
divided, will result in the same risk.
Hg/kg-day
0.05 per
Hg/kg-day
8 x 10-8
4. Alio metric
scaling
(BW )
Same as for bodv-weisht scalins. cxccot that bcnzolalDvrcnc distribution and metabolism
takes place within this volume of skin. Allometric scaling is generally regarded as
describing the relative rate of toxicokinetic processes across species. This approach also
is used by EPA to scale oral exposures.
Hg/day
0.005 per ng
/day
5 x 10"7
3 * Nominal exposure calculated as a geometric mean of average daily doses (ng/day) calculated from a range of benzo[a]pyrene soil concentrations (1- 1000 ppb) reported
4 from non-contaminated rural/agricultural soils (ATSDR, 1995) and a range of standard exposure assumptions.
401
DRAFT - DO NOT CITE OR QUOTE
------- |