DRAFT—DO NOT CITE OR QUOTE

&EPA
                                                           EPA/635/R-10/006C
                                                          www.epa.gov/iris
                Toxicological Review of Benzo[a]pyrene

                             (CASRN 50-32-8)
               In Support of Summary Information on the
               Integrated Risk Information System (IRIS)

                       Supplemental Information
                                June 2012
                                 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.
                  National Center for Environmental Assessment
                      Office of Research and Development
                     U.S. Environmental Protection Agency
                              Washington, DC

-------
                                            Toxicological Review ofbenzo[a]pyrene

                                  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 of
recommendation for use.
         This document is a draft for review purposes only and does not constitute Agency policy.
                                        ii          DRAFT—DO NOT CITE OR QUOTE

-------
                                         Toxicological Review ofbenzo[a]pyrene
CONTENTS
APPENDIX A.  OTHER AGENCY AND INTERNATIONAL ASSESSMENTS	A-l
APPENDIX B.  INFORMATION IN SUPPORT OF HAZARD IDENTIFICATION AND DOSE-
            REPONSE ANALYSIS	B-3
            TOXICOKINETICS	B-3
            HUMAN STUDIES	B-16
            ANIMAL BIOASSAYS	B-36
            OTHER PERTINENTTOXICITY INFORMATION	B-88
APPENDIX C.  DOSE-RESPONSE MODELING FOR THE DERIVATION OF REFERENCE
            VALUES FOR EFFECTS OTHER THAN CANCER AND THE DERIVATION OF
            CANCER RISK ESTIMATES	C-l
            DOSE-RESPONSE MODELING FOR DERVIATION OF RFD	C-l
            INHALATION DOSIMETRY MODELING FOR RFC DERIVATION	C-23
            DOSE-RESPONSE MODELING FOR CANCER RISK VALUES	C-26
            DOSE-RESPONSE MODELING FOR THE INHALATION UNIT RISK	C-58
            DOSE-RESPONSE MODELING FOR THE DERMAL SLOPE FACTOR	C-65
            ALTERNATIVE APPROACHES FOR CROSS-SPECIES SCALING OF THE
            DERMAL SLOPE FACTOR	C-97
APPENDIX D.  SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC COMMENTS AND
            EPA'S DISPOSITION	D-l
REFERENCES FOR APPENDICES	1
        This document is a draft for review purposes only and does not constitute Agency policy.
                                     iii         DRAFT—DO NOT CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene
TABLES AND FIGURES
Table A-l. Health assessments and regulatory limits by other national and international agencies A-
       1
Figure B-l. Metabolic pathways for benzo[a]pyrene	B-8
Figure B-2. The stereospecific activation of benzo[a]pyrene	B-9
Table B-l. Exposure to benzo[a]pyrene and mortality from cardiovascular diseases in a European
       cohort of asphalt paving workers	B-17
Table B-2. Exposure to benzo[a]pyrene and mortality from cardiovascular diseases in a Canadian
       cohort of male aluminum smelter workers	B-19
Table B-3. Exposure-related effects in Chinese coke oven workers or warehouse controls exposed
       to benzo[a]pyrene in the workplace	B-25
Table B-4. Exposure-related effects in Chinese coke oven workers or warehouse controls exposed
       to benzo[a]pyrene in the workplace, stratified by urinary metabolite levels	B-26
Table B-5. Background information on Chinese coke oven workers or warehouse controls exposed
       to benzo[a]pyrene in the workplace	B-27
Table B-6. Exposure-related effects in male Wistar  rats exposed to benzo[a]pyrene by gavage 5
       days/week for 5 weeks	B-36
Table B-7. Exposure-related effects in Wistar rats exposed to benzo[a]pyrene by gavage 5
       days/week for 5 weeks	B-40
Table B-8. Means ± SDa for liver and thymus weights in Wistar rats exposed to benzo[a]pyrene by
       gavage 5 days/week for 90 days	B-42
Table B-9. Incidences of exposure-related neoplasms in Wistar rats treated by gavage with
       benzo[a]pyrene, 5 days/week, for 104 weeks	B-44
Table B-10. Incidences of alimentary tract tumors in Sprague-Dawley rats chronically exposed to
       benzo[a]pyrene in the dietor by gavage in caffeine solution	B-47
Table B-ll. Incidence of nonneoplastic and neoplastic lesions in female B6C3Fi mice fed
       benzo[a]pyrene in the diet for up to 2 years	B-49
Table B-12. Other oral exposure cancer bioassays in mice	B-51
Table B-13. Incidence of respiratory and upper digestive tract tumors in male hamsters treated for
       life with benzo[a]pyrene by inhalation	B-56
Table B-l4. Number of animals with pharynx and larynx tumors in male hamsters exposed by
       inhalation to benzo[a]pyrene  for life	B-57
Table B-15. Skin tumor incidence and time of appearance in male C57L mice dermally exposed to
       benzo[a]pyrene for up to 103 weeks	B-59
Table B-16. Skin tumor incidence and time of appearance in male SWR, CSHeB, and A/He mice
       dermally exposed to benzo[a]pyrene for life or until a skin tumor was detected	B-60
Table B-17. Tumor incidence in female Swiss mice dermally exposed to benzo[a]pyrene for up to
       93 weeks	B-61
Table B-l8. Skin tumor incidence in female NMRI and Swiss mice dermally exposed to
       benzo[a]pyrene	B-62
Table B-19. Skin tumor incidence in female NMRI mice dermally exposed to benzo[a]pyrene.... B-62
Table B-20. Skin tumor incidence in female NMRI mice dermally exposed to benzo[a]pyrene.... B-63
Table B-21. Skin tumor incidence and time of appearance in female CFLP mice dermally exposed to
       benzo[a]pyrene for 104 weeks	B-64
Table B-22. Skin tumor incidence in female NMRI mice dermally exposed to benzo[a]pyrene for life
       	B-65
Table B-23. Skin tumor incidence in male C3H/HeJ  mice dermally exposed to benzo[a]pyrene for

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          iv           DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review ofbenzo[a]pyrene


       24 months	B-66
Table B-24. Mortality and cervical histopathology incidences in female ICR mice exposed to
       benzo[a]pyrene via gavage for 14 weeks	B-70
Table B-25. Means ± SD for ovary weight in female Sprague-Dawley rats	B-72
Table B-26. Reproductive effects in male and female CD-I Fl mice exposed in utero to
       benzo[a]pyrene	B-74
Table B-27. Effect of prenatal exposure to benzo[a]pyrene on indices of reproductive performance
       in Fl female NMRI mice	B-76
Table B-28. Exposure-related effects in Long Evans Hooded rats exposed to benzo[a]pyrene by
       gavage daily in utero from GD14-GD17	B-80
Table B-29. Exposure-related effects in Swiss Albino OF1 mice exposed as pups to benzo[a]pyrene
       in breast milk from dams treated by gavage daily from PND1 - PND14	B-82
Table B-30. Pregnancy outcomes in female F344 rats treated with benzo[a]pyrene on CDs 11-21
       by inhalation	B-84
Table B-31. In vitro genotoxicity studies of benzo[a]pyrene in non-mammalian cells	B-88
Table B-32. In vitro genotoxicity studies of benzo[a]pyrene in mammalian cells	B-89
Table B-33. In vivo genotoxicity studies ofbenzo[a]pyrene	B-l
Table C-l.  Means ± SDa for thymus weight in male Wistar rats exposed to benzo[a]pyrene by
       gavage 5 days/week for 90 days	C-2
Table C-2.  Model predictions for decreased thymus weight in male Wistar rats—90 days	C-2
Figure C-l.  Fit of linear model (nonconstant variance) to data on decreased thymus weight in male
       Wistar rats—90 days	C-3
Table C-3.  Means ± SDa for thymus weight in female Wistar rats exposed to benzo[a]pyrene by
       gavage 5 days/week for 90 days	C-6
Table C-4.  Model predictions for decreased thymus weight in female Wistar rats—90 days	C-6
Figure C-2.  Fit of linear model (constant variance) to data on decreased thymus weight in female
       Wistar rats—90 days	C-7
Table C-5.  Means ± SDs for ovary weight in female Sprague-Dawley rats	C-10
Table C-6.  Model predictions for decreased ovary weight in female Sprague-Dawley rats	C-10
Figure C-3.  Fit of linear/polynomial (1°) model to data on decreased ovary weight	C-ll
Table C-7.  Means ± SDs for Escape Latency and Time Spent in Target Quadrant	C-14
Table C-8.  Model predictions for increase in Morris water maze test for escape latency, male and
       female rats	C-14
Figure C-4.  Fit of Hill model to data on Morris water maze test escape latency	C-15
Table C-9.  Model predictions for decrease in Morris water maze test for time spent in target
       quadrant, male and female rats	C-18
Figure C-5.  Fit of Exponential 4 model to data on Morris water maze time spent in target quadrant.
       	C-18
Table C-10. Incidence of cervical epithelial hyperplasia	C-21
Table C-ll. Model predictions for increased incidence of epithelial hyperplasia in female ICR miceC-
       21
Figure C-6.  Human fractional deposition	C-23
Figure C-7.  Rat fractional deposition	C-24
Table C-12. Tumor incidence data, with time to death with tumor; male rats exposed by gavage to
       benzo[a]pyrene—Kroese etal. (2001)	C-29
Table C-13. Tumor incidence data, with time to death with tumor; female rats exposed by gavage to
       benzo[a]pyrene—Kroese etal. (2001)	C-31
Table C-14. Tumor incidence, with time to death with tumor; female mice exposed to
       benzo[a]pyrene via diet—Beland and Gulp (1998)	C-33
Table C-15. Derivation of HEDs to use for BMD modeling of Wistar rat tumor incidence data from

          This document is a draft for review purposes only and does not constitute Agency policy.
                                           v          DRAFT—DO NOT CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene


       Kroeseetal. (2001)	C-34
Table C-16. Derivation of HEDs for dose-response modeling of B6C3Fi female mouse tumor
       incidence data from Beland and Gulp (1998)	C-35
Table C-17. Summary of model selection and modeling results for best-fitting multistage-Weibull
       models, using time-to-tumor data for rats from Kroese et al. (1981)	C-36
Table C-18. Summary of human equivalent overall oral slope factors, based on male and female rat
       tumor incidence	C-55
Table C-19. Summary of model selection among multistage-We ibull models fit to alimentary tract
       tumor data for female mice	C-55
Table C-20. Individual pathology and tumor occurrence data for male Syrian hamsters exposed to
       benzo[a]pyrene via inhalation for lifetime—Thyssenetal. (1981)	C-58
Table C-21. Summary of model selection among multistage-Weibull models fit to tumor data for
       male hamsters	C-60
Table C-22. Skin tumor incidence, benign or malignant in female Swiss or NMRI mice dermally
       exposed to benzo[a]pyrene	C-68
Table C-23. Skin tumor incidence, benign or malignant, in C57L male mice dermally exposed to
       benzo[a]pyrene	C-69
Table C-24. Skin tumor incidence, benign or malignant, in female CFLP mice dermally exposed to
       benzo[a]pyrene	C-69
Table C-25. Skin tumor incidence, benign or malignant, in male C3H/HeJ mice dermally exposed to
       benzo[a]pyrene	C-70
Table C-26. 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....C-
       71
Figure C-8. Fit of multistage model to skin tumors in C57L mice exposed dermally to
       benzo[a]pyrene (Poel, 1959); graph and model output	C-72
Figure C-9. 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	C-75
Figure C-10. 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	C-77
Figure C-ll. 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	C-79
Figure C-12. Fit of multistage model to skin tumors in female NMRI mice exposed dermally to
       benzo[a]pyrene (Schmahl etal., 1977); graph and model output	C-81
Figure C-13. Fit of multistage model to skin tumors in female NMRI mice exposed dermally to
       benzo[a]pyrene (Habs etal., 1980); graph and model output	C-83
Figure C-14. Fit of multistage model to skin tumors in female NMRI mice exposed dermally to
       benzo[a]pyrene (Habs etal., 1984); graph and model output	C-85
Figure C-15. 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	C-87
Figure C-16. 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	C-89
Figure C-17. 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	C-91
Figure C-18. 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 C-93
Figure C-19. Fit of multistage model to skin tumors in male CeH/HeJ mice exposed dermally to
       benzo[a]pyrene (Sivaketal., 1997); graph and model  output	C-95
Table C-27. Alternative approaches to cross-species scaling	C-100
          This document is a draft for review purposes only and does not constitute Agency policy.
                                          vi          DRAFT—DO NOT CITE OR QUOTE

-------
ABBREVIATIONS
                                                   Toxicological Review ofbenzo[a]pyrene
3-MC    3-methylcholanthrene
8-OHdG  8-hydroxydeoxyguanosine
ADAF    age-dependent adjustment factor
Ah       aryl hydrocarbon
AHH     aryl hydrocarbon hydroxylase
AhR     Ah receptor
AIC      Akaike's Information Criterion
AKR     aldo-keto reductase
ALT     alanine aminotransferase
ANOVA  analysis of variance
ATSDR   Agency for Toxic Substances and
         Disease Registry
AUC     area under the curve
BMD    benchmark dose
BMDL    benchmark dose, 95% lower bound
BMDS    Benchmark Dose Software
BMR    benchmark response
BPDE    benzo[a]pyrene-7,8-diol-9,10-epoxide
BPQ     benzo[a]pyrene-7,8-quinone
BrdU    bromodeoxyuridine
BSM     benzene-soluble matter
BUN     blood urea nitrogen
CA       chromosomal aberration
CASRN   Chemical Abstracts Service Registry
         Number
CHO     Chinese hamster ovary
CI       confidence interval
CNS     central nervous system
CONSAAM     Conversational SAAM
COX     cyclooxygenase
CYP     cytochrome
CYP450  cytochrome P450
dG-N2-BPDE   10p-(deoxyguanosin-N2-yl>
         7p,8
-------
                                                   Toxicological Review ofbenzo[a]pyrene
RfD      reference dose
RN      reaction network
RNA     ribonucleic acid
ROS     reactive oxygen species
RR      relative risk
s.c.      subcutaneous
SAAM    Simulation, Analysis and Modeling
SAM     S-adenosylmethionine
SCC      squamous cell carcinoma
SCE      sister chromatid exchange
SD      standard deviation
SE       standard error
SEM     standard error of the mean
SIR      standardized incidence 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
TPA     12-0-tetradecanoylphorbol-13-acetate
TWA    time-weighted average
UCL     upper confidence limit
UDP     uridine diphosphate
UF      uncertainty factor
WBC     white blood cells
WT      wild type
WTC     World Trade  Center
XP      xeroderma pigmentosum
XPA     xeroderma pigmentosum group A
           This document is a draft for review purposes only and does not constitute Agency policy.
                                             viii          DRAFT—DO NOT CITE OR QUOTE

-------
                                 Toxicological Review ofbenzo[a]pyrene
APPENDIX A.  OTHER AGENCY AND
INTERNATIONAL ASSESSMENTS
     Table A-l. Health assessments and regulatory limits by other national
     and international agencies
Organization
Toxicity value or determination
Non-cancer: oral value
CalEPA(2010)
The concentration of 4 u.g/L (ADD = 1.7 x 10"3 mg/kg-day) for benzo[a]pyrene in
water for noncarcinogenic effects was derived from a LOAEL of 5 mg/kg-day for
renal toxicity from Knuckles et al. (2001), a UF of 3,000.
Non-cancer: inhalation value
WHO
(1996, 2003)
Health Canada
(1986, 2005)
The guideline value for benzo[a]pyrene in drinking water of 0.7 u.g/L was based on a
cancer slope factor of 0.46 (mg/kg-day)"1 derived from Neal and Rigdon (1967) and a
lifetime excess cancer risk of 10"5.
The Maximum Acceptable Concentration (MAC) for benzo[a]pyrene in drinking
water of 0.01 u.g/L was derived from Neal and Rigdon (1967) using a drinking water
consumption rate of 1.5 L/day, body weight of 70 kg, and a lifetime cancer risk of 5 x
10"7. The concentrations of 2, 0.2, and 0.02 jug//. benzo[a]pyrene correspond to
lifetime excess cancer risks oflO'4, 10~5, and 10~6.
Cancer: Oral value
CalEPA(2010)
Cancer slope factor of 2.9 (mg/kg-day)"1 derived from Gulp et al. (1998). This
includes an age sensitivity factor of 1.7.
Cancer: Inhalation value
WHO
(2000, 2010)
CalEPA(1994)
EU (2005)
Does not recommend specific guideline values for PAHs in air. A unit risk of 87
(mg/m3)"1 for benzo[a]pyrene, as an indicator a PAH mixtures, was derived from U.S.
EPA's IUR from coke oven emissions. The concentrations 0.0012, 1.2 x 10'4, and 1.2
x 10'5 HQ/m3 benzo[a]pyrene correspond to lifetime excess cancer risks oflO'4, 10~5,
andlO'6.
The inhalation unit risk of 1.1 (mg/m3)"1 was derived based on Thyssen et al. (1981).
Target value of 1 ng/m3 benzo[a]pyrene (averaged over one calendar year) as a
marker of PAH carcinogenic risk. Does not include information for how target value
was derived.
       This document is a draft for review purposes only and does not constitute Agency policy.
                             A-l      DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Cancer characterization
IARC(2010)
NTP (2011)
CalEPA(2000)
Health Canada
(1986, 1988)
Carcinogenic to humans (Group 1) (based on mechanistic data)
"reasonably anticipated to be a human carcinogen"
"Sufficient reason for concern regarding the carcinogenic potential of this toxicant in
humans."
Probably carcinogenic to man
This document is a draft for review purposes only and does not constitute Agency policy.
                               A-2         DRAFT—DO NOT CITE OR QUOTE

-------
                                                  Toxicological Review ofbenzo[a]pyrene
 i   APPENDIX B.  INFORMATION  IN SUPPORT OF

 2   HAZARD IDENTIFICATION AND DOSE-REPONSE

 3   ANALYSIS	


 4   TOXICOKINETICS

 5   Overview
 6          Benzo[a]pyrene is absorbed following exposure by inhalation, oral, and dermal routes. The
 7   rate and extent of absorption are dependent upon the exposure medium. The presence of
 8   benzo[a]pyrene in body fat, blood, liver, and kidney and the presence of benzo[a]pyrene
 9   metabolites in serum and excreta demonstrate wide systemic tissue distribution. Benzo[a]pyrene
10   metabolism occurs in essentially all tissues, with high metabolic capacity in the liver and significant
11   metabolism in tissues at the portal of entry (lung, skin, and gastrointestinal [GI] tract) and in
12   reproductive tissues. Stable metabolic products identified in body tissues and excreta are very
13   diverse and include phenols, quinones, and dihydrodiols. These classes of metabolites are typically
14   isolated as glucuronide or sulfate  ester conjungates in the excreta, but can also include glutathione
15   conjugates formed from quinones or intermediary epoxides. The primary route of metabolite
16   elimination is in the feces via biliary excretion, particularly following exposure by the inhalation
17   route. To a lesser degree, benzo[a]pyrene metabolites are eliminated via urine. Overall,
18   benzo[a]pyrene is eliminated quickly with a biological half-life of several hours.

19   Absorption
20          The absorption of benzo[a]pyrene has been studied in humans and laboratory animals for
21   inhalation, ingestion and dermal exposure. Studies of workers occupationally exposed to
22   benzo[a]pyrene have qualitatively demonstrated absorption via inhalation by correlating
23   concentrations of benzo [a]pyrene in the air and benzo [a]pyrene metabolites in the exposed
24   worker's urine. Occuational exposures to benzo[a]pyrene measured with personal air samplers
25   were correlated to urine concentrations of benzo[a]pyrene-9,10-dihydrodiol, a specific metabolite
26   of benzo[a]pyrene, in 24 hour aggregate urine samples by Grimmer etal., 1994. Theamountof
27   benzo[ajpyrene extracted- from personal air monitoring devices (a surrogate for ambient PAHs) of
28   coke oven workers were correlated with r-7,t-8,9,c 10 tetrahydroxy-7,8,9,10-
29   tetrahydrobenzo[a]pyrene (trans-anti-benzo[a]pyrene-tetrol, a specific metabolite of
30   benzo[a]pyrene) in the worker's urine by Wu etal. (2002). In both of these studies only a very
31   small fraction (< 1%) of the inhaled benzo[a]pyrene was recovered from urine, consistent with
32   studies in animals that find urine  is not a major route of elimination for benzo[a]pyrene (as
33   described in the excretion section below). These occupational studies cannot be used to quantify

               This document is a draft for review purposes only and does not constitute Agency policy.
                                             B-3         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    absorption through inhalation-only exposure in humans because the persistence of
 2    benzo[a]pyrene-contaminated particulate matter on surfaces and food may lead to exposures via
 3    additional routes (Bostrum et al., 2002). Nevertheless, the observation of benzo[a]pyrene
 4    metabolites in excreta of exposed humans provides qualitative evidence for benzo[a]pyrene
 5    absorption, at least some of which is likely to occur via inhalation.
 6           Results from studies of animals following intratracheal instillation of benzo[a]pyrene
 7    provide supporting, quantitative evidence that absorption by the respiratory tract is rapid (Bevan
 8    and Ulman, 1991; Gerde et al. 1993 b; Weyand and Bevan, 1986; 1987). Following intratracheal
 9    instillation of 1 |ig 3H-labeled benzo[a]pyrene/kg dissolved in triethylene glycol to Sprague-Dawley
10    rats, radioactivity rapidly appeared in the liver (reaching a maximum of about 21% of the
11    administered dose within 10 minutes). Elimination of radioactivity from the lung was biphasic,
12    with elimination half-times of 5 and 116 minutes (Weyand and Bevan, 1986). In bile-cannulated
13    rats, bile collected for 6 hours after instillation accounted for 74% of the administered radioactivity
14    (Weyand and Bevan, 1986). The results are consistent with rapid and extensive absorption by the
15    respiratory tract and rapid entry into hepatobiliary circulation following intratracheal instillation.
16    The respiratory tract absorption may also be affected by the vehicle, since higher amounts of
17    benzo[a]pyrene were excreted in bile when administered with hydrophilic triethylene glycol than
18    with lipophilic solvents ethyl laurate or tricaprylin (Bevan and Ulman, 1991). Particle-bound
19    benzo [a]pyrene deposited in the respiratory tract is absorbed and cleared more slowly than the
20    neat compound (Gerde etal., 2001).
21           Studies conducted to assess levels of benzo[a]pyrene metabolites or benzo[a]pyrene-DNA
22    adduct levels in humans exposed to benzo[a]pyreneby the oral route are not adequate to develop
23    quantitative estimates of oral bioavailability. The concentration of benzo[a]pyrene was below
24    detection limits (<0.1 |ig/person) in the feces of eight volunteers who had ingested broiled meat
25    containing approximately 8.6 |ig of benzo[a]pyrene (Hecht et al., 1979). However, studies in
26    laboratory animals demonstrate benzo[a]pyrene is absorbed via ingestion. Studies of rats and pigs
27    measured the oral bioavailability of benzo[a]pyrene in the range from 10 to 40% (Ramesh et al.,
28    2001b; Fothetal., 1988; Cavretetal., 2003; Hecht etal., 1979). The absorption of benzo [a] pyrene
29    may depend on the vehicle.  Intestinal absorption of benzo[a]pyrene was enhanced in rats when the
30    compound was solubilized in lipophilic compounds such as triolein, soybean oil, and high-fat diets,
31    as compared with fiber- or protein-rich diets (O'Neill et al., 1991; Kawamura et al., 1988). Aqueous
32    vehicles, quercetin, chlorogenic acid, or carbon particles reduced biliary excretion of
33    benzo[a]pyrene, while lipid media such as corn oil increased it (Stavric and Klassen, 1994). The
34    addition of wheat bran to the benzo[a]pyrene containing diets increased fecal excretion of
35    benzo[a]pyrene (Mirvish etal., 1981).
36           Studies of benzo[a]pyrene metabolites or DNA adducts measured in humans exposed
37    dermally to benzo[a]pyrene-containing mixtures demonstrate that benzo[ajpyrene is absorbed
38    dermally. One study of dermal absorption in human volunteers found absorption rate constants

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 B-4         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    ranging from 0.036 to 0.135/hour over a 45 minute exposure, suggesting 20-56% of the dose
 2    would be absorbed within 6 hours (Van Rooij et al., 1993). Dermal absorption rates varied 69%
 3    between different anatomical sites (forehead, shoulder, volar forearm, palmar side of the hand,
 4    groin, and ankle) and only 7% between different individual volunteers (Van Rooij et al., 1993). The
 5    overall absorbed amount of benzo[a]pyrene in explanted viable skin samples from tissue donors
 6    (maintained in short-term organ cultures) exposed for 24 hours ranged from 0.09 to 2.6% of the
 7    dose (Kao etal., 1985; Wester etal., 1990). Similar amounts of penetration were measured in skin
 8    samples from other species including marmosets, rats, and rabbits (Kao et al., 1985). Skin from
 9    mice allowed more of the dose to penetrate (more than 10%), while that of guinea pig let only a
10    negligible percentage of the dose penetrate (Kao etal., 1985). The vehicle for benzo[a]pyrene
11    exposure is an important factor in skin penetration. Exposure of female Sprague-Dawley rats and
12    female rhesus monkeys topically to benzo[a]pyrene in crude oil or acetone caused approximately 4-
13    fold more extensive absorption than benzo[a]pyrene in soil (Wester et al., 1990; Yang etal., 1989).
14    The viscosity of oil product used as a vehicle  also changed skin penetration with increased uptake
15    of benzo[a]pyrene for oils with decreased viscosity (Potter et al., 1999). Metabolism is also an
16    important determinant of permeation, with very low rates observed in nonviable skin (Kao et al.,
17    1985).

18    Distribution
19          No  adequate quantitative studies of benzo[a]pyrene tissue distribution in exposed humans
20    were identified. Obana et al.  (1981) observed low levels of benzo[a]pyrene in liver and fat tissues
21    from autopsy samples. However, prior exposure histories were not available for the donors.
22    Nevertheless, the identification of benzo[a]pyrene metabolites or DNA adducts in tissues and
23    excreta of PAH-exposed populations suggest that benzo[a]pyrene is widely distributed.
24          Distribution of benzo[a]pyrene has been studied in laboratory animals for multiple routes
25    of exposure, including inhalation, ingestion, dermal and intravenous. Exposure to benzo[a]pyrene
26    in various species (Sprague-Dawley rats, Gunn rats, guinea pigs, and hamsters) results in wide
27    distribution throughout the body and rapid uptake into well-perfused tissues (i.e. lung, kidney, and
28    liver) (Weyand and Bevan, 1987; Weyand and Bevan, 1986). Route of administration of
29    benzo[a]pyrene has little influence on the tissue distribution with similar results from studies of
30    inhalation (or intratracheal instillation), oral, i.v. and dermal exposures (Weyand and Bevan, 1987;
31    Weyand and Bevan, 1986; Morse and Carlson, 1985; Saunders et al., 2002; Neubert and Tapken
32    1988; Moir et al., 1998). Intratracheal instillation of radiolabeled benzo[a]pyrene in mice resulted
33    in increased radioactivity in lung-associated lymph nodes, suggesting distribution of
34    benzo[a]pyrene or its metabolites via the lymph (Schnizlein et al. 1987). Rats with biliary cannulas
35    had high excretion of benzo[a]pyrene and benzo[a]pyrene metabolites in bile. The benzo[a]pyrene
36    thioether and glucuronic acid-conjugated metabolites in intestines indicated enterohepatic
37    recirculation of benzo[a]pyrene and benzo[a]pyrene metabolites (Weyand and Bevan,  1986). The
38    vehicle for  delivery of inhalated benzo[a]pyrene impacts the distribution with aerosolized
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-5          DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    benzo[a]pyrene more readily absorbed directly in the respiratory tract than particle-adsorbed
 2    benzo[a]pyrene (which is cleared by the mucociliary and then ingested) (Sun et al., 1982).
 3    Exposure of pregnant rats and mice to benzo[a]pyrene via inhalation and ingestion showed a wide
 4    tissue distribution of benzo[a]pyrene, consistent with other studies and demonstrated placental
 5    transfer of benzo[a]pyrene and its metabolites (Withey et al., 1993; Neubert and Tapken 1988;
 6    Shendrikova and Aleksandrov, 1974). The reactive metabolites of benzo[a]pyrene are also
 7    transported in the blood and may be distributed to tissues incapable of benzo[a]pyrene
 8    metabolism, such. Serum of benzo[a]pyrene-treated mice incubated with splenocytes or salmon
 9    sperm DNA resulted in adduct formation, suggesting that reactive benzo[a]pyrene metabolites
10    were systemically distributed and available for interaction with target tissues (Ginsberg and
11    Atherholt, 1989).

12    Metabolism
13           The metabolic pathways of benzo[a]pyrene (Figure B-l) and variation in species, strains,
14    organ system, age and sex have been studied extensively with in vitro and in vivo experiments.  In
15    addition, there have been numerous studies of exposed humans or animals with subsequent
16    detection of benzo[a]pyrene metabolites in tissues or excreta. For example, elevated frequency of a
17    detected urinary metabolite (7,8,9,10-tetrol) was observed in patients treated with coal tar
18    medication (Bowman et al., 1997), demonstrating extensive metabolism of benzo[a]pyrene in
19    humans.
20           Phase I metabolism results in a number of reactive metabolites such as epoxide
21    intermediates, dihydrodiols, phenols, quinones, and their various combinations that are likely to
22    contribute to the toxic effects of benzo[a]pyrene (e.g. dihydrodiol epoxides and quinones). The
23    Phase II metabolism of benzo[a]pyrene metabolites protects cellular macromolecules from binding
24    with reactive benzo[a]pyrene diolepoxides and radical cations. These metabolic process include
25    glutathione conjugation of diol epoxides, sulfation and glucuronidation of phenols, and reduction of
26    quinones by NADPH:quinone oxidoreductase (NQO). Numerous reviews on the metabolism of
27    benzo[a]pyrene are available (Miller and Ramos, 2001; WHO, 1998; ATSDR, 1995; Conney et al.,
28    1994; Grover, 1986; Levin et al., 1982; Gelboin, 1980). Key concepts have been adapted largely
29    from these reviews and supplemented with recent findings.

30    Phase I metabolism
31           Phase I reactions of benzo[a]pyrene are catalyzed primarily by CYP450 and produce
32    metabolites including epoxides, dihydrodiols, phenols and quinones (Figure B-2).  The first step of
33    Phase I metabolism is reaction of benzo[a]pyrene into epoxides, the four major forms of which are
34    the 2,3-, 4,5-, 7,8-, and 9,10-isomers (Gelboin, 1980). Once formed, these epoxides may undergo
35    three different routes of metabolism: (1) spontaneous rearrangement to phenols, (2) hydration to
36    trans-dihydrodiols catalyzed by microsomal epoxide hydrolase, or (3) the Phase II detoxification of
37    binding with glutathione (either spontaneously or catalyzed by cytosolic glutathione-S-transferases

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-6          DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene

1    (IARC 1983)). The metabolism of benzo[a]pyrene to phenols occurs for 5 phenol isomers (1-, 3-, 6-,
2    7, and 9-OH benzo[a]pyrene) (Pelkonen etal. 1982). The hydration of benzo[a]pyrene epoxides to
3    trans-dihydrodiols occurs for all four major epoxide isomers (2,3-, 4,5-, 7,8-, and 9,10-). The
4    7,8-oxide is the focus of much of the study of benzo[a]pyrene metabolism, since it is a precursor to
5    the potent DNA-binding metabolite benzo[a]pyrene-7,8-diol-9,10-epoxide (BPDE). BPDE is formed
6    from benzo[a]pyrene 7,8-transdiol by multiple mechanisms including catalysis by CYPs (Deutsch
7    1979; Grover 1986), myeloperoxidase (MPO) (Mallet 1991), or prostaglandin h synthase (PHS, also
8    known as cyclooxygenase COX) (Marnett 1990), and lipid peroxidation (Byczkowski 1990). The
9    diolepoxides can react further by spontaneously hydrolyzing to tetrols (Hall and Grover 1988).
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-7          DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
          OH
        BaP7,8-dio|.9.10.epmide
   0'
6-oxo-BaP radical
    OH
BaP 1,6-hydroquinone
    Q-              0
BaP 1,6-semiquinone      BaP 1,6-quinone
                                                                    F BaP 3.6  I
                                                                    Lsemiquinone J
           BaP6,12-quinone
                                  OH
                               BaP 6.12-hydroquinone
                OH
             BaP3.6-hydroquinone
                         BaP3,6-quinone
 1
 2           Source:  Miller and Ramos (2001).

 3           Figure B-l.  Metabolic pathways for benzo[a]pyrene.

 4           The metabolism of benzo[a]pyrene, proceeds with a high degree of stereoselectivity. Liver
 5    microsomes from rats stereospecifically oxidize the 7,8-bond of benzo[a]pyrene to yield almost
 6    exclusively the (+)-benzo[a]pyrene-(7,8)-oxide (see Figure B-2).  Each enantiomer of the 7,8-oxide
 7    is stereospecifically converted by epoxide hydrolase (EH) to a different dihydrodiol and further
 8    metabolism of the (-)-benzo[a]pyrene-7,8-dihydrodiol enantiomer by rat GYP enzymes
 9    preferentially yields (+)-benzo[a]pyrene-7R,8S-diol-9S,10R-epoxide [(+)-anti- benzo[a]pyrene-7,8-
10    diol-9,10-epoxide (BPDE)], which is believed to be the most potent carcinogen among the four
11    stereoisomers (Figure B-2). Formation of these stereoisomers does not occur at equimolar ratios,
12    and the ratios differ between biological systems. For example a study in rabbit livers demonstrated
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  B-8          DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofbenzo[a]pyrene

 1    that purified microsomes oxidized the (-)-benzo[a]pyrene-7,8-dihydrodiol to isomeric diol epoxides
 2    in a ratio ranging from 1.8:1 to 11:1 in favor of the (+)-anti-BPDE isomer (Deutsch etal., 1979).
                                                                               HO
                                                                                      t
                                                                                   OH
         10
12   1
      ml2   Mixed Function
      j 3   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
         765
                                                                               HO  ^
                                                                                   OH
                                                                             (+)-BP-7S,8R-diol-9S, 10R-epoxide
                                                                                  (+) syn BPDE
                                                                                  O
                                    (-)-BP-7,8-oxide
                                                                               HO  ~m
                                                                                   OH
                                                                             (-)-BP-7S,8R-diol-9R, 10S-epoxide
 A                                                                                (-) anti BPDE
 5           Source: Grover (1986).

 6           Figure B-2. The stereospecific activation of benzo[a]pyrene.

 7           Several studies have attempted to determine which GYP isozyme is predominantly
 8    responsible for the metabolism of benzo[a]pyrene. Dermal administration of [3H]-benzo[a]pyrene
 9    to mice that have an Ah receptor (AhR) knock-out (AhR-/-) had significantly decreased formation
10    of (+)-anti-BPDE-DNA adducts compared to WT and IB I-/- mice (Kleiner etal. 2004). Gavage
11    administration of benzo[a]pyrene in AhR knock-out mice found the AhR-/- mice (with lower levels
12    of CYP1A1) had higher levels of protein adducts and unmetabolized benzo[a]pyrene than the
13    AhR+/+ or +/- mice (Sagredo et al., 2006).  Similarly, CYP1A1 (-/-) knock-out mice administered
14    benzo[a]pyrene in feed for 18 days had higher steady-state blood levels of benzo[a]pyrene and
15    benzo[a]pyrene-DNA adducts (Uno etal. 2006).  DNA post-labeling studies of mice administered by
16    gavage demonstrated higher benzo[a]pyrene-DNAadduct levels in CYP1A1(-/-) than CYP1A1(+/+)
17    mice in liver, small intestines, spleen and bone marrow (Uno et al., 2004). These findings establish
18    important roles inbenzo[a]pyrene metabolism for CYP1A1, but the relationship is not clear
19    between the GYP  enzymes and biological activation or detoxification.
20           Another important factor in evaluating variability in the metabolic activation of
21    benzo[a]pyrene by GYP P450s is the effect of functional polymorphisms, which has been the subject
22    of numerous reviews (e.g., Wormhoudt et al., 1999). Recombinant CYP1A1 allelic variants
23    produced BPDE with generally lower catalytic activity and Km values than the WT allele (Schwarz

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 B-9          DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    et al., 2001).  However, the formation of diol epoxides is stereospecific, with the allelic variants
 2    producing about three times the amount of (±)-anti-BPDE isomers as compared to the
 3    stereoisomers, (±)-syn-BPDE (Schwarz et al., 2001).  In a study of occupational exposures to
 4    benzo[a]pyrene, no relationship was observed between benzo[a]pyrene metabolite formation and
 5    the CYP1A1 Mspl polymorphism (Wu et al., 2002).
 6          Another metabolic pathway of benzo[a]pyrene metabolism is the conversion of
 7    transdihydrodiol-benzo[a]pyrene or 6-OH benzo[a]pyrene into quinones, primarily the 1,6-, 3,6-,
 8    7,8- and 6,12- isomers. Transdihydrodiol-benzo[a]pyrene such as (+/-)-anti-BPDE can be
 9    converted in  a redox cycling reaction into benzo[a]pyrene-7,8-quinone (BPQ) catalyzed by
10    dihydrodiol dehydrogenase (DD). This reaction pathway produces peroxide anion radicals,
11    benzo[a]pyrene semiquinone radicals, hydroxyl radicals, and H202 which in turn can causes
12    extensive DNA fragmentation (Penning 1999; Flowers etal., 1996; 1997).
13    6-Hydroxybenzo[a]pyrene can be oxidized into 6-oxo-benzo[a]pyrene semi-quinone radical and
14    further metabolized into 1,6-, 3,6-, or 6,12-quinones spontaneously, or catalytically by
15    prostaglandin endoperoxide synthetase (Eling, etal 1983).

16    Phase II metabolism
17          The reactive products of phase I metabolism are subject to the action of several phase II
18    conjugation and detoxification enzyme systems that display preferential activity for specific
19    oxidation products of benzo[a]pyrene. These phase II reactions play a critical role in protecting
20    cellular macromolecules from binding with reactive benzo[a]pyrene diolepoxides, radical cations,
21    or ROS. Therefore, the balance between Phase I activation of benzo[a]pyrene and its metabolites
22    and detoxification by Phase II processes is an important determinant of toxicity.
23          The diol epoxides formed from benzo[a]pyrene metabolism by Phase I reactions are not
24    usually found as urinary metabolites. Rather, they are detected as adducts of nucleic acids or
25    proteins o further metabolized by glutathione (GSH)  conjugation, glucuronidation, and sulfation.
26    These metabolites make up a significant portions of total metabolites in excreta or tissues For
27    example, the  identified metabolites in bile  6 hours after a 2 [J.g/kg benzo[a]pyrene dose by
28    intratracheal instillation to male Sprague-Dawley rats were 49% glucuronides (quinol
29    diglucuronides or monglucuronides), 30.4% thioether conjugates, 6.2% sulfate conjugates, and
30    14.4% unconjugated metabolites (Bevan and Sadler,  1992).
31          Conjugation of benzo[a]pyrene with GSH is catalyzed by GSTs. Numerous studies using
32    human GSTs  expressed in mammalian cell lines have demonstrated the ability of GST to metabolize
33    benzo[a]pyrene diol epoxides. Isolated human GST have significant catalytic activity toward
34    benzo[a]pyrene-derived diol epoxides and (±)anti-BPDE with variation in activity across GST
35    isoforms (Dreij etal. 2002; Robertson etal. 1986; Rojas etal. 1998).  Benzo[a]pyrene quinones can
36    also be conjugated with glutathione (Agarwal et al. 1991; IARC 1983). This compelling evidence for
37    a role of GSTs in the metabolism of reactive benzo[a]pyrene metabolites has triggered several
3 8    molecular epidemiology studies. However, recent studies on the impact of polymorphism on
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-10        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    adduct levels in PAH-exposed human populations did not show a clear relationship between the
 2    Phase I (CYP1A1, EH), or Phase II (GST) enzyme polymorphisms and formation of DNA adducts
 3    (Hemminki et al., 1997) or blood protein adducts (Pastorelli et al., 1998).
 4          Conjugation with UDP-glucuronide catalyzed by UGT enzymes is another important
 5    detoxification mechanism for oxidative benzo[a]pyrene metabolites. UGT isoforms, as well as their
 6    allelic variants, are expressed and have glucuronidation activity toward benzo[a]pyrene-derived
 7    phenols and diols in the aerodigestive tract (tongue, tonsil, floor of the mouth, larynx, esophagus),
 8    but not lung or liver (Zheng et al., 2002; Fang and Lazarus 2004). UGT activity also shows
 9    significant interindividual variability.  Incubation of lymphocytes with benzo [a]pyrene  resulted in
10    covalent binding to protein with a 143-fold interindividual variability and a statistically significant
11    inverse correlation between glucuronidation and protein binding (Hu and Wells, 2004).
12          Sulfotransferases can catalyze the formation of sulfates of benzo[a]pyrene metabolites. In
13    rat or mouse liver, cytosolic sulfotransferase (in the presence of 3'-phosphoadenosine 5'-
14    phosphosulfate) catalyzes formation of sulfates of three benzo[a]pyrene metabolites:
15    benzo[a]pyrene-7,8,9,10-tetrahydro-7-ol, benzo[a]pyrene-7,8-dihydrodiol, and benzo [a]pyrene-
16    7,8,9,10-tetrol. The benzo[a]pyrene-7,8,9,10-tetrahydro-7-ol-sulfate is able to form potentially
17    damaging DNA adducts (Surh and Tannenbaum, 1995). In human lung tissue 3-
18    hydroxybenzo[a]pyrene conjugation to sulfate produces benzo[a]pyrene-3-yl-hydrogen sulfate,  a
19    very lipid soluble compound  that would not be readily excreted in the urine (Cohen et al.  1976).
20          Although not specific for benzo[a]pyrene, there is now considerable evidence that genetic
21    polymorphisms of the GST, UGT, and EH genes impart an added risk to humans for developing
22    cancer. Of some significance  to the assessment of benzo [a]pyrene maybe that smoking, in
23    combination with genetic polymorphism at several gene loci, increases the risk for bladder cancer
24    (Moore etal., 2004; Choietal., 2003; Parketal., 2003) and lung cancer (Alexandrie etal.,  2004; Lin
25    et al., 2003). Coke oven workers (who are exposed to PAHs, including benzo[a]pyrene)
26    homozygous at the P187S site of the NQ01 gene (an inhibitor of benzo[a]pyrene-quinone adducts
27    with DNA), or carrying the null variant of the GSTM1 gene, had a significantly increased risk of
28    chromosomal damage in peripheral blood lymphocytes. Meanwhile, the risk was much lower than
29    controls  in subjects with a variant allele at the HllSYsite of the EH gene (Lengetal., 2004).

30    Tissue-specific Metabolism
31          Benzo[ajpyrene metabolism has been demonstrated in vivo in laboratory animals for
32    various tissues via multiple routes including inhalation, ingestion and dermal absorption. Nasal
33    instillation or inhalation of benzo[a]pyrene in monkeys, dogs, rats and hamsters resulted in the
34    formation of dihydrodiols, phenols,  quinones, and tetrols in the nasal mucus and lung (Petridou-
35    Fischer etal. 1988; Weyand and Bevan 1986,1987a,  1988; Dahl etal. 1985; Wolff etal. 1989b).  In
36    rats, the  fractions of metabolites in the lung at 6 hours after instillation were: 20% unmetabolised
37    benzo[a]pyrene, 16% conjugates or polyhydroxylated compounds, 10.7% 4,5-, 7,8-, and 9,10-
38    dihydrodiols, 9.3% 1,6-, 3,6-,  6,12- quinone, and 6.9% 3- and 9-hydroxybenzo[a]pyrene (Weyand
                This document is a draft for review purposes only  and does not constitute Agency policy.
                                                B-ll        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    and Bevan 1986). In hamsters, approximately 50% of the benzo[a]pyrene instilled was
 2    metabolized in the nose (nasal tissues had the highest metabolic acitivity per-gram of the
 3    respiratory tract tissues), and the metabolites produced were similar to other species (Dahl et al.
 4    1985).
 5          In vitro studies of human and laboratory cells and cell lines provide further quantitative and
 6    mechanistic details of the metabolism of benzo[a]pyrene in the cells of the respiratory tract, skin,
 7    liver and other tissues. Tracheobronchial tissues in culture of several species (including humans,
 8    mice, rats, hamsters, and bovines) were all found to metabolize benzo[a]pyrene extensively to
 9    phenols, diols, tetrols, quinones, and their conjugates (Autrup et al., 1980). The results show a high
10    degree of interindividual variability (a 33-fold difference in human bronchus, a 5-fold variation in
11    human trachea, and a 3-fold difference in bovine bronchus), but minimal variation among
12    individuals of the laboratory animal species (Autrup  et al.,  1980). Human bronchial epithelial and
13    lung tissue conjugated benzo [ajpyrene metabolites to glutathione and sulfates, but not with
14    glucuronide (Autrup et al. 1978; Cohen et al. 1976; Kiefer et al. 1988). The binding of
15    benzo [a]pyrene metabolites with DNA in primary human hepatocytes was associated with the
16    amount of unconjugated 7,8-dihydrodiol (Monteith et al. 1987).
17          Human and animal skin is able to metabolize  benzo[a]pyrene. Human skin samples
18    maintained in short term organ culture (i.e., human epithelial tissue, samples from human hair
19    follicles, and melanocytes isolated from adult human skin) can metabolize benzo[a]pyrene into
20    dihydrodiols, phenolas, quinones and  glucuronide and sulfate conjugates (Hall & Griver, 1988; Merk
21    etal., 1987; Alexandrov et al., 1990; Agarwal etal., 1991). The permeation of benzo[a]pyrene in
22    skin is linked to benzo[a]pyrene metabolism. Nonviable skin is unable to metabolize
23    benzo[a]pyrene (the permeation into nonviable skin  is lower than viable skin) as measured in a
24    range of species including humans, rat, mouse, rabbit and marmoset (Kao etal., 1985). Viable
25    human skin samples treated with 2 [ig/cm2 [14C]-benzo[a]pyrene in acetone and incubated for
26    24 hours produced the following proportions of benzo[a]pyrene metabolites; 52% water-soluble
27    compounds, 8% polar compounds, 17% diols, 1% phenols, 2.5% quinones and 18% unmetabolized
28    benzo[ajpyrene (Kao etal., 1985).
29          Benzo [ajpyrene is also metabolized by multiple reproductive tissues including prostate,
30    endometrium, cervical epithelial and styromal, and testes (Williams et al., 2000; Bao et al., 2002;
31    Melikian etal., 1999; Ramesh etal., 2003). Exposure  of fetal tissues to reactive benzo[a]pyrene
32    metabolites in utero is a concern. Transport of benzo[a]pyrene and benzo[a]pyrene metabolites to
33    fetal tissues including plasma, liver, hippocampus and cerebral cortex has been demonstrated in
34    multiple studies (McCabe and Flynn, 1990; Neubert and Tapken, 1988; Shendrikova and
35    Aleksandrov, 1974), and benzo[a]pyrene is metabolized by human fetal esophageal cell culture
36    (Chakradeo et al. 1993).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-12         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    Elimination
 2          Benzo[a]pyrene metabolites have been detected in the urine of exposed humans, but the
 3    fecal excretion has not been investigated in any detail. Studies of benzo[a]pyrene elimination in
 4    animals following exposure via inhalation, ingestion and dermal routes have shown benzo[a]pyrene
 5    is excreted preferentially in the feces in multiple species of laboratory animals including rat, mice,
 6    hamsters, guinea pigs, monkeys and dogs (Petridou-Fischer et al., 1988; Wolff et al., 1989; Sun et al.,
 7    1982; Wang et al., 2003; Weyand and Bevan, 1987; Yang et al., 1989; Hecht et al., 1979; Likhachev
 8    et al., 1992). The metabolites in bile are primarily benzo[a]pyrene conjugates, predominately
 9    thioether conjugates of varying extent in different species (Weyand and Bevan, 1987).  Six hours
10    after a single intratracheal instillation of benzo[a]pyrene (2 |J.g/kg) to male Sprague-Dawley rats,
11    relative metabolite levels were 31.2% diglucuronides, 30.4% thioether conjugates, 17.8%
12    monoglucuronides, 6.2% sulfate conjugates, and 14.4% unconjugated metabolites (Bevan and
13    Sadler, 1992). Rats administered benzo[a]pyrene via i.v. excrete a larger fraction in urine than via
14    inhalation or oral exposure, suggesting an important role for enterohepatic circulation of
15    benzo[a]pyrene metabolite conjugates (Moir etal., 1998; Weyand and Bevan, 1986; Hirom etal.,
16    1983). The vehicle impacts the amount of benzo[a]pyrene excreted and may in part be due to the
17    elimination rate or to other factors such as the absorption rate. For [3H]-benzo[a]pyrene
18    administered to Sprague-Dawley rats in hydrophilic triethylene glycol, 70.5% of the dose was
19    excreted into bile within 6 hours. If lipophilic solvents ethyl laurate and tricaprylin were used as
20    vehicles, 58.4 and 56.2% of the dose were excreted (Bevan and Ulman, 1991). In addition to
21    benzo[a]pyrene and its metabolites, adducts of benzo[a]pyrene with nucleotides have also been
22    identified as a small fraction of the administered dose in feces and urine of animals. The level of
23    BPDE adducts with guanine detected in urine of male Wistar rats was dose-dependent. 48 hours
24    after dosing with 100 [J.g/kg tritiated benzo[a]pyrene, 0.15% of the administered benzo[a]pyrene
25    dose was excreted in the urine as an adduct with guanine (Autrup and Seremet, 1986). Overall, the
26    data in humans and laboratory animals are sufficient to describe benzo[a]pyrene elimination
27    qualitatively but to limited to estimate quantitative rates of elimination.

28    Physiologically based pharmacokinetic models
29          Several toxicokinetic or pharmacokinetic models of benzo[a]pyrene have been developed
30    for rodents (rat and hamster). However, human models have only been developed via allometric
31    scaling, and metabolic parameters in humans have not been calibrated against in vivo toxicokinetic
32    data or in vitro experiments.
33          Bevan and Weyand (1988) performed compartmental pharmacokinetic analysis of
34    distribution of radioactivity in male Sprague-Dawley rats, following the intratracheal instillation of
35    benzo[a]pyrene to normal and bile duct-cannulated animals (Weyand and Bevan, 1987,1986).
36    However, implicit simulation approaches were used, as opposed to physiologically-based
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-13        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    approaches. The model calculated linear rate constants among compartments, and assumed the
 2    kinetics of benzo[a]pyrene and its metabolites were the same
 3          Roth and Vinegar (1990) reviewed the capacity of the lung to impact the disposition of
 4    chemicals and used benzo[a]pyrene as a case study. A PBPK model was presented based on data
 5    from Wiersma and Roth (1983a, b)  and was evaluated against tissue concentration data from
 6    Schlede et al. (1970). The model was structured with compartments for arterial blood, venous
 7    blood, lung, liver, fat, and slowly as  well as rapidly perfused tissues. Metabolism in liver and lung
 8    was estimated using kinetic data from control rats and rats pretreated with 3-MC to induce
 9    benzo[a]pyrene metabolism. The results of PBPK simulations showed that induction of
10    metabolizing enzymes increased the amount of benzo[a]pyrene cleared by the lungs relative to the
11    liver. An adequate fit was obtained for some compartments; however tissue-level data for
12    calibration and validation of this model were limited.
13          Moir et al. (1998) conducted a pharmacokinetic study on benzo[a]pyrene to obtain data for
14    model development Rats were injected with varying doses of [14C]-benzo[a]pyrene to 15 mg/kg
15    and blood, liver, fat,  and richly perfused tissue were sampled varying time points after dosing. Moir
16    (1999) then described a model for lung, liver, fat, richly and slowly perfused tissues, and venous
17    blood, with saturable metabolism occurring in the liver. The fat and richly perfused tissues were
18    modeled as diffusion-limited, while the other tissues were flow-limited. The model predicted the
19    blood benzo[a]pyrene concentrations well, although it overestimated the 6 mg/kg results at longer
20    times (>100 minutes). The model also produced a poor fit to the liver data. The model simulations
21    were also compared to data of Schlede etal. (1970), who had injected rats with 0.056 mg/kg body
22    weight of benzo[a]pyrene. The model predicted blood and fatbenzo[a]pyrene concentrations well,
23    but still poorly predicted liver benzo[a]pyrene concentrations. The model included only one
24    saturable metabolic  pathway, and only parent chemical concentrations were used to establish the
25    model. No metabolites were included in the model. This model was re-calibrated by Crowell etal.
26    (2011) by optimizing against additional rodent data and altering partition coefficient derivation.
27    However, it still did  not incorporate metabolites, and some tissues continued to exhibit poor model
28    fits.
29          An attempt to scale the Moir et al. (1998) rodent PBPK model to humans, relevant to risk
30    assessment of oral exposures to benzo[a]pyrene, was presented by Zeilmaker et al. (1999a, b). The
31    PBPK model for benzo[a]pyrene was derived from an earlier model for TCDD in rats (Zeilmaker and
32    van Eijkeren, 1997). Most compartments were perfusion-limited, and tissues modeled included
33    blood, adipose (with diffusion limitation), slowly and richly perfused tissues, and the liver.
34    However, there was  no separate compartment for the lung. The liver compartment featured the
35    AhR-dependent CYP450 induction mechanism and DNA adduct formation as a marker for
36    formation of genotoxic benzo[a]pyrene metabolites. It was assumed that DNA adduct formation
37    and the bulk benzo[a]pyrene metabolism were mediated by two different metabolic pathways. The
38    model was experimentally calibrated in rats with the data for EROD and formation of DNA adducts

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-14        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    in the liver after i.v. administration of a single dose and per os (p.o.) administration of a single or
 2    repeated doses of benzo[a]pyrene (Zeilmaker et al., 1999a).
 3          Zeilmaker et al. (1999b) assumed identical values for several parameters in rats and
 4    humans (i.e. benzo[a]pyrene tissue partition coefficients, AhR concentration in liver, rate constant
 5    for the decay of the benzo[a]pyrene-CYP450 complex, half-life of the CYP450 protein, fraction and
 6    rate of GI absorption of benzo[a]pyrene, and rates of formation and repair of DNAadducts in liver).
 7    The basal GYP45 0 activity in humans was assumed to be lower than that in rat liver. The
 8    mechanism of AhR-dependent induction of CYP450 dominated the simulated benzo[a]pyrene-DNA
 9    adduct formation in the liver. The results of PBPK model simulations indicated that the same dose
10    of benzo[a]pyrene administered to rats or humans might produce one order of magnitude higher
11    accumulation of DNA adducts in human liver when compared with the rat (Zeilmaker et al., 1999b).
12          Even though the model of Zeilmaker et al. (1999b) represents a major  improvement in
13    predictive modeling of benzo[a]pyrene toxicokinetics, the interspecies extrapolation introduce
14    significant uncertainties. As emphasized by the authors, the conversion of benzo[a]pyrene to its
15    mutagenic and carcinogenic metabolites could not be explicitly modeled in human liver because no
16    suitable experimental data were available.  According to the authors, improvement of the model
17    would require direct measurements of basal activities of CYP1A1 and CYP1A2  and formation of
18    benzo[a]pyrene-DNA adducts in human liver. Metabolic clearance of benzo[a]pyrene in the lungs
19    was also not addressed. Additionally, the toxicokinetic modeling by Zeilmaker et al. (1999b)
20    addressed only one pathway of benzo[a]pyrene metabolic activation, a single target organ (the
21    liver), and one route of administration (oral). In order to model health outcomes of exposures to
22    benzo[a]pyrene, the PBPK model needs to simulate rate of accumulation of benzo[a]pyrene-DNA
23    adducts and/or the distribution and fate of benzo[a]pyrene metabolites (e.g., BPDE) that bind to
24    DNA and other macromolecules. Alternatively, stable toxic metabolites (e.g., trans-anti-tetrol-
25    benzo[a]pyrene) may be used as an internal dose surrogate. While the metabolic pattern of
26    benzo[a]pyrene has been relatively well characterized qualitatively in animals, the quantitative
27    kinetic relationships between the more complex metabolic reactions in potential target organs are
28    not yet well defined.

29    Recommendations for the use of PBPK models in toxicity value derivation
30          PBPK models for benzo[a]pyrene were evaluated to determine the capability to extrapolate
31    from rats to humans, or between oral and inhalation exposure routes. Due to significant
32    uncertainties with respect to the inter-species scaling of the metabolic parameters between rats
33    and humans, these models were not used for cross-species extrapolation. Furthermore, no
34    complete mechanistic PBPK model for the inhalation route was identified, nor  was there a model
35    for humans that simulates the typical inhalation exposure to benzo[a]pyrene on poorly soluble
36    carbonaceous particles. This precluded the model's use for cross-route extrapolation to the
37    inhalation pathway.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-15         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    HUMAN STUDIES

 2    Non-Cancer Endpoints

 3    Cardiovascular Endpoints
 4          Burystn et al. (2005) reported the association of death from cardiovascular disease with
 5    B[a]P exposure in a cohort of 12,367 male European asphalt workers (Table B-l). These workers
 6    were first employed in asphalt paving between 1913 and 1999, and worked at least one season.
 7    Average  duration of follow-up was 17 ± 9 years (mean ± SD), encompassing 193,889 person-years
 8    of observation. Worker exposure to coal tar was estimated using industrial process and hygiene
 9    information and modeling (presented in a previous report), and coal tar exposure was found to be
10    the strongest determinant of exposure to B[a]P. Benzo[a]pyrene exposure was assessed
11    quantitatively using measurement-driven mixed effects exposure models, using data collected from
12    other asphalt industry workers, and this model was constructed and validated previously.  Due to
13    limited data availability, only information regarding the primary cause of death was collected, and
14    this analysis was limited to diseases of the circulatory system (ICD codes 390 - 459), specifically
15    ischemic heart disease (IHD: ICD codes 410 - 414). Diesel exhaust exposure was also assessed in
16    this cohort, but varied little among the asphalt pavers, and was not associated with risk of death
17    from cardiovascular disease.  0.25% of the cohort was lost to follow-up, and 0.38% emigrated
18    during the course of observation. Relative risks and associated 95% confidence intervals were
19    estimated using Poisson regression, and all models included exposure index for agent of interest
20    (coal tar or B[a]P), age, calendar period of exit from cohort, total duration of employment and
21    country, using the category of lowest exposure as the reference. Confounding by tobacco smoke
22    exposure was considered in relation to the strength of its association with cardiovascular disease
23    and the smoking prevalence in the population. The RR attributed to cigarette smoking in former
24    and current smokers was assumed to be 1.2 and 2, respectively, based upon literature reports.
25    From analysis of smoking incidence in a sub-cohort, the following smoking distribution was
26    proposed: in the lowest exposure group, 40% never smokers, 30% former smokers and 30%
27    current smokers; among the highest exposed, the proportion shifted to 20/30/50%, respectively.
28          Exposed subjects were stratified into quintiles based upon IHD mortality, with 83 - 86
29    deaths per exposure category, composing approximately 2/3 of the 660 cardiovascular disease-
30    related deaths. Both cumulative and average exposure indices for B[a]P were positively associated
31    with IHD mortality, with a RR of approximately 1.6 in the highest exposure quintile from both
32    metrics,  independent of total employment duration. Similar monotonic trends were observed for
33    all cardiovascular diseases (combined), although a dose-response relationship was evident only for
34    IHD and  not hypertension or other individual heart disease categories.  Similar trends were also
35    observed for coal tar exposure and IHD. Adjusting the RR to account for possible confounding by
36    smoking yields a RR of 1.39 under the assumptions mentioned above, and is still elevated (1.21) if
37    the contribution of smoking to cardiovascular disease etiology was greater than the original

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-16        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1   assumptions. Furthermore, the RR for the high vs. low exposure quintile is 1.24 even if the
 2   distribution of non-smokers/former smokers/current smokers shifts to 0/30/70%, using the
 3   original assumptions of cigarette smoke casual potency.
 4
 5
       Table B-l. Exposure to benzo[a]pyrene and mortality from
       cardiovascular diseases in a European cohort of asphalt paving workers
Effect measured
Cumulative exposure (ng/m3 -years)
0 - 189a
189 - 501
502-931
932 - 2012
>2013
Pfor
trend
Diseases of the circulatory system
Deaths
RR
95% Cl
137
1.00
145
1.08
0.85-1.38
118
1.06
0.80-1.42
132
1.24
0.89-1.71
128
1.42
0.96-2.09
0.09
Ischemic heart disease
Deaths
RR
95% Cl
Effect measured
83
1.00
83
0.99
0.72-1.36
84
1.22
0.86-1.74
83
1.24
0.82-1.85
85
1.58
0.98-2.55
Average exposure (ng/m3)
0-68a
68 - 105
106 - 146
147 - 272
>273
0.06
Pfor
trend
Diseases of the circulatory system
Deaths
RR
95% Cl
128
1.00
142
1.30
1.01-1.67
143
1.55
1.18-2.05
139
1.45
1.09-1.93
108
1.58
1.16-2.15
<0.001
Ischemic heart disease
Deaths
RR
95% Cl
83
1.00
83
1.13
0.82-1.55
83
1.33
0.94-1.90
86
1.20
0.84-1.71
83
1.64
1.13-2.38
0.02
 6
 7
 8
 9
10
11
12
13
14
15
16
17
 a Reference category

 Source: Burstyn et al. (2005).

       Friesen et al. (2010) examined the association between B [a]P exposure and deaths from
chronic non-malignant disease in a cohort of 6.423 male and 603 female Canadian aluminum
smelter workers (Table B-2). Inclusion criteria required at least 3 years of continuous employment
in either the smelter facility or power-generating station from 1954 - 1997, with worker history
collected up through 1999. This cohort was probabilistically linked to the Canadian national
mortality database for external comparison to the British Columbia population and calculation of
standardized mortality ratios, which were adjusted for age, sex and time period.  Ninety-five %
confidence intervals were calculated for the SMRs assuming a Poisson distribution. Internal
comparisons were also made during the analysis of IHD mortality in male workers, calculating
hazard ratios (HR) for IHD with or without acute myocardial infarction (AMI) after 1969, as AMI
could not be differentiated from other IHD on death certificates issued previously. HRs were

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-17         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    calculated using Cox regression models, with age as a metamarker of time, also including smoking
 2    status, time since 1st employed and work location status.  Smoking information for 77% of this
 3    updated cohort was collected by questionnaire, and workers categorized as 75% ever-smokers and
 4    25% never-smokers. Quantitative exposure to coal tar pitch volatiles were estimated by B[a]P
 5    measurements, calculated by a job classification and time-based exposure matrix, as described in a
 6    previous report; annual arithmetic mean values were calculated for exposures from 1977 - 2000,
 7    while pre-1977 levels were backwards-extrapolated from 1977 values, incorporating major
 8    technological changes in time periods as appropriate.
 9           Cumulative exposure metrics were highly skewed. Cumulative B[a]P with a 5-year lag (past
10    B[a]P exposure) and cumulative B[a]P in the most recent 5 years (recent B [a]P exposure) were only
11    slightly positively correlated (r = 0.10, P < 0.001). Current B[a]P exposure was highly correlated
12    with cumulative exposure for the most recent 5 years of exposure (r = 0.86, P < 0.001), but not with
13    5-year lagged cumulative exposure (r = 0.03, P < 0.001). Lagged cumulative exposure metrics (0 -
14    10 years) were all highly correlated with each other (r = 0.96, all P's  < 0.001); lagged metrics for
15    cumulative exposure were used to distinguish between effects of current versus long-term
16    exposure.
17           When exposed workers were pooled and compared externally to non-exposed referents, the
18    IHD and AMI standardized mortality ratios were all < 1.00 for males, and the only significant
19    association in females was an SMR of 1.27 for AMI. For internal comparisons, exposed males were
20    stratified into quintiles based upon IHD mortality, with approximately 56 deaths per exposure
21    category. 5-year lagged cumulative B[a]P exposure was significantly associated with elevated risk
22    of IHD mortality, HR = 1.62 (95% CI: 1.06, 2.46)  in the highest exposure quintile, while no
23    association was observed between most recent (5 years) exposure and mortality. Restricting IHD
24    events to only AMI (1969 onward) resulted in similar monotonic trends, albeit of lower statistical
25    significance.  No association was observed between B [a] P exposure and non-AMI IHD. While there
26    was little difference in the exposure-response association among 0, 2 and 5-year lagged data,  10-
27    year lagged data resulted in a weaker association. All risk estimates were strengthened by the
28    incorporation of work status and time-since-hire to account for the healthy worker effect, as
29    evidenced by the SMR of 0.87 (95% CI: 0.82, 0.92) for all chronic non-malignant diseases combined
30    in male exposed workers versus external referents. Using a continuous variable, the authors
31    calculated thatthe risk of death from IHD to be 1.002 (95% CI: 1.000,1.005) per [ig/m3  from
32    cumulative B[a]P exposure; however, visual inspection of the categorical relationships indicated
33    that the association is nonlinear, suggesting that this value may be an underestimate. Restricting
34    the cohort to only members who died within 30 days of active employment at the worksite,
35    cumulative B[a]P exposure was not significantly associated with IHC or AMI, although the HR for
36    the highest exposure group was 2.39 (95% CI: 0.95, 6.05). Exposure-response relationships were
37    similarly examined in male smelter workers for chronic obstructive pulmonary disease  (COPD) and
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-18        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1   cerebrovascular disease, but neither was significantly associated with cumulative B[a]P exposure in
 2   either internal or external comparisons.
 4
 5
       Table B-2. Exposure to benzo[a]pyrene and mortality from
       cardiovascular diseases in a Canadian cohort of male aluminum smelter
       workers

Effect measured
Categorical cumulative exposure with a 5-year lag (u.g/ms -year)
0
0 - 7.79
7.79 - 24.3
24.3 - 66.7
>66.7
Pfor
trend3
Continuous13

All ischemic heart disease (1957 onward)
Deaths
P-YC of follow-
up
HR
95% Cl
56
33,111
1
referent

56
37,581
1.11
0.76-1.62

57
34,838
1.48
1.01-2.17

56
31,533
1.28
0.86-1.91

56
13,688
1.62
1.06-2.46


0.053



281
150,751
1.002
1.000,
1.005
Acute myocardial infarction (1969 onward)

Deaths
P-YC of follow-
up
HR
95% Cl
0
35
25,071
1
referent

0-7.51
37
30,454
1.14
0.71, 1.82

7.51-27.7
37
34,621
1.21
0.75, 1.96

27.7 -67 A
38
24,081
1.36
0.84, 2.45

>67.4
37
13,261
1.46
0.87, 2.45



0.19



184
127,488
1.001
0.997,
1.005
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
 a Two-sided test for trend using the person-year-weighted mean value for each category as a linear,
 continuous variable.
 b Exposure variable was entered as a continuous, linear variable in the model
 c P-Y, person-years

 Source: Friesen et al. (2010).

Reproductive and Developmental Endpoints
       Wu etal. (2010) conducted a study of benzo[a]pyrene-DNAadduct 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 (i.e., a fetal death that remained in utero and therefore
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 1 year was also an eligibility
criterion. The participation rate was high: 81/84 eligible cases participated and 81/89 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

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-19        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    was no correlation between blood and aborted tissue levels of benzo[a]pyrene adducts (r = -0.12
 2    for the 102 blood-tissue pairs, r = -0.02 for the 51 case pairs and r = -0.21 for the 51 control pairs).
 3    (The authors noted that there was little difference between women with and without blood
 4    samples in terms of the interview-based measures collected or in terms of the DNA-adduct levels in
 5    aborted tissue.)  Benzo[a]pyrene-adduct levels were similar but slightly lower in the aborted tissue
 6    of cases compared with controls (mean ± SD 4.8 ± 6.0 incases and 6.0 ± 7.4 in controls, p = 0.29). In
 7    the blood samples, however, benzo[a]pyrene-adduct levels were higher in cases (6.0 ± 4.7 and 2.7 ±
 8    2.2 in cases and controls, respectively, p < 0.001). In logistic regression analyses using a continuous
 9    adduct measure, the OR was 1.35 (95% CI 1.11-1.64) per adduct/108 nucleotide.  These results
10    were adjusted for education and household income, but were very similar to the unadjusted results.
11    Categorizing exposure at the median value resulted in an adjusted OR of 4.27 (95% CI 1.41-12.99)
12    in the high compared with low benzo[a]pyrene-adduct group. There was no relation between
13    benzo[a]pyrene-adduct levels in the aborted tissue and missed abortion in the logistic regression
14    analyses using either the continuous (adjusted OR 0.97, 95% CI 0.93-1.02) or dichotomous
15    exposure measure (adjusted OR 0.76, 95% CI 0.37-1.54). Associations between missed abortion
16    and several interview-based measures of potential PAH exposure were also seen: adjusted OR 3.07
17    (95% CI 1.31-7.16) for traffic congestion near residence, 3.52 (95% CI 1.44-8.57) for commuting
18    by walking, 3.78 (95% CI 1.11-12.87) for routinely cooked during pregnancy, and 3.21  (95% CI
19    0.98-10.48) for industrial site or stack near residence, but there was no association with other
20    types of commuting (e.g., by bike, car, or bus).
21          Perera etal. (2005a) studied 329 nonsmoking pregnant women (30 ± 5 years old) possibly
22    exposed to PAHs from fires during the 4 weeks after 09/11/2001. Maternal and umbilical cord
23    blood levels of benzo[a]pyrene (BPDE)-DNA adducts were highest in study participants who lived
24    within 1 mile of the WTC, with an inverse correlation between cord blood levels and distance from
25    the WTC. Neither cord blood adduct level nor ETS alone was positively correlated with adverse
26    birth outcomes.  However, the interaction between ETS exposure and cord blood adducts was
27    significantly associated with reduced birth weight and head circumference.  Among babies exposed
28    to ETS in utero, a doubling of cord blood benzo[a]pyrene-DNA adducts was associated with an 8%
29    decrease in birth weight (p = 0.03) and a 3% decrease in head circumference (p = 0.04).
30          Perera et al. (2005b) compared various exposures—ETS, nutrition, pesticides, material
31    hardship—with birth outcomes (length, head circumference, cognitive development).  ETS
32    exposure and intake of PAH-rich foods by pregnant women were determined by questionnaire.
33    Levels of benzo[a]pyrene diol epoxide (BPDE)-DNA adducts were determined in umbilical cord
34    blood collected at delivery. The study population consisted of Dominican or African-American
35    nonsmoking pregnant women (n = 529;24±5 years old) free of diabetes, hypertension, HIV, and
36    drug or alcohol abuse. Benzo[a]pyrene adducts, ETS, and dietary PAHs were not significantly
37    correlated with each other.  However, the interaction between benzo[a]pyrene-DNA adducts and
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-20         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    ETS exposure was significantly associated with reduced birth weights (-6.8%; p = 0.03) and
 2    reduced head circumference (-2.9%; p = 0.04).
 3          Tang et al. (2006) measured benzo[a]pyrene diol epoxide (BPDE)-DNA adducts in maternal
 4    and umbilical cord blood obtained at delivery from a cohort of 150 nonsmoking women and their
 5    newborns in China.  Exposure assessment was related to the seasonal operation of a local, coal-fired
 6    power plant; however, airborne PAH concentrations were not measured. Dietary PAH intake was
 7    not included as a covariate because it did not significantly contribute to the final models, but ETS,
 8    sex, and maternal height and weight were considered as covariates.  DNA adduct levels were
 9    compared to several birth outcomes and physical development parameters, such as gestational age
10    at birth; infant sex, birth weight, length, head circumference, and malformations; maternal height
11    and pregnancy weight total weight gain; complications of pregnancy and delivery; and medications
12    used during pregnancy.
13          High cord blood adduct levels were significantly associated with reduced infant/child
14    weight at 18 months (P = -0.048, p = 0.03), 24 months  (P = -0.041, p  = 0.027), and 30 months of age
15    (P = -0.040, p = 0.049); decreased birth head circumference was marginally associated with DNA
16    adductlevels (P = -0.011, p = 0.057). Maternal adduct levels were correlated neither with cord
17    blood adduct levels nor with fetal and child growth. Among female infants, cord blood adduct levels
18    were significantly associated with smaller birth head circumference (p = 0.022) and with lower
19    weight at 18 months (p = 0.014), 24 months (p = 0.012), and 30 months of age (p = 0.033), and with
20    decreased body length at 18 months of age (p = 0.033). Among male infants, the corresponding
21    associations were also inverse but were not statistically significant
22          Considerable evidence of a deleterious effect of smoking on male and female fertility has
23    accumulated from epidemiological studies of time to pregnancy, ovulatory disorders, semen
24    quality, and spontaneous abortion (reviewed in Waylen et al., 2009;  Cooper and Moley, 2008;
25    Scares and Melo, 2008).  In addition, the effect of smoking, particularly during the time of the
26    perimenopausal transition, on acceleration of ovarian senescence (menopause) has also been
27    established (Midgette and Baron,  1990). More limited data are available pertaining specifically to
28    measures of benzo[a]pyrene and reproductive outcomes.
29          Neal etal. (2008, 2007) examined levels of benzo[a]pyrene and other PAHs infollicular
30    fluid and serum sample from 36 women undergoing in vitro fertilization at a clinic in Toronto, and
31    compared the successful conception rate in relation to benzo[a]pyrene levels. The women were
32    classified by smoking status, with 19 current cigarette smokers, 7 with passive or sidestream
33    smoke exposure (i.e., nonsmoker with a partner who smoked), and 10 nonsmokers exposed. An
34    early follicular phase blood sample and follicular fluid sample from the follicle at the time of ovum
35    retrieval were collected and analyzed for the presence of benzo[a]pyrene, acenapthelene,
36    phenanthrene, pyrene, and chrysene using gas chromatography/MS (detection limit 5 pg/mL). The
37    frequency of nondectable levels of serum benzo[a]pyrene was highest in the nonsmoking group
38    (60.0,14.3, and 21.0% below detection limit in nonsmoking, sidestream smoke, and active smoking

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-21        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    groups, respectively). A similar pattern was seen with follicular fluid benzo[a]pyrene (30.0,14.3,
 2    and 10.5% below detection limit in nonsmoking, sidestream smoke, and active smoking groups,
 3    respectively).  In the analyses comparing mean values across groups, an assigned value of 0 was
 4    used for nondetectable  samples. Follicular fluid benzo[a]pyrene levels were higher in the active
 5    smoking group (mean ± SE, 1.32 ± 0.68 ng/mL) than in the sidestream (0.05 ± 0.01 ng/mL) or
 6    nonsmoking (0.03 ± 0.01 ng/mL) groups (p = 0.04). The between-group differences in serum
 7    benzo[a]pyrene levels were not statistically significant (0.22  ± 0.15, 0.98 ± 0.56, and 0.40 ±
 8    0.13 ng/mL in nonsmoking, sidestream smoke, and active smoking groups, respectively), and there
 9    were no differences in relation to smoking status. Among active smokers, the number of cigarettes
10    smoked per day was strongly correlated with follicular fluid benzo[a]pyrene levels (r = 0.7, p <
11    0.01).  Follicular fluid benzo[a]pyrene levels were significantly higher among the women who did
12    not conceive (1.79 ng/mL ± 0.86) compared with women who did get pregnant (mean
13    approximately 0.10 ng/mL, as estimated from graph) (p < 0.001), but serum levels of
14    benzo[a]pyrene were not associated with successful conception.
15           A small case-control study conducted between August 2005 and February 2006 in Lucknow
16    city (Uttar Pradesh), India examined PAH concentrations in placental tissues (Singh et al., 2008) in
17    relation to risk of preterm birth. The study included 29 cases (delivery between 28 and <36 weeks
18    of gestation) and 31 term delivery controls. Demographic data smoking history, reproductive
19    history, and other information were collected by interview, and a 10 g sample  of placental tissue
20    was collected from all participants. Concentration of specific PAHs in placental tissue was
21    determined using HPLC. In addition to benzo[a]pyrene, the PAHs assayed were naphthalene,
22    acenapththylene, phenanthrene, fluorene, anthracene, benzo(a)anthracene, fluoranthene, pyrene,
23    benzo(k)fluoranthene, benzo(b)fluoranthene, benzo(g,h,i)perylene, and dibenzo(a,h)anthracene.
24    PAH exposure in this population was from environmental sources and from cooking.  The age of
25    study participants ranged from 20 to 35 years. There was little difference in birth weight between
26    cases and controls (mean 2.77 kg and 2.75 kg in the case and control groups, respectively).
27    Placental benzo[a]pyrene levels were lower than the levels of the other PAHs detected (mean 8.83
28    ppb in controls for benzo[a]pyrene compared with 25-30 ppb for anthracene,
29    benzo(k)fluoranthene, benzo(b)fluoranthene, and dibenzo(a,h)anthracene, 59 ppb for
30    acenaphthylene, and 200-380 ppm for naphthalene, phenanthrene, fluoranthene, and pyrene;
31    nondetectable levels of fluorine, benzo(a)anthracene, and benzo(g,h,i)perylene were found).  There
32    was little difference in benzo[a]pyrene levels between cases (mean ± SE 13.85 ± 7.06 ppb)  and
33    controls (8.83 ± 5.84 ppb), but elevated levels of fluoranthene (325.91 ± 45.14 and 208.6 ± 21.93
34    ppb in cases and controls, respectively, p < 0.05) and benzo(b)fluoranthene (61.91 ± 12.43 and
35    23.84 ± 7.01 ppb in cases and controls, respectively, p < 0.05) were seen.
36           Wuetal. (2010) conducted a study of benzo[a]pyrene-DNAadduct levels in relation to risk
37    of fetal death in Tianjin, China. This case-control study included women who experienced a missed
38    abortion before 14 weeks gestational age (i.e., a fetal death that remained in utero and therefore

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-22        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    required surgical intervention).  Cases were matched by age and gravidity to controls (women
 2    undergoing induced abortion due to an unplanned or unwanted pregnancy). The study excluded
 3    women who smoked, women with chronic disease and pregnancy complications, and women with
 4    occupational exposures to PAHs. Residency within Tianjin for at least 1 year was also an eligibility
 5    criterion. The participation rate was high: 81/84 eligible cases participated and 81/89 eligible
 6    controls participated.  Data pertaining to demographic characteristics, reproductive history, and
 7    factors relating to potential PAH exposure were collected using a  structured interview, and samples
 8    from the aborted tissue were obtained. In two of the four hospitals used in the study, blood
 9    samples from the women (n = 51 cases and 51 controls) were also collected. The presence of
10    benzo[a]pyrene-BPDE adducts was assessed in the blood and tissue samples using HPLC. There
11    was no correlation between blood and aborted tissue levels of benzo[a]pyrene adducts (r = -0.12
12    for the 102 blood-tissue pairs, r = -0.02 for the 51 case pairs and r = -0.21 for the 51 control pairs).
13    (The authors noted that there was little difference between women with and without blood
14    samples in terms of the interview-based measures collected or in terms of the DNA-adduct levels in
15    aborted tissue.)  Benzo[a]pyrene-adduct levels were similar but slightly lower in the aborted tissue
16    of cases compared with controls (mean ±  SD 4.8 ± 6.0 in cases and 6.0 ± 7.4 in controls, p = 0.29). In
17    the blood samples, however, benzo[a]pyrene-adduct levels were higher in cases (6.0 ± 4.7 and 2.7 ±
18    2.2 in cases and controls, respectively, p < 0.001). In logistic regression analyses using a continuous
19    adduct measure, the OR was 1.35 (95% CI 1.11-1.64) per adduct/108 nucleotide. These results
20    were adjusted for education and household income, but were very similar to the unadjusted results.
21    Categorizing exposure at the median value resulted in an adjusted OR of 4.27 (95% CI 1.41-12.99)
22    in the high compared with low benzo[a]pyrene-adduct group. There was no relation between
23    benzo[a]pyrene-adduct levels in the aborted tissue and missed abortion in the logistic regression
24    analyses using either the continuous (adjusted OR 0.97, 95% CI 0.93-1.02) or dichotomous
25    exposure measure (adjusted OR 0.76, 95% CI 0.37-1.54). Associations between missed abortion
26    and several interview-based measures  of potential PAH exposure were also seen: adjusted OR 3.07
27    (95% CI 1.31-7.16) for traffic  congestion  near residence, 3.52 (95% CI 1.44-8.57) for commuting
28    by walking, 3.78 (95% CI 1.11-12.87) for routinely cooked during pregnancy, and 3.21 (95% CI
29    0.98-10.48) for industrial site or stack near residence, but there was no association with other
30    types of commuting (e.g., by bike, car, or bus).

31    Neurotoxicity
32          Niu et al. (2010) studied 176 Chinese coke-oven workers with elevated B[a]P exposure and
33    compared them against 48 referents (workers in a supply warehouse), matched by socioeconomic
34    status, lifestyle and health. Blood levels of monoamine, amino acid and chloine neurotransmitters
35    were measured, and the WHO Neurobehavioral Core Test Battery (NCTB) was administered to
36    assess emotional state, learning, memory and hand-eye coordination. The authors self-designed a
37    study questionnaire to gather information on worker education, vocational history, smoking and
38    drinking habits, personal habits, personal and family medical history, as well as any current
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-23         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    symptoms and medications used in the pervious several weeks. Workers were excluded from the
 2    study for any of the following criteria: reported feeling depressed at any point during the previous
 3    6 months; had taken medicine in the previous 2 weeks which could affect nervous system function;
 4    or if they reported undertaking vigorous exercise less than 48 hrs previously. "Smoking" was
 5    defined as > 10 cigarettes/day during the past year. Similarly, "drinking" was defined as
 6    wine/beer/spirits consumed > 3 times/week for the past 6 months. Workplace environmental
 7    sampling stations were established at each of the physical work locations, including the referent's
 8    warehouse, and dual automatic air sampling pumps collected samples at personal breathing zone
 9    height for 6 hours/day, over 3 consecutive days. B[a]P content was determined by HPLC, and
10    relative exposure was compared to post-shift urine levels of aB[a]P metabolite, 1-hydroxypyrene
11    (1-OH-Py). Blood was collected in the morning before breakfast; monoamine (norepinephrine and
12    dopamine) and amino acid (Glu, Asp, Gly, and GABA) neurotransmitter levels were determined by
13    HPLC, acetylcholine (Ach) levels determined by hydroxyamine chromometry, and Ach esterase
14    (AchE) levels measured in lysed RBCs using activity kits.
15          B[a]P mean concentrations were 19.56 ± 13.2,185.96 ± 38.6 and 1623.56 ± 435.8 ng/m3 at
16    the bottom, side and top of the coke oven, respectively, all of which were higher than the mean at
17    the referents' warehouse (10.26 ± 7.6 ng/m3).  The authors did not report stratified analysis by
18    different levels of B[a]P exposure, and reported only comparisons between the referents and all
19    exposed workers combined (Table B-3), or between workers grouped by urinary B[a]P metabolite
20    1-OH-Py levels (Table B-4).  There were no significant differences in age, education, smoking or
21    alcohol use between the coke oven and warehouse workers.  Urinary 1-OH-Py levels were 32%
22    higher in coke oven workers compared to the referent group, corresponding to the higher levels of
23    B[a]P detected in all coke oven workstation compared to the supply warehouse.  Performance in
24    two neurobehavioral function tests, digit span and forward digit span, were significantly decreased
25    in the exposed oven workers versus control group; when stratified by urinary metabolite level,
26    scores significantly decreased with increasing 1-OH-Py levels. Of the neurotransmitters assessed,
27    norepinephrine, dopamine, Asp  and GABA were significantly decreased in exposed versus control
28    workers; norepinephrine and Asp were also significantly and inversely related with 1-OH-Py levels.
29    Dopamine levels appeared to decrease with increased urinary metabolite levels, although the
30    relationship was not statistically significant  GABA levels were highly variable, and appeared to
31    increase with increasing 1-OH-Py levels, although this relationship was statistically significant.
32    Acetylcholine levels were 4-fold higher in coke oven workers compared to referents, and AchE
33    actiivty 30% lower; both Ach and AchE were significantly associated with urinary B[a]P metabolite
34    levels, although Ach increased and AchE activity decreased with increasing 1-OH-Py.  The authors
3 5    reported results of correlation analysis, indicating that digit span scores correlated negatively with
36    Ach and positively with AchE (coefficients of-0.230, -0.276 and 0.120, 0.170, respectively),
37    although no indication of statistical significance was given. No other associations were  reported.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-24         DRAFT—DO NOT CITE OR QUOTE

-------
1
2
                                      Toxicological Review ofbenzo[a]pyrene


Table B-3. Exposure-related effects in Chinese coke oven workers or
warehouse controls exposed to benzo[a]pyrene in the workplace
Effect measured
Exposure Group
Controls (n=48)
Exposed workers (n=176)
P value
Background information (mean ± SD, incidence or %)
Age (yr)
Education (junior/senior)
Smoking
Drinking
39.71 ±7.51
23/25
77%
27%
37.86 ±6.51
110/66
64%
39%
0.098
0.068
0.093
0.140
Urine B[a]P metabolite (u.mol/mol Cr; mean ± SD)
1-OH-Py
2.77 ± 1.45
3. 66 ±0.67
0.000
Neurobehavioral function tests (mean ± SD)
Simple reaction time
Digit span
Forward digit span
413.88 ±95.40
17.31 ±4.54
10.65 ± 2.42
437.39 ± 88.44
15.47 ±4.08
9.25 ± 2.64
0.109
0.006
0.001
Neurotransmitter concentrations (mean ± SD)
Norepinephrine (ng/ml)
Dopamine (ng/ml)
Asp (u.g/ml)
Glu (u.g/ml)
GABA (ng/ml)
Ach (u.g/ml)
AchE activity (U/mg protein)
62.54 ±58.07
1566.28 ±317.64
2.13 ± 1.66
11.21 ±5. 28
2.52 ±5. 16
172.60 ±67. 19
71.31 ±46. 18
40.62 ± 29.78
1425.85 ±422.66
1.58 ±0.99
9.68 ±5. 72
1.01 ±2.21
704.00 ± 393.86
50. 27 ±34.02
0.000
0.029
0.004
0.074
0.004
0.000
0.012
     Source:  Niu et al. (2010).
             This document is a draft for review purposes only and does not constitute Agency policy.
                                           B-25       DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
1
2
            Table B-4. Exposure-related effects in Chinese coke oven workers or
            warehouse controls exposed to benzo[a]pyrene in the workplace,
            stratified by urinary metabolite levels



Effect measured
Number of subjects
Exposure Group categoried by 1-OH-Py level
0 - 3.09

u,mol/mol Cr
33
3.09-3.90

u,mol/mol Cr
72
3.90-5.53

u,mol/mol Cr
36


P value


Neurobehavioral function tests (mean ± SD)
Digit span
Forward digit span
Backward digit span
Right dotting
18.24 ±4.58
10.85 ±2. 12
7. 20 ±3.07
152.15 ±35.43
16.04 ±4.24
9.80 ± 2.86
6.38 ±2.55
153.80 ±31.55
15. 78 ±3.71
9.58 ±2.33
6. 20 ±2. 15
167.22 ±59.21
0.003
0.019
0.089
0.094
Neurotransmitter concentrations (mean ± SD)
Norepinephrine (ng/ml)
Dopamine (ng/ml)
Asp (u.g/ml)
Glu (u.g/ml)
GABA (ng/ml)
Ach (u.g/ml)
AchE activity (U/mg
protein)
67.31 ±67.45
1614.45 ± 683.57
2. 29 ±2. 13
11.56 ±8.92
1.40 ±3.59
334.66 ±83. 75
68.17 ±9.28

36.97 ±23. 58
1482.30 ±323. 66
1.61 ±0.71
9.93 ±4.14
1.42 ± 3.44
483.71 ±57.87
54.98 ±4.23

46.75 ±35.88
1405.06 ±332. 23
1.47 ±0.58
9.06 ±3.30
1.56 ±3. 24
665.85 ± 94.34
52.64 ±4.60

0.002
0.134
0.001
0.070
0.964
0.030
0.043

 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
     Source:  Niu et al. (2010).

    Immuno toxicity
           Zhang et al. (2012) studied 129 Chinese coke-oven workers with elevated B[a]P exposure
    and compared them against 37 referents (workers in a supply warehouse), matched by
    socioeconomic status, lifestyle and health. Area B[a]P levels were quantified in the various work
    areas, and the primary endpoint was the level of early and late apoptosis in PBMCs isolated from
    each worker sub-group the morning following an overnight fast  The authors self-designed a study
    questionnaire to gather information on worker education, vocational history, smoking and drinking
    habits, personal habits, personal and family medical history, as well as any current symptoms and
    medications used in the pervious several weeks. "Smoking" was defined as > 10 cigarettes/day
    during the past year, with "smoking index" defined as cigarettes/day x years smoking.  Similarly,
    "drinking" was defined as wine/beer/spirits consumed > 3 times/week for the past 6 months, and
    "drinking index" defined as grams of alcohol consumed/day x years drinking. Exposed workers
    were categorized by physical worksite location and expected differences in B[a]P exposure: 34
    oven bottom workers, 48 oven side workers, and 47 oven top workers.  Workplace environmental
    sampling stations were established at each of the physical work locations, including the referent's
    warehouse, and dual automatic air sampling pumps collected samples at personal breathing zone
    height for 6 hours/day, over 3 consecutive days. B[a]P content was determined by HPLC, and

              This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-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
                                                      Toxicological Review ofbenzo[a]pyrene

      relative exposure was compared to post-shift urine levels of a B[a]P metabolite, 1-hydroxypyrene
      (1-OH-Py). Collected and purified PBMCs were incubated with Annexin-V and PI prior to analysis
      by flow cytometry; early apoptotic cells were considered to be Annexin V+/PI-, while late apoptotic
      cells were considered Annexin V+/PI+.
            All apoptosis data was displayed graphically, and in all groupings early:late apoptotic
      PBMCs occurred at an approximate 2:1 frequency. PBMC apoptosis was similar in each of the three
      coke oven worker groups, which were all statistically significantly higher than referents
      (approximately 2-fold) for both early and late apoptosis. While self-reported smoking incidence
      varied significantly among the 4 worker groups, stratification by smoking years or smoking index
      did not reveal any significant association with PBMC apoptosis. Multiple linear stepwise regression
      analysis suggested that urine 1-OH-Py levels and years of coke oven operation were positively
      associated with increased early and late  PMBC apoptosis (Table B-5), and that years of ethanol
      consumption was negatively associated with only early apoptosis. These associations were tested
      by stratifying workers into three groups by urinary 1-OH-Py levels or coke oven operation years,
      and in both cases, the groups with the highest urinary metabolite levels or longest oven operating
      experience had statistically significantly higher levels of both early and late apoptotic PBMCs, vs.
      the lowest or shortest duration groups, respectively. Likewise, when sorted into groups based
      upon years of ethanol consumption, the highest ethanol "years of consumption" group had
      statistically significantly lower early apoptosis rates when compared to the lowest ethanol
      consuming group.

            Table B-5. Background information on Chinese coke oven workers or
            warehouse controls  exposed to benzo[a]pyrene in the workplace
Effect measured
Number of subjects
Exposure Group (ng/m3; mean ± SD)
10.2 ±7.6
37
19.5 ± 13.2
34
185.9 ± 38.6
48
1623.5 ± 435.8
47
P value
Background information (mean ± SD or %)
Age (yr)
Working years (yr)
Smoking
Drinking
37. 16 ±6.00
17.35 ±7.19
62.2
24.3
39.09 ±5.53
18.58 ±7.23
64.7
41.2
36.98 ±6.40
16.78 ±6.90
83.3
39.6
37.34 ±6.78
17.26 ± 7.44
53.2
44.7
0.451
0.742
0.017
0.259
Urine B[a]P metabolite (u.mol/mol Cr; mean ± SD)
1-OH-Py
2.78 ±1.04
3. 22 ±0.81*
3.51 ±0.55*
3. 66 ±0.58*
0.000
       * p < 0.05 significantly different from control mean
       Source: Zhang et al. (2012).
23
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-27         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    Cancer-related Endpoints

 2    Benzo[a]pyrene-Induced Cytogenetic Damage
 3          Many studies measure cytogenetic damage as biomarkers of early biological effects which
 4    also reflect exposure to genotoxic chemicals. Standard cytogenetic end points include
 5    chromosomal aberration (CA), sister chromatid exchange (SCE), micronucleus (MN) formation,
 6    hypoxanthine guanine phosphoribosyl transferase (hprt) mutation frequency, and glycophorin A
 7    mutation frequency (Gyorffy et al., 2008). These biomarkers are often incorporated in multi-
 8    endpoint studies with other biomarkers of exposure.  Because they indicate related but different
 9    endpoints, there is often a lack of correlation between the different categories of biomarkers.
10          Merlo et al. (1997) evaluated DNA adduct formation (measured by [32P]-postlabelling) and
11    MN in WBCs of 94 traffic policemen versus 52 residents from the metropolitan area of Genoa, Italy.
12    All study subjects wore personal air samplers for 5 hours of one work shift, and levels of
13    benzo[a]pyrene and other PAHs were measured. Policemen were exposed to 4.55 ng
14    benzo[a]pyrene/m3 air, compared with urban residents who were exposed to 0.15 ng/m3. DNA
15    adduct levels in policemen were 35% higher than in urban residents (p = 0.007), butMN in urban
16    residents were 20% higher than in policemen (p = 0.02). Linear regressions of DNA adducts and
17    MN incidence, respectively, versus benzo[a]pyrene exposure levels did not reveal significant
18    correlations.
19          Perera and coworkers assessed DNA damage in Finnish iron foundry workers in two
20    separate studies and using three methodologies. Based on results from personal sampling and
21    stationary monitoring in both studies, three levels of benzo[a]pyrene air concentrations were
22    defined:  low (<5 ng/m3 benzo[a]pyrene), medium (5-12 ng/m3), and high (>12 ng/m3) (Perera et
23    al., 1994,1993). In the first study, involving 48 workers, several biomarkers were analyzed for
24    dose-response and interindividual variability (Perera et al., 1993). PAH-DNA adducts were
25    determined in WBCs using an immunoassay as described in Section 4.1.2.2.1 and enzyme-linked
26    immunosorbent assay with fluorescence detection. Mutations at the hprt locus were also measured
27    in WBC DNA. The latter assay is based on the fact that each cell contains only one copy of the hprt
28    gene, which is located on the X-chromosome. While male cells have only one X-chromosome,
29    female cells inactivate one of the two X-chromosomes at random. The  gene is highly sensitive to
30    mutations such that in the event of a crucial mutation in the gene, enzyme activity disappears
31    completely from the cell. In addition, mutations at the glycophorin A gene locus were measured in
32    red blood cells (RBCs).  The glycophorin A mutation frequency was not correlated with either
33    benzo[a]pyrene exposure or PAH-DNA adduct formation. However, both PAH-DNA adductlevels
34    and hprt mutation frequency increased with increasing benzo[a]pyrene exposure. In addition,
35    there was a highly significant correlation between incidence of hprt mutations and PAH-DNA
36    adduct levels (p = 0.004).
37          In a second study, Perera et al. (1994) surveyed 64 iron foundry workers with assessments
38    conducted in 2 successive years;  24 of the workers provided blood samples in both years.  Exposure

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-28        DRAFT—DO NOT CITE  OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    to benzo[a]pyrene, collected by personal and area sampling in the first year of the study, ranged
 2    from <5 to 60 ng/m3 and was estimated to have decreased by 40% in the second year. The levels of
 3    PAH-DNA adducts were roughly 50% lower in the 2nd year, presumably reflecting decreased
 4    exposure. The longer-lived hprt mutations were not as strongly influenced by the decreasing
 5    exposure to benzo[a]pyrene. Study subjects who did not have detectable levels of DNA adducts
 6    were excluded from the study.  As in the previous study, a strong correlation between DNA adduct
 7    levels and incidence of hprt mutations was observed (Perera etal., 1993).
 8          Kalina et al. (1998) studied several cytogenetic markers in 64 coke oven workers and
 9    34 controls employed at other locations within the same plant. Airborne benzo[a]pyrene and seven
10    other carcinogenic PAHs were collected by personal air samplers, which showed ambient
11    benzo[a]pyrene concentrations ranging widely from 0.002 to 50 ug/m3 in coke oven workers and
12    from 0.002 to 0.063 ug/m3 in controls.  CAs, SCEs, high-frequency cells (HFCs), and SCE
13    heterogeneity index were  all significantly increased with benzo[a]pyrene exposure. Except for
14    increases in HFCs, no effect of smoking was observed. Consistent with studies of PAH-DNA adduct
15    formation, reduced cytogenetic response at high exposure levels produced a nonlinear dose-
16    response relationship. The authors also evaluated the potential influence of polymorphisms in
17    enzymes involved in the metabolism of benzo[a]pyrene. Glutathione S-transferase Ml (GSTM-1)
18    and N-acetyl transferase-2 polymorphisms were studied and no evidence of the two gene
19    polymorphisms having any influence on the incidence of cytogenetic damage was found.
20          Motykiewicz et al.  (1998) conducted a similar study of genotoxicity associated with
21    benzo[a]pyrene exposure  in 67 female  residents of a highly polluted industrial urban area of Upper
22    Silesia, Poland, and compared the results to those obtained from 72 female residents of another
23    urban but less polluted area in the same province of Poland.  Urinary mutagenicity and 1-
24    hydroxypyrene levels, PAH-DNA adducts in oral mucosa cells (detected by immunoperoxidase
25    staining), SCEs, HFCs, CAs, bleomycin sensitivity, and GSTM-1 and CYP1A1 polymorphisms in blood
26    lymphocytes were investigated. High volume air samplers and gas chromatography were used to
27    quantify ambient benzo[a]pyrene levels, which were 3.7 ng/m3 in the polluted area and 0.6 ng/m3
28    in the control area during  the summer. During winter, levels rose to 43.4 and 7.2 ng/m3 in the two
29    areas, respectively. The cytogenetic biomarkers (CA and SCE/HFC), urinary mutagenicity, and
30    urinary 1-hydroxypyrene  excretion were significantly increased in females from the polluted area,
31    and differences appeared to be more pronounced during winter time.  PAH-DNA adduct levels were
32    significantly increased in the study population, when compared to the controls, only in the winter
33    season. No difference in sensitivity to bleomycin-induced lymphocyte chromatid breaks was seen
34    between the two populations. As with the study by Kalina et al. (1998), genetic polymorphisms
3 5    assumed to affect the metabolic transformation of benzo [a]pyrene were not associated with any
36    difference in the incidence of DNA damage.
37          In a study of Thai school boys in urban (Bangkok) and rural areas, bulky (including but not
38    limited to BPDE-type) DNA adduct levels were measured in lymphocytes along with DNA SSBs,

               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-29         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    using the comet assay, and DNA repair capacity (Tuntawiroon et al., 2007). Ambient air and
 2    personal breathing zone measurements indicated that Bangkok school children experienced
 3    significantly higher exposures to benzo[a]pyrene and total PAHs.  A significantly higher level of
 4    SSBs (tail length 1.93 ± 0.09 versus 1.28 ± 0.12 ^m, +51%; p < 0.001) was observed in Bangkok
 5    school children when compared with rural children, and this parameter was significantly
 6    associated with DNA adduct levels. A significantly reduced DNA repair capacity (0.45 ±0.01 versus
 7    0.26 ± 0.01 y-radiation-induced deletions per metaphase, -42%; p < 0.001) was also observed in the
 8    city school children, again significantly associated with DNA adduct levels. It was not evident why
 9    higher environmental PAH exposure would be associated with lowered DNA repair capacity.
10    However, because the personal breathing zone PAH levels and DNA adduct levels were not
11    associated with each other, it is conceivable that the city school children had a priori lower DNA
12    repair capacities that contributed significantly to the high adduct levels. The authors considered
13    genetic differences between the two study populations as a possible reason for this observation.

14    Epidemiologic Findings in Humans
15           The association between human cancer and contact with PAH-containing substances, such
16    as soot, coal tar, and pitch, has been widely recognized since the early 1900s (Bostrom et al., 2002).
17    Although numerous epidemiology studies establish an unequivocal association between PAH
18    exposure and human cancer, defining the causative role for benzo[a]pyrene and other specific PAHs
19    remains a challenge. In essentially all reported studies, either the benzo[a]pyrene exposure and/or
20    internal dose are not known, or the benzo[a]pyrene carcinogenic  effect cannot be distinguished
21    from the effects of other PAH and non-PAH carcinogens.  Nevertheless, three types of investigations
22    provide support for the involvement of benzo[a]pyrene in some human cancers: molecular
23    epidemiology studies; population- and hospital-based case-control studies; and occupational cohort
24    studies. In some cohort studies, benzo[a]pyrene exposure  concentrations were measured and thus
25    provide a means to link exposure intensity with observed cancer rates.  In case-control studies, by
26    their nature, benzo[a]pyrene and total PAH doses can only be estimated.

27    Molecular Epidemiology and Case-Control Cancer Studies
28           Defective DNA repair capacity leading to genomic instability and, ultimately, increased
29    cancer risk is well documented (Wu et al., 2007, 2005). Moreover, sensitivity to mutagen-induced
30    DNA damage is highly heritable and thus represents an important factor that determines individual
31    cancer susceptibility. Based on studies comparing monozygotic and dizygotic twins, the genetic
32    contribution to BPDE mutagenic sensitivity was estimated  to be 48.0% (Wu et al., 2007). BPDE has
33    been used as an etiologically relevant mutagen in case-control studies to examine the association
34    between elevated lung and bladder cancer risk and individual sensitivity to BPDE-induced DNA
35    damage. Mutagen sensitivity is determined by quantifying chromatid breaks or DNA adducts in
36    phytohemagglutinin-stimulated peripheral blood lymphocytes as an indirect measure of DNA
37    repair capacity.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-30         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1          In a hospital-based, case-control study involving 221 lung cancer cases and 229 healthy
 2    controls, DNA adducts were measured in stimulated peripheral blood lymphocytes after incubation
 3    with BPDE in vitro (Li et al., 2001). Lung cancer cases showed consistent statistically significant
 4    elevations in induced BPDE-DNA adducts in lymphoctes, compared with controls, regardless of
 5    subgroup by age, sex, ethnicity, smoking history, weight loss, or family history of cancer. The
 6    lymphocyte BPDE-induced DNA adduct levels, when grouped by quartile using the levels in controls
 7    as cutoff points, were significantly dose-related with lung cancer risk (odds ratios [ORs] 1.11,1.62,
 8    and 3.23; trend test, p < 0.001).  In a related hospital-based, case-control study involving 155 lung
 9    cancer patients and 153 healthy controls, stimulated peripheral blood lymphocytes were exposed
10    to BPDE in vitro  (Wu et al., 2005). DNA damage/repair was evaluated in lymphocytes using the
11    comet assay, and impacts on cell cycle checkpoints were measured using a fluorescence-activated
12    cell-sorting method. The lung cancer cases exhibited significantly higher levels of BPDE-induced
13    DNA damage than the controls (p < 0.001), with lung cancer risk positively associated with
14    increasing levels of lymphocyte DNA damage when grouped in quartiles (trend test, p < 0.001). In
15    addition, lung cancer patients demonstrated significantly shorter cell cycle delays in response to
16    BPDE exposure to lymphocytes, which correlated with increased DNA damage.
17          Sensitivity to BPDE-induced DNA damage in bladder cancer patients supports the results
18    observed in lung cancer cases. In a hospital-based, case-control study involving 203 bladder cancer
19    patients and 198 healthy controls, BPDE-induced DNA damage was specifically evaluated at the
20    chromosome 9p21 locus in stimulated peripheral blood lymphocytes (Gu et al., 2008). Deletions of
21    9p21, which includes critical components of cell cycle control pathways, are associated with a
22    variety of cancers.  After adjusting for age, sex, ethnicity, and smoking status, individuals with high
23    BPDE-induced damage at 9p21 were significantly associated with increased bladder cancer risk
24    (OR 5.28; 95% confidence interval [CI] 3.26-8.59). Categorization of patients into tertiles for BPDE
25    sensitivity relative to controls demonstrated a dose-related association between BPDE-induced
26    9p21 damage and bladder cancer risk. Collectively, the results of molecular epidemiology studies
27    with lung and bladder cancer patients indicate that individuals with a defective ability to repair
28    BPDE-DNA adducts are at increased risk for cancer and, moreover, that specific genes linked to
29    tumorigenesis pathways may be molecular targets for benzo[a]pyrene and other carcinogens.
30          Due to the importance of the diet as a benzo[a]pyrene exposure source, several population-
31    and hospital-based, case-control studies have investigated the implied association between dietary
32    intake of benzo[a]pyrene and risk for several tumor types. In a study involving 193 pancreatic
33    cancer cases and 674 controls (Anderson et al., 2005), another involving 626 pancreatic cancer
34    cases and 530 controls (Li etal., 2007), and a third involving 146 colorectal adenoma cases and 228
35    controls (Sinha et al., 2005), dietary intake of benzo[a]pyrene was estimated using food frequency
36    questionnaires. In all studies, the primary focus was on estimated intake of benzo[a]pyrene (and
37    other carcinogens) derived from cooked meat  Overall, cases when compared with controls had
3 8    higher intakes of benzo [ajpyrene and other food carcinogens, leading to the conclusion that

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-31         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    benzo[a]pyrene plays a role in the etiology of these tumors in humans. In a supportive follow-up
 2    case-control study of colorectal adenomas, levels of leukocyte PAH-DNA adducts were significantly
 3    higher in cases when compared with controls (p = 0.02), using a method that recognizes BPDE and
 4    several other PAHs bound to DNA (Gunter et al., 2007).

 5    Cohort Cancer Studies
 6          Epidemiologic studies of workers in PAH-related occupations indicate  increased human
 7    cancer risks associated with iron and steel production, roofing, carbon black production, and
 8    exposure to diesel exhaust (Bosetti etal., 2007). Exposure to benzo[a]pyrene is only one of
 9    numerous contributors to the cancer risk from complex PAH-containing mixtures that occur in the
10    workplace. Although some occupational cohort studies report measured or estimated inhalation
11    exposure concentrations for benzo[a]pyrene, none report biomarkers of internal benzo[a]pyrene
12    dose in study subjects (reviewed in Bosetti et al., 2007; Armstrong et al., 2004). Several of these
13    cohort studies (summarized below) demonstrate a positive exposure-response relationship with
14    cumulative PAH exposure using benzo[a]pyrene—or a proxy such as benzene-soluble matter (BSM)
15    that can be converted to benzo[a]pyrene—as an indicator substance. These studies provide insight
16    and support for the causative role of benzo[a]pyrene in human cancer.
17
18    Cancer incidence in aluminum and electrode production plants
19          Exposure to benzo[a]pyrene and BSM in aluminum smelter workers is strongly  associated
20    with bladder cancer and weakly associated with lung cancer (Boffetta et al., 1997; Tremblay et al.,
21    1995; Armstrong et al., 1994; Gibbs, 1985; Theriault et al., 1984).  In an analysis of pooled data from
22    nine cohorts of aluminum production workers,  688 respiratory tract cancer cases were  observed
23    versus 674.1 expected (pooled RR 1.03; CI 0.96-1.11) (Bosetti et al., 2007). A total of 196 bladder
24    cancer cases were observed in eight of the cohorts, compared with 155.7 expected (pooled relative
25    risk [RR] 1.29; CI 1.12-1.49). Based on estimated airborne benzo[a]pyrene exposures from a meta-
26    analysis of eight cohort studies, the predicted lung cancer RR per 100 ug/m3-years of cumulative
27    benzo[a]pyrene exposure was 1.16 (95% CI 1.05-1.28) (Armstrong etal.,  2004).
28          Spinelli et al. (2006) reported a 14-year update to a previously published historical cohort
29    study (Spinelli etal., 1991) of Canadian aluminum reduction plant workers. The results confirmed
30    and extended the findings from the earlier epidemiology study. The study surveyed a total of 6,423
31    workers with >3 years of employment at an aluminum reduction plant in British Columbia, Canada,
32    between the years 1954 and 1997, and evaluated all types of cancers. The focus  was on cumulative
33    exposure to coal tar pitch volatiles, measured as BSM and as benzo[a]pyrene. Benzo[a]pyrene
34    exposure categories were determined from the range of predicted exposures over time  from
35    statistical exposure models. There were 662 cancer cases, of which approximately 98% had
36    confirmed diagnoses. The overall cancer mortality rate (standardized mortality ratio 0.97; CI 0.87-
37    1.08) and cancer incidence rate (standardized incidence ratio [SIR] 1.00; CI 0.92-1.08) were not
3 8    different from that of the British Columbia general population. However, this study identified
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-32         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    significantly increased incidence rates for cancers of the bladder (SIR 1.80; CI 1.45-2.21) and the
 2    stomach (SIR 1.46; CI 1.01-2.04). The lung cancer incidence rate was only slightly higher than
 3    expected (SIR 1.10; CI 0.93-1.30).  Significant dose-response associations with cumulative
 4    benzo[a]pyrene exposure were seen for bladder cancer (p trend < 0.001), stomach cancer (p trend
 5    < 0.05), lung cancer (p trend < 0.001), non-Hodgkin lymphoma (p trend < 0.001), and kidney cancer
 6    (p trend < 0.01), although the overall incidence rates for the latter three cancer types were not
 7    significantly elevated versus the general population. Similar cancer risk results were obtained
 8    using BSM as the exposure measure; the cumulative benzo[a]pyrene and BSM exposures were
 9    highly correlated (r = 0.94).
10          In several occupational cohort studies of workers in Norwegian aluminum production
11    plants, personal and stationary airborne PAH measurements were performed.
12          In a study covering 11,103 workers and 272,554 person x years of PAH exposure, cancer
13    incidence was evaluated in six Norwegian aluminum smelters (Romundstad et al., 2000a, b).
14    Reported estimates of PAH exposure concentrations reached a maximum of 3,400 [J.g/m3 PAH
15    (680 [ig/m3 benzo[a]pyrene). The overall number of cancers observed in this study did not differ
16    significantly from control values (SIR 1.03; CI 1.0-1.1). The data from this study  showed
17    significantly increased incidences for cancer of the bladder (SIR 1.3; CI 1.1-1.5) and elevated, but
18    not significant, SIRs for larynx (SIR 1.3; CI 0.8-1.9), thyroid (SIR 1.4; CI 0.7-2.5), and multiple
19    myeloma (SIR 1.4; CI 0.9-1.9). Incidence rates for bladder, lung, pancreas, and kidney cancer (the
20    latter three  with SIRs close to unity) were subjected to a cumulative exposure-response analysis.
21    The incidence rate for bladder cancer showed a trend with increasing cumulative exposure and
22    with increasing lag times (up to 3 0 years) at the highest exposure level. The incidence of both lung
23    and bladder cancers was greatly increased in smokers. The authors reported that using local
24    county rates rather than national cancer incidence rates as controls increased the SIR for lung
25    cancer (SIR 1.4; CI 1.2-1.6) to a statistically significant level.
26
27    Cancer incidence in coke oven, coal gasification, and iron and steel foundry workers
28          An increased risk of death from lung and bladder cancer is reported in some studies
29    involving coke oven, coal gasification, and iron and steel foundry workers (Bostrom et al., 2002;
30    Boffetta et al., 1997). An especially consistent risk of lung cancer across occupations is noted when
31    cumulative exposure is taken into consideration (e.g., RR of 1.16 per 100 unity-years for aluminum
32    smelter workers, 1.17 for coke oven workers, and 1.15 for coal gasification workers). In an analysis
33    of pooled data from 10 cohorts of coke production workers, 762 lung cancer cases were observed
34    versus 512.1 expected (pooled RR 1.58; CI 1.47-1.69) (Bosetti et al., 2007). Significant variations in
35    risk estimates among the studies were reported, particularly in the large cohorts (RRs of 1.1,1.2,
36    2.0, and 2.6). There was no evidence for increased bladder cancer risk in the coke production
37    workers. Based on estimated airborne benzo[a]pyrene exposures from a meta-analysis of  10
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 B-33         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    cohort studies, the predicted lung cancer RR per 100 ug/m3-years of cumulative benzo[a]pyrene
 2    exposure was 1.17 (95% CI 1.12-1.22) (Armstrong et al., 2004).
 3          A meta-analysis of data from five cohorts of gasification workers reported 251 deaths from
 4    respiratory tract cancer, compared with 104.7 expected (pooled RR 2.58; 95% CI 2.28-2.92)
 5    (Bosetti et al., 2007). Pooled data from three of the cohorts indicated 18 deaths from urinary tract
 6    cancers, versus 6.0 expected (pooled RR 3.27; 95% CI 2.06-5.19). Based on estimated airborne
 7    benzo[a]pyrene exposures from a meta-analysis of four gas worker cohort studies, the predicted
 8    lung cancer RRper 100 ug/m3-years of cumulative benzo[a]pyrene exposure was 1.15 (95% CI
 9    1.11-1.20) (Armstrong et al., 2004).
10          Increased risks were reported in iron and steel foundry workers for cancers of the
11    respiratory tract, bladder, and kidney. In an analysis of pooled data from 10 cohorts,
12    1,004 respiratory tract cancer cases were observed versus 726.0 expected (pooled RR 1.40;
13    CI 1.31-1.49) (Bosetti etal., 2007). A total of 99 bladder cancer cases were observed in seven of the
14    cohorts, compared with 83.0 expected (pooled RR 1.29; CI 1.06-1.57). For kidney cancer, 40 cases
15    were observed compared with 31.0 expected based on four studies (pooled RR 1.30; 95% CI 0.95-
16    1.77).
17          Xu et al. (1996) conducted a nested case-control study, surveying the cancer incidence
18    among 196,993 active or retired workers from the Anshan Chinese iron and steel production
19    complex. A large number of historical benzo[a]pyrene  measurements (1956-1995) were available.
20    The study included 610 cases of lung cancer and 292 cases of stomach cancer, with 959 age- and
21    gender-matched controls from the workforce. After adjusting for nonoccupational risk factors such
22    as smoking and diet, significantly elevated risks for lung cancer and stomach cancer were identified
23    for subjects employed for >15  years, with ORs varying among job categories. For either type of
24    cancer, highest risks were seen among coke oven workers: lung cancer, OR = 3.4 (CI 1.4-8.5);
25    stomach cancer, OR = 5.4 (CI 1.8-16.0).
26          There were significant trends for long-term, cumulative benzo[a]pyrene exposure versus
27    lung cancer (p = 0.004) or stomach cancer (p = 0.016) incidence.  For cumulative total
28    benzo[a]pyrene exposures of <0.84, 0.85-1.96,1.97-3.2, and >3.2, respectively, the ORs for lung
29    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), respectively. For
30    cumulative total benzo[a]pyrene exposures of <0.84, 0.85-1.96,1.97-3.2, and >3.2, the ORs for
31    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),
32    respectively.  However, the investigators noted that additional workplace air contaminants were
33    measured, which might have influenced the outcome. Of these, asbestos, silica, quartz, and iron
34    oxide-containing dusts may have been confounders.  For lung cancers, cumulative exposures to
35    total dust and silica dust both showed significant dose-response trends (p = 0.001 and 0.007,
36    respectively), while for stomach cancer,  only cumulative total dust exposure showed a marginally
37    significant trend (p = 0.061). For cumulative total dust exposures of <69, 69-279, 280-882, and
38    >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

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-34        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    1.9 (CI 1.3-2.5), respectively. For cumulative silica dust exposures of <3.7, 3.7-10.39,10.4-27.71,
 2    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),
 3    and 1.8 (CI 1.2-2.5), respectively. For cumulative total dust exposures of <69, 69-279, 280-882,
 4    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
 5    1.6 (CI 1.1-2.5), respectively.
 6           Exposure-response data from studies of coke oven workers in the United States have often
 7    been used to derive quantitative risk estimates for PAH mixtures, and for benzo[a]pyrene as an
 8    indicator substance (Bostrom et al., 2002). However, there are numerous studies of coke oven
 9    worker cohorts that do notprovide estimates of benzo[a]pyrene exposure. An overview of the
10    results of these and other studies can be obtained from the review of Boffetta et al. (1997).
11
12    Cancer incidence in asphalt workers and roofers
13           These groups encompass different types of work (asphalt paving versus roofing) and also
14    different types of historical exposure that have changed from using PAH-rich coal tar pitch to the
15    use of bitumen or asphalt, both of which are rather low in PAHs due to their source (crude oil
16    refinery) and a special purification process. Increased risks for lung cancer were reported in large
17    cohorts of asphalt workers and roofers; evidence for increased bladder cancer risk is weak
18    (Burstynetal., 2007; Partanen and Boffetta, 1994; Chiazze et al., 1991; Hansen, 1991,1989;
19    Hammond etal., 1976). In an analysis of pooled data from two cohorts of asphalt workers, 822 lung
20    cancer cases were observed versus 730.7 expected (pooled RR 1.14; 95% CI 1.07-1.22)  (Bosetti et
21    al., 2007). In two cohorts of roofers, analysis of pooled data indicated that 138 lung cancer cases
22    were observed, compared with 91.9 expected (pooled RR 1.51; 95% CI 1.28-1.78) (Bosetti etal.,
23    2007).
24
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-35        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1   ANIMAL BIOASSAYS

 2   Oral Bioassays

 3   Subchronic Studies
 4          De Jong et al. (1999) treated male Wistar rats (eight/dose group) with benzo[a]pyrene
 5   (98.6% purity) dissolved in soybean oil by gavage 5 days/week for 35 days at doses of 0, 3,10, 30,
 6   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
 7   exposure period, rats were necropsied, organ weights were determined, and major organs and
 8   tissues were prepared for histological examination (adrenals, brain, bone marrow, colon, caecum,
 9   jejunum, heart, kidney, liver, lung, lymph nodes, esophagus, pituitary, spleen,  stomach, testis, and
10   thymus). Blood was collected for examination of hematological endpoints, but there was no
11   indication that serum biochemical parameters were analyzed. Immune parameters included
12   determinations of serum immunoglobulin (Ig) levels (IgG, IgM, IgE, and IgA), relative spleen cell
13   distribution, and spontaneous cytotoxicity of spleen cell populations determined in a natural-killer
14   (NK) cell assay.
15          Body weight gain was decreased beginning at week 2 at the high dose of 90 mg/kg-day;
16   there was no effect at lower doses (De Jong et al., 1999). Hematology revealed a dose-related
17   decrease in RBC count, hemoglobin, and hematocrit at >10 mg/kg-day (Table B-6).  A minimal but
18   significant increase in mean cell volume and a decrease in mean cell hemoglobin concentration
19   were noted at 90 mg/kg-day, and may indicate dose-related toxicity for the RBCs and/or RBC
20   precursors in the bone marrow. A decrease in WBCs, attributed to a decrease in the number of
21   lymphocytes (approximately 50%) and eosinophils (approximately 90%), was observed at
22   90 mg/kg-day; however, there was no effect on the number of neutrophils or  monocytes. A
23   decrease in the cell number in the bone marrow observed in the 90 mg/kg-day dose group was
24   consistent with the observed decrease in the  RBC and WBC counts at this dose level. In the
25   90 mg/kg-day dose group, brain, heart, kidney, and lymph node weights were decreased and liver
26   weight was increased (Table B-6).  Decreases in heart weight at 3 mg/kg-day  and in kidney weight
27   at 3 and 30 mg/kg-day were also observed, but these changes did not show dose-dependent
28   responses. Dose-related decreases in thymus weight were statistically significant at >10 mg/kg-
29   day (Table B-6).

30          Table B-6.  Exposure-related effects in male Wistar rats exposed to
31          benzo[a]pyrene by gavage 5 days/week for 5 weeks

Effect
Hematologic effects
(mean ± SD; n = 7-8)
WBCs (109/L)
RBCs (109/L)
Dose (mg/kg-d)
0


14.96 ± 1.9
8.7 ±0.2
3


13. 84 ±3.0
8.6 ±0.2
10


13.69 ±1.8a
8.3 ±0.2
30


13.58 ±2.9a
7.8 ±0.4
90


8.53±l.la
7.1±0.4a
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-36         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

Effect
Hemoglobin (mmol/L)
Hematocrit (L/L)
Serum Ig levels
(mean ± SD; n = 7-8)
IgM
IgG
IgA
IgE
Cellularity (mean ± SD; n = 7-8)
Spleen (cell number x 107)
Bone marrow (G/L)
Spleen cell distribution (%)
B cells
T cells
Th cells
Ts cells
Body (g) and organ (mg) weights
(means; n = 7-8)
Body weight
Brain
Heart
Kidney
Liver
Thymus
Spleen
Mandibular lymph nodes
Mesenteric lymph nodes
Popliteal lymph nodes
Thymus cortex surface area
(% of total surface area of thymus;
mean ± SD; n = 6-8)
Dose (mg/kg-d)
0
10.5 ±0.2
0.5 ±0.01


100 ± 13
100 ± 40
100 ± 28
100 ± 65

59 ±15
31 ±7

39±4
40 ±9
23 ±7
24 ±5


305
1,858
1,030
1,986
10,565
517 ±47
551
152
165
19
77.9 ±3. 8


3
10.4 ±0.3
0.5 ±0.01


87 ±16
141 ± 106
73 ±29
50 ±20

71 ±14
36 ±5

36 ±2
48 ±12
26 ±7
26 ±6


282a
1,864
934a
l,761a
9,567
472 ± 90
590
123
148
18
74.4 ±2. 2


10
9.8±0.2a
0.47±0.01a


86 ±31
104 ± 28
78 ±67
228 ±351

59 ±13
31 ±8

34±3a
40 ±9
24 ±5
24 ±7


300
1,859
1,000
1,899
11,250
438 ± 64a
538
160
130a
19
79. 2 ±5. 9


30
9.5 ± 0.4a
0.46 ± 0.02a


67 ± 16a
106 ± 19
72 ±22
145 ± 176

63 ±10
27 ±8

32±4a
36 ±2
22 ±4
19 ±2


293
1,784
967
l,790a
11,118
388 ± 71a
596
141
158
17
75.8 ±4.0


90
8.6±0.6a
0.43±0.02a


81 ±26
99 ±29
39 ± 19a
75 ±55

41 ± 10a
19±4a

23±4a
44 ±6
26 ±4
27 ±5


250a
l,743a
863a
l,626a
12,107a
198 ± 65a
505
89a
107a
10a
68.9 ± 5. 2a


1
2
3
4
5
6
7
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).

       Statistically significant reductions were also observed in the relative cortex surface area of
the thymus and thymic medullar weight at 90 mg/kg-day, but there was no difference in cell
proliferation between treated and control animals using the proliferating cell nuclear antigen
(PCNA) technique.  Changes in the following immune parameters were noted: dose-related and
statistically significant decrease in the relative number of B cells in the spleen at 10 (13%),
30 (18%), and 90 mg/kg-day (41%); significant decreases in  absolute number of cells harvested in
the spleen (31%), in the number of B cells in the spleen (61%), and NK cell activity in the spleen
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-37        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    (E:T ratio was 40.9 ± 28.4% that of the controls) at 90 mg/kg-day; and a decrease in serum IgM
 2    (33%) and IgA (61%) in rats treated with 30 and 90 mg/kg-day, respectively. The decrease in the
 3    spleen cell count was attributed by the study authors to the decreased B cells and suggested a
 4    possible selective toxicity of benzo[a]pyrene to B cell precursors in the bone marrow. The study
 5    authors considered the decrease in IgA and IgM to be due to impaired production of antibodies,
 6    suggesting a role of thymus toxicity in the decreased (T-cell dependent) antibody production.  In
 7    addition to the effects on the thymus and spleen, histopathologic examination revealed treatment-
 8    related lesions only in the liver and forestomach at the two highest dose levels, but the incidence
 9    data for these lesions were not reported by De Jong et al. (1999). Increased incidence for
10    forestomach basal cell hyperplasia (p < 0.05 by Fisher's exact test) was reported at 30 and
11    90 mg/kg-day, and increased incidence for oval cell hyperplasia in the liver was reported at
12    90 mg/kg-day (p < 0.01, Fisher's exact test). The results indicate that 3 mg/kg-day was a no-
13    observed-adverse-effect level (NOAEL) for effects on hematological parameters (decreased RBC
14    count, hemoglobin, and hematocrit) and immune parameters (decreased thymus weight and
15    percent of B cells in the spleen) noted in Wistar rats at 10 mg/kg-day (the lowest-observed-
16    adverse-effect level [LOAEL]) and above. Lesions of the liver (oval cell hyperplasia)  and
17    forestomach (basal cell hyperplasia) occurred at doses >30 mg/kg-day.
18          Knuckles etal. (2001) exposed male and female F344 rats (20/sex/dose group) to
19    benzo[a]pyrene (98% purity) at doses of 0, 5, 50, or 100 mg/kg-day in the diet for 90 days. Food
20    consumption and body weight were monitored, and the concentration of benzo[a]pyrene in the
21    food was adjusted every 3-4 days to maintain the target dose. The authors indicated that the actual
22    intake of benzo[a]pyrene by the rats was within 10% of the calculated intake, and the nominal
23    doses were not corrected to actual doses. Hematology and serum chemistry parameters were
24    evaluated. Urinalysis was also performed. Animals were examined for gross pathology, and
25    histopathology was performed on selected organs (stomach, liver, kidney, testes, and ovaries).
26    Statistically significant decreases in RBC counts and hematocrit level  (decreases  as much as 10 and
27    12%, respectively) were observed in males at doses >50 mg/kg-day and in females at 100 mg/kg-
28    day. A maximum 12% decrease (statistically significant) in hemoglobin level was noted in both
29    sexes at 100 mg/kg-day. Blood chemistry analysis  showed a significant increase in blood urea
30    nitrogen (BUN) only in high-dose (100 mg/kg-day) males.  Histopathology examination revealed an
31    apparent increase in the incidence of abnormal tubular casts in the kidney in males at 5 mg/kg-day
32    (40%), 50 mg/kg-day (80%), and 100 mg/kg-day (100%), compared to 10% in the controls.  Only
33    10% of the females showed significant kidney tubular changes at the  two high-dose levels
34    compared to zero animals in the female control group. The casts were described as molds of distal
3 5    nephron lumen and were considered by the study authors to be indicative of renal dysfunction.
36    From this study, male F344 rats appeared to be affected more severely by benzo[a]pyrene
37    treatment than the female rats. However, the statistical significance of the kidney lesions are
38    unclear.  Several reporting gaps and inconsistencies regarding the reporting of kidney

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-38         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    abnormalities in Knuckles et al. (2001) make interpretation of the results difficult Results of
 2    histopathological kidney abnormalities (characterized primarily as kidney casts) were presented
 3    graphically and the data were not presented numerically in this report. No indication was given in
 4    the graph that any groups were statistically different than controls, although visual examination of
 5    the magnitude of response and error bars appears to indicate a fourfold increase in kidney casts in
 6    males compared to the control group (40 compared to 10%). The figure legend reported the data
 7    as "percentage incidence of abnormal kidney tissues" and reported values as mean ± SD. However,
 8    the text under the materials and methods section stated that Fisher's exact test was used for
 9    histopathological data, which would involve the pairwise comparison of incidence and not means.
10    There are additional internal inconsistencies in the data presented. The data appeared to indicate
11    that incidences for males were as follows: control, 10%; 5 mg/kg-day, 40%; 50 mg/kg-day, 80%;
12    and 100 mg/kg-day, 100%; however, these incidences are inconsistent with the size of the study
13    groups, which were reported as 6-8 animals per group. The study authors were contacted, but did
14    not respond to EPA's request for clarification of study design and/or results. Due to issues of data
15    reporting, aLOAEL could not be established for the increased incidence of kidney lesions. Based on
16    the statistically significant hematological effects including decreases in RBC counts, hematocrit, and
17    BUN, the NOAEL in males was 5 mg/kg-day and the LOAEL was 50 mg/kg-day, based on in F344
18    rats.  No exposure-related histological lesions were identified in the stomach, liver, testes, or
19    ovaries in this study.
20          In a range-finding study, Wistar (specific pathogen-free [SPF] Riv:TOX) rats (10/sex/dose
21    group) were administered benzo[a]pyrene (97.7% purity) dissolved in soybean oil by gavage at
22    dose levels of 0,1.5, 5,15, or 50 mg/kg body weight-day, 5 days/week for 5 weeks (Kroese et al.,
23    2001). Behavior, clinical symptoms, body weight, and food and water consumption were
24    monitored. None of the animals died during the treatment period. Animals were  sacrificed
25    24 hours after the last dose. Urine and blood were collected for standard urinalysis and
26    hematology and clinical chemistry evaluation. Liver enzyme induction was monitored based on
27    EROD activity in plasma. Animals were subjected to macroscopic examination, and organ weights
28    were recorded. The esophagus, stomach, duodenum, liver, kidneys, spleen, thymus, lung, and
29    mammary gland (females only) from the highest-dose and control animals were evaluated for
30    histopathology. Intermediate-dose groups were examined if abnormalities were observed in the
31    higher-dose groups.
32          A significant, but not dose-dependent, increase in food consumption in males at >1.5 mg/kg-
33    day and a decrease in food consumption in females at >5 mg/kg-day was observed (Kroese et al.,
34    2001). Water consumption was statistically significantly altered in males only: a decrease at 1.5, 5,
35    and 15 mg/kg-day and an increase at 50 mg/kg-day. Organ weights of lung, spleen, kidneys,
36    adrenals, and ovaries were not affected by treatment.  There was a dose-related, statistically
37    significant decrease in thymus weight in males at 15 and 20 mg/kg-day (decreased by 28 and 33%,
38    respectively) and a significant decrease in thymus weight in females at 50 mg/kg-day (decreased by

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-39         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1    17%) (Table B-7). In both sexes, liver weight was statistically significantly increased only at
 2    50 mg/kg-day by about 18% (Table B-7).

 3          Table B-7.  Exposure-related effects in Wistar rats exposed to benzo[a]-
 4          pyrene by gavage 5 days/week for 5 weeks

Organ
Liver weight (g; mean ± SD)
Males
Females
Thymus weight (mg; mean ± SD)
Males
Females
Basal cell hyperplasia of the
forestomach (incidence with slight
severity)
Males
Females
Dose (mg/kg-d)
0

6.10 ±0.26
4.28 ±0.11

471 ± 19
326 ±12



1/10
0/10
1.5

6.19 ±0.19
4.40 ± 0.73

434 ± 20
367 ± 23



1/10
1/10
5

6.13 ±0.10
4.37 ±0.11

418 ± 26
351 ±25



4/10
1/10
15

6.30 ±0.14
4.67 ±0.17

342 ± 20a
317 ±30



3/10
3/10a
50

7.20±0.18a
5.03±0.15a

317±21a
271 ± 16a



7/10
7/10a
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Significantly (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 5-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 etal., 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 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  etal. (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-day,
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-40         DRAFT— DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    5 days/week for 90 days. The rats were examined daily for behavior and clinical symptoms and by
 2    palpation. Food and water consumption, body weights, morbidity, and mortality were monitored.
 3    At the end of the exposure period, rats were subjected to macroscopic examination and organ
 4    weights were recorded.  Blood was collected for hematology and serum chemistry evaluation, and
 5    urine was collected for urinalysis. All gross abnormalities, particularly masses and lesions
 6    suspected of being tumors, were evaluated. The liver, stomach, esophagus, thymus, lung, spleen,
 7    and mesenteric lymph node were examined histopathologically. In addition, cell proliferation in
 8    forestomach epithelium was measured as the prevalence of S-phase epithelial cells displaying
 9    bromodeoxyuridine (BrdU) incorporation.
10          There were no obvious effects on behavior of the animals, and no difference was observed
11    in survival or food consumption between exposed animals and controls (Kroese et al., 2001).
12    Higher water consumption and slightly lower body weights than the controls were observed in
13    males but not females at the high dose of 30 mg/kg-day. Hematological investigations showed only
14    nonsignificant, small dose-related decreases in RBC count and hemoglobin level in both sexes.
15    Clinical chemistry evaluation did not show any treatment-related group differences or dose-
16    response relationships for alanine aminotransferase (ALT), serum aspartate transaminase (AST),
17    lactate dehydrogenase (LDH), or creatinine, but a small dose-related decrease in y-glutamyl
18    transferase (GGT) activity was observed in males only.  Urinalysis revealed an increase in urine
19    volume in males at 30 mg/kg-day, which was not dose related. At the highest dose, both sexes
20    showed increased levels of urinary creatinine and a dose-related increase in urinary protein.
21    However, no further investigation was conducted to determine the underlying mechanisms for
22    these changes. At necropsy, reddish to brown/gray discoloration of the mandibular lymph nodes
23    was consistently noted in most rats; occasional discoloration was also observed in other regional
24    lymph nodes (axillary). Statistically significant increases in liver weight were observed at 10 and
25    30 mg/kg-day in males (15 and 29%) and at 30 mg/kg-day in females (17%).  A decrease  in thymus
26    weight was seen in both  sexes at 30 mg/kg-day (17 and 33% decrease in females and males,
27    respectively, compared with controls) (Table B-8). At 10 mg/kg-day, thymus weight  in males was
28    decreased by 15%, but the decrease did not reach statistical significance.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-41         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1
 2
       Table B-8. Means ± SDa for liver and thymus weights in Wistar rats
       exposed to benzo[a]pyrene by gavage 5 days/week for 90 days
Organ
Liver weight (g)
Males
Females
Thymus weight (mg)
Males
Females
Dose (mg/kg-d)
0
7.49 ± 0.97
5.54 ±0.70
380 ± 60
320 ±60
3
8.00 ±0.85
5.42 ±0.76
380 ± 110
310 ±50
10
8.62 ± 1.30b
5.76 ±0.71
330 ± 60
300 ± 40
30
9.67±1.17b
6.48±0.78b
270 ± 40b
230±30b
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
aReported as SE, but judged to be SD (and confirmed by study authors).
Significantly (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; occasional early nodule development;
infiltration by inflammatory cells (mainly histiocytes); and capillary hyperemia, often in
combination with the previous changes (Kroese etal., 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, respectively, in female rats and 2/10, 0/10, 6/10, and 7/10, respectively, in
male rats. Nodular hyperplasia was noted in one animal of each 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-fourfold 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). Respective
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
resultof the gavage dosing rather than of benzo[a]pyrene treatment
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-42         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1          The target organs of benzo[a]pyrene toxicity in this 90-day dietary study of Wistar rats
 2    were the forestomach, thymus, and liver. The LOAEL for forestomach hyperplasia, decreased
 3    thymus weight, and thymus atrophy was 30 mg/kg-day and the NOAEL was 10 mg/kg-day.

 4    Chronic Studies and Cancer Bioassavs
 5          Kroese etal. (2001) exposed Wistar (Riv:TOX) rats (52/sex/dose group) to benzo[a]pyrene
 6    (98.6% purity) in soybean oil by gavage at nominal doses of 0, 3,10, or 30 mg/kg-day, 5 days/week,
 7    for 104 weeks. Mean achieved dose levels were 0, 2.9, 9.6, and 29 mg/kg-day. Additional rats
 8    (6/sex/group) were sacrificed after 4 and 5 months of exposure for analysis of DNA adduct
 9    formation in blood and major organs and tissues. The rats were 6 weeks old at the start of
10    exposure. The rats were examined daily for behavior and clinical symptoms and by palpation.
11    Food and water consumption, body weights, morbidity, and mortality were monitored during the
12    study.  Complete necropsy was performed on all animals that died during the course of the study,
13    were found moribund, or at terminal sacrifice (organ weight measurement was not mentioned in
14    the report by Kroese et al., 2001).  The organs and tissues collected and prepared for microscopic
15    examination included: brain, pituitary, heart, thyroid, salivary glands, lungs, stomach, oesophagus,
16    duodenum, jejunum, ileum, caecum, colon, rectum, thymus, kidneys, urinary bladder, spleen, lymph
17    nodes, liver pancreas, adrenals, sciatic nerve, nasal cavity, femur, skin including mammary tissue,
18    ovaries/uterus, and testis/accessory sex glands. Some of these tissues were examined only when
19    gross abnormalities were detected. All gross abnormalities, particularly masses and lesions that
20    appeared to be tumors, were also examined.
21          At 104 weeks, survival in the control group was 65% (males) and 50% (females), whereas
22    mortality in the 30 mg/kg-day dose group was 100% after about week 70.  At 80 weeks, survival
23    percentages were about 90, 85 and 75% in female rats in the 0, 3, and 10 mg/kg-day groups,
24    respectively; in males, respective survival percentages were ~95, 90, and 85% at 80 weeks.
25    Survival of 50% of animals occurred at 104,104, ~90, and 60 weeks for control through high-dose
26    females; for males, the respective times associated with 65% survival were 104,104,104, and ~60
27    weeks. The high mortality rate in high-dose rats was attributed to liver or  forestomach tumor
28    development, not to noncancer systemic effects. After 2 0 weeks, body weight was decreased
29    (compared with controls by >10%) in 30-mg/kg-day males, but not in females. This decrease was
30    accompanied by a decrease in food consumption. Body weights and food consumption were not
31    adversely affected in the  other dose groups compared to controls. In males, there was a dose-
32    dependent increase in water consumption starting at week 13, but benzo[a]pyrene treatment had
33    no significant effects on water consumption in females.
34          Tumors were  detected at significantly elevated incidences at several tissue sites in female
35    and male rats at doses >10 and >3 mg/kg-day, respectively (Table 4-5; Kroese etal., 2001).  The
36    tissue sites with the highest incidences of tumors were the liver (hepatocellular adenoma and
37    carcinoma) and forestomach (squamous cell papilloma and carcinoma) in both sexes (Table B-9).
38    The first liver tumors were detected in week 35 in high-dose male rats. Liver tumors were
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-43        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
 1   described as complex, with a considerable proportion (59/150 tumors) metastasizing to the lungs.
 2   At the highest dose level, 95% of rats with liver tumors had malignant carcinomas (95/100; Table
 3   B-9). Forestomach tumors were associated with the basal cell proliferation observed (without
 4   diffuse hyperplasia) in the forestomach of rats in the preliminary range-finding and 90-day
 5   exposure studies described previously in Section 4.2.1. At the highest dose level, 59% of rats with
 6   forestomach tumors had malignant carcinomas (60/102; Table B-9). Other tissue sites with
 7   distinctly elevated incidences of tumors in the 30 mg/kg-day dose group included the oral cavity
 8   (papilloma and squamous cell carcinoma [SCC]) in both sexes, and the jejunum (adenocarcinoma),
 9   kidney (cortical adenoma), and skin (basal cell adenoma and carcinoma) in male rats (Table B-9).
10   In addition, auditory canal tumors (carcinoma or squamous cell papilloma originating from pilo-
11   sebaceous units including the Zymbal's gland) were also detected in both sexes at 30 mg/kg-day,
12   but auditory canal tissue was not histologically examined in the lower dose groups and the controls
13   (Table B-9).  Gross examination revealed auditory canal tumors only in the high-dose group.

14          Table B-9. Incidences of exposure-related neoplasms in Wistar rats
15          treated by gavage with benzo[a]pyrene, 5 days/week, for 104  weeks


Site
Oral cavity
Papilloma
SCC
Basal cell adenoma
Sebaceous cell carcinoma
Oesophagus
Sarcoma undifferentiated
Rhabdomyosarcoma
Fibrosarcoma
Forestomach
Squamous cell papilloma
SCC
Liver
Hepatocellular adenoma
Hepatocellular carcinoma
Cholangiocarcinoma
Anaplastic carcinoma
Auditory canal
Benign tumor
Squamous cell papilloma
Carcinoma

Dose (mg/kg-d)
0
3
10
30a
Females'1

0/19
1/19
0/19
0/19

0/52
0/52
0/52

1/52
0/52

0/52
0/52
0/52
0/52

0/0
0/0
0/0

0/21
0/21
0/21
0/21

0/52
1/52
0/52

3/51
3/51

2/52
0/52
0/52
0/52

0/0
0/1
0/1

0/9
0/9
1/9
0/9

2/52
4/52
3/52

20/5 lc
10/5 lc

7/52c
32/52c
1/52
1/52

0/0
0/0
0/0

9/3 lc
9/3 lc
4/31
1/31

0/52
0/52
0/52

25/52c
25/52c

1/52
50/52C
0/52
0/52

1/20
1/20
13/20C
Males"
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-44        DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofbenzo[a]pyrene


Oral cavity
Papilloma
sec
Basal cell adenoma
Sebaceous cell carcinoma
Forestomach
Squamous cell papilloma
sec
Jejunum
Adenocarcinoma
Liver
Hepatocellular adenoma
Hepatocellular carcinoma
Cholangiocarcinoma
Kidney
Cortical adenoma
Adenocarcinoma
Urothelial carcinoma
Auditory canal
Benign
Squamous cell papilloma
Carcinoma
Sebaceous cell adenoma
Skin and mammary
Basal cell adenoma
Basal cell carcinoma
sec
Keratoacanthoma
Trichoepithelioma
Fibrosarcoma
Fibrous histiocytoma (malignant)
Dose (mg/kg-d)
0

0/24
1/24
0/24
0/24

0/52
0/52

0/51

0/52
0/52
0/52

0/52
0/52
0/52

0/1
0/1
0/1
0/1

2/52
1/52
0/52
1/52
0/52
0/52
0/52
3

0/24
0/24
0/24
0/24

7/5 2C
1/52

0/50

3/52
1/52
0/52

0/52
0/52
0/52

0/0
0/0
0/0
0/0

0/52
1/52
1/52
0/52
1/52
3/52
0/52
10

2/37
5/37
0/37
0/37

18/5 2C
25/52c

1/51

15/52C
23/52c
0/52

7/52c
2/52
0/52

1/7
0/7
2/7
0/7

1/52
0/52
1/52
1/52
2/52
5/52
1/52
30a

10/38C
11/38C
2/38
2/38

17/52C
35/52c

8/49c

4/52
45/52c
1/52

8/5 2C
0/52
3/52

0/33
4/33
19/33C
1/33

10/5 lc
4/51
5/51
4/51
8/5 lc
0/51
1/52
     aThis group had significantly decreased survival.
     blncidences 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.
     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).
1
2           Kroese et al. (2001) did not systematically investigate nonneoplastic lesions detected in rats
3    sacrificed during the 2-year study, because the focus was to identify and quantitate tumor
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 B-45         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    occurrence. However, incidences were reported for nonneoplastic lesions in tissues or organs in
 2    which tumors were detected (i.e., oral cavity, oesophagus, forestomach, jejunum, liver, kidney, skin,
 3    mammary, and auditory canal). The reported nonneoplastic lesions associated with exposure were
 4    the forestomach basal cell hyperplasia and clear cell foci of cellular alteration in the liver.
 5    Incidences for forestomach basal cell hyperplasia in the control through high-dose groups were
 6    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
 7    hepatic clear cell foci of cellular alteration were 22/52, 33/52, 4/52, and 2/52 for females and
 8    8/52, 22/52,1/52, and 1/52 for males. These results indicate that the lowest dose group, 3 mg/kg-
 9    day, was a LOAEL for increased incidence of forestomach hyperplasia and hepatic histological
10    changes in male and female Wistar rats exposed by gavage to benzo[a]pyrene for up to 104 weeks
11    (see Table 4-5). The lack of an increase in incidence of these nonneoplastic lesions in the
12    forestomach and liver at the intermediate and high doses (compared with controls) were
13    associated with increased incidences of forestomach and liver tumors at these dose levels.  The
14    authors of this study note that non-neoplastic effects were not quantified in organs with tumors.
15           As an adjunct study to the 2-year gavage study with Wistar rats, Kroese et al. (2001)
16    sacrificed additional rats (6/sex/group) after 4 and 5 months of exposure (0,1, 3,10, or 30 mg/kg-
17    day) for analysis of DNA adduct formation in WBCs and major organs and tissues. Additional rats
18    (6/sex/time period) were exposed to 0.1 mg/kg-day benzo[a]pyrene for 4 and 5 months for
19    analysis of DNA adduct formation.  Usingthe [32P]-postlabeling technique, five benzo[a]pyrene-DNA
20    adducts were identified in all of the examined tissues at 4 months (WBCs, liver, kidney, heart, lung,
21    skin, forestomach, glandular stomach, brain). Only one of these adducts (adduct 2) was identified
22    based on co-chromatography with  a standard. This adduct, identified as 10p-(deoxyguanosin-N2-
23    yl)-7p,8a,9a-trihydroxy-7,8,9,10 tetrahydro-benzo[a]pyrene (dG-N2-BPDE), was the predominant
24    adduct  in all organs of female rats exposed to 10 mg/kg-day, except the liver and kidney, in which
25    another adduct (unidentified adduct 4) was predominant. Levels of total adducts (number of
26    benzo[a]pyrene-DNA adducts per 1010 nucleotides) in examined tissues (from the single 10 mg/kg-
27    day female rat) showed the following order:  liver > heart > kidney > lung > skin > forestomach «
28    WBCs > brain. Mean values for  female levels of total benzo[a]pyrene-DNA adducts (number per
29    1010 nucleotides) in four organs showed the same order, regardless of exposure group: liver > lung
30    > forestomach « WBCs; comparable data for males were not reported).  Mean total benzo[a]pyrene-
31    DNA adduct levels in livers increased in both sexes from about 100 adducts per 1010 nucleotides at
32    0.1 mg/kg-day to about 70,000  adducts per 1010 nucleotides at 30 mg/kg-day. In summary, these
33    results  suggest that total benzo[a]pyrene-DNA adduct levels in tissues at 4 months were not
34    independently associated with the  carcinogenic responses noted after 2 years of exposure to
35    benzo[a]pyrene. The liver showed the highest total DNA adduct levels and a carcinogenic response,
36    but total DNA adduct levels in heart, kidney, and lung (in which no carcinogenic responses were
37    detected) were higher than levels in forestomach and skin (in which carcinogenic responses were
38    detected).

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-46         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1          Groups of Sprague-Dawley rats (32/sex/dose) were fed diets delivering a daily dose of
 2    0.15 mg benzo[a]pyrene/kg body weight every ninth day or 5 times/week (Brune et al., 1981).
 3    Other groups (32/sex/dose) were given gavage doses of 0.15 mgbenzo[a]pyrene (in aqueous 1.5%
 4    caffeine solution)/kg every ninth day, every third day, or 5 times/week.  The study included an
 5    untreated control group (to compare with the dietary exposed groups) and a gavage vehicle control
 6    group (each with 32 rats/sex). Rats were treated until moribundity or death occurred, with
 7    average annual doses are reported in Table 4-6 (mg/kg-year, calculated by Brune et al. [1981]).
 8    The following tissues were prepared for histopathological examination:  tongue, larynx, lung, heart,
 9    trachea, esophagus, stomach, small intestine, colon, rectum, spleen, liver, urinary bladder, kidney,
10    adrenal gland, and any tissues showing tumors or other gross changes. Survival was similar among
11    the groups, with the exception that the highest gavage-exposure group showed a decreased median
12    time of survival (Table B-10).  Increased incidences of portal-of-entry tumors (forestomach,
13    esophagus, and larynx) were observed in all of the gavage-exposed groups and in the highest
14    dietary exposure group (Table B-10).  Following dietary administration,  all observed tumors were
15    papillomas. Following gavage administration, two malignant forestomach tumors were found (one
16    each in the mid- and high-dose groups) and the remaining tumors were benign. The data in Table
17    4-6 show that the carcinogenic response to benzo[a]pyrene was stronger with the gavage protocol
18    compared with dietary exposure, and that no distinct difference in response was apparent between
19    the sexes. Tumors at distant sites (mammary gland, kidney, pancreas, lung, urinary bladder, testes,
20    hematopoietic, and soft tissue) were not considered treatment-related as they were also observed
21    at similar rates in the control group (data not provided). The study report did not address
22    noncancer systemic effects.

23          Table B-10. Incidences of alimentary tract tumors in Sprague-Dawley
24          rats chronically exposed to benzo[a]pyrene in the diet or by gavage in
25          caffeine solution
Average annual
dose (mg/kg-yr)
Estimated average
daily dose3
(mg/kg-d)
Forestomach tumors'5
Total alimentary tract
tumors0 (larynx,
esophagus,
forestomach)
Median
survival time
(wks)
Benzo[a]pyrene by gavage in 1.5% caffeine solution
0
6
18
39
0
0.016
0.049
0.107
3/64 (4.7%)
12/64 (18.8%)d
26/64 (40.1%)e
14/64 (21.9%)e
6/64 (9.4%)
13/64 (20.3%)
26/64 (40.6%)
14/64 (21.9%)
102
112
113
87
Benzo[a]pyrene in diet
0
6
39
0
0.016
0.107
2/64 (3.1%)
1/64 (1.6%)
9/64(14.1%)d
3/64 (4.7%)
3/64 (4.7%)
10/64 (15.6%)
129
128
131
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-47        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
      aAverage annual dose divided by 365 days.
      bNo sex-specific forestomach tumor incidence data were reported by Brune et al. (1981).
      cSex-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).
 1
 2          In the other modern cancer bioassay with benzo[a]pyrene, female B6C3Fi mice (48/dose
 3    group) were administered benzo[a]pyrene (98.5% purity) at concentrations of 0 (acetone vehicle),
 4    5, 25, or 100 ppm in the diet for 2 years (Beland and Gulp, 1998; Gulp et al., 1998). This study was
 5    designed to compare the carcinogenicity of coal tar mixtures with that of benzo[a]pyrene and
 6    included groups of mice fed diets containing one of several concentrations of two coal tar mixtures.
 7    Benzo[a]pyrene was dissolved in acetone before mixing with the feed.  Control mice received only
 8    acetone-treated feed.  Female  mice were chosen because they have a lower background incidence of
 9    lung tumors than male B6C3Fi mice. Gulp et al. (1998) reported that the average daily intakes of
10    benzo[a]pyrene in the 25- and 100-ppm groups were  104 and 430 ug/day,  but did not report
11    intakes for the 5-ppm group. Based on the assumption that daily benzo[a]pyrene intake at 5 ppm
12    was one-fifth of the 25-ppm intake (about 21 ug/day), average daily doses for the three
13    benzo[a]pyrene groups are estimated at 0.7, 3.3, and 16.5 mg/kg-day. Estimated doses were
14    calculated using time-weighted average (TWA) body weights of 0.032 kg for the control, 5- and 25-
15    ppm groups and 0.026 kg for the 100-ppm group (estimated from graphically presented data).
16    Food consumption, body weights, morbidity, and mortality were monitored at intervals, and lung,
17    kidneys, and liver were weighed at sacrifice. Necropsy was performed on all mice that died during
18    the experiment or survived to the end of the study period. Limited histopathologic examinations
19    (liver, lung, small intestine, stomach, tongue, esophagus) were performed on all control and high-
20    dose mice and on all mice that died during the experimental period, regardless of treatment group.
21    In addition, all gross lesions found in mice of the low-  and mid-dose groups were examined
22    histopathologically.
23          None of the mice administered  100 ppm benzo[a]pyrene survived to the end of the study,
24    and morbidity/mortality was  100% by week 78. Decreased survival was also observed at 25 ppm
25    with only 27% survival at 104 weeks, compared with  56 and 60%, in the 5-ppm and control groups,
26    respectively. In the mid- and high-dose group, 60% of mice were alive at about 90 and 60 weeks,
27    respectively. Early deaths in exposed mice were attributed to tumor formation rather than other
28    causes of systemic toxicity.  Food consumption was not statistically different in
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-48         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
 1   benzo[a]pyrene-exposed and control mice. Body weights of mice fed 100 ppm were similar to
 2   those of the other treated and control groups up to week 46, and after approximately 52 weeks,
 3   body weights were reduced in 100-ppm mice compared with controls. Body weights for the 5- and
 4   25-ppm groups were similar to controls throughout the treatment period. Compared with the
 5   control group, no differences in liver, kidney, or lung weights were evident in any of the treated
 6   groups (other organ weights were not measured).
 7          Papillomas and/or carcinomas of the forestomach, esophagus, tongue, and larynx at
 8   elevated incidences occurred in groups of mice exposed to 25 or 100 ppm, butno exposure-related
 9   tumors occurred in the liver or lung (Table B-ll; Beland and Gulp, 1998; Gulp et al., 1998). The
10   forestomach was the most sensitive tissue, and demonstrated the highest tumor incidence among
11   the examined tissues and was the only tissue with an elevated incidence of tumors at 25 ppm
12   (Table B-ll). In addition, most of the forestomach tumors in the exposed groups were carcinomas,
13   as 1, 31, and 45 mice had forestomach carcinomas in the 5-, 25-, and 100-ppm groups respectively.
14   Nonneoplastic lesions were also found in the forestomach at significantly (p < 0.05) elevated
15   incidences: hyperplasia at >5 ppm and hyperkeratosis at >25 ppm (Table B-ll).  The esophagus
16   was the only other examined tissue showing elevated incidence of a nonneoplastic lesion (basal cell
17   hyperplasia, see Table B-ll). Tumors (papillomas and carcinomas) were also significantly elevated
18   in the esophagus and tongue at 100 ppm (Table B-ll). Esophogeal carcinomas were detected in 1
19   mouse at 25 ppm and in 11 mice at 100 ppm. Tongue carcinomas were detected in seven 100-ppm
20   mice; the remaining tongue tumors were papillomas.  Although incidences of tumors of the larynx
21   were not significantly elevated in any of the exposed groups, a significant dose-related trend was
22   apparent (Table B-ll).

23          Table B-ll.  Incidence of nonneoplastic and neoplastic lesions in female
24          B6C3Fi mice fed benzo[a]pyrene in the diet for up to 2 years
Tissue and lesion
Liver (hepatocellular adenoma)
Lung (alveolar/bronchiolar adenoma and/or carcinoma)
Forestomach (papilloma and/or carcinoma)
Forestomach (hyperplasia)
Forestomach (hyperkeratosis)
Incidence (%)
Benzo[a]pyrene concentration (ppm) in diet
0
5
25
100
Average daily doses (mg/kg-d)
0
2/48
(2)
5/48
(10)
l/48b
(2)
13/48b
(27)
13/48b
0.7
7/48
(15)
0/48
(0)
3/47
(6)
23/47
(49)
22/47
3.3
5/47
(11)
4/45
(9)
36/46a
(78)
33/46a
(72)
33/46a
16.5
0/45
(0)
0/48
(0)
46/47a
(98)
37/47a
(79)
38/47a
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-49        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

Esophagus (papilloma and/or carcinoma)
Esophagus (basal cell hyperplasia)
Tongue (papilloma and/or carcinoma)
Larynx (papilloma and/or carcinoma)
(27)
0/48b
(0)
l/48b
(2)
0/49b
(0)
0/35b
(0)
(47)
0/48
(0)
0/48
(0)
0/48
(0)
0/35
(0)
(72)
2/45
(0)
5/45
(11)
2/46
(4)
3/34
(9)
(81)
27/46a
(59)
30/46a
(65)
23/48a
(48)
5/38
(13)
      Significantly different from control incidence (p < 0.05); using a modified Bonferonni procedure for
      multiple comparisons to the same control.
      Significant (p < 0.05) dose-related trend calculated for incidences of these lesions.

      Sources: Beland and Gulp (1998); Gulp et al. (1998).
 1
 2          Neal and Rigdon (1967) fed benzo[a]pyrene (purity not reported) at concentrations of 0,1,
 3    10, 20, 30, 40, 45, 50,100, and 250 ppm to male and female CFW-Swiss mice in the diet
 4    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, and
 5    44.4 mg/kg-day. The age of the mice ranged from 17 to 180 days old and the treatment time was
 6    from 1 to 197 days; the size of the treated groups ranged from 9 to 73. There were 289 mice
 7    (number of mice/sex not stated) in the control group. No forestomach tumors were reported at 0,
 8    0.2, or 1.8 mg/kg-day. The incidence of forestomach tumors at 20, 30, 40, 45, 50,100, and 250 ppm
 9    dose groups (3.6, 5.3, 7.1, 8, 8.9,17.8, and 44.4 mg/kg-day) were 1/23, 0/37,1/40, 4/40, 23/34,
10    19/23, and 66/73, respectively.

11    Other Oral Exposure Cancer Bioassavs in Mice
12          Numerous other oral exposure cancer bioassays in  mice have limitations that restrict their
13    usefulness for characterizing dose-response relationships between chronic-duration oral exposure
14    to benzo[a]pyrene and noncancer effects or cancer, but collectively, they provide strong evidence
15    that oral exposure to benzo[a]pyrene can cause portal-of-entry site tumors (see Table B-12 for
16    references).
            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 kg (females) were used (U.S. EPA, 1988) and resulting
     doses were averaged  between males and females.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-50         DRAFT—DO NOT CITE OR QUOTE

-------
                                      Toxicological Review ofbenzo[a]pyrene
Table B-12. Other oral exposure cancer bioassays in mice
Species/strain
Rat/Sprague-
Dawley














Mouse/HalCR







Mouse/HalCR









Mouse/HalCR










Exposure
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.
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.
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.


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.





Results
Dose larynx, esophagus,
and forestomach
(gavage) tumors
0 6/64
0.016 13/64
0.049 26/64
0.107 14/64
(diet)
0 3/64
0.016 3/64
0.107 10/64





Incidence with
forestomach tumors:
Low 11/20 (18 wks)
Mid 13/19 (20 wks)
High 12/12 (12 wks)



Incidence with
forestomach tumors:
Control 0/9
Low 6/9
High 9/9





8/20 exposed mice had
forestomach tumors.









Comments
Doses are annual
averages.
Nonstandard
treatment
protocol involved
animals being
treated for <5
days/week;
relatively high
control incidence
compared to
other gavage
studies.



Less-than-lifetime
exposure
duration; only
stomachs were
examined for
tumors; tumors
found only in
forestomach.
Less-than-lifetime
exposure
duration;
glandular
stomach, lung,
and livers from
control and
exposed mice
showed no
tumors.
Less-than-lifetime
exposure
duration; only
stomachs were
examined for
tumors; tumors
found only in
forestomach; no
nonexposed
controls were
mentioned.
Reference
Brune et al.,
1981














Wattenberg
,1972






Triolo et al.,
1977








Wattenberg
,1974









   This document is a draft for review purposes only and does not constitute Agency policy.
                                 B-51        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Species/strain
Mouse/CD-I









Mouse/BALB











Mouse/C3H




Mouse/albino














Mouse/albino







Exposure
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
was 33 mg/kg-d, using an
average body weight of
0.030 kg from reported
data.
25 mice (8 wks old) were
given 0.5 mg
benzo[a]pyrene 2
times/wk for 15 wks.








19 mice (about 3 mo old)
were given 0.3 mLof
0.5% benzo[a]pyrene in
polyethylene glycol-400
by gavage, once/d for 3 d.
Groups of 17-18 mice
were given single doses
of benzo[a]pyrene and
allowed to survive until
terminal sacrifice at
569 d.









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.

Results
Incidence with
forestomach tumors:
Exposed 17/20 (85%)
Controls 0/24






5/25 mice had squamous
carcinomas of the
forestomach; tumors
were detected 28-65 wks
after treatment.







By 30 wks, 7/10 mice had
papillomas; no
carcinomas were evident.


Incidence of mice (that
survived at least to 60 d)
with forestomach
papillomas:
Dose (u.g) Incidence
(Experiment 1)
(Experiment 2)
Control 0/17
0/18
12.5 3/17
2/18
50 0/17
1/17
200 8/17
NE
Gastric tumors were
observed at the following
incidence:
Control 0/158
8 mg benzo[a]pyrene
total 13/160


Comments
Less-than-lifetime
exposure
duration; only
stomach were
examined for
tumors; tumors
found only in
forestomach.


Less-than-lifetime
exposure
duration; the
following details
were not
reported:
inclusion of
controls,
methods for
detecting tumors,
and body weight
data.
Less-than-lifetime
exposure
duration.


Less-than-lifetime
exposure
duration; Gl tract
examined for
tumors with hand
lens; body weight
data not
reported.







Close to lifetime
exposure
duration; daily
dose levels and
methods of
detecting tumors
were not clearly
reported.
Reference
EI-Bayoumy,
1985








Biancifiori
etal., 1967










Berenblum
and Haran,
1955


Field and
Roe, 1965













Chouroulink
ovet al.,
1967





This document is a draft for review purposes only and does not constitute Agency policy.
                               B-52        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Species/strain
Mouse/CFW














Mouse/Swiss
albino







Mouse/ICR












Exposure
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.







Groups of mice (9-14 wks
old) were given single
doses of 0 or 0.05 mg
benzo[a]pyrene in
polyethylene glycol-400
bygavage. Surviving mice
were killed at 18 mo of
age and examined for
macroscopic tumors.
Groups of 20 or 24 mice
(71 d old) were given
1.5 mg benzo[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.
Results
ppm Exposure
Forestomach tumor
(d)
incidence
1 110 0/25
10 110 0/24
20 110 1/23
30 110 0/37
40 110 1/40
45 110 4/40
50 152 24/34
100 110 19/23
250 118 66/73


Forestomach tumor
incidence:
Dose (u.g) - Carcinoma
Papilloma
0 0/65
2/65
50 1/61
20/61

Incidence of mice with
forestomach neoplasms
Experiment 1 23/24
Experiment 2 19/20









Comments
Less-than-lifetime
exposure
duration; no
vehicle control
group; animals
ranged from 3
wks to 6 mo old
at the start of
dosing; only
alimentary tract
was examined for
tumors (see also
Rigdon and Neal,
1969, 1967,
1966).
Less-than-lifetime
duration of
exposure;
exposure-related
tumors only
found in
forestomach.


Less-than-lifetime
duration of
exposure; only
stomachs were
examined for
tumors; tumors
found only in
forestomach;
nonexposed
controls were not
mentioned.


Reference
Neal and
Rigdon,
1967












Roeetal.,
1970







Benjamin et
al., 1988











This document is a draft for review purposes only and does not constitute Agency policy.
                               B-53        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Species/strain
Mouse/white













Mouse/A/HeJ








Mouse/A/J












Mouse/A/J








Exposure
Groups of 16-30 mice
were given
benzo[a]pyrene in
triethylene glycol (0.001-
10 mg) wkly for 10 wks
and observed until 19
mo.







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.
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 u.g/mouse/d.
Estimated doses were 0,
1.6, and 9.9 mg/kg-d
using a chronic reference
body weight value of
0.026 kg (U.S. EPA, 1988).
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.
Results
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
12/12 exposed mice had
lung tumors.







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



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

Comments
Less-than-lifetime
exposure
duration.











Less-than-lifetime
exposure
duration; only
lungs examined
for tumors; no
nonexposed
controls were
mentioned.

Close to lifetime
exposure
duration; A/J
strain of mice
particularly
sensitive to
chemically
induced cancer;
only lungs and
stomachs were
examined for
tumors.

Less-than-lifetime
duration of
exposure; only
lungs and Gl tract
were examined
for tumors.



Reference
Fedorenko
and
Yansheva,
1967; as
cited in U.S.
EPA, 199 la








Wattenberg
,1974







Weyand et
al., 1995











Robinson et
al., 1987







 NE = not evaluated
This document is a draft for review purposes only and does not constitute Agency policy.
                               B-54        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
 1    Inhalation Studies
 2    Short-term and Subchronic Studies
 3          Wolff et al. (1989) exposed groups of 40 male and 40 female F344/Crl rats, via nose only, to
 4    7.5 mg benzo[a]pyrene/m3 for 2 hours/day, 5 days/week for 4 weeks (corresponding to a TWA of
 5    0.45 mg/m3). Rats were 10-11 weeks old atthe beginning of the experiment Benzo[a]pyrene
 6    (>98% pure) aerosols were formed by heating and then condensing the vaporized benzo[a]pyrene.
 7    The particle MMAD was 0.21 |im.  Subgroups of these animals (six/sex/dose) were exposed for
 8    4 days or 6 months after the end of the 4-week exposure to radiolabeled aluminosilicate particles.
 9    Lung injury was assessed by analyzing clearance of radiolabeled aluminosilicate particles and via
10    histopathologic evaluations. Body and lung weights, measured in subgroups from 1 day to 12
11    months after the exposure did not differ between controls and treated animals. Radiolabeled
12    particle clearance did not differ between the control and treated groups, and there were no
13    significant lung lesions. This study identified a NOAEL for lung effects of 0.45 mg/m3 for a short-
14    term exposure.

15    Chronic Studies and Cancer Bioassavs
16          Thyssen et al. (1981) conducted an inhalation study in which male Syrian golden hamsters
17    were exposed to benzo[a]pyrene for their natural lifetime. Groups of 20-30 animals (8 weeks old)
18    were exposed by nose-only inhalation to NaCl aerosols (controls; 240 [ig NaCl/m3) or
19    benzo[a]pyrene condensed onto NaCl aerosols atthree nominal concentrations of 2,10, or 50 mg
20    benzo[a]pyrene/m3 for 3-4.5 hours/day,  5 days/week for 1-41 weeks, followed by 3 hours/day,
21    7 days/week for the remainder of study (until hamsters died or became moribund). Thyssen et al.
22    (1981) reported average measured benzo[a]pyrene concentrations to be 0, 2.2, 9.5, or 46.5 mg/m3.
23    More than 99% of the particles were between 0.2 and 0.5 |im in diameter, and over 80% had
24    diameters between 0.2 and 0.3 |im. The particle analysis of the aerosols was not reported to
25    modern standards (MMAD and geometric SD were  not reported).  Each group initially consisted of
26    24 hamsters; final group sizes were larger as animals dying during the first 12 months of the  study
27    were replaced.
28          Survival was similar in the control, low-dose, and mid-dose groups, but was significantly
29    decreased in the high-dose group. Average survival times in the control, low-, mid-, and high-dose
30    groups were 96.4 ± 27.6, 95.2 ± 29.1, 96.4 ± 27.8, and 59.5 ± 15.2 weeks, respectively. After the 60*
31    week, body weights decreased and mortality increased steeply in the highest dose group.
32    Histologic examination of organs (a complete list of organs examined histologically was not
33    reported by Thyssen et al. [1981]) revealed a dose-related increase in tumors in the upper
34    respiratory tract, including the nasal cavity, pharynx, larynx, and trachea, and in the digestive tract
35    in the mid- and high-dose groups  (Table B-13). A statistical analysis was not included in the
36    Thyssen et al. (1981) report. No lung tumors were observed. Squamous cell tumors in the
37    esophagus and forestomach were also observed in  the high-dose group, presumably as a

                This document is a draft for review purposes  only and does not constitute Agency policy.
                                                B-55         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1    consequence of mucociliary particle clearance. Tumors were detected in other sites, but none of
 2    these appeared to be related to exposure. The results indicated that the pharynx and larynx,
 3    including the epiglottis, were the main cancer targets (Table B-13).
 4
 5
       Table B-13. Incidence of respiratory and upper digestive tract tumors
       in male hamsters treated for life with benzo[a]pyrene by inhalation
Tumor site
Nasal cavity
Larynx
Trachea
Lung
Pharynx
Esophagus
Forestomach
Reported benzo[a]pyrene concentration (mg/m3)
Oa
2b
10
50
Tumor incidence (latency in wksc)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3/26 (116 ±1.5)
8/26(107.1115.5)
1/26 (115)
0
6/26 (97.2 ± 16.9)
0
1/26 (119)
1/25 (79)
13/25(67.6112.1)
3/25(63.3133.3)
0
14/25 (67.5 ± 12.2)
2/25 (70, 79)
1/25 (72)
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Effective number of animals in control group: n = 27.
bEffective number of animals in 2 mg/m3 dose group: n = 27.
cMean±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 original individual
animal data to calculate average continuous lifetime exposures for each individual hamster. Group
averages of individual average continuous lifetime exposure concentrations were 0, 0.25,1.01, and
4.29 mg/m3 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 B-14
presents the number of animals with tumors in the larynx and pharynx and the numbers of animals

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-56        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
      in each exposure group.  Numbers of animals with either laryngeal or pharyngeal tumors are also
      noted in Table B-14, 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 B-14. 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
concentration3
(mg/m3)
Control
0.25
1.01
4.29
Number of
hamsters in
group15
27
27
26
34
Larynxb
Malignant
0
0
8
9
All
0
0
11
12
Pharynx15
Malignant
0
0
7
17
All
0
0
9
18
Larynx or pharynx,
combined0
Malignant
0
0
11
17
All
0
0
16
18
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
     aAs calculated by Clement Associates (1990) from air monitoring data collected by Thyssen and
     colleagues.
     bAs counted from information in Table E-l in Appendix E, which was obtained from examination of
     individual animal pathology reports prepared by Thyssen and colleagues and obtained by Clement
     Associates.
     cAs counted from information in Table E-l in Appendix E. 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 thatbenzo[a]pyrene is
     tumorigenic, but do not provide data useful for characterizing dose-response relationships because
     of their design (Kobayashi, 1975; Reznik-Schuller and Mohr, 1974; Henry etal., 1973; Mohr, 1971;
     Saffiotti etal., 1968; Gross etal., 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 mgbenzo[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 atportal-of-entry sites.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-57         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
 1    Dermal studies
 2    Skin-Tumor Initiation-Promotion Assays
 3          Results from numerous studies indicate that acute dermal exposure to benzo[a]pyrene
 4    induces skin tumors in mice when followed by repeated exposure to a potent tumor promoter
 5    (Weyand et al., 1992; Cavalieri et al., 1991,1981; Rice et al., 1985; El-Bayoumy et al., 1982; LaVoie
 6    et al., 1982; Raveh et al., 1982; Slaga et al., 1980,1978; Wood et al., 1980; Hoffmann et al., 1972).
 7    The typical exposure protocol in these studies involved the application of a single dose of
 8    benzo[a]pyrene (typically >20 nmol per mouse) to dorsal skin of mice followed by repeated
 9    exposure to a potent tumor promoter, such as 12-0-tetradecanoylphorbol-13-acetate (TPA).

10    Carcinogenicity Bioassays
11          Repeated application of BaP to skin (in the absence of exogenous promoters) has been
12    variously demonstrated to induce skin tumors in mice, rats, rabbits, and guinea pigs (IARC, 2010,
13    1983,1973; WHO, 1998; ATSDR, 1995). Mice have been most extensively studied, presumably
14    because of early evidence that they may be more sensitive than other animal species, but
15    comprehensive comparison of species differences in sensitivity to lifetime dermal exposure are not
16    available. Early studies of complete dermal carcinogenicity in other species (rats, hamsters, guinea
17    pigs, and rabbits) have several limitations which make them not useful for dose-response analysis
18    (see IARC,  1973 for  descriptions of studies by Nakano et al., 1937, Shubik et al., 1960; Oberling et
19    al., 1937; Schiirch and Winterstein, 1935; Wynder et al., 1957).  The limitations in these studies
20    include inadequate reporting of the amount of BaP applied, use of the carcinogen benzene as a
21    vehicle, and less than lifetime exposure duration.
22          This section discusses complete carcinogenicity bioassays in mice that provide the best
23    available dose-response data for skin tumors caused by repeated dermal exposure to BaP (Sivak et
24    al., 1997; Higginbotham et al., 1993; Albert et al., 1991; Habs et al., 1984,1980; Grimmer et al.,
25    1984, 1983; Schmahl et al., 1977; Schmidt et al., 1973; Roe et al., 1970; Poel, 1960, 1959). Early
26    studies of BaP complete carcinogenicity in mouse skin (Wynder and Hoffman 1959; Wynder et al,
27    1957) are notfurther described herein, because the investigators applied solutions of BaP at
28    varying concentrations on the skin, but did not report volumes applied. As such, applied doses in
29    these studies cannot be determined. Other complete carcinogenicity mouse skin tumor bioassays
30    with BaP are available, but these are not described further in this review, because:  (1) they only
31    included one BaP dose level (e.g., Emmett et al., 1981) or only dose levels inducing 90-100%
32    incidence of mice with tumors (e.g., Wilson and Holland, 1988; Warshawsky and Barkley, 1987) and
33    thus provide no information about the shape  of the dose-response relationship; (2) they used a 1-
34    time/week (e.g., Nesnow et al., 1983) or 1-time every 2 weeks (e.g., Levin et al., 1977) exposure
35    protocol, which is less useful for extrapolating to daily human exposure; or (3) they used a vehicle
36    demonstrated to interact with or enhance benzo[a]pyrene carcinogenicity (Bingham and Falk,
37    1969).

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-58       DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 9
10
       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 ug, 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 B-15). Doses >3.8 ugwere 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 B-15.  Skin tumor incidence and time of appearance in male C57L
       mice dermally exposed to benzo[a]pyrene for up to 103 weeks
Dose (ug)a
0 (Toluene)
0.15
0.38
0.75
3.8
19
94
188
376
752
Incidence of mice with
gross skin tumors
0/33 (0%)
5/55 (9%)
11/55 (20%)
7/56 (13%)
41/49 (84%)
38/38 (100%)
35/35 (100%)
12/14 (86%)
14/14 (100%)
13/13 (100%)
Time of first tumor
appearance (wks)
-
42-44c
24
36
21-25
11-21
8-19
9-18
4-15
5-13
Incidence of mice
with epidermoid
carcinoma11
0/33 (0%)
0/55 (0%)
2/55 (4%)
4/56 (7%)
32/49 (65%)
37/38 (97%)
35/35 (100%)
10/14 (71%)
12/14 (86%)
13/13 (100%)
Length of exposure
period (wks)
92
98
103
94
82
25-44c
22-43
20-35
19-35
19-30
11
12
13
14
15
16
17
18
19
20
Indicated doses were applied to interscapular skin 3 times/week for up to 103 weeks or until time of
appearance of a grossly detected skin tumor.
bCarcinomas were histologically confirmed.
cRanges 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, CSHeB, 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 ug 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 B-16).  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 CSHeB mice showed skin tumors at

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-59         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene

 1    doses >0.38 ugbenzo[a]pyrene, whereas AH/e mice showed tumors at doses >19 ug
 2    benzo[a]pyrene (Table B-16). Except for metastases of the skin tumors to lymph nodes and lung,
 3    Poel (1960) did not mention the appearance of exposure-related tumors in tissues other than
 4    interscapular skin.

 5          Table B-16.  Skin tumor incidence and time of appearance in male SWR,
 6          C3HeB, and A/He mice dermally exposed to benzo [ajpyrene for life or
 7          until a skin tumor was detected
Dose (ug)a
0 (Toluene)
0.15
0.38
0.75
3.8
19.0
94.0
470.0
SWR mice
Tumor
incidence13
0/20 (0%)
0/25 (0%)
2/22 (9%)
15/18 (83%)
12/17 (70%)
16/16 (100%)
16/17 (94%)
14/14 (100%)
Time of first
tumor
appearance
(wks)

-
55-55
25-72
25-51
12-28
9-17
5-11
CSHeB mice
Tumor
incidence13
0/17 (0%)
0/19 (0%)
3/17 (18%)
4/17 (24%)
11/18 (61%)
17/17 (100%)
18/18 (100%)
17/17 (100%)
Time of first
tumor
appearance
(wks)
-
-
81-93
51-93
35-73
13-32
10-22
4-19
A/He mice
Tumor
incidence13
0/17 (0%)
0/18 (0%)
0/19 (0%)
0/17 (0%)
0/17 (0%)
21/23 (91%)
11/16 (69%)
17/17 (100%)
Time of fist
tumor
appearance
(wks)
-
-
-
-
-
21-40
14-31
4-21
      Indicated doses were applied 3 times/week for life or until a skin tumor was detected. Mice were 10-
      14 weeks old at initial exposure.
      blncidence of mice exposed >10 weeks with a skin tumor.

      Source: Poel (1960).
 9          Roe et al. (1970) treated groups of 50 female Swiss mice with 0 (acetone vehicle), 0.1, 0.3,1,
10    3, or 9 ug benzo[a]pyrene applied to the shaved dorsal skin 3 times/week for up to 93 weeks; all
11    surviving mice were killed and examined for tumors during the following 3 weeks. The dorsal skin
12    of an additional control group was shaved periodically but was not treated with the vehicle. Mice
13    were examined every 2 weeks for the development of skin tumors at the site of application.
14    Histologic examinations included: (1) all skin tumors thought to be possibly malignant; (2) lesions
15    of other tissues thought to be neoplastic; and (3) limited nonneoplastic lesions in other tissues. As
16    shown in Table B-17, markedly elevated incidences of mice with skin tumors were only found in the
17    two highest dose groups (3 or 9  ug), compared with no skin tumors in the control groups.
18    Malignant skin tumors (defined as tumors with invasion or penetration of the panniculus carnosus
19    muscle) were detected in 4/41 and 31/40 mice in the 3- and 9-ug groups, respectively, surviving to
20    atleastSOO days. Malignant lymphomas were detected in all groups, but the numbers of cases were
21    not elevated compared with expected numbers after adjustment for survival differences.  Lung

               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-60        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1   tumors were likewise detected in control and exposed groups at incidences that were not
 2   statistically different

 3          Table B-17. Tumor incidence in female Swiss mice dermally exposed to
 4          benzo[a]pyrenefor up to 93 weeks
Dose (ug)a
No treatment
Acetone
0.1
0.3
1
3
9
Cumulative number of mice with skin
tumor/survivors
200 d
0/48
0/49
0/45
0/46
0/48
0/47
0/46
300 d
0/43
0/47
1/42
0/42
0/43
0/41
4/40
400 d
0/40
0/45
1/35
0/37
0/37
1/37
21/32
500 d
0/31
0/37
1/31
0/30
1/30
7/35
28/21
600 d
0/21
0/23
1/22
0/19
1/18
8/24
33/8
700 d
0/0
0/0
1/0
0/0
1/0
8/0
34/0
Skin tumor
incidence13
0/43 (0%)
0/47 (0%)
1/42 (2%)
0/42 (0%)
1/43 (2%)
8/41 (20%)
34/46 (74%)
Malignant
lymphoma
incidence0
19/44 (43%)
12/47 (26%)
11/43 (26%)
10/43 (23%)
16/44 (36%)
23/42 (55%)
9/40 (23%)
Lung tumor
incidence0
12/41 (29%)
10/46 (22%)
10/40 (25%)
13/43 (30%)
15/43 (35%)
12/40 (30%)
5/40 (13%)
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
aDoses were applied 3 times/week for up to 93 weeks to shaved dorsal skin.
bNumerator:  number of mice detected with a skin tumor. Denominator: number of mice surviving to
300 days for all groups except the highest dose group. For the highest dose group (in which skin tumors
were first detected between 200 and 300 days), the number of mice surviving to 200 days was used as
the denominator.
Numerator:  number of mice detected with specified tumor. Denominator:  number of mice surviving
to 300 days unless a tumor was detected earlier, in which case, the number dying before 300 days
without a tumor was subtracted from the number of animals reported to have been examined.

Source:  Roeetal. (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 [ig until spontaneous death occurred or until an
advanced carcinoma was observed. Skin carcinomas were identified by the presence of crater-
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 B-18), with induction
periods of 53.0 and 75.8 weeks for the 0.8 and 2.0 ug 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.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-61        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
 1
 2
       Table B-18. Skin tumor incidence in female NMRI and Swiss mice
       dermally exposed to benzo[a]pyrene
Dose (ug)a'b
Skin tumor incidence (all
types)
Incidence of papilloma
Incidence of carcinoma
Female NMRI mice
0 (Acetone)
0.05
0.2
0.8
2
0/100 (0%)
0/100 (0%)
0/100 (0%)
2/100 (2%)
30/100 (30%)
0/100 (0%)
0/100 (0%)
0/100 (0%)
0/100 (0%)
2/100 (2%)
0/100 (0%)
0/100 (0%)
0/100 (0%)
2/100 (2%)
28/100 (28%)
Female Swiss mice
0 (Acetone)
0.05
0.2
0.8
2
0/80 (0%)
0/80 (0%)
0/80 (0%)
5/80 (6%)
45/80 (56%)
0/80 (0%)
0/80 (0%)
0/80 (0%)
0/80 (0%)
3/80 (4%)
0/80 (0%)
0/80 (0%)
0/80 (0%)
5/80 (6%)
42/80 (52%)
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

15
16
     aMice were exposed until natural death or until
     blndicated doses were applied 2 times/week to
      Source:  Schmidt et al. (1973).
                                       they developed a carcinoma at the site of application.
                                       shaved skin of the back.
       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 u,g in 20 uL 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 B-19).  Carcinoma was the
primary tumor type observed following chronic dermal exposure to benzo[a]pyrene, and skin
papillomas occurred infrequently. Dermal sarcoma was not observed.

       Table B-19. Skin tumor incidence in female NMRI mice dermally
       exposed to benzo[a]pyrene
Dose (ug)a'b
0
1
Skin tumor incidence
(all types)
1/81 (1%)
11/77 (14%)
Incidence of papilloma
0/81 (0%)
1/77 (1%)
Incidence of carcinoma
0/81 (0%)
10/77 (13%)
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-62         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
1.7
3
25/88 (28%)
45/81 (56%)
0/88 (0%)
2/81 (3%)
25/88 (28%)
43/81 (53%)
 1
 2
 3
 4
 5
 6
 7
 8
 9
10

11
12
aMice were exposed until natural death or until they developed a carcinoma at the site of application.
blndicated doses were applied 2 times/week to shaved skin of the back.

Source:  Schmahl et al. (1977).

       Habs etal. (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 ugin 20 uL 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 as 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 B-20).

       Table B-20. Skin tumor incidence in female NMRI mice dermally
       exposed to benzo[a]pyrene
Dose (ug)a'b
0 (acetone)
1.7
2.8
4.6
Skin tumor incidence
0/35 (0%)
8/34 (24%)
24/35 (68%)
22/36 (61%)
Age-standardized tumor incidence0
0%
24.8%
89.3%
91.7%
13
14
15
16
17
18
19
20
21
22
aMice were exposed until natural death or until they developed a carcinoma at the site of application.
blndicated doses were applied 2 times/week to shaved skin of the back.
cMortality data of the total study population were used to derive the age-standardized tumor incidence.

Source:  Habs et al. (1980).

       Grimmer etal. (1984,1983) 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 u,g in the 1983 experiment, and 0, 3.4,
6.7, and 13.5 u,g 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 B-21).

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-63         DRAFT—DO  NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
 1   Carcinoma was the primary tumor type observed and a dose-response relationship was evident for
 2   carcinoma formation and incidence of all types of skin tumors.

 3          Table B-21. Skin tumor incidence and time of appearance in female
 4          CFLP mice dermally exposed to benzo[a]pyrene for 104 weeks
Dose (ug)a
Skin tumor incidence
(all types)
Incidence of
papilloma
Incidence of
carcinoma
Tumor appearance
in weeks
Grimmer etal. (1983)
0 (1:3 Solution of
acetone:DMSO)
3.9
7.7
15.4
0/80 (0%)
22/65 (34%)
39/64 (61%)
56/64 (88%)
0/80 (0%)
7/65 (11%)
5/64 (8%)
2/64 (3%)
0/80 (0%)
15/65 (23%)
34/64 (53%)
54/64 (84%)
—
74.6 ± 16.78b
60.9 ±13.90
44.1 ±7.66
Grimmer etal. (1984)
0 (1:3 Solution of
acetone:DMSO)
3.4
6.7
13.5
0/65 (0%)
43/64 (67%)
53/65 (82%)
57/65 (88%)
0/65 (0%)
6/64 (9%)
8/65 (12%)
4/65 (6%)
0/65 (0%)
37/64 (58%)
45/65 (69%)
53/65 (82%)
—
61 (53-65)c
47 (43-50)
35 (32-36)
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
       Indicated doses were applied twice/week to shaved skin of the back.
       bMean±SD.
       'Median with 95% Cl.

       Sources:  Grimmer et al. (1984,1983)

       Habs etal. (1984) appliedbenzo[a]pyrene (in 0.01 mL acetone) to the shaved interscapular
skin of female NMRI mice at doses of 0, 2, or 4 ug, 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 ug/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 B-22). Hematopoietic tumors (at 6/20, 3/20, and 3/20) and
lung adenomas (at 2/20,1/20, and 0/20) were observed in the controls and in the benzo[a]pyrene
treatment groups, but did not appear to be treatment related according to the study authors.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              B-64         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1
 2
       Table B-22. Skin tumor incidence in female NMRI mice dermally
       exposed to benzo[a]pyrene for life
Dose (u,g)a'b
0 (Acetone)
2
4
Skin tumor
incidence (all types)
0/20 (0%)
9/20 (45%)
17/20 (85%)
Incidence of
papilloma
0/20 (0%)
2/20 (10%)
0/20 (0%)
Incidence of
carcinoma
0/20 (0%)
7/20 (35%)
17/20 (85%)
Mean survival time,
days (95% Cl)
691 (600-763)
648 (440-729)
528 (480-555)
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
aMice were exposed until natural death or until they developed an invasive tumor at the site of
application.
blndicated doses were applied 2 times/week to shaved interscapular skin.

Source:  Habs et al. (1984).

       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 ug/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.
       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 |ig) benzo[a]pyrene in a 50 uL volume of cyclohexanone/acetone (1:1) for 104
weeks beginning at 8 weeks of age. Mice dying during the exposure period or sacrificed at the 24
month termination were necropsied; mice with skin tumors that persisted for 4 consecutive weeks
with diameters > 3 cm were sacrificed before the study termination and also necropsied. Skin
samples and any grossly observed lesions were subjected to histopathological examination.
Carcinomas and sarcomas were referred to as carcinomas, whereas papillomas, keratoacanthomas,
and fibromas were referred to as papillomas. The incidences of mice with skin tumors and mean
survival times for each group are shown in Table B-23. All high-dose mice died before the final
sacrifice, and 80% showed scabs and sores at the site of application.  The time of first tumor
appearance was not reported for the tumor-inducing groups, but from a plot of the tumor incidence

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-65        DRAFT—DO NOT  CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene
 2
 3
 5
 6
in the high-dose group versus treatment days, an estimate of ~320 days (~43 weeks) is obtained
for this group. The extent of deaths prior to 1 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 B-23. Skin tumor incidence in male C3H/HeJ mice dermally
       exposed to benzo [ajpyrene for 24 months
Dose (u,g)a
0 cyclohexanone/acetone
(1:1)
0.05
0.5
5.0
Skin tumor incidence
(all types)"
0/30 (0%)
0/30 (0%)
5/30 (20%)
27/30 (90%)
Number of mice that
died before final
sacrifice
19
15
15
30
Mean survival time,
days
607
630
666
449
 7
 8
 9
10
1 1
12
13
14
15
16
1 7
18
19
20
21
22
23
24
25
Indicated doses were applied twice/week to shaved dorsal skin.
b Number of skin tumor-bearing mice. In the high-dose group, 1 papilloma and 28 carcinomas were
detected. In the 0.5 u.g group, 2 papillomas and 3 carcinomas were detected.

Source: Sivak et al. (1997).

      To examine dose-response relationships and the time course of benzo [a] pyre ne- 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 u,g of benzo [ajpyrene in 50 uL 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 u,g radiolabeled benzo [ajpyrene to determine 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

         This document is a draft for review purposes only and does not constitute Agency policy.
                                         B-66        DRAFT— DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    (more severe in the high-dose group than in the low- and mid-dose groups) and included
 2    statistically significant increases (compared with controls) in: [3H]-thymidine labeling and mitotic
 3    indices; incidence of pyknotic and dark cells (signs of apoptosis); and epidermal thickness. Only a
 4    minor expansion of the epidermal cell population was observed. In the high-dose group, indices of
 5    epidermal damage increased to a plateau by 2 weeks of exposure. The early time course of
 6    epidermal damage indices was not described in the low- or mid-dose groups, since data for these
 7    endpoints were only collected at 20, 24, and 30 weeks of exposure. An increased level of BPDE-
 8    DNA adducts, compared with controls, was apparent in all exposed groups after 4 weeks of
 9    exposure in the following order: 64>32>16>8 [ig/week. The time-course data indicate that
10    benzo[a]pyrene-induced increases in epidermal damage indices and BPDE-DNA adducts preceded
11    the appearance of skin tumors.

12    Reproductive and Developmental Toxicity Studies

13    Oral
14          In a study evaluating the combined effects of DBF and benzo[a]pyrene on the male
15    reproductive tract, Chen et al. (2011) administered benzo[a]pyrene alone in corn oil via daily
16    gavage at 5 mg/kg-day to 30 male Sprague-Dawley rats (28-30 days old); a group of 30 rats
17    received only vehicle.  Body weight was measured weekly.  Groups of 10 rats per group were
18    sacrificed after 4, 8, and 12 weeks of exposure. At sacrifice, blood was collected for analysis of
19    serum testosterone levels by radioimmunoassay. The testes and epidiymides were weighed, and
20    the right testis and epididymis were examined microscopically. The left epididymis was used for
21    evaluation of sperm parameters (sperm count and morphology). Oxidative stress, as measured by
22    superoxide dismutase, glutathione peroxidase, and catalase activity and malondioaldehyde levels,
23    was evaluated in the left testis of each rat Exposure to benzo [a]pyrene did not affect body weight,
24    and no signs of toxicity were seen.  Testes and epididymides weights of exposed rats were similar
25    to controls at all time points. Sperm counts and percent abnormal sperm were also similar to
26    controls at 4 and 8 weeks of exposure, but were significantly (p<0.05) different from controls after
27    12 weeks of exposure to benzo[a]pyrene (29% decrease in sperm count and 54% increase in
28    percent abnormal sperm). Serum testosterone levels were significantly increased relative to
29    controls after 4 weeks (>two-fold higher) and 8 weeks (~1.5-fold higher) of benzo [ajpyrene
30    exposure, but were comparable to controls after 12 weeks. Histopathology evaluation of the testes
31    revealed irregular and disordered arrangement of germ cells in the seminiferous tubules of treated
32    rats; the authors did not report incidence or severity of these changes. Among measures of
33    testicular oxidative stress, only catalase activity was significantly affected by benzo[a]pyrene
34    exposure, showing an increase of ~50% after 12 weeks of exposure. These data suggest a LOAEL of
35    5 mg/kg-day (the only dose tested)  for decreased sperm count, increased percentage of abnormal
36    sperm, altered testosterone levels, and histopathology changes in the testes following 13 weeks of
37    exposure.

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-67        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1           Chung etal. (2011) evaluated the effects of low-dose benzo[a]pyrene exposure on
 2    spermatogenesis, and the role of altered steroidogenesis on the sperm effects.  Groups of 20-25
 3    male Sprague-Dawley rats (8 wks old) were given daily gavage doses of 0, 0.001, 0.01, or 0.1
 4    mg/kg-day benzo[a]pyrene in DMSO for 90 consecutive  days.  At the end of exposure, the animals
 5    were sacrificed for removal of the pituitary, testes, and epididymides, and collection of serum and
 6    testicular interstitial fluid. Subgroups of each exposure group were used for various analyses.
 7    Serum levels of testosterone and LH were measured, as was testosterone concentration in the
 8    interstitial fluid (ELISA assays).  Body and testes weights were recorded. Sections of the testis
 9    were analyzed for apoptotic germ cells using TUNEL assay.  Evaluation of the epididymis included
10    histopathology as well as measurement of caput and caudal epididymal tubule diameters. In
11    addition, sperm were isolated from the cauda epididymis for analysis of sperm number and
12    motility, acrosomal integrity, and immunocytochemistry for ADAMS (a disintegrin and
13    metallopeptidse domain 3; a sperm surface protein associated with fertilization).
14           Leydig cells were isolated from the right testis of animals from each dose group and
15    cultured with or without human chorionic gondatotropin (hCG) or dibutyl cyclic adenosine
16    monophosphate (dbcAMP) to evaluate testosterone production (Chung et al., 2011). Cultured
17    Leydig cells were  also subjected to western blot and immunocytochemistry analyses to evaluate
18    changes in the expression of genes involved in steroidogenesis (StAR[steroidogenic acute
19    regulatory protein], p450scc  [p450 side-chain cleavage], and 3p-HSD[3p-hydroxysteroid
20    dehydrogenase isomerase]).  Finally, pituitary gland extracts were evaluated for LH protein
21    content using immunohistochemistry. Data were reported graphically and analyzed by ANOVA
22    followed by Duncan's post hoc test, using a p-value cutoff of 0.05 for significant difference.
23           At termination of exposure, body weights of treated animals were similar to controls, as
24    were absolute testes weights (Chung et al., 2011). Testosterone concentrations in both serum and
25    testicular interstitial fluid were significantly reduced at the high dose of benzo[a]pyrene (0.1
26    mg/kg-day); based on visual inspection of the data, the mean serum concentration in  this group
27    was ~20% of the control and the mean intersitital fluid concentration was ~60% of the control
28    (n=9 animals/dose for these evaluations).  In addition, baseline production of testosterone by
29    cultured Leydig cells was significantly decreased (~50% based on data shown graphically) at 0.1
30    mg/kg-day. Both hCG- and dbcAMP-stimulated testosterone production measurements were lower
31    (~60% lower than controls) in Leydig cells from rats exposed to either 0.01 or 0.1 mg/kg-day.
32    Serum LH was significantly increased at both 0.01 and 0.1 mg/kg-day (~65-75% higher than
33    controls based on visual inspection of graphs); concordant increases in the intensity of LH
34    immunoreactivity were evident in pituitary extracts from exposed rats.
35           Dose-related increases in the number of apoptotic germ cells, primarily spermatogonia,
36    were demonstrated both via TUNEL assay and caspase-3 staining; the number per tubule was
37    significantly increased over control at all doses (Chung etal., 2011). Numbers of sperm were lower
38    in the treatment groups, but did not differ significantly from the control group. However, sperm

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-68        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    motility was significantly reduced in exposed groups compare with control. The authors did not
 2    report sperm motility for all dose groups, but showed only the significant decrease in the 0.01
 3    mg/kg-day mid-dose group (~30% lower than controls based on visual inspection of graph).
 4    Acrosomal integrity (measured by LysoTracker staining) was diminished in sperm heads from
 5    exposed rats; likewise, the expression of ADAMS protein was downregulated by exposure to
 6    benzo[a]pyrene; the authors reported a significant decrease in the 0.01 mg/kg-day group but did
 7    not provide details of the analysis of other exposure groups. Histopathology examination of the
 8    caput and cauda epididymides revealed dose-related decreases in both cauda and caput tubule
 9    diameters that were statistically significantly lower than controls at all doses (~10-30% smaller
10    mean diameter than control based on measurements of 175 tubules collected from 5 samples in
11    each group; data reported graphically).
12           Statistically significant effects observed at the lowest dose (0.001 mg/kg-day) of
13    benzo[a]pyrene in this study included decreased caput and cauda epididymal tubule diameters
14    (~10-15% lower than controls) and increased numbers of apoptotic germ cells (~twofold higher
15    than controls) by TUNEL assay (Chung etal., 2011). The authors reported that "sperm motility was
16    significantly reduced in the benzo[a]pyrene-exposed groups in comparison to that of the control"
17    but provided quantitative data only for the middle dose group, which exhibited a ~30% decrease in
18    percent motile sperm. No statistically significant decrease in sperm count was reported at any
19    dose. The middle dose (0.01 mg/kg-day) is considered to be a LOAEL, based on reduced sperm
20    motility.
21           Gao etal. (2011) examined effects of benzo[a]pyrene exposure via on cervical cell
22    morphology.  Female ICRmice (18-22 g) were exposed to doses of 0, 2.5, 5, or 10 mg/kg twice per
23    week for 14 weeks, either by oral gavage or by intraperitoneal injection  (for this review, only oral
24    results are reported). After adjustment for equivalent continuous dosing (2/7 days/week), the
25    equivalent daily doses are estimated to be 0.7,1.4, 2.9 mg/kg-day.  Both  vehicle (sesame oil) and
26    untreated control groups were maintained.  Body weights were determined weekly.  Groups of 26
27    mice per dose per exposure route were sacrificed at the end of exposure for evaluation of cervical
28    weight and histopathology. Additional groups of 10 mice were exposed  for 14 weeks and used for
29    determination of lipid peroxidation (malondialdehyde and glutathione-S-transferase levels) and
30    CYP1A1 activity (EROD) in both liver and cervix, as well as creatine kinase activity, AST activity, and
31    IL-6 levels in cervix and serum.
32           Mortality was observed in all exposure groups with the exception of the low dose oral
33    exposure group; the authors did not indicate the timing or causes of death (Gao etal., 2011). There
34    were no control deaths.  Mortality incidences in the oral exposure groups (low to high dose) were
35    0/26 (untreated control), 0/26 (vehicle control), 0/26,1/36, and 2/26.  Benzo[a]pyrene treatment
36    resulted in dose-dependent decreases in body weight gain. In the high dose group of both
37    treatments, body weight began to decline after ~7 weeks of exposure. Based on visual examination
38    of data presented graphically, mean terminal body weights in the low, mid-, and high-dose  oral

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-69        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
10
11
exposure groups were ~10,15, and 30% lower (respectively) than the vehicle control mean. The
untreated control mean body weight for the oral exposure group was similar to the vehicle control
mean body weight. Cervical weight as a function of body weight was not affected by oral
benzo[a]pyrene exposure. Microscopic examination of the cervix revealed increased incidences of
epithelial hyperplasia and inflammatory cells in the cervix of all groups of exposed mice, and
atypical hyperplasia of the cervix in mice exposed to 1.4 or 2.9 mg/kg benzo[a]pyrene. Statistical
analysis of the findings was conducted, but was poorly reported in the publication. Table B-24
shows the incidences in the oral exposure  groups, along with the results of Fisher's exact tests
performed for this review.

       Table B-24.  Mortality and cervical histopathology incidences in female
       ICR mice exposed to benzo[a]pyrene via gavage for 14 weeks
Endpoint
Mortality
Cervical epithelial hyperplasia
Atypical hyperplasia of cervix
Inflammatory cells in cervix
Dose (mg/kg-d)
Untreated
control
0/26
0/26
0/26
2/26
Vehicle
control
0/26
0/26
0/26
3/26
0.7
0/26
4/26
0/26
10/26a
1.4
1/26
6/25a
2/25
12/25a
2.9
2/26
7/24a
4/24a
18/24a
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Significantly different from vehicle control by Fisher's exact test performed for this review (one-sided p
<0.05).

Source:  Gao et al. (2011).

       Levels of malondialdehyde in both the cervix and liver were significantly higher than
controls in all dose groups of animals treated by either oral (1.5 to 2-fold higher in the cervix and
~3-fold to 7-fold higher in the liver after oral exposure p<0.05) or intraperitoneal exposure.
Concomitant decreases in GST activity (~15% to 50% lower than controls in the cervix and ~30%
to 60% lower in the liver after oral exposure; p<0.05) were also observed at all doses and in both
organs and both treatments. EROD activity was increased in the cervix (~4- to ~12-fold) and liver
(~12- to ~35-fold) of all exposure groups. Measurement of CK and AST activity in the cervix and
serum also showed significant increases at all doses and after both exposures (~1.5- to 2-fold in the
cervix, and ~20% to 50% higher than controls in the liver after oral exposure). Finally, levels of the
inflammatory cytokine IL-6 were significantly (p<0.05) increased in the cervix of all treated mice,
and were markedly increased (from more than two-fold higher than untreated or vehicle controls
at the low dose, to ~six-fold higher at the high dose) in the serum of treated mice.
       Based on the observations of decreased body weight and increased cervical epithelial
inflammation and hyperplasia, a LOAEL of 0.7 mg/kg-day (the lowest dose tested) is identified for
this study.

          This document is a draft for review purposes only and does not constitute Agency policy.
                                          B-70        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1          Mohamed etal. (2010) investigated multi-generational effects in male mice following
 2    exposure of 6-week old-C57BL/6 mice (10/group) to 0 (corn oil), 1, or 10 mg/kg-day
 3    benzo[a]pyrene for 6 weeks by gavage. Following final treatment, male mice were allowed to
 4    stabilize for 1 week prior to being mated with two untreated female mice to produce an
 5    Fl generation.  Male mice were sacrificed 1 week after mating. Fl males were also mated with
 6    untreated female mice as were F2 males. The mice of the Fl, F2, and F3 generations were not
 7    exposed to benzo[a]pyrene. The FO, Fl,  F2, and F3 mice were all sacrificed at the same age
 8    (14 weeks) and endpoints including testis histology, sperm count, sperm motility, and in vitro
 9    sperm penetration (of hamster oocytes)  were evaluated. These endpoints were analyzed
10    statistically using analysis of variance (ANOVA) and Tukey's honest significance test and results
11    were reported graphically as means ± SD.
12          Testicular atrophy was observed in the benzo[a]pyrene treatment groups, but was not
13    statistically different than controls. Statistically significant reductions were observed in epididymal
14    sperm counts of FO and Fl generations treated with the high or low dose of benzo[a]pyrene. For FO
15    and Fl generations, epididymal sperm counts were reduced approximately 50 and 70%,
16    respectively, in the low- and high-dose groups. Additionally, sperm motility was statistically
17    significantly decreased at the high dose in the FO and Fl generations. Sperm parameters of the F3
18    generation were not statistically different from controls. An in vitro sperm penetration assay
19    revealed statistically significantly reduced fertilization in FO and Fl generations of the low- and
20    high-dose groups. However, the value of this in vitro test is limited as it bypasses essential
21    components of the intact animal system  (U.S. EPA, 1996). Based on decreased epididymal sperm
22    counts of FO and Fl generations, a LOAEL of 1 mg/kg-day was established from this study (no
23    NOAEL was identified).
24          Arafa et al. (2009) exposed groups of 12 male Swiss albino rats to benzo[a]pyrene in olive
25    oil (0 or  50 mg/kg-day via gavage) for 10 consecutive days, either alone or after similar treatment
26    with  200 mg/kg-day of the flavonoid hesperidin, which has been shown to exert anti-inflammatory,
27    antioxidant, and anticarcinogenic activity. One day after the final dose, the animals were sacrificed
28    for removal of the cauda epididymides and testes. Epididymal sperm count and motility were
29    assessed, as was daily sperm production in the testes. The study authors also investigated the
30    testicular activity of LDH, SOD, and GST,  as well as GSH, malondialdehyde, and protein content. The
31    testes were examined under light microscope.
32          Relative testes weights (normalized to body weight) of benzo[a]pyrene exposed-animals
33    were significantly decreased compared with controls (35% lower, p < 0.05) (Arafa et al., 2009).  In
34    addition, exposure to benzo[a]pyrene alone resulted in significantly decreased sperm count,
35    numbers of motile sperm, and daily sperm production (~40% decrease from control in each
36    parameter, p <  0.05).  Effects on sperm count and production were abolished by hesperidin
37    pretreatment, but the  number of motile sperm remained significantly depressed (compared with
38    the control group) in the group exposed  to both benzo[a]pyrene and hesperidin. Measures of

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-71         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    antioxidant enzymes and lipid peroxidation showed statistically significant induction of oxidative
 2    stress in the testes of benzo[a]pyrene-exposed rats. With the exception of the decrease in testicular
 3    GSH content (which was partially mitigated), pretreatment with hesperidin eliminated the effects of
 4    benzo[a]pyrene on lipid peroxidation and antioxidant enzymes.
 5          Xu et al. (2010) treated female Sprague-Dawley rats (6/group) to 0 (corn oil only),  5, or 10
 6    mg/kg-day benzo[a]pyrene by gavage every other day for a duration of 60 days. This resulted in
 7    TWA doses of 0, 2.5, and 5 mg/kg-day over the study period of 60 days. Endpoints examined
 8    included ovary weight, estrous cycle, 17B-estradiol blood level, and ovarian follicle populations
 9    (including primordial, primary, secondary, atretic, and corpora leutea). Animals were observed
10    daily for any clinical signs of toxicity and following sacrifice, gross pathological examinations were
11    made and any findings were recorded. All animals survived to necropsy. A difference in clinical
12    signs was not observed for the treated groups and body weights were not statistically different in
13    treated animals (although they appear to be depressed 6% at the high dose).  Absolute ovary
14    weight was statistically significantly reduced in both the low- and high-dose groups (11 and 15%,
15    respectively) (see Table B-25). Animals treated with the high dose were noted to have a
16    statistically significantly prolonged duration of the estrous cycle and nonestrus phase compared to
17    controls. Animals in the high-dose group also had statistically significantly depressed levels of
18    estradiol (by approximately 25%) and decreased numbers of primordial follicles (by approximately
19    20%). This study also indicated a strong apoptotic response of ovarian granulosa cells as visualized
20    through TUNEL labeling; however, the strongest response was seen at the low dose; decreased
21    apoptosis was also observed at the high dose. Based on decreased ovary weight following a 60-day
22    oral exposure to benzo[a]pyrene, a LOAEL of 2.5 mg/kg-day was established from this study (no
23    NOAEL was identified).

24          Table B-25. Means ± SD for ovary weight in female Sprague-Dawley rats

Ovary weight (g)
Body weight (g)
Dose (mg/kg-d)a
0
0.160 ±0.0146
261.67 ±12.0
2.5
0.143 ±0.0098b
249. 17 ±11. 2
5
0.136 ±0.0098b
247.25 ± 11.2
      aTWA doses over the 60-day study period.
      Statistically different from controls (p < 0.05) using one-way ANOVA.

      Source: Xuetal. (2010).
25
26          Zheng et al.  (2010) treated male Sprague-Dawley rats to 0 (corn oil only), 1, or 5 mg/kg-day
27    benzo[a]pyrene by daily gavage for a duration of 30 (8/group) or 90 days (8/group). At necropsy,
28    the left testis of each animal was collected and weighed. Testes testosterone concentrations were
29    determined by radioimmunassay and results were expressed as ng/g testis and reported
30    graphically. Testicular testosterone was statistically significantly decreased in the high-dose group

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 B-72         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    approximately 15% following 90 days of exposure. The low-dose group also appeared to have a
 2    similar average depression of testosterone levels; however, the change did not reach statistical
 3    significance. Testosterone levels measured in animals sacrificed following 30 days of
 4    benzo[a]pyrene exposure were not statistically different than controls. Based on decreased
 5    testicular testosterone levels following a 90-day oral exposure to benzo[a]pyrene, a LOAEL of 5
 6    mg/kg-day and a NOAEL of 1 mg/kg-day were identified.
 7          McCallister et al. (2008) administered 0 or 300 [J.g/kg benzo[a]pyrene by gavage in peanut
 8    oil to pregnant Long Evans rats (n = 5 or 6) on CDs 14-17.  At this exposure level, no significant
 9    changes were see in number of pups per litter, pup growth, or liver to body weight ratios in control
10    compared to benzo[a]pyrene exposed offspring. Treatment-related differences in brain to body
11    weight ratios were observed only on PNDs 15 and 30.  Decreases in cerebrocortical mRNA
12    expression of the glutamatergic N-methyl-D-aspartate (NMDA) receptor subunitwas significantly
13    reduced (50%)  in treated offspring compared to controls. In addition, in utero exposed offspring
14    exhibited decreased evoked cortical neuronal activity in the barrel field cortex when tested at PNDs
15    90-120.
16          Rigdon and Neal (1965) administered diets containing 1,000 ppm benzo[a]pyrene to
17    pregnant mice (nine/group) on CDs 10-21 or 5-21. The pups were reported as appearing
18    generally normal at birth, but cannibalism was elevated in the exposed groups.  These results are in
19    contrast with an earlier study (Rigdon and Rennels, 1964) in which rats (strain not specified) were
20    fed diets containing benzo[a]pyrene at 1,000 ppm for approximately 28 days prior to mating and
21    during gestation. In the earlier study, five of eight treated females mated with untreated males
22    became pregnant, but only one delivered live young. The treated dam that delivered had two live
23    and two stillborn pups; one dead pup was grossly malformed. In the remaining treated females,
24    vaginal bleeding was observed on CDs 23 or 24.  In the inverse experimental design, three of six
25    controls mated  to benzo[a]pyrene-treated males became pregnant and delivered live young.
26    Visceral and skeletal examinations of the pups were not conducted. These studies were limited by
27    the small numbers of animals, minimal evaluation of the pups, lack of details on days of treatment
28    (food consumption, weight gain), and the occurrence of cannibalism.

29    Reproductive effects of in utero exposure via oral route
30          MacKenzie and Angevine (1981) conducted a two-generation reproductive and
31    developmental toxicity study for benzo[a]pyrene in CD-I mice. Benzo[a]pyrene was administered
32    by gavage in 0.2 mL of corn oil to groups of 30 or 60 pregnant (the FO generation) mice at doses of
33    0,10, 40, or 160 mg/kg-day on CDs 7-16 only. Therefore, unlike the standard two-generation
34    study, Fl animals were exposed only in utero. Fl offspring were evaluated for postnatal
35    development and reproductive function as follows. Fl pups (four/sex when possible) were allowed
36    to remain with their mothers until weaning on PND 20. Crossover mating studies were then
37    conducted. Beginning at 7 weeks of age, each Fl male mouse (n = 20-45/group) was allowed to
38    mate with two untreated virgin females for 5-day periods for 25  days (for a total exposure of 10
               This document is a draft for review purposes only and  does not constitute Agency policy.
                                                B-73        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1    untreated females/Fl male), after which time the males were separated from the females.
 2    Fourteen days after separation from the males (i.e., on days 14-19 of gestation), the females were
 3    sacrificed and the numbers of implants, fetuses, and resorptions were recorded. The F2 fetuses
 4    were then examined for gross abnormalities.  Similarly, each Fl female mouse (n = 20-55/group),
 5    beginning at 6 weeks of age, was paired with an untreated male for a period of 6 months.  Males
 6    were replaced if the females failed to produce a litter during the first 30-day period. All F2 young
 7    were examined for gross abnormalities on day 1 of life and their weights were recorded on day 4 of
 8    age. This F2 group was sacrificed on day 20 postpartum, while the Fl female was left with a male
 9    until the conclusion of the study. At 6 weeks of age, gonads of groups of 10 male and 10 female Fl
10    mice exposed to 0,10, or 40 mg/kg-day benzo[a]pyrene in utero were subjected to gross pathology
11    and histologic examinations.
12          No maternal toxicity was observed. The number of FO females with viable litters at
13    parturition at the highest dose was statistically significantly reduced by about 35% (Table B-2 6),
14    but progeny were normal by gross observation. Parturition rates of the low- and mid-dose groups
15    were unaffected by treatment, and litter sizes of all treated groups were similar to the control group
16    throughout lactation. However, body weights of the Fl pups in the mid- and high-dose groups were
17    statistically significantly decreased on PND 20, by 7 and 13%, respectively, and in all treated pups
18    on PND 42, 6, 6, and 10% for the low, mid, and high dose, respectively (Table B-26). The number of
19    Fl pups surviving to PNDs 20 and 42 was significantly reduced at the high dose (p < 0.01), by 8 and
20    16%, respectively. When Fl males were bred to untreated females and Fl females were mated
21    with untreated males, a marked dose-related decrease in fertility of >3 0% was observed in both
22    sexes, starting at the lowest exposure.  There were no treatment-associated gross abnormalities or
23    differences in body weights in the F2 offspring.

24          Table B-26. Reproductive effects in male and female CD-I Fl mice
25          exposed in utero to benzo[a]pyrene
Effect
FO mice with viable litters at parturition
Mean ± SEM pup weight (g) at PND 20
Mean ± SEM pup weight (g) at PND 42
Fl male fertility indexc
Fl female fertility indexd
Dose (mg/kg-d)a
0
46/60 (77%)
11.2 ±0.1
29.9 ±0.2
80.4
100.0
10
21/30 (70%)
11.6 ±0.1
28.2±0.3b
52.0b
65.7b
40
44/60 (73%)
10.4±0.1b
28.0±0.2b
4.7b
0.0b
160
13/30 (43%)b
9.7±0.2b
26.8 ± 0.4b
0.0b
0.0b
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-74        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
      aPregnant FO mice were administered daily doses of benzo[a]pyrene in corn oil on GDs 7-16.
      Significantly (p < 0.05) different from control by unspecified tests.
      Beginning at 7 weeks of age, each Fl male mouse (20-45/group) was exposed to 10 untreated females
      over a period of 25 days.  Index = (females pregnant/females exposed to males) x 100.
      dBeginning at 6 weeks of age, each Fl female mouse (20-55/group) was cohabitated with an untreated
      male for a period of 6 months.

      SEM = standard error of the mean

      Source:  MacKenzie and Angevine (1981).
 1
 2          Exposure to benzo[a]pyrene caused a marked dose-related decrease in the size of the
 3    gonads.  In Fl males, testes weights were statistically significantly reduced. Testes from animals
 4    exposed in utero to 10 and 40 mg/kg-day weighed approximately 60 and 18%, respectively, of the
 5    weight of testes from the control animals (no F2 offspring were produced in the high-dose group).
 6    This was confirmed by histopathologic observation of atrophic seminiferous tubules in the
 7    40 mg/kg-day group that were smaller than those of controls and were empty except for a basal
 8    layer of cells. The number of interstitial cells in the testes was also increased in this group. Males
 9    from the 10 mg/kg-day group showed limited testicular damage; although all  exhibited evidence of
10    tubular injury, each animal had some seminiferous tubules that displayed active spermatogenesis.
11    Ovarian tissue was absent or reduced in Fl females such that organ weights were not possible to
12    obtain. Examination of available tissue in these females revealed hypoplastic  ovaries with few
13    follicles and corpora lutea (10 mg/kg-day) or with no evidence of folliculogenesis (40 mg/kg-day).
14    Ovarian tissue was not examined in highest-dose females.
15          The LOAEL in this study was 10 mg/kg-day, based on decreases in mean pup weight (<5%)
16    atPND 42 of Fl offspring of dams treated with 10, 40, or 160 mg/kg-daybenzo[a]pyrene, marked
17    decreases in the reproductive capacity (as measured by fertility index) of both male and female Fl
18    offspring exposed at all three treatment levels of benzo[a]pyrene (by approximately 30% in males
19    and females), decreased litter size (by about 20%) in offspring of Fl dams, and the dramatic
20    decrease in size and alteration in anatomy of the gonads of both male and female Fl mice exposed
21    to 10 and 40 mg/kg-day benzo[a]pyrene in utero. A NOAEL was not identified.
22          In another reproductive and developmental toxicity study, benzo[a]pyrene was
23    administered by gavage in corn oil to nine female NMRI mice at a dose of 10 mg/kg-day on GDs 7-
24    16; a group of nine controls received corn oil (Kristensen et al., 1995). Body weights were
25    monitored. FO females were kept with their offspring until after weaning (21  days after delivery).
26    At 6 weeks of age, one Fl female from each litter (n = 9) was caged with an untreated male. The
27    F2 offspring were inspected for gross deformities at birth, weight and sex were recorded 2 days
28    after birth, and the pups were sacrificed. The Fl females were  sacrificed after 6  months of
29    continuous breeding. The effects of benzo[a]pyrene treatment on fertility, ovary weights, follicles,
30    and corpora lutea were evaluated. FO females showed no signs of general toxicity, and there was no

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-75        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
      effect on their fertility. Fl 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 B-27). At necropsy, the Fl females from
      treated FO 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 B-27). Only one dose group was used in this study, with decreased Fl female fertility
      observed following in utero exposure at the LOAEL of 10 mg/kg-day; no NOAEL was identified.

             Table B-27. Effect of prenatal exposure to benzo[a]pyrene on indices of
             reproductive performance in Fl female NMRI mice
Endpoint (median with range in parentheses)
Number of F2 offspring
Number of F2 litters
F2 litter size (number of pups per litter)
Number of d between F2 litters
Fl ovary weight (mg)
Number of small follicles
Number of medium follicles
Number of large follicles
Number of corpora lutea
Control3
92 (26-121)
8 (3-8)
11.5 (6-15)
20.5 (20-21)
13 (13-20)
44 (1-137)
9 (5-25)
14 (6-23)
16 (6-35)
Benzo[a]pyrene
exposeda(10 mg/kg-d)
22b (0-86)
3b (0-8)
8b (3-11)
21b (20-23)
9b (7-13)
Ob (0-68)
Ob (0-57)
Ob (0-19)
Ob (0-14)
10
11
12
13
14
15
16
17
18
19
20
21
22
     aGroups of nine female NMRI FO mice were administered 0 or 10 mg benzo[a]pyrene/kg-day by gavage
     in corn oil on GDs 7-16. One Fl female from each litter was continuously bred with an untreated male
     for 6 months.
     Significantly (p < 0.05) different from control group by Wilcoxon rank sum test or Kruskall-Wallis two-
     tailed test.

     Source: Kristensen et al. (1995).

           Chen et al., (2012) treated male and female neonatal Sprague-Dawley rats (10/sex/group)
     with benzo[a]pyrene (unspecified purity) dissolved in peanut oil by gavage daily from post-natal
     day (PND) 5 -  11,  at doses of 0.02, 0.2 or 2 mg/kg in 3 mL vehicle/kg b.w., determined individually
     based upon daily measurements. This time period was described as representing the brain growth
     spurt in rodents, analogous to brain developmental occurring from the third trimester to 2 years of
     age in human infants. Breeding was performed by pairs of nine week old rats, with delivery
     designated as PNDO.  Litters were culled to 8 pups/dam (4/ea male and female, when possible) and
     randomly redistributed at PND1 among the nursing dams; dams themselves were rotated every 2-3
     days to control for caretaking differences, and cage-side observations of maternal behavior were
     made daily. One male and female from each litter were assigned per treatment group, and the
     following physical maturation landmarks were assessed daily in all treatment groups until weaning
     atPND21: incisor  eruption, eye opening, development of fur, testis decent and vaginal opening.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-76         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1          Neonatal sensory and motor developmental tests were administered to pups during the
 2    preweaning period at PNDs 12,14,16 and 18, and were behavioral tests administered to rats as
 3    adolescents (PND 35, 36) or as adults (PND 70, 71): each rat was only tested during one
 4    developmental period. All dosing was performed from 1300 - 1600 hrs, and behavioral testing was
 5    during the "dark" period from 1900 - 2300 hrs, although tests were performed in a lighted
 6    environment.  Pups were observed individually and weighed daily, the order of testing litters was
 7    randomized each day, and all observations were recored by investigators blinded to group
 8    treatment.
 9          Sensory and motor developmental tests including the surface righting reflex test, negative
10    geotaxis test, and cliff aversion test were performed only once, while the forelimb grip strength test
11    was assessed during three 60 second trials on PND12. Rat movements during the open-field test
12    were recorded by camera, and two blinded investigators scored movement and rearing separately
13    during a 5 min. evaluation period. Blinded investigators directly observed video monitoring of rat
14    movements during the elevated plus maze, and after a 5  min. free exploration period, recored
15    number of entries into the closed and open arms, the time spent in the open arms, and latency to
16    the first arm entry. Assessment of the Morris water maze was slightly different, in that the rats
17    were habituated to the testing pool by a 60 second swim without a platform on the day prior to
18    testing. The rats were then tested during a 60 second swim with a hidden platform present at a
19    constant position each day for four days; on the fifth day, the rats were evaluated during a 60
20    second probe swim without a platform. The number of times each animal crossed the original
21    platform location and the duration of time spent in the platform quadrant were recorded during
22    this final evaluation.  One pup/sex/litter were assigned for behavioral testing to each of four tracks:
23    Track 1, surface righting reflex test, cliff aversion test, and open-field test (PND 12-18); Track 2,
24    negative geotaxis test, forelimb grip strength test, and open-field test (PND 12 - 20); Track 3,
25    elevated plus maze, Morris water maze, and open-field test (PND 34 - 36); Track 4, elevated plus
26    maze, Morris water maze, and open-field test (PND 69 -  71). All results were presented in
27    graphical  form only.
28          No significant effects on pup body weight were observed during the 7-day treatment period
29    (PND 5 - 11).  Three-way ANOVA (time xB[a]P treatment x sex) indicated that effects of B[a]P were
30    not sex-dependent throughout the 71 day experiment, so both sexes were pooled together.  From
31    this pooled analysis, pups in the 2 mg/kg treatment group gained significantly less weight at both
32    PND36 and PND71.  There were no differences among treatment groups in incisor eruption, eye
33    opening, development of fur, testis decent or vaginal opening.
34          For all measurements of neonatal sensory and motor development, results from both sexes
35    were analyzed together since B[a]P was reported to have no significant interaction with sex by 3-
36    way ANOVA. No significant differences were observed in either the cliff aversion or forelimb grip
37    strength tests. In the surface righting reflex test, latency was increased in the 0.2 mg/kg group at
38    PND12, in the  0.02 and 2 mg/kg groups at PND14, in only the high dose 2 mg/kg group at PND16,

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-77        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    and was not significantly different in any group at PND18. At PND12 there was a dose-related
 2    increase in negative geotaxis latency associated with 0.02, 2 and 2 mg/kg B[a]P, which was also
 3    present in the 2 mg/kg group at PND14, but returned to control levels at PND16 and PND18. In the
 4    open field test, there were no significant differences in either locomotion or rearing activity at
 5    PND18 or 20. At PND34, the 2 mg/kg group exhibited significantly increased movement, but
 6    increases in rearing were not significant. At PND69, increased locomotion was observed in both the
 7    0.2 and 2 mg/kg groups, while rearing was significantly increased in only the 2 mg/kg treatment
 8    group.
 9          The elevated plus maze performance was only evaluated in adolescent and adult rats.
10    Unlike the previous tests, 3-way ANOVA revealed a statistically significant interaction between
11    neonatal B[a]P treatment and sex, so male and female performance was analyzed independently.
12    No significant differences in PND35 males were observed, and the only significant observation in
13    PND35 females was increased time spent in the open maze arms by 2 mg/kg treatment group.
14    Significantly decreased latency time to first open arm entry was observed  in PND70 males and
15    females in both 0.2 and 2 mg/kg treatment groups; these groups also spent significantly more time
16    in open maze arms, along with the 0.02 mg/kg female group. PND70 2 mg/kg males, along with 0.2
17    and 2 mg/kg females, entered more frequently into open arms and less frequently into closed arms
18    than vehicle controls. In the Morris water maze, escape latency (time to reach the platform during
19    each of the four testing days) was consistently increased in the 2 mg/kg treatment group of both
20    sexes, in both adolescent and adult animals. These increases were statistically significant in both
21    males and females treated with 2 mg/kg B[a]P at both PND39 and PND74, and were also
22    significantly elevated in 0.2 mg/kg animals of both sexes at PND74. Likewise, performance during
23    the fifth test day, in the absence of the escape platform, was significantly adversely affected by both
24    metrics (decreased time spent in the target quadrant and decreased number of attempts to cross
25    the platform location) in 2 mg/kg rats of both sexes at both PND40 and PND75.  PND75 females
26    treated with 0.2 mg/kg B[a]P also showed significant decreases in both performance metrics, while
27    PND75 0.2 mg/kg males only demonstrated significant differences in "time spent in target
28    quadrant". Swim speed was also assessed, but there were no differences among any treatment
29    group at either age evaluated.
30          Jules et al., (2009) treated pregnant Long Evans Hooded (LEH) rats with benzo[a]pyrene
31    (unspecified purity) dissolved in 0.875 mL peanut oil by gavage daily from GD14 - GD17, at doses of
32    150, 300, 600 and 1,200 |ig B[a]P /kgb.w., with animals weighed daily. Cage-side observations
33    were performed until pup weaning, and litter size evaluated for each treatment group. Pups from 4
34    - 5 individual litters were analyzed for each endpoint, which was independently repeated for a total
35    of 3 replicates. Delivery was designated PNDO, and pups were harvested from PNDO - 15 for B[a]P
36    metabolite identification, or for other endpoints as young adults at PND53. Systolic/diastolic blood
37    pressure and heart rate was recorded by a volume pressure recording sensor and occlusion tail-cuff
38    applied to conscious, non-anesthetized animals. Animals were preconditioned to the restraint

                This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-78        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    device and tail-cuff by daily acclimatization sessions during PND46 - 50, to minimize stress effects
 2    during data collection. Cardiac function values were averaged from 15 readings each collected over
 3    a 1 minute interval every other minute for 30 minutes on PND53. Whole blood was collected from
 4    the heart and aorta prior to surgical resection and tissue processing. Plasma and heart tissue B[a]P
 5    metabolite content was quantified by reverse-phase HPLC with UV and fluorescence detection,
 6    while heart and aortic tissue was subjected to SDS-PAGE for qualitative protein analysis, and RNA
 7    extraction. Quantitative RT-PCR was performed for levels of angiotensin II (Angll), neuronal NOS
 8    (nNOS), endothelial NOS (eNOS) and 7,8-Dihydrobiopterin oxidoreductase (BH4/BH2
 9    oxidoreductase). Total RNA was also used to probe a cDNA microarray, and targets with > 2-fold
10    changes in expression were subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG) and
11    Gene Ontology (GO) biological process pathway analysis.
12          No significant differences in litter size or pup weight gain from PNDO - 15 were reported in
13    any treatment group, and no convulsions, tremors or abnormal movements were reproducibly
14    observed. Most analytical data was reported graphically, as mean ± SEM of three replicates of 3 - 5
15    offspring measured/group. Plasma and heart tissue total B[a]P metabolite levels were maximal at
16    PNDO (the first time point sampled) and progressively decreased from PNDO - 13. Compared to the
17    low-dose group (150 |J.g/kg), plasma metabolite levels were significantly elevated in the 600 and
18    1,200 |ig/kgB[a]P groups through PND13, while heart metabolite levels  were significantly
19    increased through PND11.  Metabolites in mid-dose group, 300 [J.g/kg, trended between the 150
20    and 600 [ig/kg group levels from PNDO - 7, while not achieving statistically significant differences
21    in pair-wise comparisons.  Three principle groups of B [a]P  metabolites were identified. More than
22    70% of the total heart metabolite burden was composed of diol metabolites through PND13, while
23    the more reactive hydroxyl metabolites increased in relative composition from PND9 - 13, and the
24    dione population remained constant at < 5%.
25          Cardiovascular function was evaluated in pups exposed in utero to 600 or 1,200 [ig/kg B[a]P
26    vs. controls. A dose-related and statistically significant increase in both systolic (20, 50%) and
27    diastolic pressure (30, 80%) was observed in mid and high-dose pups, respectively. Heart rate was
28    also significantly altered; a 10% increased heart rate was reported in the 600 [ig/kg B[a]P group,
29    while the average heart rate of the 1,200 [ig/kg B[a]P groups decreased 8%. Cardiac tissue eNOS
30    protein levels fluctuated as a result of B [a] P  treatment; in both the 600 and 1,200 [J.g/kg groups,
31    eNOS expression by semi-quantitative SDS-PAGE was significantly decreased at PNDO and PND5,
32    while it was significantly elevated above controls atPNDIO and PND15.  While eNOS expression in
33    the 600 [ig/kg B[a]P group had returned to control levels by PND53, eNOS expression was
34    significantly higher (approximately 2-fold) in the 1,200 [ig/kg group. Compared to vehicle-treated
35    controls, cardiac message levels of nNOS and eNOS were not significantly affected by B[a]P
36    treatment at PNDO, and while nNOS mRNA levels were 2-fold higher at PND53 in the 600 [ig/kg
37    group, and eNOS mRNA was 3-fold higher in the 1,200 [J.g/kg group, consistent with the increased
38    eNOS protein levels detected at PND53.  Message levels of BH4/BH2 oxidoreductase were

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-79         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    suppressed in both B[a]P treatment groups at PNDO, and while mRNA expression remained
 2    suppressed at PND53 in the 600 ug/kg group, BH4/BH2 message returned to control levels in the
 3    1,200 ug/kg group. Angiotensin II mRNA levels were 1.8-fold higher in both B[a]P groups at PNDO,
 4    and while expression increased to 5-fold more than controls at PND53 in the 600 ug/kg group,
 5    Angll expression remained closer to 1.5-fold greater in the high-dose group. The following
 6    pathways were identified as being enriched by 1,200 ug/kg B[a]P treatment in utero using KEGG
 7    analysis, and correcting for multiple comparisons using the false-discovery rate method: PPARy,
 8    renin-angiotensin system (Angll, adiponectin C1Q and collagen domain, adrenergic(33R, tachykinin
 9    Rl), hematopoietic cell lineage, CYP450 metabolism (CYP2a2, CYP7al and CYP2bl2), retinol
10    metabolism, cell adhesion molecules-CAMs, primary bile acid biosynthesis and tight junctions.

11          Table B-28. Exposure-related effects in Long Evans Hooded  rats
12          exposed to benzo[a]pyrene by gavage daily in utero from GD14 - GDI?
Effect measured
Heart rate (bmp; mean ± SEM)
Dose (mg/kg-d)
0
504.6 ±15. 7
0.600
554.6 ±26. 2*
1.20
466.3 ± 16.9*
Blood pressure measured by tail cuff (mmHg; mean ± SEM)
Systolic pressure
Diastolic pressure
131.6 ±1.2
85.0 ±4.2
151.6 ±45*
113.0 ±3.3*
200.4 ±2.4*
155. 6 ±3. 2*
      *Significantly (p < 0.05) different from control mean; n = 4-5/replicate, 3 replicates performed.

      Source: Jules et al. (2012).
13
14          Bouayed et al., (2009) treated nursing female Swiss Albino OF1 mice (5/dose group) with
15    benzo[a]pyrene (unspecified purity) dissolved in avocado oil by gavage daily while nursing pups
16    from PND1 - 14 at 0, 2 or 20mg/kg-day in 10 mL/kg b.w., individually determined each day. Prior
17    to benzo[a]pyrene treatment, Swiss Albino litters were culled to 10 pups (5/sex when possible),
18    and nurturing females assigned to litters that were stratified randomly to achieve equivalent mean
19    pup litter body weights across the designated treatment groups. As the effects of B[a]P on maternal
20    nurturing behavior was unknown, dam behavior was visually monitored daily until weaning.
21    Furthermore, maternal nurturing performance from PNDO - 21 was assessed by two methods: a
22    nest-building test administered q.2.d., where nest quality/complexity was scored 15 minutes after
23    cotton material was supplied; and pup retrieval, in which latency to return the displaced pup to the
24    nest was measured twice and averaged, was evaluated q.d. At the indicated times 2 mice/sex/litter
25    were randomly selected,  weighed, and brains resected for later mRNA expression analysis (n =
26    20/group).
27          Pup neuromotor  maturation and behavior was assessed during pre-weaning by four
28    standard methods (administered between 1000 - 1300 on testing days, and in temporal order as
29    indicated): 1) righting reflex test, maximum duration 120 seconds, administered on PNDs 3, 5, 7 and

                This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-80         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    9; 2) negativegeotaxis test, maximum duration 120 seconds, administered on PNDs 5, 7, 9 and 11;
 2    3) forelimb grip test, duration until failure, administered on PNDs 9 and 11; and 4) open field test, 6
 3    minute evaluation of locomotor activity and rearing following a 1 minute habituation period,
 4    administered on PND15. Adolescent function was evaluated by three methods: water escape pole
 5    climbing (WESPOC) test, administered at PND20, in which the time to find the pole, time to climb the
 6    pole, and the time to reach the safety platform were reported; elevated plus maze, administered at
 7    PND32 for 5 minutes, in which the latency time to first open arm entry, number of entries into open
 8    arms, total number of entries, percent of time spent in open arms, and percent of entries into open
 9    arms was determined; and Y-maze spontaneous alternation test, administered at PND40 for 5
10    minutes, in which the % spontaneous alternation was calculated by: [(the number of successful
11    overlapping triplets)/(total number of arm entries - 2) x 100%].
12          Benzo[a]pyrene treatment did not significantly affect the body weight of nursing mothers
13    during the 2 week treatment period. Since three-way AN OVA indicated that changes in pup weight
14    as a result of B[a]P treatment were not sex-dependent, data from male and female pups were
15    combined. B[a]P treatment of nursing mothers was associated with a 8-9% weight gain in pups
16    nursing from the 2 mg/kg group, and a 10-12% weight gain in pups from the 20 mg/kg group at
17    PND12 - 20. While not significantly different from PND26 - 40, pup weight in the 20 mg/kg group
18    was continuously higher than either the 2 mg/kg group or vehicle-treated controls. There were no
19    significant differences in pup brain weight or eye opening observed. Likewise, B[a]P treatment of
20    nursing mothers did not affect nest-building interest or quality, and while not significantly
21    impacting pup retrieval time, the retrieval latency period was observed to increase with increasing
22    treatment duration in both B[a]P groups vs. controls.
23          Behavioral test data was reported graphically, as mean ± SEM of n = 20/group. For the pre-
24    weaning neuromotor developmental tests, B[a]P treatment was found to not depend on sex, and so
25    data from male and female pups was combined.  Pups nursing from mothers administered 2 or 20
26    mg/kg-day B [a]P had significantly elevated righting reflex times at PNDS - 5, which decreased to
27    control times at PND7 - 9.  Only pups from the 20 mg/kg treatment group demonstrated
28    significantly increased negative geotaxis latency, which was 2-fold greater than controls at PNDs 5,
29    7 and 9, but returned to control levels at PND11. Interestingly, mice in the 20 mg/kg group had
30    increased forelimb grip strength, which was significantly greater than control mice at PND9 and 11,
31    corresponding to increased body weight in the B[a]P-treated mice vs. controls.  Mice in the 2 mg/kg
32    group  also performed better than controls at PND9, but were equivalent at PND11. No treatment or
33    sex-related effects were reported on locomotion or rearing activity during the open field test Sex-
34    dependency on test performance became evident during the analysis of the WESPOC test data:
3 5    female pups were not significantly affected using any metric, while males in the 2 0 mg/kg group
36    demonstrated a statistically significantly longer pole-grasping latency (3-fold), and took 13-times
37    longer to escape the pole and board the safety platform, vs. vehicle controls. While performance of
38    male pups from the 2 mg/kg group was not statistically significantly worse than vehicle controls by

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-81        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15

16
17
18
      pair-wise comparison, latency for both pole-grasping and escape in this treatment group
      contributed to a significant trend for treatment dose and these effects. In the evaluation of the
      elevated plus maze, treatment effects did not appear to depend upon sex, so both male and female
      performance was analyzed together. Mice in both B[a]P treatment groups demonstrated
      significantly decreased latency time to first entering an open arm (30 - 50%), as well as entered
      open arms 2-times more frequently and spent twice as much time there vs. vehicle controls. While
      mice in the 2 mg/kg treatment group entered into closed arms 20% less frequently than controls,
      mice in the 20 mg/kg group were not significantly different. Likewise, mice nursing from mothers
      treated with 2 mg/kg B[a]P performed 15% more spontaneous alternations in the Y-maze
      spontaneous alternation test compared to controls, while mice in the high-dose group were not
      significantly different The brains of pups nursing from the 20 mg/kg group expressed
      approximately 50% lower levels of 5-hydroxytryptamine (serotonin) 1A (5HT1A), and mu 1-opioid
      (MORI) mRNA, and a trend was observed in the low-dose group as well. No significant changes in
      alpha-ID adrenergic (ADRAID) or gamma-aminobutyric acid A (GABAA) mRNA levels were
      detected.

            Table B-29. Exposure-related effects in Swiss Albino OF1 mice exposed
            as pups to benzo[a]pyrene in breast milk from dams treated by gavage
            daily from PND1 - PND14
Effect measured
Pup body weight (g; mean ± SEM, n = 20)
PNDO
PND4
PND8
PND12
PND20
PND26
PND32
PND40
Dose (mg/kg-d)
0

1.70 ±0.02
3.01 ±0.08
5. 08 ±0.1
6.57 ±0.12
12.51 ±0.24
17.71 ±0.49
24.47 ± 0.55
30.55 ±0.94
2

1.73 ±0.02
3. 08 ±0.06
5. 26 ±0.09
7. 16 ±0.06***
13. 55 ±0.25**
18.60 ±0.36
25.59 ±0.57
30.90 ±0.93
20

1.74 ±0.02
3. 16 ±0.04
5. 30 ±0.08
7.39 ±0.05***
13.79 ±0.14***
18.35 ±0.34
25.38 ±0.54
31.78 ±0.97
19
20
21
22
23
24
25
      ** p < 0.01, *** p < 0.001 significantly different from control mean

      Source: Bouayed et al. (2009).

     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

               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-82         DRAFT—DO  NOT  CITE  OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    mating, but interpretation of this finding was marred by large variability in numbers of pregnant
 2    females and litter sizes for both treated and control mice. In separate experiments, the fertility of
 3    five male mice/group was not affected by exposure to 1,000 ppm in food for up to 30 days prior to
 4    mating with untreated females. Histologic examinations showed that male mice fed 500 ppm
 5    benzo[a]pyrene for 30 days had spermatozoa present in their testes; further details were not
 6    provided. The only treatment-related effect was a lack of weight gain related to feed unpalatability.
 7    While this study suggests that premating exposure of male or female mice to doses up to
 8    122 mg/kg-day for 20 days may not affect fertility, the sample sizes were too small and study
 9    designs were too inconsistent to provide reliable NOAELs and LOAELs for
10    reproductive/developmental toxicity.
11          In an earlier study (Rigdon and Rennels, 1964), rats (strain not specified) were fed diets
12    containing benzo[a]pyrene at 1,000 ppm for approximately 28 days prior to mating and during
13    gestation. In this study, five of eight treated females mated with untreated males became pregnant,
14    but only one delivered live young.  The treated dam that delivered had two live and two stillborn
15    pups; one dead pup was grossly malformed. In the remaining treated females, vaginal bleeding was
16    observed on CDs 23 or 24. In the inverse experimental design, three of six controls mated to
17    benzo[a]pyrene-treated males became pregnant and delivered live young. Visceral and skeletal
18    examinations of the pups were not conducted. These studies are insufficiently reported and of
19    insufficient design (e.g., inadequate numbers of animals for statistical analysis) to provide reliable
20    NOAELs or LOAELs for reproductive effects from repeated oral exposure to benzo[a]pyrene.

21    Inhalation

22    Reproductive toxicity and in utero exposure via inhalation
23          Archibong et al. (2002) evaluated the effect of exposure to inhaled benzo[a]pyrene on fetal
24    survival and luteal maintenance in timed-pregnant F344 rats. Prior to exposure on GD 8,
25    laparotomy was performed to determine the number of implantation sites, and confirmed pregnant
26    rats were divided into three groups, consisting of rats  that had four to six, seven to nine, or more
27    than nine conceptuses in utero. Rats in these groups were then assigned randomly to the treatment
28    groups or control groups to ensure a similar distribution of litter sizes. Animals (10/group) were
29    exposed to benzo[a]pyrene:carbon black aerosols at concentrations of 25, 75, or 100 [ig/m3 via
30    nose-only inhalation, 4 hours/day on CDs 11-20. Control animals were either sham-exposed to
31    carbon black or remained entirely unexposed. Results of particle size analysis of generated
32    aerosols were reported by several  other reports from  this laboratory (Inyang et al., 2003; Ramesh
33    etal., 2001a; Hoodetal., 2000). Aerosols showed a trimodal distribution with averages of 95%
34    cumulative mass with diameters <15.85 |im; 89% <10 |im; 55%  <2.5 |im;  and 38% <1 |im (Inyang
35    etal., 2003). Ramesh etal. (2001a) reported that the (MMAD ± geometric SD) for the 55% mass
36    fraction with diameters <2.5 [im was 1.7 ± 0.085. Progesterone, estradiol-17p, and prolactin
37    concentrations were determined in plasma collected on CDs 15 and 17. Fetal survival was

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-83         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

15
16
      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
      100 ug/m3, respectively, and 96.7% with carbon black or 98.8% per litter in untreated controls (see
      Table 4-24). Consequently, the number of pups per litter was also decreased in a concentration-
      dependent manner. The decrease was ~50% at 75 ug/m3 and ~65% at 100 ug/m3, compared with
      sham-exposed and unexposed control groups. No effects on hormone levels were observed on
      CDs 15 or 17 at the low-dose. Biologically significant decreases in mean pup weights (expressed as
      g per litter) of >5% were observed at doses >75 ug/m3 (14 and 16% decreases at 75 and 100
      ug/m3, respectively, p < 0.05). Exposure to benzo[a]pyrene did not affect crown-rump length (see
      Table B-30).

            Table B-30. Pregnancy outcomes in female F344 rats treated with
            benzo[a]pyrene on CDs 11-21 by inhalation
Parameter3
Implantation sites
Pups per litter
Survival (litter %)
Pup weight (g/litter)
Crown-rump length
(mm/litter)
Administered concentration of benzo[a]pyrene (u,g/m3)
0 (unexposed
control)
8.6 ±0.2
8.5 ±0.2
98.9 ±1.1
10.6 ±0.1
29.4 ±0.6
0
(carbon black)
8.8 ±0.1
8.7 ±0.2
96.7 ±1.7
8.8 ±0.1
29.3 ±0.5
25
8.8 ±0.5
7.4±0.5b
78.3±4.1b
10.5 ±0.2
28.0 ±0.6
75
9.0 ±0.2
4.2±0.1b
38.0 ± 2. lb
9.1±0.2b
27.3 ±0.7
100
8.8 ±0.1
3.0±0.2b
33.8±1.3b
8.9±0.1b
27.9 ±0.7
17
18
19
20
21
22
23
24
25
      aValues presented as means ± SEM.
      br-
      'Significantly 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 ug/m3 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 ug/m3 (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 GD 15 to 17, but
      this increase did not occur in the rats exposed to 75 ug/m3.  Although the progesterone
      concentration at 75 ug/m3 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
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-84        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    loss of fetuses was thought to be caused by the lower prolactin levels rather than progesterone
 2    deficiency. The reduced circulating levels of progesterone and estradiol-17(3 among
 3    benzo[a]pyrene-treated rats were thought to be a result of limited decidual luteotropic support for
 4    the corpora lutea. The authors proposed the following mechanism for the effects of benzo[a]pyrene
 5    on fertility: benzo[a]pyrene or its metabolites decreased prolactin and decidual luteotropin levels,
 6    compromising the luteotropic support for the corpora lutea and thereby decreasing the plasma
 7    levels of progesterone and estradiol-17p. The low estradiol-17p may decrease uterine levels of
 8    progesterone receptors, thereby resulting in fetal mortality. Based on biologically and statistically
 9    significant decreases in pups/litter and percent fetal survival/per litter, the LOAEL was 25 ug/m3;
10    no NOAEL was identified.

11    Neurotoxicity and in utero exposure via inhalation
12          To evaluate the effects of benzo[a]pyrene on the developing CNS, Wormley et al. (2004)
13    exposed timed-pregnant F344 rats (10/group) to benzo[a]pyrene:carbon black aerosols by nose-
14    only inhalation on CDs 11-21 for 4 hours/day at a concentration of 100 ug/m3. Results of particle
15    size analysis of genenerated aerosols were reported by other reports from this laboratory (Ramesh
16    etal., 2001a; Hoodetal., 2000). Particle size analysis of a 100-ug/m3 aerosol showed a trimodal
17    distribution with averages of 95% cumulative mass with diameters <15.85 um; 90% <10 um;
18    67.5% <2.5 um; and 66.2% <1 um; the MMAD ± geometric SD for the latter fraction was 0.4 ± 0.02
19    um (Hoodetal., 2000). Dams were maintained to term and pups were weaned on PND 30.
20    Benzo[a]pyrene reduced the number of live pups to one-third of control values without affecting
21    the number of implantation sites. During PNDs 60-70, electrical stimulation and evoked field
22    potential responses were recorded via electrodes implanted into the brains of the animals.  Direct
23    stimulation of perforant paths in the entorhinal region revealed a diminution in long-term
24    potentiation of population spikes across the perforant path-granular cell synapses in the dentate
25    gyrus of the hippocampus of Fl generation benzo[a]pyrene-exposed animals; responses in exposed
26    offspring were about 25% weaker than in control offspring. Additionally, NMDA receptor subunit 1
27    protein (important for synaptic functioning) was down-regulated in the hippocampus of
28    benzo[a]pyrene exposed Fl pups. The authors interpreted their results as suggesting that
29    gestational exposure to benzo[a]pyrene inhalation attenuates the capacity for long-term
30    potentiation (a cellular correlate of learning and memory) in the Fl generation.
31          In another study by this same group of investigators, Wu et al. (2003) evaluated the
32    generation of benzo[a]pyrene  metabolites in Fl generation pups, as well as the developmental
33    profile for AhR and mRNA. In this study, confirmed pregnant F344 rats were exposed to
34    benzo[a]pyrene:carbon black aerosols at 25, 75, or 100 ug/m3 via nose-only inhalation,
35    4 hours/day, for 10 days (CDs 11-21). Control animals were exposed to carbon black (sham) to
36    control for inert carrier effects or they remained untreated. Each benzo[a]pyrene concentration
37    had its own set of controls (carbon black and untreated). Two randomly selected pups were
38    sacrificed on each of PND 0, 3, 5,10,15, 20, and 30.  Body, brain, and liver weights were recorded.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-85         DRAFT—DO  NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    Benzo[a]pyrene metabolites were analyzed in the cerebral cortex, hippocampus, liver, and plasma.
 2    A dose-related increase in plasma and cortex benzo[a]pyrene metabolite concentrations in pups
 3    was observed. Dihydrodiols (4,5-; 7,8-; 9,10-) dominated the metabolite distribution profile up to
 4    PND 15 and the hydroxy (3-OH-benzo[a]pyrene; 9-OH-benzo[a]pyrene) metabolites after PND 15
 5    at 100 [ig/m3 (the only exposure concentration reported). Results indicated a dose-related
 6    decrease in the ratio of the total number of pups born per litter to the total number of implantation
 7    sites per litter.  The number of resorptions at 75 andlOO [ig/m3, butnotat25 [J.g/m3,was
 8    statistically significantly increased compared with controls.

 9    Adult male reproductive effects and repeated inhalation exposure
10          Inyangetal. (2003) evaluated the effect of subacute exposure to inhaled benzo[a]pyrene on
11    testicular steroidogenesis and epididymal function in rats. Male F344 rats (10/group), 13 weeks of
12    age, were exposed to benzo[a]pyrene:carbon black aerosols at 25, 75, or 100 [ig/m3 via nose-only
13    inhalation, 4 hours/day for 10 days. Control animals were either exposed to carbon black (sham) to
14    control for exposure to the inert carrier, or they remained untreated. Each benzo[a]pyrene
15    concentration had its own set of controls (carbon black and untreated). Aerosols showed a
16    trimodal distribution with averages of 95% cumulative mass <15.85  |im; 89% <10 |im; 55% <2.5
17    |im; and38% <1 |im (Inyangetal., 2003); an earlier report from this  laboratory indicated that the
18    55% mass fraction had a MMAD ± geometric SD of 1.7 ± 0.085 (Rameshetal., 2001a). Blood
19    samples were collected at 0, 24, 48, and 72 hours after cessation of exposure to assess the effect of
20    benzo[a]pyrene on systemic concentrations of testosterone and luteinizing hormone (LH),
21    hormones that regulate testosterone synthesis. Reproductive endpoints such as testis weight and
22    motility and density of stored (epididymal) sperm were evaluated.
23          Regardless of the exposure concentration, inhaled benzo[a]pyrene did not affect testis
24    weight or the density of stored  sperm compared with controls.  However, inhaled benzo[a]pyrene
25    caused a concentration-dependent reduction in the progressive motility of stored sperm.
26    Progressive motility was similar at 75 and 100 [ig/m3, but these values were significantly lower (p <
27    0.05) than in any other group. The reduction in sperm motility postcessation of exposure was
28    thought to be the result of benzo[a]pyrene limiting epididymal function. Benzo[a]pyrene exposure
29    to 75 [J.g/m3 caused a decrease in circulating concentrations of testosterone compared with controls
30    from the time of cessation of exposure (time 0) to 48 hours posttermination of exposure (p < 0.05).
31    However, the decrease was followed by a compensatory increase in testosterone concentration at
32    72 hours postcessation of exposure. Exposure to 75 [ig/m3 caused a nonsignificant increase in
33    plasma LH concentrations at the end of exposure compared with controls, which increased further
34    and turned significant (p < 0.05) for the remaining time of the study period. The decreased plasma
3 5    concentration of testosterone, accompanied by an increased plasma LH level, was thought to
36    indicate thatbenzo[a]pyrene did not have a direct effect on LH. The authors also noted thatthe
37    decreased circulating testosterone may have been secondary to induction of liver CYP450 enzymes
38    by benzo[a]pyrene. The authors concluded that subacute exposure to benzo[a]pyrene contributed
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-86       DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    to impaired testicular endocrine function that ultimately impaired epididymal function. Based on
 2    this study, the NOAEL was 25 [J.g/m3 and the LOAEL was 75 [ig/m3, based on a statistically
 3    significant reduction in the progressive motility of stored sperm and impairment of testicular
 4    function with 10 days of exposure at 75  [ig/m3.
 5           In a follow-up study with longer exposure duration, adult male F344 rats (10 per group)
 6    were exposed to benzo[a]pyrene:carbon black aerosols at 75 [ig/m3 via nose-only inhalation,
 7    4 hours/day for 60 days (Archibong et al., 2008; Ramesh et al., 2008). Rats in the control group
 8    were subjected to the nose-only restraint, but were not exposed to the carbon black carrier. Blood
 9    samples were collected at 0, 24, 48, and 72 hours after exposure terminated, and the animals
10    sacrificed for tissue analyses following the last blood sampling. Data were analyzed statistically for
11    benzo[a]pyrene effects on weekly body weights, total plasma testosterone and LH concentrations,
12    testis weights, density of stored spermatozoa, sperm morphological forms and motility,
13    benzo[a]pyrene metabolite concentrations and AHH activity, and morphometric assessments of
14    testicular histologies. Relative to controls, the results indicated 34% reduced testis weight (p <
15    0.025), reduced daily sperm production (p < 0.025) and reduced intratesticular testosterone
16    concentrations (p < 0.025). Plasma testosterone concentrations were reduced to about one-third of
17    the level in controls on the last day of exposure (day 60) and at 24, 48, and 72 hours later (p < 0.05).
18    However, plasma LH concentrations in benzo[a]pyrene  exposed rats were elevated throughout the
19    blood sampling time  periods compared with controls  (p < 0.05). In testis, lung, and liver, AHH
20    activity, and benzo[a]pyrene-7,8-dihydrodiol (precursor to the DNA-reactive BPDE) and
21    benzo[a]pyrene-3,6-dione metabolites were significantly (p < 0.05) elevated relative to controls.
22    Progressive motility  and mean density of stored spermatozoa were significantly reduced (p < 0.05).
23    Weekly body weight  gains were not affected by benzo[a]pyrene exposure. These results indicate
24    that 60-day exposure of adult male rats to benzo[a]pyrene:carbon black aerosols at 75 [J.g/m3
25    produced decreased  testis weight; decreased intratesticular and plasma testosterone
26    concentrations; and decreased sperm production, motility, and density.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-87         DRAFT—DO NOT CITE OR QUOTE

-------
                                              Toxicological Review ofbenzo[a]pyrene
1   OTHER PERTINENT TOXICITY INFORMATION

2         Table B-31. In vitro genotoxicity studies of benzo[a]pyrene in non-
3         mammalian cells

Result
S9
S9
Reference
Endpoint/test system: prokaryotic cells
Forward mutation
Salmonella typhimurium TM677
S. typhimurium TM677
+
+
-
ND
Rastetter et al., 1982
Babson et al., 1986
Reverse mutation
S. typhimurium TA98, TA1538
S. typhimurium TA98, TA100, TA1538
S. typhimurium TA1538, TA98
S. typhimurium TA98, TA100, TA1537
S. typhimurium TA98, TA100
S. typhimurium TA98
S. typhimurium TA98, TA100
S. typhimurium TA98, TA100
S. typhimurium TA98
S. typhimurium TA98, TA100
S. typhimurium TA98, TA100, TA1538
S. typhimurium TA97, TA98, TA100
S. typhimurium TA97, TA98, TA100, TA1537
S. typhimurium TA97, TA98, TA100
S. typhimurium TA98
S. typhimurium TA98, TA100
S. typhimurium TA98
S. typhimurium TA98, TA100
S. typhimurium TA98
S. typhimurium TA98
S. typhimurium TA98
S. typhimurium TA98
S. typhimurium TA100
S. typhimurium TA100
S. typhimurium TA100
+
+
+
+
+
+
+
+
+
+
ND
+
+
+
+
+
+
+
+
+
-
+
+
+
+
ND
ND
-
-
	
-
ND
-
ND
ND
-
-
-
ND
-
-
ND
ND
ND
ND
ND
ND
ND
-
ND
Ames et al., 1975
McCann et al., 1975
Wood et al., 1976
Epleretal., 1977
Obermeier and Frohberg,
1977
Pitts et al., 1978
LaVoie et al., 1979
Simmon, 1979a
Hermann, 1981
Alfheim and Randahl, 1984
Glatt et al., 1985
Sakaietal., 1985
Glatt et al., 1987
Marino, 1987
Alzieu et al., 1987
Prasanna et al., 1987
Ampyetal., 1988
Bos et al., 1988
Lee and Lin, 1988
Antignac et al., 1990
Gaoetal., 1991
Balanskyetal., 1994
Norpoth et al., 1984
Carver et al., 1986
Pahlman and Pelkonen, 1987
             This document is a draft for review purposes only and does not constitute Agency policy.
                                         B-88        DRAFT—DO NOT CITE OR QUOTE

-------
                                                 Toxicological Review ofbenzo[a]pyrene

S. typhimurium TA100
S. typhimurium TA100
S. typhimurium TA100
S. typhimurium TA100
S. typhimurium TA1537, TA1538
S. typhimurium TA1537, TA1538
S. typhimurium TA1537
S. typhimurium TA1538
S. typhimurium TA1538
S. typhimurium TA1535
S. typhimurium TA 1535
S. typhimurium TA 1535
S. typhimurium TA1535
Result
S9
+
+
+
-
+
+
+
+
+
-
-
-
-
S9
ND
ND
ND
ND
-
-
ND
ND
-
-
-
ND
-
Reference
Tang and Friedman, 1977
Bruce and Meddle, 1979
Phillipson and loannides, 1989
Balanskyetal., 1994
Ames et al., 1973
Glatt et al., 1975
Oesch et al., 1976
Egert and Greim, 1976
Rosenkranz and Poirier, 1979
Ames et al., 1973
Glatt et al., 1975
McCann et al., 1975
Epleretal., 1977
DIMA damage
E. co///pol A
E. co///differential killing test
E. co// WP2-WP100/rec-assay
E. CO///SOS chromotest Pq37
+
+
+
+
-
-
ND
	
Rosenkranz and Poirier, 1979
Tweats, 1981
Mamberetal., 1983
Mersch-Sundermann et al.,
1992
Endpoint/test system: nonmammalian eukaryotes
Mitotic recombination
S. cerevisiae D4-RDII
S. cerevisiae D3
ND
-
-
-
Siebert et al., 1981
Simmon, 1979b
1
2
+ = positive; - = negative; ND = not determined

     Table B-32. In vitro genotoxicity studies of benzo[a]pyrene in
     mammalian cells
dssay/test system
Result
+S9
-S9
Reference
Forward mutation
Human AHH-1 lymphoblastoid cells
Human lymphoblast (AHH-1) cells (hprt)
Human lymphoblastoid (AHH-1) cell line
ND
ND
ND
+
+
+
Danheiseretal., 1989
Crespi et al., 1985
Chen et al., 1996
             This document is a draft for review purposes only and does not constitute Agency policy.
                                            B-89        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Assay /test system
Human fibroblast (MRC5CV1) cell line
(hprt)
Human lymphoblast (TK) cells
Human lymphoblast (TK6) cells
Human embryonic epithelial (EUE) cells
Human HSC172 lung fibroblasts
Human Q3-wp normal lung keratinocytes
Human SCC-13Y lung keratinocytes
Mouse /ocZtransgenic Muta™Mouse
primary hepatocytes
Mouse L5178Y/HGPRT
Mouse lymphoma (L5178Y/TK+/-) cells
Mouse lymphoma (L5178Y/TK+/-) cells
Mouse lymphoma (L5178Y/TK+/-) cells
Mouse lymphoma (L5178Y/TK+/-) cells
Chinese hamster ovary (CHO) cells (aprt)
CHOcells(5 marker loci)
Chinese hamster V79 cells (co-cultured
with irradiated HepG2 cells)
Chinese hamster V79 lung epithelial cells
Chinese hamster V79 lung epithelial cells
Chinese hamster V79 lung epithelial cells
Rat/Fischer, embryo cells/OuaR
Result
+S9
—
ND
+
ND
+
+
ND
ND
+
+
+
+
+
+
+
+
+
+
+
ND
-S9
ND
+
ND
+
-
ND
+
+
-
-
ND
-
ND
ND
+
ND
ND
ND
ND
+
Reference
Haneltetal., 1997
Barfknechtetal., 1982
Crespi et al., 1985
Rocchi et al., 1980
Gupta and Goldstein, 1981
Allen-Hofmann and
Rheinwald, 1984
Allen-Hofmann and
Rheinwald, 1984
Chen etal., 2010
Clive et al., 1979
Clive et al., 1979
Amacher and Turner, 1980;
Amacheretal., 1980
Amacher and Paillet, 1983
Arce et al., 1987
Yang et al., 1999
Gupta and Singh, 1982
Diamond etal., 1980
Huberman, 1976
Arce et al., 1987
O'Donovan, 1990
Mishraetal., 1978
DNA damage
DNA adducts
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Human fibroblast (MRC5CV1) cell line
Human hepatoma (HepG2) cell line
Hamster tracheal cells
Chinese hamster V79 lung epithelial cells
ND
ND
ND
ND
+
ND
ND
+
+
+
+
+
ND
+
+
ND
Wiencke et al., 1990
Li et al., 2001
Wu et al., 2005
Gu et al., 2008
Haneltetal., 1997
Tarantini et al., 2009
Roggeband et al., 1994
Arce et al., 1987
This document is a draft for review purposes only and does not constitute Agency policy.
                               B-90        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Assay /test system
Virus transformed SHE and mouse
C3H10T1/2 cells
Mouse lymphoma (L5178Y/TK+/-) cells
Rat tracheal cells
Result
+S9
ND
+
ND
-S9
+
ND
+
Reference
Arce et al., 1987
Arce et al., 1987
Roggeband et al., 1994
Unscheduled DNA synthesis
HeLa cells
Human fibroblasts
Human fibroblasts
Human HepG2
Hamster primary embryo cells
Hamster tracheal cells
Rat hepatocytes
Rat tracheal cells
+
+
+
ND
ND
ND
ND
ND
ND
ND
—
+
+
+
+
-
Martin et al., 1978
Agrelo and Amos, 1981
Robinson and Mitchell,
1981
Valentin-Severin et al., 2004
Casto et al., 1976
Roggeband et al., 1994
Michalopoulos et al., 1978
Roggeband et al., 1994
DNA repair
Human mammary epithelial cells
Human skin fibroblasts
Baby hamster kidney (BHK21/cl3) cells
secondary mouse embryo fibroblasts
(C57BL/6) and human lymphocytes
Rat/F344 hepatocytes
ND
ND
ND
ND
ND
+
+
+
+
+
Leadonetal., 1988
Miloetal., 1978
Feldmanetal., 1978
Shinohara and Cerutti, 1977
Williams etal., 1982
Cytogenetic damage
CAs
Human blood cells
Human WI38 fibroblasts
Chinese hamster lung cells
Chinese hamster V79-4 lung epithelial
cells
Mouse lymphoma (L5178Y/TK+/-) cells
Rat Liver RL1 cells
ND
+
+
—
+
+
+
-
-
—
ND
ND
Salamaetal., 2001
Weinstein et al., 1977
Matsuoka et al., 1979
Popescu et al., 1977
Arce et al., 1987
Dean, 1981
MN
Human AHH-1 lymphoblastoid cells
Human HepG2 liver cells
Human lymphoblastoid (TK) cells
Human MCL-5 lymphoblastoid cells
Human peripheral blood lymphocytes
Chinese hamster V79 cells
ND
ND
ND
ND
+
ND
+
+
+
+
ND
+
Crofton-Sleigh et al., 1993
Wu et al., 2003
Fowler etal., 2010
Crofton-Sleigh et al., 1993
LoJaconoetal., 1992
Whitwell etal., 2010
This document is a draft for review purposes only and does not constitute Agency policy.
                               B-91        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Assay /test system
Chinese hamster V79-MZ cells
Result
+S9
ND
-S9
+
Reference
Matsuoka et al., 1999
DNA strand breaks
Human sperm
Human peripheral blood lymphocytes
Human fibroblast (MRC5CV1) cell line
Human hepatoma (HepG2) cell line
Human prostrate carcinoma (DU145) cell
line
Mouse embryo fibroblast (C3H/10T1/2 CL
8) cells
Rat CIS trachea epithelial cells
Rat lymphocytes
+
+
+
ND
ND
ND
ND
ND
+
+
ND
+
+
+
+
+
Sipinenetal., 2010
Rodriguez-Romero et al.,
2012
Haneltetal., 1997
Tarantini et al., 2009
Nwagbara, 2007
Lubet et al., 1983
Cosma and Marchok, 1988;
Cosma et al., 1988
Gaoetal., 1991
SCEs
Human C-HC-4 and C-HC-20 hepatoma
cells
Human diploid fibroblast (TIG-II) cell line
Human fibroblasts
Human blood cells
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Human peripheral blood lymphocytes
Chinese hamster Don-6 cells
Chinese hamster V79 lung epithelial cells
Chinese hamster V79 lung epithelial cells
Chinese hamster V79 lung epithelial cells
Chinese hamster V79 lung epithelial cells
Chinese hamster V79 lung epithelial cells
CHO cells
CHO cells
CHO cells
ND
+
ND
ND
ND
ND
ND
ND
+
+
ND
+
+
+
+
ND
+
+
ND
+
+
+
+
+
+
+
+
-
ND
+
-
ND
ND
ND
+
-
—
+
Abe et al., 1983a, b
Huh etal., 1982
Juhletal., 1978
Salamaetal., 2001
Rudigeretal., 1976
Craig-Holmes and Shaw,
1977
Schonwald et al., 1977
Wiencke et al., 1990
Tohda et al., 1980
LoJaconoetal., 1992
Abe et al., 1983a, b
Popescu et al., 1977
Mane etal., 1990
Wojciechowski et al., 1981
Arce et al., 1987
Kulka et al., 1993a
de Raat, 1979
Husgafvel-Pursiainen et al.,
1986
Wolff and Takehisa, 1977
This document is a draft for review purposes only and does not constitute Agency policy.
                               B-92        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Assay /test system
CHO cells
Chinese hamster lung cells
Rabbit peripheral blood lymphocytes
Rat ascites hepatoma AH66-B
Rat esophageal tumor Rl
Rat hepatocyte (immortalized) cell lines
(NRLcl-B, NRLcl-C, and ARL)
Rat hepatoma (Reuber H4-II-E) cells
Rat liver cell lineARLlS
Rat pleural mesothelial cells
Result
+S9
ND
ND
ND
ND
ND
+
ND
ND
ND
-S9
+
+
+
+
+
ND
+
+
+
Reference
Pal et al., 1978
Shimizu etalv 1984
Takehisa and Wolff, 1978
Abe et al., 1983a, b
Abe et al., 1983a, b
Kulkaetal., 1993b
Dean et al., 1983
long et al., 1981
Achard et al., 1987
Aneuploidy
Chinese hamster V79-MZ cells
ND
+
Matsuoka et al., 1998
Cell transformation
Human BEAS-2B lung cells
Human breast epithelial (MCF-10F, MCF-7,
T24) cell lines
Baby hamster kidney (BHK21/cl3) cells
Golden hamster embryo cells
Syrian hamster embryo (SHE) cells
SHE cells
SHE cells
SHE cells/focus assay
Fetal Syrian hamster lung (FSHL) cells
Virus infected rat embryo RLV/RE and RAT
cells; mouse embryo AKR/Me cells; Syrian
hamster embryo cells
Virus transformed SHE and mouse
C3H10T1/2 cells
Mouse C3H/10T1/2 embryo fibroblasts
Mouse embryo fibroblast (C3H/10T1/2 CL
8) cells
Mouse embryo fibroblast (C3H/10T1/2 CL
8) cells
Mouse SHE cells; BALB/c-3t3 cells;
C3H/10T1/2 cells; prostate cells
Mouse BALB/c-313 cells
Mouse BALB/c-313 cells
Mouse BALB/c-313 clone A31-1-1
ND
ND
+
+
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
+
+
ND
ND
+
+
+
+
+
+
+
+
+
+
+
+
+
+
van Agen et al., 1997
Calaf and Russo, 1993
Greb et al., 1980
Mager et al., 1977
DiPaoloetal., 1971,1969
Dunkeletal., 1981
LeBoeuf etal., 1990
Casto et al., 1977
Emura et al., 1987, 1980
Heidelberger et al., 1983
Arce et al., 1987
Nesnow et al., 2002, 1997
Peterson et al., 1981
Lubet et al., 1983
Heidelberger et al., 1983
Dunkeletal., 1981
Matthews, 1993
Little and Vetroys, 1988
This document is a draft for review purposes only and does not constitute Agency policy.
                               B-93        DRAFT—DO NOT CITE OR QUOTE

-------
                                              Toxicological Review ofbenzo[a]pyrene
Assay /test system
Rat embryo cells/SA7 virus transformation
Rat/Fischer, embryo cells (leukemia virus
transformed)
Rat/Fischer, embryo cells/OuaR
Result
+S9
ND
ND
ND
-S9
+
+
+
Reference
DiPaolo and Casto, 1976
Dunkeletal., 1981
Mishra et al., 1978
+ = positive; - = negative; ND = not determined; SHE = Syrian hamster embryo; TK = thymidine kinase
         This document is a draft for review purposes only and does not constitute Agency policy.
                                        B-94         DRAFT—DO NOT CITE OR QUOTE

-------
                                                                                Toxicological Review ofbenzo[a]pyrene
1

2
Table B-33. In vivo genotoxicity studies of benzo[a]pyrene
Endpoint
Mutation










Mutation,
germline










Test system
Human, blood
T lymphocytes
(smokers and
nonsmokers); hprt
locus mutation assay






Mouse, T-stock, (SEC x
C57BL)F1, (C3H x
101)F1, or(C3Hx
C57BL)F1 for females;
(101xC3H)Flor(C3H
x 101)F1 for males;
dominant-lethal
mutation assay




Test conditions
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.






12-wk-old males dosed with
benzo[a]pyrene i.p. and mated 3.5-6.5
d posttreatment with 12-wk-old
females from different stocks;
sacrificed on d 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
experiment; fertilized eggs collected
from 9 to 11 hrs after mating and first-
cleavage metaphase chromosomes
prepared 20 hrs after mating.
Results
+










+











Dose
Smokers and
nonsmokers









500 mg/kg











Comment
Splicing mutations, base-pair
substitutions, frameshift, and
deletion mutations observed.
Smokers and nonsmokers had
GC->TAtransversions (13 and
6%, respectively) and GC->AT
transitions (24 and 35%,
respectively) in hprt gene
consistent with in vitro
mutagenicity of
benzo[a]pyrene
The percent of dominant lethal
mutations were in the order of
T-stock = (C3Hxl01)Fl>
(SECxC57BL)Fl >
(C3HxC57BL)Fl







Reference
Hackman
et al., 2000









Generoso
et al., 1979










                             This document is a draft for review purposes only and does not constitute Agency policy.
                                           B-l         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
Mutation,
germline









Mutation,
germline







Mutations
and BPDE-
DNA
adducts,
germline



Test system
Mouse, male stocks:
(101 xC3H)Fl; female
stocks (A):
(101xC3H)Fl, (B):
(C3Hxl01)Fl, (C):
(C3HxC57BL)Fl,
(D):(SECxC57BL)Fl,
(E):T-stock females;
dominant lethal
mutations

Mouse, male stocks:
(101 xC3H)Fl; female
stocks (A): (101 x
C3H)F1, (B):(C3Hx
101)F1, (C): (C3H x
C57BL)F1, (D):(SECx
C57BL)F1, (E): T-stock
females; heritable
translocations
Mouse, C57BL/6, ell
transgenic (Big Blue®)






Test conditions
In dominant lethal assay, 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 d postdosing, or with (#3b)
with stocks B, C, D, and E females 3.5-
6.5 d postdosing. Control group mated
at time corresponding to 1.5-4.5 d
posttreatment in the test groups.
For heritable translocation assay,
males were mated with stocks B and D
females 3.5-7.7 d post-benzo[a]pyrene
treatment and male progeny screened
for translocation heterozygosity.




Benzo[a]pyrene administered i.p. in
corn oil on d 0, 1, and 2; sacrificed at d
4, 16, 30, 44, or 119. Caput and cauda
epididymal spermatozoa analyzed for
ell mutation frequency, and DNA
adducts analyzed in testis by LC-
MS/MS SRM with 15N-deoxyguanosine
labeling.
Results
+










-








+







Dose
500 mg/kg










500 mg/kg








50 mg/kg







Comment
Dominant lethal effects were
observed in early to middle
(4.5-5.5 and 6.5-7.5 d
posttreatment, respectively)
spermatozoa and in
preleptotene spermatocytes
(32.5-33.5 and 34.5-35.5 d
post-treatment).



No significant differences were
observed between treated and
control progeny.






Exposed spermatocytes
acquired persistent BPDE-DNA
adducts; exposed
spermatogonia gave rise to
spermatocytes with mutations
consistent with a
benzo[a]pyrene spectrum
(GOTA transversions).
Reference
Generoso
et al., 1982









Generoso
et al., 1982







Olsen et
al., 2010






This document is a draft for review purposes only and does not constitute Agency policy.
               B-2         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
Mutations
and BPDE-
DNA
adducts,
germline



Mutations
and BPDE-
DNA
adducts







Mutation












Test system
Mouse, C57BL/6 males,
wild type and Xpc-/~
with pUR288/ocZ
reporter gene




Mouse, C57BL/6 lacZ
transgenic









Mouse, C57BL female x
T-strain male; somatic
mutation assay










Test conditions
Benzo[a]pyrene given via gavage in
sunflower oil 3 times/wk for 1, 4, or 6
wks (Xpc7") or 6 wks (Wt). Spleen,
testis, and sperm cells analyzed for
lacZ mutation frequency, and DNA
adducts analyzed in testis by 32P-
postlabeling.

Mice dosed with single i.p. injection of
benzo[a]pyrene in DMSO; sacrificed 1,
3, 5, 7, 14, 21, and 28 d posttreatment;
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.



Mice mated for a 5-d period; 10.25 d
post-appearance of vaginal plug,
females injected i.p. with
benzo[a]pyrene or vehicle; offspring
(pups) scored for survival, morphology,
and presence of white near-midline
ventral spots and recessive spots.






Results
+







+










+












Dose
13 mg/kg







50 mg/kg










100 or
500 mg/kg











Comment
Statistically significant
increases in lacZ mutation
frequencies in Xpc-/- spleen at
4 and 6 wks (dose dependent)
and in Wt spleen and sperm at
6 wks; DNA adducts were
statistically significant in testis
in all exposed groups.
BPDE-dG adduct levels peaked
between 5 and 7 days
posttreatment, 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.
Induced coat color mosaics
represent genetic changes
(e.g., point mutations) in
somatic cells. White near-
midline ventral spots and
recessive spots represent
melanocyte cell killing and
mutagenicity, respectively.
Benzo[a]pyrene caused high
incidence of recessive spots
but did not correlate with
white near-midline ventral
spots.
Reference
Verhofstad
etal., 2011






Boerrigter,
1999









Russell,
1977











This document is a draft for review purposes only and does not constitute Agency policy.
               B-3         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
Mutation
Mutation
Mutation
Mutation
Mutation
Test system
Mouse, /ocZtransgenic
(Muta™Mouse)
Mouse, /ocZtransgenic
(Muta™Mouse)
Mouse, C57BL/6J Dlb-1
congenic; Dlb-1 locus
assay
Mouse, C57BL/6 (lacZ
negative and XPA+/+
and XPA~'~); hprt
mutations in T
lymphocytes
Mouse, Cockayne
syndrome-deficient
(Csfo^V; heterozygous
(Csb+/~) and WT
controls (Csb+/+); hprt
mutation frequency
assay
Test conditions
Benzo[a]pyrene given via gavage in
olive oil daily for 28 consecutive d;
sacrificed 3 d after last dosing; 4
organs analyzed for lacZ mutation
frequency.
Benzo[a]pyrene given orally in corn oil
for 5 consecutive d; sacrificed 14 d
after last dosing; 11 organs analyzed
for lacZ mutation frequency.
Animals dosed: (1) i.p. with vehicle or
benzo[a]pyrene two, four, or six doses
at 96-hr intervals; or (2) single dose of
benzo[a]pyrene given i.p. or p.o. alone
or 96 hrs following a single i.p. dosing
with 10 u.g/kgTCDD.
Gavage in corn oil 3 times/wk for 0, 1,
5, 9, or 13 wks; sacrificed 7 wks after
last treatment.
Csb-/yiacZ+/- and Csb+/yiacZ+/- mice
were dosed i.p. with benzo[a]pyrene 3
times/wk for 5, 9, or 13 wks; for hprt
mutation frequency analysis mice were
sacrificed 3 wks after last treatment;
splenocytes collected; for lacZ
mutation frequency analysis, mice
were sacrificed 3 d after last treatment
and liver, lung, and spleen collected.
Results
+
+
+
+
+
Dose
25, 50, and 75
mg/kg-day
125 mg/kg-day
40 mg/kg
13 mg/kg
13 mg/kg
Comment
Highest lacZ mutation
frequency observed in small
intestine, followed by bone
marrow, glandular stomach,
and liver
Highest mutation frequency
observed in colon followed by
ileum > forestomach > bone
marrow = spleen > glandular
stomach > liver = lung >
kidney = heart
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.
Mutation sensitivity:
XPA~'~ > XPA+/+.
lacZ mutation frequency
detected in all tissues but no
differences between WT and
Csb~f~ mice; hprt mutations
significantly higher in Csb~f~
mice than control mice. BPDE-
dGuo adducts in hprt gene are
preferentially removed in WT
mice than Csb~f~ mice.
Reference
Lemieux et
al., 2011
Hakura et
al., 1998
Brooks et
al., 1999
Bol et al.,
1998
Wijnhoven
et al., 2000
This document is a draft for review purposes only and does not constitute Agency policy.
               B-4         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
Mutation



Mutation






Mutation



Mutation


Mutation


Mutation








Test system
Mouse, B6C3F1;
forestomach H-ros,
K-ros, and p53
mutations
Mouse, lacZ/galE
(Muta™ Mouse); skin
painting study




Mouse, T-strain



Mouse, 129/Ola (WT);
hprt mutations in
splenic T lymphocytes
Mouse, A/J, male


Mouse, CD-I; skin
papillomas (Ha-ros
mutations)






Test conditions
Benzo[a]pyrene given in feed in a 2-yr
chronic feeding study.


Mice topically treated with a single
dose or in five divided doses daily;
sacrificed 7 or 21 d after the single or
final treatment; DNAfrom skin, liver,
and lung analyzed for mutations.


Benzo[a]pyrene given to pregnant
mice by gavage in 0.5 ml corn oil on
GDs 5-10.

Single i.p. injection followed by
sacrifice 7 wks posttreatment.

Single i.p. injection followed by
sacrifice 28 days posttreatment.

Female mice were initiated topically
with a single dose of benzo[a]pyrene
and 1 wk after initiation promoted
twice weekly with 5 nmol TPA for 14
wks. One month after stopping TPA
application, papillomas were collected
and DNAfrom 10 individual papillomas
were analyzed for Ha-ros mutations by
PCR and direct sequencing.
Results
+



+skor
Li,Lu





+



+


+


+








Dose
5, 25, or
100 ppm


1.25 or
2.5 mg/kg
(25 or
50 u.g/mouse)



10 mg/mouse
(5x2 mg)


0, 50, 100,
200, or 400
mg/kg
0, 0.05, 0.5, 5,
or 50 mg/kg

600
nmol/mouse







Comment
68% K-ros (codons 12,13), 10%
H-ros (codon 13), 10% p53
mutations; all G->T
transversions
Skin showed significant dose-
and time-dependent increase
in mutation frequency; liver
and lung showed no
mutations; mutation frequency
for single- or multiple-dose
regimens was similar.




Dose-dependent increase in
hprt mutation frequency.

Dose-dependent increase in
lung tissue K-ros codon 12
G->T mutation frequency.
About 90% of papillomas
contained Ha-ros mutations, all
of them being transversions at
codons 12 (20% GGA->GTA),
13 (50% GGC->GTC), and 61
(20% CAA->CTA).



Reference
Gulp et al.,
2000


Dean et al.,
1998





Davidson
and
Dawson,
1976
Bol et al.,
1998

Meng et
al., 2010

Colapietro
et al., 1993







This document is a draft for review purposes only and does not constitute Agency policy.
               B-5         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
Mutation









BPDE-
DNA
adducts




BPDE-
DNA
adducts




Test system
Rat, Wistar









Human, WBCs






Human, WBCs






Test conditions
Single dose by gavage; urine and feces
collected 0-24, 24-48, and 48-72 hrs
posttreatment; urine and extracts of
feces tested in S. typhimurium TA100
strain with or without S9 mix and (3-
glucuronidase.




96 people occupationally or medically
exposed to PAH mixtures (psoriatic
patients, coke oven workers, chimney
sweeps, and aluminum anode plant
workers); adducts measured by
HPLC/fluorescence analysis.

67 highly exposed coke oven workers
were tested for genetic factors that
can modulate individual responses to
carcinogenic PAHs; adducts measured
by HPLC/fluorescence analysis.


Results
+









+






+






Dose
0, 1, 5, 10, or
100 mg/kg






















Comment
Fecal extracts and urine
showed mutagenicity >1 and
10 mg/kg body weight
benzo[a]pyrene, respectively.
Highest mutagenic activity
observed for 0-24 hrs
posttreatment for feces and
24-48 hrs posttreatment for
urine with (3-glucuronidase ±
S9 mix.
Percentages of subjects with
adduct levels > the 95th
percentile control value were
47% (7/15), 21% (4/19) and 3%
(1/34) in coke oven workers,
chimney sweeps, and controls,
respectively.
Levels of BPDE-DNA adducts
were significantly associated
with workplace PAH exposure
(as correlated with urinary
excretion of 1-pyrenol), lack of
GSTM1 activity, and low NER
capacity.
Reference
Willems et
al., 1991








Pavanello
et al., 1999





Pavanello
et al., 2005





This document is a draft for review purposes only and does not constitute Agency policy.
               B-6         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
BPDE-
DNA
adducts













BPDE-
DNA
adducts





Test system
Human, peripheral
lymphocytes














Human, maternal and
umbilical cord blood






Test conditions
585 Caucasian municipal workers (52%
males, 20-62 years old) from
northeast Italy environmentally
exposed to PAH mixtures were
screened for adducts measured by
HPLC/fluorescence analysis.










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.
Results
+















+







Dose
























Comment
Forty-two percent of the
participants had elevated anti-
BPDE-DNAadduct levels,
defined as >0.5 adducts/108
nucleotides (mean, 1.28 ± 2.80
adducts/108 nucleotides).
Comparison of adduct levels
with questionnaire responses
indicated that smoking,
frequent consumption of PAH-
rich meals (>52 versus <52
times/year), and longtime
periods spent outdoors (>4
versus <4 hours/day) were risk
factors as all increased BPDE-
DNA adduct levels significantly.
BPDE-DNA adduct levels in
cord and maternal blood were
highest in study participants
who lived within 1 mile of the
WTC, with inverse correlation
between cord blood levels and
distance from WTC.

Reference
Pavanello
et al., 2006














Perera et
al., 2005a






This document is a draft for review purposes only and does not constitute Agency policy.
               B-7         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
BPDE-
DNA
adducts






BPDE-
DNA
adducts









BPDE-
DNA
adducts



BPDE-
DNA
adducts
Test system
Human, WBCs








Human, WBCs











Mouse, /ocZtransgenic
(Muta™Mouse)




Mouse, (Ahr+/+, Ahr+/~,
Ahr'~)

Test conditions
Workers were exposed for 6-8 hrs/d
for at least 4-6 mo before blood
collection; leukocyte DNA isolated and
digested, and benzo[a]pyrene tetrols
analyzed by HPLC with fluorescent
detection. Low, medium, and high
exposure groups correspond to <0.15,
0.15-4, and >4 mg/m3 of
benzo[a]pyrene, respectively.
Coke oven workers were exposed to
PAHs and benzo[a]pyrene-WBC DNA
analyzed by HPLC-fluorescence
detection for BPDE-DNA adducts.








Benzo[a]pyrene given via gavage in
olive oil daily for 28 consecutive d;
sacrificed 3 d after last dosing; 4
organs analyzed for DNA adducts using
32P-postlabeling with nuclease PI
digestion enrichment.
Gavage; sacrificed 24 hrs
posttreatment.

Results
+








+











+





+


Dose
<0. 15, 0.15-4,
or >4 u.g/m3 of
benzo[a]pyren
e





0.14 u.g/m3











25, 50, and 75
mg/kg-day




100 mg/kg


Comment
PAH exposure, CYP1A1 status
and smoking significantly
affected DNA adduct levels,
i.e., CYPlAl(*l/*2 or *2A/*2a)
> CYP1A1*1/*1; occupational >
environmental exposure;
smokers > nonsmokers;
adducts increased with dose
and duration of smoking.
Median detectable BPDE-DNA
adducts in workers vs. controls
not significant due to low
number of subjects (9 workers,
26 controls); 4/9 workers had
adducts substantially higher
than all controls. No significant
difference between smokers
and nonsmokers; no
correlation with air
benzo[a]pyrene levels and
adduct levels.
Highest adduct levels observed
in liver, followed by glandular
stomach, small intestine, and
bone marrow


No induction of CYP in Ahr'7',
but all alleles positive for
adduct formation.
Reference
Rojas et
al., 2000







Mensing et
al., 2005










Lemieux et
al., 2011




Sagredo et
al., 2006

This document is a draft for review purposes only and does not constitute Agency policy.
               B-8         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
BPDE-
DNA
adducts
BPDE-
DNA
adducts



BPDE-
DNA
adducts


BPDE-
DNA
adducts







BPDE-
DNA
adducts


Test system
Mouse, C57BL/6J
Cyplal(+/-) and
Cyplal (-/-)
Mouse, B6C3Fi




Mouse, BALB/c




Mouse, BALB/cAnN
(BALB), CBA/JN (CBA);
[32P]-postlabeling assay







Mouse, BALB/c, skin




Test conditions
Single i.p. injection; sacrificed 24 hrs
posttreatment; liver DNA analyzed by
[32P]-postlabeling assay.
Benzo[a]pyrene fed in diet for 4 wks
(100 ppm) or for 1, 2, 8, 16, and 32 wks
(5 ppm); sacrificed and liver, lungs,
forestomach, and small intestine
collected; DNA analyzed by [32P]-
postlabeling assay.
Single i.p. injection; sacrificed 12 hrs
postinjection; liver and forestomach
collected; DNA binding of [3H]-
benzo[a]pyrene analyzed by
scintillation counting.
Animals dosed i.p. with or without 24
hr pretreatment with TCDD.








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.
Results
+


+




+




+









+




Dose
500 mg/kg


5 ppm (32
wks) and 100
ppm (4 wks)



140 u.Ci/100 g
body weight



50 and
200 mg/kg








4x 1.2 u.mol/
animal



Comment
BPDE-DNAadduct levels
fourfold higher in Cyplal(-f-)
mice than Cyplal(+/-) mice.
Linear dose-response in 4-wk
study; the 5 ppm groups
showed a plateau after 4 wks
of feeding.


Liver DNA had threefold higher
binding of benzo[a]pyrene than
that of forestomach.


Adduct levels similar in both
strains dosed with
benzo[a]pyrene alone. TCDD
pretreatment had a greater
suppressive effect on adduct
formation in BALB relative to
CBA mice at low dose but
resulted in no significant
difference in adduct levels at
high dose.
Five adducts spots detected.




Reference
Uno et al.,
2001

Gulp et al.,
2000




Gangar et
al., 2006



Wu etal.,
2008








Reddy et
al., 1984



This document is a draft for review purposes only and does not constitute Agency policy.
               B-9         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
BPDE-
DNA
adducts
BPDE-
DNA
adducts
BPDE-
DNA
adducts
BPDE-
DNA
adducts
BPDE-
DNA
adducts
Test system
Mouse, Swiss,
epidermal and dermal
skin
Rat, CD, peripheral
blood lymphocytes,
lungs, and liver
Rat, Sprague-Dawley,
liver
Rat, Lewis, lung and
liver
Rat, F344;
[32P]-postlabeling assay
Test conditions
Single topical application on shaved
backs; sacrificed 1, 3, and 7 d
posttreatment; epidermal and dermal
cells separated; DNA isolated, digested
with DNAsel, and estimated DNA
binding; adducts separated by HPLC.
Single i.p. injection; sacrificed 3 d
posttreatment; DNA analyzed by
Nuclease Pl-endhanced [32P]-
postlabeling assay.
Single i.p. injection followed by
sacrifice at 4 hrs posttreatment; liver
DNA isolated and analyzed by [ 32P]-
postlabeling assay.
Animals received a single oral dose of
benzo[a]pyrene in tricaprylin;
sacrificed 1, 2, 4, 11, and 21 d
postdosing; analyzed liver and lung
DNA for BP-DNA adducts by [32P]-
postlabeling assay and urine for
8-oxodG adducts by HPLC-
electrochemical detection.
Benzo[a]pyrene given in the diet for
30, 60, or 90 d; animals sacrificed and
liver and lung isolated and DNA
extracted and analyzed for adducts.
Results





Dose
250 nmol in
150 ul
acetone
2.5 mg/animal
100 mg/kg
10 mg/kg
0, 5, 50, or
100 mg/kg
Comment
Both cells positive for
benzo[a]pyrene adducts;
epidermis > dermis; adducts
persisted up to 7 d with a
gradual decline in levels.
BPDE-dG as major adducts and
several minor adducts
detected in all tissues.
Two adduct spots detected.
BPDE-dG levels peaked 2 d
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.
Adduct levels linear at low and
intermediate doses, nonlinear
at high dose.
Reference
Oueslati et
al., 1992
Ross et al.,
1991
Reddy et
al., 1984
Briede et
al., 2004
Ramesh
and
Knuckels,
2006
This document is a draft for review purposes only and does not constitute Agency policy.
               B-10        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
BPDE-
DNA
adducts




CAs













CAs


CAs





Test system
Rat, Wistar; liver and
peripheral blood
lymphocyte adducts




Mouse, C57 (high AHH
inducible) and DBA
(lowAHHinducible)
strains; 11-d-old
embryos; adult bone
marrows








Mouse, 1C3F1 hybrid
(101/ElxC31xEl)Fl;
CAs in bone marrow
Rat, Wistar; peripheral
blood lymphocytes




Test conditions
Single dose by gavage; sacrificed 24 hrs
post-dosing; peripheral blood
lymphocytes and liver DNA analyzed
by [32P]-postlabelingfor BP-DNA
adducts.


Study used four matings (female x
male): C57 x C57; DBA x DBA; C57 x
DBA; and DBA x C57; pregnant mice
treated orally on GD 11 with
benzo[a]pyrene; sacrificed 15 hrs
posttreatment; material liver, bone
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.


Single dose by gavage; sacrificed 30 hrs
of post-dosing; bone marrow from
femur isolated and analyzed for CAs.
Single dose by gavage; sacrificed 6, 24,
and 48 hrs posttreatment; blood from
abdominal aorta collected, whole
blood cultures set up, CAs scored in
100 first-division peripheral blood
lymphocytes per animal.
Results
+






+













+


-





Dose
0, 10, or
100 mg/kg





150 mg/kg













63 mg/kg


0, 10, 100, or
200 mg/kg




Comment
At 100 mg/kg dose, total
adduct levels in peripheral
blood lymphocytes were
twofold higher than the levels
in liver; adduct profiles differed
between peripheral blood
lymphocytes and liver.
Levels of CAs: hybrid embryos
> homozygous DBA embryos >
homozygous C57 embryos;
tissue AHH activity: C57
mothers and their embryos >
DBA females and their
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.
Significant increase in CAs in
benzo[a]pyrene-treated
animals compared to controls.
No difference between control
and treatment groups at any
dose or at any sampling time
observed.


Reference
Willems et
al., 1991





Adler et
al., 1989












Adler and
Ingwersen,
1989
Willems et
al., 1991




This document is a draft for review purposes only and does not constitute Agency policy.
               B-ll        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
CAs
MN
MN
MN
MN
MN
MN
Test system
Hamster; bone
marrow
Mouse, /ocZtransgenic
(Muta™Mouse)
Mouse, B6C3Fi
(hybrid)
Mouse, CD-I and
BDF1; bone marrow
Mouse, CD-I and
BDF1, peripheral blood
reticulocytes
Mouse, ICR[Hsd:
(ICR)Br]
Mouse, Swiss albino;
bone marrow
Test conditions
Single, i.p. injection of benzo[a]pyrene
dissolved in tricapryline; animals
sacrificed 24 hrs post-exposure.
Benzo[a]pyrene given via gavage in
olive oil daily for 28 consecutive d;
blood samples were collected 48 h
after last dose; % of PCEs and NCEs
reported.
i.p. injection; several doses given to
calculate LD50.
Dosed orally once, twice, or thrice at
24-hr intervals; sacrificed 24 hrs after
last treatment.
Given single i.p injection; tail blood
collected at 24-hr intervals from 0 to
72 hrs.
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.
Given orally in corn oil; sacrificed 24
hrs post-exposure.
Results
+
+
+
+
+
+
+
Dose
25, 50, or
100 mg/kg
25, 50, and 75
mg/kg-day
232 mg/kg
(LD50/7);
259 mg/kg
(LD50/4)
250, 500,
1,000, or
2,000 mg/kg
62.5, 125, 250,
or 500 mg/kg
150 mg/kg
75 mg/kg
Comment
Benzo[a]pyrene induced CAs at
50 mg/kg body weight only,
with negative responses at the
low and high dose.
Statistically significant, dose-
dependent increases in % PCEs
and NCEs at all doses.
Study conducted to determine
the toxicity of benzo[a]pyrene
(LD50).
Significant increase at all
doses; no dose-response;
double dosing at 500 mg/kg
dose gave best response.
Maximum response seen at 48
hrs posttreatment.
All groups significantly higher
than controls for MN; fetal
liver more sensitive than any
other group.

Reference
Bayer,
1978
Lemieux et
al., 2011
Salamone
etal., 1981
Shimada et
al., 1990
Shimada et
al., 1992
Harper et
al., 1989
Koratkar et
al., 1993
This document is a draft for review purposes only and does not constitute Agency policy.
               B-12        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
MN
MN
MN
MN
MN
MN
MN
MN
Test system
Mouse, Swiss; bone
marrow polychromatic
erythrocytes
Mouse, CD-I and
MS/Ae strains
Mouse, BDF1, bone
marrow
Mouse, HRA/Skh
hairless, keratinocytes
Mouse, HOS:HR-1,
hairless; skin
micronuclei
Mouse, HR-1 hairless,
skin (benzo[a]pyrene
with slight radiation)
Rat, Sprague-Dawley,
peripheral blood
reticulocytes
Rat, Sprague-Dawley,
pulmonary alveolar
macrophages
Test conditions
Given by gavage and sacrificed 36 hrs
posttreatment.
i.p. and p.o. administration.
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 posttreatment;
bone marrow smears prepared,
stained with May-Grunwald-Giemsa
technique and scored for MN
polychromatic erythrocytes.
Single topical application.
Topical application once daily for 3 d;
sacrificed 24 hrs after last treatment.

Given single i.p injection; tail blood
collected at 24-hr intervals from 0 to
96 hrs.
Intratracheal instillation, once/day for
3d.
Results
+
+
+
+
+
+
+
+
Dose
75 mg/kg
62.5, 125, 250,
or 500 mg/kg
0, 25, 50, or
60 mg/kg
0.5, 5, 50, 100,
or
500 mg/mous
e
0.4, 1, 2, or
4mg

62.5, 125, 250,
500, or
1,000 mg/kg
25 mg/kg
Comment

Good dose response by both
routes, strains; i.p. better than
P.O.; MS/Ae strain more
sensitive than CD-I strain.
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
postinjection.


Exposure to sunlight simulator
to evaluate photogenotoxicity
and chemical exposure.
Maximum response seen at 72
hrs posttreatment.

Reference
Rao and
Nandan,
1990
Awogi and
Sato, 1989
Balansky et
al., 1994
He and
Baker,
1991
Nishikawa
et al., 2005
Hara et al.,
2007
Shimada et
al., 1992
De Flora et
al., 1991
This document is a draft for review purposes only and does not constitute Agency policy.
               B-13        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
MN

MN


MN


DNA
strand
breaks








DNA
strand
breaks






DNA
strand
breaks
Test system
Rat, Sprague-Dawley,
bone marrow cells
Hamster; bone marrow


Fish (carp, rainbow
trout, clams); blood
and hemolymph
Rat, Sprague-Dawley;
comet assay









Aquatic organisms:
carp (Cyprinus carpio),
rainbow trout
(Oncorhynchus my kiss),
and clams (Spisula
sachalinensis); Comet
assay


Rat, Brown Norway


Test conditions
Intratracheal instillation, once/day for
3d.
Single, i.p. injection of benzo[a]pyrene
dissolved in tricaprylin; animals
sacrificed 30 hours post-exposure.



Instilled intratracheally with: (1) single
dose of benzo[a]pyrene in aqueous
suspension; sacrificed at 3, 24, and 48
hrs posttreatment; alveolar
macrophages, lung cells, and
lymphocytes, hepatocytes collected or
(2) dose-response study and sacrificed
at 24 hrs posttreatment; lungs
collected; controls received normal
saline instillation; all cells analyzed by
comet assay.
All organisms acclimatized in tanks for
2 d, 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 posttreatment; cell suspensions
prepared from liver (carp and trout) or
digestive gland (clam) for comet assay.
UDS determined after 5 and 18 hrs of a
single intragastric dosing.

Results
-

-


+


+










+








-


Dose
25 mg/kg

100, 300, or
500 mg/kg

0.05, 0.25, 0.5,
orl ppm

Experiment
#1: 3mgof
benzo[a]pyren
e; Experiment
#2: dose-
response
study with
0.75, 1.5, or 3
mg
benzo[a]pyren
e
0.05, 0.25, 0.5,
and 1 ppm







62.5 mg/kg


Comment








All time points showed
significant increase in SSBs
(Experiment #1); a dose-
response in SSBs was observed
(Experiment #2).






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
concentration.


Negative at both time points.


Reference
De Flora et
al., 1991
Bayer,
1978

Kim and
Hyun, 2006

Garry et
al., 2003a,
b








Kim and
Hyun, 2006







Mullaart et
al., 1989

This document is a draft for review purposes only and does not constitute Agency policy.
               B-14        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
UDS
UDS
UDS
UDS
UDS
SCEs
Test system
Rat, F344
Mouse, HOS:HR-1
hairless; skin
Rat, Brown Norway;
liver
Mouse, (C3Hfxl01)Fl
hybrid, germ cells
Mouse, early
spermatid
Hamster; SCEs in bone
marrow
Test conditions
Single i.p. injection of benzo[a]pyrene
or DMSO; sacrificed at 2 or 12 hrs
post-exposure; liver isolated,
hepatocyte cultures were set up and
incubated with 10 mCi/mL [3H]-
thymidine for 4 hrs; washed and
autoradiography performed.
Single topical application on two spots
on the backs after stripping stratum
corneum with adhesive tape to
enhance penetration; sacrificed 24 hr
posttreatment, skin isolated
[3H]thymidine; cultured; epidermal
UDS measured.
Single intragastric injection; sacrificed
at 5 and 18 hrs post-injection.
i.p. injection of benzo[a]pyrene;
[3H]-thymidine injection later.
i.p. injection.
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.
Results

+
—
—

+
Dose
100 mg/kg
0, 0.25, 0.5,
and 1% (w/v)
in acetone
62.5 mg/kg
0.3 ml
250-500
mg/kg
450 mg/kg
Comment
Benzo[a]pyrene was negative
at both time points.
UDS index showed a dose-
dependent increase up to 0.5%
benzo[a]pyrene dose and then
plateaued.
Benzo[a]pyrene was negative
at both time points.
Concentration not specified.
Reviewed by Sotomayor and
Sega (2000).
Significant increase in
metaphase SCEs in
benzo[a]pyrene-treated
animals compared to vehicle-
treated controls.
Reference
Mirsalis et
al., 1982
Mori et al.,
1999
Mullaart et
al., 1989
Sega, 1979
Sega, 1982
Roszinsky-
Kocher et
al., 1979
This document is a draft for review purposes only and does not constitute Agency policy.
               B-15        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
SCEs
SCEs
SCEs
SCEs
SCEs
SCEs
Test system
Hamster
Hamster; fetal liver
Hamster; bone marrow
Mouse, DBA/2 and
C57BL/6, bone marrow
cells
Mouse, DBA/2 and
C57BL/6, splenic
lymphocytes
Rat, Wistar; peripheral
blood lymphocytes
Test conditions
Animals implanted s.c. with BrdU
tablet; 2 hrs later given phorone (125
or 250 mg/kg) i.p.; another 2 hrs later
dosed i.p. with benzo[a]pyrene; 24 hrs
post-BrdU dosing, animals injected
with colchicine 10 mg/kg body weight,
sacrificed 2 hrs later; bone marrow
from femur prepared for SCE assay.
i.p. injection to pregnant animals on
GDs 11, 13, or 15; fetal liver SCEs were
analyzed.
NA
Two intragastric injections given; mice
implanted with BrdU tablets, sacrificed
on d 5, SCEs estimated.
Two intragastric injections given; mice
killed on 5th day and cells cultured for
48 hrs with BrdU.
Single dose by gavage; sacrificed 6, 24,
and 48 hrs posttreatment; blood from
abdominal aorta collected, whole
blood cultures set up, SCEs scored in
50 second-division metaphases in
peripheral blood lymphocytes per
animal.
Results
+
+
+
+
+
+
Dose
50 or
100 mg/kg
50 and
125 mg/kg
2.5, 25, 40, 50,
75, or
100 mg/kg
10 or
100 mg/kg
10 or
100 mg/kg
0, 10, 100, or
200 mg/kg
Comment
SCEs increased with low dose
of phorone significantly.
Produced doubling of SCE
frequency.
Frequency of SCEs increased
>40 mg/kg body weight
SCEs and BP-DNA adducts in
the order of C57BI/6(AHH-
inducible) < DBA/2 (AHH-
noninducible).
SCEs and BP-DNA adducts in
the order of C57BI/6 (AHH-
inducible) < DBA/2 (AHH-
noninducible).
Linear dose-response at any
sampling time; however,
significant at the highest dose
only; no interaction between
dose and sampling time.
Reference
Bayer et
al., 1981
Pereira et
al., 1982
Bayer,
1978
Wielgosz
etal., 1991
Wielgosz
etal., 1991
Willems et
al., 1991
This document is a draft for review purposes only and does not constitute Agency policy.
               B-16        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
Endpoint
Mutation



Mutation


Mutation



Mutation


Mutation








Mutation







Test system
Drosophila
melanogaster, sex-
linked recessive lethal
test
D. melanogaster, sex-
linked recessive lethal
test
D. melanogaster,
Berlin-K and Oregon-K
strains; sex-linked
recessive lethal test
D. melanogaster, sex-
linked recessive lethal
test
D. melanogaster,
Canton-S (WT) males,
FM6 (homozygousfor
an X chromosome)
females; sex-linked
recessive lethal test



D. melanogaster;
somatic mutation, eye
color mosaicism





Test conditions
Base males exposed to benzo[a]pyrene
were mated with virgin females of
Berlin K or me/-9L1strains.

Adult Berlin males treated orally with
benzo[a]pyrene.

Benzo[a]pyrene dissolved in special fat
and injected into the abdomen of flies.


Male Berlin K larvae treated with
benzo[a]pyrene for 9-11 d.

Adult male flies were fed on filters
soaked in benzo[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 d; four broods
obtained; a group of 100 daughters of
each male were mated again; scored
for percent lethal.
Fifty 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
mosaic eye sectors.
Results
+



+


-



+


-








+







Dose
10 mM



5 or 7.5 mM


2 or 5 mM



0.1-4 mM


250 or 500
ppm







1, 2, or 3 mM







Comment
Data inconclusive due to low
fertility rates of mei-9L1
females.

Low mutagenic activity.


Negative at both doses.



Threefold enhancement in
lethals in treated versus
controls.
Authors report incomplete
dissolution of benzo[a]pyrene
in DMSO as a possible cause of
negative result.





Benzo[a]pyrene was effective
as a mutagen; no dose-
response observed.





Reference
Vogel et
al., 1983


Vogel et
al., 1983

Zijlstra and
Vogel,
1984

Vogel et
al., 1983

Valencia
and
Houtchens,
1981





Fahmy and
Fahmy,
1980





This document is a draft for review purposes only and does not constitute Agency policy.
               B-17        DRAFT—DO NOT CITE OR QUOTE

-------
                                                                                      Toxicological Review ofbenzo[a]pyrene
Endpoint
Cell trans-
formation





Test system
Hamster, LVG:LAK
strain (virus free);
transplacental host-
mediated assay



Test conditions
Pregnant animals dosed i.p. with
benzo[a]pyrene on GD 10; sacrificed
on GD 13, fetal cell cultures prepared,
10 x 10s cells/plate; 5 d post-culture
trypsinized; subcultured every 4-6 d
thereafter and scored for plating
efficiency and transformation.
Results
+






Dose
3 mg/100 g
body weight





Comment







Reference
Quarles et
al., 1979





1
2
     CSB = Cockayne syndrome; 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 and white and narrow eyes; Li = liver; Lu = lung; Sk = skin; DOS = unscheduled DNA synthesis; XPA = xeroderma
     pigmentosum group A
                               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-18        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene


 1    Tumor Promotion and Progression
 2    Cytotoxicity and inflammatory response
 3          The cytotoxicity of benzo[a]pyrene metabolites may contribute to tumor promotion via
 4    inflammatory responses leading to cell proliferation (Burdick et al., 2003). Benzo[a]pyrene is
 5    metabolized to o-quinones, which are cytotoxic, and can generate ROS (Bolton et al., 2000; Penning,
 6    1999). Benzo[a]pyrene o-quinones reduce the viability and survival of rat and human hepatoma
 7    cells (Flowers-Geary et al., 1996,1993).  Cytotoxicity was also induced by benzo[a]pyrene and
 8    BPDE in a human prostate carcinoma cell line (Nwagbara et al., 2007). Inflammatory responses to
 9    cytotoxicity may contribute to the tumor promotion process. For example, benzo[a]pyrene
10    quinones (1,6-, 3,6-, and 6,12-benzo[a]pyrene-quinone) generated ROS and increased cell
11    proliferation by enhancing the epidermal growth factor receptor pathway in cultured breast
12    epithelial cells (Burdick et al., 2 0 0 3).
13          Several studies have demonstrated that exposure to benzo[a]pyrene increases the
14    production of inflammatory cytokines, which may contribute to cancer progression. Garcon et al.
15    (2001a, b) exposed Sprague-Dawley rats by inhalation to benzo[a]pyrene with or without ferrous
16    oxide (Fe20s) particles. They found that benzo[a]pyrene alone or in combination with Fe20s
17    particles elicited mRNA and protein synthesis of the inflammatory cytokine, IL-1. Tamaki et al.
18    (2004) also demonstrated a benzo[a]pyrene-induced increase in IL-1 expression in a human
19    fibroblast-like synoviocyte cell line (MH7A).  Benzo[a]pyrene increases the expression of the mRNA
20    for CCL1, an inflammatory chemokine, in human macrophages (N'Diaye et al., 2006). The
21    benzo[a]pyrene-induced increase in CCL1 mRNA was inhibited by the potent AhR antagonist,
22    3'-methoxy-4'-nitroflavone.

23    AhR-mediated effects
24          The promotional effects of benzo[a]pyrene may also be related to AhR affinity and the
25    upregulation of genes related to biotransformation (i.e., induction of CYP1A1), growth, and
26    differentiation (Bostrom et al., 2002). Figure B-3 illustrates the function of the AhR and depicts the
27    genes regulated by this receptor as belonging to two major functional groups (i.e., induction of
28    metabolism or regulation cell differentiation and proliferation). PAHs bind to the cytosolic AhR in
29    complex with heat shock protein 90 (Hsp90). The ligand-bound receptor is then transported to
30    nucleus in complex with the Ah receptor nuclear translocater. The AhR complex interacts with the
31    Ah responsive elements of the DNA to increase the transcription of proteins associated with
32    induction of metabolism and regulation of cell differentiation and proliferation.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-l          DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
        *
        PAH
Hsp90
Hsp90

ARNT

                                            Enhanced
                                            specific
                                            mRNA
                                            production
•^ 	



IB
^r
AHR
ARNT
AHREDNA


                                    Increased
                                    synthesis of
                                    PAH metabolizing
                                    enzymes
                                                 Increased
                                                 synthesis of
                                                 proteins that
                                                 regulate cell
                                                 differentiation and
                                                 proliferation
AHREoNA = Ah-responsive elements of DNA; ARNT = Ah receptor nuclear translocater; Hsp90 = heat
shock protein 90
Source:  Okey etal. (1994).

Figure B-3. 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 (Nebertetal., 1993). Ma and Lu (2007)
reviewed several studies of benzo[a]pyrene toxicity and tumorigenicity 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 etal. (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 CYP1A1 expression in the skin and liver of
AhR-positive mice (+/- or +/+), but CYP1A1 expression was not altered by benzo[a]pyrene
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-2         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    treatment in AhR knock-out mice (-/-)• Talaskaetal. (2006) also showed that benzo[a]pyrene
 2    adduct levels in skin were reduced by 50% in CYP1A2 knock-out mice and by 90% in AhRknock-
 3    out mice compared with WT C57B16/J mice following a single dermal application of 33 mg/kg
 4    benzo[a]pyrene for 24 hours. Ma and Lu (2007) further noted that Ah-nonresponsive mice were at
 5    greater risk of toxicity and tumorigenicity in remote organs, distant from the site of exposure (i.e.,
 6    bone marrow). As an example, Uno et al. (2006) showed thatbenzo[a]pyrene  (125 mg/kg-day, p.o.
 7    for 18 days) caused marked wasting, immunosuppression, and bone marrow hypocellularity in
 8    CYP1A1 knock-out mice, but not in WT mice.
 9          Some studies have demonstrated the formation of DNA adducts in the liver of AhR knock-
10    out mice following i.p. or oral exposure to benzo[ajpyrene (Sagredo etal., 2006; Uno etal., 2006;
11    Kondraganti et al., 2003).  These findings suggest that there may be alternative (i.e., non-AhR
12    mediated) mechanisms  of benzo [a] pyrene activation in the mouse liver. Sagredo etal. (2006)
13    studied the relationship between the AhR genotype and GYP metabolism in different organs of the
14    mouse. AhR+/+, +/~, and -/- mice were treated once with 100 mg/kg benzo[a]pyrene by gavage.
15    CYP1A1, CYP1B1, and AhR expression was evaluated in the lung, liver, spleen,  kidney, heart, and
16    blood, via real-time or reverse transcriptase polymerase chain reaction, 24 hours after treatment.
17    CYP1A1 RNA was increased in the lung and liver and CYP1B1 RNA was  increased in the lung
18    following benzo[a]pyrene treatment in AhR+/+ and+/- mice (generally higher in heterozygotes).
19    Benzo [a]pyrene treatment did not induce CYP1A1  or CYP1B1 enzymes  in AhR-/- mice. The
20    expression of CYP1A1 RNA, as standardized to (3-actin expression, was generally about 40 times
21    that of CYP1B1. The concentration of benzo[a]pyrene metabolites and the levels of DNA and
22    protein adducts were increased in mice lacking the AhR, suggesting that there  may be an
23    AhR-independent pathway for benzo[a]pyrene metabolism and activation. The high levels of
24    benzo [ajpyrene DNA adducts in organs other than the liver of AhR-/- mice may be the result of slow
25    detoxification of benzo[a]pyrene in the liver, allowing high concentrations of the parent compound
26    to reach distant tissues.
27          Uno et al. (2006) also demonstrated a paradoxical increase in liver DNA adducts in AhR
28    knock-out mice following oral exposure to benzo[ajpyrene. WT C57BL/6 mice and several knock-
29    out mouse strains (CYP1A2-/- and CYP1B1-/- single knock-out, CYP1A1/1B1-/- and CYP1A2/1B1-/-
30    double knock-out) were studied. Benzo[ajpyrene was administered in the feed at 1.25,12.5, or 125
31    mg/kg for 18 days (this  dose is well tolerated by WT C57BL/6 mice for  1 year, but lethal within 30
32    days to the CYP1A1-/- mice). Steady-state blood levels of benzo[a]pyrene, reached within 5 days of
33    treatment, were -25 times higher in CYP1A1-/- and -75 times higher in CYP1A1/1B1-/- than in WT
34    mice, while clearance was similar to WT mice in the other knock-out mouse strains. DNA adduct
35    levels, measured by [32P]-postlabeling in liver, spleen, and bone marrow, were highest in the
36    CYP1A1-/- mice at the two higher doses, and in the CYP1A1/1B1-/- mice at the mid dose only.
37    Adduct patterns, as revealed by 2-dimensional chromatography, differed substantially between
3 8    organs in the various knock-out types.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                               B-3         DRAFT—DO NOT CITE OR  QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1           Dertinger etal. (2001, 2000) demonstrated that AhR signaling may play a role in
 2    cytogenetic damage caused by benzo[a]pyrene.  The in vivo formation of MN in peripheral blood
 3    reticulocytes of C57B1/6J mice induced by a single i.p. injection of benzo[a]pyrene (150 mg/kg) was
 4    eliminated by prior treatment with the potent AhR antagonist 3'-methoxy-4'-nitroflavone. This
 5    antagonist also protected AhR null allele mice from benzo[a]pyrene-induced increases in MN
 6    formation, suggesting that 3'-methoxy-4'-nitroflavone may also act through a mechanism
 7    independent of the AhR (Dertinger etal., 2000).
 8           Several in vitro studies have suggested that the AhR plays a role in the disruption of cell
 9    cycle control, possibly leading to cell proliferation and tumor promotion following exposure to
10    benzo[a]pyrene (Andrysik et al., 2007; Chung etal., 2007; Chen etal., 2003). Chung etal. (2007)
11    showed that benzo[a]pyrene-induced cytotoxicity and apoptosis  in mouse hepatoma (Hepalclc7)
12    cells occurred through a p53 and caspase-dependent process requiring the AhR. An accumulation
13    of cells in the S-phase of the cell cycle  (i.e., DNA synthesis and replication) was also observed,
14    suggesting that this process may be related to cell proliferation. Chen et al. (2003) also
15    demonstrated the importance of the AhR in benzo[a]pyrene-7,8-dihydrodiol- and BPDE-induced
16    apoptosis in human HepG2 cells. Both the dihydrodiol and BPDE affected Bcl2 (a member of a
17    family of apoptosis suppressors) and activated caspase and p3 8 mitogen-activated protein (MAP)
18    kinases, both enzymes that promote apoptosis. When the experiments were conducted in a cell line
19    that does not contain Ah receptor nuclear translocator (see Figure 4-1), the dihydrodiol was not
20    able to initiate apoptotic event sequences, indicating that activation to BPDE by CYP1A1 was
21    required. BPDE did not induce apoptosis-related events in a p38-defective cell line, illustrating the
22    importance of MAP kinases in this process. In rat liver epithelial  cells (WB-F344 cells), in vitro
23    exposure to benzo[a]pyrene resulted in apoptosis, a decrease in cell  number, an increase in the
24    percentage of cells in S-phase (comparable to a proliferating population of WB-F334 cells), and
25    increased expression of cell cycle proteins (e.g., cyclin A) (Andrysik et al., 2007). Benzo[a]pyrene-
26    induced apoptosis was attenuated in cells transfected with a dominant-negative mutation of the
27    AhR.

28    Inhibition of gap junctional intercellular communication (GJIC)
29           Gap junctions are channels between cells that allow substances of a molecular weight up to
30    roughly 1 kDa to pass from one cell to the other. This process of metabolic cooperation is crucial
31    for differentiation, proliferation, apoptosis, and cell death and consequently for  the two epigenetic
32    steps of tumor formation, promotion,  and progression. Chronic exposure to many toxicants results
33    in down-regulation of gap junctions. For tumor promoters, such  as TPA or TCDD, inhibition of
34    intercellular communication is correlated with their promoting potency (Sharovskaya et al., 2006;
35    Yamasaki, 1990).
36           Blaha et al. (2002)  surveyed the potency of 35 PAHs, including benzo[a]pyrene, to inhibit
37    GJIC. The scrape loading/dye transfer assay was employed using a rat liver epithelial cell line that
38    was incubated in vitro for 15, 30, or 60 minutes with 50 [M benzo[a]pyrene.  After incubation, cells
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-4          DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1   were washed, and then a line was scraped through the cells with a surgical blade. Cells were
 2   exposed to the fluorescent dye lucifer yellow for 4 minutes and then fixed with formalin.  Spread of
 3   the dye from the scrape line into cells remote from the scrape was estimated under a fluorescence
 4   microscope. Benzo[a]pyrene reduced spread of the dye after 30 minutes of exposure
 5   (approximately 50% of control). Recovery of GJIC was observed 60 minutes after exposure.
 6          Sharovskaya et al. (2006) studied the effects of carcinogenic and noncarcinogenic PAHs on
 7   GJIC in HepG2 cells. Individual carcinogenic PAHs inhibited GJIC in a temporary fashion (70-100%
 8   within 24 hours), but removal of the PAH from culture reversed the effect Noncarcinogenic PAHs
 9   had very little effect on GJIC. Benzo[a]pyrene at 20 [M inhibited GJIC completely within 24 hours,
10   while its noncarcinogenic homolog, benzo[e]pyrene, produced <20% inhibition. The effect was not
11   AhR-dependent, because benzo[a]pyrene inhibited GJIC in HepG2 cells to the same extent as in
12   hepatoma G27 cells, which express neither CYP1A1 nor AhR. The authors concluded that the
13   effects of benzo[a]pyrene and benzo[e]pyrene on GJIC were direct (i.e., not caused by metabolites).
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                B-5         DRAFT—DO NOT CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene
 i   APPENDIX C.  DOSE-RESPONSE MODELING FOR
 2   THE DERIVATION  OF  REFERENCE VALUES FOR
 3   EFFECTS OTHER THAN CANCER AND THE
 4   DERIVATION OF CANCER RISK  ESTIMATES
 5         This appendix provides technical detail on dose-response evaluation and determination of
 6   points of departure (POD) for relevant toxicological endpoints. Except where other software is
 7   noted, all endpoints were modeled using the U.S. EPA's Benchmark Dose Software (BMDS; U.S. EPA,
 8   2012; version 2.0 or later). The preambles for the cancer and non-cancer parts below describe the
 9   common practices used in evaluating the model fit and selecting the appropriate model for
10   determining the POD, as outlined in the draft Benchmark Dose Technical Guidance Document (U.S.
11   EPA, 2000).

12   DOSE-RESPONSE MODELING FOR DERVIATION OF RFD

13   Evaluation of Model Fit
14         For each dichotomous endpoint, BMDS dichotomous models were fitted to the data using
15   the maximum likelihood method.  Each model was tested for goodness-of-fit using a chi-square
16   goodness-of-fit test (x2 p-value < 0.10 indicates lack of fit). Other factors were also used to assess
17   model fit, such as scaled residuals, visual fit, and adequacy of fit in the low-dose region and in the
18   vicinity of the BMR.
19         For each continuous endpoint, BMDS continuous models were fitted to the data using the
20   maximum likelihood method. Model fit was assessed by a series of tests as follows.  For each model,
21   first the homogeneity of the variances was tested using a likelihood ratio test (BMDS Test 2). If Test
22   2 was not rejected (x2 p-value > 0.10), the model was fitted to the data assuming constant variance.
23   If Test 2 was rejected (x2 p-value < 0.10), the variance was modeled as a power function of the
24   mean, and the variance model was tested for adequacy of fit using a likelihood ratio test (BMDS
25   Test 3). For fitting models using either constant variance or modeled variance, models for the mean
26   response were tested for adequacy of fit using a likelihood ratio test (BMDS Test 4, with x2 p-value <
27   0.10 indicating inadequate fit). Other factors were also used to assess the model fit, such as scaled
28   residuals, visual fit, and adequacy of fit in the low-dose region and in the vicinity of the BMR.

29   Model Selection
30         For each endpoint, the BMDL estimate (95% lower confidence limit on the BMD, as
31   estimated by the profile likelihood method) and AIC value were used to select a best-fit model from
32   among the models exhibiting adequate fit. If the BMDL estimates were "sufficiently close," that is,

              This document is a draft for review purposes only and does not constitute Agency policy.
                                          C-l        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
1   differed by at most threefold, the model selected was the one that yielded the lowest AIC value. If
2   the BMDL estimates were not sufficiently close, the lowest BMDL was selected as the POD.

3   Decreased thvmus weight, males fKroese et al, 20011
4
5
Table C-l. Means ± SDa for thymus weight in male Wistar rats exposed
to benzo[a]pyrene by gavage 5 days/week for 90 days
Organ
Thymus weight
(mg), males
Dose (mg/kg-d)
0
380 ± 60
3
380 ± 110
10
330 ± 60
30
270 ± 40b
6
7
aReported as SE, but judged to be SD (and confirmed by study authors).
Significantly (p < 0.05) different from control mean; student t-test (unpaired, two-tailed); n
10/sex/group.

      Table C-2. Model predictions for decreased thymus weight in male
      Wistar rats—90 days
Model
Variance
p-valuea
Goodness-of-
fit
p-value
AIC
BMD1SD
(mg/kg-d)
BMDL1SD
(mg/kg-d)
Constant variance
Linear
0.01
0.74
384.84
12.97
8.97
Nonconstant variance
Hill"
Linear, Polynomial (2-
degree), Power"
Insufficient degrees of freedom
0.30
0.23
380.71
16.40
11.30
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-2          DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
                                       Linear Model with 0.95 Confidence Level
 2

 3
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
               450
               400
               350
               300
               250
                           Linear
                                                   BMDL
                                                                BMD
                                                   10

                                                    dose
                                                            15
20
        15:3310/152009
       Figure C-l. Fit of linear model (nonconstant variance) to data on
       decreased thymus weight in male Wistar rats—90 days.


       BMDs and BMDLs indicated are associated with a change of 1 SD from the

       control, and are in units of mg/kg-day.
d)
 BMDS Model Run
   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



          This document is a draft for review purposes only and does not constitute Agency policy.

                                          C-3         DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
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
                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       -0.048       0.061


   beta_0        0.048       -0.048            1       -0.84


   beta  1       -0.061        0.061        -0.84           1
                              Parameter Estimates
      Variable

        lalpha

          rho

        beta_0

        beta 1
   95.0% Wald Confidence Interval

Lower  Conf. Limit   Upper Conf.  Limit

       -37.9473           0.288754

       1.38062            7.94967

       346.558            411.351

       -7.11189           -3.17249
Model  Descriptions for  likelihoods  calculated
                     Likelihoods of  Interest
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-4          DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
9
10
11
12
13
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
Model Log (likelihood) #
Al -189.116991
A2 -183.673279
A3 -184.883626
fitted -186.353541
R -196.353362

Explanation of Tests

Param's AIC
5 388.233982
8 383.346558
6 381.767253
4 380.707081
2 396.706723



Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al v

s A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit?
(Note: When rho=0 the results of Test 3

Tests of Interest

Test -2*log (Likelihood Ratio) Test

Test 1 25.3602 6
Test 2 10.8874 3
Test 3 2.42069 2
Test 4 2.93983 2

The p-value for Test 1 is less than .05.
difference between response and/or varianc
It seems appropriate to model the data

The p-value for Test 2 is less than .1. A
model appears to be appropriate

The p-value for Test 3 is greater than .1.
to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data


Benchmark Dose Computation

Specified effect = 1

Risk Type = Estimated standard

Confidence level = 0.95

BMD = 16.4008


BMDL = 11.2965
(A3 vs. fitted)
and Test 2 will be the same.)



df p-value

0.0002928
0. 01235
0.2981
0.2299

There appears to be a
es among the dose levels


non-homogeneous variance


The modeled variance appears


The model chosen seems







deviations from the control mean







53
               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-5          DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
1    Decreased thvmus weight, females CKroese et al. 20011
2
3
Table C-3. Means ± SDa for thymus weight in female Wistar rats
exposed to benzo[a]pyrene by gavage 5 days/week for 90 days
Organ
Thymus weight
(mg) - Females
Dose (mg/kg-d)
0
320 ± 60
3
310 ±50
10
300 ± 40
30
230±30b
4
5
aReported as SE, but judged to be SD (and confirmed by study authors).
Significantly (p < 0.05) different from control mean; student t-test (unpaired, two-tailed); n =
10/sex/group.

      Table C-4. Model predictions for decreased thymus weight in female
      Wistar rats—90 days
Model (constant variance)
Hillb
Linear0
Polynomial (2-degree)c'd
Powerb
Variance
p-valuea
Mean
p- value3
AIC
BMD1SD
(mg/kg-d)
BMDL1SD
(mg/kg-d)
NA
0.17
0.17
0.81
0.77
349.12
350.80
10.52
13.29
7.64
7.77
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 (MLE) of the dose/concentration associated with the
     selected BMR; NA = not applicable; model failed to generate
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-6          DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
                                        Linear Model with 0.95 Confidence Level
 1
 2
 4

 5

 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
               360
               340
               320
               300
               280
               260
               240
               220
               200
                            Linear
                                         BMDL
                                                BMD
                      0
                                              10

                                               dose
15
20
   16:2710/152009


       BMDs and BMDLs indicated are associated with a change of 1 SD from the

       control, and are in units of mg/kg-day.



       Figure C-2. 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:
^:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese2001\9Oday\thymusweight\female\durationadjusted\2Linkrolin
. (d)
        Gnuplot  Plotting File:
^:\USEPA\IRIS\benzo[a]pyrene\RfD\Kroese2001\9Oday\thymusweight\female\durationadjusted\2Linkrolin
.pit
                                                Thu Oct 15  16:27:44 2009
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                 C-7          DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
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
  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
                              Parameter Estimates
                                                     95.0% Wald Confidence Interval

                                                  Lower Conf. Limit   Upper Conf.  Limit

                                                         1098.16            2811.69

                                                         303.558             340.73

                                                        -5.84334           -2.56026
                                                     44.

                                                     44.

                                                     44.

                                                     44.
Model  Descriptions for likelihoods calculated
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-8          DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
8
9
10
11
12
13
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


Likelihoods of Int

Model Log (likelihood)
Al -171.357252
A2 -168.857234
A3 -171.357252
fitted -171.562118
R -181.324151


Explanation of Tests



erest

# Param's AIC
5 352.714504
8 353.714467
5 352.714504
3 349.124237
2 366.648303




Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al

vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit
(Note: When rho=0 the results of Test 3

Tests of Interest

Test -2*log (Likelihood Ratio) Test

Test 1 24.9338 6
Test 2 5. 00004 3
Test 3 5.00004 3
Test 4 0.409733 2

The p-value for Test 1 is less than .05.
difference between response and/or varian
It seems appropriate to model the data

The p-value for Test 2 is greater than .1
model appears to be appropriate here


The p-value for Test 3 is greater than .1
to be appropriate here

The p-value for Test 4 is greater than .1
to adequately describe the data


Benchmark Dose Computation

Specified effect = 1

Risk Type = Estimated standard

Confidence level = 0.95

BMD = 10.5228


BMDL = 7.64037
? (A3 vs. fitted)
and Test 2 will be the same.)



df p-value

0.0003512
0.1718
0.1718
0. 8148

There appears to be a
ces among the dose levels


A homogeneous variance



The modeled variance appears


The model chosen seems







deviations from the control mean







58
               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-9          DRAFT—DO NOT CITE OR QUOTE

-------
                                                 Toxicological Review ofbenzo[a]pyrene


1   Decreased ovary weight—female rats. 60 days (Xu et al, 2010)

2          Table C-5. Means ± SDs for ovary weight in female Sprague-Dawley rats
Organ
Ovary weight (mg)
Dose (mg/kg-d)a
0
0.160 ±0.0146
2.5
0.143 ±0.0098b
5
0.136 ±0.0098b
    aTWA doses over the 60-day study period.
    Statistically different (p < 0.05) from controls using one-way ANOVA.

           Table C-6. Model predictions for decreased ovary weight in female
           Sprague-Dawley rats
Model
Power
Linear, Polynomial (1°)
Goodness-of-fit
p-value
AIC
BMD1SD
(mg/kg-d)
BMDL1SD
(mg/kg-d)
NA
0.39
-138.67
2.27
1.49
           NA = not applicable; model failed to generate
             This document is a draft for review purposes only and does not constitute Agency policy.
                                            C-10        DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
                                          Linear Model with 0.95 Confidence Level
                 0.1
                 0.1
                 0.13
                              Linear
         16:0312/142010
                                       BMDL
                                                BMD
                                                  2            3

                                                       dose
 4
 5
 6
 1
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
       Figure C-3. Fit of linear/polynomial (1°) model to data on decreased

       ovary weight.




         Polynomial Model. (Version: 2.16;  Date:  05/26/2010)
         Input  Data File:
C:/USEPA/BMDS212/Data/benzo[a]pyrene/Bap_AbsOvaryWeight/Xu2010_AbsOvaryWeight_Linear_lSD.(d)
         Gnuplot  Plotting File:
C:/USEPA/BMDS212/Data/benzo[a]pyrene/Bap_AbsOvaryWeight/Xu2010_AbsOvaryWeight_Linear_lSD.plt
                                                 Tue Dec 14  13:51:32 2010
   Total number of dose  groups = 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
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-ll         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
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
                Default Initial Parameter Values

                       alpha =     0.000136

                         rho =           0   Specified

                      beta_0 =     0.158333

                      beta  1 =      -0.0048
                                                   95.0% Wald Confidence Interval

                                                Lower Conf. Limit   Upper Conf. Limit

                                                   4.12162e-005        0.000196562

                                                       0.150369           0.166298

                                                    -0.00726768        -0.00233232
Model  Descriptions for likelihoods calculated
          Model

           Al

           A2

           A3

       fitted
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                C-12        DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
8
9
10
11
12
13
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
R 67.008505


Explanation of Tests

2 -130.017010




Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)

Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit?
(Note: When rho=0 the results of Test 3

Tests of Interest

Test -2*log (Likelihood Ratio) Test

Test 1 12.9201 4
Test 2 1.40394 2
Test 3 1.40394 2
Test 4 0.861408 1

The p-value for Test 1 is less than .05.
difference between response and/or varianc
It seems appropriate to model the data

The p-value for Test 2 is greater than .1.
model appears to be appropriate here


The p-value for Test 3 is greater than .1.
to be appropriate here

The p-value for Test 4 is greater than .1.
to adequately describe the data


Benchmark Dose Computation

Specified effect = 1

Risk Type = Estimated standard

Confidence level = 0.95

BMD = 2.27159


BMDL = 1.49968
(A3 vs. fitted)
and Test 2 will be the same.)



df p-value

0.01167
0.4956
0.4956
0.3533

There appears to be a
es among the dose levels


A homogeneous variance



The modeled variance appears


The model chosen seems







deviations from the control mean







49
               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-13        DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 3
 4
 5
 6
 7
 9
10
                                              Toxicological Review ofbenzo[a]pyrene

Morris water maze results—male and female Sprague-Dawley rats. Chen et al. (2012)
      Data from Morris water maze was presented graphically in Chen et al., 2012, but dose group
means and standard deviations were provided upon request by the study authors which enabled
modeling of this endpoint In addition, the data for male and female rats were combined for dose-
response analysis because there was no substantive difference between males and females for each
dose group (supported by statistical testing using two-way ANOVA, and allowing for interactions),
and because there was no rationale or information available suggesting there would be sex-
mediated differences for these neurologic tests.

      Table C-7. Means ± SDs for Escape Latency and Time Spent in Target
      Quadrant
Test
Escape latency
(sec)
Time spent in
target quadrant
(sec)
Dose (mg/kg-d)
0
9.89 ±5. 76
33.6 ±8.92
0.02
12.5 ±5. 10
31.9 ±8.63
0.2
19.1 ±5. 85
16.6 ±5.74
2.0
33. 5 ±9.93
11.1 ±5. 12
11
12
      Table C-8. Model predictions for increase in Morris water maze test for
      escape latency, male and female rats
Model3
Hillb
Exponential 4, 5
Polynomial (2°)
Linear, Power
Exponential 2, 3
Goodness-of-fit
p-value
0.515
0.466
0.423
0.002
<0.001
AIC
386.3
386.4
386.6
396.7
400.3
BMD1SD
(mg/kg-d)
0.106
0.115
0.123
0.543
0.815
BMDL1SD
(mg/kg-d)
0.061
0.071
0.083
0.421
0.687
     a Includes modeling of heterogeneous variances (BMDS Test 3, p = 0.313).
     b Power parameter n was estimated to be 1 (boundary of parameter space).
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-14        DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
              20
              15  -
              10
         14:41 04/242012
                                                        Toxicological Review ofbenzo[a]pyrene

                                      Hill Model with 0.95 Confidence Level
       Figure C-4. Fit of Hill model to data on Morris water maze test escape
       latency.
 4
 5
 6

 8
 9
10
11
12
13
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
         Hill Model.  (Version:  2.16;  Date: 04/06/2011)

         Input Data File: C:\Documents and Settings\jfox\My Documents\_CURRENTWORK\_CAST

plus\BaP\BMDS\hil_Chen.FM.latency_Hil-ModelVariance-BMRlStd-Restrict.(d)

         Gnuplot  Plotting File:  C:\Documents and Settings\jfox\My Documents\_CURRENTWORK\_CAST

plus\BaP\BMDS\hil_Chen.FM.latency_Hil-ModelVariance-BMRlStd-Restrict.plt

                                                  Tue Apr 24 14:41:26 2012
   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
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-15         DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
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
                                                     95.0% Wald  Confidence Interval

                                                  Lower Conf.  Limit   Upper Conf. Limit

                                                          -1.0229

                                                          0.33391

                                                          8.86283

                                                          20.9783
 Variable

   lalpha

      rho

intercept
NA - Indicates that this parameter has hit a bound

     implied by some inequality constraint and thus

     has no standard error.
     Table of Data and Estimated Values of Interest


 Dose      N    Obs  Mean     Est Mean   Obs Std Dev  Est Std Dev   Scaled Res.
 Model  Descriptions for  likelihoods  calculated
 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
                      Likelihoods of  Interest
           Model

            Al
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-16         DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
Q
o
9
10
11
12
13
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
A2 -186.795503
A3 -187.957975
fitted -188.169983
R -234.549118


Explanation of Tests
8 389.591006
6 387.915949
5 386.339965
2 473.098237



Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al

vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4 : Does the Model for the Mean Fit
(Note: When rho=0 the results of Test 3

Tests of Interest

Test -2*log (Likelihood Ratio) Test

Test 1 95.5072 6
Test 2 12.008 3
Test 3 2.32494 2
Test 4 0.424016 1

The p-value for Test 1 is less than .05.
difference between response and/or varian
It seems appropriate to model the data

The p-value for Test 2 is less than .1.
model appears to be appropriate

The p-value for Test 3 is greater than .1
to be appropriate here

The p-value for Test 4 is greater than .1
to adequately describe the data


Benchmark Dose Computation

Specified effect = 1

Risk Type = Estimated standard

Confidence level = 0.95

BMD = 0. 106284

BMDL = 0.0609511

? (A3 vs. fitted)
and Test 2 will be the same.)



df p-value

<.0001
0. 007356
0.3127
0.5149

There appears to be a
ces among the dose levels


A non-homogeneous variance


The modeled variance appears


The model chosen seems







deviations from the control mean







This document is a draft for review purposes only and does not constitute Agency policy.
                               C-17        DRAFT—DO NOT CITE OR QUOTE

-------
 1

 2
                                                Toxicological Review ofbenzo[a]pyrene



       Table C-9. Model predictions for decrease in Morris water maze test for

       time spent in target quadrant, male and female rats
Model3
Exponential 4
Exponential 5
Hill
Linear, Power, Polynomial (1°, 2°, 3°)
Goodness-of-fit
p-value
0.576
NAb
NAb
<0.001
AIC
395.4
397.1
397.1
433.1
BMDisD
(mg/kg-d)
0.065
0.084
0.071
1.23
BMDL1SD
(mg/kg-d)
0.043
0.044
0.038
0.98
      ' Includes modeling of heterogenous variances (BMDS Test 3, p = 0.919).

      5 NA: insufficient degrees of freedom to evaluate chi-square.
                                Exponential Model 4 with 0.95 Confidence Level
             40
             35
             30
             25
             20
             15
             10
                            Exponential
                BMDL BMD
                                0.5
                                               1


                                              dose
                                                             1.5
        14:3504/242012
 5

 6
I?
;2
 3
:4

 6

:s
 9
20
21
22
23
       Figure C-5. Fit of Exponential 4 model to data on Morris water maze

       time spent in target quadrant.




         Exponential Model.  (Version:  1.7;  Date:  12/10/2009)
         Input Data File:  C:\Documents  and Settings\...\exp_Chen.FM.target_Exp-ModelVariance-
BMRlStd-Down.(d)                                       ~             ~



 BMDS Model Run
    Note: Y[dose] is the median response for exposure = dose;

         sign = +1 for increasing trend in data;
               This document is a draft for review purposes only and does not constitute Agency policy.

                                                C-18         DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1              sign = -1 for decreasing  trend.
 2
 3          Model  2 is nested within Models 3 and 4.
 4          Model  3 is nested within Model 5.
 5          Model  4 is nested within Model 5.


 8       Dependent variable = Mean
 9       Independent variable = Dose
 0       Data  are  assumed to be distributed: normally
 1       Variance Model: exp(lnalpha +rho *ln(Y[dose]))
 2       The variance is to be modeled as Var(i) = exp(lalpha  +  log(mean(i))

 4       Total number of dose groups = 4
 5       Total number of records with missing values = 0
 6       Maximum number of iterations = 250
 7       Relative  Function Convergence has been set to:  le-008
 8       Parameter Convergence has been set to: le-008

20       MLE solution provided: Exact
21


24
25
26
27
28
29
30
31
32
33
34
35
36                        Parameter Estimates
37
38                      Variable         Model 4
 7
L
49
50                Table of Stats From Input Data
51
52         Dose      N         Obs Mean     Obs Std Dev

54         --    -       	   	
55
56
57
58
59
60                     Estimated Values  of Interest
61
62          Dose      Est Mean      Est Std     Scaled Residual
63        	    	    	    	
64
65
66
67
68
71
72
73         Model Al:        Yij = Mu(i)  +  e(ij)
74                  Var{e(ij)} = SigmaA2
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   C-19          DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene
 2
 3
 4         Model A3:        Yij  = Mu(i) + e(ij)
 5                  Var{e(ij)}  = exp(lalpha + log(mean(i)
 9
 0
 1                                  Likelihoods  of  Interest
 2
 3                        Model      Log(likelihood)      DF         AIC
 4                       	    	    	   	
 5                           Al
 6                           A2
 7                           A3
 8                            R
_9                            4
20
21
22       Additive constant for all  log-likelihoods =     -73.52.  This constant added to the
23       above values gives the log-likelihood including the term that does not
24       depend on the model parameters.
25
26
27                                   Explanation of Tests
28
29
30
31
32
33
34
35
36                              Tests of Interest
37
38         Test          -2*log (Likelihood Ratio)       D. F.
 7
 8
 9
50
51
52

54
55
56
57
58
59
60       Benchmark Dose Computations:
61
62         Specified Effect = 1.000000
63
64               Risk Type = Estimated standard deviations from control
65
66         Confidence Level = 0.950000
67
68                     BMD =    0.0650194
 9
 0                    BMDL =    0.0432761
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   C-20         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
 1


 2
 5

 6
 7
 8
 9
10
11
12
13
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
Cervical epithelial hyperplasia -female ICR mice (Gao et al. 2011)


       Table C-10. Incidence of cervical epithelial hyperplasia
Observation
Cervical epithelial hyperplasia
Dose (mg/kg-day)a
0
0/26
0.71
4/26
1.4
6/25
2.9
7/24
a doses converted to mg/kg-day after adjustment for equivalent continuous dosing (2/7
days/week)


       Table C-ll. Model predictions for increased incidence of epithelial

       hyperplasia in female ICR mice
Model3
Gamma
Logistic
Log-logistic
Probit
Log-Probit
Multistage
Goodness-of-fit
p-value
0.6874
0.1422
0.8360
0.1544
0.0775
0.6874
AIC
82.2821
88.4607
81.7004
88.1151
88.2004
82.2821
BMD1SD
(mg/kg-d)
0.659
1.422
0.578
1.326
1.012
0.659
BMDL1SD
(mg/kg-d)
0.452
1.052
0.369
0.979
0.686
0.452
         Logistic Model. (Version: 2.13; Date: 10/28/2009)

         Input Data  File: C:\Users\hclynch\Documents\_Active  Projects\_FA498 IRIS\xBaP\IASC Aug

2011\bmd modeling\lnl_gao 2011 inflamm cells_0pt.(d)

         Gnuplot Plotting File: C:\Users\hclynch\Documents\_Active Projects\_FA498

IRIS\xBaP\IASC Aug 2011\bmd modeling\lnl_gao 2011 inflamm cells Opt.pit
 BMDS Model Run
   User has chosen the log transformed model
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                C-21         DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
8
9
10
11
12
13
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
—* ^x
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58


Asymptotic Correlation Matrix of Parameter Estimates

( ***



int

intercept






Variable
background
intercept
slope


The model parameter (s) -background -slope
have been estimated at a boundary point, or have been specified by the user
and do not appear in the correlation matrix )

ercept

1



Parameter Estimates

95.0% Wald Confidence Interval
Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit
0 * * *
-1.6502 * * *
1 * * *

* - Indicates that this value is not calculated.





Model
Full model
Fitted model
Reduced model

AIC:




Dose Est

0.0000 0.
0 . 7100 0 .
1.4000 0.
2 . 9000 0 .

ChiA2 = 0.86


Benchmark Dose

Specified effect

Risk Type

Confidence level

BMD

BMDL



Analysis of Deviance Table

Log (likelihood) # Param's Deviance Test d.f. P-value
-39.4267 4
-39.8502 1 0.847034 3 0.8382
-45.7739 1 12.6945 3 0.005346

81.7004


Goodness of Fit
Scaled
. Prob. Expected Observed Size Residual

0000 0.000 0.000 26 0.000
1200 3.119 4.000 26 0.532
2119 5.297 6.000 25 0.344
3577 8.584 7.000 24 -0.675

d.f. = 3 P-value = 0.8360


Computation

0.1

= Extra risk

0. 95

0.578668

0.368701
59
60
               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-22        DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
                                                     Toxicological Review ofbenzo[a]pyrene
INHALATION DOSIMETRY MODELING FOR RFC DERIVATION
                                              Wed, 03/17/2010. 02:07:20 PM EOT
                            Region: Entire Lung
           0.750
           0.600
        «  0.450
        LJ_
        E
        O

        S  0.300

        Q


           0.150
           0.0
                                                0.621
                   0.449
                                                     Species & Model Info:
                                                    Species/Geometry: Human Limited
                                                    FRC Volume: 3300.00 ml
                                                    Head Volume: 50.00 ml
                                                    Breathing Route: nasal
                                                     Breathing Parameters:
                                                    Tidal Volume: 860.00 nil
                                                    Breathing Frequency: 16.00 1/min
                                                    Inspiratory Fraction: 0.50
                                                    Pause Fraction: 0.00
                                                     Particle Properties:
                                                    Diameter: MVIAD: 1.70 urn
                                                    OSD: 1.00
                                                    Concentration: 4.20 ug/m'S
 4          Figure C-6. Human fractional deposition.

 5    Species  =  humanlimited
 6    FRC = 3300.0
 7    Head volume  = 50.0
 8    Density  =1.0
 9    Number of  particles  calculated =  single
10    Diameter = 1.7000000000000002 urn  MMAD
11    Inhalability =   yes
12    GSD =1.0
13    Breathing  interval:  One single breath
14    Concentration =4.2
15    Breathing  Frequency  = 16.0
16    Tidal Volume = 860.0
17    Inspiratory  Fraction =  0.5
18    Pause Fraction = 0.0
19    Breathing  Route = nasal
20
21    Region:  Entire Lung
22    Region:  Entire Lung
23    Region       Deposition  Fraction
24    —
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-23        DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofbenzo[a]pyrene
 1    Head   0.449
 2    TB     0.045
 3     P     0.127
 4    Total  0.621
                                                Wed, 03/17/2010, 02:15:27 PM EOT
                            Region: Entire Lung
           0.250 i-
           0.200
           0.150
         sz
         o
         £ 0.100
           0.050
           0.0
                                                 0.181
                    Head      TB
Total
                                  Region
 Species 6 Model Info:
Species/Geometry: Rat
FRC Vblume: 4.00 ml
Head Vblume: 0.42 ml
Breathing Route: nasal
 Breathing Parameters:
Tidal Vblume: 1.80 ml
Breathing Frequency: 102.00 1/rnin
Inspiratory Fraction: 0.50
Pause Fraction: 0.00
 Particle Properties:
Diameter: MMAD: 1.70 urn
GSD: 1.00
Concentration: 4.20 ugAri"3
 8           Figure C-7. Rat fractional deposition.

 9    Species  = rat
10    FRC =4.0
11    Head volume =0.42
12    Density  =1.0
13    Number of particles  calculated =  single
14    Diameter = 1.7000000000000002 urn   MMAD
15    Inhalability =   yes
16    GSD =1.0
17    Breathing interval:  One single breath
18    Concentration  = 4.2
19    Breathing Frequency  = 102.0
20    Tidal  Volume =1.8
21    Inspiratory Fraction  = 0.5
22    Pause  Fraction = 0.0
23    Breathing Route = nasal
24
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 C-24         DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
Region: Entire  Lung
Region: Entire  Lung
Region       Deposition Fraction


Head  0.072
TB    0.041
 P    0.068
Total 0.181
             This document is a draft for review purposes only and does not constitute Agency policy.
                                           C-25        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    DOSE-RESPONSE MODELING FOR CANCER RISK VALUES
 2          The EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend thatthe
 3    method used to characterize and quantify cancer risk from a chemical is determined by what is
 4    known about the mode of action of the carcinogen and the shape of the cancer dose-response curve.
 5    No biologically based models for BaP carcinogenicity following oral, inhalation, or dermal exposure
 6    were identified.

 7    Methods for the Oral Slope Factor and Inhalation Unit Risk
 8          Due to the occurrence of multiple tumor types, earlier occurrence with increasing exposure,
 9    and early termination of the high-dose group in each of the oral and inhalation carcinogenicity
10    studies (see Appendix B for study details), methods that can reflect the influence of competing risks
11    and intercurrent mortality on site-specific tumor incidence rates are preferred.  EPA has generally
12    used a model that incorporates the time at which death-with-tumor occurred as well as the dose;
13    the multistage-Weibull model is multistage in dose and Weibull in time, and has the form:
14
15          P(d, t) = l- exp[-(q0 + qid + q2tf +... + q*d*) x(t± t0)c],
16
17    where P(d, t) represents the lifetime risk (probability) of cancer at dose d (i.e., human equivalent
18    exposure in this case) and age t (in bioassay weeks); parameters qt > 0, for i = 0,1,..., k; t is the time
19    at which the tumor was observed; and c is a parameter which characterizes the change in response
20    with age. The parameter to represents the time between when a potentially fatal tumor becomes
21    observable and when it causes death, and is generally set to 0 either when all tumors are
22    considered incidental or because of a lack of data to estimate the time reliably. The dose-response
23    analyses were conducted using the computer software program MultiStage-Weibull (U.S. EPA,
24    2010), which is based on Weibull models drawn from Krewskietal. (1983). Parameters were
25    estimated using the method of maximum likelihood. From specific model fits using stages up to n-1,
26    where n is the number of dose groups, the model fit with the lowest AIC was selected.
27          Two general characteristics of the observed tumor types were considered prior to
28    modeling; allowance for different, although unidentified modes of action, and allowance for relative
29    severity of tumor types. First, etiologically different tumor types were not combined across sites
30    prior to modeling (that is, overall counts of tumor-bearing animals were not tabulated) in order to
31    allow for the possibility that different tumor types could have different dose-response relationships
32    due to different underlying mechanisms or factors, such as latency. Consequently, all of the tumor
33    types were also modeled separately.
34          Additionally, the multistage-Weibull model can address relative severity of tumor types by
3 5    distinguishing between tumors as being either fatal or incidental to the death of an animal in order
36    to adjust partially for competing risks.  In contrast to fatal tumors, incidental tumors are those
37    tumors thought not to have caused the death of an animal. Cause-of-death information for most

                This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-26         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    early animal deaths was provided by the investigators of both of the bioassays. In the rat study,
 2    tumors of the forestomach or liver were the principal cause of death for most animals dying or
 3    sacrificed (due to moribundity) before the end of the study, while tumors of the forestomach were
 4    the most common cause of early deaths in the mouse study.
 5           PODs for estimating low-dose risk were identified at doses at the lower end of the observed
 6    data, generally corresponding to 10% extra risk, where extra risk is defined as [P(d) - P(0)]/[l -
 7    P(0)]. The lifetime oral cancer slope factor for humans is defined as the slope of the line from the
 8    lower 95% bound on the exposure at the POD to the control response (slope factor = 0.1/BMDLio).
 9    This slope, a 95% upper confidence limit (UCL) represents a plausible upper bound on the true risk.
10           Although the time-to-tumor modeling helps account for competing risks associated with
11    decreased survival times and other tumors, considering the tumor sites individually still does not
12    convey the total amount of risk potentially arising from the sensitivity of multiple sites—that is, the
13    risk of developing any combination of the increased tumor types,  not just the risk of developing all
14    simultaneously. One  approach suggested in the Guidelines for Carcinogen Risk Assessment (U.S. EPA,
15    2005a) would be to estimate cancer risk from tumor-bearing animals. EPA traditionally used this
16    approach until the National Resource Council (NRC) document Science and Judgment (NRC, 1994)
17    made a case that this  approach would tend to underestimate overall risk when tumor types occur in
18    a statistically independent manner. In addition, application of one model to a composite data set
19    does not accommodate biologically relevant information that may vary across sites or may only be
20    available for a subset of sites. For instance, the time courses of the multiple tumor types evaluated
21    varied, as is suggested by the variation in estimates of c, from  1.5  (e.g., male rat skin or mammary
22    gland basal cell tumors), indicating relatively little effect of age on tumor incidence, to 3.7 (e.g., male
23    mouse alimentary tract tumors), indicating a more rapidly increasing response with increasing age
24    (in addition to exposure level).  The result of fitting a model with parameters that can reflect
25    underlying mechanisms, such as z in the multistage-Weibull model, would be difficult to interpret
26    with composite data (i.e., counts of tumor-bearing animals). A simpler model, such as the
27    multistage model, could be used for the composite data but relevant biological information would
28    then be ignored.
29           Following the recommendations of the NRC (1994) regarding combining risk estimates,
30    statistical methods that can accommodate the underlying distribution of slope factors are optimal,
31    such as through maximum likelihood estimation or through bootstrapping or Bayesian analysis.
32    However, these methods have not yet been extended to models such as the multistage-Weibull
33    model.  A method involving the assumption that the variability in  the slope factors could be
34    characterized by a normal distribution is detailed below (U.S. EPA, 2010). Using the results in
35    female rats to illustrate, the overall risk estimate involved the following steps:
36
37       (1) It was assumed that the tumor groupings modeled above were statistically independent—
3 8           that is, that the occurrence of a liver tumor was not dependent upon whether there was a

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                C-27         DRAFT—DO  NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1          forestomach tumor. This assumption cannot currently be verified, and if not correct, could
 2          lead to an overestimate of risk from summing across tumor sites. However, NRC (1994)
 3          argued that a general assumption of statistical independence of tumor-type occurrences
 4          within animals was not likely to introduce substantial error in assessing carcinogenic
 5          potency from rodent bioassay data.
 6
 7       (2) The models previously fitted to estimate the BMDs and BMDLs were used to extrapolate to a
 8          lower level of risk (R), in order to reach the region of each estimated dose-response
 9          function where the slope was reasonably constant and upper bound estimation was still
10          numerically stable. For these data, a 10-3 risk was generally the lowest risk necessary. The
11          oral slope factor for each site was then estimated by R/BMDL.R, as for the estimates for each
12          tumor site above.
13
14       (3) The maximum likelihood estimates (MLE) ofunit potency (thatis, risk per unit of exposure)
15          estimated by R/BMDR, were summed across the alimentary tract, liver, and
16          jejunum/duodenum in female rats.
17
18       (4) An estimate of the 95% (one-sided) upper bound on the summed oral slope factor was
19          calculated by assuming a normal distribution for the individual risk estimates, and deriving
20          the variance of the risk estimate for each tumor site from its 95% UCL according to the
21          formula:
22
23                        95% UCL = MLE  + 1.645 x SD,
24          rearranged to:
25                        s.d. = (UCL - MLE) / 1.645,
26
27    where 1.645 is the t-statistic corresponding to a one-sided 95% CI and >120 degrees of freedom,
28    and the SD is the square root of the variance of the MLE. The variances (variance = SD2) for each
29    site-specific estimate were summed across tumor sites to obtain the variance of the sum of the
30    MLEs. The 95% UCL on the sum of MLEs was calculated from the expression above for the UCL,
31    using the variance of the sum of the MLE to obtain the relevant SD (SD  = variance1/2).
32          The results of this analysis are provided in Table C-17.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                C-28        DRAFT—DO NOT CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene
1
2
Table C-12. Tumor incidence data, with time to death with tumor; male
rats exposed by gavage to benzo[a]pyrene—Kroese et al. (2001)



Dose
(mg/kg-d)
0


















3
















Week of
death
44
80
82
84
89
90
91
92
93
94
95
96
97
98
100
104
105
108
109
29
40
74
76
79
82
92
93
94
95
98
107
108
109



Total
examined
1
1
1
1
1
3
1
1
1
1
2
2
1
1
3
1
1
7
22
1
1
1
1
1
1
2
1
1
2
1
10
15
14
Numbers of animals with
Oral cavity or
forestomach
tumors
Incidental3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
4
2
1
Fatal"
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0


Liver tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
3
0
Fatal
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Duodenum
or jejunum
tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Skin or mammary gland
Basal cell
tumors
Incidental
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
Squamous
cell tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
Kidney
urothelial
carcinoma
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
             This document is a draft for review purposes only and does not constitute Agency policy.
                                          C-29       DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene



Dose
(mg/kg-d)
10



















30






























Week of
death
39
47
63
68
69
77
80
81
84
86
90
95
97
100
102
103
104
107
108
109
32
35
37
44
45
47
48
49
50
51
52
53
56
58
59
60
61
62
63
64
65
66
67
68
70
71
73
76



Total
examined
1
2
1
2
1
1
1
1
1
1
1
3
1
1
1
1
3
12
11
6
1
1
1
1
2
1
1
1
1
1
4
1
2
2
2
2
3
5
5
2
3
1
3
1
2
1
1
1
Numbers of animals with
Oral cavity or
forestomach
tumors
Incidental3
0
0
1
2
1
0
0
1
1
0
1
3
1
1
1
1
3
12
11
5
1
1
1
0
2
1
1
1
1
1
3
1
1
2
2
1
2
5
5
2
2
1
1
1
2
1
0
1
Fatal3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
1
0
0
1
1
0
0
0
1
0
2
0
0
0
1
0


Liver tumors
Incidental
0
0
0
0
0
1
1
0
0
1
0
2
0
1
1
1
3
11
11
3
0
1
0
1
2
1
1
1
1
1
3
1
1
2
2
1
1
0
4
1
1
0
2
1
1
1
1
0
Fatal
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
1
2
4
1
1
2
1
1
0
1
0
0
1
Duodenum
or jejunum
tumors
Incidental
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
1
3
1
0
0
0
1
0
1
0
0
0
Skin or mammary gland
Basal cell
tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
1
0
1
0
0
0
0
2
0
3
0
1
0
1
1
1
1
Squamous
cell tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
2
0
1
0
0
1
0
0
Kidney
urothelial
carcinoma
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
0
0
0
0
0
0
0
0
 "Incidental" denotes presence of tumors not
tumors reported by the study investigators to
known to have caused death of particular animals.
have caused death of particular animals.
'Fatal" denotes incidence of
            This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-30         DRAFT—DO NOT CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene
1
2
Table C-13. Tumor incidence data, with time to death with tumor;
female rats exposed by gavage to benzo[a]pyrene—Kroese et al. (2001)


Dose
(mg/kg-d)
0







3
















10


























Week of
death
64
69
75
104
106
107
108
109
8
47
52
60
65
76
77
83
85
86
88
93
94
97
107
108
109
42
43
44
45
48
55
59
75
76
77
80
81
82
83
85
86
87
88
89
91
95
96
98
99
102


Total
examined
1
1
1
1
4
7
7
30
1
1
1
1
1
1
1
2
1
1
1
2
1
1
6
9
21
1
1
1
1
1
1
1
1
2
2
1
1
1
1
2
1
1
2
1
1
1
1
2
3
1
Numbers of animals with
Oral cavity or forestomach
tumors
Incidental3
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
2
1
0
0
0
0
0
0
0
0
0
0
1
1
1
0
1
1
0
1
1
0
0
0
2
3
1
Fatal"
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

Liver tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
1
1
0
1
0
1
1
1
0
1
1
0
0
0
0
1
1
0
Fatal
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
0
1
1
1
0
0
1
2
1
Duodenum or
jejunum tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
             This document is a draft for review purposes only and does not constitute Agency policy.
                                          C-31        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene


Dose
(mg/kg-d)






30






















Week of
death
104
105
106
107
108
109
26
44
47
48
54
55
56
57
58
59
60
61
62
63
64
66
67
68
69
71
72


Total
examined
1
2
1
5
7
4
1
4
3
1
1
3
2
2
4
2
1
2
2
3
5
3
2
1
4
4
2
Numbers of animals with
Oral cavity or forestomach
tumors
Incidental3
1
1
1
5
7
2
0
4
3
1
0
3
2
2
3
1
0
2
2
3
5
3
1
1
3
3
1
Fatal3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
0
0
0
0
0
1
0
1
1
1

Liver tumors
Incidental
1
1
0
5
7
2
0
3
2
0
1
1
0
2
0
0
1
0
1
0
0
0
0
0
1
1
0
Fatal
0
1
1
0
0
0
0
1
1
1
0
2
2
0
4
2
0
2
1
3
5
3
2
1
3
3
2
Duodenum or
jejunum tumors
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
1
0
0
a"lncidental" 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.
            This document is a draft for review purposes only and does not constitute Agency policy.
                                                C-32          DRAFT—DO NOT CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene
1
2
Table C-14. Tumor incidence, with time to death with tumor; female
mice exposed to benzo[a]pyrene via diet—Beland and Culp (1998)
Dose group
(ppm in diet)
0












5














25
























Week of death
31
74
89
91
93
94
97
98
99
100
101
104
105
25
55
83
86
87
88
90
94
95
96
97
98
101
102
105
44
47
64
70
77
80
81
84
85
86
88
89
90
93
94
96
97
98
99
100
101
102
104
105

Total examined
1
1
2
1
2
2
2
2
1
2
2
1
29
1
1
1
1
2
2
1
1
2
1
2
2
2
2
27
1
1
1
1
1
1
1
2
1
1
1
1
4
3
2
3
1
1
2
1
1
2
1
13
Number of animals with alimentary tract squamous cell tumors
Fatal3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
0
1
1
1
1
1
0
4
2
2
0
1
1
1
1
0
2
1
0
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
2
0
0
1
0
0
0
0
10
             This document is a draft for review purposes only and does not constitute Agency policy.
                                          C-33        DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofbenzo[a]pyrene
Dose group
(ppm in diet)
100


























Week of death
39
40
42
47
49
50
53
55
56
57
58
59
60
61
62
63
64
65
66
68
69
70
71
72
73
74
79
Total examined
1
1
1
2
1
1
1
3
1
1
1
3
1
3
5
4
3
2
3
1
2
2
1
1
1
1
1
Number of animals with alimentary tract squamous cell tumors
Fatal3
1
1
1
2
0
1
0
3
1
1
1
3
1
3
5
4
3
2
3
1
2
2
1
1
1
1
1
Incidental
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
   a"lncidental" 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.
1
2
Table C-15. 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
factor3
HEDb (mg/kg-d)
Male
3
10
30
0.349
0.349
0.288
0.27
0.27
0.25
0.54
1.81
5.17
Female
3
10
30
0.222
0.222
0.222
0.24
0.24
0.24
0.49
1.62
4.85
     aScaling 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)0'25 = scaling factor.
     bHED = administered dose x scaling factor.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 C-34         DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene
1
2
Table C-16.  Derivation of HEDs for dose-response modeling of B6C3Fi
female mouse tumor incidence data from Beland and Culp (1998)
Benzo[a]pyrene
dose in diet
(ppm)
5
25
100
Intake (ng/d)
21
104
430
TWA body weight
average (kg)
0.032
0.032
0.027
Administered
dosea (mg/kg-d)
0.7
3.3
16.5
Scaling factorb
0.15
0.15
0.14
HEDc (mg/kg-d)
0.10
0.48
2.32
     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.
     bScaling 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)
     CHED = administered dose x scaling factor.
                                                       1 = scaling factor.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-35         DRAFT—DO NOT CITE OR QUOTE

-------
                                     Toxicological Review ofbenzo[a]pyrene
1
2
Table C-17.  Summary of model selection and modeling results for best-
fitting multistage-Weibull models, using time-to-tumor data for rats
from Kroese etal. (1981)


Male rats

















Female rats









Endpoints
Oral cavity and
forestomach:
squamouscell tumors
Hepatocellular tumors


Duodenum and
jejunum tumors

Kidney: uroethelial
carcinoma

Skin and mammary
gland: basal cell
tumors
Skin and mammary
gland: squamouscell
tumors
Oral cavity and
forestomach:
squamous cell tumors
Hepatocellular tumors


Duodenum and
jejunum tumors

Model
stages
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3

AIC
577.8
407.6
229.0
367.3
301.5
289.1
69.6
65.9
66.9
31.9
31.7
32.8
110.6
105.1
104.7
63.5
64.3
65.3
277.1
211.6
201.0
595.5
774.9
468.3
37.9
37.0
37.8

BMD10
0.104
0.678
0.453
0.181
0.472
0.651
2.64
3.04
3.03
9.16
5.71
4.65
1.88
2.58
2.86
3.36
2.75
2.64
0.245
0.428
0.539
0.146
0.370
0.575
6.00
4.33
3.43

BMDL10 - BMDU10


0.281-0.612


0.449 -0.772


2.38 - 3.87


2.50-9.01


2.35-3.62
1.77 - 4.42




0.328-0.717


0.507 - 0.630


1.95 - 5.70

Model selection rationale


Lowest AIC, best fit to low dose data


Lowest AIC, best fit to low dose data


Best fit to data


Best fit to data


Lowest AIC, best fit to low dose data
Lowest AIC, best fit to low dose data




Lowest AIC, best fit to low dose data


Lowest AIC, best fit to low dose data


Best fit to low dose data
   This document is a draft for review purposes only and does not constitute Agency policy.
                                C-36       DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene
 1    Male Rat (Kroese et al, 2001): Squamous Cell Papilloma or Carcinoma in Oral Cavity or Forestomach

 2
 O

 4             Multistage Weibull  Model.  (Version:  1.6.1;   Date: 11/24/2009)
 5             Solutions are obtained using donlp2-intv,  (c) by P. Spellucci
 6             Input Data File:  OralForstKroeseM3.(d)
 ^

 9
10
11
12
13
14       Dependent  variable = CONTEXT
15       Independent variables = DOSE,  TIME
16
17     Total number of observations = 208
18     Total number of records with missing  values = 0
19     Total number of parameters in model = 6
20     Total number of specified parameters  =  0
21     Degree of  polynomial = 3
22
23
24
25
26
27
28                      Default Initial  Parameter Values
2.y                            c      =           3.6
30                            t_0    =       39.1111
31                            beta_0 =            0
32                            beta_l =   8.8911e-009
33                            beta_2 =  1.60475e-031
34                            beta 3 =  1.95818e-008
35
36
37               Asymptotic Correlation  Matrix of Parameter Estimates
38               (  *** The model parameter(s)  -beta_0    -beta_2
39                    have been estimated at  a boundary point,  or  have been specified by the  user,
40                    and do not appear in  the correlation matrix  )
41
42
43
44
45
46        t  o
47
48        beta 1
49
50        beta 3
51
52
53                                    Parameter  Estimates
54                                                           95.0% Wald Confidence Interval
55           Variable         Estimate        Std. Err.     Lower  Conf. Limit   Upper Conf.  Limit
56             c               3.74559        0.447309            2.86888              4.6223
57             tO             41.4581         2.14975            37.2447             45.6716
58
59
60
61
62
63    NA - Indicates that this parameter has hit  a bound implied by some  inequality constraint
64         and thus has  no standard error.
65
66
67                   Log (likelihood)    #  Param
68       Fitted Model        -108.512         6
69
70
71
72


                 This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-37         DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
18
                                                         Toxicological Review ofbenzo[a]pyrene
                                         U  Total  Expected Response
              0
          Minimum observation time for F tumor  context =
         Benchmark  Dose Computation

      Risk Response   =     Incidental

      Risk Type        =          Extra

      Confidence  level =            0.9
      Time            =            104
                Specified

                 effect  =

                    BMD  =

                   BMDL  =


                   BMDU  =
                                                               44
                                       Incidental Risk: OralForstKroeseM3
                            points show nonparam. est. for Incidental (unfilled) and Fatal (filled)

                           Dose = 0.00                                      Dose = 0.54
19
20
OD
CD
OD
CD
CD ~~
CN
CD
CD









I I I I I I
0 20 40 60 80 100
S 	 1
OD
CD

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1    Male Rat (Kroese et al, 2001):  Hepatocellular Adenoma or Carcinoma

 2
 3    =======================================================================
 4             Multistage  Weibull Model.  (Version:  1.6.1;  Date: 11/24/2009)
 5             Solutions are obtained using donlp2-intv,  (c) by P. Spellucci
 6             Input Data  File: LiverKroeseMS.(d)
 7
 8
 9
10
11
12
13
14
15       Dependent variable = CONTEXT
16       Independent variables = DOSE,  TIME
17
18     Total number of observations  = 208
19     Total number of records with  missing values = 0
20     Total number of parameters  in model =  6
21     Total number of specified parameters = 0
22     Degree of  polynomial = 3
23
24
25
26
27
28
29
30                      Default Initial  Parameter Values
31                            c      =          3.6
32                            t_0    =       34.6667
33                            beta_0 =            0
34                            beta_l =  2.73535e-009
35                            beta_2 =    8.116e-028
36                            beta  3 =  1.43532e-008
37
38
39               Asymptotic Correlation  Matrix of Parameter Estimates
40               ( *** The model parameter(s)  -beta_0    -beta_2
41                    have been estimated at a  boundary point,  or  have been specified by the  user,
42                    and do not  appear in the  correlation matrix  )
43
44
45
46
47
48        t  o
49
50        beta 1
51
52        beta 3
53
54
55                                    Parameter Estimates
56
57           Variable
58
59
60
61
62
63
64
65    NA - Indicates that this parameter has  hit a bound implied by some inequality constraint
66         and thus has  no standard  error.
67
68
69                   Log (likelihood)    #  Param
70       Fitted Model        -138.544        6
71
72
73
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   C-39         DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
18
                                                       Toxicological Review ofbenzo[a]pyrene
                                          Total  Expected Response
         Benchmark Dose Computation

      Risk Response   =     Incidental

      Risk Type       =          Extra
      Confidence level =            0.9

      Time            =            104
               Specified

                effect =

                   BMD =

                  BMDL =


                  BMDU =
19
20
                            Incidental Risk: Hepatocellular_Kroese_M3
                   points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
                       Dose = 0.00
                                                                Dose = 0.54
Probability
oo
o
0 ~
o
o
\
1 1 1 1 1 1
0 20 40 60 80 100
Probability
oo
o
0 ~
o

«-> i i " T i r "i
0 20 40 60 80 100
                          Time
                                                                   Time
                       Dose= 1.81
                                                                Dose= 5.17
            oo
                \     \    I     I    I    T

                0   20  40   60  80  100
                                                2
                                                CL
                                                    p
                                                    O
                                                                           \    \


                                                         0  20  40  60   80  100
                          Time
                                                                   Time
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                 C-40         DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1    Male Rat (Kroese et al, 2001):  Duodenum or Jejunum Adenocarcinoma

 2
 o

 4             Multistage  Weibull Model. (Version:  1.6.1;  Date: 11/24/2009)
 5             Solutions are  obtained using donlp2-intv,  (c) by P.  Spellucci
 6             Input Data  File: DuoJejKroeseM3.(d)
 ^

 9
10
11
12
13
14
15       Dependent variable = CONTEXT
16       Independent variables = DOSE,  TIME

18     Total number of observations  = 208
19     Total number of records with  missing values = 0
20     Total number of parameters  in model =  6
21     Total number of specified parameters = 1
22     Degree of  polynomial = 3

24
25
26       User specifies  the following parameters:
27             t 0               0
28
29
30
31
32
33
34                      Default Initial  Parameter Values
35                            c             1.63636
36
37
38
39
40
41
42
43               Asymptotic Correlation  Matrix of Parameter Estimates
44               ( *** The model parameter(s)  -t_0       -beta_0    -beta_l    -beta_2
45                    have been estimated at a boundary point,  or  have been specified by the user,
46                    and do not  appear in the correlation matrix  )
47
48                    c            beta 3
49
50        c                 1-1
51
52        beta 3-11
53
54
55                                    Parameter Estimates
56                                                           95.0% Wald Confidence Interval
57           Variable         Estimate       Std. Err.     Lower Conf. Limit   Upper Conf.  Limit
58
59
60
61
62
63
64    NA - Indicates that this parameter has  hit a bound implied by some inequality constraint
65         and thus has  no standard  error.
66
67
68                   Log (likelihood)    # Param
69       Fitted Model        -28.4387        5
70
71
72                       Data Summary
73                           CONTEXT
74                  C      F      I     U   Total  Expected Response

                 This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-41         DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
                  51

                  43
       Specified effect =

                  BMD =

                 BMDL =

                 BMDU =
                                                     Toxicological Review ofbenzo[a]pyrene
                             Incidental Risk: DuoJej_Kroese_M3
17

18

19
                     Dose= 0.00
                                                           Dose= 0.54
t
la
ro
-Q
o
Dl




if)

CD

	
0
0







t
la
ro
-Q
o
Dl


CD 1 1 1 1 ~~l 1 '
0 20 40 60 80 100
—
T — 	
CD

	
0
0
0


I
0

I I
20 40

I I I
60 80 100
                        Time
                                                              Time
                     Dose= 1.81
                                                           Dose= 5.17
                   \    \    \    \    \

                  20  40  60  80  100
                                                    \

                                                    0
 \    \    \    \    \

20  40  60  80  100
                        Time
                                                              Time
               This document is a draft for review purposes only and does not constitute Agency policy.

                                               C-42        DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene
 2    Male Rat (Kroese et al. 2001): Skin or Mammary Gland Basal Cell Tumors

 O

 4    =======================================================================
 5             Multistage  Weibull Model.  (Version:  1.6.1;  Date: 11/24/2009)
 6             Solutions are  obtained using donlp2-intv,  (c)  by P.  Spellucci
 7             Input Data  File: SKinMamBasalKroeseM3.(d)
 ^

10
11
12
13
14
15
16       Dependent variable = CONTEXT
17       Independent variables = DOSE,  TIME
18
19     Total  number of observations =  208
20     Total  number of records with missing values = 0
21     Total  number of parameters  in model =  6
22     Total  number of specified parameters = 1
23     Degree of  polynomial = 3
24
25
26
27       User specifies  the following  parameters:
28             t o    =          o
29
30
31
32
33
34
35                      Default Initial Parameter Values
36                            c      =       1.38462
37                            t_0    =            0   Specified
38                            beta_0  = 3.84298e-005
39                            beta_l  = 1.06194e-028
40                            beta_2  =            0
41                            beta 3  = 6.84718e-006
42
43
44               Asymptotic Correlation Matrix of Parameter Estimates
45               ( *** The model parameter(s)  -t_0       -beta_l     -beta_2
46                    have been estimated at a boundary point,  or have been specified  by the user,
47                    and do not  appear in the correlation matrix )
48
49                    c           beta 0       beta 3
50                                    -
51        c                 1-1           -i
52
53
54
55
56
57
58
59
60           Variable
61
62
63
64
65
66
67
68
69
70
71                   Log (likelihood)   # Param            AIC
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   C-43          DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
       Specified effect =

                  BMD =

                 BMDL =

                 BMDU =
                                                     Toxicological Review ofbenzo[a]pyrene
                                        Total  Expected Response
                         Incidental Risk: Skin  Mam  Basal Kroese M3
                      Dose= 0.54
                                                            Dose= 1.81
J^
—
!5
CD
.a

a.


oo
0
	
•
^^
0
—
0







- - mmm ^m t
^
±^
!5
CD
r2
2
CL

O 1 ill 1 1
oo
CD
	
•
^^
0
—
0








o ~S 	 1 	 1 - r -i - 1 — •
               0   20  40  60  80
                                                      0  20  40  60  80
                         Time
                                                               Time
23
                      Dose= 5.17
      .Q
       CD
      .Q
       O
           oq
           C3
           q
           c>
                         Time
               0   20  40  60  80
               This document is a draft for review purposes only and does not constitute Agency policy.

                                               C-44        DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1    Male Rat (Kroese et al, 2001): Skin or Mammary Gland Squamous Cell Tumors

 2
 o

 4             Multistage Weibull Model.  (Version: 1.6.1;  Date:  11/24/2009)
 5             Solutions are obtained using  donlp2-intv, (c)  by P.  Spellucci
 6             Input Data File: SKinMamSCCKroeseMS.(d)
 ^

 9
10
11
12
13
14
15       Dependent variable  = CONTEXT
16       Independent variables  = DOSE, TIME

18     Total  number of observations =  208
19     Total  number of records  with missing values =  0
20     Total  number of parameters  in model = 6
21     Total  number of specified parameters = 1
22     Degree of polynomial  = 3

24
25
26       User specifies the  following  parameters:
27             t 0               0
28
29
30
31
32
33
34                     Default Initial Parameter Values
35                            0=3
36
37
38
39
40
41
42
43              Asymptotic  Correlation Matrix of Parameter Estimates
44               ( *** The  model parameter(s)  -t_0      -beta_0    -beta_2
45                    have  been estimated at a boundary point, or have been specified by the user,
46                    and  do not  appear in the correlation matrix )
47
48
49
50
51
52
53
54
55
56                                   Parameter Estimates
57                                                           95.0% Wald Confidence Interval
58           Variable         Estimate       Std. Err.     Lower Conf. Limit  Upper Conf.  Limit
59
60
61
62
63
64
65
66
67
68
69                   Log (likelihood)   # Param
70       Fitted Model         -27.652         5
71
72
73                       Data  Summary
74                           CONTEXT

                This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-45         DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene
 1
 2
 O
 4
 5
 6
 8
 9
10
11
12
13
14
15
                                         U  Total  Expected Response
         Benchmark Dose  Computation
      Risk Response   =     Incidental
      Risk Type        =         Extra
      Confidence  level =           0.9
      Time            =           104
16
                Specified
                 effect =
                    BMD =
                   BMDL =
                                      Incidental Risk: OralForstKroeseM3
                           points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
                          Dose= 0.00                                    Dose= 0.54
17
           OD
           CD

           UD
           CD
           rsj
           CD
           O
           CD
                    ^     I      I     I      I

                     20    40    60   80   100
                                                         OD
                                                         CD

                                                         GD
                                                         CD
                                                         OJ
                                                         CD
                                                         CD
                                                         CD
                                                                    I     I      I     I     I

                                                                   20    40    60    80   100
                             Time
                                                                           Time
                          Dose= 1.81
                                                                        Dose= 5.17
           OD
           CD
       o
       CL
           CD
           CD
                                 O 'If  O-I-
                      I     I      I     I      I

                     20    40    60   80   100



:=-!
,±±
LE
CTJ
0
CL


• — i
OD
CD

tfl
CD

CD ~~
CM
CD _
i — i
0» 0«^^«»4
/
/
/ +*
i
f f
J

1 J
._--/ .»^rf^
                                                                   20    40    60    80   100
                             Time
                                                                           Time
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    C-46         DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1    Male Rat (Kroese et al, 2001): Kidney Urothelial Carcinomas

 2
 o

 4             Multistage Weibull  Model.  (Version:  1.6.1;  Date: 11/24/2009)
 5             Solutions are obtained using donlp2-intv,  (c) by P. Spellucci
 6             Input Data File:  KidneyUrothelialCarKroeseM3.(d)
 ^

 9
10
11
12
13
14
15       Dependent  variable = CONTEXT
16       Independent variables = DOSE,  TIME

18     Total number of observations = 208
19     Total number of records with missing values = 0
20     Total number of parameters in model = 6
21     Total number of specified parameters = 1
22     Degree of  polynomial = 3

24
25
26       User specifies  the following parameters:
27             t 0              0
28
29
30
31
32
33
34                      Default Initial  Parameter Values
35                            c             1.63636
36
37
38
39
40
41
42
43               Asymptotic Correlation  Matrix of Parameter Estimates
44               (  *** The model parameter(s)  -t_0       -beta_0     -beta_l    -beta_2
45                    have been estimated at a boundary point, or  have  been specified by the user,
46                    and do not appear in the correlation matrix  )
47
48                    c            beta 3
49
50        c                 1-1
51
52        beta 3-11
53
54
55                                    Parameter Estimates
56                                                           95.0% Wald  Confidence Interval
57           Variable         Estimate       Std. Err.     Lower  Conf.  Limit   Upper Conf.  Limit
58             c               1.74897         3.79403            -5.68719             9.18512
59             beta_0                0              NA
60             beta_l                0              NA
61             beta 2                0              NA
62
63
64
65
66
67
68
69                   Log (likelihood)    #  Param
70       Fitted Model        -11.3978        5
71
72
73                       Data Summary
74                           CONTEXT

                 This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-47         DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
                                                         Toxicological Review ofbenzo[a]pyrene
                                        U  Total  Expected Response
         Benchmark Dose Computation

      Risk Response    =     Incidental
      Risk Type        =          Extra

      Confidence level =            0.9

      Time            =            104
                Specified

                 effect =
                    BMD =

                   BMDL =
                                   Incidental Risk: Kidney_Kroese_M3
                        Dose = 0.00
                                                                 Dose= 0.54



^
o
(TJ
O
2
n





in
o
o
o



in
o _
o
o
o












^
.a
(TJ
_Q
2
CL



"-1 1 1 1 1 1 1
0 20 40 60 80 100
in
o
o
o



in
o _
o
o
o










"-1 1 1 1 1 1 1
0 20 40 60 80 100
17

18
                          Time
                                                                   Time
                        Dose= 1.81
                                                                 Dose= 5.17
          in

          o
      =  o
      .a
          §
          o


          8
          o
                                 TTi
                   I

                   20
                        I

                        40
 I

60
 I

80
                                                   in

                                                   o
                                               =  o
                                               .a
                                                2.
                                                CL
                       s J
                       o


                       8 J
                       o
                                      100
                                                            20
 I

40
 I

60
                                                                          80   100
                          Time
                                                                   Time
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-48          DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1    Female Rat (Kroese et al, 2001):  Oral Cavity or Forestomach, Squamous Cell Papilloma or Carcinoma

 2
 3    =======================================================================
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15       Dependent variable = CONTEXT
16       Independent variables = DOSE,  TIME
17
18     Total  number of observations  =  208
19     Total  number of records with  missing values = 0
20     Total  number of parameters in model = 6
21     Total  number of specified parameters = 0
22     Degree of polynomial = 3
23
24
25
26
27
28
29
30                     Default Initial  Parameter Values
31                            c      =          3.6
32                            t_0    =      45.1111
33                            beta_0  =  1.11645e-009
34                            beta_l  =  4.85388e-009
35                            beta_2  =            0
36                            beta  3  =  1.95655e-008
37
38              Asymptotic Correlation  Matrix of Parameter  Estimates
39               ( *** The model parameter(s)  -beta_2
40                    have been estimated at a boundary  point,  or have been  specified by the user,
41                    and do not appear in the correlation  matrix )
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57                                                           95.0% Wald Confidence  Interval
58           Variable         Estimate       Std. Err.      Lower Conf. Limit   Upper Conf. Limit
59             c               3.52871        0.701117             2.15454
60             t_0              46.553         5.93306             34.9244
61             beta_0     1.53589e-009    5.40523e-009        -9.05817e-009
62             beta~l     7.57004e-009     2.9647e-008        -5.05369e-008
63
64
65
66
67
68
69
70                   Log (likelihood)    # Param
71       Fitted Model        -94.5119         6
72
73


                This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-49         DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
                                      U  Total  Expected Response
         Minimum observation time for F tumor  context =


        Benchmark Dose Computation

     Risk Response    =    Incidental

     Risk Type        =         Extra
     Confidence level =           0.9

     Time            =           104
       Specified effect =

                  BMD =

                 BMDL =

                 BMDU =
                                 Incidental Risk: OralForstKroeseFS

                   points show nonparam. est. for Incidental (unfilled) and Fatal (filled)

                        Dose = 0.00                                Dose = 0.49
18

±i
'o
CD
O
p
Q.

00
0
_
1
^ 	
o
0
0 ~"






\
1 1 1 1 1 1

-1—*
&
CD

O
Q.

00
0
_
1
^
0
0
0 ~"






O IB
1 1 1 1 1 1
                0   20   40  60  80  100
                                                           0   20   40   60  80  100
                           Time
                                                                      Time
                        Dose = 1 .62
                                                                   Dose = 4.58

-i— •
o
CD
.a
o
00
o
_
«.
/•
fr
:~-\ ^C'
>,

S3.
CD
.a
e
Q.
00
o
_
o ~
o
or'?
,
•
!
e *
                 I    I     I    I     I     I

                0   20   40  60  80  100
                                                            1    1     1    1     1     1

                                                            0   20   40   60  80  100
                           Time
                                                                      Time
               This document is a draft for review purposes only and does not constitute Agency policy.

                                                C-50         DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1    Female Rat [Kroese etal., 2001]: Hepatocellular Adenoma or Carcinoma

 2
 O

 4             Multistage Weibull Model.  (Version:  1.6.1;  Date:  11/24/2009)
 5             Solutions  are obtained using donlp2-intv,  (c)  by P.  Spellucci
 6             Input Data File: LiverKroeseF3.(d)
 7             Fri Apr 16 09:08:03 2010
 8
 9
10     Timer to Tumor Model, Liver Hepatocellular Tumors,  Kroese et al, Female

12
13
14
15
16
17
18
19
20       Dependent variable  = CONTEXT
21       Independent variables  = DOSE, TIME
22
23     Total number of observations =  208
24     Total number of records  with missing values = 0
25     Total number of parameters  in model = 6
26     Total number of specified parameters = 0
27     Degree of polynomial  = 3
28
29
30
31
32
33
34
35                     Default Initial Parameter Values
36                            c                3.6
37                            t_0    =      31.7778
38                            beta_0  =            0
39                            beta_l  =  4.9104e-031
40                            beta_2  = 5.45766e-030
41                            beta 3  = 3.44704e-008
42
43
44              Asymptotic  Correlation Matrix of Parameter Estimates
45               ( *** The model parameter(s)  -beta_0     -beta_l    -beta_2
46                    have  been estimated at a boundary point, or have been specified by the user,
47                    and do not  appear in the correlation matrix )
48
49                    c           t  o
50
51
52
53        t o
54
55        beta 3
56
57
58                                   Parameter Estimates
59
60           Variable
61
62
63
64
65
66
67
68
69
70
71
72                   Log (likelihood)   # Param
73       Fitted Model         -228.17         6


                This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-51        DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
18
19
20
                                                     Toxicological Review ofbenzo[a]pyrene
                                        Total  Expected Response
                                                          44
       Specified effect =

                  BMD =

                 BMDL =

                 BMDU =
21
                             Incidental Risk: Hepatocellular_Kroese_F3
                   points show nonparam. est. for Incidental (unfilled) and Fatal (filled)
                        Dose = 0.00
                                                                  Dose = 0.49

-1— '
CD
2
CL



oo
o
,-f.
0 ~
_
0







-i— '
CD
2
Q.

^ 1 1 1 1 1 1
0 20 40 60 80 100
oo
O
^.
0

0
o


1 1 1 1
0 20 40 60
=«CMt4» — •
1 1
80 100
                           Time
                                                                     Time
                        Dose = 1 .62
                                                                  Dose = 4.58
                 \    \     \    \     \    \

                0   20   40  60   80  100
                                                 >s   °°.
                                                 =    °
                                                 !Q

                                                 s    -*
                                                 2    °
                                                 Q_
                                                      C3
                                                      O
                                                          n    i     i    i     i    i

                                                          0   20   40  60   80  100
                           Time
                                                                     Time
22
               This document is a draft for review purposes only and does not constitute Agency policy.

                                               C-52         DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene



 1    Female Rat [Kroese et al.. 2001):  Duodenum or lejunum Adenocarcinoma

 2

 3    =======================================================================
 4             Multistage Weibull Model.  (Version:  1.6.1;  Date:  11/24/2009)
 5             Solutions are obtained using donlp2-intv,  (c)  by P.  Spellucci
 6             Input Data File:  DuoJejKroeseFS.(d)
 7
 8
 9
10
11
12
13
14       Dependent variable  = CONTEXT
15       Independent variables =  DOSE,  TIME
16
17     Total number of observations = 208
18     Total number of records with missing  values = 0
19     Total number of parameters in model = 6
20     Total number of specified  parameters  =  1
21     Degree  of polynomial  = 3
22
23
24
25
26
27
28
29
30
31
32                     Default Initial  Parameter Values
33
34
35
36
37
38
39
40
41              Asymptotic  Correlation  Matrix of Parameter Estimates
42               ( *** The model  parameter(s)  -t_0       -beta_0    -beta_l   -beta_2
43                    have  been  estimated at  a boundary point, or have been specified by the  user,
44                    and do not appear in  the correlation matrix )
45
46                    c           beta 3
47
48        c                 1-1
49
50        beta  3-11
51
52
53
54
55          Variable
56
57
58
59
60
61
62
63
64
65
66                   Log (likelihood)   #  Param
67       Fitted Model       -13.8784         5
68
69
70
71
72
73


                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-53        DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
                   48
        Benchmark Dose  Computation

      Risk Response   =    Incidental

      Risk Type       =         Extra

      Confidence level =           0.9
      Time            =           104
       Specified effect =

                   BMD =

                  BMDL =

                  BMDU =
13
                                      Incidental Risk: DuoJej_Kroese_F3
                          Dose = 0.00
                                                                        Dose = 0.49
      s
       o
      CL
           LO

           d
           LT>
           O
           d
           O
           O
                     I


                    20
40
60
                                                    S
                                                    o
                                                    CL
                                     80    100
d
o
d
in
o _
d
o
o
d









I I I I I I
0 20 40 60 80 100
                             Time
                                                                          Time
                          Dose= 1.62
                                                                        Dose = 4.58
      s
       o
      D.
           LT>
           o _
           o
           o _
                     \

                    20
40
60
                                                    JD
                                                    s
                                                    o
                                                    CL
                         o

                         o

                         d


                         p _
                         d
                                                        o
                                                        o _
                                     80    100
 \

20
40
 I

60
 \

80
100
                             Time
                                                                          Time
14

15
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-54         DRAFT—DO NOT CITE OR QUOTE

-------
                                                  Toxicological Review ofbenzo[a]pyrene
1
2
      Table C-18. Summary of human equivalent overall oral slope factors,
      based on male and female rat tumor incidence
Data set
Males
Females
Tumor site
Oral cavity/forestomach
Liver
Duodenum/jejunum
Skin/mammary gland:
basal cell
Skin/mammary gland:
squam. cell
Kidney
BMDooi
6.37 x 10"3
2.00 x 10"2
6.42 x 10"1
6.06 x 10"1
7.06 x 10"2
9.84 x 10"1
BMDLooi
2.86 x 10"3
5.30 x 10"3
4.21 x 10"2
4.24 x 10"2
2.11 x 10"2
7.48 x 10"2
Sum, risk values at BMD001:
Risk value3 at
BMDooi
1.57 x 10"1
5.00 x 10"2
1.56 x 10"3
1.65 x 10"3
1.42 x 10"2
1.02 x 10"3
2.25 x 10"1
BMDLooi
3.50 x 10"1
1.89 x 10"1
2.38 x 10"2
2.36 x 10"2
4.75 x 10"2
1.34 x 10"2
SD
1.17 x 10"1
8.42 x 10"2
1.35 x 10"2
1.33 x 10"2
2.03 x 10"2
7.51 x 10"3
Sum, SD2:
Overall SDb:
Upper bound on sum of risk estimates0:
Oral cavity/forestomach
Liver
Duodenum/jejunum
3.45 x 10"3
1.53 x 10"2
5.85 x 10"2
1.01 x 10"2
1.22 x 10"1
7.27 x 10"1
Sum, risk values at BMD001:
2.90 x 10"1
6.54 x 10"2
1.71 x 10"2
1.09 x 10"1
SD2
1.38 x 10"2
7.09 x 10"3
1.82 x 10"4
1.78 x 10"4
4.10 x 10"4
5.64 x 10"5
2.17 x 10"2
1.47 x 10"1
Properties
of total
variance
0.64
0.33
0.01
0.01
0.02
0.00


4.68 x 10"1
9.92 x 10"2
8.21 x 10"3
1.38 x 10"3
1.16 x 10"1
3.48 x 10"2
9.56 x 10"3
Sum, SD2:
Overall SD:
Upper bound on sum of risk estimates0:
1.35 x 10"2
1.21 x 10"3
9.13 x 10"5
1.48 x 10"2
1.22 x 10"1
0.91
0.08
0.01


3.09E-01
3
4
"Risk value = 0.001/BMDL001.
bOverall SD = (sum, SD2)0 5.
"Upper bound on the overall risk estimate = sum of BMD0oi risk values + 1.645 x overall SD.

Source of data: Kroese et al. (2001).

      Table C-19.  Summary of model selection among multistage-Weibull
      models fit to alimentary tract tumor data for female mice
5
6
7
Model
stages
1
2
3
AIC
688.5
629.2
624.5
BMD10
0.104
0.102
0.127
BMDL10-BMDU10
0.071 -0.179
Model selection rationale
Lowest AIC, best fit to low dose data
      Source of data: Beland and Gulp (1998)
              This document is a draft for review purposes only and does not constitute Agency policy.
                                            C-55         DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
 1    Female Mice fBeland and Culv, 19981: Alimentary Tract Sauamous Cell Tumors
 4
 5
 6

 8
 9
10
11
12
13
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
        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:\mswlO-09\benzo[a]pyrene_FemaleSquamF3i.(d)
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
               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           tO         beta 0      beta 1
      Variable

        c

        t_0
        beta_0

        beta_l

        beta_2

        beta 3
Upper Conf. Limit

        9.54705

         23.677

   3.14019e-015

   2.55825e-014

   7.93813e-013

   1.19919e-012
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-56         DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
18
19
20
          Minimum observation time for F tumor context =
      dp

      Confidence level  =
      Time
                   BMD =

                  BMDL =

                  BMDU =
21

22
                                   Incidental Risk: BaP_FemaleSquamF3i
                          points show nonpararm. est for Incidental (unfilled) and Fatal (filled)
                         Dose= 0.00
                                                                   Dose= 0.10


-Jr
•^
cc
o
CL





i — i
OD
CD
UD
CD

^t
CD
fN
CD
CD









,,r--ii
i — i
OD
CD
;t± UD
15 CD
C13
O ^ —
d °
r-i
CD
CD








o
	 « 	 * 	 !•>••-»«
° 1 1 1 1 1 1 ° 1 1 1 1 1 1
0 20 40 60 80 100 0 20 40 60 80 100
                            Time
                                                                      Time
                         Dose= 0.48
                                                                   Dose= 2.32
           OD

           CD
           r-i
           CD

           p
           CD
                     I     I     I      I     I

                    20   40    60    80   100


                            Time
                                                      OD

                                                      CD
                                                  I   3
                                                      p
                                                      CD
 I

20
40
                                                                          I     I      I

                                                                         60   80   100
                                                                      Time
                 This document is a draft for review purposes only and does not constitute Agency policy.

                                                    C-57          DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
 1   DOSE-RESPONSE MODELING FOR THE INHALATION UNIT RISK
 2          As with the tumor data used for the oral slope factor (see Dose Response-modeling for the
 3   Oral Slope Factor Section), there was earlier occurrence of tumors with increasing exposure, and
 4   early termination of the high-dose group (Thyssen et al., 1981; see Appendix B for study details).
 5   The computer software program MSW (U.S. EPA, 2010) was used as described in the analysis of the
 6   oral carcinogenicity data.
 7          Thyssen et al. (1981) did not determine cause of death for any of the animals. Bounding
 8   estimates for the Thyssen et al. (1981) data were developed by treating the tumors alternately as
 9   either all incidental or all fatal. In either case, therefore, an estimate of to (the time between a
10   tumor first becoming observable and causing death) could not be estimated. The data analyzed are
11   summarized in Table C-20, the results are summarized in Table C-21, and the modeling details
12   follow.

13          Table C-20.  Individual pathology and tumor occurrence data for male
14          Syrian hamsters exposed to benzo [ajpyrene via inhalation for lifetime-
15          Thyssen etal. (1981).
Nominal
exposure
concentration
(mg/m3)
0
























2








Time on
study
17
39
45
79
83
85
86
88
89
90
101
102
103
106
108
109
112
115
116
122
123
124
125
127
132
14
35
53
59
71
78
80


Number
examined
1
1
1
1
1
1
1
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Papillomas, Polyps, Papillary polyps, Squamous cell carcinomas


Larynx
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
oa
0
0
0
oa
0
0
0
0
0
0

Pharynx
oa
0
0
0
0
oa
0
0
0
0
0
0
0
0
0
0
0
0
oa
0
0
0
0
oa
0
oa
0
0
0
0
0
0

Trachea
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

Esophagus
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

Forestomach
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

Nasal cavity
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-58        DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
Nominal
exposure
concentration
(mg/m3)


















10





















50

















Time on
study
85
87
88
93
98
99
102
103
108
111
113
114
115
116
117
120
122
133
31
32
52
67
73
76
80
85
94
100
102
105
111
113
114
115
116
117
118
122
124
125
20
21
25
29
30
34
36
37
40
41
43
47
48
51
56
57


Number
examined
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
4
1
1
1
1
1
2
1
1
1
2
1
2
1
1
1
1
1
1
1
Papillomas, Polyps, Papillary polyps, Squamous cell carcinomas


Larynx
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
oa
0
0
0
0
0
0
0
1
0
1
0
0
1
0
0
1
1
0
1
3
1
1
0
oa
oa
oa
oa
oa
oa
oa
oa
la
0
0
1
0
0
1
0

Pharynx
0
0
0
0
oa
0
0
0
0
0
0
0
0
0
0
0
oa
0
0
0
0
0
0
2
0
0
0
0
1
1
1
1
1
oa
0
0
lb
0
1
0
oa
oa
oa
oa
oa
oa
oa
oa
la
0
0
1
1
oa
1
1

Trachea
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
oa
oa
oa
oa
oa
oa
oa
oa
la
0
0
0
0
0
0
0

Esophagus
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

Forestomach
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

Nasal cavity
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
oc
0
0
1
1
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
This document is a draft for review purposes only and does not constitute Agency policy.
                               C-59        DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene
Nominal
exposure
concentration
(mg/m3)












Time on
study
60
63
64
66
68
70
71
72
73
79


Number
examined
1
1
1
1
1
1
1
1
2
4
Papillomas, Polyps, Papillary polyps, Squamous cell carcinomas


Larynx
0
0
0
1
0
1
1
1
2
3

Pharynx
1
0
1
1
1
1
1
1
2
4

Trachea
0
0
0
0
0
0
1
0
0
1

Esophagus
0
0
0
0
0
1
0
0
0
1

Forestomach
0
0
1
0
0
0
0
0
0
0

Nasal cavity
0
0
0
0
0
0
0
0
0
1
      aTissue was not examined for one animal of total examined.

      bTissue was not examined for two animals of total examined.

      cAn adenocarcinoma was observed in this tissue, but not included in the dose-response analysis because it was of a different

      cell type than the other tumors listed. It was judged to be an isolated finding not clearly associated with exposure.
 1

 2
      Table C-21.  Summary of model selection among multistage-Weibull
      models fit to tumor data for male hamsters
Tumor context
All tumors considered incidental to
cause of death
All tumors considered to be cause
of death
Model
stages
1
2
1
2
3
AIC
58.0
47.9
327.3
302.9
299.0
BMD10
0.090
0.285
0.136
0.421
0.648
BMDL10
0.064
0.198
0.104
0.343
0.461
Model selection rationale
Lowest AIC, best fit to data
(BMDU10 = 0.350)
Lowest AIC; best fit to data
(BMDU10 = 0.719)
 3           Data source: Thyssen etal. (1981)



 4    Output for squamous cell neoplasia following inhalation exposure to BaP: all tumors considered

 5    incidental to cause of death
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
        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*doseAl+beta 2*doseA2)}
  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
                 This document is a draft for review purposes only and does not constitute Agency policy.

                                                   C-60         DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
18

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
                        c                3.6
                        t_o    =            o
                        beta_0 = 1.18657e-031
                        beta_l =    1.49e-030
                        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 )
 1

-1
                                -1

                                 1
       Variable
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         -19.967         4
                   Data Summary
                       Class
                     F     I
                                     Total  Expected Response
   Benchmark Dose Computation
Risk Response    =     Incidental
Risk Type        =          Extra
Specified  effect =            0.1
Confidence level =            0.9
Time
            BMD
           BMDL
           BMDU
          This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-61         DRAFT—DO NOT CITE OR QUOTE

-------
                                         Toxicological Review ofbenzo[a]pyrene
                 Incidental Risk: BaP-Thyssen_inc2st
           Dose = 0.00
                                   Dose = 0.25
Probability
oo
o
o ~"
o

«-» i i r i T~ r™r*
0 20 60 100
Probability
oo
o
o ~"
o

o | | '| H "j~"|"— y '
0 20 60 100
              Time
                                      Time
           Dose = 1.00
                                   Dose = 4.29
p
o
    n   i   r
    o  20
 \   \
60

Time
 i   r
100
100
                                                   Time
    This document is a draft for review purposes only and does not constitute Agency policy.
                                    C-62        DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofbenzo[a]pyrene



 1    Output for respiratory tract tumors: all tumors considered to be cause of death

 2
 3    =======================================================================
 4             Multistage Weibull Model.  (Version:  1.6.1;  Date:  11/24/2009)
 5             Solutions are  obtained using donlp2-intv,  (c)  by  P.  Spellucci
 6             Input Data File:  C:\msw\benzo[a]pyrene-Thyssen allfatal noU 3st.(d)
 7
 8
 9
10
11
12
13
14
15
16
17
18     Total number  of observations  = 96
19     Total number  of records with  missing values  = 0
20     Total number  of parameters in model =  6
21     Total number  of specified parameters =  1
22     Degree  of polynomial = 3
23
24
25
26       User  specifies  the following parameters:
27              t 0    =         0
28
29
30
31
32
33
34                      Default Initial  Parameter Values
35                            c      =          4.5
36                            t_0    =             0   Specified
37                            beta_0 =             0
38                            beta_l =  1.37501e-010
39                            beta_2 =  2.84027e-010
40                            beta  3 =  1.44668e-037
41
42
43               Asymptotic Correlation  Matrix  of Parameter Estimates
44               ( *** The model parameter(s)   -t_0       -beta_0    -beta_l    -beta_2
45                    have been estimated at  a boundary point, or have been specified by the user,
46                    and do not appear in the correlation matrix )
47
48                    c           beta 3
49
50        c                1-1
51
52        beta 3-11
53
54
55                                    Parameter Estimates
56
57          Variable
58
59
60
61
62
63
64    NA - Indicates that this parameter has hit a
65         bound implied by some inequality constraint
66         and thus  has  no standard  error.
67
68
69                   Log (likelihood)   # Param
70       Fitted Model        -144.522         5
71
72
73
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   C-63         DRAFT—DO NOT CITE OR QUOTE

-------
23

24
                                                   Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
  Class

 F     I


 0     0

 0     0

18     0

18     0
                                    U  Total
                 BMD

                BMDL

                BMDU
                           Fatal Risk: BaP-Thv$sen_allfatal_noU_3st
                      Dose= 0.00
                                                            Dose= 0.25
^
=
~s
o
n



CO
o
,-j.
o
—
o






.^
=
OJ
o
D_

<-> \ \ I I I 1 1
0 20 60 100
CO
O
^
o
—
o






<-> I I I I I I I
0 20 60 100
                         Time
                                                               Time
                      Dose=  1.00
                                                            Dose= 4.29
^
.~t±
-Q
OJ
-Q
O
Q.


CO
o
—
CD ~
0
*
*
J
f
/

^
."t±
-D
(C
-Q
O
a.
'-> I I I I I I I
0 20 60 100
CO
o
—
CD ~
0
*
/
I
/
/
XV
L-1 I I I I I I I
0 20 60 100
                         Time
                                                               Time
               This document is a draft for review purposes only and does not constitute Agency policy.

                                              C-64        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1    DOSE-RESPONSE MODELING FOR THE DERMAL SLOPE FACTOR

 2    Modeling methods:
 3          For each endpoint, multistage models (BMDS; U.S. EPA, 2012;v2.1) were fitted to the data
 4    using the maximum likelihood method. Each model was tested for goodness-of-fit using a chi-
 5    square goodness-of-fit test (x2 p-value < 0.05 indicates lack of fit). Other factors were used to
 6    assess model fit, such as scaled residuals, visual fit, and adequacy of fit in the low-dose region and in
 7    the vicinity of the BMR. The BMDL estimate (95% lower confidence limit on the BMD, as estimated
 8    by the profile likelihood method) and AIC value were used to select a best-fit model from among the
 9    models exhibiting adequate fit. The data modeled are summarized in Tables C-22 through C-25.
10    The modeling results are summarized in Table C-26. The modeling details are provided with
11    Figures C- 8 through C-19.

12    Data adjustments prior to modeling:
13          Roe et al. (1970) applied benzo[a]pyrene dermally for 93 weeks or until natural death; with
14    the exception of the highest dose group, each group still had approximately 20 animals at 86 weeks
15    (see Table C-22). The tumors were first observed in the lowest and highest dose groups during the
16    interval of weeks 29-43.  Mice that died before week 29 were likely not at risk of tumor
17    development However, because tumor incidence and mortality were reported in 100-day
18    intervals, mice that had not been on study long enough to develop tumors were not easily
19    identifiable. Incidence denominators re fleet the number of animals alive at week 29, and may thus
20    tend to lead to underestimates of tumor risk if the number of animals at risk has been
21    overestimated.
22          Schmidt et al. (1973) did not report survival information; instead, the authors provided
23    incidences based on the numbers of mice initially included in each dose group at the start of the
24    study. Overall latency was reported  for the two high-dose groups in each series, but these data only
25    describe the survival  of mice with tumors (animals were removed from study when a tumor
26    appeared). It is not clear how long exposures lasted overall in each dose group, or whether some
27    mice may have died on study from other causes before tumors appeared. While it is possible that
28    no mice died during the study, all of the other studies considered here demonstrate mortality.
29    However, the data were modeled as reported, recognizing the possibility of underestimating risk
30    associated with incidences reported  and lack of duration of exposure. (See Table C-22.)
31          Schmahl et al. (1977) reported that reduced numbers of animals at risk (77-88 mice per
32    dose group compared with the initial group sizes of 100) resulted from varying rates of autolysis.
33    No other survival or latency information was provided, so all exposures were assumed to have
34    lasted for 104 weeks and were modeled as reported. Given the results of the other studies, it seems
35    possible that the numbers at risk in each group may be overestimated, which could  lead to an
36    underestimate of lifetime risk. (See Table C-22.)
37          Habs et al. (1980) reported age-standardized skin tumor incidence rates, indicating earlier
38    mortality in the two highest dose groups (2.8 and 4.6 |ig/application). These rates were used to
39    estimate the number  at risk in the dose-response modeling, by dividing the number of mice with

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                C-65        DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofbenzo[a]pyrene

 1    tumors by the age-standardized rates. Exposure lasted longer than 104 weeks in the two lower
 2    exposure groups, at about 120 and 112 weeks, and until about 88 weeks in the highest exposure
 3    group. Incidence in the two lower exposure groups may be higher than if the exposure had lasted
 4    just 104 weeks. There was mortality in the first 52 weeks of exposure, about 10-15% in the three
 5    exposure groups, but because there was no information concerning when tumors first appeared, it
 6    is not possible to determine how much the early mortality may have impacted the number of mice
 7    at risk in each group. (See Table C-22.)
 8          Habs et al. (1984) reported mean survival times (with 95% CIs) for each dose group. The
 9    CIs supported the judgment that the control and lower dose groups were treated for 104 weeks.
10    The higher dose group (4 [ig/application) was probably treated for <104 weeks, because the upper
11    95% confidence limit for the mean survival was approximately 79 weeks. However, since it was
12    not possible to estimate a more realistic duration for this group, an estimate of 104 weeks was
13    used. (See Table C-22.)
14          The studies by Poel (1960,1959) were conducted in male mice and used toluene as the
15    vehicle. In addition to a control group, the 1959 study included nine dose groups of one mouse
16    strain (C57L) and the 1960 study included seven dose groups of three other mouse strains. Both
17    studies demonstrated high mortality and tumor incidence at higher exposure levels. All C57L mice
18    in dose groups with >3.8 |ig/application died by week 44 of the study (Poel, 1959). Therefore,
19    these five dose groups were omitted prior to dose-response modeling because of the relatively
20    large uncertainty in extrapolating cancer risk as a result of lifetime exposure. Four dose groups in
21    addition to control remained. Among these groups, mice survived and were exposed until weeks
22    83-103.  According to the lifespan ranges provided, at least one mouse in each dose group died
23    before the first appearance of tumor, but insufficient information was available to determine how
24    many; consequently, the incidence  denominators were not adjusted. The dose-response data are
25    summarized in Table C-23.
26          For the Poel (1960) studies, all tumors in the highest three dose groups for each of the three
27    mouse strains had occurred by week 40.  While these observations support concern for cancer risk,
28    as noted above such results are relatively uncertain for estimating lifetime cancer  risk. In addition,
29    there was no information indicating duration of exposure for the mice without tumors; although
30    exposure was for lifetime, it might have been as short as for the mice with tumors. Overall, these
31    datasets did not provide sufficient information to estimate the extent of exposure associated with
32    the observed tumor incidence. Consequently, the experiments reported by Poel (1960) were not
33    used for dose-response modeling.
34          Grimmer et al. (1984,1983), studied female CFLP mice, using acetone:DMSO (1:3) as the
35    vehicle. Mean or median latency times were reported (as well as measures of variability), but no
36    information concerning overall length of exposure or survival was included in the  results. The total
37    of tumor-bearing mice and the reported percentages of mice with any skin tumors was reported
3 8    and varied, at most, one animal from the number of animals initially placed on study. The
39    decreasing latency and variability and increasing tumor incidence with increasing benzo[a]pyrene
40    exposure suggests that exposure probably did not last for 104 weeks in at least the high-dose
41    group, but the available information did not provide duration of exposure.  The data reported were
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                C-66         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene

1   modeled under the assumption that at least some animals in each group were treated and survived
2   until week 104. (See Table C-24.)
3          Sivak et al. (1997), exposed male C3H/HeJ mice dermally to benzo[a]pyrene in
4   cyclohexanone/acetone (1:1) for 24 months, and reported mean survival times for each group. All
5   high-dose mice died before the final sacrifice. From the information provided, it is apparent that
6   the animals in the control and lower two dose groups survived until study termination at week 104.
7   The study authors did not report how long treatment in the highest dose group lasted, but
8   estimation of the figure from the publication suggest that exposure duration was 74 weeks. (See
9   Table C-25).
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-67         DRAFT—DO NOT CITE OR QUOTE

-------
2
3
                                                         Toxicological Review ofbenzo[a]pyrene
Table C-22.  Skin tumor incidence, benign or malignant in female Swiss
or NMRI mice dermally exposed to benzo[a]pyrene



Study
Roeetal.,
1970a,b




Schmidt et
al., 1973°








Schmahl et
al., 1977°


Habsetal.,
1980c'f


Habsetal.,
1984°



Mouse
strain
Swiss





NMRI




Swiss




NMRI



NMRI



NMRI





Dose (ug)
0 (acetone)
0.1
0.3
1
3
9
0 (acetone)
0.05
0.2
0.8
2
0 (acetone)
0.05
0.2
0.8
2
0 (acetone)
1
1.7
3
0 (acetone)
1.7
2.6
4.6
0 (acetone)
2
4
Average
daily
dose
(ug/d)
0
0.04
0.13
0.43
1.29
3.86
0
0.01
0.06
0.23
0.57
0
0.01
0.06
0.23
0.57
0
0.29
0.49
0.86
0
0.49
0.74
1.31
0
0.57
1.14
First
appearance
of tumor
(wks)
—
29-43
-
57-71
43-57
29-43
-
-
-
53e
76e
-
-
-
58e
61e
—
NR
NR
NR
—
NR
NR
NR
-
NR
NR

Length of
exposure
(wks)
93
93
93
93
93
93
104d
104
104
104
104
104
104
104
104
104
104
104
104
104
128
120
112
88
104
104
104
Lifetime
average
daily dose
(ug/d)
0.00
0.03
0.09
0.31
0.92
2.76
0
0.01
0.06
0.23
0.57
0
0.01
0.06
0.23
0.57
0
0.29
0.49
0.86
0
0.49
0.74
0.80
0
0.57
1.14

Skin tumor
incidence (all
types)
0/49 (0%)
1/45 (2%)
0/46 (0%)
1/48 (2%)
8/47 (20%)
34/46 (74%)
0/100 (0%)
0/100 (0%)
0/100 (0%)
2/100 (2%)
30/100 (30%)
0/80 (0%)
0/80 (0%)
0/80 (0%)
5/80 (6%)
45/80 (56%)
1/81 (1%)
11/77 (14%)
25/88 (28%)
45/81 (56%)
0/35 (0%)
8/34 (24.8%)
24/27 (89.3%)
22/24 91.7%)
0/20 (0%)
9/20 (45%)
17/20 (85%)
     aDoses were applied 3 times/week for up to 93 weeks to shaved dorsal skin.
     bNumerator:  number of mice detected with a skin tumor. Tumors were thought to be malignant based on
     invasion or penetration of the panniculus carnosus muscle.  Denominator:  number of mice surviving to 29 weeks
     (200 days).
     cDoses 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.  Schmidt et al. (1973): At 0.23 u.g/d, all tumors were malignant in
     both strains; at 0.57 u.g/d, tumors were predominately malignant: 28/30 for NMRI and 42/45 for Swiss. Schmahl
     et al.,  (1977): malignant/total tumors were 10/11, 25/25, and 43/45 for the 1-, 1.7-, and 3-u.g/d groups. Habs et al.
     (1984): malignant/total tumors were 7/9 and 17/11 for the 2- and 4-u.g/d groups.
     Exposure periods not reported were assumed to be 104 weeks; indicated in italics.
     eCentral tendency estimates; range or other variability measure not reported.
     The percentages were reported by the authors as age-standardized incidences of animals with local tumors,
     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. The authors did not
     report the percentages of local tumors which were carcinomas or papillomas.
     NR = not reported
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   C-68          DRAFT—DO NOT CITE OR QUOTE

-------
1
2
                                                  Toxicological Review ofbenzo[a]pyrene

       Table C-23. Skin tumor incidence, benign or malignant, in C57L male
       mice dermally exposed to benzo[a]pyrene


Study
Poel, 1959





Mouse
strain
C57L






Dose (u.g)a
0 (toluene)
0.15
0.38
0.75
3.8
Average
daily dose
(ug/d)
0
0.06
0.16
0.32
1.63
First
appearance of
tumor (wks)
-
42
24
36
21-25
Length of
exposure
(wks)
92
98
103
94
82
Lifetime
average daily
doseb
0.00
0.05
0.16
0.24
0.80
Skin tumor
incidence (all
types)c
0/33 (0%)
5/55 (9%)
11/55 (20%)
7/56 (13%)
41/49 (84%)
3
4
aDoses were applied to interscapular skin 3 times/week for up to 103 weeks or until time of appearance of a
grossly detected skin tumor. See Table B-15 for data of five highest dose groups (19-752 u.g) in which all mice
died by week 44. These groups were not considered for dose-response modeling.
bSee Section 2.5.2. of Toxicological Reivew for discussion of extrapolation to lifetime average daily doses.
cTumors were histologically confirmed as epidermoid carcinomas.

       Table C-24. Skin tumor incidence, benign or malignant, in female CFLP
       mice dermally exposed to benzo[a]pyrene




Study
Grimmer et al.,
1983


Grimmer et al.,
1984






Dose (u,g)a
0(l:3acetone:DMSO)
3.9
7.7
15.4
0(l:3acetone:DMSO)
3.4
6.7
13.5


Average
daily dose
(ug/d)
0
1.1
2.2
4.4
0
0.97
1.9
3.9
Mean or
median time
of tumor
appearance
(wks)
—
74.6 ± 16.8b
60.9 ±13.9
44.1 ±7.7
—
61 (53-65)c
47 (43-50)
35 (32-36)


Length of
exposure
(wks)d
104
104
104
104
104
104
104
104

Lifetime
average daily
dose
(ug/d)
0
1.1
2.2
4.4
0
0.97
1.9
3.9


Skin tumor
incidence (all
types)0
0/80 (0%)
22/65 (34%)
39/64 (61%)
56/64 (88%)
0/80 (0%)
43/64 (67%)
53/65 (82%)
57/65 (88%)
5

6
    Indicated doses were applied twice/week to shaved skin of the back for up to 104 weeks.
    bMean±SD.
    cMedian and 95% confidence limit.
    dAssumed exposure period is indicated in italics.
    Incidence denominators were calculated from reported tumor-bearing animals and reported percentages.
    Grimmer et al. (1983): malignant/total tumors were 15/22, 34/39, and 54/56 for the low- through high-dose groups.
    Grimmer et al. (1984): malignant /total tumors were 37/43, 45/53, and 53/57 for the low- through high-dose
    groups.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 C-69         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
1
2
Table C-25.  Skin tumor incidence, benign or malignant, in male
C3H/HeJ mice dermally exposed to benzo[a]pyrene



Dose (u,g)a
0 (1:1 cyclohexanone/acetone)
0.05
0.5
5.0

Average
daily dose
(Hg/d)
0
0.01
0.14
1.4
First
appearance
of tumor
(wks)
-
-
NR
~43

Length of
exposure
(wks)b
104
104
104
74
Lifetime
average daily
dose
(ug/d)
0.0
0.01
0.14
0.51

Skin tumor
incidence (all
typesjc
0/30 (0%)
0/30 (0%)
5/30 (17%)
27/30 (90%)
     Indicated doses were applied twice/week to shaved dorsal skin.
     bAssumed exposure period is indicated in italics.
     °Number of skin tumor-bearing mice. In the high-dose group, 1 papilloma and 28 carcinomas were detected.  In
     the 0.5 u.g group, 2 papillomas and 3 carcinomas were detected.
     NR = not reported
     Source: Sivaketal. (1997).
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-70         DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
1
2
3
Table C-26.  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
Poel, 1959
maleC57L
Roeetal., 1970
female Swiss
Schmidt etal., 1973
female NMRI
Schmidt etal., 1973
female Swiss
Schmahletal., 1977
female NMRI
Habsetal., 1980
female NMRI
Habsetal., 1984
female NMRI
Grimmer etal., 1983
female CFLP
Grimmer etal., 1984b
female CFLP
Sivak etal., 1997
male CeH/HeJ
Model
Multistage 1°
Multistage 2°
Multistage 3°
Multistage 4°
Multistage 1°
Multistage 2°
Multistage 3°
Multistage 1°
Multistage 2°
Multistage 3°
Multistage 1°
Multistage 2°
Multistage 3°
Multistage 4°
Multistage 1°
Multistage 2°
Multistage 3°
Multistage 1°
Multistage 2
Multistage 3°
Multistage 1°
Multistage 2°
Multistage 1°
Multistage 2°
Multistage 3°
Multistage 1°
Log Logistic
Dichotomous-Hill
LogProbit
Gamma, Weibull
Logistic
Probit
Multistage 1°, high
dose dropped
Multistage 1°
Multistage 2°
Multistage 3°
Goodness-of-fit
p-value
0.011
0.027
0.053
0.068
0.110
0.485
0.485
0.008
0.609
0.999
<0.01
0.514
0.983
0.983
0.136
0.939
0.939
0.0
0.009
0.207
0.577
1.000
0.850
0.972
0.972
0.003
0.919
1.000
0.047
0.003
0.0
0.0
0.499
0.059
0.998
0.998
AIC
191.5
188.6
186.9
186.2
131.1
123.6
123.6
162.7
147.4
143.9
178.0
153.3
151.3
151.3
298.4
296.3
296.3
96.5
84.4
76.7
48.4
47.6
219.9
221.1
221.1
205.3
195.8
197.7
200.2
205.3
250.5
255.4
—
57.8
48.6
48.6
BMD10
(ug/d)
0.070
0.134
0.127
0.123
0.318
0.748
0.748
0.256
0.329
0.381
0.116
0.216
0.282
0.282
0.140
0.233
0.233
0.063
0.198
0.294
0.078
0.171
0.245
0.292
0.292
0.132
1.07
0.902
1.33
0.132
2.03
2.29
1.21
0.036
0.109
0.109
BMDL10
(ug/d)
0.057
0.078
0.078
0.077
0.249
0.480
0.480
0.194
0.287
0.326
0.093
0.192
0.223
0.223
0.117
0.149
0.143
0.050
0.143
0.215
0.056
0.060
0.208
0.213
0.213
0.113
0.479
0.533
1.11
0.113
1.76
2.03
1.01
0.026
0.058
0.052
Basis for Model Selection3
No significant improvement in model fit
with higher stage
No significant improvement in model fit
with higher stages
No significant improvement in model fit
with higher stages
No significant improvement in model fit
with higher stage
No significant improvement in model fit
with higher stage
Only model with adequate fit
No significant improvement in model fit
with higher stage
No significant improvement in model fit
with higher stages
(Higher stages did not provide better fit)
Lowest AIC among adequately fitting
models.
(Same as Multistage 1°)

No significant improvement in model fit
with higher stage
Figure
number
C-9
C-10
C-ll
C-12
C-13
C-14
C-15
C-16
C-17
C-18
C-19
C-20
     3 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.
     bThe POD for Grimmer et al. (1984), using a BMR of 70% (near response at the lowest dose), was based on the LogLogistic model. For
     comparison purposes, the multistage model was it fit to the Grimmer et al. (1984) data with the highest dose dropped (AIC not provided
     because it is not comparable to fits of the full dataset).
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                  C-71         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
                             Multistage Cancer Model with 0.95 Confidence Level
             0.6
             0.4
             0.2
                                     Multistage Cancer
                                    Linear extrapolation
                         0.1
 4
 5
 6
 1
 8
 9
10
11
12
13
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 C-8. Fit of multistage model to skin tumors in C57L mice exposed
            dermally to benzo[a]pyrene (Poel, 1959); graph and model output.
        Multistage Cancer Model.  (Version: 1.7;  Date: 05/16/2008)
        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
   The form of the probability function  is:


   P[response]  = background +  (1-background)*[1-EXP(
                 -betal*dose/xl-beta2*dose/x2-betaS*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
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-72        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
 4
 5
 6
 7
 8
 9
10
11
12
13
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
                     Background =
                       Beta(l) =
                       Beta(2) =
                       Beta(3) =
                          0.0449589
                           0.490451
                                 0
                            2.68146
the user,






Background


   Beta(l)


   Beta (3)
Asymptotic Correlation Matrix  of Parameter Estimates


 ( *** The model parameter(s)   -Beta(2)
      have been estimated  at a boundary point, or have been specified by


      and do not appear in the correlation matrix )


  Background      Beta(l)      Beta(3)


           1        -0.87         0.74


       -0.87            1        -0.92


        0.74        -0.92             1
Interval
       Variable
Limit
                                Parameter Estimates
Estimate        Std.  Err.
                                             95.0% Wald Confidence


                                          Lower Conf. Limit   Upper Conf.
Background
Beta(l)
Beta(2)
Beta(3)
0.0176699
0.79766
0
2.17146
* - Indicates that this value is not

Model
Full model
Fitted model
Reduced model
AIC:
Dose Est
0.0000 0.
0.0500 0.
0.1600 0.
0.2400 0.
0.8000 0.
Analysis of
Log (likelihood) #
-87.1835
-90.4265
-141.614
186.853
*
*
*
*
calculated.
Deviance Table
Param' s Deviance Test
5
3 6.48606
1 108.86

Goodness of Fit
. Prob. Expected Observed Size
0177 0.583
0563 3.098
1430 7.866
2128 11.917
8293 40.635
0.000 33
5.000 55
11.000 55
7.000 56
41.000 49
*
*
*
*


d.f. P-value
2 0.03905
4 <.0001

Scaled
Residual
-0.770
1.112
1.207
-1.605
0.139
     2  = 5.88       d.f. = 2        P-value = 0.0528



   Benchmark Dose  Computation


Specified effect =           0.1


Risk Type       =     Extra risk



          This document is a draft for review purposes only and does not constitute Agency policy.

                                         C-73        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene
 1   Confidence level =           0.95
 2
 3                BMD =       0.126567
 4
 5               BMDL =      0.0777875
 6
 7               BMDU =       0.272961
 8
 9   Taken together,  (0.0777875,  0.272961)  is  a  90     % two-sided confidence
10   interval for the BMD
11
12   Multistage Cancer Slope Factor =       1.28555
13
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-74         DRAFT—DO NOT CITE OR QUOTE

-------
             -a
             a)
             "5

             I
             c
             o
                   0.8
                   0.6
       0.4
                   0.2
                                                   Toxicological Review ofbenzo[a]pyrene
                                 Multistage Cancer Model with 0.95 Confidence Level
                                          Multistage Cancer
                                         Linear extrapolation
                                  0.5
                                          1.5
2.5
                                                  dose
 2

 3

 4


 6
 1
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Figure C-9. Fit of multistage model to skin tumors in female Swiss mice
exposed dermally to benzo[a]pyrene (Roe etal., 1970); graph and
model output.
               This document is a draft for review purposes only and does not constitute Agency policy.

                                              C-75        DRAFT—DO NOT CITE OR QUOTE

-------
                                               Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
        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 )
                                   Std.  Err.
   95.0% Wald Confidence Interval

Lower  Conf. Limit   Upper Conf.  Limit
- Indicates that this value is not calculated.
     Model

   Full model
 Fitted model

Reduced model


        AIC:
           Est.  Prob.
                               # Param's
                                                  Test  d.f.
                                                             P-value
                d.f. =3
          BMD =

         BMDL =

         BMDU =
        This document is a draft for review purposes only and does not constitute Agency policy.

                                          C-76         DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review ofbenzo[a]pyrene
 o

 I
 c
 o
 ••8
 ro
                         Multistage Cancer Model with 0.95 Confidence Level
         0.4
                                   Multistage Cancer
                                  Li near extrapolation
                                        BMDL
                                                        BMD
                         0.1
                                         0.2
  0.3

dose
0.4
0.5
 4

 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
       Figure C-10. 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.  (Version: 1.7;   Date: 05/16/2008)

        Input Data  File:

C:\USEPA\IRIS\benzoUlpyrene\dermalslopefactor\Schmidtl973femaleNMRI\2MulSchMS_. (d)

        Gnuplot Plotting File:

C:\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmidtl973femaleNMRI\2MulSchMS_.pit
          This document is a draft for review purposes only and does not constitute Agency policy.

                                          C-77        DRAFT—DO NOT CITE OR QUOTE

-------
                                                 Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
          Asymptotic Correlation Matrix of  Parameter Estimates
                and do not appear in the correlation matrix )


               Beta (2)


                     1




                              Parameter Estimates
                                                     95.0% Wald Confidence Interval

                                     Std. Err.     Lower Conf.  Limit   Upper Conf. Limit
      Model

     Full model

   Fitted model

  Reduced model


          AIC:
                                                               P-value
                               Goodness   of  Fit
             Est.  Prob.
                  d.f.  = 4
Specified effect =


Risk Type


Confidence level =


            BMD =


           BMDL =


           BMDU =
      0.1


Extra  risk


     0. 95


 0.329464


 0.286624


 0.384046
                                             % two-sided confidence
          This document is a draft for review purposes only and does not constitute Agency policy.

                                           C-78        DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review ofbenzo[a]pyrene
I
C
o

'•8
       0.7
       0.6
       0.5
       0.4
             0.3
       0.2
       0.1
                       Multistage Cancer Model with 0.95 Confidence Level
                                       Multistage Cancer
                                      Linear extrapolation
                                BMDL        BMP
                        0.1
                                         0.2
  0.3

dose
0.4
0.5
 4

 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
      Figure C-ll. 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: 05/16/2008)

              Input Data File:

 \USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmidtl973swissmice\3MulSchMS_. (d)

              Gnuplot Plotting File:

 \USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmidtl973swissmice\3MulSchMS_.pit
         This document is a draft for review purposes only and does not constitute Agency policy.

                                          C-79         DRAFT—DO NOT CITE OR QUOTE

-------
                                    Toxicological Review ofbenzo[a]pyrene
j
4
5
6
7
8
9
10
11
12
13
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
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
**** incorporate









Asympt

( ***

user,


B

Beta (2)

Beta (3)





Variable
Limit
Background
Beta (1)
Beta (2)
Beta (3)

these convergence criterion. Default values used. ****


Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0.338951
Beta(3) = 3.8728


otic Correlation Matrix of Parameter Estimates

The model parameter (s) -Background -Beta(l)
have been estimated at a boundary point, or have been spe

and do not appear in the correlation matrix )

eta(2) Beta(3)

1 -0.99

-0.99 1


Parameter Estimates

95.0% Wald Confid'
Estimate Std. Err. Lower Conf. Limit

0 * *
0 * *
0.108125 * *
4.31441 * *

* - Indicates that this value is not calculated.



Model
Full model
Fitted model
Reduced model

AIC:



Dose Est
0.0000 0.
0 . 0100 0 .
0.0600 0.
0 . 2300 0 .
0.5700 0.

ChiA2 = 0.16

Benchmark Dose

Specified effect

Risk Type

Confidence level

BMD
BMDL
BMDU


Analysis of Deviance Table

Log (likelihood) # Param's Deviance Test d.f. P-valu
-73.5285 5
-73.6628 2 0.268637 3 0.
-150.708 1 154.359 4 <.0

151.326

Goodness of Fit
Scaled
. Prob. Expected Observed Size Residual
0000 0.000 0.000 80 0.000
0000 0.001 0.000 80 -0.035
0013 0.106 0.000 80 -0.325
0566 4.524 5.000 80 0.230
5657 45.260 45.000 80 -0.059

d.f. =3 P-value = 0. 9833

Computation

0.1

= Extra risk

0. 95

0.282007
0.223401
0.309888

Taken together, (0.223401, 0.309888) is a 90 % two-sided confidence
interval for the

Multistage Cancer
BMD

Slope Factor = 0.447626
This document is a draft for review purposes only and does not constitute Agency policy.
                               C-80        DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review ofbenzo[a]pyrene
                        Multistage Cancer Model with 0.95 Confidence Level
I
c
o
'•8
       0.7
       0.6
       0.5
       0.4
              0.3
       0.2
       0.1
                                       Multistage Cancer
                                      Linear extrapolation
                   BMDL
                                      BMD
                    0.1
                                  0.2
0.3
0.4     0.5

 dose
0.6
0.7
0.8
0.9
 4

 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
      Figure C-12. Fit of multistage model to skin tumors in female NMRI
      mice exposed dermally to benzo[a]pyrene (Schmahl etal., 1977); graph
      and model output.
              Multistage  Cancer Model.  (Version:  1.7;  Date:  05/16/2008)

              Input Data  File:

 \USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmahll977femaleNMRI\2MulschMS_. (d)

              Gnuplot Plotting File:

 \USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Schmahll977femaleNMRI\2MulschMS_.plt
         This document is a draft for review purposes only and does not constitute Agency policy.

                                          C-81         DRAFT—DO NOT CITE OR QUOTE

-------
                                                  Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
Limit
       ****  We  are sorry but  Relative Function  and Parameter Convergence    ****

       ****  are currently unavailable in this model.  Please keep  checking  ****
                 Asymptotic Correlation Matrix  of Parameter Estimates


                   Background     Beta(l)      Beta(2)


       Background            1       -0.67         0.47


          Beta(l)        -0.67           1        -0.94


          Beta(2)         0.47       -0.94            1
                                      Parameter Estimates
     Variable         Estimate        Std. Err.


   Background

      Beta(1)
      Beta(2)


- Indicates that this  value is not calculated.
                                                             95.0% Wald Confidence Interval
                                                          Lower Conf. Limit   Upper Conf.
                              Analysis of Deviance Table
              Model
            Full model

          Fitted model

         Reduced model


                 AIC:
                                                                       P-value
                    Est. Prob.
                                                           81
                                                           88

                                                           81
                         d.f.  =  1
       Confidence level =


                   BMD =

                  BMDL =

                  BMDU =
                                                     % two-sided confidence
          This document is a draft for review purposes only and does not constitute Agency policy.

                                            C-82         DRAFT—DO NOT CITE OR QUOTE

-------
                                         Toxicological Review ofbenzo[a]pyrene
                 Multistage Cancer Model with 0.95 Confidence Level
 0.8
 0.6
 0.4
 0.2
                         Multistage Cancer
                        Linear extrapolation
       0      0.1     0.2
 4

 5
 6
 7
 8
 9
10
11
12
13
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
Figure C-13. Fit of multistage model to skin tumors in female NMRI
mice exposed dermally to benzo[a]pyrene (Habs etal., 1980); graph
and model output.
  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.plt
   This document is a draft for review purposes only and does not constitute Agency policy.

                                   C-83        DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
J
4
5
6
7
8
9
10
11
12
13
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
" J
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


Asymptotic C
( ***
The mo


orrelation Matrix of Parameter Estimates
del parameter
have been estimated
user,




Beta (3)




and do

Beta (3)

1




(s) -Background -Beta(l) -Beta (2)
at a boundary point, or have been specified by the

not appear in the correlation matrix )














Parameter Estimates


Variable
Limit
Background
Beta (1)
Beta (2)
Beta (3)










* - Indicates that this





Model
Full model
Fitted model
Reduced model

AIC:









Estimate

0
0
0
4 . 1289

value is not



Analysis of

Log (likelihood) #







-34.8527
-37.3373
-82.5767

76. 6745



95.0% Wald Confidence Interval
Std. Err. Lower Conf. Limit Upper Conf.

Tt- Tt- Tt-
* * *
* * *
* * *

calculated.



Deviance Table

Param's Deviance Test d.f. P-value
4
1 4.96903 3 0.1741
1 95.4478 3 <.0001




Goodness of Fit

Dose Es

0.0000 0
0 . 4900 0
0.7400 0
0 . 8000 0

ChiA2 = 4.56


Benchmark Dos

Specified effect

Risk Type

Confidence level

BMD

BMDL

BMDU

Taken together,
interval for the

Multistage Cance

t. Prob

.0000
.3848
.8123
.8792

d.f.



Expected

0.000
13. 082
21. 933
21. 102

= 3 P


Scaled
Observed Size Residual

0.000 35 0.000
8.000 34 -1.791
24.000 27 1.019
22 . 000 24 0 . 563

-value = 0.2067


e Computation

=

=

=

=

=

=


0.1

Extra risk

0. 95

0.294407

0.215151

0.320955

(0.215151, 0.320955)
BMD

r Slope


Factor =













is a 90 % two-sided confidence


0.46479
70
               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-84        DRAFT—DO NOT CITE OR QUOTE

-------
                                              Toxicological Review ofbenzo[a]pyrene
                       Multistage Cancer Model with 0.95 Confidence Level
I
c
o
'•8
       0.8
       0.6
             0.4
       0.2
                                Multistage Cancer
                              Linear extrapolation
 4

 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
      Figure C-14. Fit of multistage model to skin tumors in female NMRI
      mice exposed dermally to benzo[a]pyrene (Habs etal., 1984); graph
      and model output.
       Multistage Cancer Model.  (Version:  1.7;  Date: 05/16/2008)

       Input Data File: C:\Usepa\BMDS21\mscDax_Setting.(d)

       Gnuplot  Plotting File:  C:\Usepa\BMDS21\mscDax_Setting.plt
         This document is a draft for review purposes only and does not constitute Agency policy.

                                         C-85         DRAFT—DO NOT CITE OR QUOTE

-------
                                                 Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
                 Default Initial  Parameter Values

                   Background =            0

                      Beta(l)  =      1.66414
   Beta (1)
                and do not appear in the correlation matrix  )


               Beta(1)


                     1
      Variable

     Background

       Beta(1)
   95.0% Wald Confidence Interval

Lower  Conf. Limit   Upper Conf.  Limit
      Model

     Full model
   Fitted model

  Reduced model
                               Goodness  of  Fit


     Dose     Est._Prob.    Expected    Observed     Size
                  d.f. =2




   Benchmark Dose Computation


                            0.1


                      Extra risk


Confidence level =          0.95


            BMD =


           BMDL =


           BMDU =
Multistage Cancer Slope Factor =
          This document is a draft for review purposes only and does not constitute Agency policy.

                                           C-86         DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review ofbenzo[a]pyrene
                        Multistage Cancer Model with 0.95 Confidence Level
 I
 c
 o
 '•8
       0.8
       0.6
             0.4
       0.2
                                 Multistage Cancer
                                Linear extrapolation
                                                                          4.5
 4

 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
      Figure C-15. Fit of multistage model to skin tumors in female CFLP mice
      exposed dermally to benzo[a]pyrene (Grimmer etal., 1983); graph and
      model output.
        Multistage Cancer Model.  (Version:  1.7;  Date:  05/16/2008)

        Input  Data File:

::\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Grimmerl983CFLPmice\lMulGriMS_.(d)

        Gnuplot Plotting File:

::\USEPA\IRIS\benzo[a]pyrene\dermalslopefactor\Grimmerl983CFLPmice\lMulGriMS_.pit
          This document is a draft for review purposes only and does not constitute Agency policy.

                                          C-87         DRAFT—DO NOT CITE OR QUOTE

-------
                                                 Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
               Beta(1)
   Beta(1)
      Variable

     Background

       Beta(1)
                              Parameter Estimates
                                     Std. Err.
  - Indicates that this value is not calculated.
                      Analysis of Deviance Table
      Model

     Full model
   Fitted model

  Reduced model


          AIC:
                                                              P-value
             Est. Prob.
                 d.f.  =3
Specified effect =


Risk Type


Confidence level =


            BMD =


           BMDL =


           BMDU =
      0.1


Extra  risk


    0. 95


 0.244816


 0.208269


 0.289606
                                             %  two-sided confidence
          This document is a draft for review purposes only and does not constitute Agency policy.

                                           C-88         DRAFT—DO NOT CITE OR QUOTE

-------
                                                 Toxicological Review ofbenzo[a]pyrene
                        Multistage Cancer Model with 0.95 Confidence Level
        0.6
        0.4
        0.2
            EMDL,  3MD
                                 Multistage Cancer
                                Linear extrapolation
              0
                     0.5
                                    1.5
                                            2

                                         dose
                                                  2.5
                                                                 3.5
 4

 5
 6
 7
 8
 9
10
11
12
13
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
       Figure C-16.  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.  (Version: 1.7;   Date: 05/16/2008)

         Input Data File:

C:\Usepa\BMDS21\Data\msc_benzoUlpyrene_Grimmerl984_MultiCancl_0.1. (d)

         Gnuplot Plotting File:

C:\Usepa\BMDS21\Data\msc_benzoUlpyrene_Grimmerl984_MultiCancl_0.l.plt

                                                Wed Apr 27  17:11:28 2011
 Total number of observations  =  4

 Total number of records with  missing value

 Total number of parameters in model = 2

 Total number of specified parameters = 0

 Degree of polynomial = 1
          This document is a draft for review purposes only and does not constitute Agency policy.

                                           C-89         DRAFT—DO NOT CITE OR QUOTE

-------
                                                  Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
   Beta (1)
            *** The  model parameter(s)  -Background

                have been estimated  at a boundary point, or have been specified by the


                and  do not appear in the correlation matrix )


               Beta(1)


                    1
Limit
  Variable


Background

   Beta(1)
                      Estimate
                                     Std. Err.
      Model

     Full model
   Fitted model

  Reduced model


          AIC:
                            # Param's  Deviance   Test d.f.

                                 4

                                 1         11.61      3

                                 1       158.797      3
                                                               P-value
                  Prob.
                  d.f. =3
Specified effect =


Risk Type


Confidence level =


            BMD =


           BMDL =


           BMDU =
                       0.1


                  Extra risk


                      0. 95


                   0.132272


                   0.113427


                   0.154848
Taken together, (0.113427, 0.154848)  is a 90
interval  for the BMD
          This document is a draft for review purposes only and does not constitute Agency policy.

                                            C-90         DRAFT—DO NOT CITE OR QUOTE

-------
                                       Toxicological Review ofbenzo[a]pyrene
                  Log-Logistic Model with 0.95 Confidence Level
 0.6
 0.4
 0.2
                Log-Logistic
         BMDL
                      BMD
             0.5
                          1.5
                                  2

                                dose
                                        2.5
                                                     3.5
 4

 5
 6
 7
 8
 9
10
11
12
13
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
Figure C-17. Fit of log-logistic model to skin tumors in female CFLP
mice exposed dermally to benzo[a]pyrene (Grimmer etal., 1984); graph
and model output.
   This document is a draft for review purposes only and does not constitute Agency policy.

                                 C-91        DRAFT—DO NOT CITE OR QUOTE

-------
                                           Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
          Asymptotic  Correlation Matrix of Parameter Estimates
                and do not appear in  the correlation matrix )


             intercept        slope


 intercept            1        -0.68


     slope        -0.68            1





                               Parameter Estimates
Interval

      Variable

Conf.  Limit

     background
      intercept

         slope
      Estimate
                     Std. Err.
      Model

     Full model

   Fitted model

  Reduced model


          AIC:
      Analysis of Deviance Table


Log(likelihood)   # Param's  Deviance  Test d.f.    P-value
                                Goodness  of  Fit
             Est.  Prob.
   Benchmark Dose Computation


Specified effect =           0.7


Risk Type        =     Extra risk


Confidence level =          0.95


            BMD =       1.07152


           BMDL =      0.478669
   This document is a draft for review purposes only and does not constitute Agency policy.

                                     C-92         DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
                               Multistage Cancer Model with 0.95 Confidence Level
              0.8
              0.6
              0.4
              0.2
                                       Multistage Cancer
                                      Li near extrapolation
                                   0.5
                                                dose
 1

 2

 3

 4

 6
 7
 8
 9
10
11
12
13
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
       Figure C-18. Fit of multistage model to skin tumors in female CFLP mice

       exposed dermally to benzo[a]pyrene (Grimmer etal., 1984), highest

       dose dropped; graph and model output.



         Multistage Cancer Model.  (Version:  1.9;  Date: 05/26/2010)
         Input Data File: C: /Usepa/_BaP/msc_BaP_Griminerl984_drophidose_MultiCancl_0 . 7 . (d)
         Gnuplot  Plotting File:
C: /Usepa/_BaP/msc_BaP_Griminerl984_drophidose_MultiCancl_0 . 7 .pit



 [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 = 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
                 Default  Initial Parameter Values

                    Background =    0.0806622

                       Beta(l) =      0.88595
                Asymptotic Correlation Matrix of Parameter Estimates
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                  C-93          DRAFT—DO NOT CITE OR QUOTE

-------
                                                   Toxicological Review ofbenzo[a]pyrene
1
2
3
4
5
6
7
9
10
11
12
13
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
—J O
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

( ***



The model parameter (s) -Background
have been estimated
user,



and do


not appear in

at a boundary point, or have been specified by the

the correlation matrix )

Beta(l)
Beta(l)



1











Parameter Estimates


Variable
Limit
Background
Beta(l)








* - Indicates that this








Estimate

0
0.997117

value is not




95.0% Wald Confidence Interval
Std. Err. Lower Conf. Limit Upper Conf.

* * *
* * *

calculated.



Analysis of Deviance Table

Model
Full model
Fitted model
Reduced model

AIC:




Log (likelihood) #







-71.5928
-72.2756
-134.46

146.551



Param's Deviance Test d.f. P-value
3
1 1.36568 2 0.5052
1 125.735 2 <.0001




Goodness of Fit


Dose Est. Prob

0.0000 0
0.9700 0
1.9100 0

ChiA2 =1.39



.0000
.6199
.8511

d.f .



Expected

0.000
39.671
55.322

Scaled
Observed Size Residual

0.000 65 0.000
43.000 64 0.857
53.000 65 -0.809

= 2 P-value = 0.4992




Benchmark Dose Computation

Specified effect

Risk Type

Confidence level

BMD

BMDL

BMDU

Taken together,
interval for the


=

=

_

=

=

=


0.7

Extra risk

0.95

1.20745

1.00734

1.45789

(1.00734, 1.45789) is
BMD

Multistage Cancer Slope


Factor =













a 90 % two-sided confidence


0.6949
64
               This document is a draft for review purposes only and does not constitute Agency policy.
                                             C-94        DRAFT—DO NOT CITE OR QUOTE

-------
       0.6
       0.4
       0.2
                                                 Toxicological Review ofbenzo[a]pyrene
                        Multistage Cancer Model with 0.95 Confidence Level
                                 Multistage Cancer
                                Linear extrapolation
                         0.1
                                    0.2
                                                          0.4
                                                                     0.5
                                         dose
 4

 5
 6
 7
 8
 9
10
11
12
13
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
       Figure C-19.  Fit of multistage model to skin tumors in male CeH/HeJ

       mice exposed dermally to benzo[a]pyrene (Sivak et al.,  1997); graph

       and model output.



         Multistage Cancer Model.  (Version: 1.7;   Date: 05/16/2008)
         Input Data File: C:\Usepa\BMDS21\Data\msc_benzo[a]pyrene_Sivak!993_MultiCanc2_0.1.(d)
         Gnuplot Plotting File:
C:\Usepa\BMDS21\Data\msc_benzoUlpyrene_Sivakl993_MultiCanc2_0.l.plt
 Total number of observations  =  4

 Total number of records with  missing value

 Total number of parameters in model = 3

 Total number of specified parameters = 0

 Degree of polynomial = 2
          This document is a draft for review purposes only and does not constitute Agency policy.

                                           C-95         DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofbenzo[a]pyrene
 1
 2
 O

 4
 5
 6

 8
 9
10
11
12
13
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
                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 )
            Variable        Estimate        Std.  Err.

          Background

             Beta(1)

             Beta(2)


        -  Indicates  that this value is not calculated.
            Model

          Full model
         Fitted model

        Reduced model


                AIC:
                                                                     P-value
                        Prob.
                       d.f.  =3
      Specified effect  =


      Risk Type


      Confidence level  =


                  BMD  =


                 BMDL  =


                 BMDU  =
      0.1


Extra  risk


     0. 95


 0.108575


 0.058484


 0.129641
                This document is a draft for review purposes only and does not constitute Agency policy.

                                                 C-96         DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene

 1   ALTERNATIVE APPROACHES FOR CROSS-SPECIES SCALING OF THE DERMAL SLOPE
 2   FACTOR
 3          Several publications which develop a dermal slope factor for benzo[a]pyrene are available
 4   in the peer reviewed literature (Knafla et al., 2010; 2006; Hussain et al., 1998; LaGoy and Quirk
 5   1994; Sullivan et al., 1991). With the exception of the 2010 Knafla etal. publication, none of these
 6   approaches applied quantitative adjustments to account for interspecies differences, though the
 7   proposed slope factors were developed to account for human risk. Knafla et al. (2010) qualitatively
 8   discuss processes which could affect the extrapolation between mice and humans including skin
 9   metabolic activity adduct formation, stratum corneum thickness, epidermal thickness, etc.
10   Ultimately, the authors apply an adjustment based on the increased epidermal thickness of human
11   skin on the arms and hands compared to mouse interscapular epidermal thickness.  They
12   hypothesize that the  carcinogenic potential of benzo[a]pyrene may be related to the thickness of
13   the epidermal layer.
14          Because there is no established methodology for cross-species extrapolation of dermal
15   toxicity, several alternative approaches were evaluated. Each approach begins with the POD of
16   0.066 [J.g/day that was based on a 10% extra risk for skin tumors in male mice. Based on the
17   assumptions of each approach, a dermal slope factor for humans is calculated. The discussion of
18   these approaches uses the following abbreviations:
19
20              DSF = dermal slope factor
21              PODM = point of departure (for 10% extra risk) from mouse bioassay, in [ig/day
22              BWM= mouse body weight = 0.035 kg (assumed)
23              BWn = human body weight = 70 kg (assumed)
24              SAn = total human surface area = 19,000 cm2 (assumed)
25              SAM = total mouse surface area = 100 cm2 (assumed)
26
27   Approach 1. No interspecies adjustment to daily applied dose (POD) in mouse model
28          Under this approach, a given mass of benzo[a]pyrene, applied daily, would pose the same
29   risk in an animal or in humans, regardless of whether it is applied to a small surface area or to a
30   larger surface area at a proportionately lower concentration.
31
32                DSF=0.1/PODM
33
34                DSF= 0.1/0.068 ^g/day = 1.5 (ug/day)1
35
36          Assumptions: The same mass of benzo[a]pyrene, applied daily, would have same potency in
37   mice as in human skin regardless of treatment area.
38
39   Approach 2. Cross-species adjustment based on whole body surface-area scaling
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               C-97        DRAFT—DO NOT CITE OR QUOTE

-------
                                                    Toxicological Review ofbenzo[a]pyrene

 1          Under this approach, animals and humans are assumed to have equal lifetime cancer risk
 2   with equal average whole body exposures in loading units ([ig/cm2-day).  As long as doses are low
 3   enough that risk is proportional to the mass of applied compound, the daily dermal dose of
 4   benzo[a]pyrene can be normalized over the total surface area.
 5
 6          POD Og/cm2-day) = PODM/sA Og/cm2-day) = PODM Og/day) / SAM (cm*)
 7
 8          POD = (0.068 ng/day) / 100 cm2
 9              = 0.00068 [ig/cm2-day
10
11          DSF = 0.1/(0.00068 [ig/cm2-day) * 147 (ng/cm2-day) *
12
13          Assumptions: Mouse and human slope factors are equipotent if total dermal dose is
14   averaged over equal fractions of the entire surface area. Tumor potency of benzo[a]pyrene is
15   assumed to be related to overall dose and not dose per unit area.  For example, a human exposed to
16   0.01 [J.g/day on 10 cm2 would be assumed to have the same potential to form a skin tumor as
17   someone treated with 0.01 [ig/day over 19,000 cm2 (assumed human surface area).
18
19   Approach 3.  Cross-species adjustment based on body weight
20          Under this approach, a given mass of benzo[a]pyrene is normalized relative to the body
21   weight of the animal or human. This approach has been used for oral doses for noncancer effects.
22
23                PODM/ BWM= 0.068 ng/0.035 kg-day = 1.9 ng/kg-day
24
25                DSF = 0.1/1.9 ng/kg-day = 0.051 (ug/kg-day) 1
26
27          Assumptions: The potency of point of contact skin tumors is related to bodyweight and
28   humans and mice would have an equal likelihood of developing skin tumors based on a dermal dose
29   per kg basis.
30
31          Issues: Skin cancer following benzo[a]pyrene exposure is a local effect and not likely
32   dependent on body weight.
33
34   Approach 4.  Cross-species adjustment based on allometric scaling using body weight to the
35   3/4 power
36          Under this approach, rodents and humans exposed to the same daily dose of a carcinogen,
37   adjusted for BW3/4, would be expected to have equal lifetime risks of cancer. That is, a lifetime dose
38   expressed as |ig/kg3/4-day would lead to an equal risk in rodents and humans. This scaling reflects
39   the empirically observed phenomena of more rapid distribution, metabolism, and clearance in
40   smaller animals. The metabolism of benzo[a]pyrene to reactive intermediates is a critical step in
41   the carcinogenicity of benzo[a]pyrene, and this metabolism occurs in the skin.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                              C-98        DRAFT—DO NOT CITE OR QUOTE

-------
                                                     Toxicological Review ofbenzo[a]pyrene

 1
 2          POD (ug/day) = PODM (^g/day) x (BWH / BWM)3/4
 3
 4          POD (tig/day) = 0.068 ng/day x (70 kg / 0.035 kg)3/4
 5                        = 20.3[ig/day
 6
 7          DSF = 0.1/(20.3 Lig/day) * 0.0049 (ng/day)-1
 8
 9          Assumptions: Risk at low doses of benzo[a]pyrene is dependent on absolute dermal dose
10    and not dose per unit of skin, meaning a higher exposure concentration of benzo[a]pyrene
11    contacting a smaller area of exposed skin could carry the same risk of skin tumors as a lower
12    exposure concentration of benzo [a] pyrene that contacts a larger area of skin.
13
14          Issues: It is unclear if scaling of doses based on bodyweight ratios will correspond to
15    differences in metabolic processes in the skin of mice and humans.
16
17    Synthesis of the alternative approaches to cross-species scaling
18          A comparison of the above approaches is provided in Table C-27 below. The lifetime risk
19    from a nominal human dermal exposure to benzo [ajpyrene over a 5% area of exposed skin
20    (approximately 950 cm2), estimated at 1  x 10 ~4 [ig/day*, is calculated for each of the approaches in
21    order to judge whether the method yields risk estimates that are unrealistically high.
22
23    Other potential interspecies adjustments
24          The above discussion presents several mathematical approaches that result from varying
25    assumptions about what is the relevant dose metric for  determining equivalence across species.
26    Biological information (that is not presently comprehensive or detailed enough to develop robust
27    models) that could be used in future biologically based models for cross-species extrapolation
28    include:
29       a.  Quantitative information on interspecies differences in partitioning from exposure medium
30          to the skin and absorption through the skin
31       b.  Thickness of the stratum corneum between anatomical sites and between species
32       c.  Thickness of epidermal layer
33       d.  Skin permeability
34       e.  Metabolic activity of skin
35       f.  Formation of DNA adducts in skin
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                C-99         DRAFT—DO NOT CITE OR QUOTE

-------
                                                                                         Toxicological Review ofbenzo[a]pyrene
            Table C-27. Alternative approaches to cross-species scaling

Approach



1. Mass-per-
day scaling




2. Surface-
area scaling





3. Body-
weight
scaling



4. Allometric

scaling
(BW3/4)
\1_» V V /

Assumptions

Equal mass per day (u.g /d), if applied to equal 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 (u.g /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 (u.g /cm2-day) and skin area exposed (cm2) that have the same product when
multiplied, will result in the same risk.
Equal mass per day (U.R /d), if applied to equal fractions of total skin surface (cm2) will
have similar cancer risks. That is, lifetime exposure normalized over the whole body
[e.g., 5%-of-the-body lifetime exposure] at the same loading rate (u.g /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.
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 (u.g /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 (u.g /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 (u.g /day) and body weight (kg) that have the same result when
divided, will result in the same risk.
Same as for body-weight scaling, except that benzo[a]pyrene 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.

Dose metric




M-g/day




u.g/cm2-day






u.g/kg-day






ug/day


DSF



1.3 per
u.g/day




128 per
u.g/cm2-day





f\ f\JI C n mr
0.045 per
us/ks-dav
r*o/ lxo « M y




0.0043 per u.g
/day

Risk at nominal
exposure
(0.0001 ng/day)*


1 x 10"4





7 x 10"7






6 x 10"8





4 x 10"7


2    * Nominal exposure calculated as a geometric mean of average daily doses (u.g/day) calculated from a range of benzo[a]pyrene soil concentrations (1-1000
3    ppb) reported from non-contaminated rural/agricultural soils (ATSDR, 1995) and a range of standard exposure assumptions.
                                This document is a draft for review purposes only and does not constitute Agency policy.
                                                C-100        DRAFT—DO NOT CITE OR QUOTE

-------
                                 Toxicological Review ofbenzo[a]pyrene

i   APPENDIX D.   SUMMARY OF EXTERNAL PEER
2     REVIEW AND PUBLIC COMMENTS AND EPA'S
3     DISPOSITION
4
5
         This document is a draft for review purposes only and does not constitute Agency policy.
                             D-l       DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofbenzo[a]pyrene
 1
 2    REFERENCES FOR APPENDICES
 3    Multiple references published in the same year by the same author(s) have been assigned a letter
 4    (e.g., 1986a, 1986b) based on order of appearance in the text of the document Those same letters
 5    have been retained for the appendices.
 6
 7    Achard, S; Perderiset, M; Jaurand, MC. (1987) Sister chromatid exchanges in rat pleural mesothelial cells
 8    treated with crocidolite, attapulgite, or benzo 3-4 pyrene. Br J Ind Med 44:281-283.
 9
10    Adler, ID; Ingwersen, I. (1989) Evaluation of chromosomal aberrations in bone marrow of 1C3F1 mice.  Mutat
11    Res224(3):343-345.
12
13    Adler, ID; Kliesch, U; Kiefer, F. (1989) Clastogenic effects of benzo [ajpyrene in postimplantation embryos with
14    different genetic background. Teratog Carcinog Mutagen 9(6):383-392.
15
16    Agarwal, R; Medrano, EE; Khan, IU; et al. (1991) Metabolism of benzo [ajpyrene by human melanocytes in
17    culture. Carcinogenesis 12(10):1963-1966.
18
19    Agrelo, C; Amos, H. (1981) DNA repair in human fibroblasts. In: De Serres, FJ.Ashby, J eds.: Elsevier/North-
20    Holland, 528-532.
21
22    Albert, RE; Miller, ML; Cody, T; et al. (1991) Benzo [ajpyrene-induced skin damage and tumor promotion in
23    the mouse. Carcinogenesis 12(7):1273-1280.
24
25    Alexandrie, AK; Nyberg, F; Warholm, M; et al. (2004) Influence of CYP1A1, GSTM1, GSTT1, and NQ01
26    genotypes and cumulative smoking dose on lung cancer risk in a Swedish population. Cancer Epidemiol
27    BiomarkersPrevl3(6):908-914.
28
29    Alexandrov, K; Rojas, M; Goldberg, M; et al. (1990) A new sensitive fluorometric assay for the metabolism of
30    (±)-7,8-dihydroxy-7,8-dihydrobenzo[ajpyrene by human hair follicles. Carcinogenesis 11:2157-2161.
31
32    Alfheim, I; Ramdahl, T. (1984) Contribution of wood combustion to indoor air pollution as measured by
33    mutagenicity in Salmonella and polycyclic aromatic hydrocarbon concentration. Environ Mutagen 6(2):121-
34    130.
35
36    Alzieu, P; Cassand, P; Colin, C; et al. (1987) Effect of vitamins A, C and glutathione on the mutagenicity of
37    benzo [ajpyrene mediated by S9 from vitamin A-deficient rats. Mutat Res 192(4):227-231.
38
39    Ames, BN; Durston,  WE; Yamasaki, E; et al. (1973) Carcinogens are mutagens: a simple test system combining
40    liver homogenates for activation and bacteria for detection. Proc Natl Acad Sci USA 70(8):2281-2285.
41
42    Ames, BN; McCann,  J; Yamasaki, E. (1975) Methods for detecting carcinogens and mutagens with the
43    Salmonella/mammalian-microsome mutagenicity test. Mutat Res 31(6):347-364.
44
45    Amacher, DE; Paillet, SC. (1983) The activation of procarcinogens to mutagens by cultured rat hepatocytes in
46    the L5178Y/TK mutation assay. Mutat Res 113:77-88.
47
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Amacher, DE; Paillet, SC; Turner, GN; et al. (1980) Point mutations at the thymidine kinase locus in L5178Y
 2    mouse lymphoma cells. II. Test validation and interpretation. Mutat Res 72:447-474.
 o
 4    Amacher, DE; Turner, GN. (1980) Promutagen activation by rodent-liver postmitochondrial fractions in the
 5    L5178Y/TK cell mutation assay. Mutat Res 74:485-501.
 6
 7    Ampy, FR; Saxena, S; Verma, K. (1988) Mutagenicity of benzo(a)pyrene in uninduced tissues from BALB/c
 8    mice and Sprague-Dawley rats as an index of possible health risks using the Salmonella mutagenicity assay.
 9    Cytobios 56(225):81-87.
10
11    Andrysik, Z; Vondracek, J; Machala, M; et al. (2007) The aryl hydrocarbon receptor-dependent deregulation of
12    cell cycle control induced by polycyclic aromatic hydrocarbons in rat liver epithelial cells. Mutat Res 615(1-
13    2):87-97.
14
15    Antignac, E; Koch, B; Grolier, P; et al. (1990) Prochloraz as potent inhibitor of benzo[a]pyrene metabolism
16    and mutagenic activity in rat liver fractions. Toxicol Lett. 54(2-3):309-315.
17
18    Arce, GT; Allen, JW;  Doerr, CL; et al. (1987) Relationships between benzo(a)pyrene-DNA adduct levels and
19    genotoxic effects in  mammalian cells.  Cancer Res 47(13):3388-3395.
20
21    Archibong, AE; Inyang, F; Ramesh, A; et al. (2002) Alteration of pregnancy related hormones and fetal
22    survival in F-344 rats exposed by inhalation to benzo(a)pyrene.  Reprod Toxicol 16(6):801-808.
23
24    Archibong, AE; Ramesh, A; Niaz, MS; et al. (2008) Effects of benzo(a)pyrene on intra-testicular function in F-
25    344 rats. IntJ Environ Res Public Health 5(1):32-40.
26
27    ATSDR (Agency for  Toxic Substances and Disease Registry). (1995) Toxicological profile for polycyclic
28    aromatic hydrocarbons. Atlanta, GA: Agency for Toxic Substances and Disease Registry.
29
30    Autrup, H; Seremet, T. (1986)  Excretion of benzo[a]pyrene-Gua adduct in the urine of benzo[a]pyrene-treated
31    rats. ChemBiol Interact 60(2):217-226.
32
33    Autrup, H; Harris, CC; Stoner, GD; et al. (1978) Metabolism of [3H]benzo[a]pyrene by cultured human
34    bronchus and cultured human pulmonary alveolar macrophages. Lab Invest 38(3):217-224.
35
36    Autrup, H; Wefald, FC; Jeffrey, AM; et al. (1980) Metabolism of benzo[a]pyrene by cultured tracheobronchial
37    tissues from mice, rats, hamsters, bovines and humans. IntJ Cancer 25(2):293-300.
38
39    Awogi, T; Sato, T. (1989) Micronucleus test with benzo[a]pyrene using a single peroral administration and
40    intraperitoneal injection in males of the MS/Ae and CD-I mouse strains.  Mutat Res 223(4):353-356.
41
42    Babson, JR; Russo-Rodriguez, SE;  Rastetter, WH; et al. (1986) In vitro DNA-binding of microsomally-activated
43    fluoranthene: evidence that the major product is a fluoranthene N2-deoxyguanosine adduct. Carcinogenesis
44    7:859-865.
45
46    Balansky, R; Mircheva, Z; Blagoeva, P. (1994) Modulation of the mutagenic activity of cigarette smoke,
47    cigarette smoke condensate and benzo[a]pyrene in vitro and in vivo. Mutagenesis 9:107-112.
48
49    Bao, H; Vepakomma, M; Sarkar, MA. (2002) Benzo(a)pyrene exposure induces CYP1A1 activity and
50    expression in human endometrial cells. J  Steroid Biochem Mol Biol 81(l):37-45.
51
52    Barfknecht, TR; Hites, RA; Cavaliers, EL; et al. (1982) Human cell mutagenicity of polycyclic aromatic
53    hydrocarbon components of diesel emissions. Dev Toxicol Environ Sci 10:277-294.
54
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    2             DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Bayer, R. (1978) In vivo induction of sister chromatid exchanges by three polyaromatic hydrocarbons.
 2    Carcinogenesis 3:423-428.
 o
 4    Beland, F; Gulp, S. (1998) Chronic bioassay of two composite samples from selected manufactured gas plant
 5    waste sites. Jefferson, AK:  Research, NCfT; Technical Report 6722.02 (unpublished).
 6
 7    Benjamin, H; Storkson, J; Tallas, PG; et al. (1988) Reduction of benzo[a]pyrene-induced forestomach
 8    neoplasms in mice given nitrite and dietary soy sauce. Food Chem Toxicol 26(8):671-678.
 9
10    Berenblum, I; Haran, N. (1955) The influence of croton oil and of polyethylene glycol-400 on carcinogenesis
11    in the forestomach of the mouse. Cancer Res 15(8):510-516.
12
13    Bevan, DR; Sadler, VM. (1992) Quinol diglucuronides are predominant conjugated metabolites found in bile of
14    rats followingintratracheal instillation of benzo[a]pyrene. Carcinogenesis 13(3):403-407.
15
16    Bevan, DR; Ulman, MR. (1991) Examination of factors that may influence disposition ofbenzo[a]pyrene in
17    vivo: vehicles and asbestos. Cancer Lett 57(2):173-179.
18
19    Bevan, DR; Weyand, EH. (1988) Compartmental analysis of the disposition of benzo[a]pyrene in rats.
20    Carcinogenesis 9(11):2027-2032.
21
22    Biancifiori, C; Caschera, F; Giornelli-Santulli, F; et al. (1967) The action of oestrone and four chemical
23    carcinogens in intact and ovariectomized BALB/c/Cb/Se mice. Br J Cancer 21:452-459.
24
25    Bingham, E; Falk, HL. (1969) Environmental carcinogens. The modifying effect of cocarcinogens on the
26    threshold response. Arch Environ  Health 19(6):779-783.
27
28    Blaha, L; Kapplova, P; Vondracek, J; et al. (2002) Inhibition of gap-junctional intercellular communication by
29    environmentally occurring polycyclic aromatic hydrocarbons. Toxicol Sci 65(1):43-51.
30
31    Boerrigter, ME. (1999) Treatment of lacZ plasmid-based transgenic mice with benzo[a]pyrene: measurement
32    of DNA adduct levels, mutant frequencies, and mutant spectra. Environ Mol Mutagen 34(2-3):140-147.
33
34    Bol, SA; van Steeg, H; Jansen, JG; et al. (1998)  Elevated frequencies of benzo(a)pyrene-induced HPRT
35    mutations in internal tissue ofXPA-deficientmice. Cancer Res 58(13):2850-2856.
36
37    Bos, RP; Theuws, JL; Jongeneelen, FJ; et al. (1988) Mutagenicity of bi-, tri- and tetra-cyclic aromatic
3 8    hydrocarbons in the "taped-plate assay" and in the conventional salmonella mutagenicity assay. Mutat Res
39    204:203-206.
40
41    Bouayed, J; Desor, F;  Rammal, H; et al. (2009) Effects of lactational exposure to benzo[alpha]pyrene
42    (B[alpha]?) on postnatal neurodevelopment, neuronal receptor gene expression and behaviour in mice.
43    Toxicology 259(3):97-106.
44
45    Bowman, ED; Rothman, N; Hackl, C; et al. (1997) Interindividual variation in the levels of certain urinary
46    polycyclic aromatic hydrocarbon metabolites following medicinal exposure to coal tar ointment.  Biomarkers
47    2:321-327.
48
49    Briede, JJ; Godschalk, RW; Emans, MT; et al. (2004) In vitro and in vivo studies on oxygen free radical and
50    DNA adduct formation in rat lung and liver during benzo[a]pyrene metabolism. Free Radic Res 38(9):995-
51    1002.
52
53    Brooks, RA; Gooderham, NJ; Edwards, RJ; et al. (1999) The mutagenicity of benzo[a]pyrene in mouse small
54    intestine. Carcinogenesis 20(1):109-114.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    3             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Bruce, WR; Heddle, JA. (1979) The mutagenic activity of 61 agents as determined by the micronucleus,
 2    Salmonella, and sperm abnormality assays. Can J Genet Cytol 21(3):319-334.
 o
 4    Brune, H; Deutsch-Wenzel, RP; Habs, M; et al. (1981) Investigation of the tumorigenic response to
 5    benzo(a)pyrene in aqueous caffeine solution applied orally to Sprague-Dawley rats. J Cancer Res Clin Oncol
 6    102(2):153-157.
 7
 8    Burstyn, I; Kromhout, H; Partanen, T; et al. (2005) Polycyclic aromatic hydrocarbons and fatal ischemic heart
 9    disease. Epidemiology 16(6):744-750.
10
11    Byczkowski, JZ; Kulkarni, AP. (1990) Lipid peroxidation-coupled co-oxygenation of benzo(a)pyrene and
12    benzo(a)pyrene-7,8-dihydrodiol in human term placental microsomes. Placenta ll(l):17-26.
13
14    Calaf, G; Russo, J. (1993) Transformation of human breast epithelial cells by chemical carcinogens.
15    Carcinogenesis 14:483-492.
16
17    Carver, JH; Machado, ML; MacGregor, JA. (1986) Application of modified Salmonella/microsome prescreen to
18    petroleum-derived complex mixtures and polynuclear aromatic hydrocarbons (PAH). Mutat Res 174(4): 247-
19    253.
20
21    Casto, BC; Pieczynski, WJ; Janosko, N; et al. (1976) Significance of treatment interval and DNA repair in the
22    enhancement of viral transformation by chemical carcinogens and mutagens. Chem Biol Interact 13(2):105-
23    125.
24
25    Casto, BC; Janosko, N; DiPaolo, JA. (1977) Development of a focus assay model for transformation of hamster
26    cells in vitro by chemical carcinogens. Cancer Res 37:3508-3515.
27
28    Cavalieri, E; Rogan, E; Toth, B; et al. (1981) Carcinogenicity of the environmental pollutants cyclopenteno-
29    [cdjpyrene and cyclopentano[cd]pyrene in mouse skin.  Carcinogenesis 2(4):277-281.
30
31    Cavalieri, EL; Higginbotham, S; RamaKrishna, NV; et al. (1991) Comparative dose-response tumorigenicity
32    studies of dibenzo[alpha,l]pyrene versus 7,12-dimethylbenz[alpha]anthracene, benzo[alpha]pyrene and two
33    dibenzo [alpha,!] pyrene dihydrodiols in mouse skin and rat mammary gland. Carcinogenesis 12(10):1939-
34    1944.
35
36    Cavret, S; Laurent, C; Feidt, C; et al. (2003) Intestinal absorption of 14C from 14C-phenanthrene,
37    14C-benzo[a]pyrene and 14C-tetrachlorodibenzo-para-dioxin: approaches with the Caco-2 cell line and with
38    portal absorption measurements in growing pigs. Reprod Nutr Dev 43(2):145-154.
39
40    Chen, C; Tang, Y; Jiang, X; et al. (2012) Early postnatal benzo(a)pyrene exposure in Sprague-Dawley rats
41    causes persistent neurobehavioral impairments that emerge postnatally and continue into adolescence and
42    adulthood. Toxicol Sci. Jan;125(l):248-61.
43
44    Chen, X; An, H; Ao, L; et al. (2011) The combined toxicity of dibutyl phthalate and benzo(a)pyrene on the
45    reproductive system of male Sprague Dawley rats in vivo. J Hazard Mater 15 (Feb);186(l):835-41.
46
47    Chen, L; Devanesan, PD; Higginbotham, S; et al. (1996) Expanded analysis of benzo[a]pyrene-DNA adducts
48    formed in vitro and in mouse skin: their significance in tumor initiation. Chem Res Toxicol 9(5):897-903.
49
50    Chen, S; Nguyen, N; Tamura, K; et al. (2003) The role of the Ah receptor and p38 in benzo[a]pyrene-
51    7,8-dihydrodiol and benzo[a]pyrene-7,8-dihydrodiol-9,10-epoxide-induced apoptosis. J Biol Chem
52    278(21):19526-19533.
53
54    Choi, JY; Lee, KM;  Cho, SH; et al. (2003) CYP2E1 and NQ01 genotypes, smoking and bladder cancer.
55    Pharmacogenetics 13(6):349-355.

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                    4             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene
 2    Chouroulinkov, I; Gentil, A; Guerin, M. (1967) Study of the carcinogenic activity of 9,10-dimethyl-
 3    benzanthracene and of 3,4-benzopyrene given orally. Bull Cancer 54(l):67-78.
 4
 5    Chung, JY; Kim, YJ; Kim, JY; et al. (2011) Benzo[a]pyrene reduces testosterone production in rat Leydig cells
 6    via a direct disturbance of testicular steroidogenic machinery. Environ Health Perspect.ll9(ll):1569-74.
 7
 8    Chung, JY; Kim, JY; Kim, WR; et al. (2007) Abundance of aryl hydrocarbon receptor potentiates
 9    benzo[a]pyrene-induced apoptosis in Hepalclc7 cells via CYP1A1 activation. Toxicology 235(l-2):62-72.
10
11    Clement Associates. (1990) Development of a dose-response model for inhaled B[a]P. Fairfax, VA: Clement
12    Associates.
13
14    Clive, D; Johnson, KO; Spector, JF; et al. (1979) Validation and characterization of the L5178Y/TK+/- mouse
15    lymphoma mutagen assay system. MutatRes 59(1):61-108.
16
17    Colapietro, AM; Goodell, AL; Smart, RC. (1993) Characterization of benzo[a]pyrene-initiated mouse skin
18    papillomas for Ha-ras mutations and protein kinase C levels. Carcinogenesis 14:2289-2295.
19
20    Cohen, GM; Haws, SM; Moore, BP; et al. (1976) Benzo(a)pyrene-3-yl hydrogen sulphate, a major ethyl acetate-
21    extractable metabolite of benzo (a)pyrene in human, hamster and rat lung cultures.  Biochem Pharmacol
22    25(23):2561-2570.
23
24    Conney, AH; Chang, RL; Jerina, DM; et al. (1994) Studies on the metabolism of benzo [ajpyrene and dose-
25    dependent differences in the mutagenic profile of its ultimate carcinogenic metabolite. Drug Metab Rev
26    26(1-2):125-163.
27
28    Cooper, AR; Moley, KH. (2008) Maternal tobacco use and its preimplantation effects on fertility: more reasons
29    to stop smoking.  Semin ReprodMed 26:204-212.
30
31    Cosma, GN; Marchok, AC. (1988) Benzo [ajpyrene- and formaldehyde-induced DNA damage and repair in rat
32    tracheal epithelial cells. Toxicology 51(2-3):309-320.
33
34    Cosma, GN; Jamasbi, R; Marchok, AC. (1988) Growth inhibition and DNA damage induced by benzo [ajpyrene
35    and formaldehyde in primary cultures of rat tracheal epithelial cells. Mutat Res 201(1):161-168.
36
37    Craig-Holmes, AP; Shaw, MW. (1977) Effects of six carcinogens on SCE frequency and cell kinetics in cultured
38    human lymphocytes. Mutat Res 46:375-384.
39
40    Crespi, CL; Altman, JD; Marietta, MA. (1985) Xenobiotic metabolism and mutation in a human lymphoblastoid
41    cell line. Chem Biol Interact 53:257-271.
42
43    Crofton-Sleigh, C; Doherty, A; Ellard, S; et al. (1993) Micronucleus assays using cytochalasin-blocked MCL-5
44    cells, a proprietary human cell line expressing five human cytochromes P-450 and microsomal epoxide
45    hydrolase. Mutagenesis 8(4):363-372.
46
47    Crowell, SR; Amin, SG; Anderson, KA; et al. (2011) Preliminary physiologically based pharmacokinetic models
48    for benzo [ajpyrene and dibenzo[def,p]chrysene in rodents. Toxicol Appl Pharmacol. Dec 15;257(3):365-76.
49
50    Gulp, SJ; Gaylor, DW; Sheldon, WG; et al. (1998) A comparison of the tumors induced by coal tar and
51    benzo [ajpyrene in a 2-year bioassay. Carcinogenesis 19(1):117-124.
52
53    Gulp, SJ; Warbritton, AR; Smith, BA; et al. (2000) DNA adduct measurements, cell proliferation and tumor
54    mutation induction in relation to tumor formation in B6C3F1 mice fed coal tar or benzo [ajpyrene.
55    Carcinogenesis 21(7):1433-1440.

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    5             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene
 2    Dahl, AR; Coslett, DS; Bond, JA; et al. (1985) Metabolism of benzo[a]pyrene on the nasal mucosa of Syrian
 3    hamsters: comparison to metabolism by other extrahepatic tissues and possible role of nasally produced
 4    metabolites in carcinogenesis. J Natl Cancer Inst75(l):135-139.
 5
 6    Danheiser, SL; Liber, HL; Thilly, WG. (1989) Long-term, low-dose benzo[a]pyrene-induced mutation in human
 7    lymphoblasts competent in xenobiotic metabolism. MutatRes 210(1):143-147.
 8
 9    Davidson, GE; Dawson, GW. (1976) Chemically induced presumed somatic mutations in the mouse. Mutat Res
10    38(2):151-154.
11
12    Dean, BJ. (1981) Activity of 27 coded compounds in the RL1 chromosome assay. Evaluation of short-term
13    tests for carcinogens: Report of the internationalcollaborative program. Prog Mutat Res 1:570-579.
14
15    Dean, JH; Luster, MI; Boorman, GA; et al. (1983) Selective immunosuppression resulting from exposure to the
16    carcinogenic congener of benzopyrene in B6C3F1 mice. Clin Exp Immunol 52(1):199-206.
17
18    Dean, SW; Coates, A; Brooks, TM; etal. (1998)  Benzo[a]pyrene site of contact mutagenicity in skin of Muta
19    mouse. Mutagenesis 13(5):515-518.
20
21    De Flora, S; D'Agostini, F; Izzotti, A; et al. (1991) Prevention by N-acetylcysteine of benzo[a]pyrene
22    clastogenicity and DNAadducts in rats. MutatRes 250(l-2):87-93.
23
24    De Jong, WH; Kroese, ED; Vos, JG; et al. (1999)  Detection of immunotoxicity of benzo[a]pyrene in a subacute
25    toxicity study after oral exposure in rats. Toxicol Sci 50:214-220.
26
27    de Raat, WK. (1979) Comparison of the induction by cigarette smoke condensates of sister-chromatid
28    exchanges in Chinese hamster cells and of mutations in Salmonella typhimurium. Mutat Res 66(3):253-259.
29
30    Dertinger, SD; Lantum, HB; Silverstone, AE; et al. (2000) Effect of 3'-methoxy-4'-nitroflavone on
31    benzo[a]pyrene toxicity. Aryl hydrocarbon receptor-dependent and -independent mechanisms. Biochem
32    Pharmacol60(2):189-196.
33
34    Dertinger, SD; Nazarenko, DA; Silverstone, AE; et al. (2001) Aryl hydrocarbon receptor signaling plays a
35    significant role in mediating benzo[a]pyrene- and cigarette smoke condensate-induced cytogenetic damage in
36    vivo. Carcinogenesis 22(1): 171-177.
37
38    Deutsch, J; Vatsis, KP; Coon, MJ; et al. (1979) Catalytic activity and stereoselectivity of purified forms of rabbit
39    liver microsomal cytochrome P-450 in the oxygenation of the (-) and (+) enantiomers of trans-7,8-dihydroxy-
40    7,8-dihydrobenzo[alpha]pyrene. Mol Pharmacol 16(3):1011-1018.
41
42    Diamond, L; Kruszewski, F; Aden, DP; et al. (1980) Metabolic activation of benzo [ajpyrene by a human
43    hepatoma cell line.  Carcinogenesis l(10):871-875.
44
45    DiPaolo, JA; Donovan, P; Nelson, R. (1969) Quantitative studies of in vitro transformation by chemical
46    carcinogens. J Natl Cancer Inst 42(5):867-874.
47
48    DiPaolo, JA; Donovan, PJ; Nelson, RL. (1971) X-irradiation enhancement of transformation by benzo(a)pyrene
49    in hamster embryo cells. Proc Natl Acad Sci USA 68:1734-1737.
50
51    Dreij, K; Sundberg, K; Johansson, AS; et al. (2002) Catalytic activities of human alpha class glutathione
52    transferases toward carcinogenic dibenzo[a,l]pyrene diol epoxides. Chem Res Toxicol 15(6):825-831.
53
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    6             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Dunkel, VC; Pienta, RJ; Sivak, A; et al. (1981) Comparative neoplastic transformation responses of Balb/3T3
 2    cells, Syrian hamster embryo cells, and Rauscher murine leukemia virus-infected Fischer 344 rat embryo cells
 3    to chemical compounds. J Natl Cancer Inst 67:1303-1312.
 4
 5    Egert, G; Greim, H. (1976) Formation of mutagenic N-nitroso compounds from the pesticides prometryne,
 6    dodine and carbaryl in the presence of nitrite at pH 1. Mutat Res 37(2-3):179-186.
 7
 8    El-Bayoumy, K. (1985) Effects of organoselenium compounds on induction of mouse forestomach tumors by
 9    benzo(a)pyrene.  Cancer Res 45:3631-3635.
10
11    El-Bayoumy, K; Hecht, SS; Hoffmann, D. (1982) Comparative tumor initiating activity on mouse skin of
12    6-nitrobenzo[a]pyrene, 6-nitrochrysene, 3-nitroperylene, 1-nitropyrene and their parent hydrocarbons.
13    Cancer Lett 16(3):333-337.
14
15    Emmett, EA; Bingham, EM; Barkley, W. (1981) A carcinogenic bioassay of certain roofing materials. Am J Ind
16    Med2(l):59-64.
17
18    Emura, M; Richter-Reichhelm, HB; Schneider, P; et al. (1980) Sensitivity of Syrian golden hamster fetal lung
19    cells to benzo[a]pyrene and other polycyclic hydrocarbons in vitro. Toxicology 17(2):149-155.
20
21    Emura, M; Mohr, U; Riebe, M; et al. (1987) Predisposition of cloned fetal hamster lung epithelial cells to
22    transformation by a precarcinogen, benzo(a)pyrene, using growth hormone supplementation and collagen
23    gel substratum. Cancer Res 47(4):1155-1160.
24
25    Epler, JL; Winton, W; Ho, T; et al. (1977) Comparative mutagenesis of quinolines. Mutat Res 39(3-4):285-
26    296.
27
28    Fahmy, MJ; Fahmy, OG. (1980) Altered control of gene activity in the soma by carcinogens. Mutat Res 72:165-
29    172.
30
31    Fang, JL; Lazarus, P. (2004) Correlation between the UDP-glucuronosyltransferase (UGT1A1) TATAA box
32    polymorphism and carcinogen detoxification phenotype: significantly decreased glucuronidating activity
33    against benzo(a)pyrene-7,8-dihydrodiol(-) in liver microsomes from subjects with the UGT1A128 variant.
34    Cancer Epidemiol Biomarkers Prev 13(1):102-109.
35
36    Fedorenko, Z; Yansheva, N. (1967) Experimental reproduction of tumors of the antral part of the stomach in
37    mice by administration of various dose of 3,4-benzpyrene.  Hyg Sanit 32(5):168-173.
38
39    Feldman, G; Remsen, J; Shinohara, K; et al. (1978) Excisability and persistence of benzo(a)pyrene DNA
40    adducts in epithelioid human lung cells. Nature 274(5673):796-798.
41
42    Feron, VJ; Kruysse, A. (1978) Effects of exposure to furfural vapour in hamsters simultaneously treated with
43    benzo[alpha] pyrene or diethylnitrosamine. Toxicology 11(2):127-144.
44
45    Feron, VJ; de Jong, D; Emmelot, P. (1973) Letter: Dose-response correlation for the induction of respiratory-
46    tract tumours in Syrian golden hamsters by intratracheal instillations of benzo(a)pyrene. Eur J Cancer
47    9(5):387-390.
48
49    Field,  E; Roe, F. (1965) Tumor promotion in the forestomach epithelium of mice by oral administration of
50    citrus oils. J Natl Cancer Inst 35(5):775-784.
51
52    Flowers, L; Bleczinski, WF; Burczynski, ME; et al. (1996)  Disposition and biological activity of
53    benzo[a]pyrene-7,8-dione. A genotoxic metabolite generated by dihydrodiol dehydrogenase. Biochemistry
54    35(42):13664-13672.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    7             DRAFT—DO NOT CITE  OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Flowers, L; Ohnishi, ST; Penning, TM. (1997) DNA strand scission by polycyclic aromatic hydrocarbon
 2    o-quinones: role of reactive oxygen species, Cu(II)/Cu(I) redox cycling, and o-semiquinone anion radicals.
 3    Biochemistry 36(28):8640-8648.
 4
 5    Flowers-Geary, L; Harvey, RG; Penning, TM. (1993) Cytotoxicity of polycyclic aromatic hydrocarbon o-
 6    quinones in rat and human hepatoma cells.  Chem Res Toxicol 6(3):252-260.
 7
 8    Flowers-Geary, L; Bleczinki, W; Harvey, RG; et al. (1996) Cytotoxicity and mutagenicity of polycyclic aromatic
 9    hydrocarbon ortho-quinones produced by dihydrodiol dehydrogenase.  Chem Biol Interact 99(l-3):55-72.
10
11    Foth, H; Kahl, R; Kahl, GF. (1988) Pharmacokinetics of low doses of benzo[a]pyrene in the rat. Food Chem
12    Toxicol 26(1):45-51.
13
14    Fowler, P; Whitwell, J; Jeffrey, L; et al. (2010) Cadmium chloride, benzo[a]pyrene and cyclophosphamide
15    tested in the in vitro mammalian cell micronucleus test (MNvit) in the human lymphoblastoid cell line TK6 at
16    Covance laboratories, Harrogate UK in support of OECD draft Test Guideline 487.  Mutat Res 702(2):171-174.
17
18    Friesen, MC; Demers, PA; Spinelli, JJ; etal. (2010) Chronic and acute effects of coal tar pitch exposure and
19    cardiopulmonary mortality among aluminum smelter workers. Am J Epidemiol 172(7):790-799.
20
21    Gangar, SC; Sandhir, R; Rai, DV; et al. (2006) Preventive effects of Azadirachta indica on benzo(a)pyrene-DNA
22    adduct formation in murine forestomach and hepatic tissues. Phytother Res 20:889-895.
23
24    Gao, M; Li, Y; Sun, Y; et al. (2011) Benzo[a]pyrene exposure increases toxic biomarkers and morphological
25    disorders in mouse cervix. Basic Clin Pharmacol Toxicol. 109(5):398-406.
26
27    Gao, N; Aidoo, A; Heflich, RH. (1991) Analysis of rat lymphocyte activation of benzo[a]pyrene,
28    2-acetylaminofluorene, and several of their metabolites to mutagenic and DNA-damaging species in vitro.
29    Teratog Carcinog Mutagen ll(2):65-76.
30
31    Garfon, G; Cosset, P; Garry, S; et al. (2001a) Pulmonary induction of proinflammatory mediators following the
32    rat exposure to benzo(a)pyrene-coated onto Fe203 particles. Toxicol Lett 121(2):107-117.
33
34    Garfon, G; Zerimech, F; Hannothiaux, M; et al. (2001b) Antioxidant defense disruption by polycyclic aromatic
35    hydrocarbons-coated onto Fe(2)0(3) particles in human lung cells (A549). Toxicology 166(3):129-137.
36
37    Garry, S; Nesslany, F; Aliouat, E; et al. (2003a) Assessment of genotoxic effect of benzo[a]pyrene in
38    endotracheally treated rat using the comet assay. Mutat Res 534:33-43.
39
40    Garry, S; Nesslany, F; Aliouat, E; et al. (2003b) Hematite (Fe(2)0(3)) enhances benzo[a]pyrene genotoxicity in
41    endotracheally treated rat, as determined by Comet Assay. Mutat Res 538:19-29.
42
43    Gelboin, HV. (1980) Benzo[alpha]pyrene metabolism, activation and carcinogenesis: role and regulation of
44    mixed-function oxidases and related enzymes. Physiol Rev 60(4):1107-1166.
45
46    Generoso, WM; Cain, KT; Krishna, M; et al. (1979) Genetic lesions induced by chemicals in spermatozoa and
47    spermatids of mice are repaired in the egg.  Proc Natl Acad Sci USA 76(l):435-437.
48
49    Generoso, WM; Cain, KT; Hellwig, CS; et al. (1982) Lack of association between induction of dominant-lethal
50    mutations and induction of heritable translocations with benzo[a]pyrene in postmeiotic germ cells of male
51    mice.  Mutat Res 94(1):155-163.
52
53    Gerde, P; Muggenburg, BA; Sabourin, PJ; et al. (1993b) Disposition of polycyclic aromatic hydrocarbons in the
54    respiratory tract of the beagle dog. II. The conducting airways. Toxicol Appl Pharmacol 121(2):319-327.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    8             DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Gerde, P; Muggenburg, BA; Lundborg, M; et al. (2001) The rapid alveolar absorption of diesel soot-adsorbed
 2    benzo[a]pyrene: bioavailability, metabolism and dosimetry of an inhaled particle-borne carcinogen.
 3    Carcinogenesis 22(5):741-749.
 4
 5    Ginsberg, GL; Atherholt, TB. (1989) Transport of DNA-adducting metabolites in mouse serum following
 6    benzo[a]pyrene administration. Carcinogenesis 10(4):673-679.
 7
 8    Glatt, HR; Oesch, F; Frigerio, A; et al. (1975) Epoxides metabolically produced from some known carcinogens
 9    and from some clinically used drugs. I. Differences in mutagenicity. Int J Cancer 16(5):787-797.
10
11    Glatt, H; Bucker, M; Platt, KL; et al.  (1985) Host-mediated mutagenicity experiments with benzo[a]pyrene and
12    two of its metabolites. MutatRes 156(3):163-169.
13
14    Glatt, H; Seidel, A; Ribeiro, 0; et al.  (1987) Metabolic activation to a mutagen of 3-hydroxy-trans-7,8-
15    dihydroxy-7,8-dihydrobenzo[a]pyrene, a secondary metabolite ofbenzo[a]pyrene. Carcinogenesis
16    8(11):1621-1627.
17
18    Greb, W; Strobel, R; Rohrborn, G. (1980) Transformation of BHK 21/CL13 cells by various polycyclic
19    aromatic hydrocarbons using the method of Styles. Toxicol Lett 7(2):143-148.
20
21    Grimmer, G; Brune, H; Deutsch-Wenzel, R; et al. (1983) On the contribution of polycyclic aromatic
22    hydrocarbons to the carcinogenic impact of automobile exhaust condensate evaluated by local application
23    onto mouse skin.  Cancer Lett 21(1):105-113.
24
25    Grimmer, G; Brune, H; Deutsch-Wenzel, R; et al. (1984) Contribution of polycyclic aromatic hydrocarbons to
26    the carcinogenic impact of gasoline engine exhaust condensate  evaluated by implantation into the lungs of
27    rats. J Natl Cancer Inst72(3):733-739.
28
29    Grimmer, G; Dettbam, G; Naujack, K; et al. (1994) Relationship between inhaled PAH and urinary excretion of
30    phenanthrene, pyrene and benzo(a)pyrene metabolites in coke plant workers. Polycyclic Aromat Compd
31    5:269-277.
32
33    Gross, P; Tolker, E; Babyak, MA; et  al. (1965) Experimental lung cancer in hamsters. Repetitive intratracheal
34    applications of two carcinogenic hydrocarbons. Arch Environ Health 11:59-65.
35
36    Grover, PL. (1986) Pathways involved in the metabolism and activation of polycyclic hydrocarbons.
37    Xenobiotica 16(10-11):915-931.
38
39    Gupta, RS; Goldstein, S. (1981) Mutagen testing in the human fibroblast diphtheria toxin resistance (HF Dipr)
40    system.  Evaluation of short-term tests for carcinogens: report of the international collaborative programs.
41    Prog MutatRes 1:614-625.
42
43    Habs, M; Schmahl, D; Misfeld, J. (1980) Local carcinogenicity of some environmentally relevant polycyclic
44    aromatic hydrocarbons after lifelong topical application to mouse skin. Arch Geschwulstforsch 50(3):266-
45    274.
46
47    Habs, M; Jahn, SA; Schmahl, D. (1984) Carcinogenic activity of condensate from coloquint seeds (Citrullus
48    colocynthis) after chronic epicutaneous administration to mice.  J Cancer Res Clin Oncol 108(1): 154-156.
49
50    Hackman, P; Hou, SM; Nyberg, F; et al. (2000) Mutational spectra at the hypoxanthine-guanine
51    phosphoribosyltransferase (HPRT) locus in T-lymphocytes of nonsmoking and smoking lung cancer patients.
52    Mutat Res 468(1):45-61.
53
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    9             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Hakura, A; Tsutsui, Y; Sonoda, J; et al. (1998) Comparison between in vivo mutagenicity and carcinogenicity in
 2    multiple organs by benzo[a]pyrene in the lacZ transgenic mouse (Muta Mouse). Mutat Res 398(1-2):123-
 3    130.
 4
 5    Hall, M; Grover, PL. (1988) Stereoselective aspects of the metabolic activation of benzo[a]pyrene by human
 6    skin in vitro. Chem Biol Interact 64(3):281-296.
 7
 8    Hanelt, S; Helbig, R; Hartmann, A; et al. (1997) A comparative investigation of DNA adducts, DNA strand
 9    breaks and gene mutations induced by benzo[a]pyrene and (+/-)-anti-benzo[a]pyrene-7,8-diol 9,10-oxide in
10    cultured human cells.  Mutat Res 390:179-188.
11
12    Hara, T; Nishikawa, T; Sui, H; et al. (2007) In vivo photochemical skin micronucleus test using a sunlight
13    simulator: detection of 8-methoxypsoralen and benzo[a]pyrene in hairless mice. Mutat Res 631:1-8.
14
15    Harper, BL; Ramanujam, VM; Legator, MS. (1989)  Micronucleus formation by benzene, cyclophosphamide,
16    benzo(a)pyrene, and benzidine in male, female, pregnant female, and fetal mice. Teratog Carcinog Mutagen
17    9(4):239-252.
18
19    He, SL; Baker, R. (1991) Micronuclei in mouse skin cells following  in vivo exposure to benzo[a]pyrene,
20    7,12-dimethylbenz[a]anthracene, chrysene, pyrene and urethane.  Environ Mol Mutagen 17(3):163-168.
21
22    Hecht,  SS; Grabowski, W; Groth, K. (1979) Analysis of faeces for benzo[a]pyrene after consumption of
23    charcoal-broiled beef by rats and humans.  Food Cosmet Toxicol 17(3):223-227.
24
25    Heidelberger, C; Freeman, AE; Pienta, RJ; et al. (1983) Cell transformation by chemical agents—a review and
26    analysis of the literature. A report of the U.S. Environmental Protection Agency Gene-Tox Program. Mutat Res
27    114(3):283-385.
28
29    Hemminki, K; Dickey, C; Karlsson, S; et al. (1997) Aromatic DNA adducts in foundry workers in relation to
30    exposure, life style and CYP1A1 and glutathione transferase Ml genotype. Carcinogenesis 18(2):345-350.
31
32    Henry, MC; Port, CD; Bates, RR; et al. (1973) Respiratory tract tumors in hamsters induced by
33    benzo(a)pyrene. Cancer Res 33(7):1585-1592.
34
35    Hermann, M. (1981) Synergistic effects of individual polycyclic aromatic hydrocarbons on the mutagenicity of
36    their mixtures. Mutat Res 90(4):399-409.
37
38    Herrold, KM; Dunham, LJ. (1962) Induction of carcinoma and papilloma of the Syrian hamster by
39    intratracheal instillation of benzo[a]-pyrene.  J Natl Cancer Inst 28:467-491.
40
41    Higginbotham, S; RamaKrishna, NV; Johansson, SL; et al. (1993) Tumor-initiating activity and carcinogenicity
42    of dibenzo[a,l]pyrene versus 7,12-dimethylbenz[a]anthracene and benzo [ajpyrene at low doses in mouse
43    skin. Carcinogenesis 14(5):875-878.
44
45    Hirom, PC; Chipman, JK; Millburn, P; et al. (1983) Enterohepatic circulation of the aromatic hydrocarbons
46    benzo [ajpyrene and naphthalene. In: Extrahepatic drug metabolism and chemical Carcinogenesis.
47    Amsterdam: Elsvier Science Publishers; pp. 275-281.
48
49    Hoffmann, D; Rathkamp, G; Nesnow, S; et al. (1972) Fluoranthenes: quantitative determination in cigarette
50    smoke, formation by pyrolysis, and tumor-initiating activity. J Natl Cancer Inst 49(4):1165-1175.
51
52    Hood, DB;  Nayyar, T; Ramesh, A; et al. (2000) Modulation in the developmental expression profile of Spl
53    subsequent to transplacental exposure of fetal rats to desorbed benzo [ajpyrene following maternal
54    inhalation. Inhal Toxicol 12(6):511-535.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    10            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Hu, Z; Wells, PG. (2004) Human interindividual variation in lymphocyte UDP-glucuronosyltransferases as a
 2    determinant of in vitro benzo[a]pyrene covalent binding and cytotoxicity. Toxicol Sci 78(1):32-40.
 o
 4    Huberman, E; Sachs, L; Yang, SK; et al. (1976) Identification of mutagenic metabolites of benzo(a)pyrene in
 5    mammalian cells. Proc Natl Acad Sci U S A 73:607-611.
 6
 7    Huh, N; Nemoto, N; Utakoji, T. (1982) Metabolic activation of benzo[a]pyrene, aflatoxin Bl, and
 8    dimethylnitrosamine by a human hepatoma cell line. Mutat Res 94:339-348.
 9
10    Husgafvel-Pursiainen, K; Sorsa, M; Moller, M; et al. (1986) Genotoxicity and polynuclear aromatic
11    hydrocarbon analysis of environmental tobacco smoke samples from restaurants. Mutagenesis 1(4):287-
12    292.
13
14    IARC (International Agency for Research on Cancer). (1973) IARC monographs on the evaluation of
15    carcinogenic risks to humans. Vol. 3. Certain polycyclic aromatic hydrocarbons and heterocyclic compounds.
16    Lyon, France:  International Agency for Research on Cancer
17
18    IARC (International Agency for Research on Cancer). (1983) IARC monographs on the evaluation of
19    carcinogenic risks to humans. Vol. 32. Polynuclear aromatic hydrocarbons, Part 1, chemical environmental
20    and experimental data. Lyon, France: International Agency for Research on Cancer.
21
22    IARC (International Agency for Research on Cancer). (2010) IARC monographs on the evaluation of
23    carcinogenic risks to humans. Vol. 92. Some non-heterocyclic aromatic hydrocarbons and some related
24    exposures: Lyon, France. International Agency for Research on Cancer.
25
26    IARC (International Agency for Research on Cancer). (2012) IARC monographs on the evaluation of
27    carcinogenic risk to humans. Vol.  100F. A review of human carcinogens: chemical agents and related
28    occupations: Lyon, France. International Agency for Research on Cancer.
29
30    Inyang, F; Ramesh, A; Kopsombut, P; et al. (2003) Disruption of testicular steroidogenesis and epididymal
31    function by inhaled benzo(a)pyrene. Reprod Toxicol 17(5):527-537.
32
33    Juhl, HJ; Schurer, CC; Bartram, CR; et al. (1978) Retinoids induce sister-chromatid exchanges in human diploid
34    fibroblasts. Mutat Res 58(2-3):317-320.
35
36    Jules, GE; Pratap, S; Ramesh, A; et al. (2011 )In utero exposure to benzo(a)pyrene predisposes offspring to
37    cardiovascular dysfunction in later-life. Toxicology 16;295(l-3):56-67.
38
39    Kao, J; Patterson, FK; Hall, J. (1985) Skin penetration and metabolism of topically applied chemicals in six
40    mammalian species, including man: an in vitro study with benzo[a]pyrene and testosterone. Toxicol Appl
41    Pharmacol 81(3 Pt 1):502-516.
42
43    Kawamura, Y; Kamata, E; Ogawa, Y; et al. (1988) The effect of various foods on the intestinal absorption of
44    benzo-a-pyrene in rats. J Food Hyg Soc Jpn 29(l):21-25.
45
46    Ketkar, M; Reznik, G; Schneider, P; et al. (1978) Investigations on the carcinogenic burden by air pollution in
47    man. Intratracheal instillation studies with benzo(a)pyrene in bovine serum albumin in Syrian hamsters.
48    Cancer Lett 4(4):235-239.
49
50    Kiefer, F; Cumpelik, 0; Wiebel, FJ. (1988) Metabolism and cytotoxicity of benzo(a)pyrene in the human lung
51    tumour cell line NCI-H322. Xenobiotical8(6):747-755.
52
53    Kim, IY; Hyun, CK. (2006) Comparative evaluation of the alkaline comet assay with the micronucleus test for
54    genotoxicity monitoring using aquatic organisms. Ecotoxicol Environ Saf 64:288-297.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    11            DRAFT—DO NOT CITE  OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Kleiner, HE; Vulimiri, SV; Hatten, WB; et al. (2004) Role of cytochrome p4501 family members in the
 2    metabolic activation of poly cyclic aromatic hydrocarbons in mouse epidermis. Chem Res Toxicol
 3    17(12):1667-1674.
 4
 5    Knuckles, ME; Inyang, F; Ramesh, A. (2001) Acute and subchronic oral toxicities of benzo[a]pyrene in F-344
 6    rats. Toxicol Sci 61(2):382-388.
 7
 8    Kobayashi, N. (1975) Production of respiratory tract tumors in hamsters by benzo(a)pyrene. Gann
 9    66(3):311-315.
10
11    Kondraganti, SR; Fernandez-Salguero, P; Gonzalez, FJ; et al. (2003) Polycyclic aromatic hydrocarbon-
12    inducible DNA adducts: evidence by 32P-postlabeling and use of knockout mice for Ah receptor-independent
13    mechanisms of metabolic activation in vivo. Int] Cancer 103(1):5-11.
14
15    Koratkar, R; Das, UN; Sagar, PS; et al. (1993) Prostacyclin is a potent anti-mutagen. Prostaglandins Leukot
16    Essent Fatty Acids 48(2):175-184.
17
18    Kristensen, P; Eilertsen,  E; Einarsdottir, E; et al. (1995) Fertility in mice after prenatal exposure to
19    benzo[a]pyrene and inorganic lead. Environ Health Perspectl03(6):588-590.
20
21    Kroese, ED; Muller, JJA; Mohn, GR; et al. (2001) Tumorigenic effects in Wistar rats orally administered
22    benzo(a)pyrene for two  years: Implications for human cancer risks associated with oral exposure to
23    polycyclic aromatic hydrocarbons. Bilthoven, Netherlands:  RIVM 658603.
24
25    Kulka, U; Doehmer, J; Glatt, HR; et al. (1993a) Cytogenetic effects of promutagens in genetically engineered
26    V79 Chinese hamster cells expressing cytochromes P450. Eur J Pharmacol 228:299-304.
27
28    Kulka, U; Paul, D; Bauchinger, M. (1993b) Development of short-term mutagenicity test systems in vitro:
29    metabolic activation of indirectly acting mutagens by three immortal rat hepatocyte lines. Mutagenesis 8:193-
30    197.
31
32    Kunstler, K. (1983) Failure to induce tumors by intratracheal instillation of automobile exhaust condensate
33    and fractions thereof in Syrian golden hamsters. Cancer Lett 18(1):105-108.
34
35    LaVoie, E; Bedenko, V; Hirota, N; et al., eds. 1979. A comparison of the mutagenicity, tumor-initiating activity
36    and complete carcinogen!city of polynuclear aromatic hydrocarbons. Edited by Jones, PW.Leber, P,
37    Polynuclear Aromatic Hydrocarbons. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.
38
39    LaVoie, EJ; Amin, S; Hecht, SS; et al. (1982) Tumour initiating activity of dihydrodiols of benzo[b]fluoranthene,
40    benzo[j]fluoranthene, andbenzo[k]fluoranthene. Carcinogenesis 3(l):49-52.
41
42    Leadon, SA; Stampfer, MR; Bartley, J. (1988) Production of oxidative DNA damage during the metabolic
43    activation of benzo[a]pyrene in human mammary epithelial cells correlates with cell killing. Proc Natl Acad
44    Sci USA 85:4365-4368.
45
46    LeBoeuf, RA; Kerckaert,  GA; Aardema, MJ; et al. (1996) The pH 6.7 Syrian hamster embryo cell transformation
47    assay for assessing the carcinogenic potential of chemicals.  Mutat Res 356(1):85-127.
48
49    Lee, H; Lin, JY. (1988) Antimutagenic activity of extracts from anticancer drugs in Chinese medicine. Mutat
50    Res 204:229-234.
51
52    Leng, SG; Zheng, YX; Pan, ZF; et al. (2004) [A study on the inherited susceptibility of chromosomal damage in
53    peripheral blood lymphocytes among coke oven workers]. ZhonghuaYu FangYiXue ZaZhi 38(2):94-98.
54
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    12           DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Levin, W; Wood, AW; Wislocki, PG; et al. (1977) Carcinogenicity of benzo-ring derivatives of benzo(a)pyrene
 2    on mouse skin. Cancer Res 37(9):3356-3361.
 o
 4    Levin, W; Wood, A; Chang, R; et al. (1982) Oxidative metabolism of polycyclic aromatic hydrocarbons to
 5    ultimate carcinogens. Drug Metab Rev 13(4):555-580.
 6
 7    Liao, KH. (2004) Reaction network model for the prediction of mammalian metabolism of benzo(a)pyrene.
 8    Unpublished Ph.D. Dissertation, Colorado State University, Fort Collins, CO, Fall 2004.
 9
10    Likhachev, AJ; Beniashvili, DS; Bykov, VJ; et al. (1992) Relevance of quantitation of benzo(a)pyrene
11    metabolites in animal excretes to evaluate individual human cancer risk.  Prog Clin Biol Res 374:435-452.
12
13    Lin, P; Hsueh, YM; Ko, JL; et al. (2003) Analysis of NQ01, GSTP1, and MnSOD genetic polymorphisms on lung
14    cancer risk in Taiwan. Lung Cancer 40(2):123-129.
15
16    Lo Jacono, F; Stecca, C; Duverger, M. (1992) Mutagenic activation of benzo[a]pyrene by human red blood cells.
17    MutatRes268:21-26.
18
19    Lubet, RA; Kiss, E; Gallagher, MM; et al. (1983) Induction of neoplastic transformation and DNA single-strand
20    breaks in C3H/10T1/2 clone 8 cells by polycyclic hydrocarbons and alkylating agents. J Natl Cancer Inst
21    71(5):991-997.
22
23    Ma, Q; Lu, AY. (2007) CYP1A induction and human risk assessment: an evolving tale of in vitro and in vivo
24    studies.  Drug Metab Dispos 35(7):1009-1016.
25
26    MacKenzie, KM; Angevine, DM. (1981) Infertility in mice exposed in utero to benzo(a)pyrene. Biol Reprod
27    24(1):183-191.
28
29    Mager, R; Huberman, E; Yang, SK; et al. (1977) Transformation of normal hamster cells by benzo(a)pyrene
30    diol-epoxide. Int.] Cancer 19(6):814-817.
31
32    Mallet, WG; Mosebrook, DR; Trush, MA. (1991) Activation of (+-)-trans-7,8-dihydroxy-
33    7,8-dihydrobenzo[a]pyrene to diolepoxides by human polymorphonuclear leukocytes or myeloperoxidase.
34    Carcinogenesisl2(3):521-524.
35
36    Mamber, SW; Bryson, V; Katz, SE. (1983) The Escherichia coli WP2/WP100 rec assay for detection of
37    potential chemical carcinogens. Mutat Res 119:135-144.
38
39    Mane, SS; Purnell, DM; Hsu, 1C. (1990) Genotoxic effects of five polycyclic aromatic hydrocarbons in human
40    and rat mammary epithelial cells. Environ Mol Mutagen 15(2):78-82.
41
42    Marnett, LJ. (1990) Prostaglandin synthase-mediated metabolism of carcinogens and a potential role for
43    peroxyl radicals as reactive intermediates.  Environ Health Perspect 88:5-12.
44
45    Martin, CN; McDermid, AC; Garner, RC. (1978) Testing of known carcinogens and noncarcinogens for their
46    ability to induce unscheduled DNA synthesis in HeLa cells. Cancer Res 38(8):2621-2627.
47
48    Matsuoka, A; Hayashi, M; Ishidate, M, Jr. (1979) Chromosomal aberration tests on 29 chemicals combined
49    with S9 mix in vitro.  Mutat Res 66(3):277-290.
50
51    Matsuoka, A; Matsuura, K; Sakamoto, H; et al. (1998) Spindle disturbances induced by benzo[a]pyrene and
52    7,12-dimethylbenz[a]anthracene in a Chinese hamster cell line (V79-MZ) and the stability of the numerical
53    chromosome aberrations that follow. Mutat Res 419(1-3):1-12.
54
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    13            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Matsuoka, A; Matsuura, K; Sakamoto, H; et al. (1999) A proposal for a simple way to distinguish aneugens
 2    from clastogens in the in vitro micronucleus test.  Mutagenesis 14(4):385-389.
 o
 4    McCabe, DP; Flynn, EJ. (1990) Deposition of low dose benzo(a)pyrene into fetal tissue: influence of protein
 5    binding. Teratology 41(l):85-95.
 6
 7    McCallister MM; Maguire M; Ramesh A; et al. (2008) Prenatal exposure to benzo(a)pyrene impairs later-life
 8    cortical neuronal function. Neurotoxicology 29(5):846-854.
 9
10    McCann, J; Spingarn, NE; Kobori, J; et al. (1975) Detection of carcinogens as mutagens: bacterial tester strains
11    with R factor plasmids. Proc Natl Acad Sci USA 72(3):979-983.
12
13    Melikian, AA; Sun,  P; Prokopczyk, B; et al. (1999) Identification of benzo[a]pyrene metabolites in cervical
14    mucus and DNA adducts in cervical tissues in humans by gas chromatography-mass spectrometry. Cancer
15    Lettl46(2):127-134.
16
17    Meng, F; Knapp, GW; Green, T; etal. (2010) K-Ras mutant fraction in A/J mouse lung increases as a function of
18    benzo[a]pyrene dose.  Environ Mol Mutagen. 51(2):146-155.
19
20    Mensing, T; Marczynski, B; Engelhardt, B; et al. (2005) DNA adduct formation of benzo[a]pyrene in white
21    blood cells of workers exposed to polycyclic aromatic hydrocarbons.  Int] Hyg Environ Health 208:173-178.
22
23    Merk, HF; Mukhtar, H; Kaufmann, I; et al. (1987) Human hair follicle benzo[a]pyrene and benzo[a]pyrene 7,8-
24    diol metabolism: effect of exposure to a coal tar-containing shampoo. J Invest Dermatol 88(l):71-76.
25
26    Mersch-Sundermann, V; Mochayedi, S; Kevekordes, S. (1992)  Genotoxicity of polycyclic aromatic
27    hydrocarbons in Escherichia coli PQ37. Mutat Res 278:1-9.
28
29    Michalopoulos, G; Sattler, GL; O'Connor, L; et al. (1978) Unscheduled DNA synthesis induced by
30    procarcinogens in  suspensions and primary cultures of hepatocytes on collagen membranes. Cancer Res
31    38:1866-1871.
32
33    Midgette, AS; Baron, JA. (1990) Cigarette smoking and the risk of natural menopause. Epidemiology 1:474-
34    479.
35
36    Miller, KP; Ramos,  KS. (2001) Impact of cellular metabolism on the biological effects of benzo[a]pyrene and
37    related hydrocarbons. Drug Metab Rev 33(l):l-35.
38
39    Milo, GE; Blakeslee, J; Yohn, DS; et al. (1978) Biochemical activation of aryl hydrocarbon hydroxylase activity,
40    cellular distribution of polynuclear hydrocarbon metabolites, and DNA damage by polynuclear hydrocarbon
41    products in human cells in vitro. Cancer Res 38(6):1638-1644.
42
43    Mirsalis, JC; Tyson, CK; Butterworth, BE.  (1982) Detection of genotoxic carcinogens in the in vivo-in vitro
44    hepatocyte DNA repair assay. Environ Mutagen 4(5):553-562.
45
46    Mirvish, SS; Ghadirian, P;  Wallcave, L;  et al. (1981) Effect of diet on fecal excretion and gastrointestinal tract
47    distribution of unmetabolized benzo(a)pyrene and 3- methylcholanthrene when these compounds are
48    administered orally to hamsters. Cancer Res 41(6):2289-2293.
49
50    Mishra, NK; Wilson, CM; Pant, KJ; et al. (1978)  Simultaneous determination of cellular mutagenesis and
51    transformation by chemical carcinogens  in Fischer rat embryo cells. J Toxicol Environ Health 4:79-91.
52
53    Mohamed, ESA; Song, WH; Oh, SA; et al. (2010) The transgenerational impact of benzo(a)pyrene on murine
54    male fertility. HumReprod25(10):2427-2433.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    14           DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Mohr, U. (1971) Diethylnitrosamine-induced carcinogenesis. Diaplacental effect.  Fortschr Med 89(6):251-
 2    253. (German)
 o
 4    Moir, D. (1999) Physiological modeling of benzo(a)pyrene pharmacokinetics in the rat. In: Salem, H; Katz, SA;
 5    eds.  Toxicity assessment alternatives: methods, issues, opportunities.  Totowa, NJ: Humana Press Inc., pp. 79-
 6    95.
 7
 8    Moir, D; Viau, A; Chu, I; et al. (1998) Pharmacokinetics of benzo[a]pyrene in the rat. J Toxicol Environ Health
 9    A53(7):507-530.
10
11    Monteith, DK; Novotny, A; Michalopoulos, G; et al. (1987) Metabolism of benzo[a]pyrene in primary cultures
12    of human hepatocytes: dose-response over a four-log range. Carcinogenesis 8(7):983-988.
13
14    Moore, LE; Wiencke, JK; Bates, MN; et al. (2004) Investigation of genetic polymorphisms and smoking in a
15    bladder cancer case-control study in Argentina. Cancer Lett211(2):199-207.
16
17    Mori, M; Kobayashi, H; Sugiyama, C; et al. (1999) Induction of unscheduled DNA synthesis in hairless mouse
18    epidermis by skin carcinogens. J Toxicol Sci 24(3):217-226.
19
20    Morse, MA; Carlson, GP. (1985) Distribution and macromolecular binding of benzo[a]pyrene in SENCAR and
21    BALB/c mice following topical and oral administration. J Toxicol Environ Health  16(2):263-276.
22
23    Mullaart, E; Buytenhek, M; Brouwer, A; et al. (1989) Genotoxic effects of intragastrically administered
24    benzo[a]pyrene in rat liver and intestinal cells. Carcinogenesis 10(2):393-395.
25
26    National Toxicology Program (NTP). (2011) Report on carcinogens, Twelfth Edition. U.S. Department of
27    Health and Human Services.
28
29    N'Diaye, M; Le, FE; Lagadic-Gossmann, D; et al. (2006) Aryl hydrocarbon receptor- and calcium-dependent
30    induction of the chemokine CCL1 by the environmental contaminant benzo[a]pyrene. J Biol Chem
31    281(29):19906-19915.
32
33    Neal, J; Rigdon, H. (1967) Gastric tumors in mice fed benzo[a]pyrene: a quantitative study. Tex Rep Biol Med
34    25(4):553-557.
35
36    Neal, MS; Zhu, J; Holloway, AC; et al. (2007) Follicle growth is inhibited by benzo-[a]-pyrene, at
37    concentrations representative of human exposure, in an isolated rat follicle culture assay.  Hum Reprod
38    22:961-967.
39
40    Neal, MS; Zhu, J; Foster, WG. (2008) Quantification of benzo[a]pyrene and other PAHs in the serum and
41    follicular fluid of smokers versus non-smokers. Reprod Toxicol 25(1):100-106.
42
43    Nebert, DW; Puga, A; Vasiliou, V. (1993) Role of the Ah receptor and the dioxin-inducible [Ah] gene battery in
44    toxicity, cancer, and signal transduction. Ann N Y Acad Sci 685:624-640.
45
46    Nesnow, S; Triplett, LL; Slaga, TJ. (1983) Mouse skin tumor initiation-promotion and complete carcinogenesis
47    bioassays: mechanisms and biological activities of emission samples. Environ Health Perspect 47:255-268.
48
49    Nesnow, S; Davis, C; Nelson, G; et al. (1997) Comparison of the morphological transforming activities of
50    dibenzo[a,l]pyrene and benzo[a]pyrene in C3H10T1/2CL8 cells and characterization of the
51    dibenzo[a,l]pyrene-DNAadducts. Carcinogenesis 18(10):1973-1978.
52
53    Nesnow, S; Davis, C; Nelson, GB; et al. (2002) Comparison of the genotoxic activities of the K-region
54    dihydrodiol of benzo[a]pyrene with benzo [ajpyrene in mammalian cells: morphological cell transformation;
55    DNA damage; and stable covalent DNA adducts. Mutat Res 521(1-2):91-102.

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                    15            DRAFT—DO NOT CITE  OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene
 2    Neubert, D; Tapken, S. (1988) Transfer of benzo(a)pyrene into mouse embryos and fetuses. Arch Toxicol
 3    62(2-3):236-239.
 4
 5    Nishikawa, T; Nakamura, T; Fukushima, A; et al. (2005) Further evaluation of the skin micronucleus test:
 6    results obtained using 10 polycyclic aromatic hydrocarbons. Mutat Res 588(l):58-63.
 7
 8    Norpoth, K; Kemena, A; Jacob, J; et al. (1984) The influence of 18 environmentally relevant polycyclic
 9    aromatic hydrocarbons and Clophen A50, as liver monooxygenase inducers, on the mutagenic activity of
10    benz[a]anthracene in the Ames test. Carcinogenesis 5(6):747-752.
11
12    Nwagbara, 0; Darling-Reed, SF; Tucker, A; et al. (2007) Induction of cell death, DNA strand breaks, and cell
13    cycle arrest in DU145 human prostate carcinoma cell line by benzo[a]pyrene and benzo[a]pyrene-7,8-diol-
14    9,10-epoxide. IntJ Environ Res Public Health 4(1):10-14.
15
16    Obana, H; Hori, S; Kashimoto, T; et al. (1981) Polycyclic aromatic hydrocarbons in human fat and liver.  Bull
17    Environ Contam Toxicol 2 7(1): 2 3-2 7.
18
19    Obermeier, J; Frohberg, H. (1977) Mutagenicity studies with praziquantel, a new anthelmintic drug: tissue-
20    ,host, and urine-mediated mutagenicity assays. Arch Toxicol 38(3):149-161.
21
22    O'Donovan, MR. (1990) Mutation assays of ethyl methanesulphonate, benzidine and benzo[a]pyrene using
23    Chinese hamster V79 cells. Mutagenesis 5 Suppl:9-13.
24
25    Oesch, F; Bentley, P; Glatt, HR. (1976) Prevention of benzo(a)pyrene-induced mutagenicity by homogeneous
26    epoxide hydratase. Int J Cancer 18(4):448-452.
27
28    Okey, AB; Riddick, DS; Harper, PA. (1994) Molecular biology of the aromatic hydrocarbon (dioxin) receptor.
29    Trends Pharmacol Sci 15(7):226-232.
30
31    O'Neill, IK; Goldberg, MT; el Ghissassi, F; et al. (1991) Dietary fibre, fat and beef modulation of colonic nuclear
32    aberrations and microcapsule-trapped gastrointestinal metabolites of benzo[a]pyrene-treated C57/B6 mice
33    consuming human diets. Carcinogenesis 12(2):175-180.
34
3 5    Oueslati, R; Alexandrov, K; Chouikha, M; et al. (1992) Formation and persistence of DNA adducts in epidermal
36    and dermal mouse skin exposed to benzo(a)pyrene in vivo. In Vivo 6(2):231-235.
37
38    Pahlman, R; Pelkonen, 0. (1987) Mutagenicity studies  of different polycyclic aromatic hydrocarbons: the
39    significance of enzymatic factors and molecular structure. Carcinogenesis 8(6):773-778.
40
41    Park, SJ; Zhao, H; Spitz, MR; et al. (2003) An association between NQ01 genetic polymorphism and risk of
42    bladder cancer.  Mutat Res 536(1-2):131-137.
43
44    Pastorelli, R;  Guanci, M; Cerri, A; et al. (1998) Impact of inherited polymorphisms in glutathione S-transferase
45    Ml, microsomal epoxide hydrolase, cytochrome P450  enzymes on DNA, and blood protein adducts of
46    benzo(a)pyrene-diolepoxide. Cancer Epidemiol Biomarkers Prev 7(8):703-709.
47
48    Penning, TM; Burczynski, ME; Hung, CF; et al. (1999) Dihydrodiol  dehydrogenases and polycyclic aromatic
49    hydrocarbon activation: generation of reactive and redox active o-quinones. Chem Res Toxicol 12(1): 1-18.
50
51    Pereira, MA; McMillan, L; Kaur, P; et al. (1982) Effect of benzo[a]pyrene on sister-chromatid exchange in fetal
52    hamster liver exposed in utero. Mutat Res 105:343-347.
53
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    16           DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Perera, FP; Tang, D; Rauh, V; et al. (2005a) Relationships among polycyclic aromatic hydrocarbon-DNA
 2    adducts, proximity to the World Trade Center, and effects on fetal growth. Environ Health Perspect
 3    113(8):1062-1067.
 4
 5    Perera, FP; Rauh, V; Whyatt, RM; et al. (2005b) A summary of recent findings on birth outcomes and
 6    developmental effects of prenatal ETS, PAH, and pesticide exposures. Neurotoxicology 26(4):573-587.
 7
 8    Peterson, AR; Landolph, JR; Peterson, H; et al. (1981) Oncogenic transformation and mutation of C3H/10T
 9    1/2 clone 8 mouse embryo fibroblasts by alkylating agents. Cancer Res 41:3095-3099.
10
11    Petridou-Fischer, J; Whaley, SL; Dahl, AR.  (1988) In vivo metabolism of nasally instilled benzo[a]pyrene in
12    dogs and monkeys.  Toxicology 48(1):31-40.
13
14    Phillipson, CE; loannides, C. (1989) Metabolic activation of polycyclic aromatic hydrocarbons to mutagens in
15    the Ames test by various animal species including man. Mutat Res 211:147-151.
16
17    Pitts, JN, Jr.; Van Cauwenberghe, KA; Grosjean, D; et al. (1978) Atmospheric reactions of polycyclic aromatic
18    hydrocarbons: facile formation of mutagenic nitro derivatives. Science 202(4367):515-519.
19
20    Poel, WE. (1959) Effect of carcinogenic  dosage and duration of exposure on skin-tumor induction in mice. J
21    Natl Cancer Inst22(l):19-43.
22
23    Poel, WE. (1960) Skin as a test site for the bioassay of carcinogens and carcinogen precursors. J Natl Cancer
24    InstMonogr 10:611-631.
25
26    Popescu, NC; Turnbull, D; DiPaolo, JA. (1977) Sister chromatid exchange and chromosome aberration analysis
27    with the use of several carcinogens and noncarcinogens.  J Natl Cancer Inst 59(l):289-293.
28
29    Potter, D; Booth, ED; Brandt, HC; et al. (1999) Studies on  the dermal and systemic bioavailability of polycyclic
30    aromatic compounds in high viscosity oil  products. Arch Toxicol 73(3):129-140.
31
32    Quarles, JM; Sega, MW; Schenley, CK; et al. (1979) Transformation of hamster fetal cells by nitrosated
33    pesticides in a transplacental assay. Cancer Res 39(ll):4525-4533.
34
35    Ramesh, A; Knuckles, ME. (2006) Dose-dependent benzo(a)pyrene [B(a)P]-DNA adduct levels and persistence
36    in F-344 rats following subchronic dietary exposure to B(a)P. Cancer Lett 240:268-278.
37
38    Ramesh, A; Greenwood, M; Inyang, F; et al. (2001a) Toxicokinetics of inhaled benzo[a]pyrene: plasma and
39    lung bioavailability. Inhal Toxicol 13(6):533-555.
40
41    Ramesh, A; Inyang, F; Hood, DB; et al. (2001b) Metabolism, bioavailability, and toxicokinetics of
42    benzo(alpha)pyrene in F-344 rats following oral administration. Exp Toxicol Pathol 53(4):275-290.
43
44    Ramesh, A; Greenwood, M; Inyang, F; et al. (2003) Aryl hydrocarbon hydroxylase (AHH) and benzo(a)pyrene
45    (BaP) metabolite in F-344 rats subchronically exposed to inhaled BaP. Toxicologist 72(S-1):325.
46
47    Ramesh, A; Inyang, F; Lunstra, DD; etal. (2008) Alteration of fertility endpoints in adult male F-344 rats by
48    subchronic exposure to inhaled benzo(a)pyrene. Exp Toxicol Pathol 60(4-5):269-280.
49
50    Rao, KP; Nandan, BD. (1990) Modification of benzo(a)pyrene induced chromosomal damage in mouse bone
51    marrow by vitamin A. Bull Environ Contam Toxicol 45(6):829-832.
52
53    Rastetter, WH; B., NJR; Russo-Rodriguez, S; et al. (1982) Fluoranthene: Synthesis and mutagenicity of four
54    diol epoxides. J Org Chem 47:4873-4878.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    17            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Raveh, D; Slaga, TJ; Huberman, E. (1982) Cell-mediated mutagenesis and tumor-initiating activity of the
 2    ubiquitous polycyclic hydrocarbon, cyclopenta[c,d]pyrene.  Carcinogenesis 3(7):763-766.
 O
 4    Reddy, MV; Gupta, RC; Randerath, E; et al. (1984) 32P-postlabeling test for covalent DNA binding of chemicals
 5    in vivo: application to a variety of aromatic carcinogens and methylating agents. Carcinogenesis 5:231-243.
 6
 7    Reznik-Schuller, H; Mohr, U. (1974) Investigations on the carcinogenic burden by air pollution in man. IX.
 8    Early pathological alterations of the bronchial epithelium in Syrian golden hamsters after intratracheal
 9    instillation of benzo (a) pyrene. 1. Morphological studies from semi-thin sections. Zentralbl Bakteriol [Orig B]
10    159(5-6):493-502.
11
12    Rice, JE; Makowski, GS; Hosted, TJ, Jr.; et al. (1985) Methylene-bridged bay region chrysene and phenanthrene
13    derivatives and their keto-analogs: mutagenicity in Salmonella typhimurium and tumor-initiating activity on
14    mouse skin. Cancer Lett 27(2):199-206.
15
16    Rigdon, RH; Neal, J. (1965) Effects of feeding benzo(a)pyrene on fertility, embryos, and young mice. J Natl
17    Cancer Inst 34:297-305.
18
19    Rigdon, RH; Neal, J. (1966) Gastric carcinomas and pulmonary adenomas in mice fed benzo (a) pyrene. Tex
20    RepBiolMed24(2):195-207.
21
22    Rigdon, RH; Neal, J. (1969) Relationship of leukemia to lung and stomach tumors in mice fed benzo (a)
23    pyrene. ProcSoc Exp Biol Medl30(l):146-148.
24
25    Rigdon, RH; Rennels, EG. (1964) Effect of feeding benzpyrene on reproduction in the rat. Experientia
26    20(4):224-226.
27
28    Robertson,  IG; Guthenberg, C; Mannervik, B; et al. (1986)  Differences in stereoselectivity and catalytic
29    efficiency of three human glutathione transferases in the conjugation of glutathione with 7 beta,8 alpha-
30    dihydroxy-9 alpha,10 alpha-oxy-7,8,9,10-tetrahydrobenzo(a)pyrene.  Cancer Res 46(5):2220-2224.
31
32    Robinson, DE; Mitchell, AD. (1981) Unscheduled DNA synthesis response of human fibroblasts, WI-38 cells, to
33    20 coded chemicals. Evaluation of Short-Term Tests for Carcinogens. Report of the International
34    Collaborative Program. In: de Serres, FJ.Ashby, J eds. Progress in Mutation Research, Amsterdam: Elsevier,
35    517-527.
36
37    Robinson, M; Laurie, RD; Bull, RJ; et al. (1987) Carcinogenic effects in A/J mice of particulate of a coal tar paint
38    used in potable water systems. Cancer Lett 34(l):49-54.
39
40    Rocchi, P; Ferreri, AM; Borgia, R; et al. (1980) Polycyclic hydrocarbons induction of diphtheria toxin-resistant
41    mutants in human cells. Carcinogenesis l(9):765-767.
42
43    Roe, FJ; Peto, R; Kearns, F; et al. (1970)  The mechanism of Carcinogenesis by the neutral fraction of cigarette
44    smoke condensate.  BrJ Cancer 24(4):788-806.
45
46    Roggeband, R; Wolterbeek, AP; van den Berg, PT; et al. (1994) DNA adducts in hamster and rat tracheas
47    exposed to benzo(a)pyrene in vitro. Toxicol Lett 72:105-111.
48
49    Rojas, M; Alexandrov, K; Cascorbi, I; et al. (1998) High benzo[ajpyrene diol-epoxide DNA adduct levels in lung
50    and blood cells from individuals with combined CYP1A1 MspI/Msp-GSTM10/0 genotypes. Pharmacogenetics
51    8(2):109-118.
52
53    Rojas, M; Cascorbi, I; Alexandrov, K; et al. (2000) Modulation of benzo [ajpyrene diolepoxide-DNA adduct
54    levels in human white blood cells by CYP1A1, GSTM1 and GSTT1 polymorphism. Carcinogenesis 21(1):35-
55    41.

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                     18           DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene
 2    Rosenkranz, HS; Poirier, LA. (1979) Evaluation of the mutagenicity and DNA-modifying activity of
 3    carcinogens and noncarcinogens in microbial systems. J Natl Cancer Inst 62:873-892.
 4
 5    Ross, J; Nelson, G; Erexson, G; et al. (1991) DNA adducts in rat lung, liver and peripheral blood lymphocytes
 6    produced by i.p. administration of benzo [ajpyrene metabolites and derivatives. Carcinogenesis 12(10):1953-
 7    1955.
 8
 9    Roszinsky-Kocher, G; Easier, A; Rohrborn, G. (1979) Mutagenicity of polycyclic hydrocarbons. V. Induction of
10    sister-chromatid exchanges in vivo. MutatRes 66(l):65-67.
11
12    Roth, RA; Vinegar, A. (1990) Action by the lungs on circulating xenobiotic agents, with a case study of
13    physiologically based pharmacokinetic modeling of benzo(a)pyrene disposition. Pharmacol Ther 48(2):143-
14    155.
15
16    Rudiger, HW; Kohl, F; Mangels, W; et al. (1976) Benzpyrene induces sister chromatid exchanges in cultured
17    human lymphocytes. Nature 262(5566):290-292.
18
19    Russell, LB. (1977) Validation of the in vivo somatic mutation method in the mouse as a prescreen for
20    germinal point mutations.  ArchToxicol 38(l-2):75-85.
21
22    Saffiotti, U; Cefis, F; Kolb, LH. (1968) A method for the experimental induction of bronchogenic carcinoma.
23    Cancer Res 28(1):104-124.
24
25    Saffiotti, U; Montesano, R; Sellakumar, AR; et al. (1972) Respiratory tract carcinogenesis induced in hamsters
26    by different dose levels of benzo-(a)pyrene and ferric oxide. J Natl Cancer Inst 49(4):1199-1204.
27
28    Sagredo, C; Ovrebo, S; Haugen, A;  et al. (2006) Quantitative analysis of benzo [ajpyrene biotransformation and
29    adduct formation in Ahr knockout mice. Toxicol Lettl67(3):173-182.
30
31    Sakai, M; Yoshida, D; Mizusaki, S. (1985) Mutagenicity of polycyclic aromatic hydrocarbons and quinones on
32    Salmonella typhimuriumTA97. Mutat Res 156:61-67.
33
34    Salama, SA; Sierra-Torres, CH; Oh, HY; et al. (2001) Variant metabolizing gene alleles determine the
35    genotoxicity of benzo [ajpyrene. Environ Mol Mutagen 37:17-26.
36
37    Salamone, MF. (1981) Toxicity of 41 carcinogens and noncarcinogenic analogs. Evaluation of short-term tests
38    for carcinogens: Report of the international collaborative program. Prog Mutat Res 1:682-685.
39
40    Santodonato, J; Howard, P; Basu, D. (1981) Health and ecological assessment of polynuclear aromatic
41    hydrocarbons.  J Environ Pathol Toxicol 5(l):l-364.
42
43    Saunders, CR; Ramesh, A; Shickley, DC. (2002) Modulation of neurotoxic behavior of F-344 rats by temporal
44    disposition of benzo [ajpyrene. Toxicol Lett 129:33-45.
45
46    Schlede, E; Kuntzman, R; Haber, S; et al. (1970) Effect of enzyme induction on the metabolism and tissue
47    distribution of benzo (alpha)pyrene. Cancer Res 30(12):2893-2897.
48
49    Schmahl, D; Schmidt, KG; Habs, M. (1977) Syncarcinogenic action of polycyclic hydrocarbons in automobile
50    exhaust gas condensates. IARC Sci Publ (16):53-59.
51
52    Schmidt, KG; Schmael, D; Misfield, J. (1973) Investigations on the carcinogenic burden in man. VI.
53    Experimental investigation to determine a dose-response relationship and to estimate a threshold dose of
54    benzo [ajpyrene in the skin of two different mouse strains.  Zbl Bakt Hyg I Abs B 158:62-68.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                     19            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Schnizlein, CT; Munson, AE; Rhoades, RA. (1987) Immunomodulation of local and systemic immunity after
 2    subchronic pulmonary exposure of mice to benzo(a)pyrene. Int J Immunopharmacol 9(1):99-106.
 o
 4    Schonwald, AD; Bartram, CR; Rudiger, HW. (1977) Benzpyrene-induced sister chromatid exchanges in
 5    lymphocytes of patients with lung cancer. Hum Genet 36:261-264.
 6
 7    Schwarz, D; Kisselev, P; Cascorbi, I; et al. (2001) Differential metabolism of benzo[a]pyrene and
 8    benzo[a]pyrene-7,8-dihydrodiol by human CYP1A1 variants. Carcinogenesis 22(3):453-459.
 9
10    Sega, GA. (1979) Unscheduled DNA synthesis (DNA repair) in the germ cells of male mice-its role in the study
11    of mammalian mutagenesis. Genetics 92:s49-s58.
12
13    Sega, GA. (1982) DNA repair in spermatocytes and spermatids of the mouse. Oak Ridge, TN: Oak Ridge
14    National Lab.
15
16    Sharovskaya, J; Kobliakova, I; Solomatina, N; et al. (2006) Effect of some carcinogenic and non-carcinogenic
17    polycyclic aromatic hydrocarbons on gap junction intercellular communication in hepatoma cell cultures. Eur
18    JCellBiol85(5):387-397.
19
20    Shendrikova, A; Aleksandrov, VA. (1974) Comparative penetration of polycyclic hydrocarbons through the rat
21    placenta into the fetus. Bull Exp Biol Med 77:169-171.
22
23    Shimada, H; Satake, S; Itoh, S; et al. (1990) Multiple-dosing effects of benzo[a]pyrene in the mouse bone
24    marrowmicronucleus text. Mutat Res 234(3-4):179-181.
25
26    Shimada, H; Suzuki, H; Itoh, S; et al. (1992) The micronucleus test of benzo[a]pyrene with mouse and rat
27    peripheral blood reticulocytes. Mutat Res 278(2-3):165-168.
28
29    Shimizu, Y; Nakatsuru, Y; Ichinose, M; et al. (2000) Benzo[a]pyrene carcinogenicity is lost in mice lacking the
30    aryl hydrocarbon receptor.  ProcNatl AcadSci USA97(2):779-782.
31
32    Shimizu, RW; Sun, JD; Li, AP; et al. (1984) The use of sister-chromatid exchange in Chinese hamster primary
33    lung cell cultures to measure genotoxicity. Mutat Res 130(5):333-42.
34
35    Shinohara, K; Cerutti, PA. (1977) Excision repair of benzo[a]pyrene-deoxyguanosine adducts in baby hamster
36    kidney 21/C13 cells and in secondary mouse embryo fibroblasts C57BL/6J. Proc Natl Acad Sci USA 74:979-
37    983.
38
39    Siebert, D; Marquardt, H; Friesel, H; et al. (1981) Polycyclic aromatic hydrocarbons and possible metabolites:
40    convertogenic activity in yeast and tumor initiating activity in mouse skin. J Cancer Res Clin Oncol 102:127-
41    139.
42
43    Simmon, VF. (1979a) In vitro mutagenicity assays of chemical carcinogens and related compounds with
44    Salmonella typhimurium. J Natl Cancer Inst 62(4):893-899.
45
46    Simmon, VF. (1979b) In vitro assays for recombinogenic activity of chemical carcinogens and related
47    compounds with Saccharomyces cerevisiae D3. J Natl Cancer Inst 62(4):901-909.
48
49    Singh, VK; Singh, J; Anand, M; et al. (2008) Comparison of polycyclic aromatic hydrocarbon levels in placental
50    tissues of Indian women with full- and preterm deliveries. Int J Hyg Environ Health 211:639-647.
51
52    Sivak, A; Niemeier, R; Lynch, D; et al. (1997) Skin carcinogenicity of condensed asphalt roofing fumes and
53    their fractions following dermal application to mice. Cancer Lett 117(1):113-123.
54
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    20            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Slaga, TJ; Huberman, E; Selkirk, JK; et al. (1978) Carcinogenicity and mutagenicity of benz(a)anthracene diols
 2    and diol-epoxides. Cancer Res 38(6):1699-1704.
 o
 4    Slaga, TJ; Iyer, RP; Lyga, W; et al. (1980) Comparison of the skin tumor-initiating activities of dihydrodiols,
 5    diol-epoxides, and methylated derivatives of various polycyclic aromatic hydrocarbons. In: Bjorseth, A;
 6    Dennis, AJ; eds. Polynuclear aromatic hydrocarbons: chemistry and biological effects. Columbus, OH: Battelle
 7    Press, pp. 753-769.
 8
 9    Scares, SR; Melo, MA. (2008) Cigarette smoking and reproductive function. Curr Opin Obstet Gynecol
10    20:281-291.
11
12    Sotomayor, RE; Sega, GA. (2000) Unscheduled DNA synthesis assay in mammalian spermatogenic cells: An
13    update. Environmental and Molecular Mutagenesis 36(4):255-265.
14
15    Stavric, B; Klassen, R. (1994) Dietary effects on the uptake of benzo[a]pyrene. Food Chem Toxicol 32(8):727-
16    734.
17
18    Sun, JD; Wolff, RK; Kanapilly, GM. (1982) Deposition, retention, and biological fate of inhaled benzo(a)pyrene
19    adsorbed onto ultrafine particles and as a pure aerosol. Toxicol Appl Pharmacol 65(2):231-244.
20
21    Surh, YJ; Tannenbaum, SR. (1995) Sulfotransferase-mediated activation of 7,8,9,10-tetrahydro-7-ol,
22    7,8-dihydrodiol, and 7,8,9,10-tetraol derivatives of benzo[a]pyrene. Chem Res Toxicol 8(5):693-698.
23
24    Swartz, WJ; Mattison, DR. (1985) Benzo(a)pyrene inhibits ovulation in C57BL/6N mice. Anat Rec 212:268-
25    276.
26
27    Takehisa, S; Wolff, S. (1978) Sister-chromatid exchanges induced in rabbit lymphocytes by 2-aminofluorene
28    and 2-acetylaminofluorene after in vitro and in vivo metabolic activation. Mutat Res 58(2-3):321-329.
29
30    Talaska, G; Ginsburg, D; LaDow, K; et al. (2006) Impact of Cypla2 or Ahr gene knockout in mice: implications
31    for biomonitoring studies. Toxicol Lett 162(2-3):246-249.
32
33    Tamaki, A; Hayashi, H; Nakajima, H; et al. (2004) Polycyclic aromatic hydrocarbon increases mRNA level for
34    interleukin 1 beta in human fibroblast-like synoviocyte line via aryl hydrocarbon receptor. Biol Pharm Bull
35    27(3):407-410.
36
37    Tang, T; Friedman, MA. (1977) Carcinogen activation by human liver enzymes in the Ames mutagenicity test.
38    Mutat Res 46(6):387-394.
39
40    Tang, D; Li, TY; Liu, JJ; et al. (2006) PAH-DNA adducts in cord blood and fetal and child development in a
41    Chinese cohort.  Environ Health Perspect 114(8):1297-1300.
42
43    Tarantini, A; Maitre, A; Lefebvre, E; et al. (2009) Relative contribution of DNA strand breaks and DNA adducts
44    to the genotoxicity of benzo[a]pyrene as a pure compound and in complex mixtures.  Mutat Res 671:67-75.
45
46    Thyssen, J; Althoff, J; Kimmerle, G; et al. (1981) Inhalation studies with benzo[a]pyrene in Syrian golden
47    hamsters.  J Natl Cancer Inst66(3):575-577.
48
49    Tohda, H; Horaguchi, K; Takahashi, K;  et al. (1980) Epstein-Barr virus-transformed human lymphoblastoid
50    cells for study of sister chromatid exchange and their evaluation as a test system.  Cancer Res 40(12):4775-
51    4780.
52
53    Tong, C; Brat, SV; Williams, GM. (1981) Sister-chromatid exchange induction by polycyclic aromatic
54    hydrocarbons in an intact cell system of adult rat-liver epithelial cells. Mutat Res 91:467-473.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    21            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Triolo, AJ; Aponte, GE; Herr, DL. (1977) Induction of aryl hydrocarbon hydroxylase and forestomach tumors
 2    by benzo (a)pyrene. Cancer Res 37:3018-3021.
 o
 4    Tweats, DJ. (1981) Activity of 42 Coded Compounds in a Differential Killing Test using Escherichia coli Strains
 5    WP2, WP67 (uvrA polA), and CM871 (uvrA lexA recA). In: de Serres, FJ.Ashby, J eds. Evaluation of Short-Term
 6    Tests for Carcinogens, New York: Elsevier/North-Holland, 199-209.
 7
 8    Uno, S; Dalton, TP; Shertzer, HG; et al. (2001) Benzo[a]pyrene-induced toxicity: paradoxical protection in
 9    Cyplal(-/-) knockout mice having increased hepatic benzo[a]pyrene-DNA adduct levels. Biochem Biophys
10    Res Commun 289:1049-1056.
11
12    Uno, S; Dalton, TP; Dragin, N; et al.  (2006) Oral benzo [a]pyrene in Cypl knockout mouse lines: CYP1A1
13    important in detoxication, CYP1B1 metabolism required for immune damage independent of total-body
14    burden and clearance rate. Mol Pharmacol 69(4):1103-1114.
15
16    U.S. EPA (Environmental Protection Agency). (1988) Recommendations for and documentation of biological
17    values for use in risk assessment. Prepared by the Environmental Criteria and Assessment Office, Office of
18    Health and Environmental Assessment, Cincinnati, OH for the Office of Solid Waste and Emergency Response,
19    Washington, DC; EPA 600/6-87/008. Available online at http://www.epa.gov/iris/backgrd.html (accessed
20    April 20, 2010).
21
22    U.S. EPA (Environmental Protection Agency). (1991a) Dose-response analysis of ingested benzo (a)pyrene
23    (CAS no 50-32-8). Cincinnati, OH: U.S. Environmental Protection Agency. EPA600R92045.
24
25    U.S. EPA (Environmental Protection Agency). (1996) Guidelines for reproductive toxicity risk assessment.
26    Federal Register 61(212):56274-56322. Available online at http://www.epa.gov/iris/backgrd.html
27    (accessed April 20,2010).
28
29    U.S. EPA (U.S. Environmental Protection Agency). (2010) Revised Draft: Multistage Weibull time-to-tumor
30    model in EPA's Benchmark Dose Software (BMDS): Methodology Description. Washington, DC: U.S.
31    Environmental Protection Agency.
32
33    Valencia, R; Houtchens, K. (1981) Mutagenic activity of 10 coded compounds in the drosophila sex-linked
34    recessive lethal test.  Evaluation of short-term tests for carcinogens: report of the international collaborative
35    program: Report of the International Collaborative Program. Prog Mutat Res 1:651-659.
36
37    Valentin-Severin, I; Thybaud, V; Le Bon, AM; et al. (2004) The autoradiographic test for unscheduled DNA
38    synthesis: a sensitive assay for the  detection of DNA repair in the HepG2 cell line. Mutat Res 559(1-2):211-
39    217.
40
41    Van Agen, B; Maas, LM; Zwingmann, IH; et al. (1997) B[a]P-DNA adduct formation and induction of human
42    epithelial lung cell transformation. Environ Mol Mutagen 30(3):287-292.
43
44    Van Rooij, JG; Bodelier-Bade, MM; Jongeneelen, FJ. (1993) Estimation of individual dermal and respiratory
45    uptake of polycyclic aromatic hydrocarbons in 12 coke oven workers.  Br J Ind Med 50(7):623-632.
46
47    Vogel, EW; Zijlstra, JA; Blijleven, WG. (1983) Mutagenic activity of selected aromatic amines and polycyclic
48    hydrocarbons in Drosophila melanogaster. Mutat Res  107(l):53-77.
49
50    Wang, JJ;  Frazer, DG; Law, B; et al. (2003) Identification and quantification of urinary benzo[ajpyrene and its
51    metabolites from asphalt fume exposed mice by microflow LC coupled to hybrid quadrupole time-of-flight
52    mass spectrometry. Analyst 128(7):864-870.
53
54    Warshawsky, D; Barkley, W. (1987) Comparative carcinogenic potencies of 7H-dibenzo[c,g]carbazole,
55    dibenz[a,j]acridine and benzo [ajpyrene in mouse skin.  Cancer Lett 37(3):337-344.

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    22           DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene
 2    Wattenberg, LW. (1972) Inhibition of carcinogenic and toxic effects of polycyclic hydrocarbons by phenolic
 3    anti-oxidants and ethoxyquin. J Natl Cancer Inst48(5):1425-1430.
 4
 5    Wattenberg, LW. (1974) Inhibition of carcinogenic and toxic effects of polycyclic hydrocarbons by several
 6    sulfur-containing compounds. J Natl Cancer Inst52(5):1583-1587.
 7
 8    Waylen, AL; Metwally, M; Jones, GL; et al. (2009) Effects of cigarette smoking upon clinical outcomes of
 9    assisted reproduction: a meta-analysis. Hum Reprod 15:31-44.
10
11    Weinstein, D; Katz, ML; Kazmer, S. (1977) Chromosomal effects of carcinogens and non-carcinogens on WI-38
12    after short term exposures with and without metabolic activation. Mutat Res 46(4):297-304.
13
14    Wester, RC; Maibach, HI; Bucks, DA; et al. (1990) Percutaneous absorption of [14C]DDT and
15    [14C]benzo[a]pyrene from soil. Fundam ApplToxicol 15(3):510-516.
16
17    Weyand, EH; Bevan, DR. (1986) Benzo(a)pyrene disposition and metabolism in rats following intratracheal
18    instillation. Cancer Res 46(11):5655-5661.
19
20    Weyand, EH; Bevan, DR. (1987) Species differences in disposition of benzo[a]pyrene. Drug Metab Dispos
21    15 (4): 442-448.
22
23    Weyand, EH; LaVoie, EJ. (1988) Comparison of PAH DNA adduct formation and tumor initiating activity in
24    newborn mice. ProcAm Assoc Cancer Res 29(10):98.
25
26    Weyand, EH; He, ZM; Ghodrati, F; et al. (1992) Effect of fluorine substitution on benzo[j]fluoranthene
27    genotoxicity. Chem Biol Interact 84(l):37-53.
28
29    Weyand, EH; Chen, YC; Wu, Y; et al. (1995) Differences in the tumorigenic activity of a pure hydrocarbon and
30    a complex mixture following ingestion: benzo[a]pyrene vs manufactured gas plant residue. Chem Res Toxicol
31    8(7):949-954.
32
33    Whitwell, J; Fowler, P; Allars, S; et al. (2010) 5-Fluorouracil, colchicine, benzo[a]pyrene and cytosine
34    arabinoside tested in the in vitro mammalian cell micronucleus test (MNvit) in Chinese hamster V79 cells at
35    Covance Laboratories, Harrogate, UK in support of OECD draft Test Guideline 487. Mutat Res 702(2):230-
36    236.
37
38    WHO (World Health Organization). (1998) Environmental health criteria. Vol. 202. Selected non-heterocyclic
39    polycyclic aromatic hydrocarbons.  International Programme on Chemical Safety, Geneva, Switzerland.
40
41    Wielgosz, SM; Brauze, D; Pawlak, AL. (1991) Ah locus-associated differences in induction of sister-chromatid
42    exchanges and in DNA adducts by benzo[a]pyrene in mice. Mutat Res 246(1):129-137.
43
44    Wiersma, DA; Roth, RA. (1983a) Total body clearance of circulating benzo(a)pyrene in conscious rats: effect
45    of pretreatment with 3-methylcholanthrene and the role of liver and lung. J Pharmacol Exp Ther 226(3):661-
46    667.
47
48    Wiersma, DA; Roth, RA. (1983b) The prediction of benzo[a]pyrene clearance by rat liver and lung from
49    enzyme kinetic data.  Mol Pharmacol 24(2):300-308.
50
51    Wijnhoven, SW; Kool, HJ; van Oostrom, CT; et al. (2000) The  relationship between benzo[a]pyrene-induced
52    mutagenesis and carcinogenesis in repair-deficient Cockayne syndrome group B mice. Cancer Res
53    60(20):5681-5687.
54
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    23           DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofbenzo[a]pyrene


 1    Willems, MI; Roggeband, R; Baan, RA; et al. (1991) Monitoring the exposure of rats to benzo[a]pyrene by the
 2    determination of mutagenic activity in excreta, chromosome aberrations and sister chromatid exchanges in
 3    peripheral blood cells, and DNA adducts in peripheral blood lymphocytes and liver. Mutagenesis 6(2):151-
 4    158.
 5
 6    Williams, GM; Laspia, MF; Dunkel, VC. (1982) Reliability of the hepatocyte primary culture/DNA repair test in
 7    testing of coded carcinogens and noncarcinogens. Mutat Res 97:359-370.
 8
 9    Williams, JA; Martin, FL; Muir, GH; et al. (2000) Metabolic activation of carcinogens and expression of various
10    cytochromes P450 in human prostate tissue. Carcinogenesis 21(9):1683-1689.
11
12    Wilson, JS; Holland, LM. (1988) Periodic response difference in mouse epidermis chronically exposed to
13    crude-oils or BaP: males vs. females. Toxicology 50(l):83-94.
14
15    Withey, JR; Shedden, J; Law, FC; et al. (1993) Distribution of benzo[a]pyrene in pregnant rats following
16    inhalation exposure and a comparison with similar data obtained with pyrene. J Appl Toxicol 13(3):193-202.
17
18    Wojciechowski, JP; Kaur, P; Sabharwal, PS. (1981) Comparison of metabolic systems required to activate pro-
19    mutagens/carcinogens in vitro for sister-chromatid exchange studies. Mutat Res 88(1): 89-9 7.
20
21    Wolff, S; Takehisa, S. (1977) Induction of sister chromatid exchanges in mammalian cells by low
22    concentrations of mutagenic carcinogens that require metabolic activation  as well as those  that do not. Dev
23    Toxicol Environ Sci 2:193-200.
24
25    Wolff, RK; Griffith, WC; Henderson, RF; et al. (1989) Effects of repeated inhalation exposures to 1-
26    nitropyrene, benzo[a]pyrene, Ga203 particles, and S02 alone and in combinations on particle clearance,
27    bronchoalveolar lavage fluid composition, and histopathology. J Toxicol Environ Health 27(1):123-138.
28
29    Wood, AW; Levin, W; Lu, AY; et al. (1976) Metabolism of benzo(a)pyrene and benzo (a)pyrene derivatives to
30    mutagenic products by highly purified hepatic microsomal enzymes.  J Biol  Chem 251(16):4882-4890.
31
32    Wood, AW; Levin, W; Chang, RL; et al. (1980) Mutagenicity and tumor-initiating activity of
33    cyclopenta(c,d)pyrene and structurally related compounds. Cancer Res 40(3):642-649.
34
35    Wormhoudt, LW; Commandeur, JN; Vermeulen, NP. (1999) Genetic polymorphisms of human N-
36    acetyltransferase, cytochrome P450, glutathione-S-transferase, and epoxide hydrolase enzymes: relevance to
37    xenobiotic metabolism and toxicity. Grit Rev Toxicol  29(1):59-124.
38
39    Wormley, DD; Chirwa, S; Nayyar, T; et al. (2004) Inhaled benzo(a)pyrene impairs long-term potentiation in
40    the Fl generation rat dentate gyrus. Cell Mol Biol 50(6):715-721.
41
42    Wu, MT; Simpson, CD; Christiani, DC; et al.  (2002) Relationship of exposure to coke-oven emissions and
43    urinary metabolites of benzo(a)pyrene and pyrene in coke-oven workers. Cancer Epidemiol Biomarkers Prev
44    11(3):311-314.
45
46    Wu, J; Ramesh, A; Nayyar, T; et al. (2003) Assessment of metabolites and AhR and CYP1A1 mRNA expression
47    subsequent to prenatal exposure to inhaled benzo(a)pyrene.  Int J Dev Neurosci 21(6):333-346.
48
49    Wu, Q; Suzuki, JS; Zaha, H; et al. (2008) Differences in gene expression and benzo[ajpyrene-induced DNA
50    adduct formation in the liver of three strains of female mice with identical AhRb2 genotype treated with
51    2,3,7,8-tetrachlorodibenzo-p-dioxin and/or benzo[ajpyrene. J Appl Toxicol 28:724-733.
52
53    Wu, J; Hou, H; Ritz, B; et al. (2010) Exposure to polycyclic aromatic hydrocarbons and missed abortion in
54    early pregnancy in a Chinese population. Sci Total Environ 408:2312-2318.
55

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    24           DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofbenzo[a]pyrene


 1    Wynder, EL; Hoffmann, D. (1959) A study of tobacco carcinogenesis. VII. The role of higher polycyclic
 2    hydrocarbons. Cancer 12:1079-1086.
 o
 4    Wynder, EL; Fritz, L; Furth, N. (1957) Effect of concentration of benzopyrene in skin carcinogenesis. J Natl
 5    Cancer Instl9(3):361-370.
 6
 7    Xu, C; Chen, JA; Qiu, Z; et al. (2010) Ovotoxicity and PPAR-mediated aromatase downregulation in female
 8    Sprague-Dawley rats following combined oral exposure to benzo[a]pyrene and di-(2-ethylhexyl) phthalate.
 9    ToxicolLettl99(3):323-332.
10
11    Yamasaki, H. (1990) Gap junction intercellular communication and carcinogenesis. Carcinogenesis
12    11(7):1051-1058.
13
14    Yang, JJ; Roy, TA; Krueger, AJ; et al. (1989) In vitro and in vivo percutaneous absorption of benzo[a]pyrene
15    from petroleum crude-fortified soil in the rat. Bull Environ Contam Toxicol 43(2):207-214.
16
17    Yang, H; Mazur-Melnyk, M; de Boer, JG; et al. (1999) A comparison of mutational specificity of mutations
18    induced by S9-activated B[a]P andbenzo[a]pyrene-7,8-diol-9,10-epoxide at the endogenous APRTgene in
19    CHO cells. Mutat Res 423:23-32.
20
21    Zeilmaker, MJ; van Eijkeren, JCH. (1997) Modeling Ah-receptor dependent P450 induction I. Cellular model
22    definition and its corporation in a PBPK model of 2,3,7,8-TCDD. RIVM Report 604138 001. National Institute
23    of Public Health and the Environment, B, The Netherlands.
24
25    Zeilmaker, MJ; van Eijkeren, JCH; Kroese, ED. (1999a) PBPK simulated DNA adduct formation: relevance for
26    the risk assessment of benzo(a)pyrene. Bilthoven, The Netherlands.
27
28    Zeilmaker, MJ; van Eijkeren, JCH; Oiling, M. (1999b) A PBPK-model for B(a)P in the rat relating dose and liver
29    DNA-adduct level. Bilthoven, The Netherlands.
30
31    Zhang, HM;  Nie, JS; Li, X; et al. (2012) Characteristic analysis of peripheral blood mononuclear cell apoptosis
32    in coke oven workers. J Occup Health 54(1):44-50.
33    Zheng, Z;  Fang, JL; Lazarus, P. (2002) Glucuronidation: an important mechanism for detoxification of
34    benzo[a]pyrene metabolites in aerodigestive tract tissues. Drug Metab Dispos 30(4):397-403.
35
36    Zheng, SJ; Tian, HJ; Cao, J; et al. (2010) Exposure to di(n-butyl)phthalate and benzo(a)pyrene alters IL-lbeta
37    secretion and subset expression of testicular macrophages, resulting in decreased testosterone production in
38    rats. Toxicol Appl Pharmacol 248(l):28-37.
39
40    Zijlstra, JA; Vogel, EW. (1984) Mutagenicity of 7,12-dimethylbenz[a]anthracene and some other aromatic
41    mutagens in Drosophila melanogaster. Mutat Res 125(2):243-261.
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                    25           DRAFT—DO NOT CITE OR QUOTE

-------