f/EPA
                                                             EPA/63 5/R-14/312a
                                                            External Review Draft
                                                                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)
                                 September 2014
                                    NOTICE

This document is an External Review draft. This information is distributed solely for the purpose
of pre-dissemination peer review under applicable information quality guidelines. It has not been
formally disseminated by EPA. It does not represent and should not be construed to represent any
Agency determination or policy. It is being circulated for review of its technical accuracy and
science policy implications.
                     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                                           ii              DRAFT—DO NOT CITE OR QUOTE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
AUTHORS | CONTRIBUTORS | REVIEWERS	ix

PREFACE	xii

PREAMBLE TO IRIS TOXICOLOGICAL REVIEWS	xvi

EXECUTIVE SUMMARY	xxxiv

LITERATURE SEARCH STRATEGY! STUDY SELECTION	xli
1.   HAZARD IDENTIFICATION	1-1

    1.1. PRESENTATION AND SYNTHESIS OF EVIDENCE BY ORGAN/SYSTEM	1-1

       1.1.1.  Developmental Toxicity	1-1

       1.1.2.  Reproductive Toxicity	1-22

       1.1.3.  Immunotoxicity	1-37

       1.1.4.  Other Toxicity	1-44

       1.1.5.  Carcinogenicity	1-52

    1.2. SUMMARY AND EVALUATION	1-81

       1.2.1.  Weight of Evidence for Effects Other than Cancer	1-81

       1.2.2.  Weight of Evidence for Carcinogenicity	1-82

2.   DOSE-RESPONSE ANALYSIS	2-1

    2.1. ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER	2-1

       2.1.1.  Identification of Studies and Effects for Dose-Response Analysis	2-1

       2.1.2.  Methods of Analysis	2-5

       2.1.3.  Derivation of Candidate Values	2-8

       2.1.4.  Derivation of Organ/System-Specific Reference Doses	2-12

       2.1.5.  Selection of the Proposed Overall Reference Dose	2-14

       2.1.6.  Confidence Statement	2-14

       2.1.7.  Previous IRIS Assessment: Reference Dose	2-15

    2.2. INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER THAN CANCER	2-15

       2.2.1.  Identification of Studies and Effects for Dose-Response Analysis	2-15

       2.2.2.  Methods of Analysis	2-17

       2.2.3.  Derivation of Candidate Values	2-19

       2.2.4.  Derivation of Organ/System-Specific Reference Concentrations	2-22


          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


       2.2.5. Selection of the Proposed Reference Concentration	2-23

       2.2.6. Confidence Statement	2-23

       2.2.7. Previous IRIS Assessment: Reference Concentration	2-24

       2.2.8. Uncertainties in the Derivation of the RfD and RfC	2-24

    2.3. ORAL SLOPE FACTOR FOR CANCER	2-25

       2.3.1. Analysis of Carcinogenicity Data	2-25

       2.3.2. Dose-Response Analysis—Adjustments and Extrapolation Methods	2-28

       2.3.3. Derivation of the Oral Slope Factor	2-28

       2.3.4. Uncertainties in the Derivation of the Oral Slope Factor	2-30

       2.3.5. Previous IRIS Assessment: Oral Slope Factor	2-32

    2.4. INHALATION UNIT RISK FOR CANCER	2-32

       2.4.1. Analysis of Carcinogenicity Data	2-33

       2.4.2. Dose-Response Analysis—Adjustments and Extrapolation Methods	2-34

       2.4.3. Inhalation Unit Risk Derivation	2-35

       2.4.4. Uncertainties in the Derivation of the Inhalation Unit Risk	2-36

       2.4.5. Previous IRIS Assessment: Inhalation Unit Risk	2-38

    2.5. DERMAL SLOPE FACTOR FOR CANCER	2-39

       2.5.1. Analysis of Carcinogenicity Data	2-39

       2.5.2. Dose-Response Analysis—Adjustments and Extrapolation Methods	2-40

       2.5.3. Derivation of the Dermal Slope Factor	2-42

       2.5.4. Dermal Slope Factor Cross-Species Scaling	2-44

       2.5.5. Uncertainties in the Derivation of the Dermal Slope Factor	2-45

       2.5.6. Previous IRIS Assessment: Dermal Slope Factor	2-47

    2.6. APPLICATION OF AGE-DEPENDENT ADJUSTMENT FACTORS (ADAFS)	2-47

REFERENCES	R-l
          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
TABLES

Table ES-1. Organ/system-specific RfDs and proposed overall RfD for benzo[a]pyrene	xxxv
Table ES-2. Organ/system-specific RfCs and proposed overall RfCfor benzo[a]pyrene	xxxvii
Table LS-1. Summary of the search strategy employed for benzo[a]pyrene	xli
Table 1-1. Evidence pertaining to developmental effects of benzo[a]pyrene in humans	1-4
Table 1-2. Evidence pertaining to developmental effects of benzo[a]pyrene in animals	1-5
Table 1-3. Evidence pertaining to the neurodevelopmental effects of benzo[a]pyrene from PAH
               mixtures	1-13
Table 1-4. Evidence pertaining to the neurodevelopmental effects of benzo[a]pyrene in animals	1-15
Table 1-5. Evidence pertaining to the male reproductive toxicity of benzo[a]pyrene in adult
               animals	1-25
Table 1-6. Evidence pertaining to the female reproductive effects of benzo[a]pyrene in humans	1-32
Table 1-7. Evidence pertaining to the female reproductive effects of benzo[a]pyrene in adult
               animals	1-33
Table 1-8. Evidence pertaining to the immune effects of benzo[a]pyrene in animals	1-41
Table 1-9. Evidence pertaining to other toxicities of benzo[a]pyrene in animals	1-49
Table 1-10. Cancer sites for PAH-related agents reviewed by IARC	1-55
Table 1-11. Summary of epidemiologic studies of benzo[a]pyrene (direct measures) in relation
               to lung cancer risk: Tier 1 studies	1-55
Table 1-12. Summary of epidemiologic studies of benzo[a]pyrene (direct measures) in relation
               to lung cancer risk: Tier 2 studies	1-56
Table 1-13. Summary of epidemiologic studies of benzo[a]pyrene (direct measures) in relation
               to bladder cancer risk	1-60
Table 1-14. Tumors observed in chronic oral animal bioassays	1-63
Table 1-15. Tumors observed in chronic inhalation animal bioassays	1-66
Table 1-16. Tumors observed in chronic dermal animal bioassays	1-68
Table 1-17. Experimental support for the postulated key events for mutagenic mode of action	1-75
Table 1-18. Supporting evidence for the carcinogenic to humans cancer descriptor for
               benzo[a]pyrene	1-87
Table 2-1. Summary of derivation of PODs	2-7
Table 2-2. Effects and corresponding derivation of candidate values	2-10
Table 2-3. Organ/system-specific RfDs and proposed overall  RfD for benzo[a]pyrene	2-13
Table 2-4. Summary of derivation of PODs	2-19
Table 2-5. Effects and corresponding derivation of candidate values	2-21
Table 2-6. Organ/system-specific RfCs and proposed overall  RfCfor benzo[a]pyrene	2-22
Table 2-7. Summary of the oral slope factor derivations	2-29
Table 2-8. Summary of uncertainties in the derivation of cancer risk values for benzo[a]pyrene
               oral slope factor	2-31
Table 2-9. Summary of the inhalation unit risk derivation	2-36
Table 2-10. Summary of uncertainties in the derivation of cancer risk values for benzo[a]pyrene
               (inhalation unit risk)	2-38
Table 2-11. Summary of dermal slope factor derivations,  unadjusted for interspecies differences	2-43
Table 2-12. Summary of uncertainties in the derivation of cancer risk values for benzo[a]pyrene
               dermal slope factor	2-47
Table 2-13. Sample application of ADAFs for the estimation of benzo[a]pyrene cancer risk
               following lifetime (70-year) oral exposure	2-48

           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


Table 2-14. Sample application of ADAFs for the estimation of benzo[a]pyrene cancer risk
              following lifetime (70-year) inhalation exposure	2-49
Table 2-15. Sample application of ADAFs for the estimation of benzo[a]pyrene cancer risk
              following lifetime (70-year) dermal exposure	2-49
FIGURES

Figure LS-1. Study selection strategy	xliii
Figure 1-1. Exposure-response array for developmental effects following oral exposure to
              benzo[a]pyrene	1-8
Figure 1-2. Exposure-response array for neurodevelopmental effects following oral exposure	1-18
Figure 1-3. Exposure-response array for male reproductive effects following oral exposure in
              adult animals	1-28
Figure 1-4. Exposure-response array for female reproductive effects following oral exposure in
              adult animals	1-34
Figure 1-5. Exposure-response array for immune effects following oral exposure	1-42
Figure 1-6. Proposed metabolic activation pathways and key events in the carcinogenic mode of
              action for benzo[a]pyrene	1-70
Figure 2-1. Candidate values with corresponding PODs and composite UFs	2-12
Figure 2-2. Candidate values with corresponding PODs and composite UFs	2-22
          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
1-OH-Py  1-hydroxypyrene                          ETS
AchE     acetylcholine esterase                       EU
ADAF    age-dependent adjustment factor             Fe203
Ah       aryl hydrocarbon                          FSH
AHH     aryl hydrocarbon hydroxylase                GABA
AhR     aryl hydrocarbon receptor                   GD
AIC      Akaike's Information Criterion                GI
AKR     aldo-keto reductase                        GJIC
AMI     acute myocardial infarction
ANOVA   analysis of variance                        GSH
ARNT    Ah receptor nuclear translocator              GST
AST     aspartate transaminase                     GSTM1
ATSDR   Agency for Toxic Substances and              hCG
         Disease Registry                           HEC
BMC     benchmark concentration                   HED
BMCL    benchmark concentration lower              HERO
         confidence limit
BMD     benchmark dose                           HFC
BMDL    benchmark dose, 95% lower bound           HPLC
BMDS    Benchmark Dose Software
BMR     benchmark response                        hprt
BPDE    benzo[a]pyrene-7,8-diol-9,10-epoxide
BPQ     benzo [ajpyrene semiquinone                 HR
BrdU     bromodeoxyuridine                        Hsp90
BSM     benzene-soluble matter                     i.p.
BUN     blood urea nitrogen                        i.v.
BW      body weight                               Ig
CA       chromosomal aberration                    IHD
CAL/EPA California Environmental Protection           IRIS
         Agency                                   LDH
CASRN   Chemical Abstracts Service Registry           LH
         Number                                  LOAEL
CERCLA  Comprehensive Environmental               MAP
         Response, Compensation, and Liability         MCL
         Act                                      MCLG
CHO     Chinese hamster ovary                      MIAME
CI       confidence interval
GYP     cytochrome                               MLE
CYP450   cytochrome P450                          MMAD
DAF     dosimetric adjustment factor                 MN
dbcAMP  dibutyl cyclic adenosine                     MPPD
         monophosphate                           mRNA
DMSO    dimethyl sulfoxide                          MS
DNA     deoxyribonucleic acid                       NCE
EC       European Commission                      NCEA
EH       epoxide hydrolase
ELISA    enzyme-linked immunosorbent assay          NIOSH
EPA     Environmental Protection Agency
EROD    7-ethoxyresorufin-O-deethylase              NK
environmental tobacco smoke
European Union
ferrous oxide
follicle stimulating hormone
gamma-aminobutyric acid
gestational day
gastrointestinal
gap junctional intercellular
communication
reduced glutathione
glutathione-S-transferase
glutathione-S-transferase Ml
human chorionic gonadotropin
human equivalent concentration
human equivalent dose
Health and Environmental Research
Online
high-frequency cell
high-performance liquid
chromatography
hypoxanthine guanine phosphoribosyl
transferase
hazard ratio
heat shock protein 90
intraperitoneal
intravenous
immunoglobulin
ischemic heart disease
Integrated Risk Information System
lactate dehydrogenase
luteinizing hormone
lowest-observed-adverse-effect level
mitogen-activated protein
Maximum Contaminant Level
Maximum Contaminant Level Goal
Minimum Information About a
Microarray Experiment
maximum likelihood estimate
mass median aerodynamic diameter
micronucleus
Multi-Path Particle Deposition
messenger ribonucleic acid
mass spectrometry
normochromatic erythrocyte
National Center for Environmental
Assessment
National Institute for Occupational
Safety and Health
natural-killer
          This document is a draft for review purposes only and does not constitute Agency policy.
                                             vii              DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
NMDA   N-methyl-D-aspartate                       SEM
NOAEL   no-observed-adverse-effect level             SHE
NPL     National Priorities List                      SIR
NQO     NADPH:quinone oxidoreductase              SMR
NRC     National Research Council                   SOAR
NTP     National Toxicology Program                SOD
OECD    Organisation for Economic                   SRBC
         Co-operation and Development              SSB
OR       odds ratio                                 TCDD
ORD     Office of Research and Development          TK
PAH     polycyclic aromatic hydrocarbon             ToxR
PBMC    peripheral blood mononuclear cell            TPA
PBPK    physiologically based pharmacokinetic        TUNEL
PCA     Principal Components Analysis
PCE     polychromatic erythrocyte                   TWA
PCNA    proliferating cell nuclear antigen             UCL
PND     postnatal day                              UDP-UGT
POD     point of departure
PUVA    psoralen plus ultraviolet-A                   UDS
RBC     red blood cell                              UF
RDDRER  regional deposited dose ratio for             UFA
         extrarespiratory effects                     UFn
RfC      inhalation reference concentration            UFH
RfD      oral reference dose                         UFi
RNA     ribonucleic acid                            UFS
ROS     reactive oxygen species
RR       relative risk                                UVA
s.c.       subcutaneous                              UVB
SCC      squamous cell carcinoma                    WBC
SCE      sister chromatid exchange                   WESPOC
SCSA     sperm chromatin structure assay             WT
SD       standard deviation                         WTC
SE       standard error                             XPA
standard error of the mean
Syrian hamster embryo
standardized incidence ratio
standardized mortality ratio
Systematic Omics Analysis Review
superoxide dismutase
sheep red blood cells
single-strand break
2,3,7,8-tetrachlorodibenzo-p-dioxin
thymidine kin as e
Toxicological Reliability Assessment
12-0-tetradecanoylphorbol-13-acetate
terminal deoxynucleotidyl transferase
dUTP nick end labeling
time-weighted average
upper confidence limit
uridine diphosphate-
glucuronosyltransferase
unscheduled DNA synthesis
uncertainty factor
interspecies uncertainty factor
database deficiencies uncertainty factor
intraspecies uncertainty factor
LOAEL-to-NOAEL uncertainty factor
subchronic-to-chronic uncertainty
factor
ultraviolet-A
ultraviolet-B
white blood cell
water escape pole climbing
wild type
World Trade Center
xeroderma pigmentosum group A
           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
AUTHORS  |  CONTRIBUTORS  |  REVIEWERS
Assessment Team
       Christine Cai, MS
       Glinda Cooper, Ph.D.
       Louis D'Amico, Ph.D.
       Jason Fritz, Ph.D.
       Martin Gehlhaus, MHS
       Catherine Gibbons, Ph.D.
       Karen Hogan, MS
       Andrew Kraft, Ph.D.
       Kathleen Newhouse, MS (Assessment Manager)
       Linda Phillips, Ph.D.
       Margaret Pratt, Ph.D.
       Keith Salazar, Ph.D.
       John Schaum, MS (retired)
       Suryanarayana Vulimiri, DVM
       Gene Ching-Hung Hsu, Ph.D. (formerly with
       EPA)

       Lyle Burgoon, Ph.D.
       John Cowden, Ph.D.
       Amanda Persad, Ph.D.
       John Stanek, Ph.D.
       Chris Brinkerhoff, Ph.D.
       Emma McConnell, MS

       Scott Glaberman, Ph.D.
U.S. EPA
Office of Research and Development
National Center for Environmental
Assessment
Washington, DC
U.S. EPA
Office of Research and Development
National Center for Environmental
Assessment
Research Triangle Park, NC

Oak Ridge Institute for Science and
Education Fellow

American Association for the
Advancement of Science Fellow
Scientific Support
       Lynn Flowers, Ph.D.
       John Fox, Ph.D.
       Paul White, Ph.D.
       Maria Spassova, Ph.D.
U.S. EPA
Office of Research and Development
National Center for Environmental
Assessment
Washington, DC
Production Team
       Taukecha Cunningham
       Maureen Johnson
       Terri Konoza
       Vicki Soto
U.S. EPA
Office of Research and Development
National Center for Environmental
Assessment
Washington, DC
Contractor Support
       Heather Carlson-Lynch, S.M.
       Peter McClure, Ph.D.
SRC, Inc., Syracuse, NY
          This document is a draft for review purposes only and does not constitute Agency policy.
                                           ix             DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofBenzo[a]pyrene
        Megan Riccardi
        Kelly Salinas
        Joe Santodonato
        Julie Stickney, Ph.D.

        George Holdsworth, Ph.D.
        Lutz W.Weber, Ph.D.

        Janusz Z. Byczkowski, Ph.D., D.Sc.
Oak Ridge Institute for Science and
Education, Oak Ridge, TN

JZB Consulting, Fairborn, OH
Executive Direction
       Kenneth Olden, Ph.D., Sc.D., L.H.D.
          (Center Director)
       Lynn Flowers, Ph.D.
          (Associate Director for Health)
       Vincent Cogliano, Ph.D.
          (IRIS Program Director—acting)
       Samantha Jones, Ph.D.
          (IRIS Associate Director for Science)
       Jamie B. Strong, Ph.D.
          (Toxic Effects Branch Chief)
U.S. EPA/ORD/NCEA
Washington, DC
Internal Review Team
       Stephen Nesnow, Ph.D. (retired)
        Rita Schoeny, Ph.D.
U.S. EPA
National Health and Environmental
Effects Research Laboratory
Research Triangle Park, NC

U.S. EPA
Office of Water
Washington, DC	
Reviewers

This assessment was provided for review to scientists in EPA's Program and Region Offices.
Comments were submitted by:

       Office of Children's Health Protection, Washington DC

       Office of Policy, Washington, DC

       Office of Solid Waste and Emergency Response, Washington DC

       Office of Water, Washington DC

       Region 2, New York, NY

       Region 3, Philadelphia, PA

       Region 8, Denver, CO

This assessment was provided for review to other federal agencies and the Executive Office of the
President  Comments were submitted by:
           This document is a draft for review purposes only and does not constitute Agency policy.
                                               x               DRAFT—DO NOT CITE OR QUOTE

-------
                                                Toxicological Review ofBenzo[a]pyrene
Agency for Toxic Substances and Disease Registry, Centers for Disease Control,
Department of Health and Human Services

Department of Defense

National Aeronautics and Space Administration

National Institute for Occupational Safety and Health, Centers for Disease Control,
Department of Health and Human Services

Office of Management and Budget, Executive Office of the President

White House Council on Environmental Quality, Executive Office of the President
   This document is a draft for review purposes only and does not constitute Agency policy.
                                       xi              DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
 1
 2    PREFACE
 3          This Toxicological Review, prepared under the auspices of the U.S. Environmental
 4    Protection Agency's (EPA's) Integrated Risk Information System (IRIS) program, critically reviews
 5    the publicly available studies on benzo[a]pyrene in order to identify potential adverse health effects
 6    and to characterize exposure-response relationships. Benzo[a]pyrene is found in the environment
 7    and in food. Benzo[a]pyrene occurs in conjunction with other structurally related chemical
 8    compounds known as polycyclic aromatic hydrocarbons (PAHs).1  Benzo[a]pyrene is universally
 9    present in these mixtures and is routinely analyzed and detected in environmental media
10    contaminated with PAH mixtures: thus it is often used as an indicator chemical to measure
11    exposure to PAH mixtures [Bostrometal.. 2002). and as an index chemical for deriving potency
12    factors for PAH mixtures.
13          Benzo[a]pyrene is listed as a hazardous substance under the Comprehensive Environmental
14    Response, Compensation, and Liability Act of 1980 (CERCLA), is found at 524 hazardous waste sites
15    on the National Priorities List (NPL) and is ranked number 8 out of 275 chemicals on the Priority
16    List of Hazardous Substances for CERCLA [ATSDR. 2011).  This ranking is based on a combination
17    of factors that include the frequency of occurrence at NPL  sites, the potential for human exposure,
18    and the potential health hazard. Benzo[a]pyrene is also listed as a drinking water contaminant
19    under the Safe Drinking Water Act and a Maximum Contaminant Level Goal (MCLG) and
20    enforceable Maximum Contaminant Level (MCL) have been established2. It is also one of the
21    chemicals included in EPA's Persistent Bioaccumulative and Toxic Chemical Program
22    [http://www.epa.gov/pbt/pubs/benzo.htm]. In air, benzo[a]pyrene is regulated as a component in
23    a class of chemicals referred to as Polycyclic Organic Matter, defined as a Hazardous Air Pollutant
24    by the 1990 amendments to the Clean Air Act.
25          This assessment updates IRIS assessment of benzo[a]pyrene that was developed in 1987.
26    The previous assessment included a cancer descriptor and oral slope factor. New information has
27    become available, and this assessment reviews information on all health effects by all exposure
28    routes. Organ/system-specific reference values are calculated based on developmental,
29    reproductive, and immune system toxicity data. These reference values may be useful for
30    cumulative risk assessments that consider the combined effect of multiple agents acting on the
31    same biological system. In addition, in consideration of the Agency's need to estimate the potential
      iPAHs are a large class of chemical compounds formed during the incomplete combustion of organic matter.
      They consist of only carbon and hydrogen arranged in two or more fused rings.
      2MCLG = 0, MCL = 0.0002 mg/L.

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

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    for skin cancer from dermal exposure [U.S. EPA, 2004], especially in children exposed to
 2    contaminated soil, this assessment includes the IRIS Program's first dermal slope factor.
 3          This assessment was conducted in accordance with EPA guidance, which is cited and
 4    summarized in the Preamble to Toxicological Reviews. Appendices for chemical and physical
 5    properties, toxicokinetic information, and summaries of toxicity studies are provided as
 6    Supplemental Information to this assessment.
 7          For additional information about this assessment or for general questions regarding IRIS,
 8    please contact EPA's IRIS Hotline at 202-566-1676 (phone), 202-566-1749 (fax),  or
 9    hotline.iris@epa.gov.

10    Chemical Properties
11          Benzo[a]pyrene is a five-ring PAH. It is a pale yellow crystalline solid with a faint aromatic
12    odor. It is relatively insoluble in water and has low volatility. Benzo[a]pyrene is released to the air
13    from both natural and anthropogenic sources and removed from the atmosphere by photochemical
14    oxidation; reaction with nitrogen oxides, hydroxy and hydroperoxy radicals, ozone, sulfur oxides,
15    and peroxyacetyl nitrate; and wet and dry deposition to land or water.  In air, benzo[a]pyrene is
16    predominantly adsorbed to particulates  but may also exist as a vapor at high temperatures (ATSDR,
17    19951

18    Uses and Pathways of Exposure
19          There  is no known commercial use for benzo[a]pyrene; it is only produced as a research
20    chemical. Benzo[a]pyrene is ubiquitous  in the environment primarily as a result  of incomplete
21    combustion emissions.  It is found in fossil fuels, crude oils, shale oils, and coal tars (HSDB, 2012). It
22    is released to the environment via both natural sources (such as forest  fires) and  anthropogenic
23    sources including stoves/furnaces burning fossil fuels (especially wood and coal), motor vehicle
24    exhaust, cigarettes, and various industrial combustion processes (ATSDR. 1995).  Benzo[a]pyrene is
25    also found in soot and coal tars. Several  studies have reported that urban run-off from asphalt-
26    paved car parks treated with coats of coal-tar emulsion seal could account for a large proportion of
27    PAHs in many watersheds (Rowe and O'Connor. 2011: Van Metre and Mahler. 2010: Mahler etal..
28    2005). Benzo[a]pyrene exposure can also occur to workers involved in the production of
29    aluminum, coke, graphite, and silicon carbide, and in coal tar distillation.  The major sources of non-
30    occupational exposure are tobacco products, inhalation of polluted air,  ingestion of contaminated
31    food and water, and through cooking processes that involve smoke (HSDB. 2012). Dermal exposure
32    can occur through contact with materials containing soot, tar, or crude  petroleum, including
33    pharmaceutical products containing coal tar, such as coal tar-based shampoos and treatments for
34    eczema and psoriasis (Gal/EPA. 2010: IARC. 2010).
35          It persists for a long period of time in the atmosphere in the particulate phase and is thus
36    efficiently transported over long distances.  It is lipophilic with low water solubility; therefore, once
37    deposited in water or sediments, it adsorbs strongly to sediments and particulate matter and

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

-------
                                                        Toxicological Review ofBenzo[a]pyrene

 1    degrades slowly over several years [Gal/EPA, 2010: GLC, 2007]. Because of its presence in high
 2    concentrations in the waters and sediments of the Great Lakes and St Lawrence river ecosystem, it
 3    is 1 of the 12 level I substances identified and targeted for reduction in the Great Lakes Region
 4    fGLC. 20071.
 5          Most aquatic organisms metabolize benzo[a]pyrene, eliminating it in days, and thus, it is not
 6    expected to bioconcentrate in these organisms; however, several aquatic organisms such as
 7    plankton, oysters, and some fish cannot metabolize benzo[a]pyrene [U.S. EPA. 2010a]. Thus, the
 8    data on benzo[a]pyrene bioconcentration in aquatic organisms varies from low to very high [HSDB.
 9    2012]. Biomagnification of benzo[a]pyrene in the food chain has not been reported [ATSDR, 1995].
10    Additional information on benzo[a]pyrene exposure and chemical properties can be found in
11    Appendix A.

12    Implementation of the 2011 National Research Council Recommendations
13          On December 23, 2011, The Consolidated Appropriations Act, 2012, was signed into law
14    [U.S. Congress. 2011]. The report language included direction to EPA for the IRIS Program related
15    to recommendations provided by the National Research Council (NRC] in their review of EPA's
16    draft IRIS assessment of formaldehyde [NRC, 2011]. The report language included the following:
17          The Agency shall incorporate, as appropriate, based on chemical-specific datasets
18          and biological effects, the recommendations of Chapter 7 of the National Research
19          Council's Review of the Environmental Protection Agency's Draft IRIS Assessment of
20          Formaldehyde into the IRIS process...For draft assessments released in fiscal year
21          2012, the Agency shall include documentation describing how the  Chapter  7
22          recommendations  of the  National Academy  of  Sciences  (NAS]   have been
23          implemented  or  addressed,   including   an  explanation  for  why   certain
24          recommendations were not incorporated.

25          The NRC's recommendations, provided in Chapter 7 of their review report, offered
26    suggestions to EPA for improving the development of IRIS assessments. Consistent with the
27    direction provided by Congress, documentation of how the recommendations from Chapter 7 of the
28    NRC report have been implemented in this assessment is provided in the table below. Where
29    necessary, the documentation includes an explanation for why certain recommendations were not
30    incorporated.
31          The IRIS Program's implementation of the NRC recommendations is following a phased
32    approach that is consistent with the NRC's "Roadmap for Revision" as described in Chapter 7 of the
33    formaldehyde review report.  The NRC stated that "the committee recognizes that the changes
34    suggested would involve a multi-year process and extensive effort by the staff at the National
35    Center for Environmental Assessment and input and review by the EPA Science Advisory Board and
36    others."
37          Phase 1 of implementation has focused on a subset of the short-term recommendations,
38    such as editing and streamlining documents, increasing transparency and clarity, and using more

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

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    tables, figures, and appendices to present information and data in assessments. Phase 1 also
 2    focused on assessments near the end of the development process and close to final posting. The
 3    IRIS benzo[a]pyrene assessment is in Phase 2 and represents a significant advancement in
 4    implementing the NRC recommendations shown in Table F-l in Appendix F. The Program is
 5    implementing all of these recommendations, but recognizes that achieving full and robust
 6    implementation of certain recommendations will be an evolving process with input and feedback
 7    from the public, stakeholders, and external peer review committees. Phase 3 of implementation
 8    will incorporate the longer-term recommendations made by the NRC as outlined in Table F-2 in
 9    Appendix F, including the development of a standardized approach to describe the strength of
10    evidence for noncancer effects.  In May 2014, the NRC released their report reviewing the IRIS
11    assessment development process. As part of this review, the NRC reviewed current methods for
12    evidence-based reviews and made several recommendations with respect to integrating scientific
13    evidence for chemical hazard and dose-response assessments. In their report, the NRC states that
14    EPA should continue to improve its evidence-integration process incrementally and enhance the
15    transparency of its process.  The committee did not offer a preference but suggests that EPA
16    consider which approach best fits its plans for the IRIS process. The NRC recommendations will
17    inform the IRIS Program's efforts in this area going forward. This effort is included in Phase 3 of
18    EPA's implementation plan.

19    Assessments by Other National and International Health Agencies
20          Toxicity information on benzo[a]pyrene has been evaluated by the World Health
21    Organization, Health Canada, the International Agency for Research on Cancer, and the European
22    Union. The results of these assessments are presented in Appendix B. It is important to recognize
23    that these assessments were prepared at different times, for different purposes, using different
24    guidelines and methods,  and that newer studies have been included in the IRIS assessment.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                xv             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofBenzo[a]pyrene
    PREAMBLE TO IRIS TOXICOLOGICAL REVIEWS
 3  1. Scope of the IRIS Program

 4     Soon  after  the EPA was established in
 5  1970, it was at the forefront of developing
 6  risk assessment as a science and applying it
 7  in decisions to protect human health and the
 8  environment. The Clean Air Act, for example,
 9  mandates that the EPA provide "an  ample
10  margin of safety to  protect public health;"
11  the   Safe   Drinking  Water  Act,  that  "no
12  adverse effects on the health of persons  may
13  reasonably be anticipated to occur, allowing
14  an adequate margin  of safety." Accordingly,
15  the  EPA  uses  information on the adverse
16  effects of chemicals and on exposure levels
17  below  which   these   effects   are  not
18  anticipated to occur.
19     IRIS  assessments  critically review the
20  publicly available studies to identify adverse
21  health  effects from  exposure  to  chemicals
22  and   to  characterize   exposure-response
23  relationships.  In terms  set  forth by  the
24  National Research Council [NRC. 1983).  IRIS
25  assessments cover the hazard  identification
26  and dose-response assessment steps of risk
27  assessment, not the exposure assessment or
28  risk   characterization   steps  that  are
29  conducted  by  the  EPA's  program  and
30  regional offices and  by other federal, state,
31  and local health agencies that evaluate  risk
32  in  specific   populations  and  exposure
33  scenarios. IRIS assessments are distinct from
34  and do not address political, economic,  and
35  technical  considerations  that  influence the
36  design and selection  of risk  management
37  alternatives.
38     An IRIS assessment may cover a single
39  chemical,   a   group  of  structurally   or
40  toxicologically   related  chemicals,  or  a
41  complex mixture. These agents  may be found
42  in air,  water,  soil, or sediment Exceptions
43  are chemicals currently used exclusively as
44  pesticides,   ionizing   and   non-ionizing
45  radiation, and  criteria air pollutants listed
46  under  Section 108  of the Clean Air Act
47  (carbon monoxide,  lead, nitrogen  oxides,
48  ozone, particulate matter, and sulfur oxides).
49     Periodically, the IRIS Program asks other
50  EPA  programs and  regions, other  federal
51  agencies, state  health agencies, and the
52  general public to  nominate chemicals and
53  mixtures   for   future   assessment   or
54  reassessment Agents may be considered for
55  reassessment as significant new studies are
56  published. Selection is based on program
57  and  regional office  priorities  and  on
58  availability   of  adequate  information  to
59  evaluate the  potential for adverse  effects.
60  Other  agents  may  also be  assessed in
61  response to an urgent public health need.

62  2. Process for developing and peer
63     reviewing IRIS assessments

64     The  process    for   developing   IRIS
65  assessments  (revised  in  May  2009  and
66  enhanced  in July 2013) involves  critical
67  analysis   of   the    pertinent   studies,
68  opportunities for public input, and multiple
69  levels of scientific review. The EPA revises
70  draft assessments after  each review,  and
71  external drafts and comments become part
72  of the public record (U.S. EPA. 20141
73     Before beginning an assessment, the IRIS
74  Program discusses the scope with other EPA
75  programs and regions to  ensure that the
76  assessment will meet  their needs. Then a
77  public  meeting on problem formulation
78  invites discussion of the  key issues and the
79  studies and analytical approaches that might
80  contribute to their resolution.
81  Step 1.   Development   of   a   draft
82     Toxicological    Review.  The    draft
83     assessment  considers  all   pertinent
84     publicly  available  studies and  applies
85     consistent  criteria  to evaluate  study
86     quality,  identify  health effects, identify
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              xvi             DRAFT—DO NOT CITE OR QUOTE

-------
                                                      Toxicological Review ofBenzo[a]pyrene
 1     mechanistic  events  and  pathways,
 2     integrate the evidence of causation  for
 3     each effect, and derive toxicity values. A
 4     public meeting prior to the integration of
 5     evidence  and  derivation  of  toxicity
 6     values promotes public discussion of the
 7     literature  search,  evidence,  and  key
 8     issues.
 9  Step 2. Internal review by scientists in
10     EPA programs and regions. The draft
11     assessment  is revised  to  address the
12     comments from within the EPA.
13  Step 3. Interagency  science consultation
14     with  other federal agencies and the
15     Executive Offices of the President. The
16     draft assessment  is revised  to address
17     the interagency comments. The  science
18     consultation     draft,      interagency
19     comments, and the EPA's response to
20     major comments  become  part of the
21     public record.
22  Step 4.  Public review  and  comment,
23     followed by external peer review. The
24     EPA releases the  draft  assessment  for
25     public review and  comment A public
26     meeting  provides  an  opportunity  to
27     discuss  the assessment prior to peer
28     review. Then the EPA releases a draft for
29     external peer review. The peer reviewers
30     also  receive written  and  oral  public
31     comments and the peer  review meeting
32     is open to the public. The peer reviewers
33     assess whether the evidence has been
34     assembled  and evaluated  according to
35     guidelines and whether the conclusions
36     are justified by the evidence. The peer
37     review draft, written public  comments,
38     and peer review report  become part of
39     the public record.
40  Step 5. Revision of  draft Toxicological
41     Review and development of draft IRIS
42     summary.   The  draft  assessment  is
43     revised  to  reflect the  peer  review
44     comments, public  comments, and newly
45     published  studies  that are critical to the
46     conclusions  of the assessment  The
47     disposition  of peer review  comments
48     and public comments becomes part of
49     the public record.
50  Step 6. Final EPA review and interagency
51     science discussion with other federal
52     agencies and the  Executive Offices of
53     the President. The draft assessment and
54     summary are revised to address the EPA
55     and interagency comments. The science
56     discussion  draft,   written  interagency
57     comments,  and the EPA's response to
58     major  comments  become part of the
59     public record.
60  Step 7.  Completion  and posting.  The
61     Toxicological Review and IRIS summary
62     are posted  on the IRIS website [http://
63     www.epa.gov/iris/].
64     The   remainder   of   this   Preamble
65  addresses step  1, the development of a draft
66  Toxicological   Review.  IRIS  assessments
67  follow  standard  practices   of   evidence
68  evaluation and peer review, many  of which
69  are discussed in EPA  guidelines [U.S. EPA.
70  2005a. b, 2000b. 1998. 1996. 1991c. 1986a.
71  b) and other methods [U.S. EPA. 2012a.  c,
72  2011. 2006a.  b, 2002. 19941 Transparent
73  application of  scientific  judgment is  of
74  paramount  importance.   To   provide  a
75  harmonized    approach    across   IRIS
76  assessments,  this  Preamble   summarizes
77  concepts   from   these   guidelines   and
78  emphasizes    principles     of    general
79  applicability.

SO  3. Identifying and selecting
81     pertinent studies

82  3.1.  Identifying studies
83     Before beginning an assessment, the EPA
84  conducts  a comprehensive search of the
85  primary scientific  literature. The  literature
86  search  follows  standard  practices  and
87  includes the PubMed and ToxNet databases
88  of the National Library of Medicine, Web of
89  Science,  and other databases  listed in the
90  EPA's   HERO   system    (Health   and
91  Environmental  Research   Online,  http://
92  hero.epa.gov/]. Searches for information on
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              xvii            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1  mechanisms  of  toxicity  are  inherently
 2  specialized and may include studies on other
 3  agents that act through related mechanisms.
 4     Each assessment  specifies the  search
 5  strategies, keywords, and cut-off dates of its
 6  literature  searches.  The  EPA  posts   the
 7  results of the literature search on the  IRIS
 8  website and requests information from the
 9  public on additional  studies  and ongoing
10  research.
11     The EPA also considers studies received
12  through  the IRIS  Submission  Desk  and
13  studies  (typically  unpublished) submitted
14  under the Toxic Substances Control  Act or
15  the  Federal  Insecticide,  Fungicide,  and
16  Rodenticide  Act  Material submitted as
17  Confidential    Business   Information   is
18  considered only if it includes health  and
19  safety data that can be publicly released. If a
20  study that may be critical to the conclusions
21  of the  assessment  has  not  been peer-
22  reviewed,  the  EPA  will  have  it peer-
23  reviewed.
24     The EPA also examines the toxicokinetics
25  of the agent to identify other chemicals (for
26  example, major metabolites of the agent) to
27  include  in  the  assessment  if  adequate
28  information  is available, in order to more
29  fully explain the toxicity of the agent and to
30  suggest  dose   metrics  for   subsequent
31  modeling.
32     In assessments of  chemical  mixtures,
33  mixture studies  are  preferred  for their
34  ability  to  reflect   interactions   among
35  components. The literature search seeks, in
36  decreasing order of preference (U.S. EPA.
37  2000b. 32.2:  1986c. 32.11:
38  -  Studies of the mixture being assessed.
39  -  Studies of a sufficiently similar mixture.
40     In evaluating similarity, the assessment
41     considers the alteration of mixtures in
42     the  environment  through  partitioning
43     and transformation.
44  -  Studies    of    individual   chemical
45     components of the mixture, if there are
46     not  adequate  studies  of sufficiently
47     similar mixtures.
48  3.2. Selecting pertinent epidemiologic
49       studies
50      Study design is the key consideration for
51  selecting pertinent  epidemiologic  studies
52  from the results of the literature search.
53  -   Cohort studies, case-control studies, and
54      some  population-based  surveys  (for
55      example, NHANES) provide the strongest
56      epidemiologic evidence, especially if they
57      collect  information  about  individual
58      exposures and effects.
59  -   Ecological      studies     (geographic
60      correlation studies) relate exposures and
61      effects  by geographic area.  They can
62      provide  strong evidence  if  there are
63      large   exposure   contrasts    between
64      geographic    areas,   relatively   little
65      exposure  variation within study areas,
66      and population migration is limited.
67  -   Case reports  of  high  or  accidental
68      exposure    lack   definition    of  the
69      population  at risk and  the expected
70      number  of  cases. They  can provide
71      information about a rare effect or about
72      the  relevance  of analogous  results  in
73      animals.
74      The   assessment   briefly    reviews
75  ecological studies  and  case  reports  but
76  reports details  only if they  suggest  effects
77  not identified by other studies.

78  3.3. Selecting pertinent experimental
79       studies
80      Exposure   route   is  a   key   design
81  consideration   for   selecting   pertinent
82  experimental   animal  studies  or  human
83  clinical studies.
84  -   Studies of oral,  inhalation,  or dermal
85      exposure  involve passage through  an
86      absorption barrier and are  considered
87      most pertinent to human environmental
88      exposure.
89  -   Injection  or  implantation studies are
90      often considered less pertinent but may
91      provide   valuable   toxicokinetic    or
92      mechanistic information. They also may
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xviii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1     be  useful  for  identifying  effects  in
 2     animals  if deposition  or absorption  is
 3     problematic (for example, for particles
 4     and fibers).
 5     Exposure duration is also a key design
 6  consideration   for   selecting   pertinent
 7  experimental animal studies.
 8  -  Studies of effects from chronic exposure
 9     are most pertinent  to lifetime  human
10     exposure.
11  -  Studies of effects from less-than-chronic
12     exposure  are   pertinent   but   less
13     preferred for identifying  effects  from
14     lifetime  human  exposure.  Such  studies
15     may be  indicative of effects from less-
16     than-lifetime human exposure.
17     Short-duration studies  involving animals
18  or humans  may  provide  toxicokinetic or
19  mechanistic information.
20     For    developmental    toxicity    and
21  reproductive toxicity,  irreversible  effects
22  may  result from a brief exposure during a
23  critical period of development. Accordingly,
24  specialized study designs are used for these
25  effects fU.S. EPA. 2006b. 1998.1996.1991cl

26  4. Evaluating the quality of
27     individual studies

28     After   the    subsets    of   pertinent
29  epidemiologic  and  experimental  studies
30  have  been  selected  from the  literature
31  searches,  the  assessment evaluates  the
32  quality  of  each  individual  study.  This
33  evaluation  considers the  design, methods,
34  conduct, and documentation of each study,
35  but not whether the results  are positive,
36  negative, or null. The objective is to identify
37  the stronger, more informative studies based
38  on   a   uniform   evaluation   of  quality
39  characteristics  across  studies of  similar
40  design.

41  4.1.  Evaluating the quality of
42       epidemiologic studies
43     The assessment evaluates  design  and
44  methodological aspects that can increase or
45  decrease  the   weight  given   to   each
46  epidemiologic study in the overall evaluation
47  [U.S. EPA. 2005a. 1998.1996.1994.1991c):

48  -   Documentation    of    study   design,
49      methods,  population characteristics, and
50      results.
51
52
53
54
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82

83
84
   Definition  and  selection  of the study
   group and comparison group.
   Ascertainment  of  exposure
   chemical or mixture.
             to  the
55  -  Ascertainment of disease or health effect
   Duration of exposure and follow-up and
   adequacy for assessing the occurrence of
   effects.
   Characterization
   critical periods.
of  exposure  during
   Sample size and statistical power to
   detect anticipated effects.
-  Participation  rates  and  potential  for
   selection bias as a result of the achieved
   participation rates.
-  Measurement  error   (can  lead  to
   misclassification  of  exposure,   health
   outcomes, and other factors) and other
   types of information bias.
-  Potential confounding and other sources
   of bias addressed in the study design or
   in the analysis of results. The basis for
   consideration  of  confounding  is  a
   reasonable    expectation   that    the
   confounder  is related to both exposure
   and outcome and is sufficiently prevalent
   to result in bias.
   For developmental toxicity, reproductive
toxicity, neurotoxicity, and cancer  there is
further  guidance  on  the   nuances  of
evaluating epidemiologic  studies of these
effects (U.S. EPA. 2005a. 1998.1996.1991c).

4.2.  Evaluating the quality of
     experimental studies
85     The assessment  evaluates  design  and
86  methodological aspects that can increase or
87  decrease  the   weight  given   to   each
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xix             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1  experimental animal study, in vitro study, or
 2  human clinical study (U.S. EPA. 2005a. 1998.
 3  1996.  1991c).  Research involving  human
 4  subjects  is  considered only if conducted
 5  according to ethical principles.
 6  -  Documentation of study design, animals
 7     or study population, methods, basic data,
 8     and results.
 9  -  Nature  of the assay and validity for its
10     intended purpose.
11  -  Characterization of the nature and extent
12     of impurities and contaminants  of  the
13     administered chemical or mixture.
14  -  Characterization  of  dose   and   dosing
15     regimen (including age at exposure) and
16     their adequacy to elicit adverse effects,
17     including latent effects.
18  -  Sample  sizes and  statistical  power to
19     detect dose-related differences or trends.
20  -  Ascertainment of survival,  vital signs,
21     disease or effects, and cause of death.
22  -  Control  of  other  variables that could
23     influence the occurrence of effects.
24     The assessment uses statistical tests to
25  evaluate  whether the observations may be
26  due to chance. The standard for  determining
27  statistical significance  of a response is a
28  trend test or comparison of outcomes in  the
29  exposed groups against those of concurrent
30  controls. In  some situations, examination of
31  historical  control  data from  the  same
32  laboratory within a few years of the study
33  may improve the analysis. For an uncommon
34  effect  that  is  not  statistically significant
35  compared    with   concurrent  controls,
36  historical controls may show that the effect
37  is  unlikely  to  be  due  to  chance.  For a
38  response that appears significant against a
39  concurrent control response that is unusual,
40  historical  controls  may offer  a different
41  interpretation [U.S. EPA. 2005a. §2.2.2.1.3).
42     For developmental toxicity, reproductive
43  toxicity, neurotoxicity, and cancer there is
44  further   guidance  on   the   nuances  of
45  evaluating  experimental studies of these
46  effects  [U.S. EPA. 2005a. 1998. 1996.  1991c).
47  In  multigeneration  studies,  agents  that
48  produce developmental effects at doses that
49  are not toxic to the maternal animal are of
50  special concern. Effects that occur at doses
51  associated with mild  maternal toxicity are
52  not assumed to result only from maternal
53  toxicity. Moreover, maternal effects may be
54  reversible, while effects on the offspring may
55  be permanent fU.S. EPA.  1998. §3.1.2.4.5.4:
56  1991c. 33.1.1.41.

57  4.3. Reporting study results
58      The assessment uses  evidence tables to
59  present  the  design  and  key  results  of
60  pertinent studies.  There may  be  separate
61  tables  for each site of toxicity or  type of
62  study.
63      If a large  number of studies observe the
64  same effect,  the assessment considers the
65  study quality characteristics in this section
66  to identify the strongest studies or types of
67  study. The tables present details from these
68  studies and  the  assessment  explains  the
69  reasons for  not reporting details  of other
70  studies or groups  of studies that do not add
71  new information.  Supplemental information
72  provides    references    to   all    studies
73  considered, including those not summarized
74  in the tables.
75      The assessment discusses strengths and
76  limitations that affect the interpretation of
77  each study. If the interpretation of a study in
78  the assessment differs from that of the study
79  authors, the assessment discusses the basis
80  for the difference.
81      As  a check   on  the  selection  and
82  evaluation of pertinent studies, the EPA asks
83  peer reviewers to identify studies that were
84  not adequately considered.

85  5.  Evaluating the overall evidence
86      of  each effect

87  5.1. Concepts of causal inference
88      For each  health effect, the assessment
89  evaluates the  evidence   as  a whole  to
90  determine whether it is reasonable to infer a
91  causal  association  between exposure to the
              This document is a draft for review purposes only and does not constitute Agency policy.
                                                xx             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1  agent and the occurrence of the effect This
 2  inference  is  based  on  information  from
 3  pertinent human studies, animal studies, and
 4  mechanistic  studies of  adequate  quality.
 5  Positive, negative, and null results are given
 6  weight according to study quality.
 7      Causal   inference  involves   scientific
 8  judgment,  and  the  considerations  are
 9  nuanced   and   complex.   Several  health
10  agencies  have  developed frameworks  for
11  causal inference,  among  them  the  U.S.
12  Surgeon General  [CDC. 2004:  HEW. 19641.
13  the International Agency for  Research  on
14  Cancer  [IARC.  2006]  ,  the  Institute   of
15  Medicine [IOM. 2008).  and the  U.S.  EPA.
16  f2010b.31.6:   2005a. 32.51.      Although
17  developed   for   different   purposes,  the
18  frameworks  are   similar in  nature  and
19  provide  an   established   structure   and
20  language   for   causal    inference.   Each
21  considers  aspects of an association that
22  suggest causation, discussed by Hill [1965]
23  and elaborated by Rothman and Greenland
24  f!998j. and U.S. EPA f2005a. §2.2.1.7: 1994.
25  §2.2.1.71
26  Strength of  association: The finding of a
27      large   relative  risk   with    narrow
28      confidence  intervals  strongly suggests
29      that an association is not due to  chance,
30      bias,  or other factors. Modest relative
31      risks, however, may reflect  a small range
32      of exposures, an agent of low potency, an
33      increase  in an  effect that is  common,
34      exposure  misclassification,  or  other
35      sources of bias.
36  Consistency of association:  An inference of
37      causation is  strengthened if elevated
38      risks   are  observed  in  independent
39      studies  of  different  populations  and
40      exposure scenarios.  Reproducibility  of
41      findings constitutes one of  the strongest
42      arguments  for  causation. Discordant
43      results  sometimes reflect differences  in
44      study design,  exposure, or confounding
45      factors.
46  Specificity  of  association:  As  originally
47      intended,  this  refers  to  one  cause
48      associated  with one  effect  Current
49     understanding that many agents cause
50     multiple effects and many effects have
51     multiple  causes  make  this   a  less
52     informative aspect of  causation, unless
53     the  effect  is  rare or  unlikely to have
54     multiple causes.
55  Temporal    relationship:   A    causal
56     interpretation  requires  that exposure
57     precede development of the effect
58  Biologic   gradient  (exposure-response
59     relationship):      Exposure-response
60     relationships strongly suggest causation.
61     A monotonic  increase  is not the only
62     pattern consistent with causation.  The
63     presence   of   an  exposure-response
64     gradient also  weighs  against bias  and
65     confounding  as  the   source  of  an
66     association.
67  Biologic  plausibility:  An  inference   of
68     causation  is   strengthened   by  data
69     demonstrating    plausible     biologic
70     mechanisms,  if  available.   Plausibility
71     may  reflect subjective prior beliefs if
72     there is insufficient understanding of the
73     biologic process involved.
74  Coherence: An inference  of  causation is
75     strengthened by supportive results from
76     animal    experiments,    toxicokinetic
77     studies, and short-term tests. Coherence
78     may  also  be  found in other  lines  of
79     evidence,   such  as  changing  disease
80     patterns in the population.
81  "Natural   experiments":   A   change   in
82     exposure that brings about a change in
83     disease  frequency   provides   strong
84     evidence,  as it  tests the hypothesis of
85     causation.  An example would be  an
86     intervention to  reduce  exposure in  the
87     workplace   or  environment  that   is
88     followed by a reduction of an  adverse
89     effect.
90  Analogy:   Information    on   structural
91     analogues  or  on chemicals that induce
92     similar  mechanistic events can  provide
93     insight into causation.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxi             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1     These considerations are consistent with
 2  guidelines  for  systematic  reviews  that
 3  evaluate the quality and weight of evidence.
 4  Confidence is increased if the magnitude of
 5  effect is  large, if there is evidence of an
 6  exposure-response  relationship,  or if an
 7  association was observed  and the plausible
 8  biases would tend to decrease the magnitude
 9  of  the  reported  effect.  Confidence  is
10  decreased     for     study    limitations,
11  inconsistency  of results,  indirectness  of
12  evidence,  imprecision,  or  reporting  bias
13  [Guyattetal.. 2008b: Guyattetal.. 2008a).

14  5.2.  Evaluating evidence in humans

15     For each effect, the assessment evaluates
16  the evidence from the epidemiologic studies
17  as a whole. The  objective is to determine
18  whether  a credible  association has been
19  observed and, if so, whether that association
20  is  consistent with causation. In doing  this,
21  the   assessment    explores    alternative
22  explanations  (such  as chance, bias,  and
23  confounding) and draws a conclusion about
24  whether these  alternatives can satisfactorily
25  explain any observed association.
26     To    make   clear   how   much   the
27  epidemiologic  evidence contributes  to  the
28  overall   weight  of   the   evidence,   the
29  assessment may select a standard descriptor
30  to characterize the  epidemiologic evidence
31  of association between exposure to the agent
32  and occurrence of a health effect.
33  Sufficient epidemiologic  evidence  of an
34     association consistent with causation:
35     The  evidence  establishes  a   causal
36     association    for   which    alternative
37     explanations such as chance, bias, and
38     confounding  can  be  ruled out with
39     reasonable confidence.
40  Suggestive epidemiologic evidence of an
41     association consistent with causation:
42     The   evidence   suggests   a   causal
43     association   but   chance,   bias,    or
44     confounding  cannot  be  ruled  out  as
45     explaining the association.
46  Inadequate epidemiologic  evidence   to
47     infer a causal association: The available
48     studies  do  not permit  a  conclusion
49     regarding the presence or absence of an
50     association.
51  Epidemiologic evidence consistent with no
52     causal association:  Several adequate
53     studies covering the full range of human
54     exposures  and  considering  susceptible
55     populations,  and for which  alternative
56     explanations   such   as    bias   and
57     confounding can  be ruled out,  are
58     mutually  consistent in not  finding an
59     association.

60  5.3.  Evaluating evidence in animals

61     For each effect, the assessment evaluates
62  the evidence from the animal experiments as
63  a whole to determine  the extent to which
64  they  indicate  a potential  for  effects in
65  humans.  Consistent  results  across  various
66  species and strains increase confidence that
67  similar  results  would  occur in  humans.
68  Several  concepts discussed by Hill [1965]
69  are pertinent to  the weight of experimental
70  results:   consistency  of  response,   dose-
71  response relationships, strength of response,
72  biologic  plausibility,  and coherence  [U.S.
73  EPA. 2005a. §2.2.1.7: 1994. Appendix C]}.
74     In  weighing evidence  from  multiple
75  experiments,    U.S.    EPA   f2005a). S2.5
76  distinguishes:
77  Conflicting evidence (that is, mixed positive
78     and negative results in the same sex and
79     strain using a similar  study protocol)
80     from
81  Differing  results (that is, positive results
82     and  negative results  are in  different
83     sexes  or strains  or use different  study
84     protocols).
85     Negative or null results do not invalidate
86  positive results  in  a different experimental
87  system.   The  EPA  regards  all as  valid
88  observations and looks to explain differing
89  results using  mechanistic information (for
90  example,    physiologic    or    metabolic
91  differences   across    test  systems)  or
92  methodological  differences  (for example,
93  relative sensitivity of the tests, differences in
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1  dose  levels,  insufficient  sample  size,  or
 2  timing of dosing or data collection).
 3     It is well established that there are
 4  critical periods for some developmental and
 5  reproductive   effects   [U.S.   EPA.  2006b.
 6  2005a. b,  1998.  1996.  1991cl Accordingly,
 7  the assessment determines whether critical
 8  periods have been adequately investigated.
 9  Similarly,   the   assessment   determines
10  whether  the  database   is  adequate  to
11  evaluate other critical sites and effects.
12     In   evaluating  evidence   of   genetic
13  toxicity:
14  -  Demonstration  of   gene   mutations,
15     chromosome aberrations, or aneuploidy
16     in humans  or experimental mammals
17     (in vivo)   provides   the    strongest
18     evidence.
19  -  This is followed  by positive results in
20     lower  organisms  or  in  cultured  cells
21     (in vitro) or for other genetic events.
22  -  Negative results carry less weight, partly
23     because  they  cannot   exclude   the
24     possibility of effects  in  other tissues
25     flARC. 20061
26     For germ-cell mutagenicity, the EPA has
27  defined categories of evidence, ranging from
28  positive   results  of   human   germ-cell
29  mutagenicity  to  negative  results  for  all
30  effects of concern (U.S. EPA. 1986a. §2.3).

31  5.4.  Evaluating mechanistic data
32     Mechanistic  data  can  be  useful  in
33  answering several questions.
34  -  The biologic  plausibility  of  a causal
35     interpretation of human studies.
36  -  The generalizability of animal studies to
37     humans.
38  -  The    susceptibility    of   particular
39     populations or lifestages.
40     The focus of the analysis is to describe, if
41  possible, mechanistic pathways that lead to a
42  health effect These pathways encompass:
43  -  Toxicokinetic processes of  absorption,
44     distribution,      metabolism,      and
45     elimination that lead to the formation of
46     an active agent and its presence at the
47     site of initial biologic interaction.
48  -  Toxicodynamic  processes that lead to a
49     health effect at this or another site (also
50     known as a mode of action}.
51     For each effect, the assessment discusses
52  the available  information  on its  modes of
53  action and associated key events (key events
54  being  empirically  observable,  necessary
55  precursor steps or biologic markers  of such
56  steps; mode of action being a series of key
57  events   involving  interaction with  cells,
58  operational  and   anatomic  changes,  and
59  resulting in  disease). Pertinent information
60  may  also come from studies of metabolites
61  or of compounds that are structurally similar
62  or that act  through  similar mechanisms.
63  Information  on  mode  of  action  is  not
64  required for a conclusion  that the agent is
65  causally related to an effect (U.S. EPA. 2005a.
66  §2.51
67     The   assessment   addresses   several
68  questions about each hypothesized mode of
69  action fU.S. EPA. 2005a. §2.4.3.41
70  1) Is the hypothesized  mode  of action
71     sufficiently supported in test animals?
72     Strong support for  a  key event being
73     necessary to a  mode of action can come
74     from  experimental  challenge   to  the
75     hypothesized mode  of action, in which
76     studies  that   suppress  a  key event
77     observe   suppression   of  the   effect
78     Support  for   a  mode  of  action  is
79     meaningfully strengthened by consistent
80     results in different experimental  models,
81     much   more   so   than  by  replicate
82     experiments  in the same model.  The
83     assessment  may   consider   various
84     aspects of causation in addressing this
85     question.
86  2) Is the hypothesized  mode  of action
87     relevant to humans? The  assessment
88     reviews the key events to identify critical
89     similarities and differences between the
90     test    animals    and   humans.   Site
91     concordance  is not assumed between
92     animals and humans, though it may hold
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              xxiii            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1     for certain  effects  or  modes  of action.
 2     Information   suggesting  quantitative
 3     differences in doses where effects would
 4     occur   in  animals   or  humans   is
 5     considered   in    the    dose-response
 6     analysis.  Current   levels  of  human
 7     exposure are not used to rule out human
 8     relevance, as IRIS  assessments may be
 9     used in evaluating new or unforeseen
10     circumstances  that may entail higher
11     exposures.
12  3) Which  populations or lifestages can
13     be  particularly  susceptible to  the
14     hypothesized  mode  of action?  The
15     assessment  reviews the key  events to
16     identify populations and lifestages that
17     might be susceptible to their occurrence.
18     Quantitative differences may  result in
19     separate toxicity values for susceptible
20     populations or lifestages.
21     The assessment discusses the  likelihood
22  that  an  agent  operates through multiple
23  modes of action. An uneven level of support
24  for different modes of action can reflect
25  disproportionate      resources      spent
26  investigating     them     [U.S.	EPA.
27  2005a. §2.4.3.3). It should be noted that in
28  clinical reviews, the credibility of a series of
29  studies  is reduced if evidence is  limited to
30  studies  funded  by one interested sector
31  [Guyattetal.. 2008a).
32     For  cancer,  the assessment  evaluates
33  evidence of a mutagenic mode of action to
34  guide  extrapolation to lower doses  and
35  consideration of susceptible lifestages. Key
36  data include the ability of  the agent or a
37  metabolite  to react with  or bind to  DNA,
38  positive results in  multiple test systems, or
39  similar  properties and  structure-activity
40  relationships to mutagenic carcinogens [U.S.
41  EPA. 2005a.§2.3.5).

42  5.5.  Characterizing the overall weight
43       of the evidence

44     After evaluating the human, animal, and
45  mechanistic evidence pertinent to an effect,
46  the assessment answers the question:  Does
47  the agent  cause the  adverse effect [NRG.
48  2009. 19831? In doing this, the assessment
49  develops  a  narrative  that  integrates the
50  evidence pertinent to causation. To provide
51  clarity  and   consistency,   the   narrative
52  includes  a standard  hazard descriptor. For
53  example, the following  standard descriptors
54  combine  epidemiologic, experimental, and
55  mechanistic evidence of carcinogenicity [U.S.
56  EPA. 2005a. §2.51
57  Carcinogenic   to   humans:   There   is
58     convincing epidemiologic  evidence of a
59     causal association   (that  is,  there  is
60     reasonable    confidence    that   the
61     association cannot be fully explained by
62     chance, bias, or confounding); or there is
63     strong human evidence of cancer  or its
64     precursors,  extensive animal  evidence,
65     identification of key precursor events in
66     animals,  and strong evidence  that they
67     are anticipated to occur in humans.
68  Likely to be carcinogenic to humans: The
69     evidence   demonstrates   a   potential
70     hazard to humans but does not meet the
71     criteria for carcinogenic. There may be a
72     plausible    association    in    humans,
73     multiple positive  results in animals, or a
74     combination of human, animal, or  other
75     experimental evidence.
76  Suggestive   evidence   of  carcinogenic
77     potential: The evidence raises concern
78     for effects in humans but is not sufficient
79     for   a   stronger   conclusion.   This
80     descriptor  covers a range of evidence,
81     from  a  positive  result  in  the  only
82     available study to a single positive result
83     in an extensive database that includes
84     negative results in other species.
85  Inadequate    information    to   assess
86     carcinogenic   potential:   No    other
87     descriptors  apply.  Conflicting  evidence
88     can    be    classified   as   inadequate
89     information if all  positive  results are
90     opposed by negative studies of  equal
91     quality in  the  same  sex and strain.
92     Differing  results,  however,   can  be
93     classified  as suggestive evidence  or as
94     likely to be carcinogenic.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxiv             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1  Wot likely to be carcinogenic to humans:
 2     There is robust evidence for  concluding
 3     that there is no basis for concern. There
 4     may be no effects in both sexes of at least
 5     two appropriate animal species; positive
 6     animal  results  and strong,  consistent
 7     evidence that  each  mode  of action  in
 8     animals does not operate in humans;  or
 9     convincing evidence that effects are not
10     likely by a particular exposure route  or
11     below a defined dose.
12     Multiple descriptors may be used if there
13  is evidence that carcinogenic effects differ by
14  dose  range  or exposure  route  [U.S. EPA.
15  2005a.32.5).
16     Another example of standard descriptors
17  comes  from  EPA's  Integrated   Science
18  Assessments, which  evaluate causation  for
19  the effects  of the  criteria  pollutants  in
20  ambient air [U.S. EPA. 2010b. 31.61.
21  Causal relationship: Sufficient evidence  to
22     conclude   that  there  is  a   causal
23     relationship.    Observational   studies
24     cannot   be  explained  by   plausible
25     alternatives, or they are  supported by
26     other lines of evidence, for example,
27     animal    studies   or   mechanistic
28     information.
29  Likely  to   be  a  causal  relationship:
30     Sufficient  evidence   that  a   causal
31     relationship  is  likely, but  important
32     uncertainties   remain.  For  example,
33     observational     studies   show    an
34     association but coexposures are difficult
35     to address or other lines of evidence are
36     limited   or  inconsistent;  or  multiple
37     animal    studies     from    different
38     laboratories  demonstrate  effects  and
39     there are limited or no human data.
40  Suggestive of a   causal  relationship:  At
41     least one   high-quality  epidemiologic
42     study shows an association but  other
43     studies are inconsistent.
44  Inadequate to infer a causal relationship:
45     The studies do not permit a  conclusion
46     regarding the presence or absence of an
47     association.
48  Not  likely to be a  causal relationship:
49     Several adequate  studies,  covering the
50     full  range  of  human  exposure  and
51     considering susceptible populations, are
52     mutually  consistent in not showing an
53     effect at any level of exposure.
54     The EPA  is investigating and may on a
55  trial  basis use  these  or other  standard
56  descriptors   to  characterize  the  overall
57  weight of the evidence for effects other than
58  cancer.

59  6. Selecting studies for derivation
60     of toxicity values
61
For each effect where there is  credible
62  evidence of an association with the agent,
63  the assessment derives  toxicity values if
64  there    are   suitable   epidemiologic   or
65  experimental data. The decision to  derive
66  toxicity values may be linked to the hazard
67  descriptor.
68     Dose-response    analysis     requires
69  quantitative measures of dose and response.
70  Then, other factors being equal:
71  -  Epidemiologic studies are preferred over
72     animal studies, if quantitative measures
73     of exposure are available and effects can
74     be attributed to the agent.
75  -  Among  experimental animal models,
76     those that respond most like humans are
77     preferred,  if  the   comparability  of
78     response can be determined.
79  -  Studies   by   a   route    of  human
80     environmental exposure are  preferred,
81     although a validated toxicokinetic model
82     can be   used to  extrapolate  across
83     exposure routes.
84  -  Studies of longer exposure duration and
85     follow-up  are  preferred, to  minimize
86     uncertainty about  whether effects  are
87     representative of lifetime exposure.
88  -  Studies with multiple exposure levels are
89     preferred  for their ability  to provide
90     information  about the  shape  of  the
91     exposure-response curve.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxv             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1  -   Studies with adequate power to detect
 2      effects  at lower  exposure  levels are
 3      preferred, to minimize  the  extent  of
 4      extrapolation to  levels  found  in the
 5      environment.
 6      Studies with nonmonotonic  exposure-
 7  response relationships are not necessarily
 8  excluded from the  analysis.  A diminished
 9  effect at higher exposure levels may be
10  satisfactorily  explained by factors such  as
11  competing toxicity, saturation of absorption
12  or  metabolism,  exposure misclassification,
13  or selection bias.
14      If a large  number of studies are suitable
15  for  dose-response analysis, the assessment
16  considers  the study characteristics in this
17  section  to  focus on the most informative
18  data. The  assessment explains  the reasons
19  for  not analyzing other groups of studies. As
20  a check on the selection of studies for dose-
21  response  analysis,   the   EPA  asks  peer
22  reviewers to  identify studies  that were not
23  adequately considered.

24  7.  Deriving toxicity values

25  7.1. General framework for dose-
26       response analysis
27      The EPA uses a two-step  approach that
28  distinguishes analysis of the observed dose-
29  response data from inferences about lower
30  doses fU.S. EPA. 2005a. 531
31      Within the observed range, the preferred
32  approach is to use modeling to incorporate a
33  wide range  of data  into the analysis. The
34  modeling  yields a point  of departure (an
35  exposure level near the  lower end  of the
36  observed    range,    without   significant
37  extrapolation to lower doses; see Sections
38  7.2  and 7.3).
39      Extrapolation to  lower doses considers
40  what is known about the modes of action for
41  each effect  (see Sections 7.4 and  7.5).  If
42  response estimates  at lower  doses  are not
43  required, an alternative is to derive reference
44  values, which are calculated by applying
45  factors to the point of departure in order  to
46  account  for  sources  of  uncertainty  and
47  variability (see Section 7.6).
48     For  a group of agents that induce an
49  effect through a common mode of action, the
50  dose-response analysis may derive a relative
51  potency factor for  each agent. A full dose-
52  response analysis is conducted for one well-
53  studied index chemical in the group, then the
54  potencies of other members are expressed in
55  relative terms based on relative toxic effects,
56  relative  absorption   or  metabolic rates,
57  quantitative structure-activity relationships,
58  or receptor binding characteristics (U.S. EPA.
59  2005a. 33.2.6: 2000b. 34.41.
60     Increasingly,  EPA   is  basing  toxicity
61  values  on  combined  analyses of  multiple
62  data sets or multiple  responses. The  EPA
63  also   considers   multiple  dose-response
64  approaches  if they  can be supported by
65  robust data.

66  7.2.  Modeling dose to sites of biologic
67       effects
68     The preferred  approach for  analysis  of
69  dose is toxicokinetic modeling because of its
70  ability to incorporate a wide range  of data.
71  The preferred dose metric would refer to the
72  active agent at the site of its biologic effect or
73  to a  close,  reliable surrogate measure. The
74  active  agent may  be the   administered
75  chemical or a metabolite. Confidence in the
76  use of a toxicokinetic model depends on the
77  robustness of its validation process and on
78  the results of sensitivity analyses (U.S. EPA.
79  2006a: 2005a. §3.1: 1994. §4.3).
80     Because  toxicokinetic   modeling  can
81  require  many parameters  and  more  data
82  than are typically available, the EPA has
83  developed standard approaches that can be
84  applied to typical data sets. These standard
85  approaches also facilitate comparison across
86  exposure patterns and species.
87  -  Intermittent   study  exposures   are
88     standardized to a daily average over the
89     duration of exposure. For chronic effects,
90     daily exposures are averaged over the
91     lifespan.  Exposures  during a  critical
92     period, however, are not averaged over a
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxvi             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1     longer duration (U.S. EPA. 2005a. 53.1.1:
 2     1991c.S3.21
 3  -  Doses are  standardized  to  equivalent
 4     human terms to facilitate comparison of
 5     results from different species.
 6     -   Oral  doses  are  scaled allometrically
 7         using mg/kg3/4-d as the equivalent
 8         dose    metric    across    species.
 9         Allometric   scaling   pertains   to
10         equivalence  across  species,  not
11         across lifestages, and is not used to
12         scale  doses from adult humans  or
13         mature animals to infants or children
14         fU.S.  EPA. 2011: 2005a. 33.1.31.
15     -   Inhalation  exposures   are   scaled
16         using dosimetry models that  apply
17         species-specific   physiologic   and
18         anatomic   factors  and   consider
19         whether the effect occurs at the site
20         of first contact or  after  systemic
21         circulation    [U.S.    EPA.   2012a:
22         1994.331.
23     It can be informative to convert  doses
24  across exposure routes. If this is  done, the
25  assessment  describes the underlying data,
26  algorithms,  and  assumptions  [U.S.  EPA.
27  2005a. 33.1.41.
28     In the absence  of study-specific data on,
29  for example, intake rates or body weight, the
30  EPA has developed recommended values for
31  use in  dose-response  analysis  [U.S. EPA,
32  19881

33  7.3.  Modeling response in the range
34       of observation
35     Toxicodynamic  ("biologically   based")
36  modeling can incorporate data on biologic
37  processes leading to an effect. Such models
38  require sufficient data to ascertain a mode of
39  action and to quantitatively support model
40  parameters  associated  with its key events.
41  Because   different  models   may  provide
42  equivalent fits  to  the  observed  data but
43  diverge substantially at lower doses, critical
44  biologic  parameters should  be  measured
45  from laboratory studies, not by model fitting.
46  Confidence in the  use of a toxicodynamic
47  model depends on the  robustness of its
48  validation process  and on  the results  of
49  sensitivity  analyses.  Peer review  of the
50  scientific basis and performance of a model
51  is essential [U.S. EPA. 2005a. 33.2.21.
52     Because  toxicodynamic  modeling can
53  require  many   parameters   and   more
54  knowledge  and  data than  are typically
55  available, the EPA has developed a standard
56  set of  empirical  ("curve-fitting")  models
57  (http://www.epa.gov/ncea/bmds/) that can
58  be applied  to typical data  sets, including
59  those that are nonlinear. The EPA has also
60  developed   guidance  on  modeling  dose-
61  response data, assessing model  fit, selecting
62  suitable  models,  and reporting modeling
63  results   (U.S.   EPA.   2012c1.   Additional
64  judgment or alternative analyses are used if
65  the procedure fails to yield reliable  results,
66  for example, if the fit is poor, modeling may
67  be restricted to the lower doses, especially if
68  there is competing toxicity at higher  doses
69  fU.S. EPA. 2005a. 33.2.31.
70     Modeling is used to  derive a point  of
71  departure fU.S. EPA. 2012c: 2005a. 33.2.41.
72  (See  Section 7.6 for alternatives if a point of
73  departure cannot be derived by modeling.)

74  -  If linear  extrapolation is used, selection
75     of a response level corresponding to the
76     point  of  departure   is   not   highly
77     influential, so standard values near the
78     low end of the  observable range are
79     generally used (for example, 10% extra
80     risk for  cancer bioassay data,  1% for
81     epidemiologic  data,   lower  for   rare
82     cancers).

83  -  For   nonlinear    approaches,     both
84     statistical  and biologic  considerations
85     are taken into account.
86
87
88
89
90
91
92
93
94
95
For dichotomous data, a response
level of 10% extra risk is generally
used for minimally adverse effects,
5% or lower for more severe effects.
For continuous data, a response level
is  ideally based on  an established
definition of biologic significance. In
the absence of such definition, one
control standard deviation from the
control  mean  is  often  used  for
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxvii            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1
 2
 3
minimally adverse  effects, one-half
standard deviation  for more  severe
effects.
 4     The point of departure is the 95% lower
 5  bound on  the dose  associated  with the
 6  selected response level.

 7  7.4.  Extrapolating to lower doses and
 8       response levels
 9     The purpose of extrapolating to lower
10  doses is to  estimate responses at exposures
11  below  the   observed   data.   Low-dose
12  extrapolation, typically used for cancer data,
13  considers what is known  about  modes  of
14  action [U.S.  EPA. 2005a. §3.3.1 and §3.3.2).
15  1) If a biologically based  model has been
16     developed and validated  for the agent,
17     extrapolation may use  the fitted model
18     below the observed range if significant
19     model uncertainty can be  ruled out with
20     reasonable confidence.
21  2) Linear extrapolation is used if the dose-
22     response  curve is  expected to have a
23     linear  component  below the  point  of
24     departure. This includes:
25     -   Agents or their metabolites that are
26         DNA-reactive   and   have   direct
27         mutagenic activity.
28     -   Agents or their metabolites for which
29         human  exposures  or  body burdens
30         are  near doses associated with key
31         events leading to an effect
32     Linear extrapolation is also used when
33     data are insufficient to establish mode of
34     action and when scientifically plausible.
35     The result of linear  extrapolation  is
36     described by an oral slope factor or an
37     inhalation unit risk, which is the slope of
38     the dose-response curve at lower doses
39     or concentrations, respectively.
40  3) Nonlinear   models   are  used   for
41     extrapolation if there are  sufficient data
42     to ascertain the mode  of action and  to
43     conclude that it is not linear at lower
44     doses,  and   the   agent  does   not
45     demonstrate mutagenic or other activity
46     consistent with linearity at lower doses.
47     Nonlinear approaches generally should
48     not  be used  in cases  where mode  of
49     action  has  not  been ascertained.  If
50     nonlinear extrapolation is  appropriate
51     but no model is developed, an alternative
52     is to calculate reference values.
53  4) Both linear and  nonlinear approaches
54     may be used if there are multiple modes
55     of action. For example, modeling to a low
56     response   level  can  be  useful  for
57     estimating the response at doses where a
58     high-dose mode of action would be less
59     important
60     If linear extrapolation  is  used, the
61  assessment  develops  a  candidate  slope
62  factor or unit risk for each suitable data set.
63  These results are arrayed, using common
64  dose metrics, to  show  the distribution  of
65  relative  potency across  various effects and
66  experimental systems. The assessment then
67  derives or selects an overall slope factor and
68  an overall unit risk for the agent, considering
69  the  various  dose-response  analyses, the
70  study preferences discussed in Section  6,
71  and the possibility of basing a more robust
72  result on multiple data sets.

73  7.5.  Considering susceptible
74       populations and lifestages
75     The  assessment analyzes the available
76  information on populations and lifestages
77  that  may be particularly susceptible  to each
78  effect. A tiered  approach is used [U.S. EPA.
79  2005a.53.5).
80  5) If an epidemiologic  or   experimental
81     study reports quantitative results for a
82     susceptible population or lifestage, these
83     data  are  analyzed  to  derive separate
84     toxicity    values    for    susceptible
85     individuals.
86  6) If data on risk-related parameters  allow
87     comparison of the  general  population
88     and  susceptible individuals,  these data
89     are used to adjust the general-population
90     toxicity   values   for  application   to
91     susceptible individuals.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              xxviii            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
 1  7)  In the absence of chemical-specific data,
 2      the  EPA  has developed  age-dependent
 3      adjustment factors for early-life exposure
 4      to  potential  carcinogens  that  have  a
 5      mutagenic  mode of  action.  There  is
 6      evidence  of early-life  susceptibility  to
 7      various carcinogenic agents,  but most
 8      epidemiologic   studies   and   cancer
 9      bioassays   do  not  include  early-life
10      exposure.  To address the potential for
11      early-life    susceptibility,    the   EPA
12      recommends [U.S. EPA. 2005b. §5):
13      -   10-fold adjustment for  exposures
14         before age 2 years.
15      -   3-fold   adjustment  for   exposures
16         between ages 2 and 16 years.

17  7.6. Reference values and uncertainty
18       factors
19      An oral reference dose or  an  inhalation
20  reference concentration is an estimate of an
21  exposure    (including    in   susceptible
22  subgroups) that is  likely to be without  an
23  appreciable  risk  of adverse health effects
24  over  a  lifetime  fU.S.  EPA.  2002.34.21.
25  Reference values are typically  calculated for
26  effects other than cancer and  for suspected
27  carcinogens if a well characterized mode of
28  action indicates that a necessary  key event
29  does  not occur below  a specific  dose.
30  Reference values  provide no information
31  about risks at higher exposure levels.
32      The  assessment  characterizes  effects
33  that form the basis for reference values as
34  adverse, considered to  be adverse, or  a
35  precursor  to   an  adverse   effect.  For
36  developmental    toxicity,    reproductive
37  toxicity, and neurotoxicity there is guidance
38  on adverse effects and their biologic markers
39  fU.S. EPA. 1998.1996.1991c1.
40      To   account   for   uncertainty   and
41  variability in  the derivation  of  a  lifetime
42  human exposure  where adverse effects are
43  not anticipated to occur, reference values are
44  calculated by applying a series  of uncertainty
45  factors to the point of departure. If a point of
46  departure cannot be derived by modeling, a
47  no-observed-adverse-effect   level   or   a
48  lowest-observed-adverse-effect level is used
49  instead. The assessment discusses scientific
50  considerations  involving  several  areas  of
51  variability or uncertainty.
52  Human variation: The assessment accounts
53      for variation in susceptibility across the
54      human population  and the possibility
55      that  the available  data  may  not  be
56      representative of individuals  who  are
57      most susceptible to the effect. A factor of
58      10 is generally used to account for this
59      variation. This factor is reduced only if
60      the  point  of departure  is  derived  or
61      adjusted  specifically   for  susceptible
62      individuals (not for a general population
63      that includes  both susceptible and non-
64      susceptible    individuals)   (U.S.  EPA,
65      2002. §4.4.5:    1998.  §4.2:    1996. §4:
66      1994. §4.3.9.1: 1991c. 33.41.
67  Animal-to-human extrapolation: If animal
68      results are  used to  make inferences
69      about humans, the  assessment adjusts
70      for cross-species  differences. These may
71      arise from differences in  toxicokinetics
72      or toxicodynamics.  Accordingly,  if  the
73      point of departure  is standardized  to
74      equivalent human terms or is based  on
75      toxicokinetic  or dosimetry modeling, a
76      factor of 101/2 (rounded to 3) is applied
77      to account for the remaining uncertainty
78      involving      toxicokinetic      and
79      toxicodynamic   differences.    If    a
80      biologically based model adjusts fully for
81      toxicokinetic     and     toxicodynamic
82      differences across species, this factor is
83      not used. In most other cases, a factor of
84      10   is  applied   fU.S.   EPA.   2011:
85      2002. §4.4.5:    1998.  §4.2:    1996. §4:
86      1994. §4.3.9.1: 1991c. §3.4).
87  Adverse-effect   level  to   no-observed-
88      adverse-effect level: If  a point  of
89      departure   is based   on  a  lowest-
90      observed-adverse-effect    level,    the
91      assessment must infer  a dose where
92      such effects are not expected. This can be
93      a  matter of great uncertainty, especially
94      if there is no evidence  available at lower
95      doses.  A factor  of 10  is  applied  to
96      account for the uncertainty in making
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxix            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
 1     this inference.  A factor  other  than 10
 2     may   be  used,  depending   on  the
 3     magnitude and  nature of the response
 4     and the  shape of  the  dose-response
 5     curve fU.S. EPA. 2002. §4.4.5: 1998. §4.2:
 6     1996. §4: 1994. §4.3.9.1: 1991c.§3.41
 7  Subchronic-to-chronic exposure: If a point
 8     of  departure  is based  on subchronic
 9     studies,   the    assessment  considers
10     whether lifetime exposure  could have
11     effects at  lower levels of exposure.  A
12     factor  of 10 is applied to account for the
13     uncertainty in using subchronic studies
14     to  make   inferences  about  lifetime
15     exposure.  This  factor  may   also  be
16     applied    for    developmental    or
17     reproductive effects if exposure covered
18     less than the full critical period. A factor
19     other than 10 may be used,  depending
20     on  the duration of the studies and the
21     nature of the response (U.S. EPA. 2002.
22     §4.4.5: 1998. §4.2: 1994. §4.3.9.11
23  Incomplete  database:   If  an  incomplete
24     database  raises  concern that  further
25     studies might identify a more sensitive
26     effect,  organ  system,  or lifestage, the
27     assessment  may   apply  a  database
28     uncertainty    factor     [U.S.     EPA.
29     2002. §4.4.5:    1998. §4.2:    1996. §4:
30     1994. §4.3.9.1:  1991c. §3.4).  The size of
31     the factor depends on the nature of the
32     database  deficiency.  For example, the
33     EPA typically follows the suggestion that
34     a  factor  of 10 be  applied  if both  a
35     prenatal  toxicity  study  and  a  two-
36     generation  reproduction   study  are
37     missing and a factor of 101/2 if either is
38     missing [U.S. EPA. 2002. §4.4.5).
39     In  this  way,  the  assessment derives
40  candidate  values for each suitable  data set
41  and effect that is credibly associated with the
42  agent  These  results  are   arrayed,  using
43  common dose metrics, to show where effects
44  occur across a range of exposures [U.S. EPA.
45  1994. §4.3.9).
46     The assessment derives  or  selects an
47  organ- or  system-specific reference value for
48  each organ or system affected by the agent.
49  The  assessment explains the rationale  for
50  each organ/system-specific reference value
51  (based on,  for example, the highest quality
52  studies, the most  sensitive outcome, or a
53  clustering  of values).  By  providing these
54  organ/system-specific reference values, IRIS
55  assessments     facilitate      subsequent
56  cumulative risk assessments  that consider
57  the combined effect of multiple agents acting
58  at a  common   site  or  through common
59  mechanisms [NRG. 2009).
60     The assessment then selects an overall
61  reference  dose  and  an overall reference
62  concentration  for  the agent to represent
63  lifetime  human  exposure levels  where
64  effects are  not  anticipated to  occur. This is
65  generally the most sensitive organ/system-
66  specific    reference    value,     though
67  consideration    of   study   quality   and
68  confidence  in  each  value  may lead  to a
69  different selection.

70  7.7.  Confidence and uncertainty in the
71       reference values
72     The  assessment  selects  a  standard
73  descriptor  to  characterize  the  level   of
74  confidence  in each reference value, based on
75  the likelihood that the value would change
76  with further testing. Confidence in reference
77  values is based on quality of the studies used
78  and completeness of the database, with more
79  weight  given  to the latter.  The  level  of
80  confidence  is increased for  reference values
81  based on human data supported by animal
82  data  [U.S. EPA. 1994. §4.3.9.2).
83  High confidence: The reference value is not
84     likely  to  change  with  further  testing,
85     except for mechanistic studies that might
86     affect  the  interpretation  of prior  test
87     results.
88  Medium confidence: This is a matter  of
89     judgment,   between  high   and   low
90     confidence.
91  Low  confidence:  The  reference value  is
92     especially  vulnerable  to  change  with
93     further testing.
              This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxx            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofBenzo[a]pyrene
 1     These   criteria  are  consistent  with
 2  guidelines  for   systematic  reviews  that
 3  evaluate the quality of evidence. These also
 4  focus on whether further research would be
 5  likely to change confidence in the estimate of
 6  effect [Guyattetal.. 2008b).
 7     All  assessments discuss the significant
 8  uncertainties encountered in  the  analysis.
 9  The    EPA   provides    guidance    on
10  characterization  of uncertainty [U.S. EPA.
11  2005a. §3.6].  For example, the discussion
12  distinguishes model  uncertainty  (lack  of
13  knowledge   about the  most  appropriate
14  experimental  or  analytic  model)  and
15  parameter uncertainty  (lack of knowledge
16  about  the   parameters  of  a   model).
17  Assessments also discuss human variation
18  (interpersonal   differences   in   biologic
19  susceptibility or  in exposures  that modify
20  the effects of the agent).
21

22
References
23  CDC   (Centers  for  Disease  Control  and
24         Prevention).   (2004).  The  health
25         consequences of smoking:  A report
26         of the Surgeon General. Washington,
27         DC:  U.S. Department of Health and
28         Human                   Services.
29         http://www.surgeongeneral.gov/lib
30         rary/smokingconsequences/
31  Guyatt. GH: Oxman. AD: Kunz. R:  Vist. GE:
32         Falck-Ytter.  Y:  Schiinemann.   HI.
33         (2008a). GRADE: What is "quality of
34         evidence" and why is it important to
35         clinicians? [Review]. BMJ 336: 995-
36         998.
37         http://dx.doi.org/10.1136/bmi.3949
38         0.551019.BE
39  Guyatt. GH: Oxman. AD: Vist. GE:  Kunz.  R:
40         Falck-Ytter.  Y:  Alonso-Coello.   P:
41         Schiinemann.  HI. (2008b).  GRADE:
42         An emerging consensus on rating
43         quality of evidence and strength  of
44         recommendations.  BMJ  336: 924-
45         926.
46         http://dx.doi.org/10.1136/bmi.3948
47         9.470347.AD
48  HEW (U.S. Department of Health, Education
49         and Welfare). (1964).  Smoking and
50         health:   Report  of  the  advisory
51         committee to the surgeon general of
52         the     public    health     service.
53         Washington, DC: U.S. Department of
54         Health,   Education,   and  Welfare.
55         http://profiles.nlm.nih.gov/ps/retrie
56         ve/ResourceMetadata/NNBBMO
57  Hill.  AB.  (1965).  The  environment  and
58         disease: Association  or  causation?
59         ProcRSocMed 58: 295-300.
60  IARC (International Agency for Research on
61         Cancer).  (2006).  Preamble to the
62         IARC  monographs.  Lyon,  France.
63         http://monographs.iarc.fr/ENG/Pre
64         amble/
65  IOM   (Institute   of  Medicine).   (2008).
66         Improving the presumptive disability
67         decision-making    process     for
68         veterans.  In JM Samet; CC Bodurow
69         (Eds.).  Washington,  DC:  National
70         Academies                  Press.
71         http://www.nap.edu/openbook.php
72         ?record id=11908
73  NRC  (National Research  Council).  (1983).
74         Risk  assessment  in  the  federal
75         government:  Managing the process.
76         Washington, DC: National Academies
77         Press.
78         http://www.nap.edu/openbook.php
79         ?record id=366&page=Rl
80  NRC  (National Research  Council).  (2009).
81         Science  and  decisions:  Advancing
82         risk  assessment.  Washington,  DC:
83         National     Academies      Press.
84         http://www.nap.edu/catalog/12209
85         .html
86  Rothman. KT: Greenland. S. (1998).  Modern
87         epidemiology (2nd ed.). Philadelphia,
88         PA: Lippincott, Williams, & Wilkins.
89  U.S.  EPA  (U.S.  Environmental  Protection
90         Agency).  (1986a). Guidelines  for
91         mutagenicity    risk   assessment
92         (EPA/630/R-98/003).  Washington,
93         DC:  U.S.  Environmental  Protection
94         Agency, Risk  Assessment  Forum.
95         http://www.epa.gov/iris/backgrd.ht
96         ml
              This document is a draft for review purposes only and does not constitute Agency policy.
                                              xxxi            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
U.S.  EPA  (U.S. Environmental  Protection
       Agency). (1986b). Guidelines for the
       health risk assessment of chemical
       mixtures.     (EPA/630/R-98/002).
       Washington, DC: U.S. Environmental
       Protection Agency, Risk Assessment
       Forum.
       http://cfpub.epa.gov/ncea/cfm/reco
       rdisplay.cfm?deid=22567
U.S.  EPA  (U.S. Environmental  Protection
U.S.
  Agency). (1986c). Guidelines for the
  health risk assessment of chemical
  mixtures. Fed Reg 51: 34014-34025.
EPA (U.S. Environmental Protection
       Agency). (1988). Recommendations
       for and documentation of biological
       values for use  in risk  assessment.
       (EPA/600/6-87/008).    Cincinnati,
       OH: U.S.  Environmental Protection
       Agency,    National   Center    for
       Environmental        Assessment
       http://cfpub.epa.gov/ncea/cfm/reco
       rdisplay.cfm?deid=34855
U.S.  EPA  (U.S.  Environmental  Protection
       Agency).  (199 Ic).   Guidelines  for
       developmental    toxicity     risk
       assessment.  (EPA/600/FR-91/001).
       Washington,  DC: U.S. Environmental
       Protection Agency, Risk Assessment
       Forum.
       http://www.epa.gov/raf/publication
       s/guidelines-dev-toxicity-risk-
       assessmenthtm
U.S.  EPA  (U.S.  Environmental  Protection
       Agency).   (1994).   Methods   for
       derivation  of inhalation  reference
       concentrations  and  application  of
       inhalation  dosimetry. (EPA/600/8-
       90/066F).  Research Triangle  Park,
       NC: U.S.  Environmental Protection
       Agency, Environmental Criteria and
       Assessment                 Office.
       http://cfpub.epa.gov/ncea/cfm/reco
       rdisplay.cfm?deid=71993
U.S.  EPA  (U.S.  Environmental  Protection
       Agency).   (1996).  Guidelines   for
       reproductive     toxicity      risk
       assessment.   (EPA/630/R-96/009).
       Washington,  DC: U.S. Environmental
       Protection Agency, Risk Assessment
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
                                           U.S.
       Forum.
       http://www.epa.gov/raf/publication
       s/pdfs/REPR051.PDF
     EPA  (U.S. Environmental  Protection
       Agency). (1998). Guidelines for ne
       urotoxicity     risk     assessment
       (EPA/630/R-95/001F). Washington,
       DC:  U.S. Environmental Protection
       Agency,  Risk  Assessment  Forum.
       http://www.epa.gov/raf/publication
       s/pdfs/NEUROTOX.PDF
U.S.  EPA  (U.S. Environmental  Protection
       Agency).  (2000b).  Supplementary
       guidance for conducting health risk
       assessment  of  chemical  mixtures.
       (EPA/630/R-00/002).  Washington,
       DC:  U.S. Environmental Protection
       Agency,  Risk  Assessment  Forum.
       http://cfpub.epa.gov/ncea/cfm/reco
       rdisplay.cfm?deid=20533
U.S.  EPA  (U.S. Environmental  Protection
                                           U.S.
                                           U.S.
       Agency).  (2002). A review of the
       reference   dose   and   reference
       concentration           processes.
       (EPA/630/P-02/002F). Washington,
       DC:  U.S.  Environmental  Protection
       Agency,  Risk  Assessment  Forum.
       http://cfpub.epa.gov/ncea/cfm/reco
       rdisplay.cfm?deid=51717
     EPA  (U.S.  Environmental  Protection
       Agency).  (2005a).   Guidelines  for
       carcinogen     risk    assessment
       (EPA/630/P-03/001F). Washington,
       DC:  U.S.  Environmental  Protection
       Agency,  Risk  Assessment  Forum.
       http: / /www. epa. gov/cancer guidelin
       eg
     EPA
	  (U.S.  Environmental  Protection
 Agency).   (2005b).   Supplemental
 guidance  for assessing susceptibility
 from   early-life    exposure    to
 carcinogens.  In US  Environmental
 Protection Agency, Risk Assessment
 Forum       (pp.      1125-1133).
 (EPA/630/R-03/003F). Washington,
 DC: U.S.  Environmental  Protection
 Agency,   Risk  Assessment  Forum.
 http://www.epa.gov/cancerguidelin
 es/guidelines-carcinogen-
 supplementhtm
              This document is a draft for review purposes only and does not constitute Agency policy.
                                             xxxii            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
U.S.  EPA  (U.S. Environmental  Protection
       Agency). (2006a). Approaches for the
       application of physiologically based
       pharmacokinetic (PBPK) models and
       supporting data in risk assessment
       (Final   Report)    [EPA   Report].
       (EPA/600/R-05/043F). Washington,
       DC:  U.S. Environmental  Protection
       Agency,    National   Center   for
       Environmental         assessment
       http://cfpub.epa.gov/ncea/cfm/reco
       rdisplay.cfm?deid=157668
U.S.  EPA  (U.S. Environmental  Protection
U.S.
U.S.
U.S.
  Agency).  (2006b). A framework for
  assessing     health     risk     of
  environmental    exposures     to
  children.     (EPA/600/R-05/093F).
  Washington, DC:  U.S. Environmental
  Protection Agency, National  Center
  for    Environmental    Assessment
  http://cfpub.epa.gov/ncea/cfm/reco
  rdisplay.cfm?deid=158363
EPA (U.S.  Environmental Protection
  Agency). (2010b). Integrated science
  assessment  for  carbon  monoxide
  [EPA     Report].     (EPA/600/R-
  09/019F). Research Triangle  Park,
  NC.
  http://cfpub.epa.gov/ncea/cfm/reco
  rdisplay.cfm?deid=218686
EPA (U.S.  Environmental Protection
  Agency). (2011). Recommended use
  of body weight 3/4  as the default
  method in derivation  of  the  oral
  reference                    dose.
  (EPA/100/R11/0001).  Washington,
  DC: U.S.  Environmental Protection
  Agency, Risk  Assessment  Forum.
  http://www.epa.gov/raf/publication
  s/interspecies-extrapolation.htm
EPA (U.S. Environmental Protection
  Agency).   (2012a).   Advances   in
  inhalation    gas   dosimetry   for
  derivation    of    a    reference
  concentration (RFC) and use in risk
  assessment.   (EPA/600/R-12/044).
  Washington,                  DC.
  http://cfpub.epa.gov/ncea/cfm/reco
  rdisplay.cfm?deid=244650
50  U
51
52
53
54
55
56
57
58  U
59
60
61
62
63
                                             S.  EPA  (U.S.  Environmental  Protection
                                             S.
  Agency).  (2012b). Benchmark dose
  technical  guidance.   (EPA/100/R-
  12/001).  Washington,  DC:  Risk
  Assessment                Forum.
  http://www.epa.gov/raf/publication
  s/pdfs/benchmark dose guidance.p
  df
EPA (U.S.  Environmental Protection
  Agency).  (2014).  EPAs Integrated
  Risk      Information      System:
  Assessment  development   process
  [EPA   Report].  Washington,   DC.
  http://epa.gov/iris/process.htm
                                            64
                                            65  August 2013
              This document is a draft for review purposes only and does not constitute Agency policy.
                                             xxxiii           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
 1
 2    EXECUTIVE SUMMARY
 3                                  Occurrence and Health Effects

 4                 Benzo[a]pyrene is a  five-ring polycyclic aromatic  hydrocarbon (PAH).
 5          Benzo[a]pyrene (along with other PAHs) is released into the  atmosphere as a
 6          component  of smoke  from forest  fires, industrial processes, vehicle  exhaust,
 7          cigarettes, and through the burning of fuel (such as wood, coal,  and petroleum
 8          products).   Oral exposure to benzo[a]pyrene can  occur by eating certain food
 9          products,  such  as  charred  meats, where benzo[a]pyrene is formed during the
10          cooking  process   or  by  eating  foods  grown  in  areas   contaminated  with
11          benzo[a]pyrene (from the air and soil).  Dermal exposure may occur from contact
12          with soils or materials that contain soot, tar, or crude petroleum  products or by
13          using certain pharmaceutical products containing coal tars, such as those used to
14          treat the skin conditions, eczema and psoriasis. The magnitude of human exposure
15          to benzo[a]pyrene  and other PAHs depends on factors such  as lifestyle (e.g., diet,
16          tobacco smoking), occupation, and living conditions (e.g., urban versus rural setting,
17          domestic heating, and cooking methods).
18                 Animal studies demonstrate  that exposure to  benzo[a]pyrene  may  be
19          associated  with  developmental,  reproductive, and immunological effects.   In
20          addition, epidemiology studies involving exposure to PAH mixtures have  reported
21          associations  between  internal  biomarkers  of  exposure   to benzo[a]pyrene
22          (benzo[a]pyrene diol epoxide-DNA adducts) and adverse birth outcomes (including
23          reduced birth weight, postnatal  body  weight,  and head  circumference)  and
24          decreased fertility.
25                 Studies in  multiple  animal species  demonstrate that benzo[a]pyrene  is
26          carcinogenic at multiple tumor sites (alimentary tract,  liver, kidney, respiratory
27          tract,  pharynx, and skin) by all routes of exposure. In addition,  there is strong
28          evidence of carcinogenicity in occupations  involving exposure  to  PAH  mixtures
29          containing benzo[a]pyrene, such as aluminum production, chimney sweeping, coal
30          gasification, coal-tar distillation, coke production, iron  and  steel  founding, and
31          paving and  roofing with coal tar  pitch.   An increasing number of occupational
32          studies  demonstrate a positive  exposure-response  relationship with cumulative
33          benzo[a]pyrene exposure and lung cancer.
34

35    Effects Other Than Cancer Observed Following Oral Exposure
36          In animals, oral exposure to benzo[a]pyrene has been shown to result in developmental
37    toxicity, reproductive toxicity, and immunotoxicity. Developmental effects in rats and mice include
38    neurobehavioral changes and cardiovascular effects following gestational  exposures. Reproductive
39    and immune effects include decreased sperm counts, ovary weight, and follicle numbers, and
40    decreased immunoglobulin and B-cell numbers and thymus weight following oral exposures in
41    adult animals. In humans, benzo[a]pyrene exposure occurs in conjunction with other PAHs and, as


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

-------
                                                         Toxicological Review ofBenzo[a]pyrene
 1    such, attributing the observed effects to benzo[a]pyrene is complicated. However, human studies
 2    report associations between particular health endpoints and internal measures of exposure, such as
 3    benzo[a]pyrene-deoxyribonucleic acid (DNA) adducts, or external measures of benzo[a]pyrene
 4    exposure. Overall, the human studies report developmental and reproductive effects that are
 5    generally analogous to those observed in animals, and provide qualitative, supportive evidence for
 6    hazards associated with benzo[a]pyrene exposure.

 7    Oral Reference Dose (RfD) for Effects Other Than Cancer
 8          Organ- or system-specific RfDs were derived for hazards associated with benzo[a]pyrene
 9    exposure where data were amenable (see Table ES-1). These organ- or system-specific reference
10    values may be useful for subsequent cumulative risk assessments that consider the combined effect
11    of multiple agents acting at a common site.
12          Developmental toxicity, represented by neurobehavioral changes following neonatal
13    exposure, was chosen as the basis for the proposed overall oral RfD as the available data indicate
14    that neurobehavioral changes represent the most sensitive hazard of benzo[a]pyrene exposure.
15    The neurodevelopmental study by Chen etal. [2012] was used to derive the RfD. The endpoint of
16    altered anxiety-like behavior, as measured in the  elevated plus maze, was selected as the critical
17    effect due to the sensitivity of this endpoint and the observed dose-response relationship of effects
18    across dose groups. Benchmark dose (BMD) modeling was utilized to derive the BMDLiso of
19    0.09 mg/kg-day that was used as the point of departure (POD) for RfD derivation.
20          The proposed overall RfD was calculated by dividing the POD for altered anxiety-like
21    behavior as measured in the elevated plus maze by a composite uncertainty factor (UF) of 300 to
22    account for the  extrapolation from animals to humans (10), for interindividual differences in
23    human susceptibility (10), and for deficiencies in  the toxicity database (3).

24          Table ES-1. Organ/system-specific RfDs and proposed overall RfD for
25          benzo[a]pyrene
Effect
Developmental
Reproductive
Immunological
Proposed Overall RfD
Basis
Neurobehavioral changes
Gavage neurodevelopmental study in rats (postnatal days
[PNDs] 5-11)
Chen etal. (2012)

Decreased ovary weight
Gavage subchronic (60 d) reproductive toxicity study in rats
Xu etal. (2010)

Decreased thymus weight and serum IgM
Gavage subchronic (35 d) study in rats
De Jong etal. (1999)

Developmental toxicity
RfD
(mg/kg-d)
3 x ID'4
4 x 10"4
2 x 10"3
3 x ID'4
Confidence
Medium
Medium
Low
Medium
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                xxxv            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    Confidence in the Overall Oral RfD
 2          The overall confidence in the RfD is medium. Confidence in the principal study [Chenetal..
 3    2012] is medium-to-high. The design, conduct, and reporting of this neurodevelopmental study
 4    was good and a wide variety of neurotoxicity endpoints were measured. Some informative
 5    experimental details were, however, omitted including the sensitivity of some assays at the
 6    indicated developmental ages and lack of reporting gender-specific data for all outcomes. Several
 7    subchronic and developmental studies covering a wide variety of endpoints are also available;
 8    however, the lack of a multigeneration toxicity study with exposure throughout development is not
 9    available. Therefore, confidence in the database is medium.

10    Effects Other Than Cancer Observed Following Inhalation Exposure
11          In animals, inhalation exposure to benzo[a]pyrene has been shown to result in
12    developmental and reproductive toxicity. Studies in rats following inhalation exposure show
13    decreased fetal survival and brain effects in offspring, and decreased testes weight and sperm
14    counts in adult animals. Overall, the available human PAH mixtures studies report developmental
15    and reproductive effects that are generally analogous to those observed in animals, and provide
16    qualitative, supportive evidence for the hazards associated with benzo[a]pyrene exposure.

17    Inhalation Reference Concentration (RfC) for Effects Other Than Cancer
18          An attempt was made to derive organ- or system-specific RfCs for hazards associated with
19    benzo[a]pyrene exposure where data were amenable (see Table ES-2). These organ- or system-
20    specific reference values may be useful for subsequent cumulative risk assessments that consider
21    the combined effect of multiple agents acting at a common site.
22          Developmental toxicity, represented by decreased fetal survival, was chosen as the basis for
23    the proposed inhalation RfC as the available data indicate that developmental effects represent a
24    sensitive hazard of benzo[a]pyrene  exposure. The developmental  inhalation study in rats by
25    Archibong et al. [2002] and the observed decreased fetal survival following exposure to
26    benzo[a]pyrene on gestation days (CDs] 11-20 were used to derive the overall RfC. The lowest-
27    observed-adverse-effect level (LOAEL] of 25 |J.g/m3 based on decreased fetal survival was selected
28    as the POD. The LOAEL was adjusted to account for the discontinuous daily exposure to derive the
29    PODADj and the human equivalent concentration (HEC] was calculated from the PODADj by
30    multiplying by the regional deposited dose ratio (RDDRER] for extrarespiratory (i.e., systemic]
31    effects, as described in Methods/or Derivation of Inhalation Reference Concentrations and
32    Application of Inhalation Dosimetry  (U.S.  EPA. 1994]. These adjustments resulted in a PODnEc of
33    4.6 [ig/m3, which was used as the POD for RfC derivation.
34          The RfC was calculated by dividing the POD by a composite UF of 3,000 to account for
35    toxicodynamic differences between animals and humans (3], interindividual differences in human
36    susceptibility (10], LOAEL-to-no-observed-adverse-effect level (NOAEL] extrapolation (10], and
37    deficiencies in the toxicity database (10].
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               xxxvi           DRAFT—DO  NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
 1
 2
       Table ES-2. Organ/system-specific RfCs and proposed overall RfC for
       benzo[a]pyrene
Effect
Developmental
Reproductive
Proposed Overall
RfC
Basis
Decreased fetal survival
Developmental toxicity study in rats (GDs 11-20)
Archibongetal. (2002)

Reductions in testes weight and sperm parameters
Subchronic (60 d) reproductive toxicity study in rats
(Archibong et al. (2008); Ramesh et al. (2008))

Developmental toxicity
RfC(mg/m3)
2 x 10"6
Not
calculated3
2 x 10"6
Confidence
Low-medium
NA
Low-medium
 3
 4

 5
 6
 7
 8
 9
10
11
12
13

14
15
16
17
18
19
20
21
22
23
24
25
aNot calculated due to UF >3,000.

Confidence in the Overall Inhalation RfC
       The overall confidence in the RfC is low-to-medium. Confidence in the principal study
[Archibong et al., 2002] is medium. The conduct and reporting of this developmental inhalation
study were adequate; however, a NOAEL was not identified. Confidence in the database is low due
to the lack of a multigeneration toxicity study and the lack of information on varied toxicity
endpoints following subchronic and chronic inhalation exposure. However, confidence in the RfC is
bolstered by consistent systemic effects observed by the oral route (including reproductive and
developmental effects) and similar effects observed in human populations exposed to PAH
mixtures.

Evidence for  Human Carcinogenicity
       Under EPA's Guidelines for Carcinogen Risk Assessment [U.S. EPA. 2005a). benzo[a]pyrene is
"carcinogenic  to humans" based on strong and consistent evidence in animals and humans. The
evidence includes an extensive number of studies demonstrating carcinogenicity in multiple animal
species exposed via all routes of administration and increased cancer risks, particularly in the lung
and skin, in humans exposed to different PAH mixtures containing benzo[a]pyrene. Mechanistic
studies provide strong supporting evidence that links the metabolism of benzo[a]pyrene to DNA-
reactive agents with key mutational events in genes that can lead to tumor development These
events include formation of specific DNA adducts and characteristic mutations in oncogenes and
tumor suppressor genes that have been observed in humans exposed to PAH mixtures. This
combination of human, animal, and mechanistic evidence provides the basis for characterizing
benzo[a]pyrene  as "carcinogenic to humans."
26    Quantitative Estimate of Carcinogenic Risk From Oral Exposure
27          Lifetime oral exposure to benzo[a]pyrene has been associated with forestomach, liver, oral
28    cavity, jejunum or duodenum, and auditory canal tumors in male and female Wistar rats,

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

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    forestomach tumors in male and female Sprague-Dawley rats, and forestomach, esophagus, tongue,
 2    and larynx tumors in female B6C3Fi mice (male mice were not tested).  Less-than-lifetime oral
 3    exposure to benzo[a]pyrene has also been associated with forestomach tumors in more than
 4    10 additional bioassays with several strains of mice. The Kroese etal. [2001] and Beland and Gulp
 5    [1998] studies were selected as the best available studies for dose-response analysis and
 6    extrapolation to lifetime cancer risk following oral exposure to benzo[a]pyrene. These studies
 7    included histological examinations for tumors in many different tissues, contained three exposure
 8    levels and controls, contained adequate numbers of animals per dose group (~50/sex/group],
 9    treated animals for up to 2 years, and included detailed reporting methods and results (including
10    individual animal data].
11          Time-weighted, average daily doses were converted to human equivalent doses (HEDs] on
12    the basis of (body weight]3/4 scaling [U.S. EPA. 1992]. EPA then used the multistage-Weibull model
13    for the derivation of the oral slope factor. This model was used because it incorporates the time at
14    which death-with-tumor occurred and can account for differences in mortality observed between
15    the exposure groups. Using linear extrapolation from the BMDLio, human equivalent oral slope
16    factors were derived for each gender/tumor site combination (slope factor = 0.1/BMDLi0] reported
17    by Kroese etal. [2001] and Beland and Gulp [1998]. The oral slope factor of 1 per mg/kg-day
18    based on the tumor response in the alimentary tract (forestomach, esophagus, tongue, and larynx]
19    of female B6C3Fi mice [Beland and Gulp. 1998] was selected as the factor with the highest value
20    (most sensitive] among a range of slope factors derived.

21    Quantitative Estimate of Carcinogenic Risk From Inhalation Exposure
22          Inhalation exposure to benzo[a]pyrene has been associated with squamous cell neoplasia in
23    the larynx, pharynx, trachea, nasal cavity, esophagus, and forestomach of male Syrian golden
24    hamsters exposed for up to 130 weeks to benzo[a]pyrene condensed onto NaCl particles [Thyssen
25    etal.. 1981]. Supportive evidence for the carcinogenicity of inhaled benzo[a]pyrene comes from
26    additional studies with hamsters exposed to benzo[a]pyrene via intratracheal instillation. The
27    Thyssen et al. [1981] bioassay represents the only study of lifetime exposure to inhaled
28    benzo[a]pyrene.
29          A time-to-tumor dose-response model was fit to the time-weighted average [TWA]
30    continuous exposure concentrations and the individual animal incidence data for the overall
31    incidence of tumors in the upper respiratory tract or pharynx.  The inhalation unit risk of
32    6 x lO-4 per ng/m3 was calculated by linear extrapolation (slope factor = 0.1/BMCLio] from a
33    BMCLio of 0.16 mg/m3for the occurrence of upper respiratory and upper digestive tract tumors in
34    male hamsters chronically exposed by inhalation to benzo[a]pyrene [Thyssen et al.. 1981].

35    Quantitative Estimate of Carcinogenic Risk From Dermal Exposure
36          Skin cancer in humans has been documented to result from occupational exposure to
37    complex mixtures of PAHs including benzo[a]pyrene, such as coal tar, coal tar pitches, unrefined

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

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    mineral oils, shale oils, and soot In animal models, numerous dermal bioassays have demonstrated
 2    an increased incidence of skin tumors with increasing dermal exposure of benzo[a]pyrene in all
 3    species tested (mice, rabbits, rats, and guinea pigs), although most benzo[a]pyrene bioassays have
 4    been conducted in mice. Due to the evidence supporting a hazard from exposure to benzo[a]pyrene
 5    by the dermal route (see Section 1.1.5) and the availability of quantitative information, a cancer
 6    slope factor for the dermal route was developed. The analysis in this assessment focuses on
 7    lifetime carcinogenicity bioassays in several  strains of mice demonstrating increasing incidence of
 8    benign and malignant skin tumors following repeated dermal exposure to benzo[a]pyrene.
 9          The National Institute for Occupational Safety and Health (NIOSH) study (Sivaketal.. 1997:
10    NIOSH, 1989)  was selected as the best available study for dose-response analysis and extrapolation
11    to lifetime cancer risk following dermal exposure to benzo[a]pyrene. This study used three
12    exposure levels that highlighted the low-dose region and reported a number of attributes not
13    available for the older studies, including single housing of mice, blinded assessment of tumor status,
14    and time of appearance of tumors for each animal.
15          This mouse skin tumor incidence data was modeled using the multistage-Weibull model.
16    The resulting BMDLio was adjusted for interspecies differences by allometric scaling. The dermal
17    slope factor of 0.006 per ng/day was calculated by linear extrapolation (slope factor =
18    0.1/BMDLio-HEo) from the human equivalent POD for the occurrence of skin tumors in male mice
19    exposed dermally to benzo[a]pyrene for 104 weeks. As this slope factor has been developed for a
20    local effect, it is not intended to estimate systemic risk of cancer following dermal absorption of
21    benzo[a]pyrene into the systemic circulation.

22    Susceptible Populations and  Lifestages
23          Benzo[a]pyrene has been determined to be carcinogenic by a mutagenic mode of action in
24    this assessment According to the Supplemental Guidance for Assessing Susceptibility from Early Life
25    Exposure to Carcinogens (U.S. EPA.  2005b). individuals exposed during early life to carcinogens with
26    a mutagenic mode of action are assumed to have an increased risk for cancer. The oral slope factor
27    of 1 per mg/kg-day, inhalation unit risk of 0.0006 per [ig/m3, and dermal slope factor of 0.006 per
28    Mg/day for benzo[a]pyrene, calculated from data applicable to adult exposures, do not reflect
29    presumed early life susceptibility to this chemical. Although some chemical-specific data exist for
30    benzo[a]pyrene that demonstrate increased  early life susceptibility to cancer, these data were not
31    considered sufficient to develop separate risk estimates for childhood exposure. In the absence of
32    adequate chemical-specific data to evaluate differences in age-specific susceptibility, the
33    Supplemental Guidance (U.S. EPA. 2005b) recommends that age-dependent adjustment factors
34    (ADAFs) be applied in estimating cancer risk. The ADAFs are 10- and 3-fold adjustments that are
35    combined with age specific exposure estimates when estimating cancer risks from early  life
36    (<16 years of age) exposures to benzo[a]pyrene.
37          Regarding effects other than cancer, there are epidemiological studies that report
38    associations between developmental effects  (decreased postnatal growth, decreased head
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                xxxix            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    circumference, and neurodevelopmental delays), reproductive effects and internal biomarkers of
 2    exposure to benzo[a]pyrene. Studies in animals also indicate alterations in neurological
 3    development and heightened susceptibility to reproductive effects following gestational or early
 4    postnatal exposure to benzo[a]pyrene.

 5    Key Issues Addressed in Assessment
 6          The overall RfD and RfC were developed based on effects observed following exposure to
 7    benzo[a]pyrene during a critical window of development The derivation of a general population
 8    toxicity value based on exposure during development has implications regarding the evaluation of
 9    populations exposed outside of the developmental period and the averaging of exposure to
10    durations outside of the critical window of susceptibility. Discussion of these considerations is
11    provided in Sections 2.1.5 and 2.2.5.
12          The dermal slope factor was developed based on data in animals. Because there is no
13    established methodology for extrapolating dermal toxicity from animals to humans, several
14    alternative approaches were evaluated (see Appendix D in Supplemental Information).  Allometric
15    scaling using body weight to the % power was selected based on known species differences in
16    dermal metabolism and penetration of benzo[a]pyrene.
17
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 xl             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
      LITERATURE SEARCH STRATEGY| STUDY SELECTION
 3          The literature search strategy used to identify primary, peer-reviewed literature pertaining
 4    to benzo[a]pyrene was conducted using the databases listed in Table LS-1 (see Appendix C for the
 5    complete list of keywords used). References from previous assessments by the U.S. Environmental
 6    Protection Agency (EPA) and other national and international health organizations were also
 7    examined.  EPA conducted a comprehensive, systematic literature search for benzo[a]pyrene
 8    through February, 2012. In addition, a search of the online database PubMed was conducted for the
 9    timeframe January 2012 through August 2014, to ensure inclusion of critical studies published
10    since the initial literature search.
11
       Table LS-1. Summary of the search strategy employed for benzo[a]pyrene
       Database
                                              Keywords
      Pubmed
      Toxcenter
      Toxline
           Chemical name (CASRN): benzo[a]pyrene (50-32-8)
           Synonyms:  benzo[d,e,f]chrysene, benzo[def]chrysene, 3,4-benzopyrene, 1,2-benzpyrene, 3,4-bp,
           benz(a)pyrene, 3,4-benzpyren, 3,4-benzpyrene, 4,5-benzpyrene, 6,7-benzopyrene, benzopirene,
           benzo(alpha)pyrene

           Standard toxicology search keywords
           Toxicity (including duration, effects to children and occupational exposure); development;
           reproduction; teratogenicity; exposure routes; pharmacokinetics; toxicokinetics; metabolism; body
           fluids; endocrinology; carcinogenicity; genotoxicity; antagonists; inhibitors
      TSCATS
      ChemID
      Chemfinder
      CCRIS
      HSDB
      GENETOX
      RTECS
           Searched by CASRNs and chemical names (including synonyms)
12
13
14
15
16
17
18
19
20
21
22
aPrimary and secondary keywords used for the Pubmed, Toxcenter, and Toxline databases can be found in the
 Supplemental Information.

       Figure LS-1 depicts the literature search, study selection strategy, and number of references
obtained at each stage of literature screening. Approximately 20,700 references were identified
with the initial keyword search. Based on a secondary keyword search followed by a preliminary
manual screen of titles or abstracts by a toxicologist, approximately 1,190 references were
identified that provided information potentially relevant to characterizing the health effects or
physical and chemical properties of benzo[a]pyrene. A more detailed manual review of titles,
abstracts, and/or papers was then conducted. Notable exclusions from the Toxicological Review

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

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    are large numbers of animal in vivo or in vitro studies designed to identify potential therapeutic
 2    agents that would prevent the carcinogenicity or genotoxicity of benzo[a]pyrene and toxicity
 3    studies of benzo[a]pyrene in nonmammalian species (e.g., aquatic species, plants).
 4           For the updated literature search conducted for the timeframe January 2012 through
 5    August 2014, the search terms included benzo(a)pyrene AND (rat OR mouse OR mice) and results
 6    were screened manually by title, abstract, and/or full text using the exclusion criteria outlined in
 7    Figure LS-1. Relevant studies that could potentially impact the hazard characterization and dose-
 8    response assessment were identified and considered. No studies were identified that would impact
 9    the assessment's major conclusions. Several pertinent studies, published since the last
10    comprehensive literature search, were identified and incorporated into the text where relevant
11
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 xlii             DRAFT—DO NOT CITE OR QUOTE

-------
                                                                  Toxicological Review ofBenzo[a]pyrene
                  References identified based on initial keyword search (see Table LS-1): "21,000 references
                                 References identified based on
                                 secondary keyword search (see Table
                                 LS-1): ~6,100 references
       30 references submitted by American
       Petroleum Institute
1

2
                                                                       Secondary keyword searching (see Table
                                                                       LS-1): ~14,600 references excluded
                                                         Manual screen of titles/abstracts:
                                                         ~4,900 references excluded
                                                         • Not relevant to BaP toxicity in mammals
                                                         (e.g., toxicity in aquatic species, plants)
                                                         • Site-specific risk assessments
                                                         • Chemical analytical methods
                                                         • Cancer chemotherapy studies
             Considered for inclusion in the Toxicological Review: - 1,000
             references; references subsequently evaluated based on
             Preamble Section 3
                                                            Manual screen of manuscripts excluded: ~ 600 references
                                                             • Not relevant to BaP toxicity in mammals
                                                             • Inadequate basis to infer exposure
                                                             • Inadequate reporting of study methods or results
                                                             • Animal toxicity studies with mixtures of chemicals
                                                             • Abstracts
                                                             • Duplicates
               Approximately 700 references cited in the Draft Toxicological Review
               •  Developmental toxicity: 37 references
               •  Reproductive toxicity: 70 references
               *  Immunotoxicity: 58 references
               •  Other Toxicological Effects: 27 references
                   •  Forestomach toxicity: 5 references
                   •  Hematological toxicity: 3 references
                   •  Liver toxicity: 3 references
                   •  Kidney toxicity: 3 references
                   •  Cardiovascular toxicity: 11 references
                   •  Neurological toxicity: 12 references
               •  Carcinogenicity: 171 references
               •Toxicokinetic: 115 references
               •Genotoxicity: 196 references
Figure LS-1. Study selection strategy.
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                       xliii              DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene

 1           Selection of studies for inclusion in the Toxicological Review was based on consideration of
 2    the extent to which the study was informative and relevant to the assessment and general study
 3    quality considerations. In general, the relevance of health effect studies was evaluated as outlined
 4    in the Preamble and EPA guidance (/I Review of the Reference Dose and Reference Concentration
 5    Processes [U.S. EPA. 2002] and Methods for Derivation of Inhalation Reference Concentrations and
 6    Application of Inhaled Dosimetry [U.S. EPA. 1994]]. The reasons for excluding epidemiological and
 7    animal studies from the references identified by the keyword search are provided in Figure LS-1.
 8           The available studies examining the health effects of benzo[a]pyrene exposure in humans
 9    are discussed and evaluated in the hazard identification sections of the assessment (Section 1], with
10    specific limitations of individual studies and of the collection of studies noted. The common major
11    limitation of the human epidemiological  studies (with respect to identifying potential  adverse
12    health outcomes specifically from benzo[a]pyrene] is that they all involve exposures to complex
13    mixtures containing other PAHs and other compounds. The  evaluation of the epidemiological
14    literature focuses on studies in which possible associations between external measures of exposure
15    to benzo[a]pyrene or biomarkers  of exposure to benzo[a]pyrene (e.g., benzo[a]pyrene-DNA
16    adducts or urinary biomarkers] and potential adverse health outcomes were evaluated.  Pertinent
17    mechanistic studies in humans (e.g., identification of benzo[a]pyrene-DNA adducts and
18    characteristics of mutations in human tumors] were also considered in assessing the weight of
19    evidence for the carcinogenicity of benzo[a]pyrene.
20           The health effects literature for benzo[a]pyrene is extensive. All animal studies of
21    benzo[a]pyrene involving repeated oral, inhalation, or dermal exposure that were considered to be
22    of acceptable quality, whether yielding positive, negative, or null results, were considered in
23    assessing the evidence for health effects  associated with chronic exposure to benzo[a]pyrene.
24    These studies were evaluated for aspects of design, conduct, or reporting that could affect the
25    interpretation of results and the overall contribution to the synthesis of evidence for determination
26    of hazard potential using the study quality considerations outlined in the Preamble. Discussion of
27    study strengths and limitations (that ultimately supported preferences for the studies and data
28    relied upon] were included in the  text where relevant.
29           Animal toxicity studies involving short-term duration and other routes of exposure were
30    also evaluated to inform conclusions about health hazards, especially regarding mode of action.
31    The references considered and cited in this document, including bibliographic information and
32    abstracts, can be found on the Health and Environmental Research Online (HERO] website2
33    (http://hero.epa.gov/benzoapyrene].
      2HERO (Health and Environmental Research On-line] is a database of scientific studies and other references
      used to develop EPA's risk assessments aimed at understanding the health and environmental effects of
      pollutants and chemicals. It is developed and managed in EPA's Office of Research and Development (ORD]
      by the National Center for Environmental Assessment (NCEA]. The database includes more than
      300,000 scientific articles from the peer-reviewed literature.  New studies are added continuously to HERO.

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

-------
                                                       Toxicological Review ofBenzo[a]pyrene
 1
 2    1.   HAZARD IDENTIFICATION
 3   1.1.  PRESENTATION AND SYNTHESIS OF EVIDENCE BY ORGAN/SYSTEM
 4          NOTE: In the environment, benzo[a]pyrene occurs in conjunction with other structurally
 5   related chemical compounds known as polycyclic aromatic hydrocarbons (PAHs).3 Accordingly,
 6   there are few epidemiologic studies designed to solely investigate the effects of benzo[a]pyrene.
 7   There are, however, many epidemiologic studies that have investigated the effects of exposure to
 8   PAH mixtures. Benzo[a]pyrene is universally present in these mixtures and is routinely analyzed
 9   and detected in environmental media contaminated with PAH mixtures, thus, it is often used an an
10   indicator chemical to measure exposure to PAH mixtures [Bostrom et al.. 2002].

11   1.1.1.  Developmental Toxicity
12          Human and animal studies provide evidence for PAH- and benzo[a]pyrene-induced
13   developmental effects. Effects on fetal survival, postnatal growth, and development have been
14   demonstrated in human populations exposed to PAH mixtures during gestation.  Animal studies
15   demonstrate various effects including changes in fetal survival, pup weight, blood pressure, fertility,
16   reproductive organ weight and histology, and neurological function in gestationally and/or early
17   postnatally treated animals.

18   Altered Birth Outcomes
19          Human and animal studies provide evidence that benzo[a]pyrene exposure may lead to
20   altered outcomes reflecting growth and development in utero or in early childhood. Two cohort
21   studies in pregnant women in China and the United States examined cord blood levels of
22   benzo[a]pyrene-7,8-diol-9,10 epoxide (BPDE)-deoxyribonucleic acid (DNA) adducts in relation to
23   measures of child growth following exposure to PAH mixtures [Tangetal.. 2006: Perera etal..
24   2005b: Perera etal., 2004] (Table 1-1). In the Chinese cohort, high benzo[a]pyrene-adduct levels
25   were associated with reduced weight at 18, 24, and 30 months of age, but not at birth [Tangetal.,
26   2006). In the U.S. cohort, an independent effect on birth weight was not observed with either
27   benzo[a]pyrene-adducts or environmental tobacco smoke (ETS) exposure; however, a doubling of
28   cord blood adducts in combination with ETS  exposure  in utero was seen, corresponding to an 8%
29   reduction in birth weight [Perera etal.. 2005b]. ETS, also called secondhand smoke, is the smoke
30   given off by a burning tobacco product and the  smoke exhaled by a smoker that contains over
     3PAHs are a large class of chemical compounds formed during the incomplete combustion of organic matter.

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

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    7,000 chemicals including benzo[a]pyrene.  No associations were seen with birth length (or height
 2    at later ages) in either of these cohort studies.
 3          A Chinese case-control study indicated that PAH exposure may be associated with increased
 4    risk of fetal death [Wu etal.. 2010).  A strong association was seen between maternal blood
 5    benzo[a]pyrene-DNA adduct levels and risk of delayed miscarriage (fetal death before 14 weeks of
 6    gestation), with a fourfold increased risk for levels above compared with below the median.
 7    However, no significant difference in adduct levels was detected between fetal tissue from cases
 8    compared to controls.
 9          Decreased fetal survival has also been noted in gestationally treated animals at relatively
10    high doses by the oral and inhalation routes. An approximate 40% decrease in fetal survival was
11    noted in mouse dams treated by gavage on CDs 7-16 at doses of 160 mg/kg-day, but no decreases
12    were observed at 10 or 40 mg/kg-day (Mackenzie and Angevine. 1981).  Several lower dose studies
13    of rats treated on CDs 14-17 with doses of up to 1.2 mg/kg-day benzo[a]pyrene did notobserve
14    any difference in fetal survival (Jules etal.. 2012: McCallister etal.. 2008: Brown et al.. 2007). By
15    the inhalation  route, fetal survival was decreased by 19% following exposure to 25 ug/m3
16    benzo[a]pyrene on CDs 11-20 in F344 rats (Archibong et al., 2002). Another publication from the
17    same group of collaborators Wuetal. (2003a)  graphically reported fetal survival as part of a study
18    analyzing metabolites of benzo[a]pyrene and activation of the aryl hydrocarbon receptor (AhR) and
19    cytochrome P450 (CYP450) 1A1 fWuetal.. 2003al  This study did notreportthe number of dams
20    or litters, and no numerical data were reported. The study authors reported statistically significant
21    decreases in fetal survival at 75 and 100 ug/m3 benzo[a]pyrene on CDs 11-20 compared to the
22    25 ug/m3 group. An apparent decrease in fetal survival was also seen at 25 ug/m3, but it was
23    unclear whether or not this change was statistically significant compared to the vehicle control.
24          In animals (Table 1-2 and Figure 1-1), reduced body weight in offspring has also been noted
25    in some developmental studies. Decreases in body weight (up to 13%) were observed in mice
26    following prenatal gavage exposure (gestation days [CDs] 7-16), and as time from exposure
27    increased (postnatal days [PNDs] 20-42), the dose at which effects were observed decreased (from
28    40 to 10 mg/kg-day, respectively) (Mackenzie  and Angevine. 1981). In addition, decreases in body
29    weight (approximately 10-15%) were observed in rats on PNDs 36 and 71 following gavage
30    exposure at only 2 mg/kg-day on PNDs 5-11 (Chen etal., 2012). At doses up to 1.2 mg/kg-day and
31    follow-up to PND 30, two developmental studies in rats did not observe decrements in pup body
32    weight following treatment from GD 14 to 17 flules etal.. 2012: McCallister etal.. 20081  Maternal
33    toxicity was not observed in mouse or rat dams exposed to up to 160 mg/kg-day benzo[a]pyrene
34    flules etal.. 2012: McCallister etal.. 2008: Brown etal.. 2007: Kristensen etal.. 1995: Mackenzie and
35    Angevine. 1981).

36    Fertility in Offspring
37          Several studies suggest that gestational exposure to maternal tobacco smoke decreases the
38    future fertility of female offspring fYe etal.. 2010:  Tensen etal.. 1998: Weinberg etal.. 19891
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-2             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    (Table 1-1). In animal models, marked effects on the development of male and female reproductive
 2    organs and the fertility of animals exposed gestationally has also been demonstrated [Kristensen et
 3    al.. 1995: Mackenzie and Angevine. 1981] (Table 1-2 and Figure 1-1). In two studies examining
 4    reproductive effects in mice, decreased fertility and fecundity in Fl animals was observed following
 5    exposure to doses >10 mg/kg-day during gestation (Kristensen etal.. 1995: Mackenzie and
 6    Angevine. 1981). When Fl females were mated with untreated males, a dose-related decrease in
 7    fertility of >30% was observed, in addition to a 20% decrease in litter size starting at the lowest
 8    dose tested of 10 mg/kg-day (Mackenzie and Angevine. 1981). A dose-related decrease in fertility
 9    was also observed in male mice treated gestationally with benzo[a]pyrene. At the lowest dose
10    tested (10 mg/kg-day), a 35% decrease in fertility was observed when gestationally exposed
11    animals were mated with untreated females (Mackenzie and Angevine. 1981). Similar effects on
12    fertility were observed in another developmental study in mice (Kristensen et al.. 1995). Fl
13    females (bred continuously for 6 months) in this study had  63% fewer litters, and litters were 30%
14    smaller as compared to control animals. The fertility of male offspring was not assessed in this
15    study.

16    Reproductive Organ Effects in Offspring
17          The above-mentioned studies also demonstrated dose-related effects on male and female
18    reproductive organs in animals exposed gestationally to benzo[a]pyrene (Table 1-2 and Figure 1-
19    1). Testicular weight was decreased and atrophic seminiferous tubules and vacuolization were
20    increased at >10 mg/kg-day in male mice exposed to benzo[a]pyrene gestationally from GD  7 to 16;
21    severe atrophic seminiferous tubules were observed at 40 mg/kg-day (Mackenzie and Angevine.
22    1981). Testicular weight was  also statistically significantly  decreased in Sprague-Dawley rats
23    treated on PNDs 1-7 at doses >10 mg/kg-day, when examined at PND 8 (Liang etal.. 2012).
24          In female mice treated with doses >10 mg/kg-day during gestation, ovarian effects were
25    observed including decreases  in ovary weight, numbers of follicles, and corpora lutea (Kristensen et
26    al.. 1995: Mackenzie and Angevine. 1981). Specifically, ovary  weight in Fl offspring was reduced
27    31% following exposure to 10 mg/kg-day benzo[a]pyrene (Kristensen et al.. 1995). while in
28    another gestational study at the same dose level, ovaries were so drastically reduced in size  (or
29    absent) that they were not weighed (Mackenzie and Angevine, 1981). Hypoplastic ovaries with few
30    or no follicles and corpora lutea (numerical data not reported), and ovaries with few or no small,
31    medium, or large follicles and  corpora lutea (numerical data not reported) have also been observed
32    in mouse offspring exposed gestationally to benzo[a]pyrene (Kristensen et al.. 1995: Mackenzie and
33    Angevine. 1981).

34    Cardiovascular Effects in Offspring
35          Increased systolic and diastolic blood pressure was  observed in adult animals following
36    gestational treatment with benzo[a]pyrene (Tules etal., 2012] (Table 1-2 and Figure 1-1).
37    Approximate elevations in systolic and diastolic blood pressure of 20-30 and 50-80% were noted

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-3             DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review ofBenzo[a]pyrene


1    in the 0.6 and 1.2 mg/kg-day dose groups, respectively.  Heart rate was decreased at 0.6 mg/kg-day,
2    but was increased at 1.2 mg/kg-day.

3    Immune Effects in Offspring
4           Several injection studies in laboratory animals suggest that immune effects may occur
5    following gestational or early postnatal exposure to benzo[a]pyrene. These studies are discussed in
6    Section 1.1.3.
7
8
Table 1-1.  Evidence pertaining to developmental effects of benzo[a]pyrene in
humans
           Study design and reference
                                                        Results
     Tangetal. (2006) (Tongliang, China)
     Birth cohort

     150 nonsmoking women who delivered
     babies between March 2002 and June 2002

     Exposure: Mean hours per day exposed to
     ETS 0.42 (SD 1.19); lived within 2.5 km of
     power plant that operated from December
     2001 to May 2002; benzo[a]pyrene-DNA
     adducts from maternal and cord blood
     samples; cord blood mean 0.33 (SD 0.14)
     (median 0.36) adducts/10 8 nucleotides;
     maternal blood mean 0.29 (SD 0.13)
     adducts/10~8 nucleotides
Relation between cord blood benzo[a]pyrene-DNA adducts and log-
transformed weight and height
                     Weight
                  Beta (p-value)
                  -0.007 (0.73)
                               Birth
                               18 mo
                               24 mo
                               30 mo
                  -0.048 (0.03)
                  -0.041 (0.027)
                  -0.040 (0.049)
Length (height)
 Beta (p-value)
 -0.001 (0.89)
 -0.005 (0.48)
 -0.007 (0.28)
 -0.006 (0.44)
                              Adjusted for ETS, sex of child, maternal height, maternal weight, and
                              gestational age (for measures at birth)
     (Perera et al. (2005b); Perera et al. (2004))
     (New York, United States)

     Birth cohort

     265 pregnant African-American and
     Dominican nonsmoking women who
     delivered babies between April 1998 and
     October 2002 (214 and 208 for weight and
     length analysis, respectively);
     approximately 40% with a smoker in the
     home

     Exposure: Benzo[a]pyrene-DNA adducts in
     cord blood samples; mean 0.22 (SD 0.14)
     adducts/10~8 nucleotides; median of
     detectable values 0.36 adducts/
     10 8 nucleotides
                               Relation between cord blood benzo[a]pyrene-DNA adducts and log-
                               transformed weight and length
                               Interaction
                               term
                               Benzo[a]-
                               pyrene-DNA
                               adducts
                     Weight
                  Beta (p-value)
                  -0.088 (0.05)

                  -0.020 (0.49)
    Length
 Beta (p-value)
 -0.014 (0.39)

 -0.005 (0.64)
                               ETS in home        -0.003 (0.90)            -0.007 (0.32)
                               Adjusted for ethnicity, sex of newborns, maternal body mass index,
                               dietary PAHs, and gestational age
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                    1-4              DRAFT—DO NOT CITE OR QUOTE

-------
                                                               Toxicological Review ofBenzo[a]pyrene
           Study design and reference
                                                                  Results
     Wu et al. (2010) (Tianjin, China)
     Case control study: 81 cases (96%
     participation rate)—fetal death confirmed
     by ultrasound before 14 wks of gestation;
     81 controls (91% participation rate)—
     elective abortions; matched by age,
     gestational age, and gravidity; excluded
     smokers and occupational PAH exposure

     Exposure: Benzo[a]pyrene in aborted tissue
     and maternal blood samples (51 cases and
     controls; 2 of 4 hospitals)
                                       Benzo[a]pyrene adduct levels (/Iff nucleotides), mean (± SD)
                                                           Cases         Controls         (p-value)
                                       Maternal blood     6.0 (±4.7)       2.7 (± 2.2)        (<0.001)
                                       Aborted tissue     4.8 (± 6.0)       6.0 (± 7.4)         (0.29)
                                       Low correlation between blood and tissue levels (r = -0.02 in cases,
                                       r = -0.21 in controls)
                                       Association between benzo[a]pyrene adducts and miscarriage3
                                                                       OR             (95% Cl)
                                       Per unit increase in adducts       1.37           (1.12,1.67)
                                       Dichotomized at median          4.56           (1.46,14.3)
                                       Conditional logistic regression, adjusted for maternal education,
                                        household income, and gestational age; age also considered as
                                        potential confounder
1
2

3
4
Cl = confidence interval; OR = odds ratio; SD = standard deviation.

        Table 1-2. Evidence pertaining to developmental effects of benzo[a]pyrene in
        animals
        Study design and reference
                                                              Results
      Birth outcomes and postnatal growth
      Mackenzie and Angevine (1981)
      CD-I mice, 30 or 60 FO females/
      dose
      0,10, 40, or 160 mg/kg-d by
      gavage
      CDs 7-16
                                4, number of FO females with viable litters: 46/60, 21/30,44/60, and 13/30*

                                4, Fl body weight at PND 20
                                % change from control: 0, 4, -7*, and -13*

                                ^ Fl body weight at PND 42
                                % change from control:  0, -6*, -6*, and -10*
                                (no difference in pup weight at PND 4)
      Kristensenetal. (1995)
      NMRI mice, 9 FO females/dose
      0 or 10 mg/kg-d by gavage
      CDs 7-16
                                Exposed FO females showed no gross signs of toxicity and no effects on
                                fertility (data not reported)
      Jules etal. (2012)
      Long-Evans rats, 6-17 FO
      females/dose
      0, 0.15, 0.3, 0.6, or 1.2 mg/kg-d by
      gavage
      CDs 14-17
                                No overt signs of toxicity in dams or offspring, differences in pup body
                                weight, or number of pups/litter
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-5               DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
  Study design and reference
                              Results
McCallisteretal. (2008)
Long-Evans Hooded rats,
5-6/group
0 or 0.3 mg/kg-d by gavage
CDs 14-17
No difference in number of pups/litter

No overt maternal or pup toxicity

No difference in liverbody weight

Increased brain:body weight ratio at PNDs 15 and 30 (data not shown)
Brown et al. (2007)
Long-Evans Hooded rats, 6/group
0, 0.025, or 0.15 mg/kg-d by
gavage
CDs 14-17
No difference in number of pups/litter or overt maternal or pup toxicity
Chen etal. (2012)
Sprague-Dawley rats, 20 pups
(10 male and 10 female)/group
0, 0.02, 0.2, or 2 mg/kg-d by
gavage
PNDs 5-11
Statistically significant decrease in pup body weight (approximate 10-15%
decrease) at 2 mg/kg-d measured on PNDs 36 and 71 (no significant
alteration of pup weight during treatment period)

No differences among treatment groups in developmental milestones:
incisor eruption, eye opening, development of fur, testis decent, or vaginal
opening
Archibong et al. (2002)
F344 rats, 10 females/group
0, 25, 75, or 100 u.g/m3 nose-only
inhalation for 4 hrs/d
CDs 11-20
•i, fetal survival ([pups/litter]/[implantation sites/litter] x 100)
% fetal survival: 97, 78*, 38*, and 34*%
Reproductive effects in offspring
Mackenzie and Angevine (1981)
CD-I mice, 30 or 60 FO females/
dose
0,10, 40, or 160 mg/kg-d by
gavage
CDs 7-16
^ number of Fl females with viable litters: 35/35, 23/35*, 0/55*, and 0/20*

4/F1 female fertility index (females pregnant/females exposed to males x
100):  100, 66*, 0*, and 0*

4, Fl male fertility index (females pregnant/females exposed to males x
100): 80, 52*, 5*, and 0*

•^ F2 litter size from Fl dams (20%) at 10 mg/kg-d (no litters were produced
at high doses)

4, size or absence of Fl ovaries (weights not collected)
hypoplastic ovaries with few or no follicles and corpora lutea (numerical data
not reported)

•^ testicular weight in Fl offspring
% change from control: 0, -42, -82, and ND (statistical significance not
reported)

/T" atrophic seminiferous tubules and vacuolization at >10 mg/kg-d; severe
atrophic seminiferous tubules at 40 mg/kg-d (numerical data not reported)
           This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-6               DRAFT—DO NOT CITE OR QUOTE

-------
                                                               Toxicological Review ofBenzo[a]pyrene
        Study design and reference
                                                              Results
      Kristensenetal. (1995)
      NMRI mice, 9 FO females/dose
      0 or 10 mg/kg-d by gavage
      CDs 7-16
                                4, number of F2 litters (-63%)

                                4, F2 litter size (-30%)

                                4' ovary weight (-31%) in Fl females

                                Few or no small, medium, or large follicles and corpora lutea
      Cardiovascular effects in offspring
      Jules etal. (2012)
      Long-Evans rats, 6-17 FO
      females/dose
      0, 0.15, 0.3, 0.6, or 1.2 mg/kg-d by
      gavage
      CDs 14-17
                                /T" systolic blood pressure (measured at PND 53)
                                        15%* increase at 0.6 mg/kg-d
                                        52%* increase at 1.2 mg/kg-d
                                        (other dose groups not reported)

                                /T" diastolic blood pressure (measured at PND 53)
                                        33%* increase at 0.6 mg/kg-d
                                        83% increase at 1.2 mg/kg-d
                                        (other dose groups not reported)

                                Altered heart rate
                                        10%* increase at 0.6 mg/kg-d
                                        8%* decrease at 1.2 mg/kg-d
1
2
3
*Statistically significantly different from the control (p< 0.05).
a% change from control calculated as: (treated value - control value)/control value x 100.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-7              DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene































1
1000 i

100 =
T 10,
ra :
E
0 0.1 !

n m























• LOAEL ANOAEL • Doses >LOAEL c
Doses  «

£
v


i
4- number
viable litters



^ ox ^
ro rn o o r^
ci 2 « Q ° „ " -0
rH t-l*jLJrsl^f*l ~
njOgZ .^" g
CJ ^r^^.li-2g^uira
= 0 Q •^c^^-^o^S
>0^rarTfOmLHSlN'-,,
MJliiJ3i3?l





m
c
1
5s


I P
o r-
l£3 V
^H C
r^ -

g

0 ^
0












x|/ ovarian size












-H LO
co cn
2 2 o
O — p
.E c * c.
? •- o ^
0) C ^H
&0 C C i

4






; mice
ID
rH
IS

Q
o






/
esticular
weight
and t
atrophy
Reproductive Effects
in Offspring












m
m
SSQ
o 2 z
(N Q_
"^ ^ O
0) rH OJ
* Q ^
^ ij n













Cardio-
vascular
Effects in
Offspring
2 Figure 1-1. Exposure-response array for developmental effects following oral
3 exposure to benzo[a]pyrene.





 4   Neurodevelopmental Effects
 5          There is evidence in humans and animals that benzo[a]pyrene induces developmental
 6   neurotoxicity.  In addition to the persistent reductions in cognitive ability observed in epidemiology
 7   studies of prenatal PAH exposure, the two epidemiology studies that examined benzo[a]pyrene-
 8   specific measures observed effects on neurodevelopment and behavior in young children. Altered
 9   learning and memory, motor activity, anxiety-like behavior, and electrophysiological changes have
10   also been observed in animals following oral and inhalation exposure to benzo[a]pyrene.
11          The mammalian brain undergoes periods of rapid brain growth, particularly during the last
12   3 months of pregnancy in humans, which has been compared to the first 1-2 weeks of life in the rat
13   and mouse neonate [Dobbing and Sands, 1979,1973]. This period is characterized by axonal and
14   dendritic outgrowth and the establishment of mature neuronal connections. Also during this
15   critical period, animals acquire many new motor and sensory abilities [Kolb and Whishaw. 1989).
16   There is a growing literature of animal studies that shows subtle changes in motor and cognitive

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-8             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    function following acute or repeated perinatal or lactational exposure to benzo[a]pyrene [Bouayed
 2    etal.. 2009a: McCallister etal.. 2008: Wormley etal.. 20041 These effects are described below.

 3    Cognitive function
 4          Head circumference at birth is associated with measures of intelligence in children, even
 5    among term infants [Broekman et al.. 2009: Gale etal.. 2006).  The two birth cohort studies that
 6    examined maternal or cord blood levels of benzo[a]pyrene-DNA adducts in relation to head
 7    circumference provide some evidence of an association, most strongly within the context of an
 8    interaction with ETS [Tang etal.. 2006: Pereraetal.. 2005b: Perera etal.. 2004] (Table 1-3). The
 9    cohort in Tongliang, China also examined intelligence quotient scores at age 5 years [Pereraetal.,
10    2012a). An interaction with ETS was seen in this analysis, with larger decrements seen on the full
11    scale and verbal scales with increased benzo[a]pyrene-DNAadduct levels in the presence of
12    prenatal exposure to ETS compared to the effects seen in the absence of prenatal exposure.
13          Animal studies have also provided evidence of altered learning and memory behaviors
14    following lactational or direct postnatal exposure to benzo[a]pyrene [Chen etal.. 2012: Bouayed et
15    al., 2009a] (Table 1-4). In mice, spatial working memory was measured using the Y-maze
16    spontaneous alternation test [Bouayed etal., 2009a). This test records alternations between arm
17    entries in a Y-shaped maze as a measure of memory, as rodents typically prefer to investigate a new
18    arm of the maze. To a lesser extent, this test can also reflect changes in sensory processing, novelty
19    preference, and anxiety-related responses in rodents. An improvement in working memory was
20    evident in mice, as exhibited by significant increases in spontaneous alternations in the Y-maze test
21    in mice on PND 40 following lactational exposure to 2 mg/kg-day benzo[a]pyrene (but not
22    20 mg/kg-day) from PND 0 to 14 [Bouayed et al., 2009a). The total number of arm entries in the
23    Y-maze was unaffected by lactational exposure, suggesting that changes in motor function were not
24    driving this response.  Similarly, CPRlox/loxtransgenic mice (a mouse model with low brain
25    expression of CYP450  reductase) exhibit deficiencies in novel object recognition tests following in
26    utero exposure to benzo[a]pyrene during late gestation [Li etal.. 2012: Shengetal.. 2010). This
27    finding suggests impairment in short-term memory, although these tests also reflect locomotor
28    exploratory behavior,  response to novelty, and attention. In rats, spatial learning and memory was
29    measured using the Morris water maze, which measures the ability of a rat to navigate to a target
30    platform using external spatial cues. Increased escape latency (time to find the hidden platform), as
31    well as decreased time in the target quadrant and decreased number of platform crossings during a
32    probe trial with the platform removed were observed in PND 39-40 rats following postnatal
33    exposure to 2 mg/kg-day benzo[a]pyrene [Chen etal.. 2012). These effects were more pronounced
34    in animals tested at PNDs 74-75, with effects observable at >0.2 mg/kg-day. No difference in swim
35    speed was observed during the probe  trial tests (swim speed did not appear to be analyzed during
36    the hidden platform trials) between treatment groups, suggesting that the observed changes are
37    not attributable to general motor impairment These observations may indicate primary effects of
38    benzo[a]pyrene on learning and/or memory processes; however, the presented data were
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-9             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    insufficient to attribute these findings to learning and memory processes alone. Specifically, visual
 2    examinations of the improvements in escape latency (slopes) over the four learning trial days were
 3    not noticeably affected by treatment dose, suggesting that all groups learned at a similar rate. As
 4    four trials/day were averaged for each animal at each trial day, it is unclear whether the dose-
 5    related increases in escape latency already observable at trial day 1 reflect effects on learning
 6    across those first four trials or other effects (e.g., altered anxiety or vision responses). As it is not
 7    clear that the groups learned to a comparable extent in hidden platform tests, the results of the
 8    probe trial cannot be conclusively attributed to memory retention; however, the decreased
 9    performance of the benzo[a]pyrene-exposed rats in the Morris water maze still represents a
10    persistent neurobehavioral change.

11    Neuromuscular function, coordination, and sensorimotor development
12          Motor behavior and coordination, assessed by locomotion, reaching, balance,
13    comprehension, drawing, and hand control was one of the specific domains assessed in the Chinese
14    birth cohort evaluated by Tangetal. (2008).  In children aged 2 years, decreased scores were seen
15    in relation to increasing benzo[a]pyrene-DNA adducts measured in cord blood, with a Beta per unit
16    increase in adducts of -16 (p = 0.04), and an approximate twofold increased risk of development
17    delay per unit increase in adducts (Table 1-3).
18          In laboratory animals (Table 1-4 and Figure 1-2), impaired performance in neuromuscular
19    and sensorimotor tests have been consistently observed in mice lactationally exposed to >2 mg/kg-
20    day benzo[a]pyrene from PND 0 to 14 (Bouayed et al.. 2009a) and in rat pups postnatally exposed
21    to >0.02  mg/kg-day benzo[a]pyrene from PND  5 to 11 (Chenetal.. 2012). In the righting reflex test,
22    significant increases in righting time were observed in PNDs 3-5 mice and in PNDs 12-16 rats.
23    These decrements did not show a monotonic dose response. In another test of sensorimotor
24    function and coordination, dose-dependent increases in latency in the negative geotaxis test were
25    observed in PND 5-9 mice and in PND 12-14 rats. The forelimb grip strength test of
26    neuromuscular strength was also evaluated in both mice and rats, but alterations were only
27    observed in mice. In mice, a dose-dependent increase in duration of forelimb grip was observed on
28    PNDs 9 and 11 during lactational exposure to benzo[a]pyrene. The Water Escape Pole Climbing
29    test was  also used to evaluate neuromuscular function and coordination in mice (Bouayed etal.,
30    2009a).  No effect on climbing time was observed, suggesting no change in muscle strength.
31    However, increased latency in pole grasping and pole escape in PND 20 male pups was observed,
32    highlighting potential decrements in visuomotor integration and/or coordination, although anxiety
33    or fear-related responses cannot be ruled out Treatment-dependent increases in pup body weight
34    around the testing period complicate the interpretation of these results.
35          Chen etal. (2012) observed statistically significant delays on the order of ~0.2-0.3 seconds
36    in the surface righting test and ~3-4 seconds in the negative geotaxis test. Differences due to
37    exposure were notable between PND 12 and 16, with equal performance in righting observed by
38    PND 18 and in geotaxis by PND 16.  The authors found no effect on gender, therefore the data for
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-10           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    male and female rats was pooled for these measures. However, it should be noted that differences
 2    in the maturation of these developmental landmarks following challenge have been shown to exist
 3    between males and females. Negative geotaxis and surface righting are discrete endpoints
 4    routinely used as part of a neurobehavioral test battery to assess acquisition of behavioral reflexes.
 5    Chenetal. [2012] used the surface righting and negative geotaxis tests as quantitative measures of
 6    sensorimotor function at PND 12 and beyond. Typically in these tests, animals are observed on
 7    consecutive days (e.g., PNDs 3-12) and time to acquisition of these phenotypes is measured. Chen
 8    etal. [2012] did not measure performance over consecutive days of development or measure
 9    baseline acquisition of these behaviors. Although these tests as conducted by Chen etal. [2012]
10    cannot discern a developmental delay, the data support a transient impairment of sensorimotor
11    function in animals that have already developed this reflex (e.g., able to orient 180 degrees and able
12    to right within 2  seconds]. Thus, these data indicate that benzo [ajpyrene may affect sensorimotor
13    function in developing animals.

14    Anxiety and activity
15          Anxiety/depression and attention/hyperactivity symptoms in children ages 6-7 years were
16    examined via questionnaire in relation to prenatal air monitoring of benzo[a]pyrene and other
17    PAHs, and in relation to benzo[a]pyrene-specific DNA adducts measured at birth in a follow-up of a
18    birth cohort study conducted in New York City [Perera et al.. 2012b].  PAH exposure levels (based
19    on personal air monitoring, n = 253] and benzo[ajpyrene-specific DNA adducts measured in cord
20    blood samples (n = 138] were both positively associated with symptoms of anxiety/depression and
21    attention problems (see Table 1-3]. Given the limited sample size, however, the cord blood results
22    are based on relatively sparse data (<5 in the borderline or clinical range in the low exposure
23    referent group].  Associations with maternal blood adducts were  similar to or slightly smaller than
24    those seen with cord blood adducts.  Exposure was treated as a dichotomy (i.e., for adducts,
25    detectable compared with non-detectable levels] in these analyses.
26          Decreased anxiety-like behavior was reported in both rats and mice weeks to months
27    following postnatal oral exposure to benzo[a]pyrene (Chenetal.. 2012: Bouayed et al.. 2009a]
28    (Table 1-4]. Anxiety-like behaviors were  measured in both species using an elevated plus maze,
29    where an increase in the time spent in the closed arms  of the maze is considered evidence of
30    anxious behavior. In mice, significant increases in the percent open arm entries and percent time
31    spent in open arms of the maze, as well as significantly decreased entries into closed arms of the
32    maze (in the  2 mg/kg-day group], were observed on PND 32 following lactational exposure to
33    >2 mg/kg-day benzo[a]pyrene (Bouayed et al.. 2009a]. To rule out potential differences in total
34    activity or general motivation and exploration, the authors expressed the open arm data as
35    percentages, and they also demonstrated that there were no exposure-related effects on the total
36    number of arm entries. The mice also exhibited decreased latency of the first entry into an open
37    arm following lactational exposure to 20 mg/kg-day benzo[a]pyrene.  Similar results were reported
38    for rats, with decreased anxiety-like behavior following oral benzo[a]pyrene exposure from PND 5
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-11            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    to 11, although sex-specific differences were observed [Chenetal., 2012]. In females, postnatal
 2    exposure to >0.2 mg/kg-day benzo[a]pyrene was associated with a significant increase in the
 3    number of open arm entries and a significant decrease in the number of closed arm entries on
 4    PND 70. Significantly increased time in open arms of the maze was reported in PND 70 female rats
 5    following postnatal exposure to >0.02 mg/kg-day. Male rats also showed decreased anxiety-like
 6    behavior on PND 70, although the doses needed to detect these responses were higher than females
 7    (i.e., increases at >2 mg/kg-day for open arm entries and >0.2 mg/kg-day for time spent in open
 8    arms). A significant decrease in latency to enter an open arm of the maze was observed in both
 9    male and female rat pups exposed to 2 mg/kg-day benzo[a]pyrene. Similar to the observations in
10    mice, exposure did not appear to have an effect on total activity or general motivation of the rats, as
11    total arm entries were unchanged by treatment
12           Increased spontaneous locomotor activity in the open field on PNDs 34 and 69 has been
13    reported in rats postnatally exposed to 2 and >0.2 mg/kg-day, respectively [Chenetal.. 2012). but
14    not in mice exposed lactationally to doses up to 20 mg/kg-day and tested on PND 15 [Bouayed et
15    al.. 2009a). Interestingly, no differences in the open field test were observed in rats that were
16    postnatally exposed and tested on PNDs 18 and 20, suggesting either that longer latencies between
17    exposure and testing may be required, or that these developmental effects may only manifest in
18    more mature rats [Chenetal.. 2012). An apparent increased sensitivity of older animals was also
19    present in the elevated plus maze and Morris water maze tests performed by Chenetal. [2012].
20    Elevated activity in an open field is attributable primarily to either increased motor activity or
21    decreased anxiety-like behavior.  However, the relative contributions of these two components
22    could not be separated in either of these studies, as the authors did not evaluate activity in central
23    versus peripheral regions of the field (i.e., anxious rodents will spend less time in the center of the
24    field).

25    Electrophysiological changes
26           Electrophysiological effects  of gestational exposure  to benzo[a]pyrene have been examined
27    in two studies (by the same research group) through implanted electrodes in the rat cortex and
28    hippocampus (Table 1-4). Maternal inhalation exposure to  0.1 mg/m3 resulted in reduced long-
29    term potentiation in the dentate gyrus of male offspring between PND 60 and 70 (Wormley etal.,
30    2004): however, significant fetal toxicity at this exposure level complicates interpretation of these
31    results. Oral exposure of dams to 0.3 mg/kg-day for 4 days  during late gestation resulted in
32    decreased evoked neuronal activity in male offspring following mechanical whisker stimulation
33    between PND 90 and 120 (McCallister etal.. 2008). Specifically, the authors noted reduced spike
34    numbers in both short and long latency responses following whisker stimulation.  These effects
35    were observed several months post-exposure, suggesting that gestational benzo[a]pyrene exposure
36    may have long-lasting functional effects on neuronal activity elicited by sensory stimuli.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-12            DRAFT—DO NOT CITE OR QUOTE

-------
                                                               Toxicological Review ofBenzo[a]pyrene
1
2
Table 1-3.  Evidence pertaining to the neurodevelopmental effects of
benzo[a]pyrene from PAH mixtures
          Reference and study design
                                                         Results
     (Tang et al. (2008): Tang et al. (2006))
     (Tongliang, China)

     Birth cohort

     150 nonsmoking women, delivered
     March 2002-June 2002; lived within
     2.5 km of power plant that operated from
     December 2001 to May 2002

     Outcomes: Head circumference at birth;
     Gesell Developmental Schedule,
     administered by physicians at 2 yrs of age
     (four domains: motor, adaptive, language,
     and social); standardized mean
     score = 100 ± SD 15 (score
     <85 = developmental delay)

     Exposure:  Mean hrs/d exposed to ETS
     0.42 (SD 1.19); lived within 2.5 km of
     power plant that operated from
     December 2001 to May 2002;
     benzo[a]pyrene-DNA adducts from
     maternal and cord blood samples; cord
     blood mean 0.33 (SD 0.14) (median 0.36)
     adducts/10~8 nucleotides; maternal blood
     mean 0.29 (SD 0.13)
     adducts/10~8 nucleotides
                              Relation between cord blood benzo[a]pyrene-DNA adducts and log-
                              transformed head circumference
                                           Birth
                                           18 mo
                                           24 mo
                                           30 mo
                      Beta (p-value)

                      -0.011 (0.057)

                      -0.012 (0.085)

                      -0.006 (0.19)

                      -0.005 (0.31)
                              High versus low, dichotomized at median, adjusted for ETS, sex of
                              child, maternal height, maternal weight, Cesarean section delivery,
                              maternal head circumference, and gestational age (for measures at
                              birth)
     (Tang et al. (2008); Tang etal. (2006)) (see
     above for population and exposure
     details)

     n = 110 for Developmental Quotient
     analysis; no differences between the
     110 participants in this analysis and the
     nonparticipants with respect to maternal
     age, gestational age, birth weight, birth
     length, or birth head circumference;
     higher maternal education (direction not
     reported, p = 0.056)

     Outcomes: Gesell Developmental
     Schedule, administered by physicians at
     2 yrs of age (four domains: motor,
     adaptive, language, and social);
     standardized mean score = 100 ± SD 15
     (score <85 = developmental delay)
                              Association between benzo[a]pyrene adducts and development
                                                   Beta (95% Cl)a
                          OR (95% Cl)
                              Motor
                              Adaptive
                              Language
                              Social
                              Average
-16.0 (-31.3, -0.72)*

 -15.5 (-35.6, 4.61)

 -16.6 (-33.7, 0.46)

 -9.29 (-25.3, 6.70)

-14.6 (-28.8, -0.37)*
1.91(1.22,2.97)*

 1.16 (0.76,1.76)

 1.31(0.84,2.05)

 1.52 (0.93, 2.50)

 1.67 (0.93, 3.00)
                              aLinear regression of change in Developmental Quotient per unit
                               increase in benzo[a]pyrene adducts
                              bLogistic regression of risk of developmental delay (defined as
                               normalized score <85) per 1 unit (0.1 adducts/10~8 nucleotides)
                               increase in adducts
                              Both analyses adjusted for sex, gestational age, maternal education,
                              ETS, and cord lead levels
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-13             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
     Reference and study design
                            Results
Perera et al. (2012a); (Tang et al. (2008);
Tang et al. (2006)) (see above for
population and exposure details)

132 (83%) followed through age 5; 100 of
these had complete data for analysis; no
differences between the 100 participants
in this analysis and the nonparticipants
with respect to adduct levels, ETS
exposure, IQ measures, maternal age,
gestational age, or infant gender; higher
maternal education (60 and 35% with
> high school,  respectively, in participants
and nonparticipants, p < 0.05)

Outcomes: Wechsler Preschool and
Primary Intelligence Quotient scale
(Shanghai version)
ETS measure not correlated with benzo[a]pyrene adduct measures
(i.e., absolute value of Spearman r < 0.10)
Relation between cord blood benzo[a]pyrene-DNA adducts, ETS
exposure, and IQ measures
                      Beta (95% Cl)

                       Main effect

Full scale            -2.42 (-7.96, 3.13)

Verbal              -1.79 (-7.61,4.03)

Performance         -2.57 (-8.92, 3.79)
Beta per 1 unit increase in log-transformed cord adducts, adjusted for
ETS exposure, gestational age, maternal education, cord lead, maternal
age, and gender
With ETS interaction term

  -10.10 (-18.90, -1.29)

  -10.35 (-19.61, -1.10)

   -7.78 (-18.03,2.48)
(Perera et al. (2012b); Perera et al.
(2005b); Perera et al. (2004)) (United
States, New York)

Birth cohort

265 pregnant women: African-American
and Dominican nonsmoking women who
delivered babies between April 1998 and
October 2002 (253 and 207 for behavior
and head circumference analysis,
respectively); approximately 40% with a
smoker in the home

Outcomes: Head circumference at birth

Exposure: Benzo[a]pyrene-DNA adducts
from maternal and cord blood samples;
mean 0.22 (SD 0.14)
adducts/10 8 nucleotides; median of
detectable values
0.36 adducts/10"8 nucleotides
Relation between cord blood benzo[a]pyrene-DNA adducts,
environmental tobacco smoke exposure (ETS), and log-transformed
head circumference
Interaction term
benzo[a]pyrene-DNA adducts
ETS in home
  Beta (p-value)

  -0.032 (0.01)

  -0.007 (0.39)

  -0.005 (0.43)
High versus low, dichotomized at 0.36 adducts/10   nucleotides,
adjusted for ethnicity, sex of newborns, maternal body mass index,
dietary PAHs, and gestational age
           This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-14              DRAFT—DO NOT CITE OR QUOTE

-------
                                                                Toxicological Review ofBenzo[a]pyrene
           Reference and study design
                                                                  Results
      Perera et al. (2012b)
      n = 215 with outcome data and no missing
      covariate data); no differences between
      the participants in this analysis and the
      nonparticipants with respect to adduct
      levels, ETS exposure, maternal age,
      gestational age, and socioeconomic
      variables; participants more likely to be
      female and African-American

      Outcomes: Child Behavior Checklist
      (118 items), completed by mothers for
      children ages 6-7 yrs. Two domains:
      anxious/depression, attention problems
      (normalized T-score <65 = borderline or
      clinical syndrome); also used for scales of
      anxiety problems and attention deficit
      hyperactivity problems based on DSM
      classification
                                      Logistic regression of risk of borderline or clinical status in relation PAH
                                      levels and to detectable levels of benzo[a]pyrene adducts
                                                                                        Cord blood
                                      Anxious/depressed
                                      Attention problems
                                      Anxiety (DSM)
                                      Attention deficit -
                                      hyperactivity (DSM)
Prevalence

  6.3 %

  6.7%
  9.5%

  7.9%
    PAH

 OR (95% Cl)

8.9(1.7,46.5)

3.8(1.1,12.7)
4.6 (1.5, 14.3)

2.3 (0.79, 6.7)
  OR (95% Cl)

 2.6(0.69,9.4)

4.1 (0.99, 16.6)
 2.5(0.84,7.7)

2.6 (0.68, 10.3)
                                      Exposure dichotomized for PAH as above and below median
                                      (2.273 ng/m3) for parent population and for cord blood benzo[a]pyrene
                                      adducts as detectable (n = 56 cord blood samples) versus non-
                                      detectable (n = 92); adjusted for sex, gestational age, maternal
                                      education, maternal IQ, prenatal ETS, ethnicity, age, heating season,
                                      prenatal demoralization, and HOME inventory
1
2
3
4
5

6
7
*Statistically significantly different from the control (p< 0.05).

DSM = Diagnostic and Statistical Manual of Mental Disorders; HOME = Home Observation for Measurement of the
 Environment; IQ= intelligence quotient.

        Table 1-4. Evidence pertaining to the neurodevelopmental effects of
        benzo[a]pyrene in animals
        Reference and study design
                                                               Results3
      Cognitive fun ction
      Chen etal. (2012)
      Sprague-Dawley rats, 20 pups
      (10 male and 10 female)/group
      0, 0.02, 0.2, or 2 mg/kg-d by gavage
      PNDsS-11
                                 Hidden Platform test in Morris water maze:
                                         Adolescent test period (PNDs 36-39): significant increase in escape
                                         latency at 2 mg/kg-d only
                                         Adult test period (PNDs 71-74): significant increase in escape
                                         latency at >0.2 mg/kg-d
                                         Increases in latency were already ~30% greater than controls at
                                         2 mg/kg-d on the first trial day (i.e., on PND 36 or 71)
                                         All experimental groups exhibited similar improvements in escape
                                         latency, as slopes were visually equivalent across the 4 trial days

                                 Probe test in the Morris water maze (d 5):
                                         Time spent in the target quadrant:
                                         PND 40: significant decrease at 2 mg/kg-d only
                                         PND 75: significant decrease at >0.2 mg/kg-d

                                         Number of platform crossings:
                                         PND 40: significant decrease at 2 mg/kg-d only
                                         PND 75: significant decrease at >0.2 mg/kg-d (in females) and
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-15             DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
  Reference and study design
                              Results3
                                         2 mg/kg-d (in males)
Bouayed et al. (2009a)
Female Swiss albino mice, 5/group
0, 2, or 20 mg/kg-d maternal
gavage
PNDs 0-14 (lactational exposure)
Significant increase in the percent of spontaneous alternations in the Y-
maze alternation test at 2 mg/kg-d but not at 20 mg/kg-d

No effect on the total number of arm entries in the Y-maze alternation test
Neuromuscularfunction, coordination, and sensohmotor development
Chen etal. (2012)
Sprague-Dawley rats, 20 pups
(10 male and 10 female)/group
0, 0.02, 0.2, or 2 mg/kg-d by gavage
PNDs 5-11
Latency in the surface righting reflex test
        PND 12: significant increase at 0.2 mg/kg-d only
        PND 14: significant increase at 0.02 and 2 mg/kg-d only
        PND 16: significant difference at 2 mg/kg-d only
        PND 18: no significant difference

Latency in the negative geotaxis test
        PND 12: significant increase at all doses
        PND 14: significant increase at 2 mg/kg-d only
        PNDs 16 and 18: no significant difference

No effect on duration of forelimb grip in forelimb grip strength test
No effect on the latency to retract from the edge in cliff aversion test

Note: Males and females were pooled for all analyses
Bouayed et al. (2009a)
Female Swiss albino mice, 5/group
0, 2, or 20 mg/kg-d by maternal
gavage
PNDs 0-14 (lactational exposure)
Significant increase in righting time in the surface righting reflex test at both
doses on PNDs 3 and 5 (but not PNDs 7 and 9)

Significant increase in latency in the negative geotaxis time for 20 mg/kg-d
dose group at PNDs 5, 7, and 9 (no significant difference at PND 11)

Significant increase in duration of forelimb grip in forelimb grip strength
test at both dose groups on PND 9 (statistically significant at PND 11 only at
high dose)

Significant increase in pole grasping latency in male pups in the water
escape pole climbing test at 20 mg/kg-d

No effect on climbing time in the water escape pole climbing test

Significant increase in pole escape latency in the water escape pole climbing
test in male rats at 20 mg/kg-d
Anxiety and/or motor activity
Chen etal. (2012)
Sprague-Dawley rats, 20 pups
(10 male and 10 female)/group
0, 0.02, 0.2, or 2 mg/kg-d by gavage
PNDs 5-11
Elevated plus maze:
       Significant increase in the number of entries into open arms at PND
       70 at >0.2 mg/kg-d (in females) and 2 mg/kg-d (in males) (no
       difference at PND 35)

       Significant decrease in the number of entries into closed arms at
       PND 70 at >0.2 mg/kg-d (in females) and 2 mg/kg-d (in males) (no
           This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-16              DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
  Reference and study design
                             Results3
                                        difference at PND 35)

                                        Significant increase in the time spent in open arms at PND 35 at
                                        2 mg/kg-d in females and at PND 70 at doses >0.02 mg/kg-d in
                                        females and >0.2 mg/kg-d in males

                                        Significant decrease in latency time to first enter an open arm on
                                        PND 70 at >0.2 mg/kg-d (no difference at PND 35)

                                        No effect on the total number of arm entries between treatment
                                        groups
                                        (calculated by EPA from graphically reported open and closed arm
                                        entries)

                                 Open field test:
                                        Significant increase in the number of squares: PND 34, significant
                                        increase at 2 mg/kg-d; PND 69, significant increase at >0.02 mg/kg-
                                        d (no difference at PNDs 18 and 20)

                                        Significant increase in rearing activity at 0.2 mg/kg-d on PND 69 (no
                                        difference at PNDs 18, 20, and 34)
Bouayed et al. (2009a)
Female Swiss albino mice, 5/group
0, 2, or 20 mg/kg-d by maternal
gavage
PNDs 0-14 (lactational exposure)
Elevated plus maze:
       Significantly increased time in open arms at >2 mg/kg-d

       Significantly increased percentage of entries into open arms at
       >2 mg/kg-d

       Significantly decreased entries into closed at 2 mg/kg-d, but not at
       20 mg/kg-d

       Significantly decreased latency time to enter an open arm at
       20 mg/kg-d

       No effect on the total number of arm entries

       No significant effect of gender on performance was detected, so
       males and females were pooled for analyses

Open field test:
       No significant change in activity on PND 15, but data not provided
Electrophysiological changes
McCallister et al. (2008)
Long-Evans Hooded rats, 5-6/group
0 or 0.3 mg/kg-d by gavage
CDs 14-17
Statistically significant decreases in stimulus-evoked cortical neuronal
activity on PNDs 90-120

Reduction in the number of spikes in both the short and long latency
periods on PNDs 90-120 (numerical data not presented)
           This document is a draft for review purposes only and does not constitute Agency policy,
                                               1-17             DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene
         Reference and study design
                                                           Results3
      Wormley et al. (2004)
      F344 rats, 10 females/group
      0 or 100 u.g/m3 by nose-only
      inhalation for 4 hrs/d
      CDs 11-21
                               Electrophysiological changes in the hippocampus:
                                       Consistently lower long term potentiation following gestational
                                       exposure (statistical analysis not reported)
                                       % change relative to control: -26%

                               Note: significant fetal toxicity observed (99 versus 34% birth index)
 1
 2
a% change from control calculated as: (treated value - control value)/control value x 100.
        100
         10 ,
       o
       Q
        0.1 -.
       0.01
                                          ILOAEL ANOAEL  • Doses > LOAEL  O Doses < NOAEL
                                     •      A      A
                     y. Q
                     I?
                                   j~- LTi "C3
                                   O ,/! CJ
                                   = 0 !=
•5 "S
Q '=
c Q '-
5-|
                                                                             p
                                                                          5 Q '5
                                                                          3? 3
                                                                          O   r-
                                                                          CD   =
                                                                      elevated plus
                                                                       maze test
 4
 5

 6
 7
 8
 9
10
11
12
13
       Figure 1-2. Exposure-response array for neurodevelopmental effects
       following oral exposure.

Mode of Action Analysis—Developmental Toxicity and Neurodevelopmental Toxicity
       Data regarding the potential mode of action for the various manifestations of developmental
toxicity associated with benzo[a]pyrene exposure are limited, and the mode of action for
developmental toxicity is not known.  General hypothesized modes of action for the various
observed developmental effects include, but are not limited to, altered cell signaling, cytotoxicity,
and oxidative stress.
       It is plausible that developmental effects of benzo[a]pyrene may be mediated by altered cell
signaling through the AhR. Benzo[a]pyrene is a ligand for the AhR, and  activation of this receptor
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                   1-18             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    regulates downstream gene expression including the induction of GYP enzymes important in the
 2    conversion of benzo[a]pyrene into reactive metabolites. Studies in AhR knock-out mice indicate
 3    that AhR signaling during embryogenesis is essential for normal liver, kidney, vascular,
 4    hematopoietic, and immune development [Schmidt etal.. 1996: Fernandez-Salguero etal.. 1995).
 5    In experiments in AhR responsive and less-responsive mice, the mice with the less-responsive AhR
 6    were protected from renal injury as adults following gavage treatment with 0.1 or 0.5 mg/kg-day
 7    benzo[a]pyrene from GD 10 to 13. Renal injury was indicated by an increase in urinary albumin
 8    and a decrease in glomerular number [Nanez etal.. 2011].
 9          Low birth weight has been associated with prenatal exposure to PAHs in human
10    populations [Perera et al., 2005b]. Several epidemiology studies have revealed an inverse
11    association between low birth weight and increased blood pressure, hypertension, and measures of
12    decreased renal function as adults [Zandi-Nejadetal.. 2006). It has been hypothesized that this
13    may be attributable to a congenital nephron deficit associated with intrauterine growth restriction
14    [Zandi-Nejad et al.. 2006].
15          No clear mode(s) of action for the observed neurodevelopmental changes following
16    benzo[a]pyrene exposure have been demonstrated. General hypothesized mechanisms with
17    limited support are related to altered central nervous system neurotransmission. These
18    mechanisms involve altered neurotransmitter gene expression, neurotransmitter levels, and
19    neurotransmitter receptor signaling in regions associated with spatial learning, anxiety, and
20    aggression, such as the hippocampus, striatum, amygdala, andhypothalamus [Li etal.. 2012: Oiu et
21    al.. 2011: Tang etal.. 2011: Xia etal.. 2011: Bouayed etal.. 2009a: Grova etal.. 2008: Brown etal..
22    2007: Grova etal.. 2007: Stephanou etal.. 1998).
23          Mechanistic studies in rodents exposed as adults, which exhibit some of the same
24    behavioral changes as animals exposed during development, may also inform potential mode(s) of
25    action for the observed neurodevelopmental changes. Specifically regarding potential changes in
26    spatial learning and memory processes, multiple studies in developing [Li etal.. 2012: McCallister
27    etal.. 2008: Wormley etal.. 2004] and adult [Maciel etal.. 2014: Oiu etal.. 2013: Tang etal.. 2011:
28    Grova etal.. 2008: Grova etal.. 2007] rodents suggest that changes in N-methyl-D-aspartate
29    (NMDA] receptor signaling seen with benzo[a]pyrene exposure (e.g., changes in expression
30    patterns of NR2A and NR2B subunits] may be  responsible for apparent effects on learning and
31    memory.
32          In relation to potential changes in anxiety-like behaviors (and also relevant to effects on
33    learning and memory processes], many commonly used anti-anxiety medications work by
34    increasing brain serotonin levels (e.g., selective serotonin reuptake inhibitors], increasing brain
35    dopamine levels (e.g., dopamine reuptake inhibitors], or by targeting gamma-aminobutyric acid
36    (GABA] receptors (e.g., benzodiazepines]. Although GABAA receptor messenger ribonucleic acid
37    (mRNA] in whole-brain homogenates was unchanged following lactational benzo[a]pyrene
38    exposure, exposure  at >2 mg/kg-day from PND 1 to 14 caused dose-dependent decreases in
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-19            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    serotonin receptor (5HTiA) expression [Bouayed et al., 2009a: Stephanou etal., 1998]. Additional
 2    support for identifying changes in monoamine neurotransmitter signaling (serotonin and dopamine
 3    signaling, in particular) as a potential mechanism(s) in the altered anxiety-like behaviors observed
 4    following benzo[a]pyrene exposure is provided in multiple studies of rodents exposed as adults
 5    f Bouayed etal.. 2012: Oiu etal.. 2011: Xia etal.. 2011: Stephanou etal.. 1998: Tayasekaraetal..
 6    1992] and in a single study of blood neurotransmitter levels in occupationally-exposed men [Niu et
 7    al.. 2010]. Overall, these data suggest possible effects of benzo[a]pyrene exposure on NMDA
 8    receptor expression and regulation of monoamine neurotransmitters including serotonin and
 9    dopamine, but these findings require additional studies to clarify and extend understanding of
10    these events.

11    Summary of Developmental Effects
12           Developmental effects following in utero exposure to PAH mixtures or benzo[a]pyrene
13    alone have been reported in humans and in animal models. In human populations, decreased head
14    circumference, decreased birth weight, and decreased postnatal weight have been reported.
15    Analogous effects in laboratory animals, including decreased pup weight and decreased fetal
16    survival, have been noted following gestational or early postnatal exposure to benzo[a]pyrene by
17    the oral or inhalation route [Chen etal.. 2012: Archibongetal.. 2002: Mackenzie and Angevine.
18    1981].  Reproductive function is also altered in mice treated gestationally with benzo[a]pyrene
19    [Kristensen etal.. 1995: Mackenzie and Angevine. 1981].  These effects include impaired
20    reproductive performance in Fl offspring (male and female]  and alterations of the weight and
21    histology of reproductive organs (ovaries and testes].
22           The available human and animal data also support the conclusion thatbenzo[a]pyrene is a
23    developmental neurotoxicant. Human studies of environmental PAH exposure in two cohorts have
24    observed neurotoxic effects, including suggestions of reduced head circumference (Tang etal..
25    2006: Perera etal.. 2005b: Perera etal.. 20041. impaired cognitive ability fPerera etal.. 2009: Tang
26    etal.. 2008]. impaired neuromuscular function (Tang etal.. 2008]. and increased attention
27    problems and anxious/depressed behavior following prenatal exposure (Perera etal.. 2012b].
28    These effects were seen in birth cohort studies in different populations (New York City and China],
29    in studies using specific benzo[a]pyrene measures (i.e., adduct levels measured in cord blood
30    samples] (Perera et al.. 2012b: Tang etal.. 2008: Tang etal.. 2006: Perera et al.. 2005b: Perera etal..
31    2004].  The available evidence from mice and rats also demonstrates significant and persistent
32    developmental impairments following exposure to benzo[a]pyrene. Impaired learning and
33    memory behaviors, decreased anxiety-like behaviors, and impaired neuromuscular function were
34    consistently observed in multiple neurobehavioral tests in two separate species at comparable oral
35    doses, and in the absence of maternal or neonatal toxicity (Chen etal.. 2012: Bouayed et al.. 2009a].
36           A decrease in anxiety, indicative of a change in nervous system function, can impair an
37    organism's ability to react to a potentially harmful situation.  This decreased ability of an organism
38    to adapt to the environment is considered to be an adverse effect according to EPA's Neurotoxicity
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-20            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    guidelines [U.S. EPA, 1998]. EPA's Developmental Toxicity guidelines state that an alteration in a
 2    functional outcome following developmental exposure indicates the potential for altered
 3    development in humans, although the types of developmental effects in animals will not necessarily
 4    be the same as those produced in humans [U.S. EPA. 1991c].
 5          Studies in humans also suggest that behavior is adversely affected by benzo[a]pyrene
 6    exposure. In studies where symptoms were self-reported, exposure to benzo[a]pyrene was
 7    positively associated with increased symptoms of anxiety/depression and/or attention problems in
 8    prenatally-exposed children [Pereraetal.. 2012b].  In addition, one study of an occupationally-
 9    exposed cohort of coke oven workers also reported increased symptoms of anxiety/depression
10    and/or attention problems [Qiu etal., 2013]: however, these results were inconsistent with a
11    second cohort of occupationally exposed male workers [Niu etal.. 2010). While these results could
12    suggest that benzo[a]pyrene exposure may adversely affect anxiety-related processes in a different
13    manner across species (i.e., increasing anxiety in humans, but decreasing anxiety in rodents),
14    additional studies in humans are necessary to draw such a conclusion.  Importantly, however, it is
15    recognized that animals and humans may not necessarily experience the same effects or functional
16    changes, and that limitations in the information available across species can sometimes prevent
17    endpoint-specific comparisons [Francis etal., 1990].
18          In conclusion, EPA identified developmental toxicity and developmental neurotoxicity as
19    human hazards of benzo[a]pyrene exposure.

20    Susceptible Populations and Lifestages
21          Childhood susceptibility to benzo[a]pyrene toxicity is indicated by epidemiological studies
22    reporting associations between adverse birth outcomes and developmental effects and internal
23    biomarkers of exposure to benzo[a]pyrene, presumably via exposure to complex PAH mixtures
24    fPereraetal..2012b: Pereraetal.. 2009: Tang etal.. 2008: Tang etal.. 2006: Pereraetal.. 2005b:
25    Pereraetal.. 2005a: Pereraetal.. 2004). The occurrence of benzo[a]pyrene-specific DNAadducts in
26    maternal and umbilical cord blood in conjunction with exposure to ETS was associated with
27    reduced birth weight and head circumference in offspring of pregnant women living in New York
28    City [Pereraetal.. 2005b].  In other studies, elevated levels of BPDE-DNA adducts in umbilical cord
29    blood were associated with: [1] reduced birth weights or reduced head circumference [Perera et
30    al., 2005a: Perera etal., 2004]: and (2] decreased body weight at 18, 24, and 30 months [Tang etal.,
31    2008: Tang etal.. 20061
32          Studies in humans and experimental animals indicate that exposure to PAHs in general,  and
33    benzo[a]pyrene in particular, may impact neurological development. Observational studies in
34    humans have suggested associations between gestational exposure to PAHs and later measures of
35    neurodevelopment [Pereraetal.. 2009: Tang etal..  2008]. In the Perera et al. [2009] study, the
36    exposure measures are based on a composite of eight PAHs measured in air. In Tang etal., [2008],
37    increased levels of benzo[a]pyrene-DNA adducts in cord blood were associated with decreased
38    developmental quotients in offspring [Tang etal.. 2008].
                This document is a draft for review purposes  only and does not constitute Agency policy.
                                               1-21            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          Evidence in animals of the effects of benzo[a]pyrene on neurological development includes:
 2    (1] decrements in reflex-related behaviors associated with neuromuscular coordination and
 3    sensorimotor function [Chenetal.. 2012: Bouayedetal.. 2009a]: (2) disrupted learning and/or
 4    short-term memory processes [Chenetal.. 2012: Lietal.. 2012: Shengetal.. 2010: Bouayedetal..
 5    2009a); and (3) decreased anxiety-related responses [Chenetal.. 2012: Bouayedetal.. 2009a].
 6    Mechanistic studies also support findings of benzo[a]pyrene induced alterations in
 7    electrophysiological response to stimulation of the dentate gyrus of the hippocampus [Wormley et
 8    al.. 2004: Wu etal.. 2003a] and decreased evoked response in the field cortex [McCallister etal..
 9    20081.

10    1.1.2.  Reproductive Toxicity
11          Human and animal studies provide evidence for benzo[a]pyrene-induced male and female
12    reproductive toxicity.  Effects on sperm quality and male fertility have been demonstrated in human
13    populations highly exposed to PAH mixtures [Spares and Melo. 2008: Hsu etal.. 2006]. The use of
14    internal biomarkers of exposure in humans (e.g., BPDE-DNA adducts) support associations between
15    benzo[a]pyrene exposure and these effects. In females, numerous epidemiological studies indicate
16    that cigarette smoking reduces fertility; however, few studies have specifically examined levels of
17    benzo[a]pyrene exposure and female reproductive outcomes.  Animal studies demonstrate
18    decrements in sperm quality, changes in testicular histology, and hormone alterations following
19    benzo[a]pyrene exposure in adult male animals, and decreased fertility and ovotoxic effects in adult
20    females following exposure to benzo[a]pyrene.

21    Male Reproductive Effects

22    Fertility
23          Effects on male fertility have been demonstrated in populations exposed to mixtures of
24    PAHs.  Spermatozoa from smokers have reduced fertilizing capacity, and embryos display lower
25    implantation rates [Spares and Melo.  2008). Occupational PAH exposure has been associated with
26    higher levels of PAH-DNA adducts in sperm and male infertility [Gaspari et al.. 2003]. In addition,
27    men with higher urinary levels of PAH metabolites have been shown to be more likely to be infertile
28    [Xiaetal., 2009].  Studies were not identified that directly examined the reproductive capacity of
29    adult animals following benzo[a]pyrene exposure. However, a dose-related decrease in fertility
30    was observed in male mice treated in utero with benzo[a]pyrene, as discussed in Section 1.1.1.

31    Sperm parameters
32          Effects on semen quality have been demonstrated in populations exposed to mixtures of
33    PAHs including coke oven workers and smokers [Spares and Melo, 2008: Hsu etal., 2006]. Coke
34    oven workers had higher frequency of oligospermia (19 versus 0% in controls] and twice the
35    number of morphologically abnormal sperm (Hsu etal.. 2006]. Elevated levels of BPDE-DNA
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-22            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene

 1    adducts have been measured in the sperm of populations exposed to PAHs occupationally [Gaspari
 2    etal.. 20031 and through cigarette smoke fPhillips. 2002: Zenzes etal.. 19991 A higher
 3    concentration of BPDE-DNA adducts was observed in sperm not selected for intrauterine
 4    insemination or in vitro fertilization based on motility and morphology in patients of fertility clinics
 5    [Perrinetal.. 201 Ib: Perrinetal.. 2011a). An association between benzo[a]pyrene exposure levels
 6    and increased sperm DNA fragmentation using the sperm chromatin structure assay was observed
 7    by Rubes etal. [2010]. However, it is currently unclear whether the sperm chromatin structure
 8    assay, which measures sperm fragmentation following denaturation, is predictive of fertility
 9    [Sakkas and Alvarez. 2010: ASRM. 20081.
10          In several studies in rats and mice, a decrease in sperm count, motility, and production and
11    an increase in morphologically abnormal sperm have been reported (Table 1-5 and Figure 1-3).
12    Alterations in these sperm parameters have been observed in different strains of rats and mice and
13    across different study designs and routes of exposure.
14          Decreases in epididymal sperm counts (25-50% compared to controls) have been reported
15    in Sprague-Dawley rats and C57BL6 mice treated with 1-5 mg/kg-day benzo[a]pyrene by oral
16    exposure for 42 or 90 days (Chen etal., 2011: Mohamed etal., 2010). Another subchronic study
17    noted a 44% decrease in epididymal sperm concentration at doses >50 mg/kg-day in Hsd:ICR
18    (GDI) mice (Tengetal.. 2013). Additionally, a 15% decrease in epididymal sperm count was
19    observed at a much lower dose in Sprague-Dawley rats exposed to benzo[a]pyrene for 90 days
20    (Chung etal.. 2011). However, confidence in this study is limited because the authors dosed the
21    animals with 0.001, 0.01, and 0.1 mg/kg-day benzo[a]pyrene, but only reported on sperm
22    parameters at the mid-dose, and no other available studies demonstrated findings in the range  of
23    the mid- and high-dose. In rats, an oral short-term study and a subchronic inhalation study lend
24    support for the endpoint of decreased sperm count (Arafaetal., 2009: Archibong et al., 2008:
25    Ramesh et al.. 2008). Significantly decreased sperm count and daily sperm production (20-40%
26    decrease from control in each parameter) were observed in rats following 10 days of gavage
27    exposure to 50 mg/kg-day (Arafaetal.. 2009) and following gavage dosing with 10 mg/kg-day  on
28    PNDs 1-7 (Liang etal.. 2012). In addition, a 69% decrease from controls in sperm count was
29    observed in rats following inhalation exposure to 75  |ig/m3 benzo[a]pyrene for 60 days (Archibong
30    etal.. 2008: Ramesh etal.. 20081.
31          Both oral and inhalation exposure of rodents to benzo[a]pyrene have been shown to lead to
32    decreased epididymal sperm motility and altered morphology. Decreased motility of 20-30%
33    compared to controls was observed in  Hsd:ICR (GDI) mice (>100 mg/kg-day), C57BL6 mice (>1
34    mg/kg-day), and Sprague-Dawley rats  (0.01 mg/kg-day) following subchronic oral exposure (Teng
35    etal.. 2013: Chung etal.. 2011: Mohamed etal.. 2010). The effective doses spanned several orders
36    of magnitude; however, as noted above, reporting is limited in the study that observed effects at
37    0.01 mg/kg-day benzo[a]pyrene (Chung etal., 2011). A short-term oral study in rats also reported
38    a significantly decreased number of motile sperm (~40% decrease) following 10 days of gavage
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-23            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    exposure to 50 mg/kg-day benzo[a]pyrene [Arafaetal., 2009]. In addition, decreased sperm
 2    motility was observed following inhalation exposure to 75 [ig/m3 benzo[a]pyrene in rats for 60
 3    days [Archibongetal.. 2008: Rameshetal.. 2008] and to >75 [ig/m3 for 10 days [Inyangetal..
 4    2003]. Abnormal sperm morphology was observed in Sprague-Dawley rats treated with
 5    5 mg/kg-day benzo[a]pyrene by gavage for 84 days [Chenetal.. 2011] and in rats exposed to
 6    75 [J.g/m3 benzo[a]pyrene by inhalation for 60 days [Archibongetal.. 2008: Ramesh et al.. 2008].

 7    Testicular changes
 8          Several studies have demonstrated dose-related effects on male reproductive organs in
 9    adult animals exposed subchronically to benzo[a]pyrene (Table 1-5 and Figure 1-3]. Decreases in
10    testicular weight of approximately 35% have been observed in a 60-day gavage study in Hsd:ICR
11    (GDI] mice at 100 mg/kg-day [Tengetal.. 2013]. in a 10-day gavage study in adult Swiss albino rats
12    at 50 mg/kg-day [Arafaetal.. 2009]. and following subchronic inhalational exposure of adult F344
13    rats to 75 [ig/m3 [Archibong et al.. 2008: Ramesh et al.. 2008]. No effects on testes weight were
14    observed in Wistar rats exposed for 35 days to gavage doses up to 50 mg/kg-day [Kroese etal..
15    2001], F344 rats exposed for 90 days to dietary doses up to 100 mg/kg-day [Knuckles etal., 2001],
16    or Sprague-Dawley rats exposed for 90 days to gavage doses up to 0.1 mg/kg-day [Chung etal.,
17    2011]. Strain differences may have contributed to differences in response; however, F344 rats
18    exposed to benzo[a]pyrene via inhalation showed effects on testicular weight [Archibongetal..
19    2008: Ramesh et al.. 2008].  In addition, decreased testicular weight has also  been observed in
20    offspring following in utero and early postnatal exposure to benzo[a]pyrene as discussed in Section
21    1.1.1.
22          Histological changes in the testis have often been reported to accompany decreases in
23    testicular weight.  Apoptosis, as evident by increases in terminal deoxynucleotidyl transferase dUTP
24    nick end labeling [TUNEL] positive germ cells and increases in caspase-3 staining, was evident in
25    seminiferous tubules of Sprague-Dawley rats following 90 days of exposure to >0.001 and
26    0.01 mg/kg-day, respectively, benzo[a]pyrene by gavage [Chung etal.. 2011]. However, the study
27    authors did not observe testicular atrophy or azospermia in any dose group.  Seminiferous tubules
28    were reported to look qualitatively similar between controls and animals exposed to
29    benzo[a]pyrene by inhalation doses of 75 [ig/m3 for 60 days [Archibongetal., 2008: Ramesh etal.,
30    2008]. However, when histologically examined,  statistically significantly reduced tubular lumen
31    size and length were observed in treated animals.  Seminiferous tubule diameters also appeared to
32    be reduced in exposed animals, although this difference did not reach statistical significance
33    [Archibong et al.. 2008: Ramesh etal.. 2008]. In addition, histological changes in the seminiferous
34    tubules have also been observed in offspring following in utero exposure to benzo[a]pyrene as
35    discussed in Section 1.1.1.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-24            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
 1    Epididymal changes
 2           In addition to testicular effects, histological effects in the epididymis have been observed
 3    following 90-day gavage exposure to benzo[a]pyrene [Chung etal.. 2011] (Table 1-5 and Figure 1-
 4    3). Specifically, statistically significant decreased epididymal tubule diameter (for caput and cauda)
 5    was observed at doses >0.001 mg/kg-day. At the highest dose tested (0.1 mg/kg-day), diameters
 6    were reduced approximately 25%. A 60 day gavage study in Hsd:ICR(CDl) mice observed a 27%
 7    decrease in cauda epididymis weight at 100 mg/kg-day (Jengetal.. 2013]: however, no change in
 8    epididymis weight was observed following an 84-day treatment in Sprague-Dawley rats of 5
 9    mg/kg-day benzo[a]pyrene (Chenetal., 2011].

10    Hormone changes
11           Several animal models have reported decreases in testosterone following both oral and
12    inhalation  exposure to benzo[a]pyrene (Table 1-5 and Figure 1-3]. In male Sprague-Dawley rats,
13    decreases in testosterone have been observed following 90-day oral exposures (Chung etal.. 2011:
14    Zheng etal., 2010]. Statistically significant decreases of 15% in intratesticular testosterone were
15    observed at 5 mg/kg-day in one study (Zheng etal., 2010], while a second study in the same strain
16    of rats reported statistically significant decreases of approximately 40% in intratesticular
17    testosterone and 70% in serum testosterone at 0.1 mg/kg-day (Chung etal.. 2011].  In addition,
18    Sprague-Dawley rats treated with 10 mg/kg-day by gavage on PNDs 1-7 exhibited statistically
19    significantly decreased serum testosterone (>40%] when examined at PND 8 and PND 35 (Liang et
20    al.. 2012].  Statistically significant decreases in intratesticular testosterone (80%] and serum
21    testosterone (60%] were also observed following inhalation exposure to 75 |ig/m3 benzo[a]pyrene
22    in F344 rats for 60 days (Archibongetal., 2008: Ramesh etal., 2008]. Statistically significant
23    increases in serum luteinizing hormone (LH] have also been observed in Sprague-Dawley rats
24    following gavage exposure to benzo[a]pyrene at doses of >0.01 mg/kg-day (Chung etal.. 2011] and
25    in F344 rats following inhalation exposure to 75  |ig/m3 benzo[a]pyrene for 60 days (Archibong et
26    al.. 2008: Ramesh etal.. 2008].

27           Table 1-5. Evidence pertaining to the male reproductive toxicity of
28           benzo[a]pyrene in adult animals
           Reference and study design
                        Results
      Sperm quality
      Mohamed et al. (2010)
      C57BL/6 mice, 10 males/dose (treated
      before mating with unexposed females)
      0,1, or 10 mg/kg-d by gavage (FO males
      only)
      42 d
•^ epididymal sperm count in FO mice
Approximate % change from control (data reported graphically)
       0, -50*, and -70*

4, epididymal sperm motility in FO mice
Approximate % change from control (data reported graphically):
       0, -20*, and -50*
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-25            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
     Reference and study design
                           Results
                                      4/ epididymal sperm count in untreated Fl and F2 generations (data
                                      reported graphically)

                                      No effects were observed in the F3 generation
Chen etal. (2011)
Sprague-Dawley rats, 10 males/dose
0 or 5 mg/kg-d by gavage
84 d
  epididymal sperm count (% change from control)
        0 and-29*

  % abnormal epididymal sperm
        5 and 8*
Chung etal. (2011)
Sprague-Dawley rats, 20-25 males/dose
0, 0.001, 0.01, or 0.1 mg/kg-d by gavage
90 d
4/ epididymal sperm motility (% change relative to control; reported
only for 0.01 mg/kg-d)
        0 and -30*

No statistically significant decrease in epididymal sperm count
(Archibong et al. (2008); Ramesh et al.
(2008))
F344 rats, 10 males/group
0 or 75 u.g/m3, 4 hrs/d by inhalation
60 d
•^ epididymal sperm motility (% change from control)
        0 and-73*

4, epididymal sperm count (% change from control)
        0 and-69*

/T" % abnormal epididymal sperm
        33 and 87*

4, spermatids/g testis (approximate % change from control;
numerical data not reported)
        0 and -45*
Testicular changes (weight, histology)
Mohamed et al. (2010)
C57BL/6 mice, 10 males/dose (treated
before mating with unexposed females)
0,1, or 10 mg/kg-d by gavage (FO males
only)
42 d
4/ seminiferous tubules with elongated spermatids (approximate %
change from control; numerical data not reported)
        0, -20*, and -35*

No statistically significant change in area of seminiferous epithelium
of testis (approximate % change from control; numerical data not
reported)
        0, 5, and 20
Chung etal. (2011)
Sprague-Dawley rats, 20-25 males/dose
0, 0.001, 0.01, or 0.1 mg/kg-d by gavage
90 d
/T" number of apoptoticgerm cells per tubule (TUNELor caspase 3
positive)

No change in testis weight or histology
Chen etal. (2011)
Sprague-Dawley rats, 10/dose
0 or 5 mg/kg-d by gavage
84 d
1" testicular lesions characterized as irregular arrangement of germ
cells and absence of spermatocytes (numerical data not reported)

No change in testis weight
(Archibong et al. (2008); Ramesh et al.
(2008))
F344 rats, 10 adult males/group
  decreased testis weight (% change from control)
        0 and 34*
           This document is a draft for review purposes only and does not constitute Agency policy,
                                               1-26              DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
Reference and study design
0 or 75 u.g/m3, 4 hrs/d by inhalation
60 d
Kroese et al. (2001)
Wistar rats, 10/sex/dose
0, 1.5, 5, 15, or 50 mg/kg-d by gavage,
5d/wk
35 d
Knuckles etal. (2001)
F344 rats, 20/sex/dose
0, 5, 50, or 100 mg/kg-d in diet
90 d
Epididymal changes (weight, histology)
Chung etal. (2011)
Sprague-Dawley rats, 20-25 males/dose
0, 0.001, 0.01, or 0.1 mg/kg-d by gavage
90 d
Chen etal. (2011)
Sprague-Dawley rats, 10/dose
0 or 5 mg/kg-d by gavage
84 d
Hormone changes
Chung etal. (2011)
Sprague-Dawley rats, 20-25 males/dose
0, 0.001, 0.01, 0.1 mg/kg-d by gavage
90 d
Zheng etal. (2010)
Sprague-Dawley rats, 8 males/dose
0, 1, or 5 mg/kg-d by gavage
90 d
(Archibong et al. (2008); Ramesh et al.
(2008))
F344 rats, 10 adult males/group
0 or 75 u.g/m3, 4 hrs/d by inhalation
Results
•i, size of seminiferous tubule lumens and reduced tubular length
No change in % of tubules with elongated spermatids
No change in testis weight

No change in testis weight


•i, diameter of caput epididymal tubule (n = 5; numerical data not
reported)
•^ diameter of cauda epididymal tubule (n = 5; numerical data not
reported)
No change in epididymis weight


•i, Intratesticular testosterone (approximate % change from control;
data reported graphically)
0, -12, -25, and -40*
•^ Serum testosterone (approximate % change from control;
numerical data not reported)
0, 0, -35, and -70*
/T" serum LH (approximate % change from control; numerical data not
reported)
0,33, 67*, and 87*
•i, Human chorionic gonadotropin (hCG) or dibutyl cyclic adenosine
monophosphate (dbcAMP)-stimulated testosterone production in
Leydig cells
•^ Intratesticular testosterone (approximate % change from control;
numerical data not reported)
0, -15, and -15*
•i, intratesticular testosterone (approximate % change from control;
numerical data not reported)
0 and -80*
This document is a draft for review purposes only and does not constitute Agency policy,
                                1-27            DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
1
2
3
4
5
6
Reference and study design Results
60 d 4, serum testosterone (approximate % change from control)
0 and -60*
1" serum LH (approximate % change from control)
0 and 50*
*Statistically significantly different from the control (p< 0.05).
a% change from control calculated as: (treated value - control value)/control value x 100.





100 :


1 10 =
CUD
-Si
E ! '-

o
0
0.1 =


0.01 =



n nm





A NUALL

o Doses < NOAEL

=•=
B B JL

-1- =t±=


	 T 	 T 	 T 	


^ i= ^ _i i_




O __i ^ ^H Wl . t/l ^H yi ^H yi O ^
o S g2 32 g2 32 32 32
r-j^ r\l'* rvj . fc rsj ' " r\j.* fN-* rN-*
_ d >• -.>- >- ^>- -.>- ^>*
	 : "T3 	 ^~O "~O 	 :~O 	 ^~O 	 :~C
<"Q ^^ ~3^ 3 fU ^ ~7^ 3 fO ^ <^ ^ fO 3
If || || g| || || ||
~ -JS i^i^ _c o 1*1 ^ _co _co ^=0
i^V "^ 00 LJCTi WDO O O LJCTi MC^i
{9
sU epididymal -J, epididymal %b epididymal Irregular ^/ TT (serum ^ LH \|/ TT
sperm count sperm count sperm motility germ cell and (intratesticular)
and motility INnatanormal organization intratesticular}
sperm
Sperm Quality Testicular Hormone Changes
Effects
Figure 1-3. Exposure-response array for male reproductive effects following
oral exposure in adult animals.
This document is a draft for review purposes only and does not constitute Agency policy,
                                1-28            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    Mode-of-action analysis—male reproductive effects
 2           Exposure to benzo[a]pyrene in laboratory animals induces male reproductive effects
 3    including decreased levels of testosterone and increased levels of LH, decreased sperm quality, and
 4    histological changes in the testis. Decrements in sperm quality and decreased fertility have also
 5    been demonstrated in populations highly exposed to PAH mixtures [Spares and Melo. 2008: Hsu et
 6    al.. 20061
 7           Numerous studies have indicated that benzo[a]pyrene reduces testosterone levels following
 8    oral or inhalation exposure [Chung etal., 2011: Zheng etal., 2010: Archibongetal., 2008: Ramesh et
 9    al., 2008]. It is plausible that the effects on sperm quality and histological changes of the
10    reproductive organs are secondary to an insufficiency of testosterone [Inyangetal.. 2003). One
11    study has hypothesized that benzo[a]pyrene perturbs the production of testosterone by Leydig
12    cells [Chung etal.. 2011). This study found a statistically significant reduction in testicular
13    testosterone in rats treated with 0.1 mg/kg-day benzo[a]pyrene for 90 days and found that
14    testosterone production in isolated Leydig cells was also inhibited approximately 50%, even in
15    cultures stimulated with human chorionic gonadotropin and dibutryl cyclic adenosine
16    monophosphate.
17           Leydig cell function is thought to be regulated by testicular macrophages [Hales. 2002).
18    When testicular macrophages are activated and produce inflammatory mediators, Leydig cell
19    testosterone production is inhibited [Hales. 2002).  Zheng etal.  [2010] treated rats with 5 mg/kg-
20    day benzo[a]pyrene for 90 days and reported a statistically significant increase in ED-1 type
21    testicular macrophages and a statistically significant decrease in intratesticular testosterone.
22           Arafaetal. [2009] reported that male reproductive effects observed following
23    benzo[a]pyrene exposure could be ameliorated by antioxidant pre-treatment. This study reported
24    decreased sperm count, motility, and production, in addition to  decreased testis weight following a
25    10 day oral administration in rats of 50 mg/kg-day benzo[a]pyrene. Pretreatment with the citris
26    flavonoid hesperidin protected rats from all of these effects except the decrease in sperm motility.
27           A study in tobacco smokers suggests that direct DNA damage from the reactive metabolite
28    BPDE may decrease sperm motility [Perrinetal.. 2011a]. In this study,  motile sperm were
29    separated from non-motile sperm using a "swim-up" self-migration technique. The investigators
30    found that the motile sperm selected by this method had significantly fewer BPDE-adducts than
31    non-selected sperm.
32           Other hypothesized modes of action of the observed male reproductive effects include
33    benzo[a]pyrene-mediated DNA damage to male germ cells leading to genotoxicity, cytotoxicity, and
34    apoptosis [Chungetal.. 2011: Perrinetal.. 2011b: Perrinetal.. 201 la: Olsenetal.. 2010: Revel et
35    al.. 2001]. compromised function of Sertoli cells [Raychoudhury and Kubinski. 2003]. and
36    decreased embryo viability post-fertilization associated with sperm DNA damage [Borinietal.,
37    2006:  Seli etal..  20041
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-29            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    Female Reproductive Effects

 2    Fertility
 3          In women, exposure to cigarette smoke has been shown to affect fertility, including effects
 4    related to pregnancy, ovulatory disorders, and spontaneous abortion (as reviewed in Waylen etal..
 5    2009: Cooper and Moley. 2008: Spares and Melo. 2008). In addition, several studies suggest that in
 6    utero exposure to maternal tobacco smoke also decreases the future fertility of female offspring [Ye
 7    etal.. 2010: Jensen etal.. 1998: Weinbergetal.. 1989]. Benzo[a]pyrene levels in follicular fluid and
 8    benzo[a]pyrene-DNA adducts in granulosa-lutein cells and oocytes and in human cervical cells have
 9    been associated with smoking status and with amount smoked [Neal etal.. 2008: Mancinietal..
10    1999: Melikian etal.. 1999: Zenzes etal.. 1998: Shamsuddin and Can. 19881
11          Few epidemiological studies have examined the specific influence of components of PAH
12    mixtures on fertility or other reproductive outcomes; EPA identified only two studies with specific
13    data on benzo[a]pyrene  (Table 1-6). One of these studies addressed the probability of conception
14    among women undergoing in vitro fertilization [Neal etal., 2008]. Follicular fluid benzo[a]pyrene
15    levels were significantly higher among the women who did not conceive compared with women
16    who did get pregnant.  No association was seen between conception and serum levels of
17    benzo[a]pyrene. The other study examined risk of delayed miscarriage (fetal  death before
18    14 weeks of gestation), using a case-control design with controls selected from women undergoing
19    elective abortion (Wu etal.. 2010). A strong association was seen between maternal blood
20    benzo[a]pyrene-DNA adduct levels and risk of miscarriage, with a fourfold increased risk for levels
21    above compared with below the median.  Benzo[a]pyrene-DNA adduct levels were similar in the
22    aborted tissue of cases compared with controls.
23          Experimental studies in mice also provide evidence thatbenzo[a]pyrene exposure affects
24    fertility (Table 1-7 and Figure 1-4). Decreased fertility and fecundity (decreased number of FO
25    females producing viable litters at parturition) was statistically significantly reduced by about 35%
26    in adult females exposed to 160 mg/kg-day of benzo[a]pyrene (Mackenzie and Angevine. 1981).  In
27    another study, FO females showed no signs of general toxicity or effects on fertility following gavage
28    exposure to 10 mg/kg-day on CDs 7-16 (Kristensen et al., 1995).  Decrements infertility were
29    more striking in the offspring from these studies, as described in Section 1.1.1.

30    Ovarian effects
31          Human epidemiological studies that directly relate ovotoxicity and benzo[a]pyrene
32    exposure are not available; however, smoking, especially during the time of the peri-menopausal
33    transition, has been shown to accelerate ovarian senescence (Midgette and Baron. 1990).
34    Benzo[a]pyrene-induced ovarian toxicity has been demonstrated in animal studies. In adult female
35    rats treated by gavage, statistically significant, dose-related decreases in ovary weight have been
36    observed in female rats treated for 60 days at doses >5 mg/kg (2.5 mg/kg-day adjusted)  (Xu etal..
37    2010). At 10 mg/kg in adult rats (5 mg/kg-day adjusted), ovary weight was decreased 15% (Xu et

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-30            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    al., 2010].  Changes in ovary weight were not observed in two subchronic studies in rats.
 2    Specifically, no changes in ovary weight were seen in Wistar rats exposed for 35 days to gavage
 3    doses up to 50 mg/kg-day [Kroese etal.. 2001] or in F344 rats exposed for 90 days to dietary doses
 4    up to 100 mg/kg-day (Knuckles etal.. 2001).
 5           In adult female rats treated by gavage, dose-related decreases in the number of primordial
 6    follicles have been observed in female rats treated for 60 days at doses >2.5  mg/kg-day, with a
 7    statistically significant decrease of approximately 2 0% at the high dose (Xu etal.. 2010] (Table 1-7
 8    and Figure 1-4). No notable differences in other follicle populations and corpora lutea were
 9    observed.  However, in utero studies exposing dams to the same doses produced offspring with few
10    or no follicles or corpora lutea (Kristensenetal., 1995: Mackenzie and Angevine, 1981]. Additional
11    support for the alteration of female reproductive endpoints comes from intraperitoneal (i.p.)
12    experiments in animals and in vitro experiments. Several studies have observed ovarian effects
13    (decreased numbers of ovarian follicles and corpora lutea, absence of folliculogenesis, oocyte
14    degeneration, and decreased fertility) in rats and mice exposed via i.p. injection (Borman etal..
15    2000: Miller etal.. 1992: Swartz and Mattison. 1985: Mattison etal.. 1980). Further evidence is
16    available from in vitro  studies showing inhibition of antral follicle  development and survival, as
17    well as decreased production of estradiol, in mouse ovarian follicles cultured with benzo[a]pyrene
18    for 13 days (Sadeu and Foster. 2011). Likewise, follicle stimulating hormone (FSH)-stimulated
19    growth of cultured rat ovarian follicles was inhibited by exposure  to benzo[a]pyrene (Neal etal..
20    20071

21    Hormone levels
22           Alteration of hormone levels has been observed in female rats following oral or inhalation
23    exposure to benzo[a]pyrene (Table 1-7 and Figure 1-4). Inhalation exposure to
24    benzo[a]pyrene:carbon black particles during gestation resulted in decreases in plasma
25    progesterone, estradiol, and prolactin in pregnant rats (Archibongetal.. 2002). In addition,
26    statistically significant, dose-related decreases in estradiol along with altered estrus cyclicity was
27    observed in female rats treated for 60 days at doses >2.5 mg/kg-day by gavage (Xu etal.. 2010).
28    Mechanistic experiments have also noted decreased estradiol output in murine ovarian follicles
29    cultured with benzo[a]pyrene in vitro for 13 days, but did not find any decrease in progesterone
30    (Sadeu and Foster. 2011).

31    Cervical effects
32           One subchronic animal study is available that investigated effects in  the cervix following
33    oral exposure to benzo[a]pyrene (Table 1-7 and Figure 1-4). Statistically-significant dose-related
34    increases in the incidence  of cervical inflammatory cells were observed in mice exposed twice a
35    week for 98 days to benzo[a]pyrene via gavage at doses >2.5 mg/kg (Gao etal., 2011).  Cervical
36    effects of increasing severity, including epithelial hyperplasia, atypical hyperplasia, apoptosis, and
37    necrosis, were observed at higher doses. There are no data on cervical effects in other species or in

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-31            DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review ofBenzo[a]pyrene


1    other mouse strains. Gao etal. [2011] considered the hyperplasia responses to be preneoplastic
2    lesions. Cervical neoplasia was not reported in the available chronic bioassays, but this tissue was
3    not subjected to histopathology examination in either bioassay [Kroese etal.. 2001: Beland and
4    Gulp. 1998). Thus, the relationship of the cervical lesions to potential development of neoplasia is
5    uncertain.
6
7
Table 1-6.  Evidence pertaining to the female reproductive effects of
benzo[a]pyrene in humans
              Reference and study design
                                                            Results
      Probability of conception
      Neal etal. (2008)
      36 women undergoing in vitro fertilization
      (19 smokers, 7 passive smokers, and
      10 nonsmokers)

      Exposure: benzo[a]pyrene in serum and follicular
      fluid
                                      Benzo[a]pyrene levels (ng/mL)
                                      Follicular fluid
                                      Serum
Conceived
   0.1
  0.01
 Did not
conceive
  1.7
  0.05
  p-value
  <0.001
Not reported
      Fetal death
      Wu et al. (2010) (Tianjin, China)
      Case control study: 81 cases (96% participation
      rate)—fetal death confirmed by ultrasound before
      14 wks gestation; 81 controls (91% participation
      rate)—elective abortions; matched by age,
      gestational age, and gravidity; excluded smokers
      and occupational PAH exposure

      Exposure: benzo[a]pyrene in aborted tissue and
      maternal blood samples (51 cases and controls,
      2 of 4 hospitals)
                                      Benzo[a]pyrene adduct levels (/10s nucleotides), mean
                                      (±SD)

                                                       Cases       Controls      p-value

                                      Maternal blood  6.0 (±4.7)    2.7 (± 2.2)      <0.001

                                      Aborted tissue   4.8 (± 6.0)    6.0 (± 7.4)        0.29

                                      Low correlation between blood and tissue levels (r = -0.02 in
                                      cases, r = -0.21 in controls)


                                      Association between benzo[a]pyrene adducts and
                                      miscarriage3

                                                                   OR          95% Cl

                                      Per unit increase in adducts      1.37        1.12,1.67

                                      Dichotomized at median         4.56        1.46,14.3

                                      Conditional logistic regression, adjusted for maternal
                                      education, household income, and gestational age; age also
                                      considered as potential confounder
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                    1-32             DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review ofBenzo[a]pyrene
1
2
Table 1-7.  Evidence pertaining to the female reproductive effects of
benzo[a]pyrene in adult animals
         Reference and study design
                                                       Results3
      Fertility
      Mackenzie and Angevine (1981)
      CD-I mice, 30 or 60 FO females/dose
      0,10, 40, or 160 mg/kg-d by gavage
      CDs 7-16
                             number of FO females with viable litters
                                  46/60, 21/30, 44/60, and 13/30*
      Kristensenetal. (1995)
      NMRI mice, 9 females/dose
      0 or 10 mg/kg-d by gavage
      CDs 7-16
                           No changes in fertility of FO females
      Ovarian effects (weight, histology, follicle numbers)
      Xuetal. (2010)
      Sprague-Dawley rats, 6 females/
      dose
      0, 5 or 10 mg/kg by gavage every
      other day (2.5 and 5  mg/kg-d,
      adjusted)
      60 d
                           •^ ovary weight (% change from control)
                                  0, -11*, and -15*
                           •^ number of primordial follicles (20%* decrease at high dose)

                           1" increased apoptosis of ovarian granulosa cells (approximate %
                           apoptosis)
                                  2, 24*, and 14*
      Knuckles etal. (2001)
      F344 rats, 20/sex/dose
      0, 5, 50, or 100 mg/kg-d in diet
      90 d
                           No changes in ovary weight
      Kroese et al. (2001)
      Wistar rats, 10/sex/dose
      0,1.5, 5,15, or 50 mg/kg-d by gavage
      5d/wk
      35 d
                           No changes in ovary weight
      Hormone levels
      Xuetal. (2010)
      Sprague-Dawley rats, 6 females/
      dose
      0, 5, or 10 mg/kg by gavage every
      other day (2.5 and 5  mg/kg-d,
      adjusted)
      60 d
                           •^ serum estradiol (approximate % change from control)
                                  0, -16, and -25*

                           Altered estrous cyclicity
      Archibong et al. (2002)
      F344 rats, 10 females/group
      0, 25, 75, or 100 u.g/m3 by inhalation
      4 hrs/d

      GDs 11-20 (serum hormones tested
      at GD 15 and 17 in 0, 25, and
      75 u.g/m3 dose groups)
                           4> FO estradiol, approximately 50% decrease at 75 |Jg/m3 at GD 17

                           •^ FO prolactin, approximately 70% decrease at 75 u.g/m3at GD 17

                           /T" FO plasma progesterone approximately 17% decrease at 75 u.g/m3 at
                           GD17
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                    1-33              DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
Reference and study design
Results3
Cervical effects
Gaoetal. (2011)
ICR mice, 26 females/dose
0, 2.5, 5, or 10 mg/kg by gavage
2d/wk
98 d
-t cervical epithelial hyperplasia: 0/26, 4/26, 6/25*, and 7/24*
T* cervical atypical hyperplasia: 0/26, 0/26, 2/25, and 4/24*
T* inflammatory cells in cervical epithelium: 3/26, 10/26, 12/25*, and
18/24*
1
2
3
*Statistically significantly different from the control (p< 0.05).
a% change from control calculated as: (treated value - control value)/control value x 100.
1000 T 	 1






**" 1 nn



CUD



(U
o
Q
10 -







i







3






L

<





1
/
_


A
L
\

I
3 A •
0


m .
• <



c
^H u^
00 Cl
O CTi „
^ 8
.E 01 ra .-
S 'E ° S
QD C C i
e .- oj r-
1? P
QJ O -i'
c
o
•^3
sl/FO female fertility
Fertility Effects
A '


j




V


• LOAEL
ANOAEL
• Doses > LOAEL
O Doses < NOAEL





















^ A
o ^^ ^


)



^ »H (/I T— 1
0 CJ g - 0 «
,— | - O j_ O ^
o >; ^ ^ '- ; :
:• ^ "ro 3 "ra TD
® -g 2^ ^ ^ S"
X o O LTI t_) i
LO >; ro =! O
— c tJi
4^ ovarian weight
Ovarian Effects
O ra
T— 1
^ "S
*^ >-
o ^
x 6
4- ovarian
follicles
• 	 • 	
T 	 f 	


A M
$ m

A

o 2
^ "S
O ^3
x 6
4- serum
Estradiol
Hormone
Changes









o ^_
CJ >.
(5 ob
T cervical
epithelial
hyperplasia
Cervical
Effects

5
6
       Figure 1-4. Exposure-response array for female reproductive effects following
       oral exposure in adult animals.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                 1-34            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    Mode-of-action analysis—female reproductive effects
 2          Although the mechanisms underlying female reproductive effects following benzo[a]pyrene
 3    exposure are not fully established, associations with stimulation of apoptosis, impairment of
 4    steroidogenesis, and cytotoxicity have been made. Ovarian lesions in benzo[a]pyrene-exposed rats
 5    have been associated with increased apoptosis in ovarian granulosa cells and alteration in
 6    hormone-mediated regulation of folliculogenesis [Xu etal.. 2010). and results from in vitro
 7    experiments provide support for an association between benzo[a]pyrene exposure and impaired
 8    folliculogenesis, steroidogenesis, and oocyte maturation [Sadeu and Foster, 2011: Nealetal., 2007].
 9    A growing body of research suggests that benzo[a]pyrene triggers the induction of apoptosis in
10    oocytes through AhR-driven expression of pro-apoptotic genes, including Bax [Kee etal.. 2010: Neal
11    etal.. 2010: Pru etal.. 2009: Matikainen etal.. 2002: Matikainen etal.. 2001: Robles etal.. 20001
12    Other proposed mechanisms include the impairment of folliculogenesis from reactive metabolites
13    [Takizawa  et al.. 1984: Mattison and Thorgeirsson. 1979.1977] or by a decreased sensitivity to
14    FSH-stimulated follicle growth [Neal etal.. 2007]. Based on findings that an ERa antagonist
15    counteracted effects of subcutaneously administered benzo[a]pyrene on uterine weight (decreased
16    in neonatal rats and increased in immature rats], interactions with ERa have been proposed,
17    possibly via occupation of ERa binding sites or via AhR-ER-crosstalk [Kummer et al.. 2008: Kummer
18    etal.. 2007]. However, several in vitro studies have demonstrated low affinity binding of
19    benzo[a]pyrene to the estrogen receptor and alteration of estrogen-dependent gene expression
20    [Liu etal.. 2006: van Lipzigetal.. 2005: Vondraceketal.. 2002: Fertucketal.. 2001: Charles etal..
21    2000]. so the role of the ER in benzo [a]pyrene-induced reproductive toxicity is unclear.

22    Summary of Reproductive Effects

23    Male reproductive effects
24          Exposure to benzo [a]pyrene in laboratory animals induces male reproductive effects
25    including decreased levels of testosterone and increased levels of LH, decreased sperm count and
26    motility, histological changes in the testis, and decreased reproductive success.  These findings in
27    animals are supported by decrements in sperm quality and decreased fertility in human
28    populations exposed to PAH mixtures [Spares and Melo, 2008: Hsu etal., 2006]. In laboratory
29    animals, male reproductive toxicity has been observed after oral and inhalation exposure to rats or
30    mice. Effects seen after oral exposures include impaired fertility, effects on sperm parameters,
31    decreased reproductive organ weight, testicular lesions, and hormone alterations  [Chen etal..  2011:
32    Chung etal.. 2011: Mohamedetal.. 2010: Zheng etal.. 2010: Mackenzie and Angevine. 1981]. In
33    addition to oral exposure, male reproductive effects of benzo [a] pyrene have also been observed
34    following inhalation exposure in rats [Archibongetal., 2008: Rameshetal., 2008: Inyangetal.,
35    2003]. The male reproductive effects associated with benzo[a]pyrene exposure are considered to
36    be biologically plausible and adverse.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-35            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1          In conclusion, EPA identified male reproductive system effects as a human hazard of
 2    benzo[a]pyrene exposure.

 3    Female reproductive effects
 4          A large body of mechanistic data, both in vivo and in vitro, suggests that benzo[a]pyrene
 5    impacts fertility through the disruption of folliculogenensis. This finding is supported, albeit
 6    indirectly, by observations of premature ovarian senescence in women exposed to cigarette smoke
 7    [Midgette and Baron. 1990].  Evidence for female reproductive toxicity of benzo[a]pyrene comes
 8    from studies of human populations exposed to PAH mixtures as well as laboratory animal and in
 9    vitro studies. In addition, two human studies observed associations specifically between
10    benzo[a]pyrene measures and two fertility-related endpoints:  decreased ability to conceive [Neal et
11    al.. 2008: Neal etal.. 2007] and increased risk of early fetal death (i.e., before 14 weeks of gestation]
12    [Wuetal.. 2010].  Studies in multiple strains of rats and mice indicate fertility-related effects
13    including decreases in ovarian follicle populations and decreased fecundity.  Decreased serum
14    estradiol has also been noted in two different strains of rats exposed by oral or inhalation exposure.
15    The reproductive effects associated with benzo[a]pyrene exposure are biologically supported and
16    relevant to humans.
17          In conclusion, EPA identified female reproductive effects as a human hazard of
18    benzo[a]pyrene exposure.

19    Susceptible Populations and Lifestages
20          Epidemiological studies indicate that exposure to complex mixtures of PAHs, such as
21    through cigarette smoke, is associated with measures of decreased fertility in humans [Neal etal.,
22    2008: El-Nemr etal., 1998] and that prenatal exposure to cigarette smoking is associated with
23    reduced fertility of women later in life [Weinberg et al.. 1989].  A case-control study in a Chinese
24    population has also indicated that women with elevated levels of benzo[a]pyrene-DNA adducts in
25    maternal blood were 4 times more likely to have experienced a miscarriage [Wu etal.. 2010].
26          Inhalation exposure of pregnant female rats to benzo[a]pyrene:carbon black aerosols on
27    CDs 11-20 caused decreased fetal survival and number of pups per litter associated with decreased
28    levels of plasma progesterone, estradiol, and prolactin [Archibongetal., 2002]. Decreased numbers
29    of live pups were  also seen in pregnant mice following i.p. exposure to benzo[a]pyrene [Mattison et
30    al.. 1980]. These results indicate that benzo[a]pyrene exposure can decrease the ability of females
31    to maintain pregnancy.
32          Oral multigenerational studies of benzo[a]pyrene exposure in mice demonstrated effects on
33    fertility and the development of reproductive organs (decreased ovary and testes weight] in both
34    male and female offspring of pregnant mice exposed to 10-160 mg/kg-day on CDs 7-16
35    (Kristensen etal., 1995: Mackenzie and Angevine, 1981].
36          Reductions in female fertility associated with decreased ovary weight and follicle number
37    following gestational exposure (as discussed in Section 1.1.1] are supported by observations of:

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-36            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    (1) destruction of primordial follicles [Borman et al., 2000: Mattisonetal., 1980] and decreased
 2    corpora lutea [Miller etal.. 1992: Swartz and Mattison. 1985] in adult female mice following i.p.
 3    exposure; (2] decreased ovary weight in adult female rats following oral exposure [Xu etal.. 2010]:
 4    and (3] stimulation of oocyte apoptosis [Matikainenetal.. 2002: Matikainen etal.. 2001] or by a
 5    decreased sensitivity to FSH-stimulated follicle growth [Neal etal.. 2007].
 6          Reductions in male fertility associated with decreased testes weight following gestational
 7    exposure (as discussed in Section 1.1.1] are supported by observations of: (1] decreased sperm
 8    count, altered serum testosterone levels, testicular lesions, and/or increased numbers of apoptotic
 9    germ cells in adult rats following repeated oral exposure to benzo[a]pyrene [Chung etal., 2011:
10    Chen etal., 2010: Zheng etal., 2010: Arafa etal., 2009]: (2] decreased epididymal sperm counts in
11    adult FO  and Fl generations of male mice following 6 weeks of oral exposure of the FO animals to
12    benzo[a]pyrene [Mohamed etal.. 2010]: and (3] decreased testis weight, decreased testicular or
13    plasma testosterone levels, and/or decreased sperm production, motility, and density in adult male
14    rats following repeated inhalation exposure to aerosols of benzo[a]pyrene:carbon black [Archibong
15    etal.. 2008: Ramesh etal.. 2008: Inyangetal.. 2003].

16    1.1.3. Immunotoxicity
17          Human studies evaluating immune effects following exposure to benzo[a]pyrene alone are
18    not available for any route of exposure. However, a limited number of occupational human studies,
19    particularly in coke oven workers [Zhang etal.. 2012: Wu etal.. 2003b: Winker et al.. 1997:
20    Szczekliketal.. 1994]. show effects on immune parameters associated with exposure to PAH
21    mixtures. These studies are of limited utility because effects associated specifically with
22    benzo[a]pyrene cannot be distinguished from other constituents of the PAH mixture. Subchronic
23    and short-term animal studies have reported immunotoxic effects of benzo[a]pyrene by multiple
24    routes of exposure (Table 1-8 and Figure 1-5]. Effects include changes in thymus weight and
25    histology, decreased B cell percentages and other alterations in the spleen, and immune
26    suppression. Data obtained from subchronic oral gavage studies are supported by short-term, i.p.,
27    intratracheal, and subcutaneous (s.c.] studies. Additionally, there is evidence in animals for effects
28    of benzo[a]pyrene on the developing immune system. No studies were located that examined
29    immune  system endpoints following inhalation exposure of animals to benzo[a]pyrene.

30    Thymus Effects
31          Decreased thymus weights (up to 62% compared to controls] were  observed in male and
32    female Wistar rats exposed by gavage to 10-90 mg/kg-day benzo[a]pyrene for 35 or 90 days
33    (Kroese etal.. 2001: De Jong etal.. 1999]. This effect may be due to thymic atrophy. The incidence
34    of slight thymic atrophy was increased in males (6/10] and females (3/10]  at a dose of
35    30 mg/kg-day in a 90-day study, although there was no evidence of atrophy at any lower dose
36    (Kroese et al., 2001]. Additionally, at the highest dose tested (90 mg/kg-day] in one of the 35-day
37    studies, the relative cortex surface area of the thymus and thymic medullar  weight were

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-37            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    significantly reduced [De Jongetal.. 1999]. Other histopathological changes in the thymus
 2    (increased incidence of brown pigmentation of red pulp; hemosiderin) were observed in Wistar
 3    rats of both sexes at 5 0 mg/kg-day in a 3 5-day study; however, this tissue was not examined in
 4    intermediate-dose groups [Kroese etal.. 2001). Consistent with the effects observed in these
 5    studies, decreased thymus weights and reduced thymic cellularity were observed in i.p. injection
 6    studies that exposed mice to doses ranging from 50 to 150 mg/kg in utero [Holladay and Smith.
 7    1995.1994: Urso and Johnson. 1988).

 8    Spleen Effects
 9          Reduced splenic cellularity, indicated by decreased relative and absolute number of B cells
10    in the spleen (decreased up to 41 and 61% compared to controls, respectively) and decreased
11    absolute number of splenic cells (31% decrease atthe highest dose), was observed in a subchronic
12    study in male Wistar rats administered 3-90 mg/kg-day benzo[a]pyrene by gavage for 35 days (De
13    Jongetal.. 1999). While the effect on the relative number of B cells was dose-related, the lower
14    doses did not affect the number of B cells or the absolute splenic cell number.  The reduced splenic
15    cell count at the highest dose was attributed by the study authors to the decreased B cells, and
16    suggests a possible selective toxicity of benzo[a]pyrene to B cell precursors in the bone marrow.
17    The spleen effects observed in De long et al. (1999) are supported by observations of reduced
18    spleen cellularity and decreased spleen weights following i.p. injection or in utero benzo[a]pyrene
19    exposure to doses ranging from 50 to 150 mg/kg (Holladay and Smith. 1995: Urso etal.. 1988).
20          In addition to physical effects on the spleen, several studies have demonstrated functional
21    suppression of the spleen following benzo[a]pyrene exposure.  Dose-related decreases in sheep red
22    blood cell (SRBC) specific serum IgM levels after SRBC challenge were reported in rats (10 or
23    40 mg/kg-day) and mice (5, 20,or 40 mg/kg-day) following s.c. injection of benzo[a]pyrene for
24    14 days (Temple et al.. 1993). Similarly, reduced spleen cell responses, including decreased
25    numbers of plaque forming cells and reduced splenic phagocytosis to SRBC and lipopolysaccharide
26    challenge, were observed in B6C3Fi mice exposed to doses >40 mg/kg-day benzo[a]pyrene by i.p.
27    or s.c. injection for 4-14 days (Lyte and Bick. 1985: Dean etal.. 1983: Munson and White. 1983] or
28    by intratracheal instillation for 7 days (Schnizlein et al.. 1987).

29    Immunoglobulin Alterations
30          Alterations in immunoglobulin levels have been associated with exposure to PAH mixtures
31    in a limited number of human studies. Some  occupational studies have reported evidence of
32    immunosuppression following PAH exposure. For example, reductions in serum IgM and/or IgA
33    titers were reported in coke oven workers (Wuetal.. 2003b: Szczeklik et al.. 1994). Conversely,
34    immunostimulation of immunoglobulin levels has also been observed in humans, specifically
35    elevated IgG (Karakaya et al., 1999] and elevated IgE (Wu etal., 2003b] following occupational PAH
36    exposure.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-38            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          Decreases in serum IgM (13-33% compared to controls) and IgA levels (22-61% compared
 2    to controls) were observed in male Wistar rats exposed to 3-90 mg/kg-day benzo[a]pyrene by
 3    gavage for 35 days (De long etal.. 1999): however, these reductions were not dose-dependent.
 4    Similarly, reductions in IgA (9-38% compared to controls) were also observed in male and female
 5    B6C3Fi mice exposed to doses of 5-40 mg/kgbenzo[a]pyreneby s.c. injection for 14 days (Munson
 6    and White. 1983). Reductions in serum IgG levels of 18-24%, although not statistically significant,
 7    were observed in female B6C3Fi mice exposed to doses >50 mg/kg benzo[a]pyrene by i.p. injection
 8    for 14 days (Dean etal.. 1983).

 9    Immune Suppression and Sensitization
10          Some occupational studies of coke oven emissions have reported evidence of
11    immunosuppression following PAH exposure.  Reduced mitogenic responses in T cells (Winker et
12    al.. 1997) and reduced T-lymphocyte proliferative responses (Karakaya et al.. 2004) have been
13    observed following occupational exposure to PAH. Increased levels of apoptosis were observed in
14    the peripheral blood mono nuclear cells (a population of lymphocytes and monocytes) of
15    occupationally exposed coke oven workers, which is a response that may contribute to
16    immunodeficiency in this population (Zhang etal., 2012). However, a limitation of this study is that
17    it does not attribute the proportion of apoptotic activity to a specific class of cells and does not
18    include assessment of other potential markers of immunotoxicity in peripheral blood.
19          Results of functional immune assays in laboratory animals following short-term i.p. and s.c.
20    exposures add to the evidence for benzo[a]pyrene immunotoxicity. Resistance to Streptococcus
21    pneumonia or Herpes simplex type 2 was dose dependently reduced in B6C3Fi mice following s.c.
22    injection of >5 mg/kg-day benzo[a]pyrene for 14 days (Munson etal., 1985). Reduced cell
23    proliferation, IFN-y release, and IL-4 release were observed in male and female C56BL/6 mice
24    following short-term exposure to a gavage dose of 13 mg/kg benzo[a]pyrene as measured in a
25    modified local lymph node assay (van den Berg et al.. 2005). A statistically significant decrease in
26    natural killer cell activity was observed in male Wistar rats (Effector:Target cell ratio was
27    40.9 ± 28.4% that of controls) exposed to 90 mg/kg-day by gavage for 35 days (De Jong etal..
28    1999): however, splenic natural killer cell activity was not affected in B6C3Fi mice after s.c.
29    injection of 40mg/kg-day benzo[a]pyrene for 14 days (Munson etal., 1985). The magnitude of the
30    dose and duration of the exposure may account for the discrepancy between these two studies.
31    Single i.p. injections of 50 mg/kg benzo[a]pyrene decreased pro- and/or pre-B-lymphocytes and
32    neutrophils in the bone marrow of C57BL/6J mice without affecting the numbers of immature and
33    mature B-lymphocytes or GR-1+ myeloid cells (Galvan et al.. 2006).
34          In contrast to studies that have shown immunosuppression, benzo[a]pyrene may also
35    induce sensitization responses. Epicutaneous abdominal application of 100 [ig benzo[a]pyrene to
36    C3H/HeN mice, followed by ear challenge with 20 |ig benzo[a]pyrene 5 days later, produced a
37    contact hypersensitivity (a significant ear swelling) response (Klemme etal.. 1987).
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-39            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1   Developmental Immunotoxicity
 2          As noted above, several i.p. injection studies suggest that cell-mediated and humoral
 3   immunity may be altered by exposure to high doses of benzo[a]pyrene during gestation.
 4   Suppression of the mixed lymphocyte response, the graft-versus-host response, and suppression of
 5   the plaque-forming cell response to SRBCs was observed in mice exposed in utero to 150 mg/kg
 6   during mid (CDs 11-13), late (CDs 16-18), or both (CDs 11-17) stages of gestation; these effects
 7   persisted until 18 months of age  [Urso and Gengozian. 1984.1982.1980). Fetal thymic atrophy, as
 8   assessed by reductions in cellularity (74-95%, compared to controls), was observed in mice
 9   exposed to 50-150 mg/kg benzo[a]pyrene from GD 13 to 17, when examined on GD 18 [Holladay
10   and Smith. 1994). Analysis of cell surface markers (e.g., CD4, CDS) from the same study indicate
11   that benzo [a]pyrene may inhibit and/or delay thymocyte maturation, possibly contributing to the
12   observed thymic atrophy (Holladay and Smith. 1994). Consistent with these findings, several other
13   studies have noted decreased thymocyte numbers and disrupted T cell maturation after in utero
14   exposure to benzo[a]pyrene (Rodriguez etal.. 1999: Holladay and Smith. 1995: Lummus and
15   Henningsen. 1995: Urso etal.. 1992: Urso and Tohnson. 1987).
16          The fetal liver is the primary hematopoietic organ during gestation and a major source of
17   thymocyte precursors beginning around GD 10 or 11  in mice (Landreth and Dodson. 2005: Penit
18   andVasseur. 1989). Statistically significant reductions in total cellularity in the fetal liver of 54  and
19   67% were reported in offspring after i.p. exposures of 50 or 100 mg/kg benzo[a]pyrene,
20   respectively, to the dams on CDs 13-17 (Holladay and Smith. 1994). The decreased fetal liver
21   cellularity was accompanied by decreased expression of terminal deoxynucleotidyl transferase  and
22   CD45R cellular markers, which are known to be present in cortical thymocyte progenitors in the
23   fetal liver (Holladay and Smith. 1994: Fine et al.. 1990: Silverstone et al.. 1976).  These data also
24   suggest that benzo[a]pyrene disrupts liver hematopoiesis during gestation and may interfere with
25   prolymphoid seeding of the thymus, possibly contributing to thymic atrophy and cell-mediated
26   immunosuppression. Decreased numbers of CD4+ T-cells have been reported in the spleen of
27   1-week-old mice following in utero benzo[a]pyrene exposure by i.p. injection to the dams,
28   demonstrating the potential for downstream effects on T-cell development (Rodriguez etal.. 1999).
29   The decreased numbers of CD4+ T-cells correspond with observations of decreased proliferation in
30   the presence of Concanavalin A and a weak response  compared to controls in an allogeneic mixed
31   lymphocyte reaction assay (Urso and Kramer. 2008).
32          Postnatal exposure to benzo[a]pyrene has also been suggested to cause immune effects.
33   Dose-dependent decreases in erythrocytes (attributed to reduced bone marrow erythropoiesis), as
34   well as reduced expression of IL-4 and IFN-y were observed in the pups of Wistar rats exposed to
35   0.1-10 mg/kg-day benzo[ajpyrene by subcutaneous injection for 14 days (Matiasovic et al.. 2008).
36   This finding suggests that benzo[a]pyrene may alter the immune response to infection or
37   vaccination in developing animals.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-40            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
1
2
       Table 1-8. Evidence pertaining to the immune effects of benzo[a]pyrene in
       animals
Reference and study design
Results3
Thy m us effects
Kroese et al. (2001)
Wistar rats, 10/sex/dose
0, 3, 10, or 30 mg/kg-d by gavage 5 d/wk
90 d
De Jong etal. (1999)
Wistar rats, 8 males/dose
0, 3, 10, 30, or 90 mg/kg-d by gavage 5 d/wk
35 d
Kroese et al. (2001)
Wistar rats, 10/sex/dose
0, 1.5, 5, 15, or 50 mg/kg-d by gavage 5 d/wk
35 d
4, thymus weight
Females (% change from control): 0, -3, -6, and -28*
Males (% change from control): 0, 0, -13, and -29*
/T" slight thymic atrophy
Females (incidence): 0/10, 0/10, 0/10, and 3/10
Males (incidence): 0/10, 2/10, 1/10, and 6/10*
4> thymus weight
% change from control: 0, -9, -15*, -25*, and -62*
4> thymus weight
Females (% change from control): 0, 13, 8, -3, and -17*
Males (% change from control): 0, -8, -11, -27*, and -33*
Spleen effects
De Jong etal. (1999)
Wistar rats, 8 males/dose
0, 3, 10, 30, or 90 mg/kg-d by gavage 5 d/wk
35 d
•^ relative number (%) of B cells in spleen
% change from control: 0, -8, -13*, -18*, and -41*
4, total number of B cells in spleen
% change from control: 0, 13, -13, -13, and -61*
Change in total cell number in the spleen
% change from control: 0, 20, 0, +7, and -31*
Immunoglobulin alterations
De Jong etal. (1999)
Wistar rats, 8 males/dose
0, 3, 10, 30, or 90 mg/kg-d by gavage 5 d/wk
35 d
4, serum IgM
% change from control: 0, -13, -14, -33*, and -19
4, serum IgA
% change from control: 0, -27, -22, -28, and -61*
3
4
5
*Statistically significantly different from the control (p< 0.05).
a% change from control calculated as: (treated value - control value)/control value x 100.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                 1-41             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene




1 n


"j;
ro
DO
OJ3

0
Q
•1




• LOAEL


ANOAEL
• Doses > LOAEL
0 Doses < NOAEL
1

z









.


Kroese etal., 2001
rm

k ,
A L






5 ,

90-day study; females
Krocscctdl., 2001



A




4




•

> <


Tr



L
m
!•
T







M
,


v
i_l
c





A •
Z









) <




^
A

'

> L







r.

v .
* • • «
T T




1 L

\ ji
T







k




A L









V .
f
T

.


3 z







\ <






, ,

1


>
>
t/i */> t/i
	 CD — i-H — Q —
CD '~ O ^3 O ^
£ ^-* E o E (~g E
"? -^ ~? g ra" 0 ~?
N^- Thymus Weight
Thymus Effects
>/> ^>
S| ° 1
^ >£ f >:
|| ||
^Slight Thymic
Atrophy
cr> ^
en -^
o 3
«, -
° -3
tu LO
Q m
•^ n
relative
number
{%) of B
cells in
spleen
s
cr, ^ cr, ^
CD ^ en -^
CD '"3 CD -~
CJ ^ O zi
on ^ QO ^
1 -S l-o
(LJ LTl CU |_^)
O m O ro
vL-in total Change in
number of total cell
B cells in number in
spleen the spleen
oleen Effects
Cl •"
3 E
c >-
^ ip
Q m
vUSerum
IfiM
ImmunG
Alter,

U1
0 -±
CJ ^
r&
-=., T3
Q ro
IRA
globulin
tions

 2          Figure 1-5. Exposure-response array for immune effects following oral
 3          exposure.

 4   Mode-of-Action Analysis—Immune Effects
 5          Exposure to benzo[a]pyrene induces immunosuppressive effects such as decreased
 6   numbers of B cells in the spleen and decreased thymus weight and cellularity following oral, i.p.,
 7   s.c., or intratracheal exposure in experimental animals. However, the key events underlying
 8   benzo[a]pyrene immunotoxicity have not been identified.
 9          Benzo[a]pyrene is a well-known ligand for the AhR [Okey etal.. 1994: Nebertetal.. 1993:
10   Postlind et al.. 1993).  Ligands of the AhR have been shown to have a role in regulating
11   hematopoietic stem cells in the bone marrow, a major site of B-cell proliferation and antibody
12   production [Esser. 2009). Benzo[a]pyrene reduced B-cell lymphopoiesis at concentrations as low
13   as 10~8M [Hardin et al.. 1992]. Furthermore, Ah-responsive (C57BL/6) mice showed greater dose-
14   dependent reductions in B-cell lymphopoiesis than those observed in Ah-nonresponsive (DBA/2)
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-42            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    mice [Hardinetal., 1992]. Addition of the AhR antagonist and CYP450 inhibitor, a-naphthaflavone,
 2    inhibited the benzo[a]pyrene-induced suppression of B-cell lymphopoiesis in a concentration-
 3    dependent fashion. Similarly, the CYP1A1 inhibitor, l-(l-propynyl) pyrene, blocked
 4    benzo[a]pyrene-induced B-cell growth inhibition but not growth inhibition caused by the
 5    benzo[a]pyrene metabolite, BPDE; these data suggest that a GYP 1A1-dependent metabolite of
 6    benzo[a]pyrene is responsible for the B-cell growth suppressive effects observed after
 7    benzo[a]pyrene exposure [Allan etal.. 2006]. Altogether, these data suggest that benzo[a]pyrene
 8    may regulate B-cell proliferation and antibody production in the bone marrow via the AhR.

 9    Summ ary oflmm un e Effects
10          Evidence for immunotoxic effects of benzo[a]pyrene exposure comes from animal studies
11    that vary in route and duration of exposure. There are no human epidemiological studies that
12    provide specific support for benzo [a]pyrene immunotoxicity; however, immunosuppression has
13    been observed in studies following occupational exposure to PAH mixtures.  However, these
14    findings are limited by co-exposures to other constituents of PAH mixtures.
15          Effects such as altered thymus weight and histology, spleen effects, and altered
16    immunoglobulin levels observed by the oral route reported in animal bioassays provide some
17    evidence of immunotoxicity following benzo[a]pyrene exposure; however, in vivo functional assays
18    provide stronger support for immunotoxicity [WHO.  2012]. The immunological changes observed
19    in the available subchronic gavage studies are supported by a larger database of in vivo studies of
20    benzo[a]pyrene (by parenteral exposure] indicating functional immunosuppression such as
21    decreased proliferative responses to antigens and decreased resistance to pathogens or tumor cells
22    [Kong etal.. 1994: Blanton etal.. 1986: Munson etal.. 1985: White etal.. 1985: Dean etal.. 1983:
23    Munson and White, 1983]. Although the key events underlying the mode of action of
24    benzo[a]pyrene immunotoxicity are not firmly established, there is evidence of physical alterations
25    to tissues/organs of the immune system, as well as decreases in immune function. Evidence of
26    benzo[a]pyrene-associated immunotoxicity is supported by consistent thymic effects observed in
27    two oral studies, as well as splenic effects, and varying immunosuppressive responses observed in
28    short-term or in vitro tests.
29          EPA concluded there was suggestive evidence that immunotoxicity is a potential human
30    hazard of benzo [a]pyrene exposure.

31    Suscep tible Pop illations an d Lifestages
32          The severity and persistence of immune effects observed during in utero studies suggests
33    that immunotoxicity may be greater during gestation than adulthood [Dietertand Piepenbrink.
34    2006: Holladay and Smialowicz. 2000: Urso and Gengozian. 1982]. Urso and Gengozian [1982]
35    provide experimental support demonstrating that immunosuppression from benzo  [a]pyrene
36    exposure during gestation was greater than for mice  exposed after birth to a 25-fold higher dose.
37    There is also substantial literature indicating that disruption of the immune system during certain

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-43            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    critical periods of development (e.g., initiation of hematopoiesis, migration of stem cells, expansion
 2    of progenitor cells) may have significant and lasting impacts on lifetime immune function (e.g.,
 3    Burns-NaasetaL 2008: Dietert. 2008: Landreth. 2002: DietertetaL 20001 In addition, chemical-
 4    specific studies show increased dose sensitivity and disease persistence from developmental versus
 5    adult chemical exposure (reviewed in Luebke et al.. 2006).
 6          Thymus toxicity is a sensitive and specific effect of benzo[a]pyrene and has been observed
 7    in both prenatal and adult exposure studies. The thymus serves as a major site of thymocyte
 8    proliferation and selection for maturation, and impairment can lead to cell-mediated immune
 9    suppression (Kuper etal., 2002: De Waal et al., 1997: Kuper etal., 1992]. The thymus is believed to
10    be critical for T lymphocyte production during early life and not in adulthood (Hakim etal., 2005:
11    Schonland etal.. 2003: Petrie. 2002: Mackall etal.. 1995). Therefore, the decreases in thymus
12    weight observed in studies of adult animals exposed to benzo[a]pyrene suggest that
13    immunosuppression may be a heightened concern for individuals developmentally exposed to
14    benzo[a]pyrene.

15    1.1.4. Other Toxicity
16          There is some evidence thatbenzo[a]pyrene can produce effects in the forestomach, liver,
17    kidney, and cardiovascular system, as well as alter hematological parameters. However, there is
18    less evidence for these effects compared to organ systems described earlier in Sections 1.1.1-1.1.3.
19    Overall, EPA concluded that the available  evidence does not support these noncancer effects as
20    potential human hazards.

21    Forestomach Toxicity
22          Lesions have been observed in the forestomach following subchronic and chronic oral
23    exposure to benzo[a]pyrene (Table 1-9).  Increases in the incidence of forestomach hyperplasia
24    have been observed in Wistar rats following shorter-term, subchronic, and chronic gavage exposure
25    (Kroese etal.. 2001: De long et al.. 1999] and in B6C3Fi mice following chronic dietary exposure
26    (Beland and Gulp. 1998: Gulp etal.. 1998).
27          Following chronic gavage exposure, increased incidences of forestomach hyperplasia were
28    observed in male and female rats at 3 and 10 mg/kg-day; at the highest dose, a lower incidence of
29    hyperplasia was reported (Kroese etal., 2001). However, only the highest-level lesion (hyperplasia,
30    papilloma, or carcinoma) observed in each organ was scored, such that hyperplasia observed in the
31    forestomach, in which tumors were also observed, was not scored. The majority of animals in the
32    high-dose group  exhibited forestomach tumors; therefore, the hyperplasia was not scored and the
33    incidence of forestomach hyperplasia in the study is more uncertain at the highest dose. Shorter-
34    term studies (Kroese etal.. 2001: De Jong etal.. 1999] showed dose-related increases in
35    forestomach hyperplasia at doses >10 mg/kg-day in Wistar rats. In addition, following chronic
36    dietary exposure, a dose-dependent increase in the incidence of forestomach hyperplasia and
37    hyperkeratosis was observed in female mice at >0.7 mg/kg-day (Beland and Gulp. 1998: Gulp etal..

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-44           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    1998]. Forestomach tumors were also observed at >0.7 mg/kg-day by Beland and Gulp [1998] and
 2    Gulp etal.f 19981.
 3          Although humans do not have a forestomach, forestomach effects observed in rodents are
 4    believed to be supportive of a human hazard, as humans have similar squamous epithelial tissue in
 5    their oral cavity [IARC. 2003: Wester and Kroes. 1988). Mechanistic investigations suggest that
 6    bioactivation of benzo[a]pyrene leads to reactive intermediates that can lead to mutagenic events,
 7    as well as to cytotoxic and apoptotic events. The available human, animal, and in vitro evidence
 8    best supports a mutagenic mode of action as the primary mode by which benzo[a]pyrene induces
 9    carcinogenesis. Available data indicate that forestomach hyperplasia may be a histological
10    precursor to neoplasia observed at this site after chronic exposure to benzo[a]pyrene [Kroese etal.,
11    2001: De long et al.. 1999). Dose-response data show that forestomach hyperplasia occurs at
12    shorter durations and at lower doses than tumors in rats and mice exposed to benzo[a]pyrene for
13    up to 2 years [Kroese etal.. 2001: Beland and Gulp. 1998). Kroese etal. [2001] reported that the
14    forestomach lesions demonstrated a progression over the course of intercurrent sacrifices; the
15    authors described early lesions as focal or confluent basal hyperplasia, followed by more advanced
16    hyperplasia with squamous cell papilloma, culminating in squamous cell carcinoma.  The
17    description of the progression of forestomach lesions provided by Kroese etal. [2001], coupled
18    with the observation that hyperplasia occurs before tumors and at lower doses than tumors,
19    suggests that forestomach hyperplasia induced by benzo[a]pyrene is likely a preneoplastic lesion.

20    Hematological Toxicity
21          Altered hematological parameters, including decreases in red blood cell [RBC] count,
22    hemoglobin, and hematocrit have been observed in laboratory animals following benzo[a]pyrene
23    exposure (Table 1-9]. Statistically significant decreases in RBC count, hemoglobin, and hematocrit
24    were observed in male Wistar rats at doses >10 mg/kg-day for 35 days [De long etal.. 1999]. A
25    minimal, but statistically significant increase in mean cell volume and a decrease in mean cell
26    hemoglobin were observed at the highest dose (90 mg/kg-day], which may indicate dose-related
27    toxicity for the RBCs and/or RBC precursors in the bone marrow [De Jong etal.. 1999]. Similarly,
28    male and female F344 rats also showed maximal decreases in RBC counts, hematocrit, and
29    hemoglobin levels between 10 and 12% in a 90-day dietary study [Knuckles etal., 2001].  Findings
30    were significant for RBC counts and hematocrit in males at >50 mg/kg-day, while decreased RBC
31    counts and hematocrit in females and hemoglobin levels in both sexes were only significant in the
32    100 mg/kg-day group [Knuckles et al.. 2001]. Small, but not statistically significant, decreases in
33    RBC counts and hemoglobin were observed in both 35- and 90-day studies in Wistar rats [Kroese et
34    al.. 2001]. It should be noted that when observed, the magnitudes of the decreases in RBCs,
35    hemoglobin, and hematocrit were generally small; about 18% at 90 mg/kg-day and <10% at lower
36    doses [De long etal.. 1999] and about 10% in F344 rats [Knuckles etal.. 2001].  A decrease in white
37    blood cells [WBCs], attributed to reduced numbers of lymphocytes and eosinophils, was also
38    observed at 90 mg/kg-day following gavage exposure for 35 days [De long etal.. 1999]. The mode
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-45            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    of action by which benzo[a]pyrene exposure may lead to altered hematological parameters is
 2    undetermined.

 3    Liver Toxicity
 4           Liver effects other than cancer associated with benzo[a]pyrene exposure primarily include
 5    changes in liver weight and abnormal histopathology (Table 1-9). Increased liver weight was
 6    reported in a 90-day study in both male and female Wistar rats given benzo[a]pyrene by gavage
 7    [Kroese et al.. 2001]. Both females (17% increase) and males (29% increase) demonstrated
 8    statistically significant increased liver weights at the highest dose tested (30 mg/kg-day); a
 9    statistically significant increase (15%) was also reported in males at 10 mg/kg-day. Similar to the
10    findings in the 90-day study by Kroese etal. (2001). increased liver:body weight ratios were
11    observed at the highest dose in a 90-day dietary study in male F344 rats, although there was no
12    change observed in female liver weights (Knuckles etal.. 2001).  Increased liver:body weight ratios
13    were also observed in both sexes at high doses (600 and 1,000 mg/kg) in an accompanying acute
14    study (Knuckles etal.. 2001). A statistically significant increase in liver weight was also observed in
15    male Wistar rats given 90 mg/kg-day benzo[a]pyrene by gavage for 35 days (De long et al., 1999).
16    Consistent with the findings by De long  et al. (1999), a statistically significant increased liver weight
17    (about 18%) was also observed in both  male and female Wistar rats at the highest dose
18    (50 mg/kg-day) given by gavage in a 3 5-day study (Kroese etal.. 2001).
19           Limited exposure-related differences in clinical chemistry parameters associated with liver
20    toxicity were observed; no differences in alanine aminotransferase or serum aspartate
21    transaminase levels were observed, and a small dose-related decrease in y-glutamyl transferase
22    was observed in males only exposed to benzo[a]pyrene for 90 days (Kroese etal., 2001).
23           Treatment-related lesions  in the liver (oval cell hyperplasia) were identified as statistically
24    significantly increased following exposure to 90 mg/kg-day benzo[a]pyrene for 35 days; however,
25    incidence data were not reported (De long etal.. 1999).  A 2-year carcinogenicity study (Kroese et
26    al.. 2001) observed some histopathological changes in the liver; however, organs with tumors were
27    not evaluated.  Since many of the animals in the highest two doses developed liver tumors, the dose
28    responsiveness of the histological changes is unclear.
29           A dose-dependent increase in liver microsomal ethoxyresorufin-o-deethylase (EROD)
30    activity, indicative of CYP1A1 induction, was observed in both sexes at doses >1.5 mg/kg-day in a
31    35-day study (Kroese et al.. 2001). However, at the highest dose tested, with the greatest fold
32    induction in EROD activity,  there was no evidence of associated adverse histopathologic findings.
33    The finding of increased liver weight across multiple studies of varying exposure durations, as well
34    as histopathological changes in the liver provide evidence of the liver as a target of benzo[a]pyrene-
35    induced toxicity. The mode of action by which benzo[a]pyrene induces these effects is unknown.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-46            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    Kidney Toxicity
 2           There is minimal evidence of kidney toxicity following exposure to benzo[a]pyrene
 3    (Table 1-9). Statistically significant decreases in kidney weight were observed at doses of 3, 30, and
 4    90 mg/kg-day, but not at 10 mg/kg-day, in a 35-day gavage study in male Wistar rats [De long etal..
 5    1999).  In a 35-day gavage study with a similar dose range in male and female Wistar rats, no
 6    statistically significant changes in kidney weights were observed at any dose [Kroese et al.. 2001).
 7    Histopathological analysis of kidney lesions revealed an apparent dose-responsive increase in the
 8    incidence of abnormal tubular casts in the kidney in male F344 rats exposed by diet for 90 days
 9    [Knuckles etal., 2001]. The casts were described as molds of distal nephrons lumen and were
10    considered by the study authors to be indicative of renal dysfunction. However, the statistical
11    significance of the kidney lesions is unclear.  Several gaps and inconsistencies in the reporting make
12    interpretation of the kidney effects difficult, including: (1] no reporting of numerical data; (2) no
13    indication of statistical significance in the accompanying figure for kidney lesions; (3) discrepancies
14    between the apparent incidences and sample sizes per dose group; and (4) uncertainty in how
15    statistical analysis of histopathological data was applied.  As such, the significance of the abnormal
16    tubular casts is unclear.

17    Cardiovascular Toxicity
18           Atherosclerotic vascular disease and increased risk of cardiovascular mortality have been
19    associated with cigarette smoking [Ramos and Moorthy. 2005: Miller and Ramos. 2001: Thirman et
20    al.. 1994] and, to a more limited degree, occupational exposure to PAH mixtures [Friesenetal..
21    2010: Friesen et al.. 2009: Burstyn et al.. 2005: Chau et al.. 1993]. Elevated mortality due to
22    cardiovascular  disease was observed in a PAH-exposed occupational population (coke oven plant
23    workers], but elevated cardiovascular mortality was also observed in the non-exposed or slightly
24    exposed populations [Chau etal.. 1993].  Elevated risks of ischemic heart disease (IHD] were
25    associated with past cumulative benzo[a]pyrene exposure among aluminum smelter workers (with
26    a 5-year lag], although the trend was not statistically significant; there was no observed association
27    with more recent benzo[a]pyrene exposure (Friesen et al.. 2010]. Elevated risk of mortality from
28    IHD was also associated with cumulative benzo[a]pyrene exposure in a cohort of male asphalt
29    workers (although not statistically significant]; the trend in average benzo[a]pyrene exposure and
30    association with IHD was statistically significant, with an approximately 60% increase in risk
31    between the lowest and highest exposure groups (Burstyn et al.. 2005]. The two studies that
32    associate benzo[a]pyrene exposure with cardiovascular effects (Friesenetal.. 2010: Burstyn etal..
33    2005] rely on statistical models to create exposure groups rather than direct measurement of the
34    cohort under examination.  Additionally, while these studies used benzo[a]pyrene exposure
35    groupings for analysis, they cannot address co-exposures that may have occurred in the
36    occupational setting (asphalt or aluminum smelters] or exposures that occurred outside the
37    workplace.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-47             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          Increased systolic and diastolic blood pressure has been observed in the offspring of dams
 2    exposed to increasing concentrations ofbenzo[a]pyrene (Tules etal.. 2012] (Table 1-1). At the
 3    highest dose tested (1.2 mg/kg body weight by gavage to the dams), systolic pressures were
 4    elevated approximately 50% and diastolic pressures were elevated approximately 80% above
 5    controls. An intranasal exposure of 0.01 mg/kg-day benzo[a]pyrene in adult male rats also
 6    produced an increase in blood pressure following a 7-day exposure (Centner and Weber. 2011).
 7          Reduced endothelial integrity and increased smooth muscle cell mass, both related to
 8    atherosclerosis, have been observed in Sprague-Dawley rats exposed to 10 mg/kg benzo[a]pyrene
 9    by i.p. injection (once/week for 8 weeks) (Zhang and Ramos, 1997). The molecular mechanisms
10    underlying PAH-induced vascular injury and the development of atherosclerosis are not well
11    established, but current hypotheses include cell proliferative responses to injury of endothelial cells
12    from reactive metabolites (including reactive oxygen species [ROS]) and genomic alterations in
13    smooth muscle cells from reactive metabolites leading to transformed vasculature cells and
14    eventual plaque formation (Ramos and Moorthy. 2005). However, while the link between PAHs
15    and atherosclerotic disease has been studied, experiments specifically looking at the relationship
16    between levels of exposure to benzo[a]pyrene (via environmentally relevant routes) and the
17    development of aortic wall lesions related to atherosclerosis have not generally been performed.
18          One exception to this observation comes from a series of experiments  on Apolipoprotein E
19    knock-out (ApoE-/-) mice exposed orally to benzo[a]pyrene. ApoE-/- mice develop spontaneous
20    atherosclerosis, which is thought to be due to enhanced oxidative stress from the lack of ApoE
21    (Godschalketal.. 2003).  Overall, these studies suggest that benzo[a]pyrene exposure  in ApoE-/-
22    mice enhances the progression of atherosclerosis through a general local inflammatory process.

23    Nervous System Effects
24          Neurobehavioral function and mood state were evaluated in two studies of men
25    occupationally exposed to PAH mixtures (Oiuetal.. 2013: Niu etal.. 2010). Alterations in
26    neurobehavioral function was evaluated in coke oven workers using the Neurobehavioral Core Test
27    Battery (self-reported symptoms by questionnaire). These studies also measured urinary levels of
28    the PAH metabolite, 1-hydroxypyrene, as markers of PAH exposure. In both studies, exposure was
29    associated with decrements in short-term memory and/or attention in digit span tests. In addition,
30    Qiu etal. (2013) reported an association between benzo[a]pyrene exposure and decrements in
31    tests related to sensorimotor coordination (i.e., reaction time, digit symbol, and pursuit aiming
32    tests), as well as lower health ratings for the tension-anxiety mood category; Niu etal. (2010)
33    performed the same test battery but did not detect these associations.
34          Alterations in neuromuscular, autonomic, sensorimotor, and electrophysiological endpoints
35    have been reported in rats and mice following acute or short-term exposure to benzo[a]pyrene
36    (Bouayedetal..2009b: Grova etal.. 2008: Grova etal.. 2007: Saunders etal.. 2006: Liu etal.. 2002:
37    Saunders etal.. 2002: Saunders etal.. 2001). Impaired learning and memory (as measured by
38    Morris water maze performance or novel object recognition) was observed following  subchronic
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-48            DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene
 1    gavage in adult rats [Maciel etal., 2014: Chenetal., 2011: Chengzhietal., 2011] and following
 2    subchronic or short-term i.p. exposure in adult mice (Oiuetal.. 2011: Xiaetal.. 2011: Grovaetal..
 3    2007). Decreased anxiety-like behavior in hole board and elevated plus maze tests has been
 4    observed following short-term i.p. exposure [Grovaetal.. 2008). while decreased depressive-like
 5    activity was observed in the tail suspension test (but not the forced swim test) following short-term
 6    oral exposure [Bouayedetal.. 2012). In addition, a 28-day gavage study in male mice observed an
 7    increase in aggressive behavior  (as measured by the resident intruder test) and an increase in
 8    consummatory sexual behavior  in mice treated with 0.02 mg/kg-day (Bouayed et al.. 2009b).
 9    These data are consistent with the neurobehavioral effects observed following developmental
10    exposure, and they suggest that benzo[a]pyrene exposure could be neurotoxic in adults; however,
11    only limited data are available to inform the neurotoxic potential of repeated subchronic or chronic
12    exposure to benzo[a]pyrene via the oral route (Table 1-9).

13           Table 1-9. Evidence pertaining to other toxicities of benzo[a]pyrene in
14           animals
                Reference and study design
                     Results3
       Forestomach toxicity
       Kroese et al. (2001)
       Wistar (Riv:TOX) rats: male and female (52/sex/dose
       group)
       0, 3,10, or 30 mg/kg-d by gavage 5 d/wk
       104 wks (chronic)

       Wistar (Riv:TOX) rats: male and female
       (10/sex/dose group)
       0, 3,10, or 30 mg/kg-d by gavage 5 d/wk
       90 d (subchronic)

       Wistar (specific pathogen-free Riv:TOX) rats
       (10/sex/dose group)
       0,1.5, 5,15, or 50 mg/kg body weight by gavage
       5 d/wk
       5 wks (shorter-term)
Forestomach hyperplasia (basal cell hyperplasia)
incidences'5:
M: 2/50; 8/52; 8/52; and 0/52
F: 1/52; 8/51; 13/51; and 2/52

Forestomach hyperplasia (slight basal cell hyperplasia)
incidences:
M: 2/10; 0/10; 6/10; and 7/10
F: 0/10; 2/10; 3/10; and 7/10

Forestomach hyperplasia (basal cell hyperplasia)
incidences:
M: 1/10; 1/10; 4/10; 3/10; and 7/10
F: 0/10; 1/10; 1/10; 3/10; and 7/10*
       (Beland and Gulp (1998); Gulp et al. (1998))
       BSCSFi mice: female (48/dose group)
       0, 5, 25, or 100 ppm in the diet (average daily
       doses": 0, 0.7, 3.3, and 16.5 mg/kg-d)
       2 years
Forestomach hyperplasia
Incidences: 13/48; 23/47; 33/46*; and 38/47*

Forestomach hyperkeratosis
Incidences: 13/48, 22/47, 33/46*, 38/47*
       De Jong etal. (1999)
       Wistar rats: male (8/dose group)
       0, 3,10, 30, or 90 mg/kg-d by gavage 5 d/wk
       5 wks
Forestomach hyperplasia (basal cell hyperplasia)
statistically significantly increased incidences at 30 and
90 mg/kg-d were reported, but incidence data were not
provided
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                   1-49             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
          Reference and study design
                      Results3
Hematological toxicity
Kroese et al. (2001)
Wistar rats, 10/sex/dose
0, 3,10, or 30 mg/kg-d by gavage 5 d/wk
90 d

Wistar rats, 10/sex/dose
0,1.5, 5,15, or 50 mg/kg-d by gavage 5 d/wk for
35 d
RBC count and hemoglobin changes not statistically
significant in males or females at any dose (numerical data
not reported)

RBC count: changes not statistically significant (numerical
data not reported)

Hemoglobin: changes not statistically significant
(numerical data not reported)
Knuckles etal. (2001)
F344 rats, 20/sex/dose
0, 5, 50, or 100 mg/kg-d by diet
90 d
4, RBC count
Females (% change from control): statistically significant at
100 mg/kg-d (numerical data not reported)

Males (% change from control): statistically significant at
50 and 100 mg/kg-d (numerical data not reported)

•i, hematocrit
Females (% change from control): statistically significant at
100 mg/kg-d (numerical data not reported)

Males (% change from control): statistically significant at
50 and 100 mg/kg-d (numerical data not reported)

•i, hemoglobin
Females: statistically significant at 100 mg/kg-d (numerical
data not reported)
Males: statistically significant at 100 mg/kg-d (numerical
data not reported)
De Jong etal. (1999)
Wistar rats, 8 males/dose
0, 3,10, 30, or 90 mg/kg-d by gavage 5 d/wk
35 d
4, RBC count
% change from control: 0, -1, -5*, -10*, and -18*

•^ hemoglobin
% change from control: 0, -1, -7*, -10*, and -18*

•i, hematocrit
% change from control: 0,0, -6*, -8*, and -14*

4, WBC count
% change from control: 0, -8, -9, -9, and -43*

1" mean cell volume
% change from control: 0,0, -3, 0, and 3*

•i, mean corpuscular hemoglobin concentration
% change from control: 0, -1, -1, -1, and -3*
           This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-50              DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
Reference and study design
Results3
Liver toxicity
Kroese et al. (2001)
Wistar rats, 10/sex/dose
0, 3, 10, or 30 mg/kg-d by gavage 5 d/wk
90 d
Wistar rats, 10/sex/dose
0, 1.5, 5, 15, or 50 mg/kg-d by gavage 5 d/wk for
35 d
Knuckles etal. (2001)
F344 rats, 20/sex/dose
0, 5, 50, or 100 mg/kg-d by diet
90 d
De Jong etal. (1999)
Wistar rats, 8 males/dose
0, 3, 10, 30, or 90 mg/kg-d by gavage 5 d/wk
35 d
1" liver weight
Females (% change from control): 0, -2, 4, and 17*
Males (% change from control): 0, 7, 15*, and 29*
Liver histopathology: no effects reported
/T" liver weight
Females (% change from control): 0, 3, 2, 9, and 18*
Males (% change from control): 0, 2, 1, 3, and 18*
Liver histopathology: no effects reported
1" liverbody weight ratio
Females: no change (numerical data not reported)
Males (% change from control): 23% change reported at
100 mg/kg-d (numerical data not reported)
1" liver weight
% change from control: 0, -9, 7, 5, and 15*
/T" liver oval cell hyperplasia (numerical data not reported)
reported as significant at 90 mg/kg-d;
Kidney effects
Knuckles etal. (2001)
F344 rats, 20/sex/dose
0, 5, 50, or 100 mg/kg-d by diet
90 d
De Jong etal. (1999)
Wistar rats, 8 males/dose
0, 3, 10, 30, or 90 mg/kg-d by gavage 5 d/wk
35 d
Kroese et al. (2001)
Wistar rats, 10/sex/dose
0, 1.5, 5, 15, or 50 mg/kg-d by gavage 5 d/wk
35 d
1" abnormal tubular casts
Females: not statistically significant (numerical data not
reported)
Males: apparent dose-dependent increase (numerical data
not reported)
•i, kidney weight
% change from control: 0, -11*, -4, -10*, and -18*
Kidney weight: no change (data not reported)
Nervous system effects
Chengzhietal. (2011)
Sprague-Dawley rats, male, 32/dose
0 or 2 mg/kg-d by gavage
90 d
/T" time required for treated rats to locate platform in
water maze (data reported graphically)
This document is a draft for review purposes only and does not constitute Agency policy,
                                1-51            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene
               Reference and study design
                    Results3
       Bouayed et al. (2009b)
       Swiss albino mice, male, 9/group
       0, 0.02, or 0.2 mg/kg-d by gavage
       28 d
Significant decrease in latency to attack and increase in the
number of attacks in the resident-intruder test at
0.02 mg/kg-d (but not at high dose)

Significant increase in mount number in the copulatory
behavior test at 0.02 and 0.2 mg/kg-d
 1
 2    *Statistically significantly different from the control (p < 0.05).
 3    a% change from control calculated as: (treated value - control value)/control value x 100.
 4    bReported incidences may not fully account for the occurrence of hyperplasias due to the scoring of only the
 5     highest-level lesion in an individual animal (e.g., animals with forestomach tumors that also showed hyperplasia
 6     would not have the observation of hyperplasia recorded).
 7    °Based on the assumption that daily benzo[a]pyrene intake at 5 ppm was one-fifth of the 25-ppm intake (about
 8     21 u.g/d) and using TWA body weights of 0.032 kg for the control, 5- and 25-ppm groups and 0.026 kg for the
 9     100-ppm group.

10    1.1.5.  Carcinogenicity

11    Eviden ce in Hum ans
12           Numerous epidemiologic studies indicate an association between PAH related occupations
13    and lung, bladder, and skin cancer (Table 1-10). This discussion primarily focuses on epidemiologic
14    studies that included a direct measure of benzo[a]pyrene exposure. All identified studies have co-
15    exposures to other PAHs.  The identified studies were separated into tiers according to the extent
16    and quality of the exposure analysis and other study design features:

17           Tier 1: Detailed exposure assessment conducted (using a benzo(a)pyrene metric), large
18           sample size (>~50 exposed cases), and adequate follow-up period to account for expected
19           latency (e.g., >20 years for lung cancer).

20           Tier 2: Exposure assessment, sample size, or follow-up period did not meet the criteria for
21           Tier 1, or only a  single-estimate exposure analysis was conducted.
22           For lung cancer, each of the Tier 1 studies observed increasing risks of lung cancer with
23    increasing cumulative exposure to benzo[a]pyrene (measured in |ig/m3-years), and each of these
24    studies addressed in the analysis the potential for confounding by smoking (Armstrong and  Gibbs.
25    2009: Spinelli etal.. 2006: Xu etal.. 19961 (Table 1-11). These three studies represent different
26    geographic locations and two different industries. The pattern of results in the Tier 2 studies was
27    mixed, as would be expected for studies with less precise exposure assessments or smaller sample
28    sizes: one of the  standardized mortality ratio (SMR) estimates was <1.0, with the other eight
29    estimates ranging from  1.2 to 2.9 (Table 1-12). In considering all of the available studies,
30    particularly those with the strongest methodology, there is considerable support for an association
31    between benzo[a]pyrene exposure and lung cancer, although the relative contributions of
32    benzo[a]pyrene  and of other PAHs cannot be established.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-52             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          For bladder cancer, the cohort and nested case-control studies observed a much smaller
 2    number of cases compared with lung cancer; this limits their ability to examine exposure-response
 3    relationships.  Three cohort studies with detailed exposure data, however, identified 48-90 cases
 4    fBurstynetal..20Q7: Gibbs and Sevigny. 2007a: GibbsetaL 2007: Gibbs and Sevigny. 2007bl
 5    [Spinellietal.. 2006] (Tier 1 studies, Table 1-13). Although cumulative exposure (up to
 6    approximately 2 ug/m3-years] was not related to increasing risk in the study of asphalt workers by
 7    Burstyn et al. (2007). an exposure-response was seen with the wider exposure range (i.e.,
 8    >80 ug/m3-years] examined in two studies of aluminum smelter workers by (Gibbs and Sevigny
 9    (2007aj: Gibbs etal. (2007]: Gibbs and Sevigny (2007b]]: and (Spinellietal.. 2006]. This difference
10    in response is not surprising, given that the highest exposure group in the asphalt worker studies
11    corresponded to the exposures seen in the lowest exposure categories in the studies of aluminum
12    smelter workers.  The five studies with more limited exposure information or analyses each
13    included between 2 and 16 bladder cancer cases, with relative risk (RR] estimates ranging from
14    0.6 to 2.9. None of these individual effect estimates was statistically significant (Tier 2  studies,
15    Table 1-13].
16          Two of the identified occupational studies contained information on risk of mortality from
17    melanoma. Neither of these studies observed increased risks of this type of cancer, with an SMR of
18    0.91 (95% confidence interval [CI] 0.26, 2.48] (22 cases] in Spinelli et al. f 20061 and 0.58 (95% CI
19    0.12,1.7] in Gibbs etal. (2007](3 cases]. These studies did not include information on  non-
20    melanoma skin cancers.
21          Non-melanoma skin cancer, specifically squamous cell carcinoma, is of particular interest
22    with respect to dermal PAH exposures. The literature pertaining to this kind of cancer  and PAH
23    exposure goes back to the 18th century work of Sir Percivall Pott describing scrotal cancer, a
24    squamous cell skin cancer, in English chimney sweeps (Brown and Thornton,  1957]. Recent studies
25    of chimney sweeps in several Nordic countries have not found increases in non-melanoma skin
26    cancer incidence (Hogstedtetal.. 2013: Pukkalaetal.. 2009: Evanoff etal.. 1993]. likely due to
27    greatly reduced exposure associated with better occupational hygiene (IARC. 2012]. A study
28    among asphalt workers (roofers] reported an increased risk of mortality from non-melanoma skin
29    cancer among asphalt workers (roofers], with an SMR of 4.0 (95% CI: 1.0,10.9] among workers
30    employed >20 years (Hammond etal., 1976]. In addition to this study, two studies in Scandinavian
31    countries examined non-melanoma skin cancer risk in relation to occupations with likely dermal
32    exposure to creosote (i.e., timber workers and brick makers] using incidence data from population
33    registries (Karlehagenetal.. 1992: Tornqvistetal.. 1986]. The standardized incidence  ratio (SIR]
34    estimates were 1.5 (95%  CI: 0.7, 2.6] based on five exposed cases and 2.37 (95% CI: 1.08, 4.50]
35    based on nine cases in Tornqvistetal. (1986] and Karlehagen et al. (1992]. Because non-melanoma
36    skin cancers are rarely fatal if caught early, and the preventative excision of precancerous lesions is
37    common, the available occupational studies and cancer registries likely underestimate  the risk of
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-53            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    squamous cell carcinoma (Car0e etal.. 2013: Voelter-Mahlknechtetal.. 2007: ONS. 2003: Letzel and
 2    Drexler. 19981
 3           In addition to cohorts of workers occupationally exposed to PAH mixtures, populations
 4    exposed to benzo[a]pyrene through topical coal tar formulations for the treatment of psoriasis,
 5    eczema, and dermatitis have also been studied [Roelofzenetal.. 2010: Mitropoulos and Norman.
 6    2005: Stern et al.. 1998: Stern and Laird. 1994: Lindelof and Sigurgeirsson. 1993: Torinuki and
 7    Tagami. 1988: Pittelkowetal.. 1981: MaughanetaL 1980: Stern etal.. 1980). Epidemiological
 8    studies examining skin cancer risk in relation to various types of topical tar exposure are
 9    summarized in the Supplemental Information, Table D-6.  Case reports, reviews, and studies that
10    did not include a measure of coal tar use are not included. The available studies examining
11    therapeutic topical coal tar use and risk of skin cancer were limited by low quality exposure data
12    with high potential of exposure misclassification (e.g., Roelofzenetal.. 2010: Mitropoulos and
13    Norman. 2005: Lindelof and  Sigurgeirsson.  1993). small size and short duration of follow-up up
14    [e.g.. Torinuki and Tagami. 1988]. and choice of referent rates and differences in disease
15    ascertainment between cases and the reference population (e.g., Pittelkowetal.. 1981: Maughan et
16    al., 1980]. In addition, clinic-based studies focused on the regimen of coal tar in conjunction with
17    ultraviolet-B  (UVB] therapy,  although some appear to indicate increased risk with coal tar
18    exposure, cannot distinguish effects of coal  tar from the effects of UVB (e.g., Stern etal.. 1998: Stern
19    and Laird. 1994: Lindelof and Sigurgeirsson. 1993: Stern etal.. 1980].  Therefore, because of the
20    limitations with respect to study design and analysis, EPA did not consider these studies further in
21    the evaluation of the risk of skin cancer from exposure to benzo[a]pyrene.  Although EPA does not
22    consider the available studies sufficient to evaluate the risk of skin cancer, acute studies of coal tar
23    treated patients provide in vivo evidence of benzo[a]pyrene-specific genotoxicity (increased BPDE-
24    DNA adducts] in human skin (Godschalk etal.. 2001: Rojas etal.. 2001: Zhang etal.. 1990], an early
25    key  eventin the carcinogenic mode of action of benzo[a]pyrene (see Figure 1-6 of Section 1.1.5).
26           Lung, bladder, and skin cancers are  the cancers that have been observed in occupational
27    studies of PAH mixtures (Benbrahim-Tallaa etal.. 2012: Baan etal.. 2009: Secretan etal.. 2009).
28    The reproducibility of lung, bladder,  and skin cancers in different populations and exposure
29    settings after occupational exposure to PAH mixtures (see Table 1-10) adds plausibility to the
30    hypothesis that common etiologic factors may be operating. The potential role that benzo[a]pyrene
31    may play as a causal agent is further  supported by the observation that these same sites are also
32    increased in the studies that included a direct measure of benzo[a]pyrene.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-54            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene
            Table 1-10. Cancer sites for PAH-related agents reviewed by IARC
PAH-related mixture or
occupation
Aluminum production
Carbon electrode manufacture
Coal gasification
Coal tar distillation
Coal tar pitch (paving and roofing)
Coke production
Creosotes
Diesel exhaust
Indoor emissions from household
combustion of biomassfuel (primarily
wood)
Indoor emissions from household
combustion of coal
Mineral oils, untreated or mildly
treated
Shale oils
Soot (chimney sweeping)
Sites with sufficient
evidence in humans
Lung, urinary bladder

Lung
Skin
Lung
Lung

Lung

Lung
Skin
Skin
Lung, skin
Sites with limited
evidence in humans

Lung


Urinary bladder

Skin
Urinary bladder
Lung



Urinary bladder
Reference
Baanetal. (2009)

IARC (2010)

Baanetal. (2009)

Baanetal. (2009)

Baanetal. (2009)

Baanetal. (2009)

IARC (2010)

Benbrahim-Tallaa et al.
(2012)
Secretan et al. (2009)

Secretan et al. (2009)

Baanetal. (2009)

Baanetal. (2009)

Baanetal. (2009)
2
3

4
5
Source: Adapted from IARC (2010).

       Table 1-11. Summary of epidemiologic studies of benzo[a]pyrene (direct
       measures) in relation to lung cancer risk: Tier 1 studies
              Reference and study design
                                                                   Results
     Armstrong and Gibbs (2009) (Quebec, Canada)
      Cohort, aluminum smelter workers, seven plants
      16,431 (15,703 men; 728 women); duration
      minimum 1 yr, began work 1966-1990; follow-up
      through 1999 (mean ~30 yrs); smoking information
      collected from medical records

      Exposure:  Job exposure matrix ~5,000 personal
      benzo[a]pyrene measures from the 1970s to 1999

      Related references: Lavoue et al. (2007) (exposure
      data); (Gibbs and Sevigny (2007a); Gibbs etal.
      (2007); Gibbs and Sevigny (2007b); Armstrong et al.
                                             SMR 1.32 (1.22, 1.42) [677 cases]

                                             Lung cancer risk by cumulative benzo[a]pyrene exposure

                                                Median
                                               benzo[a]-
                                                pyrene      n
                                               u.g/m3-yrs   cases   SMR(95%CI)      RR (95% Cl)

                                                  0        35   0.62(0.44,0.87)     1.0 (referent)

                                                  10       266   1.09 (0.96, 1.23)   1.75 (1.23,2.48)

                                                  30       70   1.88(1.47,2.38)   3.02(2.01,4.52)

                                                  60       53   1.21 (0.91, 1.59)   1.94 (1.27,2.97)

                                                 120       114   1.93(1.59,2.32)   3.09(2.12,4.51)
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-55             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene
(1994))
Spinelli et al. (2006) (British Columbia, Canada)
Cohort, aluminum smelter workers; 6,423 (all men);
duration minimum >3 yrs; began work 1954-1997;
follow-up through 1999 (14% loss to follow-up;
mean ~24 yrs); smoking information from self-
administered questionnaire
Exposure: Job exposure matrix using 1,275 personal
benzo[a]pyrene measures from 1977 to 2000 (69%
for compliance monitoring)
Related references: Friesen et al. (2006) (exposure
data); Spinelli etal. (1991)
Xuetal. (1996) (China)
Nested case-control in iron-steel worker cohort
610 incident cases (96% participation); 959 controls
(94% participation) (all men); duration data not
reported; smoking information collected from
interviews; next-of-kin interviews with 30% of lung
cancer cases and 5% of controls
Exposure: Job exposure matrix 82,867 historical
monitoring records, 1956-1992
240 116 1.79(1.48,2.15) 2.86(1.96,4.18)
480 23 2.36(1.49,3.54) 3.77(2.23,6.38)
No evidence of confounding by smoking
Additional modeling as continuous variable: RR 1.35 (95%
Cl 1.22, 1.51) at 100 u.g/m3-yrs (0.0035 per u.g/m3-yrs
increase); other shapes of exposure-response curve
examined.
SMR: 1.07 (0.89, 1.28) [120 cases]
SIR: 1.10 (0.93, 1.30) [147 cases]
Lung cancer risk by cumulative benzo[a]pyrene exposure
Benzo[a]pyrene
u.g/m3-yrs n cases RR (95% Cl)a
0-0.5 25 1.0 (referent)
0.5-20 42 1.23 (0.74,2.03)
20-40 23 1.35 (0.76,2.40)
40-80 25 1.36 (0.78,2.39)
>80 32 1.79 (1.04,3.01)
Adjusting for smoking category; trend p < 0.001.
Lung cancer risk by cumulative benzo[a]pyrene exposure
Benzo[a]-pyrene
(u.g/m3-yrs) n cases RR (95% Cl)a
<0.84 72 1.1(0.8,1.7)
0.85-1.96 117 1.6(1.2,2.3)
1.97-3.2 96 1.6(1.1,2.3)
>3.2b 105 1.8(1.2,2.5)
Adjusting for birth year and smoking category; trend
p < 0.004. Referent group is "nonexposed" (employed in
administrative or low-exposure occupations).
bStudy table IV unclear; could be >3.0 for this category.
1
2
Table 1-12. Summary of epidemiologic studies of benzo[a]pyrene (direct
measures) in relation to lung cancer risk: Tier 2 studies
              Reference and study design
                                                           Results
     Limited follow-up period (<20 yrs)
     Friesen et al. (2009) (Australia)
     Cohort, aluminum smelter workers; 4,316 (all men);
     duration minimum 90 d; began work after 1962;
     follow-up through 2002, mean 16 yrs (maximum
     20 yrs); Smoking information from company records
                                      RR 1.2 (0.7, 2.3) [19 cases in exposed; 20 in unexposed]

                                      Lung cancer risk by cumulative benzo[a]pyrene exposure
                                        Benzo[a]-pyrene
                                           u.g/m3-yrs           n cases        RR (95% Cl)a
                                              0                20         1.0 (referent)
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-56            DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
Reference and study design
if employed before 1995 and study interviews if
employed after 1994
Exposure: Job/task exposure matrix using TWA
benzo[a]pyrene measures (n = 655), 1977-2004 (79%
from 1990 to 2004)
Results
>0-0.41 6
0.41-10.9 6
>10.9 7
aPoisson regression, adjusting for smoking;
0.7 (0.3, 1.8)
1.4(0.6,3.5)
1.7(0.7,4.2)
trend p = 0.22.
Proxy measure
Olsson et al. (2010) (Denmark, Norway, Finland,
Israel)

Nested case-control, asphalt workers; 433 lung
cancer cases (65% participation); 1,253 controls (58%
participation), matched by year of birth, country (all
men); duration: minimum >2 seasons, median
8 seasons; began work 1913-1999; follow-up: from
1980 to 2002-2005 (varied by country); smoking
information from interviews

Exposure: Compilation of coal tar exposure
measures, production characteristics, and repeat
measures in asphalt industry in each country used to
develop exposure matrix
Related references: (Boffetta et al. (2003); Burstyn
etal. (2000))

Costantino et al. (1995) (United States, Pennsylvania)
Cohort, coke oven workers; 5,321 and 10,497
unexposed controls (non-oven steel workers;
matched by age, race, date of first employment) (all
men); duration data not reported; worked in 1953;
follow-up through 1982 (length data not reported)
Exposure: Average daily exposure coal tar pitch
volatiles: 3.15 mg/m3 top-side full-time jobs,
0.88 mg/m3 side jobs; used to calculate weighted
cumulative exposure index

Related reference: Dong et al. (1988) (exposure data)




Lung cancer risk by cumulative coal tar exposure3

Coal tar n
unit-yrsa cases RR
0.39-4.29 43 1.31
4.30-9.42 32 0.98
9.43-16.88 30 0.97
16.89-196.48 54 1.60
(trend p-value)


(95% Cl)
(0.87, 2.0)
(0.62, 1.6)
(0.61, 1.6)
(1.09, 2.4)
(0.07)
Adjusting for set, age, country, tobacco pack-years.





SMR 1.95 (1.59, 2.33) [255 cases]
Lung cancer risk by cumulative exposure
Coal tar pitch
volatiles n
(mg/m3-mo) cases
0 203
1-49 34
50-199 43

200-349 59
350-499 39

500-649 27
>650 56
Adjusting for age, race, coke plant, period
trend p< 0.001.









RR (95% Cl)a
1.0 (referent)
1.2 (0.85, 1.8)
1.6(1.1,2.3)

2.0(1.5,2.8)
2.0(1.6,3.2)

2.7(2.0,4.6)
3.1(2.4,4.6)
of follow-up;

This document is a draft for review purposes only and does not constitute Agency policy,
                                1-57            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
          Reference and study design
                       Results
Limited exposure information
 Liuetal. (1997)(China)

Cohort, various carbon plants and aluminum smelter
workers; 6,635 (all men); duration minimum 15 yrs;
began work before 1971; follow-up: through 1985
(mean ~14 yrs); smoking information from
questionnaire

Exposure: Area samples from one carbon plant,
1986-1987
SMR 2.2 (1.1, 2.8) [50 cases]

Lung cancer risk by exposure category
  Exposure
  category
    None
    Low
  Moderate
    High
  Mean
benzo[a]-
 pyrene
  u.g/m3
                                                                 0.30
                                                                 1.19
n cases
  13
  6
  5
  26
SMR(95%CI)a

1.49 (0.83, 2.5)

1.19(0.48,2.5)

1.52 (0.55, 3.4)

4.30(2.9,6.2)
                                               'Calculated by EPA from data in paper.
Bergerand Manz (1992) (Germany)
SMR 2.88 (2.28, 3.59) [78 cases]
Cohort, coke oven workers; 789 (all men); duration
minimum 10 yrs (mean 27 yrs); began work
1900-1989; follow-up through 1989 (length data not
reported); smoking information from plant records
and interviews

Exposure: Mean benzo[a]pyrene: 28 u.g/m3 (range
0.9-89 u.g/m3)
(Hansen (1991); Hansen, 1989) (Denmark)

Cohort, asphalt workers; 679 workers (applicators)
(all men); duration data not reported; employed
1959 to 1980; follow-up to 1986 (mean ~11 yrs);
smoking information from 1982 surveys of industry
and general population

Exposure: Asphalt fume condensate, 35 personal
samples during flooring: median 19.7 mg/m3 (range
0.5-260 mg/m3)
SMR 2.90 (1.88, 4.3) [25 cases] (ages 40-89)
SMR 2.46 (1.59, 3.6) [25 cases] (with smoking adjustment)
           This document is a draft for review purposes only and does not constitute Agency policy,
                                               1-58              DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
          Reference and study design
                       Results
Gustavsson etal. (1990) (Sweden)
Cohort, gas production (coke oven) workers; 295 (all
men); duration minimum 1 yr, median 15 yrs;
employed 1965-1972; follow-up: 1966-1986
(mortality); 1966-1983 (incidence; mean ~15yrs);
smoking information from interviews with older
workers

Exposure: Area sampling - top of ovens;
benzo[a]pyrene, 1,964 mean 4.3 u.g/m3 (range
0.007-33 u.g/m3); 1,965 mean 0.52 u.g/m3
(0.021-1.29 u.g/m3)
SMR 0.82 (0.22, 2.1) [4 cases] (referent group = employed
men)
SIR 1.35 (0.36, 3.5) [4 cases]
Moulin etal. (1989) (France)
Cohort and nested case-control, two carbon
electrode plants; 1,302 in Plant A (all men),
employed in 1975; follow-up 1975-1985 (incidence);
smoking information from plant records; 1,115 in
Plant B (all men); employed in 1957; follow-up
1957-1984 (mortality); duration of employment and
follow-up data not reported
Plant A: SMR 0.79 (0.32,1.6) [7 cases]
Plant B: SMR 1.18 (0.63, 2.0) [13 cases]

Internal comparison (case-control), >1 yr duration:
Plant A: OR 3.42 (0.35, 33.7) [7 cases, 21 controls]
Plant B: OR 0.49 (0.12, 2.0) [13 cases, 33 controls]
Exposure: Benzo[a]pyrene, 19 area samples and
16 personal samples in Plant A (personal sample
mean 2.7 u.g/m3; range 0.59-6.2 u.g/m3); 10 area
samples and 7 personal samples in Plant B; personal
sample mean 0.17 u.g/m3, range 0.02-0.57 u.g/m3
Hammond etal. (1976) (United States)
SMR 1.6 (1.3,1.9) [99 cases] (>20 yrs since joining union)
(CIs calculated by EPA from data in paper)
Cohort, asphalt roofers; 5,939 (all men); duration
minimum 9 yrs, began before 1960; follow-up
through 1971

Exposure: 52 personal samples (masks with filters)
during specific jobs and tasks; mean benzo[a]pyrene
16.7 u.g per 7-hr d
           This document is a draft for review purposes only and does not constitute Agency policy,
                                                1-59              DRAFT—DO NOT CITE OR QUOTE

-------
                                                                Toxicological Review ofBenzo[a]pyrene
1
2
Table 1-13. Summary of epidemiologic studies of benzo[a]pyrene (direct
measures) in relation to bladder cancer risk
              Reference and study design
                                                              Results
     Tier 1 studies
     Burstyn et al. (2007) (Denmark, Norway, Finland,
     Israel)
     Cohort, asphalt workers; 7,298 (all men); duration
     minimum >2 seasons, median 8 seasons; began
     work 1913-1999; follow-up began around 1960,
     ended around 2000 (years varied by country);
     median 21 yrs; smoking information not collected
     Exposure: Compilation of benzo[a]pyrene
     measures, production characteristics, and repeat
     measures in asphalt industry in each country used
     to develop exposure matrix
     Related references: (Boffetta etal., 2003; Burstyn
     etal. (2000))
                                       48 incident bladder cancer cases (39 cases in analyses with
                                       15-yr lag)
                                       Bladder cancer risk by cumulative benzo[a]pyrene exposure3
                                         Benzo[a]-
                                          pyrene
                                        u.g/m3-yrsa
                                         0-0.253
                                        0.253-0.895
                                        0.895-1.665
                                          >1.665
            RR (95% Cl)
n cases
  12
  12
  12
  12
                 RR (95% Cl)
                 (15-yr lag)0
   (no lag)0
 1.0 (referent)    1.0 (referent)
0.69 (0.29,1.6)   1.1 (0.44, 2.9)
1.21(0.45,3.3)   1.7(0.62,4.5)
0.84(0.24,2.9)   1.1(0.30,4.0)
                                       Adjusting for age, calendar period, total duration of
                                        employment, country.
                                       bTrend p = 0.9.
                                       'Trend p = 0.63.
                                       Stronger pattern seen with average exposure in 15-yr lag
                                       (RR 1.5, 2.7,1.9 in second through fourth quartile; trend
                                       p = 0.15)
     (Gibbs and Sevigny (2007a); Gibbs et al. (2007);
     Gibbs and Sevigny (2007b)) (Quebec, Canada)
                                       Hired before 1950: SMR 2.24 (1.77, 2.79) [78 cases]
                                       Bladder cancer risk by cumulative benzo[a]pyrene exposure
     Cohort, aluminum smelter workers, seven plants
     16,431 (15,703 men; 728 women); duration
     minimum 1 yr, began work 1966-1990; follow-up:
     through 1999 (mean ~30 yrs); smoking information
     collected from medical records
     Exposure:  Job exposure matrix using
     ~5,000 personal benzo[a]pyrene measures from the
     1970s to 1999
     Related  references: Lavoue et al. (2007) (exposure
     data); (Armstrong etal. (1994); Gibbs (1985); Gibbs
     and Horowitz (1979))
                                         Benzo[a]-
                                          pyrene
                                        u.g/m3-yrsa
                                            0
                                            10
                                            30
                                            60
                                            120
                                            240
                                            480
n cases
   3
  14
   3
   1
  15
  30
  12
                 Smoking-
                 adjusted
 SMR(95%CI)        RRb
0.73(0.15,2.1)   1.0 (referent)
0.93 (0.45, 1.4)
1.37 (0.28, 4.0)
0.35 (0.9,1.9)
 4.2 (2.4, 6.9)
 6.4 (4.3, 9.2)
  23.9 (12.2,
    41.7)
                   1.11
                   1.97
                   0.49
                   8.49
                                                    aCategory midpoint.
                                                    bCls not reported; highest category is >80 |jg/m3-yrs
                                                     (n observed = 57).
                                                    Mortality risk reduced in cohort hired in 1950-1959,
                                                    SMR =1.23.
                                                    Similar patterns seen in analysis of bladder cancer
                                                    incidence.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     1-60              DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
Reference and study design
Spinelli et al. (2006) (British Columbia, Canada)
See Table 1-11 for study details; this study is
considered a 'Tier 2") study for bladder cancer
because of the smaller number of bladder cancer
cases (n = 12) compared with lung cancer cases
(n = 120)
Tier 2 studies
Friesen et al. (2009) (Australia)
See Table 1-12 for study details
Costantino et al. (1995) (United States,
Pennsylvania)
See Table 1-12 for study details
Hammond et al. (1976) (United States)
See Table 1-12 for study details
Moulin etal. (1989) (France)
See Table 1-12 for study details
Gustavsson et al. (1990) (Sweden)
See Table 1-12 for study details
Results
SMR 1.39 (0.72, 2.43) [12 cases]
SIR 1.80; Cl 1.45-2.21 [90 cases, including in situ]
Bladder cancer risk by cumulative benzo[a]pyrene exposure
Benzo[a]-
pyrene
u.g/m3-years n cases RR (95% Cl)a
0-0.5 17 1.0 (referent)
0.5-20 20 0.83 (0.43, 1.59)
20-40 13 1.16 (0.56,2.39)
40-80 18 1.50 (0.77,2.94)
>80 22 1.92 (1.02,3.65)
Adjusting for smoking category; trend p < 0.001.

RR 0.6 (0.2, 2.0) [five cases in exposed; eight in unexposed]
Bladder cancer risk by cumulative benzo[a]pyrene exposure
Benzo[a]-
pyrene
u.g/m3-yrs n cases RR (95% Cl)a
0 8 1.0 (referent)
>0-0.41 1 0.2 (0.03,1.9)
0.41-10.9 2 0.7 (0.2,3.7)
>10.9 2 1.2 (0.2,5.6)
aPoisson regression, adjusting for smoking category; trend
p = 0.22.
SMR 1.14 (0.61, 2.12) (16 cases)

SMR 1.7 (0.94, 2.8) (13 cases) (>20 yrs since joining union)
(CIs calculated by EPA from data in paper)
Plant A: 0 observed cases; expected <1.0
Plant B: SMR 1.94 (0.40, 5.0) (3 cases)
SMR 2.85 (0.30, 10.3) (2 cases) (referent group - employed
men)
This document is a draft for review purposes only and does not constitute Agency policy,
                                1-61            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
     Evidence in Animals
 2    Oral exposure
 3          Evidence of tumorigenicity following oral exposure to benzo[a]pyrene has been
 4    demonstrated in rats and mice.  As summarized in Table 1-14, oral exposure to benzo[a]pyrene has
 5    resulted in an increased incidence of tumors in the alimentary tract in male and female rats [Kroese
 6    etal..2Q01: Brune etal.. 1981] and female mice [Beland and Gulp. 1998: Gulp etal.. 1998]. liver
 7    carcinomas in male and female rats,  kidney adenomas in male rats [Kroese etal.. 2001]. and
 8    auditory canal tumors in both sexes  [Kroese etal., 2001].
 9          Forestomach tumors have been observed in several lifetime cancer bioassays in rats and
10    mice following both gavage and dietary exposure to benzo[a]pyrene at doses ranging from
11    0.016 mg/kg-day in Sprague-Dawley rats to 3.3 and 10 mg/kg-day in B6C3Fi mice and Wistar rats,
12    respectively fKroese etal.. 2001: Beland and Gulp. 1998: Gulp etal.. 1998: Brune etal.. 19811  In
13    addition, multiple less-than-lifetime  oral exposure cancer bioassays in mice provide supporting
14    evidence that oral exposure to benzo[a]pyrene is associated with an increased incidence of
15    forestomach tumors [Weyandetal.,  1995: Benjamin etal., 1988: Robinson etal., 1987: El-Bayoumy,
16    1985: Triolo etal.. 1977:  Wattenberg. 1974: Roe etal.. 1970: Biancifiori etal.. 1967: Chouroulinkov
17    etal.. 1967: Fedorenko and Yansheva. 1967: Neal and Rigdon. 1967: Berenblum andHaran. 1955].
18    Although humans do not have a forestomach, similar squamous epithelial tissue is present in the
19    oral cavity [IARC. 2003: Wester and  Kroes. 1988]: therefore, EPA concluded that forestomach
20    tumors observed in rodents following benzo[a]pyrene exposure are relevant in the assessment of
21    carcinogenicity [Beland and Gulp, 1998].  For further discussion, see Sections 1.2 and 2.3.4.
22          Elsewhere in the  alimentary  tract, dose-related increases of benign and malignant tumors
23    were observed. In rats, oral cavity tumors were induced in both sexes and adenocarcinomas of the
24    jejunum were induced in males  [Kroese etal.. 2001].  In mice, tumors were induced in the tongue,
25    esophagus, and larynx of females (males were not tested] [Beland and Gulp. 1998: Gulp etal.. 1998].
26          Chronic oral exposure to benzo[a]pyrene resulted in a dose-dependent increased incidence
27    of liver carcinomas in both sexes of Wistar rats, with the first liver tumors detected in week 35 in
28    high-dose male rats; liver tumors were described as complex, with a considerable proportion
29    (59/150 tumors] metastasizingto the lungs (Kroese etal., 2001]. Treatment-related hepatocellular
30    tumors were not observed in mice (Beland and Gulp. 1998: Gulp etal.. 1998].
31          A statistically significantly increased incidence of kidney tumors (cortical adenomas] was
32    observed in male Wistar  rats following chronic gavage exposure (Kroese etal.. 2001] (Table 1-14].
33    The kidney tumors were  observed at the mid- and high-dose groups. Treatment-related kidney
34    tumors were not observed in two other chronic studies (Beland and Gulp. 1998: Brune etal.. 1981].
35          Lung tumors were also observed following almost nine months of dietary exposure to
36    approximately 10 mg/kg-day in female AJ mice (Weyand et al., 1995]. Other lifetime exposure
37    studies did not report treatment-related increases in lung tumors (Kroese etal.. 2001: Beland and
38    Gulp. 1998: Gulp etal.. 19981

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-62            DRAFT—DO NOT CITE  OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
       Table 1-14. Tumors observed in chronic oral animal bioassays
    Study design and reference
                           Results
Kroese et al. (2001)
Wistar (Riv:TOX) rats (52/sex/dose
group)
0, 3,10, or 30 mg/kg-d by gavage 5 d/wk
2 yrs
Forestomach
        incidences:
        M: 0/52; 7/52*; 18/52*; and 17/52* (papilloma)
        M: 0/52; 1/52; 25/52*; and 35/52* (squamous cell carcinoma)
        F: 1/52; 3/51; 20/51*; and 25/52* (papilloma)
        F: 0/52; 3/51; 10/51*; and 25/52* (squamous cell carcinoma)

Oral cavity
        incidences:
        M: 0/24; 0/24; 2/37; and 10/38* (papilloma)
        M: 1/24; 0/24; 5/37; and 11/38* (squamous cell carcinoma)
        F: 0/19; 0/21; 0/9; and 9/31*(papilloma)
        F: 1/19; 0/21; 0/9; and 9/31* (squamous cell carcinoma)

Jejunum (adenocarcinomas)
        incidences:
        M: 0/51; 0/50; 1/51; and 8/49*
        F: 0/50; 0/48; 0/50; and 2/51

Duodenum (adenocarcinomas)
        incidences:
        M: 0/51; 0/50; 0/51; and 1/49
        F: 0/49; 0/48; 0/50; and 2/51

Liver (adenomas and carcinomas)
        incidences:
        M: 0/52; 3/52; 15/52*; and 4/52 (adenoma)
        M: 0/52; 1/52; 23/52*; and 45/52* (carcinoma)
        F: 0/52; 2/52; 7/52*; and 1/52  (adenoma)
        F: 0/52; 0/52; 32/52*; and 50/52* (carcinoma)

Kidney (cortical adenoma)
        incidences:
        M: 0/52; 0/52; 7/52*;  and 8/52*
        F: increase not observed

Auditory canalb (Zymbal gland)  (carcinomas)
        incidences:
        M: 0/1; 0/0; 2/7; and 19/33*
        F: 0/0; 0/1; 0/0; and 13/20*
          This document is a draft for review purposes only and does not constitute Agency policy,
                                             1-63              DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review ofBenzo[a]pyrene
       (Beland and Gulp (1998); Gulp et al.
       (1998))
       BSCSFi mice: female (48/dose group)
       0, 5, 25, or 100 ppm (average daily
       doses3: 0, 0.7, 3.3, and 16.5 mg/kg-d) in
       the diet
       2 yrs
Forestomach (papillomas and squamous cell carcinomas)
       incidences: 1/48; 3/47; 36/46*; and 46/47*

Esophagus (papillomas and carcinomas)
       incidences: 0/48; 0/48; 2/45; and 27/46*

Tongue (papillomas and carcinomas)
       incidences: 0/49; 0/48; 2/46; and 23/48*

Larynx (papillomas and carcinomas)
       incidences: 0/35; 0/35; 3/34; and 5/38
       Bruneetal. (1981)
       Sprague-Dawley rats: male and female
       (32/sex/dose)
       Gavage: 0, 6,18, 39 mg/kg-yr (0, 0.016,
       0.049, 0.107 mg/kg-d)
       Diet: 0, 6, 39 mg/kg-yr (0, 0.016,
       0.107 mg/kg-d)
       Treated until moribund or dead
       2 yrs
Forestomach (papillomas and carcinomas0); gavage
       incidences: 3/64; 12/64*; 26/64*; and 14/64*

Forestomach (papillomas); diet
       incidences: 2/64; 1/64; and 9/64*

Larynx and esophagus (papillomas); gavage
       incidences: 3/64; 1/64; 0/64; and 0/64

Larynx and esophagus (papillomas); diet
       incidences: 1/64; 2/64; and 1/64
 1
 2    * Indicates statistical significance as identified in study.
 3    aBased on the assumption that daily benzo[a]pyrene intake at 5 ppm was one-fifth of the 25-ppm intake (about
 4     21 u.g/day) and using TWA body weights of 0.032 kg for the control, 5- and 25-ppm groups and 0.026 kg for the
 5     100-ppm group.
 6    blncidencesarefor number of rats with tumors compared with number of tissues examined histologically.
 7     Auditory canal tissue was examined histologically when abnormalities were observed on macroscopic
 8     examination.
 9    °Two malignant forestomach tumors were observed (one each in the mid- and high-dose groups).

10    Inhalation exposure
11           The inhalation database of benzo[a]pyrene carcinogenicity studies consists of one lifetime
12    inhalation bioassay in male hamsters [Thyssenetal.. 1981].  Intratracheal instillation studies in
13    hamsters  are also available [Feron and Kruysse. 1978: Ketkar etal.. 1978: Feronetal.. 1973: Henry
14    etal.. 1973: Saffiottietal.. 19721
15           Several long term intratracheal installation studies in hamsters evaluated the
16    carcinogenicity of benzo[a]pyrene [Feron and Kruysse. 1978: Feronetal.. 1973: Henry etal.. 1973:
17    Saffiotti etal.. 1972).  These studies treated animals with benzo[a]pyrene once a week in a saline
18    solution (0.5-0.9%) for >8 months and observed animals for 1-2 years following cessation of
19    exposure. Tumors  in the larynx, trachea, bronchi, bronchioles, and alveoli were observed.
20    Individual studies also reported tumors  in the nasal cavity and forestomach. These intratracheal
21    instillation studies support the carcinogenicity of benzo[a]pyrene in the respiratory tract; however,
22    direct extrapolation from a dose delivered by intratracheal instillation to an inhalation
23    concentration expected to result in similar responses is not recommended [Driscoll etal.. 2000).
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                   1-64             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          Lifetime inhalation exposure to benzo[a]pyrene resulted in the development of tumors in
 2    the respiratory tract and pharynx in Syrian golden hamsters [Thyssen et al.. 1981). The authors
 3    stated that the rates of tumors of other organs generally corresponded to the rates in controls. U.S.
 4    EPA [1990] obtained individual animal data (including individual animal pathology reports for the
 5    respiratory and upper digestive tracts, time-to-death data, and exposure chamber monitoring data)
 6    from the study authors [Clement Associates. 1990): this information is summarized in Table 1-15.
 7    Concentration-dependent increased incidences of tumors in the upper respiratory tract, including
 8    the larynx and trachea, were seen at measured exposure concentrations of >9.5 mg/m3. In addition,
 9    a decrease in mean tumor latency was observed in the larynx and trachea.  Nasal cavity tumors
10    were observed at the mid- and high-concentration, but the incidences were not dose-dependent. A
11    concentration-related increase in tumors in the upper digestive tract (pharynx and esophagus) was
12    also reported. In addition, a single forestomach tumor was observed in each of the mid- and high-
13    concentration groups. Also, in animals with a tumor in either the larynx or pharynx, forestomach
14    tumors were not observed in control animals.  The study authors suggested that the upper digestive
15    tract tumors were a consequence of mucociliary particle clearance. All nasal, forestomach,
16    esophageal, and tracheal tumors occurred in hamsters that also had tumors in the larynx or
17    pharynx, except in the mid-concentration group, where two animals with nasal tumors had no
18    tumors in the pharynx or larynx.
19          A re-analysis of the individual animal pathology reports and the exposure chamber
20    monitoring data provided by the study authors yielded estimates of average continuous lifetime
21    exposures for each individual hamster. Group averages of individual average continuous lifetime
22    exposure concentrations were 0, 0.25,1.01, and 4.29  mg/m3 for the control through high-exposure
23    groups (U.S. EPA. 1990).
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-65            DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review ofBenzo[a]pyrene
             Table 1-15. Tumors observed in chronic inhalation animal bioassays
                 Reference and study design
                    Results'1
      Thyssenetal. (1981)
      Syrian golden hamsters: male
      (26-34 animals/group placed on study)

      0, 2.2, 9.5, or 46.5 mg/m3 on NaCI particles by nose only
      inhalation for 3-4.5 hrs, 5-7 d/wk
      (TWA exposure concentrations3: 0, 0.25,1.01, and
      4.29 mg/m3)

      Treated until moribund or dead (up to 130 wks)
      MMAD: not reported
Larynx
       incidences: 0/26; 0/21; 11/26; and 11/25
       mean tumor latency0: 107 and 68 wks
Pharynx
       incidences: 0/23; 0/19; 9/22; and 18/23
       mean tumor latency: 97 and 68 wks
Trachea
       incidences: 0/27; 0/21; 2/26; and 3/25
       mean tumor latency: 115 and 63 wks

Nasal cavity
       incidences: 0/26; 0/22; 4/26; and 1/34
       mean tumor latency: 116 and 79 wks

Esophagus
       incidences: 0/27; 0/22; 0/26; and 2/34
       mean tumor latency: 71 wks

Forestomach
       incidences: 0/27; 0/22; 1/26; and 2/34
       mean tumor latency: 119 and 72 wks
 2
 3    aDuration-adjusted inhalation concentrations calculated from exposure chamber monitoring data and exposure
 4     treatment times. Daily exposure times: 4.5 hours/day, 5 days/week on weeks 1-12; 3 hours/day, 5 days/week on
 5     weeks 13-29; 3.7 hours/day, 5 days/week on week 30; 3 hours/day, 5 days/week on weeks 31-41; and
 6     3 hours/day, 7 days/week for reminder of the experiment.
 7    bThyssen et al. (1981) reported only the incidences of malignant tumors, confirmed by comparison with the
 8     original study pathology data  (Clement Associates, 1990). The incidences summarized here include relevant
 9     benign tumors (papillomas, polyps, and papillary polyps). The malignant tumors were squamous cell carcinomas,
10     with the exception of one in situ carcinoma of the larynx and one adenocarcinoma of the nasal cavity, both in the
11     9.5 mg/m3 group. Denominators reflect the number of animals examined for histopathology for each tissue.  See
12     Section D.4.2 and Table E-17 in the Supplemental Material for study details and a complete listing of individual
13     data, respectively.
14    cMean time of observation of tumor, 9.5 and 46.5 mg/m3 concentration groups.
15    dThyssen et al. (1981) did not report statistical significance testing.  See Section D.4.2.

16    Dermal exposure

17           Repeated application of benzo[a]pyrene to skin (in the absence of exogenous promoters)
18    has been demonstrated to induce skin tumors in mice, rats, rabbits, and guinea pigs. These studies
19    have been reviewed by multiple national and international health agencies [IARC. 2010: IPCS. 1998:
20    ATSDR. 1995: IARC.  1983.1973]. Mice have been the most extensively studied species in dermal
21    carcinogenesis studies of benzo[a]pyrene because of evidence thatthey may be more sensitive than
22    other animal species; however, comprehensive comparisons of species differences in sensitivity to
                 This document is a draft for review purposes only and does not constitute Agency policy,
                                                    1-66             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    lifetime dermal exposure are not available. Systemic tumors in benzo[a]pyrene-treated mice were
 2    not increased compared to controls in lifetime dermal bioassays in which macroscopic examination
 3    of internal organs was included [Higginbotham etal.. 1993: Habsetal.. 1980: Schmahl etal.. 1977:
 4    Schmidt etal.. 1973: Roe etal.. 1970: Poel. 19591
 5          The analysis in this document focuses on lifetime carcinogenicity bioassays in several
 6    strains of mice following repeated dermal exposure to benzo[a]pyrene (Table 1-16). These studies
 7    involved 2- or 3-times/week exposure protocols, at least two exposure levels plus controls, and
 8    histopathological examinations of the skin and other tissues [Sivaketal.. 1997: Grimmer etal..
 9    1984: Habsetal.. 1984: Grimmer etal.. 1983: Habsetal.. 1980: Schmahl etal.. 1977: Schmidt etal..
10    1973: Roe etal.. 1970: Poel. 1963. 19591.
11          Numerous studies in mice observed skin tumors following benzo[a]pyrene exposure, but
12    were not considered further in this assessment because of the availability of the lifetime studies
13    identified above. These studies included several "skin painting" studies in mouse skin that did not
14    report the doses applied (e.g., Wynder and Hoffmann. 1959: Wynderetal.. 1957]: several shorter-
15    term studies (Albertetal.. 1991: Nesnowetal..  1983: Emmettetal.. 1981: Levin etal.. 1977]:
16    initiation-promotion studies utilizing acute dosing of benzo[a]pyrene followed by repeated
17    exposure to a potent tumor promoter; and studies involving vehicles expected to interact with or
18    enhance benzo[a]pyrene carcinogenicity (e.g.. Bingham and Falk. 1969].
19          One study applied benzo[a]pyrene (topically once a week for 6 months] to immuno-
20    compromised mice with human skin grafts (n = 10] and did not observe tumors, whereas all three
21    control mice (mice with no skin grafts] developed skin tumors (Urano etal.. 1995]. The authors
22    concluded this result indicates that human skin is much less susceptible to benzo[a]pyrene than
23    mouse skin. Though some studies indicate that the skin grafts maintain some metabolic function
24    (Das etal., 1986], it is unclear whether the human skin grafts maintain the same viability,
25    vascularization, and full metabolic capacity as human skin in vivo (Kappes etal.. 2004]. Another
26    concern is the short amount of time allowed for tumor development All of the mice with human
27    skin grafts treated with benzo [ajpyrene died within 6 months of the start of treatment (Urano etal..
28    1995]. While 6 months is generally sufficient for the development of tumors in mouse skin, human
29    latency for squamous cell carcinoma in PAH-exposed occupational cohorts is thought to be
30    >20 years (Youngetal.. 2012: Voelter-MahlknechtetaL 2007: Everall and Dowd. 1978].  Potent
31    mutagenic carcinogens such as 7,12-dimethylbenz[a]anthracene, methylcholanthrene, and
32    methylnitronitrosoguanidine also fail to produce skin tumors in this model system (Soballe etal..
33    1996: Urano etal.. 1995: Graem. 1986]. Therefore, the ability of this model system to predict
34    hazard for human skin cancer risk (particularly from metabolically active carcinogens] is unclear.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-67            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
       Table 1-16. Tumors observed in chronic dermal animal bioassays
         Reference and study design
                       Results3
Poel (1959)
C57Lmice: male (13-56/dose)
0, 0.15, 0.38, 0.75, 3.8, 19, 94, 188, 376, or 752 u.g
Dermal; 3 times/wk for up to 103 wks or until the
appearance of a tumor by gross examination
Skin tumors (gross skin tumors and epidermoid carcinoma);
dose-dependent decreased time of tumor appearance
        incidences:
        Gross skin tumors: 0/33; 5/55; 11/55; 7/56; 41/49;
        38/38; 35/35; 12/14; 14/14; and 13/13
        Epidermoid carcinoma: 0/33; 0/55; 2/55; 4/56;
        32/49; 37/38; 35/35; 10/14; 12/14; and  13/13
Cytotoxicity: information not provided
Poel (1963)
SWR, CSHeB, or A/He mice: male (14-25/dose)
0, 0.15, 0.38, 0.75, 3.8, 19.0, 94.0, or 470 u.g
Dermal; 3 times/wk until mice died or a skin tumor
was observed
Skin tumors and dose-dependent decreased time of first
tumor appearance
        incidences:
        SWR: 0/20; 0/25; 2/22; 15/18; 12/17; 16/16; 16/17;
        and 14/14
        CSHeB: 0/17; 0/19; 3/17; 4/17; 11/18; 17/17; 18/18;
        and 17/17
        A/He mice: 0/17; 0/18; 0/19; 0/17; 0/17; 21/23;
        11/16; and 17/17
Cytotoxicity: information not provided
Roe etal. (1970)
Swiss mice: female (50/dose)
0, vehicle, 0.1,0.3,1, 3, or 9 |jg
Dermal; 3 times/wk for up to 93 wks
Skin tumors; malignant skin tumors were observed in 4/41 and
31/40 mice in the two high-dose groups, respectively
        incidences:
        0/43; 0/47; 1/42; 0/42; 1/43; 8/41; and 34/46
Cytotoxicity: information not provided
Schmidt etal. (1973)
NMRI mice: female (100/group)
Swiss mice: female (100/group)
0, 0.05, 0.2, 0.8, or 2 u.g
Dermal; 2 times/wk until spontaneous death
occurred or until an advanced carcinoma was
observed
Skin tumors (carcinomas)
        incidences:
        NMRI:
        2/100 at 2 |jg (papillomas);
        2/100 at 0.8 u.g and 30/100 at 2 u.g (carcinomas)
        Swiss:
        3/80 at 2 |jg (papillomas);
        5/80 at 0.8 u.g and 45/80 at 2 u.g (carcinomas)
Cytotoxicity: information not provided
Schmahl etal. (1977)
NMRI mice: female (100/group)
0,1,1.7, or 3 u.g
Dermal; 2 times/wk until natural death or until they
developed a carcinoma at the site of application
Skin tumors (papillomas and carcinomas)
        incidences:
        0/81; 1/77; 0/88; and 2/81 (papillomas)
        0/81; 10/77; 25/88; and 43/81 (carcinomas)
Cytotoxicity: information not provided
Habs etal. (1980)
NMRI mice: female (40/group)
0,1.7, 2.8, or 4.6 |jg
Dermal; 2 times/wk until natural death or gross
observation of infiltrative tumor growth
Skin tumors and dose-dependent increase in age-standardized
tumor incidence
        incidences:
        0/35; 8/34; 24/35; and 22/36
        age-standardized tumor incidence:
        0, 24.8, 89.3, and 91.7%
Cytotoxicity: information not provided
          This document is a draft for review purposes only and does not constitute Agency policy,
                                               1-68             DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
               Reference and study design
                      Results3
       (Grimmer et al. (1984); Grimmer et al. (1983))
       CFLP mice: female (65-80/group)
       0, 3.9, 7.7, or 15.4 u.g (1983 study)
       0, 3.4, 6.7, or 13.5 u.g (1984 study)
       Dermal; 2 times/wk for 104 wks
Skin tumors (papillomas and carcinomas) with a decrease in
tumor latency
       incidences:
       1983: 0/80; 7/65; 5/64; and 2/64 (papillomas)
            0/80; 15/65; 34/64; and 54/64 (carcinomas)
       1984: 0/65; 6/64; 8/65; and 4/65 (papillomas)
            0/65; 37/64; 45/65; and 53/65 (carcinomas)
Cytotoxicity: information not provided
       Habsetal. (1984)
       NMRI mice: female (20/group)
       0, 2, or 4 u.g
       Dermal; 2 times/wk for life
Skin tumors (papillomas and carcinomas) with a decrease in
mean survival time
       incidences:
       0/20; 2/20; and 0/20 (papillomas)
       0/20; 7/20; and 17/20 (carcinomas)
Cytotoxicity: information not provided
       (Sivak et al. (1997); Arthur D Little, 1989; NIOSH
       (1989))
       C3H/HeJ mice: male (30/group)
       0, 0.05, 0.5, or 5 u.g
       Dermal; 2 times/wk for up to 104 wks
Skin tumors (papillomas and carcinomas)
       incidences:
       0/30; 0/30; 5/30 (1 papilloma, 1 keratoacanthoma,
       3 carcinomas); and 27/30 (1 papilloma,
       28 carcinomas)
Cytotoxicity:  80% incidence of scabs and sores in highest dose
group; no Cytotoxicity noted at lower doses
 1
 2    Statistical significance not reported by study authors.

 3    Mode-of-Action Analysis—Carcinogenicity
 4           The carcinogenicity of benzo[a]pyrene, the most studied PAH, is well documented [IARC.
 5    2010: Xuetal.. 2009: TiangetaL 2007: TiangetaL 2005: Xue and Warshawsky. 2005: RameshetaL
 6    2004: Bostr6metal..20Q2: Penningetal.. 1999: IPCS. 1998: Harvey. 1996: ATSDR. 1995: Cavalieri
 7    andRogan. 1995: U.S. EPA. 1991b). The primary mode of action by which benzo[a]pyrene induces
 8    carcinogenicity is via a mutagenic mode of action. This mode of action is presumed to apply to all
 9    tumor types and is relevant for all routes of exposure. The general sequence of key events
10    associated with a mutagenic mode of action for benzo[a]pyrene is: (1) bioactivation of
11    benzo[a]pyrene to DNA-reactive metabolites via three possible metabolic activation  pathways: a
12    diol epoxide pathway, a radical cation pathway, and an o-quinone and ROS pathway; (2) direct DNA
13    damage by reactive metabolites, including the formation of DNA adducts and ROS-mediated
14    damage; (3) formation and fixation of DNA mutations, particularly in tumor suppressor genes or
15    oncogenes associated with tumor initiation; and (4) clonal expansion of mutated cells during the
16    promotion and progression phases of cancer development. These events are depicted as stages of
17    benzo[a]pyrene-induced carcinogenesis in Figure 1-6.
18           Benzo[a]pyrene is a complete carcinogen, in that it can act as both an initiator and a
19    promoter of carcinogenesis. Initiation via direct DNA damage (key event 2) can occur via all three
20    metabolites of benzo[a]pyrene. DNA damage that is not adequately repaired leads to mutation (key
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  1-69            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
event 3), and these mutations can undergo clonal expansion (key event 4) enabled by multiple
mechanisms also induced by benzo[a]pyrene, including AhR binding leading to an upregulation of
genes related to biotransformation, growth, and differentiation, and regenerative cell proliferation
resulting from cytotoxicity and a sustained inflammatory response. However, there is not sufficient
evidence that these mechanisms, which contribute to the promotion and progression phases of
cancer development, act independently of DNA damage and mutation to produce benzo[a]pyrene-
induced tumors (please see Other possible modes of action, below). The available human, animal,
and in vitro evidence supports a mutagenic mode of action as the primary mode by which
benzo[a]pyrene induces carcinogenesis.
                          Key events in the mode of action for benzo[a]pyrene carcinogenicity
                                                     Mutation
                                                    (transversion
                                                    K-ras,H-ras,
                                                    andpSJ targets
                                                               Upregulation of genes
                                                                  related to
                                                                biotransformation,
                                                                  growth, and
                                                                 differentiation
                                                     Mutation
                                                    (depurination
                                                     H-rastarget
                           ci-qumone
                            andROS
                                                Mutation
                                              (depurination
                                                oxidative
                                               damage and
                                              strand scission
                                       DNAadducts
                                       and oxidative
                                       base damage
       Figure 1-6.  Proposed metabolic activation pathways and key events in the
       carcinogenic mode of action for benzo[a]pyrene.

Data in Support of the Mode of Action
Summary of metabolic activation pathways
       Dial epoxide pathway. Benzo[a]pyrene diol epoxide metabolites, believed to be the most
potent DNA-binding metabolites of benzo[a]pyrene, are formed through a series of Phase I
metabolic reactions (see Appendix D of the Supplemental Information). The initial metabolism is
carried out primarily by the inducible activities of GYP enzymes including CYP1A1, CYP1B1, and
CYP1A2.  Further metabolism by epoxide hydrolase and the mixed function oxidase system yields
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                 1-70             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    (+)-anti-BPDE, one of the mostpotentDNA-binding metabolites of benzo[a]pyrene.
 2    Benzo[a]pyrene diol epoxide metabolites interact preferentially with the exocyclic amino groups of
 3    deoxyguanine and deoxyadenine [Geacintovetal.. 1997: TerinaetaL 1991). Adducts may give rise
 4    to mutations unless these adducts are removed by DNA repair processes prior to replication. The
 5    stereochemical nature of the diol epoxide metabolite (i.e., anti- versus syn-diol epoxides) affects the
 6    number and type of adducts and mutation that occurs [Geacintovetal.. 1997}. Transversion
 7    mutations (e.g., GC^TA or AT^TA) are the most common type of mutation found in mammalian
 8    cells following diol epoxide exposure (Bostrometal.. 2002].
 9          Radical cation pathway. Radical cation formation involves a one-electron oxidation by GYP
10    or peroxidase enzymes (i.e., horseradish peroxidase, prostaglandin H synthetase) that produce
11    electrophilic radical cation intermediates (Cavalieri and Rogan. 1995.1992}.  Radical cations can be
12    further metabolized to phenols and quinones (Cavalieri et al.. 1988e: Cavalieri etal.. 1988d). or they
13    can form unstable adducts with DNA that ultimately result in depurination. The predominant
14    depurinating adducts occur at the N-3 and N-7 positions of adenine and the C-8 and N-7 positions of
15    guanine (Cavalieri and Rogan. 1995].
16          o-Quinone/ROS pathway. The o-quinone metabolites of PAHs are formed by enzymatic
17    dehydroge nation of dihydrodiols (Bolton etal.. 2000: Penning etal.. 1999: Harvey. 1996: ATSDR.
18    1995] (see Appendix D of the Supplemental Information]. Dihydrodiol dehydrogenase enzymes are
19    members of the a-keto reductase gene superfamily. o-Quinone metabolites are potent cytotoxins,
20    are weakly mutagenic, and are capable of producing a broad spectrum of DNA damage. These
21    metabolites  can interact directly with DNA as well as result in the production of ROS (i.e., hydroxyl
22    and superoxide radicals] that may produce further cytotoxicity and DNA damage.  The
23    o-quinone/ROS pathway also can produce depurinated DNA adducts from benzo[a]pyrene
24    metabolites. In this pathway, and in the presence of NAD(P]+, aldo-keto reductase oxidizes
25    benzo[a]pyrene-7,8-diol to a ketol, which subsequently forms benzo[a]pyrene-7,8-dione. This and
26    other PAH o-quinones react with DNA to form unstable, depurinating DNA adducts.  In the presence
27    of cellular reducing equivalents, o-quinones can also activate redox cycles, which produce ROS
28    (Penning etal.. 1996].  DNA damage in in vitro systems following exposure to benzo[a]pyrene-
29    7,8-dione or other o-quinone PAH derivatives occurs through the aldo-keto reductase (AKR]
30    pathway and can involve the formation of stable DNA adducts (Baluetal., 2004], N-7 depurinated
31    DNA adducts (Mccoull etal.. 1999]. DNA damage from ROS (8-oxo-7,8-dihydro-2'-deoxyguanosine
32    adducts] (Park etal.. 2006).  and strand scission (Flowers etal.. 1997: Flowers etal.. 19961

33    Summary of genotoxicity and mutagenicity
34          The ability of metabolites of benzo[a]pyrene to cause mutations and other forms of DNA
35    damage in both in vivo and in vitro studies is well documented (see genotoxicity tables in
36    Appendix D in Supplemental Information).  With metabolic activation (e.g., the inclusion of S9),
37    benzo[a]pyrene is consistently mutagenic in the prokaryotic Salmonella/Ames and Escherichia coli
38    assays. In mammalian in vitro studies, benzo[a]pyrene is consistently mutagenic and clastogenic,
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-71            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    and induces cell transformation both with and without metabolic activation. Cytogenetic damage in
 2    the form of chromosomal aberrations (CAs), micronuclei (MN), sister chromatid exchanges (SCEs),
 3    and aneuploidy are commonplace following benzo[a]pyrene exposure as are DNA adduct
 4    formation, single-strand breaks (SSB), and induction of DNA repair and unscheduled DNA synthesis
 5    (UDS). In vitro mammalian cell assays have been conducted in various test systems, including
 6    human cell lines.
 7          In the majority of in vivo studies, benzo[a]pyrene has tested positive in multiple species and
 8    strains and under various test conditions for cell transformation, CAs, DNA adducts, DNA strand
 9    breaks, MN formation, germline mutations, somatic mutations (H-ras, K-ras, p53, lacZ, hprt), and
10    SCEs. Human studies are available following exposures to PAH mixtures through cigarette smoke
11    or occupational exposure in which benzo[a]pyrene-specific DNA adducts have been detected, and it
12    has been demonstrated qualitatively that benzo[a]pyrene metabolites damage DNA in exposed
13    humans.

14    Experimental support for the hypothesized mode of action
15          EPA's Guidelines for Carcinogen Risk Assessment [Section 2.4; [2005a]] describe a procedure
16    for evaluating mode-of-action data for cancer.  A framework for analysis of mode of action
17    information is provided and followed below.
18          Strength, consistency, and specificity of association. Strong evidence links the
19    benzo[a]pyrene diol epoxide metabolic activation pathway with key mutational events in genes that
20    are associated with tumor initiation (mutations in the p53 tumor suppressor gene and H-ras or
21    K-ras oncogenes) (Table 1-17). Results in support of a mutagenic mode of action via
22    benzo[a]pyrene diol epoxide include observations of frequent G^T transversion mutations in p53
23    and ras genes in lung tumors of human cancer patients exposed to coal smoke (Keohavongetal.,
24    2003: DeMarini et al.. 2001). These results are consistent with evidence thatbenzo[a]pyrene diol
25    epoxide is reactive with guanine bases in DNA; that G^T transversions, displaying strand bias, are
26    the predominant type of mutations caused by benzo[a]pyrene in several biological systems (Liu et
27    al.. 2005: Hainaut and Pfeifer. 2001: Marshall etal.. 1984): and that sites of DNA adduction at
28    guanine positions in cultured human HeLa or bronchial epithelial cells exposed to benzo[a]pyrene
29    diol epoxide correspond to p53 mutational hotspots observed in human lung cancers (Denissenko
30    etal., 1996: Puisieuxetal., 1991).  In addition,  mice exposed to benzo[a]pyrene in the diet (Culp et
31    al.. 2000) or by i.p. injection (Nesnowetal.. 1998a:  Nesnowetal.. 1998b: Nesnowetal.. 1996.1995:
32    Mass etal.. 1993) had forestomach or lung tumors, respectively, showing frequent G—>T or C
33    transversions in the K-ras gene. Supporting evidence includes observations that benzo[a]pyrene
34    diol epoxide (specifically (+)-anti-BPDE) is more potent than benzo[a]pyrene itself, benzo[a]pyrene
35    phenols, or benzo[a]pyrene diols in mutagenicity assays in bacterial and in vitro mammalian
36    systems (Malaveille etal., 1977: Newbold and Brookes, 1976)  and in producing lung tumors in
37    newborn mice following i.p. administration. Other  supporting evidence includes observations of
38    elevated BPDE-DNA adduct levels in WBCs of groups of coke oven workers and chimney sweeps,
               This document is a draft for review purposes  only and does not constitute Agency policy.
                                                1-72            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene

 1    occupations with known elevated risks of cancer [Vineisetal., 2007: Pavanello etal., 1999], and in
 2    lung tissue from tobacco smokers with lung cancer [Rojas etal.. 2004: Godschalketal.. 2002: Rojas
 3    etal.. 1998: Andreassen etal.. 1996: Alexandrovetal.. 1992}. Several epidemiological studies have
 4    indicated that PAH-exposed individuals who are homozygous for a CYP1A1 polymorphism, which
 5    increases the inducibility of this enzyme (thus increasing the capacity to produce benzo[a]pyrene
 6    diol epoxide), have increased levels of PAH or BPDE-DNA adducts [Aklillu etal.. 2005: Alexandrov
 7    etal.. 2002: Bartsch etal.. 2000: Perera and Weinstein. 2000).
 8          Additional supporting evidence of a mutagenic mode of action for benzo[a]pyrene
 9    carcinogenicity is the extensive database of in vitro and in vivo studies demonstrating the
10    genotoxicity and mutagenicity of benzo[a]pyrene following metabolic activation (Table 1-17). In
11    vitro studies overwhelmingly support the formation of DNA adducts, mutagenesis in bacteria, yeast,
12    and mammalian cells, several measures of cytogenetic damage (CA, SCE, MN), and DNA damage. In
13    vivo systems  in animal models are predominantly positive for somatic mutations following
14    benzo[a]pyrene exposure.
15          Support for the radical cation activation pathway contributing to tumor initiation through
16    mutagenic events includes observations that depurinated DNA adducts (expected products from
17    reactions of benzo[a]pyrene radical cations with DNA) accounted for 74% of identified DNA
18    adducts in mouse skin exposed to benzo[a]pyrene (Roganetal.. 1993] and that 9 of 13 tumors
19    examined from mice exposed to dermal applications of benzo[a]pyrene had H-ras oncogene
20    mutations attributed to depurinated DNA adducts from benzo[a]pyrene radical cations
21    (Chakravarti et al.. 1995).
22          Support for the o-quinone/ROS pathway contributing to tumor initiation via mutagenic
23    events includes in vitro demonstration that several types of DNA damage can occur from
24    o-quinones and ROS (Parketal.. 2006: Balu etal.. 2004: Mccoull etal.. 1999: Flowers etal.. 1997:
25    Flowers etal.. 1996). In addition, benzo[a]pyrene-7,8-dione can induce mutations in the p53 tumor
26    suppressor gene using an in vitro yeast reporter gene assay (Park etal.. 2008: Shenetal.. 2006: Yu
27    etal.. 2002). and dominant p53 mutations induced by benzo[a]pyrene-7,8-dione in this system
28    corresponded to p53 mutational hotspots observed in human lung cancer tissue (Park etal.. 2008).
29          Dose-response concordance and temporal relationship. Studies in humans demonstrating
30    that benzo [a]pyrene-induced mutational events in p53 or ras oncogenes precede tumor formation
31    are not available, but there is evidence linking benzo [a] pyrene exposure to  signature mutational
32    events in humans. In vitro exposure of human p53 knock-in murine fibroblasts to 1 |iM
33    benzo[a]pyrene for 4-6 days induced p53 mutations with similar features to those identified in p53
34    mutations in human lung cancer; i.e., predominance of G^T transversions with strand bias and
35    mutational hotspots atcodons 157-158 (Liu etal.. 2005).
36          Bennett etal. (1999) demonstrated a dose-response relationship between smoking
37    history/intensity and the types of p53 mutations associated with benzo[a]pyrene (G—>T
38    transversions) in human lung cancer patients (Table 1-17). In lung tumors of nonsmokers, 10% of
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-73            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene

 1   p53 mutations were G^T transversions, versus 40% in lung tumors from smokers with >60 pack-
 2   years of exposure.
 3          In mice, dose-response and temporal relationships have been described between the
 4   formation of BPDE-DNA adducts and skin and forestomach tumors (Table 1-17). In a study using
 5   mice treated dermally with benzo[a]pyrene once or twice per week for up to 15 weeks (10, 25, or
 6   50 nmol benzo[a]pyrene per application), levels of benzo[a]pyrene-DNA adducts in the skin, lung,
 7   and liver increased with increasing time of exposure and increasing dose levels (Talaskaetal..
 8   2006). Levels at the end of the exposure period were highest in the skin; levels in the lung and liver
 9   at the same time were 10- and 20-fold lower, respectively. Levels of benzo[a]pyrene-DNA adducts
10   in skin and lung increased in an apparent biphasic manner showing a lower linear slope between
11   the two lowest dose levels, compared with the slope from the middle to the highest dose.
12          Another study examined the dose-response  relationship and the time course of
13   benzo[a]pyrene-induced skin damage (Table 1-17),  DNA adduct formation, and tumor  formation in
14   female mice. Mice were treated dermally with 0,16, 32, or 64 |ig of benzo[a]pyrene once per week
15   for 29 weeks (Albert etal.. 1991). Indices of skin damage and levels of BPDE-DNA adducts in skin
16   reached plateau levels in exposed groups by 2 -4 weeks of exposure. With increasing dose level,
17   levels of BPDE-DNA adducts (fmol/ng DNA) initially increased in a linear manner and began to
18   plateau at doses >32 [ig/week. Tumors began appearing after  12-14 weeks of exposure for the
19   mid- and high-dose groups and at 18 weeks for the low-dose group. At study termination
20   (3 5 weeks after start of exposure), the mean number of tumors per mouse was approximately one
21   per mouse in the low- and mid-dose groups and eight per mouse in the high-dose group. The time-
22   course data indicate thatbenzo[a]pyrene-induced increases in BPDE-DNA adducts preceded the
23   appearance of skin tumors, consistent with the formation of DNA adducts as a precursor event in
24   benzo [a] pyrene-induced skin tumors. A follow-up to this study by the same authors (Albert etal.,
25   1996) measured DNA adducts, necrosis, and inflammation (marked by an increase in leukocytes) in
26   the skin of treated mice after 5 weeks of dermal exposure. In the 64 [ig/week dose group,
27   statistically elevated levels of DNA adducts, inflammation, and necrosis were reported; however, in
28   the lower dose group (16 [ig/week), DNA adducts were statistically significantly elevated without
29   increases in inflammation and necrosis.
30          Gulp etal. (1996) compared dose-response relationships for BPDE-DNA adducts and
31   tumors in female B6C3Fi mice exposed to benzo [a] pyrene in the diet at 0,18.5, 90, or 350  [J.g/day
32   for 28 days (to examine adducts) or 2 years (to examine tumors) (Table 1-17). The benzo[a]pyrene
33   dose-tumor response data showed a sharp increase  in forestomach tumor incidence between the
34   18.5 ^g/day group (6% incidence) and the 90 ^g/day group (78% incidence). The BPDE-DNA
35   adduct levels in forestomach showed a relatively linear dose-response throughout the
36   benzo[a]pyrene dose range tested. The appearance of increased levels of BPDE-DNA adducts in the
37   target tissue at 2 8 days is temporally consistent with the contribution of these adducts to the
38   initiation of forestomach tumors. Furthermore, about 60% of the examined tumors had mutations
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-74           DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene

 1    in the K-ras oncogene at codons 12 and 13, which were G^T or G^C transversions indicative of
 2    BPDE reactions with DNA (Gulp etal.. 1996).
 3           Biological plausibility and coherence. The evidence for a mutagenic mode of action for
 4    benzo[a]pyrene is consistent with the current understanding that mutations in p53 and ras
 5    oncogenes are associated with increased risk of tumor initiation (Table 1-17). The benzo[a]pyrene
 6    database is internally consistent in providing evidence for BPDE-induced mutations associated with
 7    tumor initiation in cancer tissue  from humans exposed to complex mixtures containing
 8    benzo[a]pyrene [Keohavongetal.. 2003: Pfeifer andHainaut. 2003: Pfeifer etal.. 2002: DeMarini et
 9    al., 2001: Hainaut and Pfeifer, 2001: Bennett etal., 1999), in animals exposed to benzo[a]pyrene
10    [Gulp etal.. 2000: Nesnowetal..  1998a: Nesnowetal.. 1998b: Nesnowetal.. 1996.1995: Mass etal..
11    1993). and in in vitro systems [Denissenko etal.. 1996: Puisieux et al.. 1991). Consistent
12    supporting evidence includes: (1) elevated BPDE-DNA adduct levels in tobacco smokers with lung
13    cancer [Rojas etal.. 2004: Godschalk etal.. 2002: Rojasetal.. 1998: Andreassen et al.. 1996:
14    Alexandrovetal.. 1992): (2) demonstration of dose-response relationships between G—>T
15    transversions in p53 mutations in lung tumors and smoking intensity [Bennett etal.. 1999): (3) the
16    extensive database of in vitro and in vivo studies demonstrating the genotoxicity and mutagenicity
17    of benzo[a]pyrene following metabolic activation;  and (4) general concordance between temporal
18    and dose-response relationships for BPDE-DNA adduct levels and tumor incidence in studies of
19    animals exposed to benzo[a]pyrene [Gulp etal.. 1996: Albert etal.. 1991). There is also supporting
20    evidence that contributions to tumor initiation through mutagenic events can be made by the
21    radical cation [Chakravartietal.. 1995: Roganetal.. 1993) and o-quinone/ROS metabolic activation
22    pathways [Park etal.. 2008: Park etal.. 2006: Shen etal.. 2006: Balu etal.. 2004: Yu etal.. 2002:
23    Mccoulletal.. 1999: Flowers etal.. 1997: Flowers etal.. 1996).

24           Table 1-17. Experimental support for the postulated key events for mutagenic
25           mode of action
      1. Bioactivation of benzo[a]pyrene to DNA-reactive metabolites via three possible metabolic activation
         pathways: a diol epoxide pathway, a radical cation pathway, and an o-quinone and ROS pathway
      Evidence that benzo[a]pyrene metabolites induce key events:
          •  Metabolism of benzo[a]pyrene via all three pathways has been demonstrated in multiple in vitro
             studies, and the diol epoxide and radical cation metabolic activation pathways have been demonstrated
             in in vivo studies in humans and animals (see Metabolic Activation Pathways section)
          •  Multiple in vivo studies in humans and animals have demonstrated distribution of reactive metabolites
             to target tissues
      Human evidence that key events are necessary for carcinogenesis:
       •  Humans with CYP polymorphisms or lacking a functional GSTM1 gene form higher levels of benzo[a]pyrene
          diol epoxides, leading to increased  BPDE-DNA adduct formation and increased risk of cancer (Vineis et al.,
          2007; Pavanello et al., 2005; Pavanello et al., 2004; Alexandrovetal., 2002; Perera and Weinstein, 2000)
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 1-75            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
2.  Direct DNA damage by the reactive metabolites, including the formation of DNA adducts and ROS-mediated
   damage
Evidence that benzo[a]pyrene metabolites induce key events:

    •   Reactive benzo[a]pyrene metabolites have demonstrated genotoxicity in most in vivo and in vitro
        systems in which they have been tested, including the bacterial mutation assay, transgenic mouse assay,
        dominant lethal mutations in mice, BPDE-DNA adduct detection in humans and animals, and DNA
        damage, CAs, MN formation, and SCE in animals (Appendix D in Supplemental Information)

    •   Multiple in vivo benzo[a]pyrene animal exposure studies have demonstrated DNA adduct formation in
        target tissues that precede tumor formation and increase in frequency with dose (Gulp et al., 1996;
        Talaska et al., 1996; Albert et al., 1991)

    •   Benzo[a]pyrene diol epoxide metabolites interact preferentially with the exocyclic amino groups of
        deoxyguanine and deoxyadenine in DNA (Geacintov et al., 1997; Jerina et al., 1991; Koreeda et al., 1978;
        Jeffrey et al., 1976)

    •   Benzo[a]pyrene o-quinone metabolites are capable of activating redox cycles and producing ROS that
        cause oxidative base damage (Park et al., 2006; Balu etal., 2004; Mccoull etal., 1999; Flowers et al.,
        1997; Flowers et al., 1996)

Human evidence that key events are necessary for carcinogenesis:

    •   Detection of benzo[a]pyrene diol epoxide-specific DNA adducts is associated with increased cancer risk
        in humans that are occupationally exposed (see Evidence in Humans section)

    •   These  benzo[a]pyrene diol epoxides formed BPDE-DNA adducts preferentially at guanine residues that
        have been detected in tissues of humans with cancer who were exposed to  PAHs (Vineis and  Perera,
        2007; Rojas et al., 2004; Godschalk et al., 2002; Li et al., 2001; Pavanello etal., 1999;  Rojaset al., 1998;
        Andreassen et al., 1996; Alexandrov et al., 1992)
3.  Formation and fixation of DNA mutations, particularly in tumor suppressor genes or oncogenes associated
   with tumor initiation
Evidence that benzo[a]pyrene metabolites induce key events:

    •   Several in vivo exposure studies have observed benzo[a]pyrene diol epoxide-specific mutational spectra
        (e.g., G->T transversion mutations) in K-ras, H-ras, and p53 in forestomach or lung tumors (Gulp et al.,
        2000; Nesnowetal., 1998a; Nesnowetal., 1998b; Nesnowetal., 1996,1995; Mass etal., 1993)

    •   Multiple animal exposure studies have identified benzo[a]pyrene-specific mutations in H-ras, K-ras, and
        p53 in target tissues preceding tumor formation (Liu etal., 2005; Wei et al., 1999; Gulp etal., 1996)
        (Chakravarti et al., 1995; Ruggeri et al., 1993)

Human evidence that key events are necessary for carcinogenesis:

    •   DNA adducts formed by the benzo[a]pyrene diol epoxide reacting with guanine bases lead
        predominantly to G->T transversion mutations; these specific mutational spectra have been identified in
        PAH-associated tumors in humans at mutational hotspots, including oncogenes (K-ras) and tumor
        suppressor genes (p53) (Liu et al., 2005;  Keohavong et al., 2003; Pfeifer and Hainaut, 2003; Pfeifer et al.,
        2002; DeMarini etal., 2001; Hainaut and Pfeifer, 2001; Bennett et al., 1999; Denissenko et al., 1996;
        Puisieuxetal., 1991; Marshall etal., 1984; Koreeda et al., 1978; Jeffrey et al., 1976)
           This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-76              DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
      4. Clonal expansion of mutated cells during the promotion and progression phases of cancer development
      Evidence that benzo[a]pyrene metabolites induce key events:
          •  Benzo[a]pyrene has been shown to be a complete carcinogen, in that skin tumors in mice, rats, rabbits,
             and guinea  pigs have  been associated with repeated application of benzo[a]pyrene to skin in the
             absence of exogenous  promoters (IPCS, 1998; Sivak et al., 1997; ATSDR, 1995; Grimmer et al., 1984;
             Habs et al., 1984; Grimmer et al., 1983; IARC, 1983, 1973; Habs et al., 1980; Schmahl et al., 1977;
             Schmidt etal., 1973; Roe et al., 1970; Poel, 1960, 1959(IPCS, 1998; ATSDR, 1995; IARC, 1983, 1973)
          •  Mice exposed dermally to benzo[a]pyrene for 26 weeks were found to have increased frequencies of H-
             ras mutations in exposure-induced hyperplastic lesions that were further increased in tumors (Wei etal.,
             1999)
          •  AhR activation by PAHs (including benzo[a]pyrene) upregulates genes responsible for tumor promotion
             and increases tumor incidence in mice (Ma and Lu, 2007; Talaska et al., 2006; Shimizu et al., 2000)
 1    Other possible modes of action
 2           The carcinogenic process for benzo[a]pyrene is likely to be related to some combination of
 3    molecular events resulting from the formation of several reactive metabolites that interact with
 4    DNA to form adducts and produce DNA damage resulting in mutations in cancer-related genes, such
 5    as tumor suppressor genes or oncogenes.  These events may reflect the initiation potency of
 6    benzo[a]pyrene. However, benzo[a]pyrene possesses promotional capabilities that may be related
 7    to AhR affinity, immune suppression, cytotoxicity and inflammation (including the formation of
 8    ROS),  as well as the inhibition of gap junctional intercellular communication (GJIC).
 9           The ability of certain PAHs to act as initiators and promoters may increase their
10    carcinogenic potency. The promotional effects of PAHs appear to be related to AhR affinity and the
11    upregulation of genes related to growth and differentiation (Bostrom etal.. 2002]. The genes
12    regulated by this receptor belong to two major functional groups (i.e., induction of metabolism or
13    regulation cell differentiation and proliferation). PAHs bind to the cytosolic AhR in complex with
14    heat shock protein 90. The ligand-bound receptor is then transported to the nucleus in complex
15    with the AhR nuclear translocator protein. The AhR complex interacts with AhR elements of DNA
16    to increase the transcription of proteins associated with induction of metabolism and regulation of
17    cell differentiation and proliferation. Following benzo[a]pyrene exposure, disparities have been
18    observed in the tumor pattern and toxicity of Ah-responsive  and Ah-nonresponsive mice, as
19    Ah-responsive mice were more susceptible to tumorigenicity in target tissues such as liver, lung,
20    and skin (Ma and Lu. 2007: Talaska etal.. 2006: Shimizu etal.. 20001.
21           Benzo[a]pyrene has both inflammatory and immunosuppressive effects that may function
22    to promote tumorigenesis. Inflammatory  responses to cytotoxicity may contribute to the tumor
23    promotion process; for example, benzo[a]pyrene quinones (1,6-, 3,6-, and 6,12-benzo[a]pyrene-
24    quinone) generated ROS and increased cell proliferation by enhancing the epidermal growth factor
25    receptor pathway in cultured breast epithelial cells  (Burdicketal.. 2003]. In addition, several
26    studies have demonstrated that exposure  to benzo[a]pyrene increases the production of
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  1-77            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    inflammatory cytokines, which may contribute to cancer progression [N'Diaye etal., 2006: Tamaki
 2    et al.. 2004: Garcon et al.. 2001b: Garcon et al.. 2001a: Albert et al.. 1996: Albert et al.. 19911 One of
 3    these studies, Albert etal. [1996]. measured DNA adducts, necrosis, and inflammation (marked by
 4    an increase in leukocytes) in the skin of benzo[a]pyrene-treated mice after 5 weeks of dermal
 5    exposure. In the highest dose group, statistically elevated levels of DNA adducts, inflammation, and
 6    necrosis were reported; however, in the lower dose group, DNA adducts were statistically
 7    significantly elevated without increases in inflammation and necrosis. It is likely that at high doses
 8    of benzo[a]pyrene, inflammation promotes the formation of tumors.
 9          In addition to inflammation, immunosuppressive effects of benzo[a]pyrene have been noted
10    [as reviewed in, Zaccaria and McClure, 2013]. Immune effects of benzo[a]pyrene exposure (see
11    Section 1.1.3] may provide an environment where tumor cells can evade detection by immune
12    surveillance mechanisms normally responsible for recognizing and eliminating nascent cancer cells
13    (Hanahan and Weinberg. 2011]. In addition, the developing fetus may be even more sensitive to
14    these effects; Urso and Gengozian (1980] found that mice exposed to benzo[a]pyrene in utero not
15    only had a significantly increased tumor incidence as adults but also a persistently suppressed
16    immune system.
17          Gap junctions are channels between cells that are crucial for differentiation, proliferation,
18    apoptosis, and cell death. Interruption of GJIC is associated with a loss of cellular control of growth
19    and differentiation, and consequently with the two epigenetic steps of tumor formation, promotion
20    and progression.  Thus, the inhibition of gap junctional intercellular communication by
21    benzo[a]pyrene, observed in vitro (Sharovskaya et al.. 2006: Blahaetal.. 2002]. provides another
22    mechanism of tumor promotion.
23          In summary, there are tumor-promoting effects of PAH exposures that are not mutagenic.
24    Although these effects are observed following benzo[a]pyrene-specific exposures, the occurrence of
25    BPDE-DNA adducts and associated mutations that precede both cytotoxicity and tumor formation
26    and increase with dose  provides evidence that mutagenicity is the primary event that initiates
27    tumorigenesis following benzo[a]pyrene exposures. A biologically plausible mode of action may
28    involve a combination of effects induced by benzo[a]pyrene, with mutagenicity as the initiating
29    tumorigenic event Subsequent AhR activation and cytotoxicity could then lead to increased ROS
30    formation, regenerative cell proliferation, and inflammatory responses, which, along with evasion
31    of immune surveillance and GJIC, would provide an environment where the selection for mutated
32    cells increases the rate of mutation, allowing clonal expansion and progression of these tumor cells
33    to occur. However, it was determined that, in comparison to the large database on the mutagenicity
34    of benzo[a]pyrene, there were insufficient data to develop a separate mode of action analysis for
35    these promotional effects.

36    Conclusions about the hypothesized mode of action
37          There is sufficient evidence to conclude that the major mode of action for benzo[a]pyrene
38    carcinogenicity involves mutagenicity mediated by DNA reactive metabolites. The evidence for a
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-78            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    mutagenic mode of action for benzo[a]pyrene is consistent with the current understanding that
 2    mutations in p53 and ras oncogenes are associated with increased risk of tumor initiation.  The
 3    benzo[a]pyrene database provides strong and consistent evidence for BPDE-induced mutations
 4    associated with tumor initiation in cancer tissue from humans exposed to complex mixtures
 5    containing benzo[a]pyrene, in animals exposed to benzo[a]pyrene, and in in vitro systems.
 6    Supporting evidence suggests that contributions to tumor initiation through potential mutagenic
 7    events can be made by the radical cation and o-quinone/ROS metabolic activation pathways. Other
 8    processes may contribute to the carcinogenicity of benzo[a]pyrene via the promotion and
 9    progression phases of cancer development (e.g., inflammation, cytotoxicity, sustained regenerative
10    cell proliferation).

11    Support for the Hypothesized Mode of Action in Test Animals
12          Benzo[a]pyrene induces gene mutations in a variety of in vivo and in vitro systems and
13    produces tumors in all animal species tested and by all routes of exposure (see Appendix D in
14    Supplemental Information). Strong, consistent evidence in animal models supports the postulated
15    key events: the metabolism of benzo[a]pyrene to DNA-reactive intermediates, the formation of
16    DNA adducts, the subsequent occurrence of mutations in oncogenes and tumor suppressor genes,
17    and the clonal expansion of mutated cells.

18    Relevance of the Hypothesized Mode of Action to Humans
19          A substantial database indicates that the postulated key events for a mutagenic mode of
20    action all occur in human tissues.  Evidence is available from studies of humans exposed to PAH
21    mixtures (including coal smoke and tobacco smoke) indicating a contributing role for
22    benzo[a]pyrene diol epoxide in inducing key mutational events in genes that are associated with
23    tumor initiation (mutations in the p53 tumor suppressor gene and H-ras or K-ras oncogenes). The
24    evidence includes observations of a spectrum of mutations in ras oncogenes and the p53 gene in
25    lung tumors of human patients exposed to coal smoke or tobacco smoke) that are similar to the
26    spectrum of mutations caused by benzo[a]pyrene diol epoxide in several biological systems,
27    including tumors from mice exposed to benzo[a]pyrene. Additional supporting evidence includes
28    correspondence between hotspots of p53 mutations in human lung cancers and sites of DNA
29    adduction by benzo[a]pyrene diol epoxide in experimental systems, and elevated BPDE-DNA
30    adduct levels in respiratory tissue of lung cancer patients or tobacco smokers with lung cancer.

31    Populations or Lifestages Particularly Susceptible to the Hypothesized Mode of Action
32          A mutagenic mode of action for benzo[a]pyrene-induced carcinogenicity is considered
33    relevant to all populations and lifestages. The current understanding of biology of cancer indicates
34    that mutagenic chemicals, such as benzo[a]pyrene, are expected to exhibit a greater effect in early
35    life exposure versus later life exposure (U.S. EPA. 2005b: Vesselinovitchetal..  1979).  The EPA's
36    Supplemental Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA.

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-79            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    2005b] recommends the application of age-dependent adjustment factors (ADAFs) for carcinogens
 2    that act through a mutagenic mode of action. Given that a determination benzo [a]pyrene acts
 3    through a mutagenic mode of carcinogenic action has been made, ADAFs should be applied along
 4    with exposure information to estimate cancer risks for early-life exposure.
 5          Toxicokinetic information suggest early lifestages may have lower levels of some GYP
 6    enzymes than adults [Ginsberg etal.. 2004: Cresteil. 1998). suggesting that lower levels of
 7    mutagenic metabolites may be formed in early lifestages.  Though expression of bioactivating
 8    enzymes is believed to be lower in the developing fetus and children, metabolism of
 9    benzo[a]pyrene still occurs, as indicated by the detection of benzo[a]pyrene-DNA or protein
10    adducts or urinary metabolites [Naufaletal., 2010: Ruchirawatetal., 2010: Suter etal., 2010:
11    Mielzynskaetal.. 2006: Pereraetal.. 2005a: Tang etal.. 1999: Whyattetal.. 1998).  While
12    expression of GYP enzymes is lower in fetuses and infants, the greater liver to body mass ratio and
13    increased blood flow to liver in fetuses and infants may compensate for the decreased expression of
14    GYP enzymes [Ginsberg et al.. 2004]. Activity of Phase II detoxifying enzymes in neonates and
15    children is adequate for sulfation but decreased for glucuronidation and glutathione conjugation
16    [Ginsberg etal., 2004]. The conjugation of benzo[a]pyrene-4,5-oxide with glutathione was
17    approximately one-third less in human fetal than adult liver cytosol [Pacifici etal.,1988].
18          In addition, newborn or infant mice develop liver and lung tumors more readily than young
19    adult mice following acute i.p. exposures to benzo[a]pyrene [Vesselinovitch et al.. 1975].  These
20    results indicate that exposure to  benzo[a]pyrene during early life stages presents additional risk for
21    cancer, compared with exposure during adulthood, despite lower metabolic activity in early
22    lifestages. Population variability in metabolism and detoxification of benzo [a]pyrene, in addition to
23    DNA repair capability, may affect cancer risk. Polymorphic variations in the human population in
24    CYP1A1, CYP1B1, and other GYP  enzymes have been implicated as determinants of increased
25    individual cancer risk  in some studies [Ickstadtetal.. 2008: Aklillu etal.. 2005: Alexandrov et al..
26    2002: Perera and Weinstein. 2000]. Some evidence suggests that humans lacking a functional
27    GSTM1 gene have higher BPDE-DNA adduct levels and are thus at greater risk for cancer [Binkova
28    etal.. 2007: Vineis etal.. 2007: Pavanello etal.. 2005: Pavanello etal.. 2004: Alexandrov et al.. 2 0 0 2:
29    Perera and Weinstein. 2000]. In addition, acquired deficiencies or inherited gene polymorphisms
30    that affect the efficiency or fidelity of DNA repair may also influence individual susceptibility to
31    cancer from environmental mutagens [Wangetal.. 2010: Ickstadtetal.. 2008: Binkova etal.. 2007:
32    Matulloetal..20Q3: Shen etal.. 2003: Cheng etal.. 2000: Perera and Weinstein. 2000: Wei etal..
33    2000: Amos etal.. 1999].  In general, however, available support for the role of single
34    polymorphisms in significantly modulating human PAH cancer risk from benzo[a]pyrene or other
35    PAHs is relatively weak or inconsistent  Combinations of polymorphisms, on the other hand, may
36    be critical determinants of a cumulative DNA-damaging dose, and thus indicate greater
37    susceptibility to cancer from benzo [ajpyrene exposure [Vineis etal., 2007].
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-80             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1   An alysis of Toxicogen omicsData
 2          An analysis of pathway-based transcriptomic data was conducted to help inform the cancer
 3   mode of action for benzo[a]pyrene (see the Supplemental Information for details of this analysis).
 4   These data support a mutagenic and cellular proliferation mode of action that follows three
 5   candidate pathways: aryl hydrocarbon signaling; DNA damage regulation of the Gl/S phase
 6   transition; and/or Nrf2 regulation of oxidative stress. Specifically, the analysis showed that
 7   benzo[a]pyrene may activate the AhR, leading to the formation of oxidative metabolites and
 8   radicals which may lead to oxidative damage and DNA damage.  Subsequently, DNA damage can
 9   occur and activate p53 andp53 target genes, including p21 and MDM2. In addition, the data
10   indicate that p53 signaling may be decreased under these conditions, as ubiquitin and MDM2 are
11   both upregulated, and work together to degrade p53. Furthermore, the transcriptional
12   upregulation of cyclin D may result in enough cyclin D protein to overcome the p21 inhibitory
13   competition for CDK4, allowing for Gl/S phase transition to occur. The data also support the
14   hypothesis that an upregulation of proliferating cell nuclear antigen (PCNA) in combination with
15   the upregulation of ubiquitin indicates that cells are moving towards the Gl/S phase transition.
16   Although the alterations to the Nrf2 pathway suggest cells are preparing for a pro-apoptotic
17   environment, there is no transcriptional evidence that the apoptotic pathways are being activated.
18          There are uncertainties associated with the available transcriptomics data. For instance,
19   the available studies only evaluate gene expression following benzo[a]pyrene exposure and do not
20   monitor changes in protein or metabolite expression, which would be more indicative of an actual
21   cellular state change. Further research is required at the molecular level to demonstrate that the
22   cellular signaling events being inferred from such data are actually operative and result in
23   phenotypic changes. In addition, this analysis relied upon two short term studies that evaluated
24   mRNA expression levels in a single tissue (liver) and species  (mouse) and were conducted at
25   relatively high doses.

26   1.2. SUMMARY AND EVALUATION

27   1.2.1.  Weight of Evidence for Effects Other than Cancer
28          The weight of the evidence from human and animal studies indicates that the strongest
29   evidence for human hazards following benzo[a]pyrene exposure is for developmental toxicity,
30   reproductive toxicity, and immunotoxicity. Most of the available human data on benzo[a]pyrene
31   report associations between particular health endpoints and concentrations of benzo[a]pyrene-
32   DNA adducts, with fewer noncancer studies correlating health effects with external measures of
33   exposure. In general, the available human studies report effects that are analogous to the effects
34   observed in animal toxicological studies, and provide qualitative, supportive evidence for the effect-
35   specific hazards identified in Sections 1.1.1-1.1.4.
36          In animals, evidence of developmental toxicity, reproductive toxicity, and immunotoxicity
37   has been observed across species and dosing regimens. The  available evidence from mice and rats

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-81            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    treated by gavage during gestation or in the early postnatal period demonstrate developmental
 2    effects including decreased body weight, decreased fetal survival, decreased fertility, atrophy of
 3    reproductive organs, and altered neurobehavioral outcomes [Chenetal.. 2012: Tules etal.. 2012:
 4    Bouayed et al.. 2009a: Kristensenetal.. 1995: Mackenzie and Angevine. 1981). Several studies in
 5    animals have indicated that exposure to benzo[a]pyrene in early life may result in altered
 6    neurobehavioral outcomes and sensorimotor development [Chenetal.. 2012: Bouayed etal..
 7    2009a]. Male and female reproductive toxicity, as evidenced by effects on sperm parameters,
 8    decreased reproductive organ weights, histological changes, and hormone alterations, have been
 9    observed after oral exposure in rats and mice [Chenetal., 2011: Chung etal., 2011: Mohamed etal.,
10    2010: Zheng etal., 2010: Mackenzie and Angevine, 1981]. Benzo[a]pyrene exposure has also been
11    shown to lead to altered immune cell populations and histopathological changes in immune system
12    organs [Kroese etal.. 2001: De long et al.. 1999). as well as thymic and splenic effects following
13    subchronic oral exposure. Varying immunosuppressive responses are also observed in short-term
14    oral and injection studies. Overall, the oral data support the conclusion that developmental toxicity
15    and reproductive toxicity are human hazards following exposure to benzo[a]pyrene and that
16    immunotoxicity is a potential human hazard of benzo[a]pyrene exposure.
17          Following inhalation exposure to benzo[a]pyrene in animals, evidence of developmental
18    and reproductive toxicity has been observed. Decreased fetal survival has been observed in rats
19    exposed to benzo[a]pyrene via inhalation during gestation [Wormley  etal.. 2004: Archibongetal..
20    2002). Male reproductive toxicity, as evidenced by effects on sperm parameters, decreased testes
21    weight, and  hormone alterations, has also been observed in rats following subchronic inhalation
22    exposure to benzo[a]pyrene [Archibongetal.. 2008: Ramesh etal.. 2008]. Female reproductive
23    toxicity, as evidenced by modified hormone levels in dams, has been observed following inhalation
24    exposure to benzo[a]pyrene during gestation [Archibongetal., 2002]. The inhalation data support
25    the conclusion that developmental toxicity and reproductive toxicity are human hazards following
26    exposure to benzo[a]pyrene.
27          Other types of effects were observed following benzo[a]pyrene exposure including
28    forestomach hyperplasia, hematological, hepatic, renal, cardiovascular, and adult neurological
29    toxicity (see Section 1.1.4]. Forestomach hyperplasia was observed following oral and inhalation
30    exposure; however, this endpoint most likely reflects early events in the neoplastic progression of
31    forestomach tumors following benzo[a]pyrene exposure (see Section  1.1.4], and was not
32    considered further for dose-response analysis and the derivation of reference values. For the
33    remaining effects, EPA concluded that the available evidence does not support these noncancer
34    effects as potential human hazards.

35    1.2.2.  Weight of Evidence for Carcinogenicity
36          Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005b], benzo[a]pyrene is
37    "carcinogenic to humans." This guidance emphasizes the importance  of weighing all of the evidence
38    in reaching conclusions about human carcinogenic potential. The descriptor of "carcinogenic to
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-82             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    humans" can be used when the following conditions are met: (a) there is strong evidence of an
 2    association between human exposure and either cancer or the key precursor events of the agent's
 3    mode of action but not enough for a causal association; (b) there is extensive evidence of
 4    carcinogenicity in animals; (c) the mode or modes of carcinogenic action and associated key
 5    precursor events have been identified in animals; and (d) there is strong evidence that the key
 6    precursor events that precede the cancer response in animals are anticipated to occur in humans
 7    and progress to tumors, based on available biological information. The data supporting these four
 8    conditions for benzo[a]pyrene are presented below and in Table 1-18.

 9    a) Strong Human Evidence of Cancer or its Precursors
10          There is a large body of evidence for human carcinogenicity for complex PAH mixtures
11    containing benzo[a]pyrene, including soot, coal tars, coal-tar pitch, mineral oils, shale oils, and
12    smoke from domestic coal burning [IARC. 2010: Baanetal.. 2009). There is also evidence of
13    carcinogenicity, primarily of the lung and skin, in occupations involving exposure to PAH mixtures
14    containing benzo[a]pyrene, such as chimney sweeping, coal gasification, coal-tar distillation, coke
15    production, iron and steel founding, aluminum production, and paving and roofing with coal tar
16    pitch [IARC, 2010: Baan etal.,  2009: Straif etal., 2005]. Increased cancer risks have been reported
17    among other occupations involving exposure to PAH mixtures such as carbon black and diesel
18    exhaust [Benbrahim-Tallaa et al..  2012: Bosetti et al.. 2007).  There is extensive evidence of the
19    carcinogenicity of tobacco smoke, of which benzo[a]pyrene is a notable constituent. The
20    methodologically strongest epidemiology studies (in terms of exposure assessment, sample size,
21    and follow-up period) provide consistent evidence of a strong association between benzo[a]pyrene
22    exposure and lung cancer. Three  large epidemiology studies in different geographic areas,
23    representing two different industries, observed increasing risks of lung cancer with increasing
24    cumulative exposure to benzo[a]pyrene (measured in |ig/m3-years), with approximately a twofold
25    increased risk at the higher exposures; each of these studies addressed potential confounding by
26    smoking f Armstrong and Gibbs. 2009: Spinelli etal.. 2006: Xu etal.. 19961  (Table 1-11). Although
27    the relative contributions of benzo[a]pyrene and of other PAHs cannot be established, the
28    exposure-response patterns seen with the benzo[a]pyrene measures make it unlikely that these
29    results represent confounding by other exposures. Similarly, for bladder cancer, two of the three
30    cohort studies with detailed exposure data observed an increasing risk with exposures
31    >80 [ig/m3-years(Gibbs and Sevigny. 2007a: Gibbs etal.. 2007: Gibbs and Sevigny. 2007b: Spinelli et
32    al.. 2006] (Table 1-13). The exposure range was much lower in the third study (Burstyn etal..
33    2007: Gibbs and Sevigny. 2007a: Gibbs etal.. 2007: Gibbs and Sevigny. 2007b). such that the highest
34    exposure group only reached the  level of exposure seen in the lowest exposure categories in the
35    other studies.  Data pertaining to  non-melanoma skin cancer is limited to studies with more indirect
36    exposure measures, e.g., based on occupations with likely dermal exposure to creosote (i.e., timber
37    workers, brick makers, and power linesmen); the RR estimates seen in the four available studies
38    that provide risk estimates for this type of cancer ranged from 1.5 to 4.6, with three of these four
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-83            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene

 1    estimates >2.5 and statistically significant [Pukkala, 1995: Karlehagenetal., 1992: Tornqvistetal.,
 2    1986: Hammond etal.. 1976). These four studies provide support for the association between
 3    dermal PAH exposure, including benzo[a]pyrene exposure, and skin cancer. Although it is likely
 4    that multiple carcinogens present in PAH mixtures contribute to the carcinogenic responses, strong
 5    evidence is available from several studies of humans exposed to PAH mixtures supporting a
 6    contributing role for benzo[a]pyrene diol epoxide in inducing key mutagenic precursor cancer
 7    events in target tissues. Elevated BPDE-DNA adducts have been reported in smokers compared to
 8    nonsmokers, and the increased adduct levels in smokers are typically increased twofold compared
 9    with nonsmokers [Phillips, 2002]. Elevated BPDE-DNA adduct levels have been observed in WBCs
10    of groups of coke oven workers and chimney sweeps, occupations with known elevated risks of
11    cancer [Rojas etal.. 2000: Bartsch etal.. 1999: Pavanello etal.. 1999: Bartschetal.. 1998: Rojas et
12    al.. 1998). and in lung tissue from tobacco smokers with lung cancer [Rojas etal.. 2004: Godschalk
13    etal.. 2002: Bartschetal.. 1999: Rojas etal.. 1998: Andreassen et al.. 1996: Alexandrov etal.. 1992).
14          Mutation spectra distinctive to diol epoxides have been observed in the tumor suppressor
15    gene p53 and the K-ras oncogene in tumor tissues taken from lung cancer patients who were
16    chronically exposed to two significant sources of PAH mixtures: coal smoke and tobacco smoke.
17    Hackman et al. [2000] reported an increase of GC^TA transversions and a decrease of GC^AT
18    transitions at the hprt locus in T-lymphocytes of humans with lung cancer who were smokers
19    compared to non-smokers. Lung tumors from cancer patients exposed to emissions from burning
20    smoky coal showed mutations in p53 and K-ras that were primarily G^T transversions (76 and
21    86%, respectively) [DeMarini etal.. 2001). Keohavong etal. [2003] investigated the K-ras
22    mutational spectra from nonsmoking women and smoking men chronically exposed to emissions
23    from burning smoky coal, and smoking men who resided in homes using natural gas;  among those
24    with K-ras mutations, 67, 86, and 67%, respectively, were G—>T transversions. Lung tumors from
25    tobacco smokers showed a higher frequency of p53 mutations that were G—>T transversions
26    compared with lung tumors in non-smokers [Pfeifer and Hainaut. 2003: Pfeifer etal..  2002: Hainaut
27    and Pfeifer. 2001]. and the frequency of these types of p53 mutations in lung tumors from smokers
28    increased with increasing smoking intensity [Bennett etal.. 1999].
29          Similarly, investigations of mutagenesis following specific exposures to benzo[a]pyrene (as
30    opposed to PAH mixtures] have consistently observed that the benzo[a]pyrene diol epoxide is very
31    reactive with guanine bases in DNA, and that G^T transversions are the predominant type of
32    mutations caused by benzo[a]pyrene diol epoxide in several biological test [Pfeifer and Hainaut.
33    2003: Hainaut and Pfeifer. 2001]. Following treatment of human HeLa cells with benzo [a] pyrene
34    diol epoxide, Denissenko etal. [1996] reported that the distribution of BPDE-DNA adducts within
35    p53 corresponded to mutational hotspots observed in p53 in human lung cancers. Benzo[a]pyrene
36    exposure induced mutations in embryonic fibroblasts from human p53 "knock-in" mice that were
37    similar to those found in smoking related human cancers, with a predominance of G—>T
38    transversions that displayed strand bias and were also located in the same mutational hotspots
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               1-84           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    found in p53 in human lung tumors [Liuetal., 2005]. These results, combined with a mechanistic
 2    understanding that mutations in p53 (which encodes a key transcription factor in DNA repair and
 3    regulation of cell cycle and apoptosis) may be involved in the initiation phase of many types of
 4    cancer, are consistent with a common mechanism for mutagenesis following exposures to PAH
 5    mixtures and provide evidence of a contributing role of benzo[a]pyrene diol epoxide in the
 6    carcinogenic response of humans to coal smoke and tobacco smoke.
 7          Therefore, while the epidemiological evidence alone does not establish a causal association
 8    between human exposure and cancer, there is strong evidence that the key precursor events of
 9    benzo[a]pyrene's mode of action are likely to be associated with tumor formation in humans.

10    b) Extensive Animal Evidence
11          In laboratory animals (rats, mice, and hamsters), exposures to benzo[a]pyrene via the oral,
12    inhalation, and dermal routes have been associated with carcinogenic responses both systemically
13    and at the site  of administration. Three 2-year oral bioassays are available that associate lifetime
14    benzo[a]pyrene exposure with carcinogenicity at multiple sites. These bioassays observed
15    forestomach, liver, oral cavity, jejunum, kidney, auditory canal (Zymbal gland), and skin or
16    mammary gland tumors in male and female Wistar rats (Kroese et al., 2001): forestomach tumors
17    in male and female Sprague-Dawley rats (Brune etal.. 1981): and forestomach, esophagus, tongue,
18    and larynx tumors in female B6C3Fi mice (Beland and Gulp. 1998: Gulp et al..  1998). Repeated or
19    short-term oral exposure to benzo[a]pyrene was associated with forestomach tumors in additional
20    bioassays with several strains of mice (Weyandetal.. 1995: Benjamin etal.. 1988: Robinson etal..
21    1987: El-Bayoumy. 1985: Triolo etal.. 1977: Wattenberg. 1974: Roe etal..  1970: Biancifiori etal..
22    1967: Chouroulinkovetal.. 1967: Fedorenko and Yansheva. 1967: Neal andRigdon. 1967:
23    Berenblum and Haran, 1955). EPA has considered the uncertainty associated with the relevance of
24    forestomach tumors for estimating human risk from benzo[a]pyrene exposure.  While humans do
25    not have a forestomach, squamous epithelial tissue similar to that seen in the rodent forestomach
26    exists in the oral cavity and upper two-thirds of the esophagus in humans  (IARC. 2003). Human
27    studies, specifically associating exposure to benzo[a]pyrene with the alimentary tract tumors are
28    not currently available. However, benzo[a]pyrene-DNA adducts have been detected in oral and
29    esophageal tissue obtained from smokers (reviewed by Phillips, 2002] and several epidemiological
30    studies have identified increased exposure to PAHs as an independent risk factor for esophageal
31    cancer (Abedi-Ardekani et al.. 2010: Szymanska et al.. 2 010: Gustavsson etal.. 1998: Liu etal..
32    1997). Thus, EPA concluded that forestomach tumors in rodents are relevant for assessing the
33    carcinogenic risk to humans.
34          Lifetime inhalation exposure to benzo[a]pyrene was associated primarily with tumors in
35    the larynx and pharynx of male Syrian golden hamsters exposed to benzo[a]pyrene:NaCl aerosols
36    (Thyssenetal., 1981). Additionally, less-than-lifetime oral exposure cancer bioassays in mice
37    provide supporting evidence that exposure to benzo[a]pyrene is associated with an increased
38    incidence of lung tumors in mice (Weyandetal.. 1995: Robinson etal.. 1987: Wattenberg. 1974). In
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-85            DRAFT—DO NOT CITE  OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    additional studies with hamsters, intratracheal instillation of benzo[a]pyrene was associated with
 2    upper and lower respiratory tract tumors [Feron and Kruysse. 1978: Ketkar etal.. 1978: Feron et
 3    al.. 1973: Henry etal.. 1973: Saffiotti etal.. 19721 Dermal application of benzo[a]pyrene
 4    (2-3 times/week) has been associated with mouse skin tumors in numerous lifetime bioassays
 5    fSivak etal.. 1997: Grimmer etal.. 1984: Habs etal.. 1984: Grimmer etal.. 1983: Habs etal.. 1980:
 6    Schmahl etal.. 1977: Schmidt etal.. 1973: Roe etal.. 1970: Poel. 1963.19591 Skin tumors in rats,
 7    rabbits, and guinea pigs have also been associated with repeated application of benzo[a]pyrene to
 8    skin in the absence of exogenous promoters [IPCS. 1998: ATSDR. 1995: IARC. 1983.1973). When
 9    followed by repeated exposure to a potent tumor promoter, acute dermal exposure to
10    benzo[a]pyrene induced skin tumors in numerous studies of mice, indicating that benzo[a]pyrene is
11    a strong tumor-initiating agent in the mouse skin model [Weyandetal.. 1992: Cavalierietal.. 1991:
12    Rice etal.. 1985:  El-Bayoumy etal.. 1982: Lavoie etal.. 1982: Ravehetal.. 1982: Cavalierietal..
13    1981: Slaga etal.. 1980: Wood etal.. 1980: Slaga etal.. 1978: Hoffmann etal.. 19721
14           Carcinogenic responses in animals exposed to benzo [a] pyrene by other routes of
15    administration include:  (1) liver or lung tumors in newborn mice given acute postnatal i.p.
16    injections (Lavoie etal.. 1994: Busby etal.. 1989: Weyand and Lavoie. 1988: Lavoie etal.. 1987:
17    Wislocki etal.. 1986: Busby etal.. 1984: Buening etal.. 1978: Kapitulnik etal.. 19781: (2) increased
18    lung tumor multiplicity in A/J adult mice given single i.p. injections [Mass etal.. 1993): (3) injection
19    site tumors in mice following s.c. injection [Nikonova. 1977: Pfeiffer. 1977: Homburger etal.. 1972:
20    Roe and Waters.  1967: Grant and Roe. 1963: Steiner. 1955: Rask-Nielsen. 1950: Pfeiffer and Allen.
21    1948: Bryan and Shimkin. 1943:  Barry etal.. 1935]: (4) injection site sarcomas in mice following
22    intramuscular injection[Sugiyama. 1973]: (5) mammary tumors in rats with intramammilary
23    administration [Cavalierietal., 1991: Cavalierietal., 1988c: Cavalieri et al., 1988b: Cavalierietal.,
24    1988a]: (6) cervical tumors in mice with intravaginal application (Naslundetal., 1987]: and
25    (7) tracheal tumors in rats with intratracheal implantation [Topping etal.. 1981: Nettesheim etal..
26    19771
27          Therefore, the animal database provides extensive evidence of carcinogenicity in animals.

28    c) Key Precursor Events have been Identified in Animals
29          There is sufficient evidence to conclude that benzo[a]pyrene carcinogenicity involves a
30    mutagenic mode of action mediated by DNA-reactive metabolites. The benzo[a]pyrene database
31    provides strong and consistent evidence for BPDE-induced mutations associated with tumor
32    initiation in cancer tissue from humans exposed to complex mixtures containing benzo [a] pyrene, in
33    animals exposed to benzo[a]pyrene, and in in vitro systems. Other processes may contribute to the
34    carcinogenicity of benzo [ajpyrene via the promotion and progression phases of cancer
35    development (e.g., inflammation, cytotoxicity, sustained regenerative cell proliferation, anti-
36    apoptotic signaling), but the available evidence best supports a mutagenic mode of action as the
37    primary mode by which benzo[a]pyrene acts.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                1-86            DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene
 1    d) Strong Evidence that the Key Precursor Events are Anticipated to Occur in Humans
 2           Mutations in p53 and ras oncogenes have been observed in tumors from mice exposed to
 3    benzo[a]pyrene in the diet [Gulp etal.. 2000] or by i.p. injection [Nesnowetal.. 1998a: Nesnow et
 4    al.. 1998b: Nesnowetal.. 1996.1995: Mass etal.. 1993). Mutations in these same genes have also
 5    been reported in lung tumors of human cancer patients, bearing distinctive mutation spectra (G—>T
 6    transversions) that correlate with exposures to coal smoke [Keohavongetal.. 2003: DeMarini etal..
 7    2001] or tobacco smoke [Pfeifer and Hainaut. 2003: Pfeifer etal.. 2002: Hainaut and Pfeifer. 2001:
 8    Bennett etal.. 19991.

 9           Table 1-18. Supporting evidence for the carcinogenic to humans cancer
10           descriptor for benzo[a]pyrene
                        Evidence
                   Reference
     a) Strong human evidence of cancer or its precursors
         •   Increased risk of lung, bladder, and skin cancer
             in humans exposed to complex PAH mixtures
             containing benzo[a]pyrene


         •   Benzo[a]pyrene-specific biomarkers detected in
             humans exposed to PAH mixtures associated
             with increased risk of cancer

                BPDE-DNA adducts in WBCs of coke oven
                workers and chimney sweeps

                BPDE-DNA adducts in smokers


         •   Benzo[a]pyrene-specific DNA adducts have
             been detected in target tissues in humans
             exposed to PAH mixtures

                BPDE-DNA adducts in non-tumor lung
                tissues of cigarette smokers with lung
                cancer and in skin of eczema patients
                treated with coal tar

                BPDE-DNA adduct formation in p53 in
                human cells in vitro corresponds to
                mutational hotspots at guanine residues in
                human lung tumors


         •   Benzo[a]pyrene-specific mutational spectra
             identified in PAH-associated tumors in humans
(IARC (2010), 2004)); Secretan et al. (2009);Baan et al.
(2009); Benbrahim-Tallaa etal. (2012)
(Rojasetal. (2000); Bartsch et al. (1999); Pavanello et al.
(1999); Bartsch et al. (1998); Rojas et al. (1998))

Phillips (2002)
Rojasetal. (2004); (Godschalk et al. (2002); Bartsch et al.
(1999); Godschalk et al. (1998b); Rojas et al. (1998);
Andreassen et al. (1996); Alexandrov et al. (1992))
(Denissenko et al. (1996); Puisieux et al. (1991))
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                   1-87            DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
Evidence

- GC->TA transversions and GC->AT
transitions at hprt locus in T-lymphocytes
of humans with lung cancer
— G-^T transversions in exposed human-p53
knock-in mouse fibroblasts at the same
mutational hotspot in p53 from smoking-
related lung tumors in humans
- G->T transversions at the same mutational
hotspot in p53 and K-ras in human lung
tumors associated with smoky coal
exposures
- Increased percentage of G->T
transversions in p53 in smokers versus
nonsmokers
b) Extensive animal evidence
Oral exposures
• Forestomach tumors in male and female rats
and in female mice following lifetime exposure

• Forestomach tumors in mice following less-
than-lifetime exposures




• Alimentary tract and liver tumors in male and
female rats following lifetime exposure
• Kidney tumors in male rats following lifetime
exposure
• Auditory canal tumors in male and female rats
following lifetime exposure
• Esophageal, tongue, and laryngeal tumors in
female mice following lifetime exposure
• Lung tumors in mice following less-than-
lifetime exposure
Reference
Hackman et al. (2000)
Liu etal. (2005)
(Keohavong et al. (2003); DeMarini et al. (2001))
(Pfeifer and Hainaut (2003); Pfeifer et al. (2002); Hainaut
and Pfeifer (2001); Bennett etal. (1999))



(Kroese et al. (2001); Beland and Gulp (1998); Gulp et al.
(1998); Brune etal. (1981))
(Weyand et al. (1995); Benjamin et al. (1988); Robinson et
al. (1987); EI-Bayoumy (1985); Triolo et al. (1977);
Wattenberg (1974); Roe et al. (1970); Biancifiori et al.
(1967); Chouroulinkov et al. (1967); Fedorenko and
Yansheva (1967); Neal and Rigdon (1967); Berenblum and
Haran (1955))
Kroese et al. (2001)
Kroese et al. (2001)
Kroese et al. (2001)
(Beland and Gulp (1998); Gulp et al. (1998))
(Weyand et al. (1995); Robinson et al. (1987); Wattenberg
(1974))
This document is a draft for review purposes only and does not constitute Agency policy,
                                1-88            DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
Evidence
Reference
Inhalation exposures
• Upper respiratory tract tumors in male
hamsters following chronic exposure
Thyssenetal. (1981)

Dermal exposures
• Skin tumors in mice following lifetime
exposures without a promoter
• Skin tumors in rats, rabbits, and guinea pigs
following subchronic exposures
(Sivak et al. (1997); Grimmer et al. (1984); Habs et al.
(1984); Grimmer et al. (1983); Habs et al. (1980); Schmahl
et al. (1977); Schmidt et al. (1973); Roe et al. (1970); Poel
(1963), 1959))
(IPCS, 1998; ATSDR, 1995; IARC, 1983, 1973)

Other routes of exposure
• Respiratory tract tumors in hamsters following
intratracheal instillation
• Liver or lung tumors in newborn mice given
acute postnatal i.p. injections
• Lung tumor multiplicity in A/J adult mice given
single i.p. injections
(Feron and Kruysse, 1978; Ketkar et al., 1978; Feron et al.,
1973; Henry et al., 1973; Saffiotti et al., 1972)
(Lavoie et al., 1994; Busby et al., 1989; Weyand and
Lavoie, 1988; Lavoie et al., 1987; Wislocki et al., 1986;
Busby et al., 1984; Buening et al., 1978; Kapitulnik et al..
1978)
Mass et al. (1993)

c) Identification of key precursor events have been identified in animals
• Bioactivation of benzo[a]pyrene to DNA-
reactive metabolites has been shown to occur
in multiple species and tissues by all routes of
exposure
• Direct DNA damage by the reactive
metabolites, including the formation of DNA
adducts and ROS-mediated damage
• Formation and fixation of DNA mutations,
particularly in tumor suppressor genes or
oncogenes associated with tumor initiation
See 'Experimental Support for Hypothesized Mode of
Action' section
d) Strong evidence that the key precursor events are anticipated to occur in humans
• Mutations in p53 or ras oncogenes have been
observed in forestomach or lung tumors from
mice exposed to benzo[a]pyrene
(Gulp et al. (2000); Nesnow et al. (1998a); Nesnow et al.
(1998b); Nesnow et al. (1996), 1995); Mass et al. (1993))

This document is a draft for review purposes only and does not constitute Agency policy,
                                1-89            DRAFT—DO NOT CITE OR QUOTE

-------
                                          Toxicological Review ofBenzo[a]pyrene
Evidence

- G->T transversions in ras oncogenes or the
p53 gene have been observed in lung
tumors of human cancer patients exposed
to coal smoke
- Higher frequency of G->T transversions in
lung tumors from smokers versus
nonsmokers
Reference
(Keohavong et al. (2003); DeMarini et al. (2001))
(Pfeifer and Hainaut (2003); Pfeifer et al. (2002); Hainaut
and Pfeifer (2001); Bennett et al. (1999))

This document is a draft for review purposes only and does not constitute Agency policy,
                                1-90            DRAFT—DO NOT CITE OR QUOTE

-------
                                                       Toxicological Review ofBenzo[a]pyrene
 1
 2    2. DOSE-RESPONSE  ANALYSIS
 3   2.1.  ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER
 4          The oral reference dose (RfD) (expressed in units of mg/kg-day) is defined as an estimate
 5   (with uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
 6   population (including sensitive subgroups) that is likely to be without an appreciable risk of
 7   deleterious effects during a lifetime. It can be derived from a no-observed-adverse-effect level
 8   (NOAEL), lowest-observed-adverse-effect level (LOAEL), or the 95% lower bound on the
 9   benchmark dose (BMDL), with uncertainty factors (UFs) generally applied to reflect limitations of
10   the data used.

11   2.1.1.  Identification of Studies and Effects for Dose-Response Analysis
12          In Section 1.2.1, developmental, reproductive, and immunological toxicities were
13   highlighted as human hazards or potential human hazards of benzo[a]pyrene exposure by the oral
14   route. Studies within each effect category were evaluated using general study quality
15   characteristics (as discussed in Section 6 of the Preamble) to help inform the selection of studies
16   from which to derive toxicity values. Rationales for selecting the studies and effects to represent
17   each of these hazards are summarized below.
18          Human studies are preferred over animal studies when quantitative measures of exposure
19   are reported and the reported effects are determined to be associated with exposure. For
20   benzo[a]pyrene, human studies of environmental polycyclic aromatic hydrocarbon (PAH) mixtures
21   across multiple cohorts have observed effects following exposure to complex mixtures of PAHs.
22   The available data suggest that benzo[a]pyrene exposure may pose health hazards other than
23   cancer including reproductive and developmental effects such as infertility, miscarriage, and
24   reduced birth weight (Wuetal.. 2010: Neal etal.. 2008: Tang etal.. 2008: PereraetaL 2005b:
25   PereraetaL, 2005a), effects on the developing nervous system (PereraetaL, 2012a: PereraetaL,
26   2009), and cardiovascular effects (Friesenetal., 2010: Burstyn etal., 2005). However, the available
27   human studies that utilized benzo[a]pyrene-deoxyribonucleic acid (DNA) adducts as the exposure
28   metric do not provide external exposure levels of benzo[a]pyrene from which to derive a value, and
29   exposure is likely to have occurred by multiple routes. In addition, uncertainty exists due to
30   concurrent exposure to other PAHs and other components of the mixture (such as metals).
31          Animal studies were evaluated to determine which provided the most relevant routes and
32   durations of exposure; multiple exposure levels to provide information about the shape of the dose-
33   response curve; and power to  detect effects at low exposure levels (U.S. EPA. 2002). The oral
34   database for benzo[a]pyrene includes a variety of studies and datasets that are suitable for use in

               This document is a draft for review purposes only and does not constitute Agency policy.
                                               2-1             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    deriving reference values.  Specifically, chronic effects associated with benzo[a]pyrene exposure in
 2    animals include observations of organ weight and histological changes and hematological
 3    parameters observed in several oral cancer bioassays [Kroese etal.. 2001: Beland and Gulp. 1998).
 4    Multiple subchronic studies are available that characterize a variety of effects other than cancer. In
 5    addition, several developmental studies are available that help inform hazards of exposure during
 6    sensitive developmental windows.

 7    Developmental Toxicity
 8          Numerous animal studies observed endpoints of developmental toxicity following oral
 9    exposure during gestational or early postnatal development [Chen etal., 2012: Tules etal., 2012:
10    Shengetal.. 2010: BouayedetaL 2009a: Kristensen etal.. 1995: Mackenzie and Angevine. 19811
11    and were considered for dose-response analysis based on the above criteria. Kristensen et al.
12    [1995]. with only one dose group, was not considered further given its concordance with
13    Mackenzie and Angevine [1981]. which had multiple groups. From the remaining studies
14    demonstrating developmental toxicity, the studies conducted by Chen etal. [2012] and Jules et al.
15    [2012] were identified as the most informative studies for dose-response analysis. The
16    neurodevelopmental study by Chen etal. [2012] was a well-designed and well-conducted study
17    that evaluated multiple developmental endpoints and measures of neurotoxicity in neonatal,
18    adolescent, and adult rats after early postnatal exposure. The study randomly assigned a total of
19    10 male and 10 female pups per treatment group, with no more than one male and one female from
20    each litter for behavioral testing.  In addition, the pups were cross-fostered with dams being rotated
21    among litters every 2-3 days to distribute any maternal caretaking differences randomly across
22    litters and treatment groups. Importantly, all tests were conducted by investigators blinded to
23    treatment, and test order was randomized each day.
24          In the neurobehavioral tests, Chen etal. [2012] observed increased locomotion in the open
25    field test, increased latency in negative geotaxis and surface righting tests, decreased anxiety-like
26    behaviors in the elevated plus maze test, and impaired performance in the Morris water maze test
27    at various time points following neonatal benzo[a]pyrene treatment  Altered behaviors and
28    locomotion in open field tests could be attributed to anxiety responses due to open spaces and
29    bright light, as well as changes to  motor system function. Chen etal. [2012] reported increased
30    quadrants crossed, which could indicate either increased motor activity or decreased anxiety (less
31    fear of the open spaces/bright lights]; thus, this response does not reflect a clear effect of exposure
32    on a discrete neurological function.  In addition, the biological significance of the impairments
33    observed in the negative geotaxis and surface righting tests were considered somewhat uncertain
34    as these effects were not persistent as compared to the effects noted in the elevated plus maze and
35    Morris water maze (which persisted into adulthood]. As a result, EPA considered the elevated plus
36    maze and Morris water maze tests to be the most informative measures of neurobehavioral
37    function performed by Chen  etal. [2012].
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-2             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1           Significant, dose-related effects were reported in an established test of spatial learning and
 2    memory (Morris water maze). Specifically, increased escape latency in hidden platform trials and
 3    decreased time spent in the target quadrant during a probe trial were observed following
 4    benzo[a]pyrene exposure in rats tested as adolescents or as adults. Due to the altered baseline
 5    performance of treated animals on day 1 of the hidden platform trials these findings cannot be
 6    specifically attributed to impaired learning.  In fact, the slopes of the lines across trial days are
 7    nearly identical for the treatment groups, suggesting the lack of a robust effect on learning. The
 8    impaired Morris water maze performance of treated animals could be due to effects on several
 9    other components of neurological function besides learning, including anxiety, vision, and
10    locomotion. Similarly, as escape latencies were not comparable across groups after learning
11    acquisition (i.e., the end of the hidden platform trials), differences in probe trial  performance are
12    difficult to attribute to impaired memory retention alone. As a result of this lack of specificity,
13    although they identify significant effects  of benzo[a]pyrene exposure, the Morris water maze data
14    were considered less informative than the results from the elevated plus maze test  Chen et al.
15    [2012) reported an increase in the number of open arm entries in the elevated plus maze test, an
16    indicator of decreased anxiety-like behavior. These results indicate effects on a single, discrete
17    neurological function that are unlikely to be complicated by changes in other processes such as
18    motor activity (total activity, calculated by summing open and closed arm entries was unchanged
19    with treatment). This neurobehavioral endpoint is supported by similar observations in developing
20    [Bouayed et al.. 2009a) and adult [Grovaetal.. 2008) mice, and may be indirectly related to
21    observations of increased aggression in mice [Bouayed etal.. 2009b) and is considered adverse (see
22    discussion in Section 1.1.1).
23           Tules etal. (2012) was also identified for dose-response analysis. This study was of
24    sufficient duration, utilized multiple doses, did not observe maternal toxicity,  and evaluated
25    multiple cardiovascular endpoints. The study  authors reported increases in both systolic
26    (approximately 20-50%) and diastolic (approximately 33-83%) pressure and heart rate in adult
27    rats that were exposed gestationally to benzo[a]pyrene. A limitation of this study is that the
28    authors only reported effects at the two highest doses.  However, given the magnitude of the
29    response and the appearance of these effects in adulthood following gestational exposure, these
30    endpoints were selected for dose-response analysis because of their sensitivity and biological
31    plausibility.
32           Bouayed et al. (2009a) and Mackenzie  and Angevine (1981) were not  selected for dose-
33    response analysis. Bouayed etal. (2009a) used the same tests as Chen etal. (2012). but the doses
34    evaluated were higher (2 and 20 mg/kg-day compared to 0.02, 0.2, and 2 mg/kg-day, respectively).
35    Similarly, Mackenzie and Angevine (1981) demonstrated developmental effects in a multi-dose
36    study with relevant routes and durations of exposure; however, the doses studied
37    (10-160 mg/kg-day) were much higher than those evaluated in other developmental toxicity
38    studies (Chenetal.. 2012: Tules etal.. 2012).
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-3             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    Reproductive Toxicity
 2           Male reproductive toxicity was demonstrated in numerous subchronic studies [Chenetal..
 3    2011: Chung etal.. 2011: Mohamedetal.. 2010: Zheng etal.. 2010). Chung etal. [2011] was not
 4    included in the dose-response analysis because numerical data were not reported or were only
 5    reported for the mid-dose of three doses. Chenetal. [2011] is a subchronic study that applied only
 6    a single dose level. This study corroborated other available multi-dose studies and is considered
 7    supportive, but was not considered for dose-response analysis due to the limited reporting of
 8    numerical data. The studies conducted by Mohamedetal. [2010] and Zheng etal. [2010] were
 9    identified as the most informative male reproductive toxicity studies for dose-response analysis.
10    Decreased sperm count and motility observed by Mohamedetal. [2010] and decreased
11    intratesticular testosterone levels observed by Zheng etal. [2010] were selected for dose-response
12    analysis as both represent sensitive endpoints of male reproductive toxicity and are indicators of
13    potentially decreased fertility. These effects are also consistent with human studies in PAH
14    exposed populations, as effects on male fertility and semen quality have been demonstrated in
15    epidemiological studies of smokers [reviewed by Spares and Melo, 2008].
16           Female reproductive toxicity was demonstrated in two subchronic studies [Gao etal., 2011:
17    Xu etal.. 2010]. Specifically, Xu etal.. 2010 demonstrated altered ovary weights and follicle
18    numbers, and Gao etal. [2011] demonstrated cervical epithelial cell hyperplasia following oral
19    exposure to benzo[a]pyrene. These studies were identified as the most informative studies on
20    female reproductive toxicity for dose-response analysis.  Gao etal. [2011] identified statistically-
21    significant, dose-related increases in the incidence of cervical inflammatory cells in mice exposed to
22    low doses of benzo[a]pyrene for 98 days [Gao etal., 2011]. Cervical effects of increasing severity
23    (including epithelial hyperplasia, atypical hyperplasia, apoptosis, and necrosis] were also observed
24    at higher doses [Gao etal.. 2011].  There are no data on cervical effects in other species or in other
25    mouse strains. However, Gao etal. [2011] also evaluated cervical effects in separate groups of mice
26    exposed via intraperitoneal [i.p.]  injection, and observed similar responses in these groups of mice,
27    providing support for the association between effects in this target organ and benzo[a]pyrene
28    exposure. Epidemiological studies have demonstrated an association between cigarette smoking
29    and increased risk of cervical cancer [Pate Capps etal., 2009].  In addition, benzo[a]pyrene
30    metabolites and benzo[a]pyrene-DNA adducts have been detected in human cervical mucus and
31    cervical tissues obtained from smokers [Phillips. 2002: Melikian et al.. 1999]. However, data to
32    support that cervical hyperplasia following oral benzo[a]pyrene exposure progresses to cervical
33    tumors were not available (no cervical tumors were noted in the two available lifetime oral cancer
34    bioassays].  Thus, in the absence of these data, cervical hyperplasia is presented as a noncancer
35    effect.
36           Xu etal. [2010] identified biologically and statistically significant decreases in ovary weight,
37    estrogen, and primordial follicles, and altered estrus cycling in treated animals. These reductions in
38    female reproductive parameters are supported by a large database of animal studies indicating that

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-4             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    benzo[a]pyrene is ovotoxic with effects including decreased ovary weight, decreased primordial
 2    follicles, and reduced fertility fBorman etal.. 2000: KristensenetaL 1995: Miller etal.. 1992:
 3    Swartz and Mattison. 1985: Mackenzie and Angevine. 1981: Mattison et al.. 1980). Additionally,
 4    epidemiology studies indicate that exposure to complex mixtures of PAHs, such as through cigarette
 5    smoke, is associated with measures of decreased fertility in humans [Neal etal.. 2008: El-Nemr et
 6    al.. 1998). Specific associations have also been made between infertility and increased levels of
 7    benzo[a]pyrene in follicular fluid in women undergoing in vitro fertilization [Neal etal.. 2008].

 8    Immunotoxicity
 9          As described in Section 1.1.3, the immune system was identified as a potential human
10    hazard of benzo[a]pyrene exposure based on findings of organ weight and immunoglobulin
11    alterations, as well as effects on cellularity and functional changes in the immune system in animals.
12    The only available studies to support development of an RfD were conducted by Kroese etal.
13    [2001] and De Jong etal. [1999]. These are subchronic studies with multiple exposure levels and
14    adequate power to detect effects. In comparing these studies, the Kroese etal. [2001] study is
15    preferred for dose-response analysis due to its longer duration (90 days].
16          Decreasedthymus weight, observed in Kroese etal. [2001], decreased IgM and IgAlevels,
17    and decreased relative numbers of B-cells, observed in De long et al. [1999]. were  selected for dose-
18    response analysis. It is recognized that thymus weight changes on their own have been noted to be
19    less reliable indicators of immunotoxicity [Luster etal.. 1992].  However, there are converging lines
20    of evidence that support the derivation of an organ/system-specific RfD for benzo[a]pyrene
21    immunotoxicity. Alterations in immunoglobulin levels have been noted in humans after exposure
22    to PAHs, as well as in animal studies after exposure to benzo[a]pyrene. Changes in B cell
23    populations in the spleen provide additional evidence of immunotoxicity.  Finally, functional effects
24    on the immune system, including dose-related decreases in SRBC-specific IgM levels and dose-
25    dependent decreases in resistance to pneumonia or Herpes simplex type 2 following short-term
26    subcutaneous [s.c.] injection have been reported [Temple etal.. 1993: Munsonetal.. 1985]. The
27    observed decreases in thymus weight, IgM and IgA levels, and number of B cells associated with
28    exposure to benzo[a]pyrene were concluded to be representative of immunotoxicity following
29    benzo[a]pyrene exposure and were selected for dose-response analysis.

30    2.1.2. Methods of Analysis
31          No biologically based dose-response models are available for benzo[a]pyrene. In this
32    situation, EPA evaluates a range of dose-response models thought to be consistent with underlying
33    biological processes to determine how best to empirically model the dose-response relationship in
34    the range of the observed data. Consistent with this approach, all models available in EPA's
35    Benchmark Dose Software [BMDS] were evaluated. Consistent with EPA's Benchmark Dose
36    Technical Guidance Document [U.S. EPA, 2012c],  the benchmark dose [BMD] and the 95% lower
37    confidence limit on the BMD [BMDL] were estimated using a benchmark response [BMR] of

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-5             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    1 standard deviation (SD) from the control mean for continuous data or a BMR of 10% extra risk for
 2    dichotomous data in the absence of information regarding what level of change is considered
 3    biologically significant, and also to facilitate a consistent basis of comparison across endpoints,
 4    studies, and assessments. The estimated BMDLs were used as points of departure (PODs). Further
 5    details including the modeling output and graphical results for the best fit model for each endpoint
 6    can be found in Appendix E of the Supplemental Information.
 7          Among the endpoints identified as representative of the hazards of benzo[a]pyrene
 8    exposure, the data for neurobehavioral changes in the elevated plus maze and Morris water maze
 9    tests [Chenetal., 2012], decreased ovary weight [Xu etal., 2010], increased cervical hyperplasia
10    [Gao etal. [2011], and decreasedthymus weight [Kroese etal., 2001] were amenable to dose-
11    response modeling.  Although the data for Morris water maze performance [Chenetal.. 2012] was
12    ultimately considered to be less informative than the elevated plus maze data (see Section 2.1.1],
13    EPA performed dose-response modeling on this endpoint to ensure that the elevated plus maze
14    data were an accurate  representation of other sensitive, behavioral changes in this study. See
15    Appendix E of the Supplemental Information for details of statistical analyses.
16          The data for the remaining endpoints identified in Section 2.1.1 were not modeled.
17    Specifically, the data for cardiovascular effects observed in Tules etal. [2012] were limited due to
18    the reporting of results at only the two highest dose groups. The data for epididymal sperm counts
19    presented in the  Mohamedetal. [2010] study were reported graphically only and requests for the
20    raw data were unsuccessful. The observed decrease in IgM and IgA [De long etal.. 1999]  was
21    inconsistent and not amenable to dose-response modeling. NOAELs or LOAELs were used as the
22    POD for these endpoints.
23          Human equivalent doses [HEDs] for oral exposures were derived from the PODs estimated
24    from the laboratory animal data as described in EPA's Recommended Use of Body Weight3/4 as the
25    Default Method in Derivation of the Oral Reference Dose [U.S. EPA. 2011]. In this guidance, EPA
26    advocates a hierarchy of approaches for deriving HEDs from data in laboratory animals, with the
27    preferred approach  being physiologically-based toxicokinetic modeling. Other approaches can
28    include using chemical-specific information in the absence of a complete physiologically-based
29    toxicokinetic model. As discussed in Appendix D of the Supplemental Information, several animal
30    physiologically based pharmacokinetic [PBPK] models for benzo[a]pyrene have been developed
31    and published, but a validated human PBPK model for benzo[a]pyrene for extrapolating doses from
32    animals to humans is not available. In lieu of either chemical-specific models or data to inform the
33    derivation of human equivalent oral exposures, a body weight scaling to the % power (i.e., B W3/4]
34    approach is applied  to extrapolate toxicologically equivalent doses of orally administered agents
35    from adult laboratory animals to adult humans for the purpose of deriving an oral RfD.  BW3/4
36    scaling was not employed for deriving HEDs from studies in which doses were administered
37    directly to early postnatal animals because of the absence of information on whether allometric
38    (i.e., body weight] scaling holds when extrapolating doses from neonatal animals to  adult humans
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-6            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
due to presumed toxicokinetic and/or toxicodynamic differences between lifestages [U.S. EPA,
2011: HattisetaL 20041
       Consistent with EPA guidance fU.S. EPA. 20111. the points of departure (PODs) estimated
based on effects in adult animals are converted to HEDs employing a standard dosimetric
adjustment factor (DAF) derived as follows:
                    DAF =
       where
                    BWa = animal body weight; and
                    BWh = human body weight.

       Using BWa of 0.25 kg for rats and 0.035 kg for mice and BWh of 70 kg for humans [U.S. EPA.
1988). the resulting DAFs for rats and mice are 0.24 and 0.15, respectively. Applying this DAF to
the POD identified for effects in adult rats or mice yields a PODnED as follows (see Table 2-1):

             PODHED = Laboratory animal dose (mg/kg-day) x DAF.

       Table 2-1 summarizes the sequence of calculations leading to the derivation of a human-
equivalent POD for each data set discussed above.

       Table 2-1. Summary of derivation of PODs
Endpoint and
reference
Species/
sex
Model3
BMR
BMD
(mg/kg-d)
BMDL
(mg/kg-d)
PODADJb
(mg/kg-d)
PODHEDC
(mg/kg-d)
Developmental
Neurobehavioral
changes
Chen etal. (2012)

Cardiovascular
effects
Jules etal. (2012)

Female
Sprague-
Dawley rats
Long-Evans
rats
Exponential
(M4)a
ISO
0.18
0.09
LOAEL(0.6mg/kg-d)
(15% T* in systolic blood pressure; 33% T* in
diastolic blood pressure)
0.09
0.6
0.09
0.15
Reproductive
Decreased ovary
weight
Xu etal. (2010)

Decreased
intratesticular
testosterone
Zheng etal. (2010)
Female
Sprague-
Dawley rats
Male
Sprague-
Dawley rats
Linear3
ISO
2.3
1.5
NOAEL (1 mg/kg-d)
(15% -^ in testosterone)
1.5
1
0.37
0.24
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                2-7             DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene
Endpointand
reference
Decreased sperm
count and motility
Mohamed et al.
(2010)
Cervical epithelial
hyperplasia
Gaoetal. (2011)

Species/
sex
Male C57BL/6
mice
Female ICR
mice
Model3
BMR
BMD
(mg/kg-d)
BMDL
(mg/kg-d)
LOAEL(lmg/kg-d)
(50% -^ in sperm count; 20% 4, in sperm motility)
Log-
logistic3
10%
0.58
0.37
PODADJb
(mg/kg-d)
1
0.37
PODHEDC
(mg/kg-d)
0.15
0.06
Immunological
Decreased thymus
weight
Kroese et al. (2001)

Decreased IgM levels
De Jong etal. (1999)

Decreased IgA levels
De Jong etal. (1999)

Decreased number
of B cells
De Jong etal. (1999)
Female
Wistar rats
Male
Wistar rats
Male
Wistar rats
Male
Wistar rats
Linear3
ISO
10.5
7.6
NOAEL(10mg/kg-d)
(14% 4, in IgM)
NOAEL(30mg/kg-d)
(28% 4, in IgA)
NOAEL(30mg/kg-d)
(7% -tin B cells at NOAEL; 31% ^ at LOAEL)
7.6
7.1
21
21
1.9
1.7
5.2
5.2
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11

12

13
14
15
16
17
18
19
20
21
22
23
3For modeling details, see Appendix E in Supplemental Information.
bFor studies in which animals were not dosed daily, administered doses were adjusted to calculate the TWA daily
 doses prior to BMD modeling.
CHED PODs were calculated using BW3/4scaling (U.S. EPA, 2011) for effects from dosing studies in adult animals
 (i.e., Gao etal., 2011; Mohamed etal., 2010; Xu et al., 2010; De Jong etal., 1999) or for developmental effects
 resulting from in utero exposures.  BW3/4 scaling was not employed for deriving HEDs from studies in which doses
 were administered directly to early postnatal animals (i.e., Chen et al., 2012) because of the absence of
 information on whether allometric (i.e., body weight) scaling holds when extrapolating doses from neonatal
 animals to adult humans due to presumed toxicokinetic and/or toxicodynamic differences between lifestages
 (U.S. EPA, 2011; Hattis et al., 2004).

2.1.3.  Derivation of Candidate Values

       Under EPA's A Review of the Reference Dose and Reference Concentration Processes [U.S. EPA.
2002: Section 4.4.5). also described in the Preamble, five possible areas of uncertainty and
variability were considered. An explanation of the five possible areas of uncertainty and variability
follows:
       An intraspecies uncertainty factor, UFn, of 10 was applied to account for variability and
uncertainty in toxicokinetic and toxicodynamic susceptibility within the subgroup of the human
population most sensitive to the health hazards of benzo[a]pyrene [U.S. EPA. 2002). In the case of
benzo[a]pyrene, the PODs were derived from studies in inbred animal strains  and are not
considered sufficiently representative of the exposure and dose-response of the most susceptible
human subpopulations (in this case, the developing fetus). In certain cases, the toxicokinetic
component of this factor may be replaced when a PBPK model is available that incorporates the

           This document is a draft for review purposes only and does not constitute Agency policy.
                                             2-8              DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    best available information on variability in toxicokinetic disposition in the human population
 2    (including sensitive subgroups). In the case of benzo[a]pyrene, insufficient information is available
 3    to quantitatively estimate variability in human susceptibility; therefore, the full value for the
 4    intraspecies UF was retained.
 5          An interspecies uncertainty factor, UFA, of 3 (101/2 = 3.16, rounded to 3) was applied, to all
 6    PODs in Table 2-2 except Chen et al. [2012]. because BW3/4 scaling is being used to extrapolate oral
 7    doses from laboratory animals to humans. Although BW3/4 scaling addresses some aspects of cross-
 8    species extrapolation of toxicokinetic and toxicodynamic processes, some residual uncertainty
 9    remains. In the absence of chemical-specific data to quantify this uncertainty, EPA's BW3/4 guidance
10    [U.S. EPA, 2011] recommends use of a UF of 3. BW3/4 scaling was not employed for deriving HEDs
11    from studies in which doses were administered directly to early postnatal animals [i.e. Chenetal..
12    2012] because of the absence of information on whether allometric (i.e., body weight] scaling holds
13    when extrapolating doses from neonatal animals to adult humans due to presumed toxicokinetic
14    and/or toxicodynamic differences between lifestages (U.S. EPA. 2011: Hattis etal.. 2004].  In this
15    case, a value of 10 was applied because of the  absence of quantitative information to characterize
16    either the toxicokinetic or toxicodynamic differences between animals and humans at this lifestage.
17    A subchronic to chronic uncertainty factor, UFS, of 1 was applied when dosing occurred during
18    gestation (Tules etal.. 2012] or the early postnatal period (Chenetal.. 2012] that is relevant to
19    developmental effects.  The developmental period is recognized as a susceptible lifestage and
20    repeated exposure is not necessary for the manifestation of developmental toxicity (U.S. EPA.
21    1991c]. A value of 10 was applied when the POD was based on a subchronic study (studies in
22    Table 2-2, other than the two developmental toxicity studies, were 42-90 days in duration] to
23    account for the possibility that longer exposure may induce effects at a lower dose.
24          A UF for extrapolation from a LOAEL to NOAEL, UFL, of 1 was applied when the POD was
25    based on a NOAEL (Zheng etal.. 2010: De long etal.. 1999]. A value of 1 was applied for LOAEL-to-
26    NOAEL extrapolation when a BMR of a 1 SD (Chenetal.. 2012: Kroese etal.. 2001] or  10% change
27    (Gao etal.. 2011] from the control was selected under an assumption that it represents a minimal
28    biologically significant response level. A NOAEL was not determined for the most sensitive effects
29    observed in Jules et al. (2012] and Mohamed etal. (2010]. At the LOAEL. Jules etal. (2012]
30    observed statistically significant increases in systolic (15%] and diastolic (33%] blood pressure
31    when measured in adulthood following gestational exposure. Regarding the study by Mohamed et
32    al. (2010]. the authors observed a statistically significant decrease sperm count (50%] and motility
33    (20%] in treated FO males at the LOAEL and the observed decrements in sperm count persisted in
34    untreated Fl male offspring. The data reported in these studies were not amenable to dose-
35    response modeling, which would have allowed for extrapolation to a minimally biologically
36    significant response level. Therefore, a full UF of 10 was applied to approximate a NOAEL for these
37    studies, which observed a high magnitude of response at the LOAEL. A database uncertainty factor,
38    UFo, of 3 was applied to account for database deficiencies, including the lack of a standard
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-9             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
 1    multigenerational study or extended 1-generation study that includes exposure from premating
 2    through lactation, considering that benzo [a]pyrene has been shown to affect fertility in adult male
 3    and female animals by multiple routes of exposure (see Section 1.1.2). Considering that decreased
 4    fertility in adult male and female mice is observed following gestational exposure, it is assumed that
 5    exposure occurring over this more comprehensive period of development could result in a lower
 6    POD. Also, the lack of a study examining functional neurological endpoints following a more
 7    comprehensive period of developmental exposure (i.e., gestation through lactation) is a data gap,
 8    considering human and animal evidence indicating altered neurological development (see
 9    Section 1.1.1).
10          Table 2-2 is a continuation of Table 2-1 and summarizes the application of UFs to each POD
11    to derive a candidate value for each data set. The candidate values presented in the table below are
12    preliminary to the derivation of the organ/system-specific reference values. These candidate
13    values are considered individually in the selection of a representative oral reference value for a
14    specific hazard and subsequent overall RfD for benzo[a]pyrene.

15          Table 2-2. Effects and corresponding derivation of candidate values


Endpoint and reference

PODHEDa
(mg/kg-d)

POD
type


UFA


UFH


UFL


UFS


UFD

Composite
UF
Candidate
value
(mg/kg-d)
Developmental
Neurobehavioral changes in
rats
Chen etal. (2012)

Cardiovascular effects in rats
Jules etal. (2012)

0.09



0.15


BMDL1SD



LOAEL


10



3


10



10


1



10


1



1


3



3


300



1,000


3 x ID'4



2 x 10"4


Reproductive
Decreased ovary weight in
rats
Xu etal. (2010)

Decreased intratesticular
testosterone in rats
Zheng etal. (2010)

Decreased sperm count and
motility in mice
Mohamed et al. (2010)

Cervical epithelial
hyperplasia in mice
Gao etal. (2011)
0.37



0.24



0.15



0.06


BMDL1SD



NOAEL



LOAEL



BMDL10


3



3



3



3


10



10



10



10


1



1



10



1


10



10



10



10


3



3



3



3


1,000



1,000



10,000



1,000


4 x 10"4



2 x ID'4



Not calculated
duetoUF
>3,000a

6 x 10"5


                This document is a draft for review purposes only and does not constitute Agency policy,
                                                2-10            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


Endpoint and reference

PODHEDa
(mg/kg-d)

POD
type


UFA


UFH


UFL


UFS


UFD

Composite
UF
Candidate
value
(mg/kg-d)
Immunological
Decreased thymus weight in
rats
Kroese et al. (2001)

Decreased serum IgM in rats
De Jong etal. (1999)

Decreased serum IgA in rats
De Jong etal. (1999)

Decreased number of B cells
in rats
De Jong etal. (1999)
1.9



1.7


5.2


5.2


BMDL1SD



NOAEL


NOAEL


NOAEL


3



3


3


3


10



10


10


10


1



1


1


1


10



10


10


10


3



3


3


3


1,000



1,000


1,000


1,000


2 x 10"3



2 x 10"3


5 x 10"3


5 x 10"3


1
2
3
4
5
6
7
aAs recommended in EPA's A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA,
 2002), the derivation of a reference value that involves application of the full 10-fold UF in four or more areas of
 extrapolation should be avoided.


       Figure 2-1 presents graphically the candidate values, UFs, and PODs, with each bar

corresponding to one data set described in Tables 2-1 and 2-2.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  2-11             DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene
                Neurodevelopmental
                 alterations in rats
                 (Chen eta!.,2012)

              Cardiovascular effects
             in uts (Julesot al.,2012)
                                                                                  Composite UF

                                                                                 A Candidate value

                                                                                 • POD(HED)
        Q
        O
        Q-
        UJ
        C£.
          4' Ovary weight in rats
            (Xuet al., 2010)

             4-- Intratesticular
            testosterone in rats
            (Zheng eta I. ,2010)
           4- Sperm count and
            motility in mice
         (Mohamedetal., 2010)
            Cervical epithelial
            hyperplasia in mice
             (Gaoetal , 2011)

        • 'is Thymus vvc-ight in rats
          (Kroesectal.,2001)
1

2

3
4
5
6
        ID
           x b< run IgM in rats
           (DeJonget al., 1999)

           \_, Seruri IgA in rats
           (DeJongetal., 1999)
           4- Number of B cells in rats
              (DeJongetal., 1999)
                              0.00001      0.0001      0.001       0.01        0.1
                                                           Doses (mg/kg-d)
                                                                                                 10
       Figure 2-1.  Candidate values with corresponding PODs and composite UFs.

2.1.4.  Derivation of Organ/System-Specific Reference Doses
       Table 2-3 distills the candidate values from Table 2-2 into a single value for each organ or
system. These organ- or system-specific reference values may be useful for subsequent cumulative
risk assessments that consider the combined effect of multiple agents acting at a common site.
                772/s document ;'s a draft for review purposes only and does not constitute Agency policy,
                                                   2-12              DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene
 1
 2
       Table 2-3. Organ/system-specific RfDs and proposed overall RfD for
       benzo[a]pyrene
Effect
Developmental
Reproductive
Immunological
Proposed overall RfD
Basis
Neurobehavioral changes
Decreased ovary weight
Decreased thymus weight and serum IgM
Developmental toxicity
RfD (mg/kg-d)
3 x irj-4
4 x 10"4
2 x 10"3
3 x itr4
Study
exposure
description
Critical
window of
development
(postnatal)
Subchronic
Subchronic
Critical
window of
development
(postnatal)
Confidence
Medium
Medium
Low
Medium
 3
 4
 5
 6
 7
 8

 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Developmental Toxicity
       The candidate value based on neurobehavioral changes in rats [Chenetal.. 2012] was
selected as the organ/system-specific RfD representing developmental toxicity.  This candidate
value was selected because it is associated with the application of the smaller composite UF and
because similar effects were replicated across other studies [Maciel etal.. 2014: Bouayedetal..
2009a: Bouayedetal.. 2009b: Grovaetal.. 20081

Reproductive Toxicity
       Among the adverse reproductive effects associated with oral benzo[a]pyrene exposure,
decrements in sperm parameters, decreases in testosterone, and effects in the ovary were
supported by a large body of evidence. The data supporting cervical effects are limited to a single
study, and were therefore given less weight compared to the other reproductive effects. The
derivation of a candidate value based on decreased sperm count and motility [Mohamedetal..
2010] involved too much uncertainty (see Table 2-2] and the study used to derive a candidate value
based on decreased testosterone [Zheng etal.. 2010] did not observe a dose-response relationship
(a 15% decrease in testosterone was seen at the low and high doses, with statistical significance at
the high dose]. The study by Xu etal. [2010] observed a dose-response relationship for decreased
ovary weight (both doses were statistically significant]. Additionally, statistically significant
decreases in primordial follicles were observed at the high dose, supporting the ovaries as a target
of toxicity. Therefore, the candidate value based on decreased ovary weight in rats from the Xu et
al. (2010] study was selected as the organ/system-specific RfD representing reproductive toxicity.
The ovarian effects are supported by a large database of animal studies and human studies of
exposure to benzo[a]pyrene and PAH mixtures. While evidence in the benzo[a]pyrene database
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                2-13            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    supports a male and female reproductive hazard, there is more confidence in the POD from Xu et al.
 2    [2010] as the basis for an organ/system-specific RfD for reproductive effects.

 3    Immunotoxicity
 4          The candidate values based on decreased thymus weight [Kroese etal.. 2001] and serum
 5    IgM levels in rats [De long et al.. 1999] were selected as the organ/system-specific RfD representing
 6    immunotoxicity. The observed decreases in thymus weight, IgM and IgA levels, and number of
 7    B cells associated with exposure to benzo[a]pyrene were determined to be representative of
 8    immunotoxicity. In combination, these effects provide more robust evidence of immunotoxicity.
 9    The candidate values for decreased thymus weight [Kroese etal., 2001] and serum IgM levels in
10    rats [De long etal.. 1999] were equal and provided the most sensitive POD; thus, these candidate
11    values were selected as the organ/system-specific RfD representing immunotoxicity.

12    2.1.5. Selection of the Proposed Overall Reference Dose
13          Multiple organ/system-specific reference doses were derived for effects identified as
14    human hazards or potential hazards from benzo[a]pyrene including developmental toxicity,
15    reproductive toxicity (representative of effects in both sexes], and immunotoxicity. To estimate an
16    exposure level below which effects from benzo[a]pyrene exposure are not expected to occur, the
17    lowest organ/system-specific RfD (3 x 1Q-4 mg/kg-day] is proposed as the overall RfD for
18    benzo[a]pyrene. This value, based on induction of neurobehavioral changes in rats exposed to
19    benzo[a]pyrene during a susceptible lifestage, is supported by several animal and human studies
20    (see Section 1.1.1].
21          The overall RfD is derived to be protective of all types of effects for a given duration of
22    exposure and is intended to protect the population as a whole including potentially susceptible
23    subgroups (U.S. EPA. 2002]. This value should be applied in general population risk assessments.
24    However, decisions concerning averaging exposures over time for comparison with the RfD should
25    consider the types of toxicological effects and specific lifestages of concern. For example,
26    fluctuations in exposure levels that result in elevated exposures during various lifestages could
27    potentially lead to an appreciable risk, even if average levels over the full exposure duration were
28    less than or equal to the RfD. Alternatively, developmental toxicity may not be a concern due to
29    exposure scenarios in which exposure is occurring outside of the critical window of development

30    2.1.6. Confidence Statement
31          A confidence level of high, medium, or low is assigned to the study used to derive the RfD,
32    the overall database, and the RfD itself, as described in Section 4.3.9.2 of EPA's Methods for
33    Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA.
34    1994].
35          Confidence in the principal study (Chen etal.,  2012] is medium-to-high. The study design
36    included randomized experimental testing, blinded observations, culling of pups to account for

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-14            DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene

 1    nutritional availability, treatment-randomization, and controls for litter and nursing bias. Some
 2    informative experimental details were, however, omitted including the sensitivity of some assays at
 3    the indicated developmental ages, gender-specific data for all outcomes, and individual animal data.
 4    Notably, these study limitations do not apply to the endpoint chosen to derive the RfD, and the
 5    overall methods and reporting are considered sufficient.  Confidence in the database is medium,
 6    primarily due to the lack of a multigenerational reproductive toxicity study given the sensitivity to
 7    benzo[a]pyrene during development. Reflecting medium-to-high confidence in the principal study
 8    and medium confidence in the database, confidence in the RfD is medium.

 9    2.1.7.  Previous IRIS Assessment: Reference Dose
10          An RfD was not derived in the previous IRIS assessment

11    2.2. INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER
12         THAN CANCER
13          The inhalation reference concentration (RfC) (expressed in units of mg/m3) is defined as an
14    estimate (with uncertainty spanning perhaps an order of magnitude) of a continuous inhalation
15    exposure to the human population (including sensitive subgroups) that is likely to be without an
16    appreciable risk of deleterious effects during a lifetime. It can be derived from a NOAEL, LOAEL, or
17    the 95 percent lower bound on the benchmark concentration (BMCL), with UFs generally applied to
18    reflect limitations of the data used.

19    2.2.1.  Identification of Studies and Effects for Dose-Response Analysis
20          In Section 1.2.1, developmental and reproductive toxicities were identified as hazards of
21    benzo[a]pyrene exposure by the inhalation route. Studies within each effect category were
22    evaluated using general study quality characteristics (as discussed in Section 6 of the Preamble) to
23    help inform the selection of studies from which to derive toxicity values. Rationales for selecting
24    the studies and effects to represent each of these hazards are summarized below.
25          Human studies of environmental PAH mixtures across multiple cohorts have observed
26    developmental and reproductive effects following prenatal exposure.  However, these studies are
27    limited by exposure to complex mixtures of PAHs; and, within individual studies, there may have
28    been more than one route of exposure. In addition, the available human studies that utilized
29    benzo[a]pyrene-DNA adducts as the exposure metric do not provide external exposure levels of
30    benzo[a]pyrene from which to derive an RfC.  Although preferred for derivation of reference values,
31    human studies were not considered because of the contribution to the observed hazard of multiple
32    PAHs across multiple routes of exposure and uncertainty due to concurrent exposure to other PAHs
33    and other components of the mixtures (such as metals).
34          Animal studies were evaluated to determine which provided the most relevant routes and
35    durations of exposure, multiple exposure levels to provide information about the shape of the dose
36    response curve, and relative ability to detect effects at low exposure levels.  The only chronic animal
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               2-15            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    inhalation study available for benzo[a]pyrene, Thyssenetal. [1981], was designed as a cancer
 2    bioassay and did not report other effects; however, the inhalation database for benzo[a]pyrene
 3    includes several shorter duration studies that are sufficient for use in deriving reference values
 4    [U.S. EPA. 2002). Specifically, several reproductive toxicity studies are available for the inhalation
 5    route, including one subchronic study Archibong et al. [2008]. Furthermore, several developmental
 6    studies are available that help identify hazards of exposure during sensitive developmental
 7    windows [Wormley etal.. 2004: Archibong et al.. 2002]. In addition, a 4-week inhalation study in
 8    rats is available that investigated, but did not detect, lung injury [Wolff etal.. 1989]. The inhalation
 9    database for benzo[a]pyrene is less extensive than the database of studies by the oral route;
10    however, the types of noncancer effects observed are consistent between routes and are supported
11    by studies in human populations  (see Sections 1.1.1,1.1.2, and 1.1.3].

12    Developmental Toxicity
13           Developmental toxicity, as represented by decreased fetal survival and developmental
14    neurotoxicity, was observed in several inhalation studies [Wormley et al.. 2004: Wuetal.. 2003a:
15    Archibong etal., 2002]. Wu etal. [2003a] was not considered for dose-response analysis due to
16    lack of study details related to number of dams and litters per group and lack of reporting of
17    numerical data. Wormley etal. [2004] was not considered for dose-response analysis, as this study
18    employed only a single exposure  group at which overt toxicity was noted (a 66% reduction in fetal
19    survival].
20           Of the studies demonstrating developmental toxicity, the study conducted by Archibong et
21    al. [2002] was identified as the most informative study for dose-response analysis. Archibong et al.
22    [2002]  observed decreased fetal survival at the lowest dose tested by the inhalation route on
23    CDs 11-20 (i.e., LOAEL of 25 [ig/m3].  This study indicates that the developing fetus is a sensitive
24    target following inhalation exposure to benzo[a]pyrene. The observed decrease in fetal survival is
25    supported by the oral database for benzo[a]pyrene (e.g., decreased survival of litters in mice
26    following in utero exposure to benzo[a]pyrene on CDs 7-16] [Mackenzie and Angevine. 1981].

27    Reproductive Toxicity
28           Reproductive toxicity, as represented by reductions in sperm quality, both count and
29    motility, and testis weights in adults, was observed by Archibong et al. [2008], Rameshetal. [2008]
30    and Archibong et al. f20021 Archibong et al. f20081 and Rameshetal. f20081 reported the results
31    of a single exposure, subchronic inhalation exposure study in male rats. This subchronic study was
32    of sufficient duration and possessed adequate power to detect effects,  but utilized a single exposure
33    concentration, which is less informative for dose-response analysis than a design using multiple
34    exposure concentrations. However, this single-dose subchronic study is consistent with male
35    reproductive effects observed across multiple studies by the oral route and with human studies in
36    PAH exposed populations (see Section 1.1.2]. The endpoints of decreased testes weight and sperm
37    count and motility reported in Archibong et al. [2008] were selected for dose-response analysis as

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-16            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    both represent sensitive endpoints of male reproductive toxicity and are indicators of potentially
 2    decreased fertility. These effects are also consistent with human studies in PAH exposed
 3    populations as effects on male fertility and semen quality have been demonstrated in
 4    epidemiological studies of smokers [Spares and Melo. 2008).

 5    2.2.2.  Methods of Analysis
 6          Data for decreased fetal survival from Archibong et al. [2002] were reported as litter means
 7    and SDs. These data were not amenable to BMD modeling due to the pattern of variability in the
 8    data set, and attempts to obtain the raw data from the study authors were unsuccessful. Therefore,
 9    the LOAEL from this study was used as the POD for dose-response analysis. The study by
10    Archibong et al. [2008]. using only one exposure level, was judged not to support dose-response
11    modeling due to the lack of understanding of the underlying dose-response relationship. LOAELs
12    were also used as the PODs for dose-response analysis.
13          By definition, the RfC is intended to apply to continuous lifetime exposures for humans [U.S.
14    EPA. 1994]. EPA recommends that adjusted continuous exposures be used for inhalation
15    developmental toxicity studies as well as for studies of longer durations [U.S. EPA, 2002]. The
16    LOAELs identified from Archibong et al. [2002] and Archibong et al. [2008] were adjusted to
17    account for the discontinuous daily exposure as follows:
18
19                 PODADj = POD x hours exposed per day/24 hours
20                        = LOAEL x (duration of exposure/24 hours]
21                        = PODAD,
22
23          Next, the human equivalent concentration [HEC] was calculated from the PODAoj by
24    multiplying by a DAF, which, in this case, was the regional deposited dose ratio [RDDRER] for
25    extrarespiratory (i.e., systemic] effects as described in Methods for Derivation of Inhalation
26    Reference Concentrations and Application of Inhalation Dosimetry [U.S. EPA. 1994]. The observed
27    developmental effects are considered systemic in nature (i.e., extrarespiratory] and the normalizing
28    factor for extrarespiratory effects of particles is body weight. The RDDRER was calculated as
29    follows:
30                            BWA   (VE)H   (FTOT)H
31          where:
32                        BW = body weight [kg];
33                        VE = ventilation rate (L/min]; and
34                        FTOT = total fractional deposition.
35
36          The total fractional deposition includes particle deposition in the nasal-pharyngeal,
37    tracheobronchial, and pulmonary regions. FTOT for both animals and humans was calculated using

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-17            DRAFT— DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene

 1   the Multi-Path Particle Dosimetry (MPPD) model, a computational model used for estimating
 2   human and rat airway particle deposition and clearance (MPPD; Version 2.0 © 2006, publicly
 3   available through the Hamner Institute). FTOT was based on the average particle size of 1.7 ± 0.085
 4   |im (mass median aerodynamic diameter [MMAD] ± geometric SD) as reported in Wu etal. (2003a)
 5   for the exposure range 25-100 |im3. For the model runs, the Yeh-Schum 5-lobe model was used for
 6   the human and the asymmetric multiple path model was used for the rat (see Appendix E for MPPD
 7   model output). Both models were run under nasal breathing scenarios with the inhalability
 8   adjustment selected. A geometric SD of 1 was used as the default by the model because the
 9   reported geometric SD of 0.085 was <1.05.
10          The human parameters used in the model for calculating FTOT and in the subsequent
11   calculation of the PODnEc were as follows: human body weight, 70 kg; VE, 13.8 L/minute; breathing
12   frequency, 16 per minute; tidal volume, 860 mL; functional residual capacity, 3,300 mL; and upper
13   respiratory tract volume, 50 mL.  Although the most sensitive population in Archibong et al. (2002)
14   is the developing fetus, the adult rat dams were directly exposed. Thus, adult rat parameters were
15   used in the calculation of the HEC. The parameters  used for the rat were body weight, 0.25 kg (a
16   generic weight for male and female rats); VE, 0.18 L/minute;  breathing frequency, 102 per minute;
17   tidal volume, 1.8 mL; functional residual capacity, 4 mL; and upper respiratory tract volume,
18   4.42 mL. All other parameters were set to default values (see Appendix E).
19          Under these  conditions, the MPPD model calculated FTOT values of 0.621 for the human and
20   0.181 for the rat Using the above equation, the RDDRER was  calculated to be 1.1.
21          From this, the PODHEc was calculated as follows:
22
23                                       PODHEc = PODAD, x RDDRER
24
25          Table 2-4 summarizes the sequence of calculations leading to the derivation of a human-
26   equivalent POD for each data set discussed above.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               2-18            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
             Table 2-4. Summary of derivation of PODs
Endpoint and
reference
Species/sex
Model
Developmental
Decreased fetal survival
Archibong et al. (2002)

Pregnant F344 rats
BMR

LOAEL(25u.g/m3)
BMC
(Hg/m3)

19% ^
BMCL
(Hg/m3)
PODADJa
(Hg/m3)
PODHECb
(Hg/m3)


4.2
4.6
Reproductive
Decreased testis weight
Archibong et al. (2008)

Decreased sperm count
and motility
Archibong et al. (2008)

Male F344 rats
Male F344 rats
LOAEL(75u.g/m3)
LOAEL(75u.g/m3
69% 4/sperm cou
73% 4, sperm mo
54% T* abnormal
34% ^
nt
tility
sperm


12.5
12.5
13.8
13.8
 2
 3    aPODs were adjusted for continuous daily exposure: PODADJ= POD x hours exposed per day/24 hours.
 4    bPODHEc calculated by adjusting the PODADJ by the RDDR calculated using particle size reported in Hood et al. (2000)
 5    using MPPD software as detailed in Section 2.2.2 and Appendix E in the Supplemental Information.

 6    2.2.3.  Derivation of Candidate Values
 7           Under EPA's A Review of the Reference Dose and Reference Concentration Processes [U.S. EPA.
 8    2002: Section 4.4.5], also described in the Preamble, five possible areas of uncertainty and
 9    variability were considered. An explanation of the five possible areas of uncertainty and variability
10    follows:
11           An intraspecies uncertainty factor, UFH, of 10 was applied to account for variability and
12    uncertainty in toxicokinetic and toxicodynamic susceptibility within the subgroup of the human
13    population most sensitive to the health hazards  of benzo[a]pyrene [U.S. EPA. 2002). In the case of
14    benzo[a]pyrene, the PODs were derived from studies in inbred animal strains and are not
15    considered sufficiently representative of the exposure and dose-response of the most susceptible
16    human subpopulations (in this case, the  developing fetus). In certain cases, the toxicokinetic
17    component of this factor may be replaced when a PBPK model is available that incorporates the
18    best available information on variability in toxicokinetic disposition in the human population
19    (including sensitive subgroups). In the case of benzo[a]pyrene, insufficient information is available
20    to quantitatively estimate variability in human susceptibility; therefore, the full value for the
21    intraspecies UF was retained.
22           An interspecies uncertainty factor, UFA, of 3 (101/2 = 3.16, rounded to 3) was applied to
23    account for residual uncertainty in the extrapolation from laboratory animals to humans in the
24    absence of information to characterize toxicodynamic differences between rats and humans after
25    inhalation exposure to benzo[a]pyrene.  This value is adopted by convention where an adjustment
26    from animal to a HEC has been performed as described in EPA's Methods for Derivation of Inhalation
27    Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA. 1994).

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                 2-19             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          A subchronic to chronic uncertainty factor, UFS, of 1 was applied when dosing occurred
 2    during gestation [Archibongetal.. 2002] or the early postnatal period that is relevant to
 3    developmental effects [U.S. EPA. 1991a). A value of 10 was applied when the POD is based on a
 4    subchronic study to account for the possibility that longer exposure may induce effects at a lower
 5    dose [Archibong et al.. 2008] was 60 days in duration). A UF for extrapolation from a LOAEL to a
 6    NOAEL, UFL, of 10 was applied when a LOAEL was used as the POD (Archibongetal.. 2008:
 7    Archibongetal.. 2002]. The data reported in these studies were not amenable to dose-response
 8    modeling, which would have allowed for extrapolation to a minimally biologically significant
 9    response level. At the LOAEL, these studies observed a high magnitude of response (see Table 2-4].
10    Therefore, a full UF of 10 was applied to approximate a NOAEL for studies that observed a high
11    magnitude of response at the LOAEL. For example, the LOAEL used as the POD for the
12    developmental effect observed in Archibongetal. (2002] was based on a 19% decrease in fetal
13    survival.
14          A database uncertainty factor, UFo, of 10 was applied to account for database deficiencies,
15    including the lack of a standard multigenerational study or extended 1-generation study that
16    includes exposure from premating through lactation, considering that benzo[a]pyrene has been
17    shown to affect fertility in adult male and female animals by multiple routes of exposure and that
18    decrements in fertility are greater following developmental exposure (see Section 1.1.2).
19          In addition, the lack of a study examining functional neurological endpoints following
20    inhalation exposure during development is also a data gap, considering human and animal evidence
21    indicating altered neurological development following exposure to benzo[a]pyrene alone or
22    through PAH mixtures (see Section 1.1.1).
23          The most sensitive POD for the RfC candidate values in Table 2-5 is based on the endpoint of
24    decreased fetal survival observed in Archibongetal. (2002). However, oral exposure studies have
25    demonstrated neurotoxicity at doses lower than those where decreased fetal survival was
26    observed. A statistically significant decrease in fetal survival was observed following treatment
27    with 160 mg/kg-day benzo[a]pyrene, but not at lower doses (Mackenzie and Angevine. 1981):
28    however, other oral studies observed statistically significant neurobehavioral effects at doses of
29    benzo[a]pyrene around 0.2-2 mg/kg-day (Chenetal.. 2012: Bouayed etal.. 2009a). Considering
30    the relative sensitivity of the systemic health effects observed in the oral database, it is likely that
31    neurodevelopmental toxicity would be expected to occur at exposure concentrations below the
32    POD for the RfC based on decreased fetal survival.
33          According to EPA's A Review of the Reference Dose and Reference Concentration Processes
34    (U.S. EPA. 2002: Section 4.4.5). the UFD is intended to account for the potential for deriving an
35    under-protective RfD/RfC as a result of an incomplete characterization of the  chemical's toxicity,
36    but also including a review of existing data that may also suggest that a lower reference value might
37    result if additional data were available. Therefore, a database UF of 10 for the benzo[a]pyrene
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-20            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
 1
 2
 3
 4
 5
 6
 7
inhalation database was applied to account for the lack of a multigenerational study and the lack of
a developmental neurotoxicity study.
       Table 2-5 is a continuation of Table 2-4 and summarizes the application of UFs to each POD
to derive a candidate values for each data set The candidate values presented in the table below
are preliminary to the derivation of the organ/system-specific reference values. These candidate
values are considered individually in the selection of an RfC  for a specific hazard and subsequent
overall RfC for benzo[a]pyrene.

       Table 2-5. Effects and corresponding derivation  of candidate values


Endpoint

PODHEC
(Hg/m3)

POD
type


UFA


UFH


UFL


UFS


UFD

Composite
UFb
Candidate
value3
(mg/m3)
Developmental
Decreased fetal survival
in rats
Archibong et al. (2002)

4.6



LOAEL



3



10



10



1



10



3,000



2 x 10"6



Reproductive
Decreased testis weight
in rats
Archibong et al. (2008)

Decreased sperm count
and motility in rats
Archibong et al. (2008)
13.8



13.8


LOAEL



LOAEL


3



3


10



10


10



10


10



10


10



10


30,000



30,000


Not calculated
duetoUF
>3,000

Not calculated
duetoUF
>3,000
 9
10
11
12
13
14
15
16
aCandidate values were converted from u.g/m3 to mg/m3.
bAs recommended in EPA's A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA,
 2002), the derivation of a reference value that involves application of the full 10-fold UF in four or more areas of
 extrapolation should be avoided.

       Figure 2-2 presents graphically these candidate values UFs, and PODs, with each bar
corresponding to one data set described in Tables 2-4 and 2-5.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                  2-21            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
               Q-
               O
 4- fetal survival in rats
(Archibong etal., 2002)
                                                                                    Composite UF
                                                                                  A Candidate value
                                                                                  • POD(HEC)
               Q
               O
               cc
                        x]/ testis weight in rats
                       (Archibongetal., 2008)
                   4- sperm count and motility
                  in rats (Archibong et al., 2008)
                                        0.001       0.01        0.1         1
                                                        Exposure concentration
                                                            10
100
 1
 2           Figure 2-2. Candidate values with corresponding PODs and composite UFs.
 3    2.2.4.  Derivation of Organ/System-Specific Reference Concentrations
 4           Table 2-6 distills the candidate values from Table 2-5 into a single value for each organ or
 5    system. These organ- or system-specific reference values may be useful for subsequent cumulative
 6    risk assessments that consider the combined effect of multiple agents acting at a common site. The
 7    candidate values for reproductive toxicity from Archibong et al. [2008] were not derived to
 8    represent reproductive toxicity because as recommended in EPA's A Review of the Reference Dose
 9    and Reference Concentration Processes [U.S. EPA. 2002]. the derivation of a reference value that
10    involves application of the full 10-fold UF in four or more areas of extrapolation should be avoided.

11           Table 2-6.  Organ/system-specific RfCs and proposed overall RfC for
12           benzo[a]pyrene
Effect
Developmental
Reproductive
Proposed Overall RfC
Basis
Decreased fetal survival
Reductions in testes weight and
sperm parameters
Decreased fetal survival
RfC (mg/m3)
2 x 10"6
Not calculated
2 x 10"6
Study exposure
description
Critical window of
development
(prenatal)
Subchronic
Critical window of
development
(prenatal)
Confidence
Low-medium
NA
Low-medium
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                 2-22            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    2.2.5.  Selection of the Proposed Reference Concentration
 2           The derivation of multiple organ/system-specific reference concentrations were considered
 3    for effects identified as human hazards of benzo[a]pyrene inhalation exposure, i.e., developmental
 4    and reproductive toxicity.  However, an organ/system-specific RfC to represent reproductive
 5    toxicity could not be derived due to high uncertainty (i.e., a composite UF of >3,000).
 6           An overall RfC of 2 x 10~6 mg/m3 was selected based on the hazard of developmental
 7    toxicity. The study by Archibong et al. [2002] was selected as the study used for the derivation of
 8    the proposed overall RfC, as it observed biologically significant effects at the lowest dose tested by
 9    the inhalation route. This study indicates that the developing fetus is a sensitive target following
10    inhalation exposure to benzo[a]pyrene and the observed decreased fetal survival/litter is the most
11    sensitive noncancer effect observed following inhalation exposure to benzo[a]pyrene. Additional
12    support for this endpoint of decreased fetal survival is provided by a developmental/reproductive
13    study conducted via the oral route [Mackenzie and Angevine. 1981].
14           This overall RfC is derived to be protective of all types of effects for a given duration of
15    exposure and is intended to protect the population as a whole, including potentially susceptible
16    subgroups [U.S. EPA, 2002]. This value should be applied in general population risk assessments.
17    However, decisions concerning averaging exposures over time for comparison with the RfC should
18    consider the types of toxicological effects and specific lifestages of concern. For example,
19    fluctuations in exposure levels that result in elevated exposures during these lifestages could
20    potentially lead to an appreciable risk, even if average levels over the full exposure duration were
21    less than or equal to the RfC. Alternatively, developmental toxicity may not be a concern due to
22    exposure scenarios in which exposure is occurring outside of the critical window of development

23    2.2.6.  Confidence Statement
24           A confidence level of high, medium, or low is assigned to the study used to derive the RfC,
25    the overall database, and the RfC itself, as described in Section 4.3.9.2 of EPA's Methods for
26    Derivation  of Inhalation Reference Concentrations and Application  of Inhalation Dosimetry [U.S. EPA.
27    1994].
28           The overall confidence in the RfC is low-to-medium. Confidence in the principal study
29    [Archibong et al., 2002] is medium. The conduct and reporting of this developmental study were
30    adequate; however, a NOAEL was not identified.  Confidence in the database is low due to the lack
31    of a multigeneration toxicity study, lack of studies on developmental neurotoxicity and immune
32    endpoints, and lack of information regarding  subchronic and chronic inhalation exposure.
33    However, confidence in the RfC is bolstered by consistent systemic effects observed by the oral
34    route (including reproductive and developmental effects] and similar effects observed in human
35    populations exposed to PAH mixtures. Reflecting medium confidence in the principal study and low
36    confidence in the database, confidence in the  RfC is low-to-medium.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-23             DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    2.2.7.  Previous IRIS Assessment: Reference Concentration
 2           An RfC was not derived in the previous IRIS assessment

 3    2.2.8.  Uncertainties in the Derivation of the RfD and RfC
 4           The following discussion identifies uncertainties associated with the RfD and RfC for
 5    benzo[a]pyrene. To derive the RfD, the UF approach (U.S. EPA. 2000a. 19941 was applied to a POD
 6    based on neurobehavioral changes in rats treated developmentally. To derive the RfC, this same
 7    approach was applied to a POD from a  developmental study for the effect of decreased fetal
 8    survival. UFs were applied to the POD  to account for extrapolating from an animal bioassay to
 9    human exposure, the likely existence of a diverse population of varying susceptibilities, and
10    database deficiencies. These extrapolations are carried out with default approaches given the lack
11    of data to inform individual steps.
12           The database for benzo[a]pyrene contains limited human data.  The observation of effects
13    associated with benzo[a]pyrene exposure in humans is complicated by several factors including the
14    existence of benzo[a]pyrene in the environment as one component of complex mixtures of PAHs,
15    exposure to benzo[a]pyrene by multiple routes of exposure within individual studies, and the
16    difficulty in obtaining accurate exposure information. Data on the effects of benzo[a]pyrene alone
17    are derived from a large database of studies in animal models. The database for oral
18    benzo[a]pyrene  exposure includes two lifetime bioassays in rats and mice, two developmental
19    studies in mice, and several subchronic studies in rats.
20           Although the database is adequate for RfD derivation, there is uncertainty associated with
21    the database including that the principal study for the RfD exposed animals during a relatively short
22    period of brain development potentially underestimating the magnitude of resulting neurological
23    effects. Also, the database lacks a comprehensive multi-generation reproductive/developmental
24    toxicity studies and immune system endpoints were not evaluated in the available chronic-duration
25    or developmental studies.  Additionally, the only available chronic studies of oral  or inhalational
26    exposure to benzo[a]pyrene focused primarily on neoplastic effects leaving non-neoplastic  effects
27    mostly uncharacterized.
28           The only chronic inhalation study of benzo[a]pyrene was designed as a lifetime
29    carcinogenicity study and did not examine noncancer endpoints  [Thyssenetal., 1981]. In addition,
30    subchronic and short-term inhalation studies are available, which examine developmental and
31    reproductive endpoints in rats.  Developmental studies by the inhalation route identified
32    biologically significant reductions in the number of pups/litter and percent fetal survival and
33    possible neurodevelopmental effects (e.g., diminished electrophysiological responses to stimuli in
34    the hippocampus) following gestational exposures. Additionally, a 60-day oral study in male rats
35    reported male reproductive effects (e.g., decreased testes weight and sperm production and
36    motility), but provides limited information to characterize dose-response relationships with
37    chronic exposure scenarios.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-24            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          The study selected as the basis of the RfC provided limited information regarding the
 2    inhalation exposures of the animals. Specifically, it is not clear whether the reported
 3    concentrations were target values or analytical concentrations and the method used to quantify
 4    benzo[a]pyrene in the generated aerosols was not provided.  Requests to obtain additional study
 5    details from the authors were unsuccessful; therefore, the assumption was made that the reported
 6    concentrations were analytical concentrations.
 7          One area of uncertainty in the database pertains to the lack of information regarding
 8    fertility in animals exposed gestationally to benzo[a]pyrene, especially in light of developmental
 9    studies by the oral route indicating reduced fertility in the Fl generation and decreased
10    reproductive organ weights.  The database also lacks a multigenerational reproductive study via the
11    inhalation route. Areas of uncertainty include the lack of chronic inhalation studies focusing on
12    noncancer effects, limited data on dose-response relationships for impaired male or female fertility
13    with gestational exposure or across several generations, and limited data on immune system
14    endpoints with chronic exposure to benzo[a]pyrene.
15          The toxicokinetic and toxicodynamic differences for benzo[a]pyrene between the animal
16    species in which the POD was derived and humans are unknown. PBPK models can be useful for
17    the evaluation of interspecies toxicokinetics; however, the benzo[a]pyrene database lacks an
18    adequate model that would inform potential differences. There is some evidence from the oral
19    toxicity data that mice may be more susceptible than rats to some benzo[a]pyrene effects (such as
20    ovo toxicity) [Borman et al.. 2000). although the underlying mechanistic basis of this apparent
21    difference is not understood. Most importantly, it is unknown which animal species may be more
22    comparable to humans.

23    2.3.  ORAL SLOPE FACTOR FOR CANCER
24          The carcinogenicity assessment provides information on the carcinogenic hazard potential
25    of the substance in question and quantitative estimates of risk from oral and inhalation exposure
26    may be derived. Quantitative risk estimates may be derived from the application of a low-dose
27    extrapolation procedure.  If derived, the oral slope factor is a plausible upper bound on the estimate
28    of risk per mg/kg-day of oral exposure.

29    2.3.1.  Analysis of Carcinogenicity Data
30          The database for benzo[a]pyrene contains numerous cancer bioassays that identify tumors,
31    primarily of the alimentary tract including the forestomach, following oral exposure in rodents.
32    Three 2-year oral bioassays are available that associate lifetime benzo[a]pyrene exposure with
33    carcinogenicity at multiple sites: forestomach, liver, oral cavity, jejunum, kidney, auditory canal
34    (Zymbal  gland) tumors, and skin or mammary gland tumors in male and female Wistar rats [Kroese
35    etal., 2001): forestomach tumors in male and female Sprague-Dawley rats [Brune etal., 1981): and
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-25             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    forestomach, esophageal, tongue, and larynx tumors in female B6C3Fi mice [Beland and Gulp, 1998:
 2    Gulp etal.. 19981
 3          In addition to these 2-year cancer bioassays, there are studies available that provide
 4    supporting evidence of carcinogenicity but are less suitable for dose-response analysis due to one
 5    or more limitations in study design: [1] no vehicle control group; (2) only one benzo[a]pyrene dose
 6    group; or (3) a one-time exposure to benzo[a]pyrene [Benjamin et al.. 1988: Robinson et al.. 1987:
 7    El-Bayoumy. 1985: Wattenberg. 1974: Roe etal.. 1970: Biancifiorietal.. 1967: Chouroulinkovetal..
 8    1967: Berenblum and Haran. 1955]. Of the controlled, multiple dose-group, repeat-dosing studies
 9    that remain, most treated animals for <1  year, which is less optimal for extrapolating to a lifetime
10    exposure [Weyandetal., 1995: Triolo etal., 1977: Fedorenko and Yansheva, 1967: Neal and Rigdon,
11    19671
12          Brune etal. [1981] dosed rats (32/sex/group] with benzo[a]pyrene in the diet or by gavage
13    in a 1.5% caffeine solution, sometimes as infrequently as once every 9th day, for approximately
14    2 years and observed increased forestomach tumors. This study was not selected for quantitation
15    due to the nonstandard treatment protocol in comparison to the Good Laboratory Practice (GLP]
16    studies conducted by Kroese etal. [2001] and Beland and Gulp [1998] and the limited reporting of
17    study methods.
18          The Kroese etal.  f20011 and Beland and Gulp T19981 studies were selected as the best
19    available studies for dose-response analysis and extrapolation to lifetime cancer risk following oral
20    exposure to benzo[a]pyrene. The rat bioassay by Kroese etal. [2001] and the mouse bioassay by
21    Beland and Gulp [1998] were conducted in accordance with GLP as established by the Organisation
22    for Economic Co-operation and Development [OECD]. These  studies included histological
23    examinations for tumors in many different tissues, contained three exposure levels and controls,
24    contained adequate numbers of animals per dose group [~50/sex/group], treated animals for up to
25    2 years, and included detailed reporting of methods and results (including individual animal data].
26          Details of the rat [Kroese et al.. 2001] and female mouse [Beland and Gulp. 1998] study
27    designs are provided in Appendix D of the Supplemental Information. Dose-related increasing
28    trends in tumors were noted at the following sites:

29       •   Squamous cell carcinomas [SCCs] or papillomas of the forestomach or oral cavity in male
30          and female rats;

31       •   SCCs or papillomas of the forestomach, tongue, larynx, or esophagus in female mice;

32       •   Auditory canal carcinomas in male and female rats;

33       •   Kidney urothelial carcinomas in male rats;

34       •   Jejunum/duodenum adenocarcinomas in female and male rats;

35       •   Hepatocellular adenomas or carcinomas in male and female rats; and

                This document is a draft for review purposes only and does not constitute Agency policy.
                                               2-26            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1       •  SCCs or basal cell tumors of the skin or mammary gland in male rats.

 2          These tumors were generally observed earlier during the study with increasing exposure
 3    levels, and showed statistically significantly increasing trends in incidence with increasing
 4    exposure level (Cochran-Armitage trend test, p < 0.001). These data are summarized in Appendix D
 5    of the Supplemental Information. As recommended by the National Toxicology Program (NTP)
 6    [McConnell etal., 1986] and as outlined in EPA's Guidelines for Carcinogen Risk Assessment [U.S.
 7    EPA. 2005a). etiologically similar tumor types (i.e., benign and malignant tumors of the same cell
 8    type) were combined for these tabulations when it was judged that the benign tumors could
 9    progress to the malignant form. In addition, when one tumor type occurred across several
10    functionally related tissues, as with squamous cell tumors in the tongue, esophagus, larynx, and
11    forestomach, or adenocarcinomas of the jejunum or duodenum, these incidences were also
12    aggregated as counts of tumor-bearing animals.
13          In the rat study [Kroese et al., 2001], the oral cavity and auditory canal were examined
14    histologically only if a lesion or tumor was observed grossly at necropsy. Consequently, dose-
15    response analysis for these sites was not straightforward. Use of the number of tissues examined
16    histologically as the number at risk would tend to overestimate the incidence, because the
17    unexamined animals were much less likely to have a tumor.  On the other hand, use of all animals in
18    a group as the number at risk would tend to underestimate if any of the unexamined animals had
19    tumors that could only be detected microscopically. The oral cavity squamous cell tumors were
20    combined with those in the forestomach because both are part of the alimentary tract, recognizing
21    that there was some potential for underestimating this cancer risk.
22          The auditory canal tumors from the rat study were not considered for dose-response
23    analysis, for several reasons.  First, the control and lower dose groups were not thoroughly
24    examined, similar to the situation described above for oral cavity tumors.  Unlike the oral cavity
25    tumors, the auditory canal tumors were not clearly related to any other site or tumor type, as they
26    were described as a mixture of squamous and sebaceous cells derived from pilosebaceous units.
27    The tumors were observed mainly in the high-dose groups and were highly coincident with the oral
28    cavity and forestomach tumors. Because the only mid-dose male with  an auditory canal tumor did
29    not also have a forestomach or oral cavity squamous cell tumor, and no auditory canal tumors were
30    observed in low-dose male or female rats, the data are insufficient to conclude that the auditory
31    canal tumors occur independently of other tumors. The investigators did not suggest that these
32    tumors were metastases from other sites (in which case, the auditory canal tumors would be
33    repetitions of other tumors, or statistically dependent). Therefore dose-response  analysis was not
34    pursued for this site, either separately or in combination with another tumor type.
35          The incidence data that were modeled are provided in Tables E-9, E-10, andE-11 (Kroese et
36    al.. 2001: Beland and Gulp. 19981
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               2-27            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    2.3.2.  Dose-Response Analysis—Adjustments and Extrapolation Methods
 2          EPA's Guidelines for Carcinogen Risk Assessment [U.S. EPA. 2005a] recommend that the
 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    The dose response is assumed to be linear in the low-dose range, when evidence supports a
 6    mutagenic mode of action because of DNA reactivity, or if another mode of action that is anticipated
 7    to be linear is applicable. In this assessment, EPA concluded that benzo[a]pyrene carcinogenicity
 8    involves a mutagenic mode of action (as discussed in Section 1.1.5).  Thus, a linear approach to low-
 9    dose extrapolation was used.
10          The high-dose groups of both the rat and mouse studies were dead or moribund by week 79
11    for female mice, week 72 for female rats, and week 76 for male rats. Due to the occurrence of
12    multiple tumor types, earlier occurrence with increasing exposure and early termination of the
13    high-dose group in each study, methods that can reflect the influence of competing risks and
14    intercurrent mortality on site-specific tumor incidence rates are preferred. In this case, EPA has
15    used the multistage-Weibull model, which incorporates the time at which death-with-tumor
16    occurred as well as the dose.
17          Adjustments for approximating human equivalent slope factors applicable for continuous
18    exposure were applied prior to dose-response modeling. First, continuous daily exposure for the
19    gavage study in rats [Kroese etal.. 2001] was estimated by multiplying each administered dose by
20    (5 days)/(7 days) = 0.71, under the assumption of equal cumulative exposure yielding equivalent
21    outcomes. Dosing was continuous in the mouse diet study [Beland and Gulp. 1998). so no
22    continuous adjustment was necessary. Next, consistent with the EPA's Guidelines for Carcinogen
23    Risk Assessment [U.S. EPA, 2005a], an adjustment for cross-species scaling was applied to address
24    toxicological equivalence across species. Following EPA's cross-species scaling methodology, the
25    time-weighted daily average doses were converted to HEDs on the basis of (body weight)3/4 (U.S.
26    EPA. 1992}.  This was accomplished by multiplying administered doses by (animal body weight
27    (kg)/70 kg)1/4 (U.S. EPA. 1992}. where the animal body weights were TWAs from each group, and
28    the U.S. EPA (1988) reference body weight for humans is 70 kg. It was not necessary to adjust the
29    administered doses for lifetime equivalent exposure prior to modeling for the groups terminated
30    early, because the multistage-Weibull model characterizes the tumor incidence as a function of
31    time, from which it provides an extrapolation to lifetime exposure.
32          Details of the modeling and the model selection process can be found in Appendix E of the
33    Supplemental Information. PODs for estimating low-dose risk were identified at doses at the lower
34    end of the observed data, generally corresponding to 10% extra risk.

35    2.3.3.  Derivation of the Oral Slope Factor
36          The PODs estimated for each tumor site are summarized in Table 2-7. The lifetime oral
37    cancer slope factor for humans is defined as the slope of the line from the lower 95% bound on the
               This document is a draft for review purposes only and does not constitute Agency policy.
                                               2-28            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene


 1    exposure at the POD to the control response (slope factor = 0.1/BMDLi0). This slope, a 95% upper
 2    confidence limit represents a plausible upper bound on the true risk. Using linear extrapolation
 3    from the BMDLio, human equivalent oral slope factors were derived for each gender/tumor site
 4    combination and are listed in Table 2-7.

 5           Table 2-7.  Summary of the oral slope factor derivations
Tumor
Forestomach, oral cavity:
squamous cell tumors
Kroese et al. (2001)

Hepatocellular adenomas or
carcinomas
Kroese et al. (2001)

Jejunum/duodenum
adenocarcinomas
Kroese et al. (2001)

Kidney: urothelial carcinomas
Kroese et al. (2001)

Skin, mammary:
Basal cell tumors
Squamous cell tumors
Kroese et al. (2001)

Forestomach, oral cavity:
squamous cell tumors
Kroese et al. (2001)

Hepatocellular adenomas or
carcinomas
Kroese et al. (2001)

Jejunum/duodenum
adenocarcinomas
Kroese et al. (2001)

Forestomach, esophagus, tongue,
larynx (alimentary tract):
squamous cell tumors
Beland and Gulp (1998)
Species/
sex
MaleWistar
rats
MaleWistar
rats
MaleWistar
rats
MaleWistar
rats
MaleWistar
rats
Female
Wistar rats
Female
Wistar rats
Female
Wistar rats
Female
B6C3Fi
Mice
Selected
model
Multistage
Weibull
Multistage
Weibull
Multistage
Weibull
Multistage
Weibull
Multistage
Weibull
Multistage
Weibull
Multistage
Weibull
Multistage
Weibull
Multistage
Weibull
BMR
10%
10%
10%
10%
10%
10%
10%
10%
10%
BMD
(mg/kg-d)
0.453
0.651
3.03
4.65
2.86
2.64
0.539
0.575
3.43
0.127
POD =
BMDL
(mg/kg-d)
0.281
0.449
2.38
2.50
2.35
1.77
0.328
0.507
1.95
0.071
Slope factor3
(mg/kg-d)-1
0.4
0.2
0.04
0.04
0.04
0.06
0.3
0.2
0.05
1
0.5b
0.3b
1
 6
 7
 8
 9
10
aHuman equivalent slope factor = 0.1/BMDL10HED; see Appendix E of the Supplemental Information for details of
 modeling results.
bSlope factor characterizing the risk of incurring at least one of the tumor types listed.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                 2-29             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          Oral slope factors derived from rat bioassay data varied by gender and tumor site
 2    (Table 2-7). Values ranged from 0.04 per mg/kg-day, based on kidney tumors in males, to 0.4 per
 3    mg/kg-day, based on alimentary tract tumors in males. Slope factors based on liver tumors in male
 4    and female rats (0.2 per mg/kg-day) were only slightly lower than slope factors based on
 5    alimentary tract tumors (0.2-0.3 per mg/kg-day). The oral slope factor for alimentary tract tumors
 6    in female mice was highest at 1 per mg/kg-day (Table 2-7), which was approximately twofold
 7    higher than the oral slope factor derived from the alimentary tract tumors in male rats.
 8          Although the time-to-tumor modeling helps to account for competing risks associated with
 9    decreased survival times and other causes of death including other tumors, considering the tumor
10    sites individually still does not convey the total amount of risk potentially arising from the
11    sensitivity of multiple sites—that is, the risk of developing any combination of the increased tumor
12    types. A method, for estimating overall risk, involving the assumption that the variability in the
13    slope factors could be characterized by a normal distribution, is detailed in Appendix E of the
14    Supplemental Information. The resulting composite slope  factor for all tumor types for male rats
15    was 0.5 per mg/kg-day, about 25% higher than the slope factor based on the most sensitive tumor
16    site, oral cavity and forestomach, while for female rats, the composite slope factor was equivalent to
17    that for the most sensitive site (Table 2-7; see Appendix E of Supplemental Information for
18    composite slope factor estimates).
19          The overall risk estimates from rats and mice spanned about a threefold range. As there are
20    no data to support any one result as most relevant for extrapolating to humans, the most sensitive
21    result was used to derive the oral slope factor. The recommended slope factor for assessing human
22    cancer risk associated with lifetime oral exposure to benzo[a]pyrene is 1 per mg/kg-day, based on
23    the alimentary tract tumor response in female B6C3Fi mice.  Note that the oral slope factor should
24    only be used with lifetime human exposures <0.1 mg/kg-day, because above this level, the dose-
25    response relationship is not expected to be proportional to benzo[a]pyrene exposure.

26    2.3.4. Uncertainties in the Derivation of the Oral Slope Factor
27          The oral slope factor for benzo[a]pyrene was based on the increased incidence of
28    alimentary tract tumors, including forestomach tumors, observed in a lifetime dietary study in mice
29    (Beland and Gulp, 1998). EPA has considered the uncertainty associated with the relevance of
30    forestomach tumors for estimating human risk from benzo[a]pyrene exposure. The rodent
31    forestomach serves to store foods and liquids for several hours before contents continue to the
32    stomach for further digestion (Clayson et al..  1990: Grice et al.. 1986). Thus, tissue of the
33    forestomach in rodents may be exposed to benzo[a]pyrene for longer durations than analogous
34    human tissues in the oral cavity and esophagus. This suggests that the rodent forestomach may be
35    quantitatively more sensitive to the development of squamous epithelial tumors in the forestomach
36    compared to oral or esophageal tumors in humans.
37          Uncertainty in the magnitude of the recommended oral slope factor is reflected to some
38    extent in the range of slope factors among tumors sites and species; the oral slope factor based on
                This document is a draft for review purposes only and does not constitute Agency policy.
                                               2-30            DRAFT—DO NOT CITE OR QUOTE

-------
                                                              Toxicological Review ofBenzo[a]pyrene


1    the mouse alimentary tract data was about threefold higher than the overall oral slope factor based
2    on male rat data (Table 2-8). These comparisons show that the selection of target organ, animal
3    species, and interspecies extrapolation can impact the oral cancer risk estimate. However, all of the
4    activation pathways implicated in benzo[a]pyrene carcinogenicity have been observed in human
5    tissues, and associations have been made between the spectra of mutations in tumor tissues from
6    benzo[a]pyrene-exposed animals and humans exposed to complex PAH mixtures containing
7    benzo[a]pyrene (see Section 1.1.5).
8
9
Table 2-8. Summary of uncertainties in the derivation of cancer risk values
for benzo[a]pyrene oral slope factor
            Consideration and
        impact on cancer risk value
                                 Decision
        Justification and discussion
     Selection of target organ
     •^ oral slope factor, up to fivefold, if
     alimentary tract tumors not selected
                          Alimentary tract tumors
                          (forestomach,
                          esophagus, tongue,
                          larynx)
Tumor site is concordant across rats and mice,
increasing support for its relevance to humans.
As there are no data to support any one result as
most relevant for extrapolating to humans, the
most sensitive result for alimentary tract tumors
was used to derive the oral slope factor.
     Selection of data set
     •^ oral slope factor ~threefold if rat
     bioassay were selected for oral slope
     factor derivation
                          Beland and Gulp (1998)
Beland and Gulp (1998) was a well-conducted
study and had the lowest HEDs of the available
cancer bioassays, reducing low-dose
extrapolation uncertainty.
     Selection of dose metric
     Alternatives could 4, or /T" slope
     factor
                          Administered dose
Experimental evidence supports a role for
metabolism in toxicity, but actual responsible
metabolites have not been identified.
     Interspecies extrapolation
     Alternatives could 4, or 1" slope
     factor (e.g., 3.5-fold 4, [scaling by
     body weight] or 1" 2-fold [scaling by
     BW2/3])
                          BW3/4 scaling (default
                          approach)
There are no data to support alternatives.
Because the dose metric was not an area under
the curve, BW   scaling was used to calculate
equivalent cumulative exposures for estimating
equivalent human risks. While the true human
correspondence is unknown, this overall
approach is expected to neither over- nor
underestimate human equivalent risks.
     Dose-response modeling
     Alternatives could 4, or /T" slope
     factor
                          Multistage-Weibull
                          model
No biologically based models for benzo[a]pyrene
were available.  Because the multistage-Weibull
model could address additional available data
(time of death with tumor, and whether a tumor
caused the death of the animal), this model was
superior to other available models.
     Low-dose extrapolation
     •i, cancer risk estimate would be
     expected with the application of
     nonlinear low-dose extrapolation
                          Linear extrapolation
                          from POD (based on
                          mutagenic mode of
                          action)
Available mode-of-action data support linearity
(mutagenicity is a primary mode of action of
benzo[a]pyrene).
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                   2-31             DRAFT—DO NOT CITE OR QUOTE

-------
                                                        Toxicological Review ofBenzo[a]pyrene
Consideration and
impact on cancer risk value
Statistical uncertainty at POD
4> oral slope factor 1.8-fold if BMD
used as the POD rather than BMDL
Sensitive subpopulations
1" oral slope factor to unknown
extent
Decision
BMDL (preferred
approach for calculating
plausible upper bound
slope factor)
ADAFs are
recommended for early
life exposures
Justification and discussion
Limited size of bioassay results in sampling
variability; lower bound is 95% Cl on administered
exposure at 10% extra risk of alimentary tract
tumors.
No chemical-specific data are available to
determine the range of human toxicodynamic
variability or sensitivity.
 1    2.3.5.  Previous IRIS Assessment: Oral Slope Factor
 2          The previous cancer assessment for benzo[a]pyrene was posted on the IRIS database in
 3    1992. At that time, benzo[a]pyrene was classified as a probable human carcinogen (Group B2)
 4    based on inadequate data in humans and sufficient data in animals via several routes of exposure.
 5    An oral slope factor was derived from the geometric mean of four slope factor estimates based on
 6    studies of dietary benzo[a]pyrene administered in the diet for approximately 2 years in 10-week-
 7    old Sprague-Dawley rats [Brune etal., 1981] and administered for up 7 months in 2-week-old to
 8    5-month-old CFW-Swiss mice [Neal and Rigdon. 1967]. A single slope factor estimate of 11.7 per
 9    mg/kg-day, using a linearized multistage procedure applied to the combined incidence of
10    forestomach, esophageal, and laryngeal tumors, was derived from the Brune etal. [1981] study (see
11    Section 1.1.5 for study details].  Three modeling procedures were used to derive risk estimates
12    from the Neal and Rigdon (1967] bioassay (see Section 1.1.5]. U.S. EPA (1991a] fit a two-stage
13    response model, based on exposure-dependent changes in both transition rates and growth rates of
14    preneoplastic cells, to derive a value of 5.9 per mg/kg-day.  U.S. EPA(1991b] derived a value of
15    9.0 per mg/kg-day by linear extrapolation from the 10% response point to the background
16    response in a re-analysis of the  1990 model. Finally, using a Weibull-type model to reflect less-
17    than-lifetime exposure to benzo[a]pyrene, the same assessment (U.S. EPA. 1991b] derived an
18    upper-bound slope factor estimate of 4.5 per mg/kg-day. The four slope factor estimates, which
19    reflected extrapolation to humans assuming surface area equivalence (BW2/3 scaling] were within
20    threefold of each other and were judged to be of equal merit Consequently, the geometric mean of
21    these four estimates, 7.3 per mg/kg-day, was recommended, in 1992, as the oral slope factor.

22    2.4. INHALATION UNIT RISK FOR CANCER
23          The carcinogenicity assessment provides information on the carcinogenic hazard potential
24    of the substance in question and quantitative estimates of risk from oral and inhalation exposure
25    may be derived. Quantitative risk estimates may be derived from the application of a low-dose
26    extrapolation procedure. If derived, the inhalation unit risk is a plausible upper bound on the
27    estimate of risk per |J.g/m3 air breathed.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                               2-32            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    2.4.1. Analysis of Carcinogenicity Data
 2          The inhalation database demonstrating carcinogenicity of benzo[a]pyrene consists of a
 3    lifetime inhalation bioassay in male hamsters [Thyssenetal.. 1981] and intratracheal instillation
 4    studies, also in hamsters [Feron and Kruysse. 1978: Ketkar etal.. 1978: Feronetal.. 1973: Henry et
 5    al.. 1973: Saffiotti etal.. 1972).  The intratracheal instillation studies provide supporting evidence of
 6    carcinogenicity of inhaled benzo[a]pyrene; however, the use of this exposure method alters the
 7    deposition, clearance, and retention of substances, and therefore, studies utilizing this exposure
 8    technique are not as useful for the quantitative extrapolation of cancer risk from the inhalation of
 9    benzo[a]pyrene in the environment [Driscolletal.,  2000].
10          The bioassay by Thyssenetal. [1981] represents the only lifetime inhalation cancer
11    bioassay available for describing exposure-response relationships for cancer from inhaled
12    benzo[a]pyrene. As summarized in Section 1.1.5, increased incidences of benign and malignant
13    tumors of the pharynx, larynx, trachea, esophagus, nasal cavity, or forestomach were seen with
14    increasing exposure concentration. In addition, survival was decreased relative to control in the
15    high-exposure group; mean survival times in the control, low-, and mid-concentration groups were
16    96.4, 95.2,  and 96.4 weeks, respectively, compared  to 59.5 weeks in the high-exposure group
17    animals [Thyssenetal.. 1981].  Overall, tumors occurred earlier in the highest benzo[a]pyrene
18    exposure group than in the mid-exposure group.
19          Strengths of the study included exposures until natural death, up to 2.5 years, multiple
20    exposure groups; histological examination of multiple organ systems, and availability of individual
21    animal pathology reports with time of death and tumor incidence data by site in the upper
22    respiratory and digestive tracts. In addition, the availability of weekly chamber air monitoring data
23    and individual times on study allowed the calculation of time-weighted average (TWA] lifetime
24    continuous exposures for each hamster. Group averages of these TWA concentrations were 0, 0.25,
25    1.01, and 4.29 mg/m3 fU.S. EPA. 19901
26          Several limitations concerning exposure conditions in the Thyssenetal. [1981] study were
27    evaluated for their impact on the derivation of an inhalation unit risk for benzo[a]pyrene. These
28    issues include minimal detail about the particle size distribution of the administered aerosols,
29    variability of chamber concentrations, and the use of a sodium chloride aerosol as a carrier.
30          First, particle distribution analysis of aerosols, in particular the MMAD and geometric SD,
31    was not reported, although the investigators did report that particles were within the respirable
32    range for hamsters, with >99% of the particles having diameters 0.2-0.5 um and >80% having
33    diameters  0.2-0.3 um.
34          Second, weekly averages of chamber concentration measurements varied two- to fivefold
35    from the overall average for each group, which exceeds the limit for exposure variability of <20%
36    for aerosols recommended by OECD [2009]. For risk assessment purposes,  EPA generally assumes
37    that cancer risk is proportional to cumulative exposure, and therefore to lifetime average exposure
38    as estimated  here, when there is no information to the contrary. Under this  assumption, the

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-33            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    variability of the chamber concentrations has little impact on the estimated exposure-response
 2    relationship. The impacts of alternative assumptions are considered in Section 2.4.4.
 3           Lastly, exposure occurred through the inhalation of benzo[a]pyrene adsorbed onto sodium
 4    chloride aerosols, which might have irritant carrier effects, and may have a different deposition
 5    than benzo[a]pyrene adsorbed onto carbonaceous particles (as is more typical in the environment).
 6    The above study design and reporting issues concerning the particle size composition, exposure
 7    variability, and deposition do not negate the robust tumor response following benzo[a]pyrene
 8    inhalation exposure. Consequently, EPA concluded that the strengths of the study supported the
 9    use of the data to derive an inhalation unit risk for benzo[a]pyrene.  See Section 2.4.4 for a
10    discussion of uncertainties in the unit risk.

11    2.4.2.  Dose-Response Analysis—Adjustments and Extrapolation Methods
12           Biologically based dose-response models for benzo[a]pyrene are not available. A simplified
13    version of the two-stage carcinogenesis model proposed by Moolgavkar and Venzon [1979] and
14    Moolgavkar and Knudson [1981] has been applied to the Thyssenetal. [1981] individual animal
15    data [U.S. EPA, 1990]. However, the simplifications necessary to fit the tumor incidence data
16    reduced that model to an empirical model (i.e., there were no biological data to inform estimates of
17    cell proliferation rates for background or initiated cells]. Sufficient data were available to apply the
18    multistage-Weibull model, as used for the  oral slope factor (described in detail in Appendix E of the
19    Supplemental Information], specifically the individual times of death for each animal. Unlike in the
20    oral bioassays, Thyssenetal. (1981] did not determine cause of death for any of the animals. Since
21    the investigators for the oral bioassays considered some of the same tumor types to be fatal at least
22    some of the time, bounding estimates of the POD for these Thyssenetal. (1981] data were
23    developed by treating the tumors alternately as either  all incidental to the death of an affected
24    animal or as causing the death of an affected animal.
25           The tumor incidence data used for dose-response modeling comprised the benign and
26    malignant tumors in the pharynx and respiratory tract (see Table E-17]. The tumors in these sites
27    were judged to be sufficiently similar to combine in overall incidences, based on the assumption
28    that the benign tumors could develop into malignancies, as outlined in EPA's  Guidelines for
29    Carcinogen Risk Assessment (Section 2.2.2.1.2: U.S. EPA, 2005a]. Specifically, while the pharynx and
30    larynx are associated with the upper digestive tract and the upper respiratory tract, respectively,
31    these sites are close anatomically and in some cases where both tissues were affected, the site of
32    origin could not be distinguished (U.S. EPA. 1990]. In addition, the benign tumors (e.g., papillomas,
33    polyps, and papillary polyps] were considered early stages of the SCCs in these tissues (U.S. EPA.
34    1990].  Consequently, the overall incidence of SCCs or benign tumors judged to originate from the
35    same cell type (papillomas, polyps, or papillary polyps] were selected for dose-response modeling.
36           A toxicokinetic model to assist in cross-species scaling of benzo[a]pyrene inhalation
37    exposure was not available. EPA's RfC default dosimetry adjustments (U.S. EPA. 1994] were
38    utilized in the benzo[a]pyrene RfC calculation (see Section 2.2.2] but could not be applied to the
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-34            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    aerosols generated for the inhalation bioassay by Thyssen et al. [1981] as the approaches
 2    presented in the RfC methodology guidelines [U.S. EPA. 1994] were developed for insoluble and
 3    nonhygroscopic particles, not the sodium chloride particle used in Thyssen etal. [1981].
 4    Consequently, without data to inform a basis for extrapolation to humans, it was assumed that
 5    equal risk for all species would be associated with equal concentrations in air, at least at anticipated
 6    environmental concentrations. This is equivalent to assuming that any metabolism of
 7    benzo[a]pyrene is directly proportional to breathing rate and that the deposition rate is equal
 8    between species.
 9          The multistage-Weibull model was fit to the TWA exposure concentrations and the
10    individual animal tumor and survival data for tumors in the larynx, pharynx, trachea, or nasal cavity
11    (tumors of the pharynx and upper respiratory tract], using the software program, MultiStage-
12    Weibull [U.S. EPA. 2010c]. Modeling results are provided in Appendix E of the Supplemental
13    Information. Because benzo[a]pyrene carcinogenicity involves a mutagenic mode of action, linear
14    low-exposure extrapolation from the BMCLio was used to derive the inhalation unit risk [U.S. EPA.
15    2005a].

16    2.4.3.  Inhalation Unit Risk Derivation
17          The results from modeling the inhalation carcinogenicity data from Thyssen etal. [1981]
18    are summarized in Table 2-9. Taking the tumors to have been the cause of death  of the
19    experimental animals with tumors, the BMCio and BMCLio values were 0.468 and 0.256 mg/m3,
20    respectively. Then, taking all of the tumors to have been incidental to the cause of death for each
21    animal with a tumor, the BMCio and BMCLio values were 0.254 and 0.163 mg/m3, respectively,
22    about twofold lower than the first case. Because the tumors were unlikely to have all been fatal, the
23    lower BMCLio from the incidental deaths analysis, 0.163 mg/m3, is recommended for the
24    calculation of the inhalation unit risk. Using linear extrapolation from the BMCLio (0.163 mg/m3],
25    an inhalation unit risk of 0.6 per mg/m3, or 6 x 1Q-* per ng/m3 (rounding to one significant digit],
26    was calculated.  Note that the inhalation unit risk should only be used with lifetime human
27    exposures <0.3  mg/m3, the human equivalent POD, because above this level, the dose-response
28    relationship is not expected to be proportional to benzo[a]pyrene exposure.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-35             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
            Table 2-9. Summary of the inhalation unit risk derivation


Tumor site and context
Upper respiratory tract and pharynx;
all treated as cause of death
Thyssenetal. (1981)

Upper respiratory tract and pharynx;
all treated as incidental to death
Thyssenetal. (1981)

Species/
sex
Male
hamsters


Male
hamsters


Selected
model
Multistage
Weibull, 2°


Multistage
Weibull, 2°



BMR
10%



10%



BMC
(mg/m3)
0.468



0.254


POD =
BMCL
(mg/m3)
0.256



0.163



Unit risk3
(mg/m3)-1
0.4



0.6


 2
 3    aHuman equivalent unit risk = 0.10/BMCL10; see Appendix E for details of modeling results.

 4    2.4.4. Uncertainties in the Derivation of the Inhalation Unit Risk
 5          Table 2-10 summarizes uncertainties in the derivation of the inhalation unit risk for
 6    benzo[a]pyrene; further detail is provided in the following discussion.  Only one animal cancer
 7    bioassay, in one sex, by the inhalation route is available that describes the exposure-response
 8    relationship for respiratory tract tumors with lifetime inhalation exposure to benzo [a]pyrene
 9    [Thyssenetal.. 1981). Although corroborative information on exposure-response relationships in
10    other animal species is lacking, the findings for upper respiratory tract tumors are consistent with
11    findings in other hamster studies with intratracheal administration of benzo[a]pyrene (upper and
12    lower respiratory tract tumors), and with some of the portal-of-entry effects in oral exposure
13    studies.
14          The hamster inhalation bioassay by Thyssen et al. [1981] observed upper respiratory tract
15    tumors, but not lung tumors. The lack of a lung tumor response in hamsters, given the strong
16    association of inhaled PAH mixtures with lung cancer in humans across many studies (see
17    Section 1.1.5) suggests that this study may not be ideal for extrapolating to humans. Hamsters have
18    an apparent lower sensitivity to lung carcinogenesis than rats and mice and a tendency to give false
19    negatives for particles classified as carcinogenic to humans by IARC (Mauderly. 1997). However,
20    hamster laryngeal tumors have been used as an indication of the carcinogenic hazard of cigarette
21    smoke for more than 50 years (IARC, 2002).  For example, a large study investigating the inhalation
22    of cigarette smoke in hamsters (n = 4,400) indicated that the larynx was the most responsive tumor
23    site, which the authors indicated was due to a large difference in particle deposition between the
24    larynx and the lung (Dontenwill etal.. 1973). EPA's Guidelines for Carcinogen Assessment (U.S. EPA.
25    2005a] stress that site concordance between animals and humans need not always be assumed.
26    Therefore,  the robust tumor response in the upper respiratory tract of Syrian golden hamsters was
27    considered to be supportive of the use of the Thyssenetal. (1981) study for the derivation of an
28    inhalation unit risk.
29          An  additional uncertainty includes the inability to apply U.S. EPA (1994) dosimetry
30    approaches to extrapolate inhaled concentrations from animals to humans, due to the use of a

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-36            DRAFT—DO NOT CITE OR QUOTE

-------
                                                          Toxicological Review ofBenzo[a]pyrene

 1    soluble hygroscopic carrier particle (sodium chloride) for the delivery of benzo[a]pyrene. One
 2    likely consequence of the use of hygroscopic carrier particles would be the growth of
 3    benzo[a]pyrene-sodium chloride particles in the humid environment of the respiratory tract
 4    resulting in increased particle diameter and resulting changes in particle deposition, specifically,
 5    increased impaction in the upper respiratory tract and less deposition in the lung [Varghese and
 6    Gangamma. 2009: Asgharian. 2004: Perron. 1994: XuandYu. 1985). In addition, sodium chloride
 7    can be irritating to the respiratory tract, depending on concentration. The Thyssenetal. [1981]
 8    study reported that vehicle controls were exposed to 240 |ig/m3, and it is unclear whether exposure
 9    to this concentration of sodium chloride could have potentiated the tumor response seen in the
10    mid- and high-concentration benzo[a]pyrene groups.  Exposure to benzo[a]pyrene in the
11    environment predominantly occurs via non-soluble, non-hygroscopic, carbonaceous particles (such
12    as soot and diesel exhaust particles). The potential impact of differences in carrier particle on the
13    magnitude of the inhalation unit risk is unknown.
14          Regarding uncertainty associated with exposure characterization,  the individual exposure
15    chamber measurements varied from about an order of magnitude less than the target concentration
16    to about twofold higher than the target concentration. Weekly average analytical concentrations
17    were documented to vary by two- to fivefold in all exposed groups, with no particular trends over
18    time. Continuous time-weighted group average concentrations were used for dose-response
19    modeling under the assumption that equal cumulative exposures are expected to lead to similar
20    outcomes.  This assumption is generally expected to lead to an unbiased estimate  of risk when there
21    is incomplete information. However, it is possible that peak exposure above some concentration
22    may be more associated with the observed effects, or that deposition of particles may have reached
23    a maximum level or plateau, such as in the high-exposure group. Regarding the role of peak
24    exposures, the higher exposures for each group were distributed evenly throughout the study for
25    the most part, suggesting that any association of risk with peak exposures would also be
26    proportional to cumulative exposure. If particle deposition reached a plateau with the high-
27    exposure group, there is relatively less impact on the unit risk because the derivation relies on the
28    dose-response at lower exposure.  But the  actual dynamics of particle deposition at these or other
29    exposure levels are not well understood. There is not enough information available to estimate a
30    more quantitative impact on the estimated unit risk due to these uncertainties.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-37            DRAFT—DO NOT CITE OR QUOTE

-------
                                                                Toxicological Review ofBenzo[a]pyrene
1
2
        Table 2-10. Summary of uncertainties in the derivation of cancer risk values
        for benzo[a]pyrene (inhalation unit risk)
             Consideration and
         impact on cancer risk value
                                          Decision
                               Justification and discussion
     Selection of data set and target organ
     No inhalation unit risk if Thyssen et al.
     (1981) not used
                                   Respiratory tract tumors
                                   from Thyssen et al.
                                   (1981)
                       The Thyssen et al. (1981) bioassay is the only
                       lifetime inhalation cancer bioassay available for
                       describing exposure-response relationships for
                       cancer from inhaled benzo[a]pyrene.
                       Intratracheal installation studies support the
                       association of benzo[a]pyrene exposure with
                       respiratory tract tumors.
Selection of dose metric
Alternatives could 4, or
                             unit risk
Administered exposure
as TWA
Experimental evidence supports a role for
metabolism in toxicity, but actual responsible
metabolites are not identified. The
recommended unit risk is a reasonable estimate
if the proportion of the carcinogenic moiety
remains the same at lower exposures.
     Interspecies extrapolation
     Alternatives could 4, or ^ slope
     factor
                                   Cross-species scaling
                                   was not applied. The
                                   carrier particle used was
                                   soluble and hygroscopic,
                                   therefore the RfC
                                   methodology (U.S. EPA,
                                   1994) dosimetric
                                   adjustments could not
                                   be applied.
                       There are no data to support alternatives.  Equal
                       risk per u.g/m3 is assumed.
Dose-response modeling
Alternatives could 4, or
factor
                             slope
Multistage-Weibull
model
No biologically based models for benzo[a]pyrene
were available. Because the multistage-Weibull
model could address additional available data
(time of death with tumor), this model was
superior to other available empirical models.
     Low-dose extrapolation
     •^ cancer risk estimate would be
     expected with the application of
     nonlinear low-dose extrapolation
                                   Linear extrapolation
                                   from the POD (based on
                                   mutagenic mode of
                                   action)
                       Available mode-of-action data support linearity
                       (mutagenicity is a primary mode of action of
                       benzo[a]pyrene).
     Statistical uncertainty at POD
     •^ inhalation unit risk 1.4-fold if BMC
     used as the POD rather than BMCL
                                   BMCL (preferred
                                   approach for calculating
                                   plausible upper bound
                                   unit risk)
                       Limited size of bioassay results in sampling
                       variability; lower bound is 95% confidence
                       interval (Cl) on administered exposure at 10%
                       extra risk of respiratory tract tumors.
     Sensitive subpopulations
     "T* inhalation unit risk to unknown
     extent
                                   ADAFs are
                                   recommended for early
                                   life exposures
                       No chemical-specific data are available to
                       determine the range of human toxicodynamic
                       variability or sensitivity.
3

4
2.4.5.  Previous IRIS Assessment: Inhalation Unit Risk
        An inhalation unit risk for benzo[a]pyrene was not previously available on IRIS.
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                     2-38              DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
 1    2.5. DERMAL SLOPE FACTOR FOR CANCER
 2          Human and animal studies of exposure to PAH mixtures or benzo[a]pyrene alone
 3    demonstrate an increased incidence of skin tumors with increasing dermal exposure. This
 4    assessment for benzo[a]pyrene derives a dermal slope factor, a quantitative risk estimate that is a
 5    plausible upper bound on the estimate of risk per [J.g/day of lifetime dermal exposure. This
 6    derivation provides the first dermal slope factor for the Integrated Risk Information System (IRIS)
 7    database.

 8    2.5.1.  Analysis of Carcinogenicity Data
 9          Skin cancer in humans has been documented to result from occupational exposure to
10    complex mixtures of PAHs including benzo[a]pyrene, such as coal tar pitches, non-refined mineral
11    oils, shale oils, and soot flARC. 2010: BaanetaL 2009: IPCS. 1998: Boffetta etal.. 1997: AT SDR.
12    1995]. Although studies of human exposure to benzo[a]pyrene alone are not available, repeated
13    application of benzo[a]pyrene to skin (in the absence of exogenous promoters) has been
14    demonstrated to induce skin tumors in guinea pigs, rabbits, rats, and mice. Given the availability of
15    lifetime bioassays of dermal benzo[a]pyrene exposure in mice, this analysis focuses on
16    carcinogenicity bioassays including repeated dermal exposure to benzo[a]pyrene for approximately
17    2 years. These studies involved 2- or 3-times/week exposure protocols, at least two exposure
18    levels plus controls, and histopathological examinations of the skin and other tissues (Sivaketal..
19    1997: Grimmer etal.. 1984: Habsetal.. 1984: Grimmer etal.. 1983: Habs etal.. 1980:  Schmahl etal..
20    1977: Schmidt etal.. 1973:  Roe etal.. 1970: Poel. 1963.1959]. These studies, in several strains of
21    mice, demonstrated primarily malignant skin tumors, as well as earlier occurrence of tumors with
22    increasing exposure levels  (see Tables D-15 to D-23 in the Supplemental Information for study
23    details).
24          Other carcinogenicity studies in mice were judged to support the studies listed above, but
25    were not considered in the dose-response analysis because of the availability of lifetime studies
26    directly relevant to deriving a dermal slope factor. These other studies included:  (1) early "skin
27    painting" studies of benzo[a]pyrene carcinogenicity in mouse skin that did not report sufficient
28    information to estimate the doses applied (e.g., Wynder and Hoffmann, 1959: Wynder etal., 1957]:
29    (2) bioassays with minimal dose-response information, either using just one benzo[a]pyrene dose
30    level or with only dose levels inducing 90-100% incidence of mice with tumors, which provide
31    relatively little information about the shape of the dose-response relationship especially for low
32    exposures (e.g.. Wilson and Holland. 1988): and (3) shorter studies (i.e., <1 year) (Higginbotham et
33    al.. 1993: Albert etal.. 1991: Nesnow etal.. 1983: Emmettetal.. 1981: Levin etal.. 1977). which
34    would tend to underestimate lifetime risk by overlooking the potential for the development of
35    tumors in later life.
36          The National Institute for Occupational Safety and Health (NIOSH) study fSivak etal.. 1997:
37    NIOSH. 1989]  is a well-conducted and  documented study and was determined to be the best

               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-39            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    available study for dose-response analysis. Specifically, mice were randomly assigned to treatment
 2    groups while maintaining comparable distributions of body weights, housed singly (minimizing
 3    grooming or other interference with application sites), and observed weekly for tumor status.
 4    Histopathology was evaluated by two pathologists, both without knowledge of treatment group,
 5    who reached a consensus diagnosis in each case. In addition, the availability of time-of-tumor
 6    appearance and time of death for all animals provided a clearer characterization of the number at
 7    risk of development of tumors and the extent of exposure associated with tumor development,
 8    supporting more accurate estimation of lifetime cancer risk.
 9          The other lifetime studies provide support for the NIOSH study. Overall, study designs
10    varied in terms of number of exposure levels used (two to nine, compared with three in typical NTP
11    bioassays) and in number of mice per group (from ~17 to 100 mice/dose group, compared with
12    50 used in most NTP bioassays). While the largest studies would be expected to have greater
13    ability to detect low responses at low doses (e.g.. Schmidt etal.. 1973). studies conducted at similar
14    doses, but smaller group sizes, showed significant dose-response trends (e.g.. Poel. 1959).  For all of
15    the supporting studies, individual animal data were not available).
16          Some aspects of study design and conduct of the supporting studies were not consistently
17    reported. These study attributes included composition and purity of the test article, randomization
18    of animals to treatment groups, frequency of tumor evaluations, exact length of treatment period,
19    and blinding of treatment group from pathologists.  As each study was affected by a similar number
20    of omissions, possibly reflecting reporting practices at the time they were conducted, all but one
21    study (Poel. 1963) were included in dose-response modeling for comparison (see Appendix E.2.4 of
22    the Supplemental Information).
23          Poel (1963) exposed three strains of mice until natural death. The study authors did not
24    report length of time that the control mice were on study, nor the duration of exposure for mice
25    that did not develop tumors following treatment  Overall, while these three datasets support a
26    finding of dermal  carcinogenicity, they did not provide sufficient information to estimate the extent
27    of exposure associated with the observed tumor incidence and were not used for dose-response
28    modeling.

29    2.5.2. Dose-Response Analysis—Adjustments and Extrapolation Methods
30          Biologically based dose-response models for dermal exposure to benzo[a]pyrene are not
31    available. As with the oral and inhalation benzo[a]pyrene carcinogenicity data, benzo[a]pyrene
32    dermal exposure carcinogenicity data were generally characterized by earlier occurrence of tumors
33    and increased mortality with increasing exposure level. Individual data were available to apply a
34    time-to-tumor model (the multistage-Weibull  model, described in detail in Appendix E.2.1  of the
35    Supplemental Information) to the NIOSH study data, while each of the other lifetime dermal data
36    sets (Sivak etal.. 1997: Grimmer etal.. 1984: Habs etal.. 1984: Grimmer etal.. 1983: Habsetal..
37    1980: Schmahl etal.. 1977: Schmidt etal.. 1973: Roe etal.. 1970: Poel. 1963.19591 was modeled
38    using the multistage model. The following discussion describes the inputs for each model.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-40           DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          For both dose-response models, following EPA's Guidelines for Carcinogen Risk Assessment
 2    [Section 2.2.2.1.2: U.S. EPA. 2005a]. incidence of malignant or benign skin tumors was preferred for
 3    dose-response modeling, based on evidence that skin papillomas can develop into malignant skin
 4    tumors. For some studies, it was not clear whether related benign tumors were included in the
 5    total tumors reported [e.g.. Habs et al.. 1980). In the NIOSH study, one tumor in the mid-dose group
 6    that had been reported as a papilloma by Sivaketal. [1997] was listed as a keratoacanthoma in the
 7    full report [NIOSH. 1989]. Since it is not clear whether keratoacanthomas can develop into
 8    malignant skin tumors, two analyses were run for the NIOSH data set, one including the
 9    keratoacanthoma and one omitting it altogether (see Table E-23 of the Supplemental Information].
10    For the other studies, the most inclusive tabulation available of mice with benign or malignant skin
11    tumors, without double-counting, was used.
12          Concerning the appropriate dose measure, we have used continuous daily exposure (see
13    Preamble, Section 7.2] unless there are data indicating that dose-rate effects are relevant. The
14    exposure protocols for the selected bioassays varied between two or three applications per week.
15    Although environmental dermal exposure may more likely occur intermittently than oral or
16    inhalation exposures, due to interruption of exposure through bathing or washing of affected areas,
17    the dermal slope factor was derived for use with estimates of constant daily lifetime exposure.
18    Therefore, all administered doses were converted to TWA daily doses using the equation:
19
20       Average daily dose/day = (ug/application] x (number of applications/week 4- 7 days/week]
21
22          A particular consideration for time-to-tumor modeling (affecting only the NIOSH study] is
23    the time of tumor observation. The NIOSH data set provided the time of first appearance of skin
24    tumors for each animal. Skin tumors were classified as incidental for the purposes of modeling
25    because death generally occurred weeks later.
26          For the lifetime studies without individual animal data, some refinements of the dose-
27    response information were made where possible before applying the multistage model, in order to
28    approximate what would be accomplished had a time-to-tumor model been viable. These
29    refinements involved adjustments for incidence (i.e., numbers of animals at risk were reduced if
30    mortality  occurred before tumors first appeared] and dose (i.e., in the event of shortened exposure
31    duration due to 100% mortality in a dose group, an equivalent, lower lifetime exposure was
32    estimated]. These adjustments are described by study in Appendix E.2.3 of the Supplemental
33    Information.
34          The multistage-Weibull and multistage-cancer models were  then fit to the respective data
35    sets and BMDi0s/BMDLi0s estimated (details in Appendix E.2.3 of the Supplemental Information].
36    Because benzo[a]pyrene is expected to cause cancer via a mutagenic mode of action,  a linear
37    approach  to low dose extrapolation from the PODs (i.e., BMDLio] was used (U.S. EPA,  2005a] to
38    derive the candidate dermal slope factors.
               This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-41            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1          A toxicokinetic model to assist in cross-species scaling of benzo[a]pyrene dermal exposure
 2    was not available, nor was there any guidance to inform a basis for extrapolation of dermal dose-
 3    response in mice to humans. Several alternative approaches to interspecies extrapolation were
 4    developed that were linear, and that could be applied with equivalent results either before or after
 5    dose-response modeling.  Consequently, these adjustments were evaluated separately from the
 6    dose-response modeling and applied after derivation of a dermal slope factor for mice (see
 7    Section 2.5.4).

 8    2.5.3.  Derivation of the Dermal Slope Factor
 9          The results from modeling the dermal carcinogenicity data separated by sex are
10    summarized in Table 2-11 (see Tables E-23 and E-24 in the Supplemental Information for more
11    details).  Adequate fits to the NIOSH data were obtained using time-to-tumor modeling, whether or
12    not the keratoacanthoma in the mid-dose group was included. Excluding the keratoacanthoma, in
13    view of its seemingly random occurrence and the uncertainty whether such tumors would progress
14    to carcinomas, linear extrapolation from the BMDLio of 0.060 [J.g/day from the NIOSH study led to a
15    candidate dermal slope factor for mice of 1.7 per [ig/day.
16          Dermal slope factors calculated from the supporting studies (Sivaketal., 1997: Grimmer et
17    al.. 1984: Habsetal.. 1984: Grimmer etal.. 1983: Habs etal.. 1980: Schmahl etal.. 1977: Schmidt et
18    al.. 1973: Roe etal.. 1970: Poel. 1963.1959) using the multistage model and linear extrapolation
19    from the BMDLio values ranged from 0.25 to 1.8 per [ig/day, a roughly sevenfold range
20    (Table 2-11). Values  ranged from 0.9 to 1.7 per [ig/day for male mice, and from 0.25 to 0.67 per
21    [ig/day for female mice. These results suggest that some female mouse strains may be as sensitive
22    as some male mouse strains, but the associated uncertainties—e.g., increased extent of low-dose
23    extrapolation and incomplete exposure information—provide less support for many of these
24    values. Four female mice data sets were considered to be the most uncertain because of dose
25    ranges covered and incomplete information regarding length of exposure (Grimmer et al.. 1984:
26    Habs etal.. 1984: Grimmer etal.. 1983: Habs etal.. 19801 In particular, the data set reported by
27    Habs etal. (1984) yielded the highest but most uncertain result, with only two dose-response
28    points; the slope estimate is particularly affected by the characterization of the high exposure level.
29    There was insufficient information to conclude that males were more sensitive because both sexes
30    were not tested for any mouse strain.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-42            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
1
2
       Table 2-11.  Summary of dermal slope factor derivations, unadjusted for
       interspecies differences
Reference
Mouse
strain
Selected
model3
BMR
BMD
(Hg/d)
POD =
BMDL
(Hg/d)
Candidate
dermal
slope
factors'5
(Hg/dr1
Comments
Male mice
Sivak et al.
(1997); NIOSH
(1989)
Poel (1959)°

C3H/HeJ
C57L
Multistage-
Weibull2°
Multistage 3°
10%
10%
0.11
0.13
0.060
0.078
1.7
1.3
Well-conducted and
reported study, including
individual times on study
and individual tumor
diagnoses
Grouped survival data
reported
Female mice
Roe etal. (1970)

Schmidt et al.
(1973)
Schmidt et al.
(1973)
Schmahl et al.
(1977)
Habs etal. (1980)

(Habsetal.,
1984)

Grimmer et al.
(1983)
Grimmer et al.
(1984)
Swiss
Swiss
NMRI
NMRI
NMRI
NMRI
CFLP
CFLP
Multistage 2°
Multistage 3°
Multistage 2°
Multistage 2°
Multistage 4°
Multistage 1°
Multistage 1°
Log-logistic
10%
10%
10%
10%
10%
30%
10%
50%
10%
40%
70%
0.69
0.28
0.33
0.23
0.36
0.49
0.078
0.51
0.24
1.2
1.07
0.39
0.22
0.29
0.15
0.24
0.44
0.056
0.37
0.21
1.0
0.48
0.25
0.45
0.34
0.67
0.42
0.69
1.8
1.4
0.48
0.40
1.5
Grouped survival data
reported
No characterization of
exposure duration
No characterization of
exposure duration
No characterization of
exposure duration
Higher overall exposure
range; unclear overall
duration of exposure
No characterization of
exposure duration for high
exposure; high response at
lowest exposure limits
usefulness of low-dose
extrapolation
No characterization of
exposure duration
No characterization of
exposure duration; high
response at lowest exposure
limits usefulness of low-dose
extrapolation
3
4
5
6
aSee Appendix E.2.4 (Supplemental Information) for modeling details.
bUnadjusted for interspecies differences. Slope factor = R/BMDLR, where R is the BMR expressed as a fraction.
cHigh exposure groups with 100% mortality omitted prior to dose-response modeling.
               This document is a draft for review purposes only and does not constitute Agency policy,
                                                 2-43             DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene
 2    2.5.4. Dermal Slope Factor Cross-Species Scaling
 3          Different methodologies have been established for interspecies scaling of PODs used to
 4    derive oral slope factors and inhalation unit risks.  Cross-species adjustment of oral doses is based
 5    on allometric scaling using the % power of body weight. This adjustment accounts for more rapid
 6    distribution, metabolism, and clearance in small animals [U.S. EPA. 2005a). Cross-species
 7    extrapolation of inhalation exposures is based on standard dosimetry models that consider factors
 8    such as solubility, reactivity, and persistence [U.S. EPA. 1994] in addition to species differences in
 9    physiology. Although no established methodology exists to adjust for interspecies differences in
10    dermal toxicity at the point of contact, allometric scaling using body weight to the % power was
11    selected based on known species differences in dermal metabolism and penetration of
12    benzo[a]pyrene. In vitro skin permeation was highest in the mouse, compared to rat, rabbit, and
13    human, and was enhanced by induction of CYP enzymes [Kao etal.. 1985). Using this approach,
14    rodents and humans exposed to the same daily dose of a carcinogen, adjusted for BW3/4, would be
15    expected to have equal lifetime risks of cancer.
16          Alternative approaches were also evaluated, including:  (1) assuming that a given mass of
17    benzo[a]pyrene, applied daily, would pose the same risk in an animal  or in humans, regardless of
18    whether it is applied to a small surface area or to a larger surface area at a proportionately lower
19    concentration (e.g., no interspecies adjustment); (2) assuming that equal mass per day, if applied to
20    equal fractions of total skin surface will have similar cancer risks; and (3) assuming that risk is
21    directly proportional to dose expressed as mass per kg body weight per day. A comparison of these
22    alternatives is provided in Appendix E of the Supplemental Information.
23          The PODM derived from the NIOSH study [Sivak etal.. 1997: NIOSH. 1989] is adjusted to a
24    HED as follows:
25          POD HED (Jig/day)     = PODM (^g/day) x (BWH / BWM)3/4
26                               = 0.060 ^g/day x (70 kg / 0.035  kg)3/4
27                               = 17.9 ^g/day
28
29          The resulting PODnED is used to calculate the dermal slope factor for benzo[a]pyrene:
30
31          Dermal slope factor = BMR/PODHED = 0.1/(17.9 ng/day) = 0.006 per Hg/day.
32
33          Note that the dermal slope factor should only be used with lifetime human exposures
34    <18 |ig/day, the human equivalent of the PODM, because above  this level the dose-response
35    relationship is not expected to be proportional to the mass of benzo[a]pyrene applied.
36          Several assumptions are made in the use of this scaling method. First, it is assumed that the
37    toxicokinetic processes in the skin will scale similarly to interspecies differences in whole-body
38    toxicokinetics. Secondly, it is assumed that the risk at low doses ofbenzo[a]pyrene is linear.

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-44            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    Although one study indicates that at high doses of benzo[a]pyrene carcinogenic potency is related
 2    to mass applied per unit skin and not to total mass [Davies. 1969}. this may be due to promotional
 3    effects, such as inflammation, that are observed at high doses of benzo[a]pyrene.
 4          The dermal slope factor has been developed for a local effect and it is not intended to
 5    estimate systemic risk of cancer following dermal absorption of benzo[a]pyrene into the systemic
 6    circulation. Although some information suggests thatbenzo[a]pyrene metabolites can enter
 7    systemic circulation following dermal exposure in humans [Godschalk et al.. 1998a]. lifetime skin
 8    cancer bioassays that have included pathological examination of other organs have not found
 9    elevated incidences of tumors at distal sites [Habs etal., 1980: Schmahl etal., 1977: Schmidt etal.,
10    1973: Roe etal., 1970: Poel, 1959].  This may be because benzo[a]pyrene tends to bind to  targets
11    within the skin rather than enter the plasma receptor fluid (a surrogate measure of systemic
12    absorption) in in vitro human skin experiments. These data are consistent with metabolism of
13    benzo[a]pyrene to reactive metabolites within the viable layers of the skin [Wester et al.. 1990).
14    Some studies indicate that the fraction of benzo[a]pyrene left within the viable layers of the skin is
15    a large portion of the applied dose  [Moody et al.. 2007: Moody and Chu. 1995]. Taken together,
16    these data support the conclusion that the risk of skin cancer following dermal exposure likely
17    outweighs  cancer risks at distal organs.

18    2.5.5. Uncertainties in the Derivation of the Dermal Slope Factor
19          Table 2-12 summarizes uncertainties in the derivation of the dermal slope factor for
20    benzo[a]pyrene; further detail is provided in the following discussion. Uncertainty in the
21    recommended dermal slope factor is partly reflected in the range of POD values derived from the
22    modeled mouse skin tumor data sets: the lowest and highest BMDLio values listed in Table 2-11
23    show a sevenfold difference (0.056-0.39  [ig/day] in magnitude. However, many of the studies
24    considered had incomplete information concerning time on study, which would tend to lead to
25    underestimates of risk. Low-dose extrapolation uncertainty for several of these studies also
26    decreases confidence in their results.  Reliance on the NIOSH study, a relatively well-conducted
27    study with the low exposure levels having low early mortality, exposures continuing for full
28    lifetimes, and individual times on study minimizes this source of uncertainty.
29          Human dermal exposure to benzo[a]pyrene in the environment likely occurs predominantly
30    through soil contact  The available mouse dermal bioassays of benzo[a]pyrene relied on delivery of
31    benzo[a]pyrene to the skin in a solvent solution (typically acetone or toluene]. The use of a volatile
32    solvent likely results in a larger dose of benzo[a]pyrene available for uptake into the skin
33    (compared to soil]. Consequently, reliance on these studies may overestimate the risk of skin
34    tumors from benzo[a]pyrene contact through soil; however, cancer bioassays delivering
35    benzo[a]pyrene through a soil matrix are not available.
36          There is uncertainty in extrapolating from the intermittent exposures in the mouse assays
37    to daily exposure scenarios. All of the dermal bioassays that were considered treated animals
38    2-3 times/week. This assessment makes the assumption that risk is proportional to total
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-45            DRAFT—DO NOT CITE OR QUOTE

-------
                                                         Toxicological Review ofBenzo[a]pyrene

 1    cumulative exposure. However, this may overestimate risk if duration-adjusted doses are below
 2    doses that saturate or diminish detoxifying metabolic steps.
 3          The relative impact of a particular vehicle on benzo[a]pyrene carcinogenicity and the
 4    relative sensitivity of male and female mice to benzo [a]pyrene exposure was difficult to ascertain in
 5    the supporting lifetime studies. Overall, the study designs for these studies included different
 6    mouse strains, sexes, and vehicles, but for any given mouse strain, even across multiple studies,
 7    only one sex and only one vehicle was tested.  Thus, it was not possible to evaluate the relative
 8    impact of particular vehicles onbenzo[a]pyrene carcinogenicity, or the relative sensitivity of male
 9    and female mice.
10          The available data were not useful to determine which animal species may be the best
11    surrogate for human dermal response to benzo[a]pyrene. In extrapolation of the animal dermal
12    information to humans, the assumption is made that equal lifetime risks of cancer would follow
13    from exposure to the same daily dose adjusted for BW3/4. Qualitatively, the toxicokinetics and
14    toxicodynamics in mouse and human skin appear to be similar [Knafla etal.. 2011: Bickers etal..
15    1984]. Specifically, all of the activation pathways implicated in benzo[a]pyrene carcinogenicity
16    have been observed in mouse and human skin, and associations have been made between the
17    spectra of mutations in tumor tissues from benzo[a]pyrene-exposed animals and humans exposed
18    to complex PAH mixtures containing benzo[a]pyrene (see Section 1.1.5).
19          The dermal slope factor for benzo[a]pyrene is based on skin cancer and does not represent
20    systemic cancer risk from dermal exposure. It is unclear whether dermal exposure to
21    benzo[a]pyrene would result in elevated risk of systemic tumors. Some studies in humans suggest
22    that although the skin may be responsible for a "first pass" metabolic effect, benzo[a]pyrene-
23    specific adducts have been detected in white blood cells (WBCs) following dermal exposure to
24    benzo[a]pyrene, indicating that dermally applied benzo[a]pyrene enters systemic circulation
25    [Godschalketal.. 1998a). Although none of the lifetime dermal bioassays in mice, which included
26    macroscopic examination of internal organs, reported an elevation of systemic tumors in
27    benzo[a]pyrene-treated mice compared to controls [Higginbothametal.. 1993: Habs etal.. 1980:
28    Schmahl etal.. 1977: Schmidt etal..  1973: Roe etal.. 1970: Poel. 1959]. most of these studies
29    attempted to remove animals with grossly observed skin tumors from the study before the death of
30    the animal, possibly minimizing the development of more distant tumors with longer latency. The
31    risk of benzo [a]pyrene-induced point-of-contact tumors in the skin possibly competes with
32    systemic risk of tumors.  Currently, the potential contribution of dermally absorbed benzo[a]pyrene
33    to systemic cancer risk is unclear.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                2-46            DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene
1
2
       Table 2-12. Summary of uncertainties in the derivation of cancer risk values
       for benzo[a]pyrene dermal slope factor
            Consideration and
        impact on cancer risk value
                                        Decision
                              Justification and discussion
     Selection of data set
     4, dermal slope factor if alternative
     data set were selected
                                 NIOSH (1989)
                      Study included lowest doses among available
                      studies (where intercurrent mortality was less
                      likely to impact the number at risk).
     Selection of target organ
     No dermal slope factor if skin tumor
     studies not used
                                 Selection of skin tumors
                      Skin tumors were replicated in numerous studies
                      of male or female mice. No studies were
                      available indicating that other tumors occur
                      following  dermal exposure.
Selection of dose metric
Alternatives could 4, or
factor
                           slope
Administered dose, as
TWA in u.g/d
Experimental evidence supports a role for
metabolism in toxicity, but actual responsible
metabolites are not identified.
     Interspecies extrapolation
     Alternatives could 4, or ^ slope
     factor
                                 Total daily dose scaled
                                 by BW3/4
                      Alternatives discussed in Appendix E. An
                      established methodology does not exist to adjust
                      for interspecies differences in dermal toxicity at
                      the point of contact. Benzo[a]pyrene
                      metabolism is known to occur in the dermal
                      layer. Viewing the skin as an organ, and without
                      evidence to the contrary, metabolic processes
                      were assumed to scale allometrically.
Dose-response modeling
Alternatives could 4, or 1
factor
                           slope
Multistage-Weibull
model
No biologically based models for benzo[a]pyrene
were available. The multistage-Weibull model is
consistent with biological processes,
incorporates timecourse information, and is
preferred for IRIS cancer assessments when
individual data are available (Gehlhaus et al.,
2011).
     Low-dose extrapolation
     •^ cancer risk estimate would be
     expected with the application of
     nonlinear low-dose extrapolation
                                 Linear extrapolation
                                 from POD (based on
                                 mutagenic mode of
                                 action)
                      Available mode of action data support linearity
                      (mutagenicity is a primary mode of action of
                      benzo[a]pyrene).
     Sensitive subpopulations
     1" dermal slope factor to unknown
     extent
                                 ADAFs are
                                 recommended for early
                                 life exposures
                      No chemical-specific data are available to
                      determine the range of human toxicodynamic
                      variability or sensitivity.
4    2.5.6.  Previous IRIS Assessment: Dermal Slope Factor
5           A dermal slope factor for benzo[a]pyrene was not previously available on IRIS.

6    2.6.  APPLICATION OF AGE-DEPENDENT ADJUSTMENT FACTORS (ADAFS)
7           Based on sufficient support in laboratory animals and relevance to humans, benzo[a]pyrene
8    is determined to be carcinogenic by a mutagenic mode of action. According to the Supplemental
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                   2-47             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
Guidance for Assessing Susceptibility from Early Life Exposure to Carcinogens ("Supplemental
Guidance"} [U.S. EPA. 2005b). individuals exposed during early life to carcinogens with a mutagenic
mode of action are assumed to have increased risk for cancer.  The oral slope factor of 1 per
mg/kg-day, inhalation unit risk of 0.6 per mg/m3, and dermal slope factor of 0.006 per  [J.g/day for
benzo[a]pyrene, calculated from data applicable to adult exposures, do not reflect presumed early
life susceptibility to this chemical. Although chemical-specific data exist for benzo[a]pyrene that
quantitatively demonstrate increased early life susceptibility to cancer [Vesselinovitchetal.. 1975].
these data were not considered sufficient to develop separate risk estimates for childhood
exposure, as they used acute i.p. exposures [U.S. EPA, 2005b].  In the absence of adequate chemical-
specific data to evaluate differences in age-specific susceptibility, the Supplemental Guidance [U.S.
EPA. 2005b) recommends that ADAFs be applied in estimating cancer risk.
       The Supplemental Guidance [U.S. EPA. 2005b) establishes ADAFs for three specific age
groups. These ADAFs and their corresponding age groupings are: 10 for individuals exposed at
<2 years of age, 3 for exposed individuals at 2-<16 years of age, and 1 for exposed individuals
>16 years of age. The 10- and 3-fold adjustments are combined with age-specific exposure
estimates when estimating cancer risks from early life (<16 years of age) exposures to
benzo[a]pyrene. To  illustrate the use of the ADAFs established in the Supplemental Guidance [U.S.
EPA. 2005b). sample calculations are presented for three exposure duration scenarios, including
full lifetime, assuming a constant benzo[a]pyrene exposure of 0.001 mg/kg-day (Table 2-13).

       Table 2-13.  Sample application of ADAFs for the estimation of benzo[a]pyrene
       cancer risk following lifetime (70-year) oral exposure
Age group
0-<2 yrs
2-<16 yrs
>16 yrs
ADAF
10
3
1
Unit risk (per
mg/kg-d)
1
1
1
Sample exposure
concentration (mg/kg-d)
0.001
0.001
0.001
Duration
adjustment
2 yrs/70 yrs
14 yrs/70 yrs
54 yrs/70 yrs
Total risk
Cancer risk for age-
specific exposure
period
0.0003
0.0006
0.0008
0.002
22
23
24
25
26
27
28
29
30
       The example exposure duration scenarios include full lifetime exposure (assuming a
70-year lifespan). Table 2-13 lists the four factors (ADAFs, cancer risk estimate, assumed exposure,
and duration adjustment) that are needed to calculate the age-specific cancer risk based on the
early age-specific group.  The cancer risk for each age group is the product of the four factors in
columns 2-5.  Therefore, the cancer risk following daily benzo[a]pyrene oral exposure in the age
group 0-<2 years is the product of the values in columns 2-5 or 10 x 1  x 0.001 x 2/70 = 3 x 1Q-4.
The cancer risk for specific exposure duration scenarios that are listed in the last column are added
together to get the total risk. Thus, a 70-year (lifetime) risk estimate for continuous exposure to
                This document is a draft for review purposes only and does not constitute Agency policy,
                                                2-48            DRAFT—DO NOT CITE OR QUOTE

-------
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13

14
15
                                                   Toxicological Review ofBenzo[a]pyrene

0.001 mg/kg-day benzo[a]pyrene is 2 x 1Q-3, which is adjusted for early-life susceptibility and
assumes a 70-year lifetime and constant exposure across age groups.
       In calculating the cancer risk for a 30-year constant exposure to benzo[a]pyrene at an
exposure level of 0.001 mg/kg-day for ages 0-30 years, the duration adjustments would be 2/70,
14/70, and 14/70, and the age-specific risks for the three age groups would be 3 x 1Q-4, 6 x 1Q-4,
and 2 x 10~4, which would result in a total risk estimate of 1 x 1Q-3.
       In calculating the cancer risk for a 30-year constant exposure to benzo[a]pyrene at an
exposure level of 0.001 mg/kg-day for ages 20-50 years, the duration adjustments would be 0/70,
0/70, and 30/70. The age-specific risks for the three groups are 0, 0, and 4 x 10~4, which would
result in a total risk estimate of 4 x 10~4.
       Consistent with the approaches for the oral route of exposure (Table 2-13), the ADAFs
should also be applied when assessing cancer risks for subpopulations with early life exposures to
benzo[a]pyrene via the inhalation and dermal routes (presented in Tables 2-14 and 2-15).

       Table 2-14. Sample application of ADAFs for the estimation of benzo[a]pyrene
       cancer risk following lifetime (70-year) inhalation exposure
Age group
0-<2 yrs
2-<16 yrs
>16 yrs
ADAF
10
3
1
Unit risk
(per ug/m3)
6 x 10"4
6 x 10"4
6 x 10"4
Sample exposure
concentration (ug/m3)
0.1
0.1
0.1
Duration
adjustment
2 yrs/70 yrs
14 yrs/70 yrs
54 yrs/70 yrs
Total risk
Cancer risk for
age-specific
exposure period
0.00002
0.00004
0.00005
0.00010
16
17
       Table 2-15. Sample application of ADAFs for the estimation of benzo[a]pyrene
       cancer risk following lifetime (70-year) dermal exposure
Age group
0-<2 yrs
2-<16 yrs
>16 yrs
ADAF
10
3
1
Unit risk
(per ng/d)
0.006
0.006
0.006
Sample exposure
concentration (ng/d)
0.001
0.001
0.001
Duration
adjustment
2 yrs/70 yrs
14 yrs/70 yrs
54 yrs/70 yrs
Total risk
Cancer risk for
age-specific
exposure period
2 x 10"6
4 x 10"6
5 x 10"6
1 x 10"5
18
               This document is a draft for review purposes only and does not constitute Agency policy,
                                               2-49            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene
 1
 2    REFERENCES
 3    Abedi-Ardekani, B; Kamangar, F; Hewitt, SM; Hainaut, P; Sotoudeh, M; Abnet, CC; Taylor, PR; Boffetta, P;
 4           Malekzadeh, R; Dawsey, SM. (2010). Polycyclic aromatic hydrocarbon exposure in oesophageal
 5           tissue and risk of oesophageal squamous cell carcinoma in north-eastern Iran. Gut 59:  1178-
 6           1183. http://dx.doi.org/10.1136/gut.2010.210609
 7    Aklillu, E; 0vreb0, S; Botnen, IV; Otter, C; Ingelman-Sundberg, M. (2005). Characterization of common
 8           CYP1B1 variants with different capacity for benzo[a]pyrene-7,8-dihydrodiol epoxide formation
 9           from benzo[a]pyrene. Cancer Res 65: 5105-5111. http://dx.doi.org/10.1158/0008-5472.CAN-05-
10           0113
11    Albert. RE; Miller. ML; Cody. T; Andringa. A; Shukla. R; Baxter. CS. (1991). Benzo[a]pyrene-induced skin
12           damage and tumor promotion in the mouse. Carcinogenesis 12: 1273-1280.
13           http://dx.doi.0rg/10.1093/carcin/12.7.1273
14    Albert, RE; Miller, ML; Cody, TE; Talaska, G; Underwood, P; Andringa, A. (1996). Epidermal cytokinetics,
15           DNA adducts, and dermal inflammation in the mouse skin in response to repeated
16           benzo[a]pyrene exposures. Toxicol Appl Pharmacol 136: 67-74.
17    Alexandrov. K; Cascorbi. I; Rojas. M; Bouvier. G; Kriek. E; Bartsch. H. (2002). CYP1A1 and GSTM1
18           genotypes affect benzo[a]pyrene DNA adducts in smokers' lung: comparison with
19           aromatic/hydrophobic adduct formation [Review]. Carcinogenesis 23: 1969-1977.
20    Alexandrov, K; Rojas, M; Geneste, O; Castegnaro, M; Camus, AM; Petruzzelli, S; Giuntini, C; Bartsch, H.
21           (1992). An improved fluorometric assay for dosimetry of benzo(a)pyrene diol-epoxide-DNA
22           adducts in smokers' lung: Comparisons with total bulky adducts and aryl hydrocarbon
23           hydroxylase activity. Cancer Res 52: 6248-6253.
24    Allan, LL; Schlezinger, JJ; Shansab, M; Sherr, PH. (2006). CYP1A1 in polycyclic aromatic hydrocarbon-
25           induced B lymphocyte growth suppression. Biochem  Biophys Res Commun 342: 227-235.
26           http://dx.doi.0rg/10.1016/i.bbrc.2006.01.131
27    Amos. Cl; Xu. W; Spitz. MR. (1999). Is there a genetic basis for lung cancer susceptibility? [Review].
28           Recent Results Cancer Res 151: 3-12.
29    Andreassen. A; Kure. EH; Nielsen. PS; Autrup. H; Haugen. A. (1996). Comparative synchronous
30           fluorescence spectrophotometry and 32P-postlabeling analysis of PAH-DNA adducts in human
31           lung and the relationship to TP53 mutations. Mutat Res 368: 275-282.
32    Arafa, HM; Aly, HA; Abd-Ellah, MF; EI-Refaey, HM. (2009). Hesperidin attenuates benzo[alpha]  pyrene-
33           induced testicular toxicity in rats via regulation of oxidant/antioxidant balance.  Toxicol Ind
34           Health 25: 417-427. http://dx.doi.org/10.1177/0748233709106624
35    Archibong, AE; Inyang, F; Ramesh, A; Greenwood, M; Nayyar, T; Kopsombut, P; Hood, DB; Nyanda, AM.
36           (2002). Alteration of pregnancy related hormones and fetal survival  in F-344 rats exposed by
37           inhalation to benzo(a)pyrene. Reprod Toxicol 16: 801-808.
38    Archibong, AE; Ramesh, A; Niaz,  MS; Brooks, CM; Roberson, SI; Lunstra, DP.  (2008). Effects of
39           benzo(a)pyrene on intra-testicular function in F-344 rats. Int J Environ Res Public Health 5: 32-
40           40. http://dx.doi.org/10.3390/iierph5010032
41    Armstrong, B; Tremblay, C; Baris, D; Theriault, G. (1994). Lung cancer mortality and polynuclear aromatic
42           hydrocarbons: A case-cohort study of aluminum production workers in Arvida, Quebec, Canada.
43           Am J Epidemiol 139: 250-262.

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

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1    Armstrong, BG; Gibbs, G. (2009). Exposure-response relationship between lung cancer and polycyclic
 2           aromatic hydrocarbons (PAHs). Occup Environ Med 66: 740-746.
 3           http://dx.doi.org/10.1136/oem.2008.043711
 4    Arthur D Little (Arthur D Little Inc.). (1989). Assessment of the cocarcinogenic/promoting activity of
 5           asphalt fumes. National Institute for Occupational Safety and Health.
 6           http://www.ntis. gov/search/product.aspx?ABBR=PB91110213
 7    Asgharian, B. (2004). A model of deposition of hygroscopic particles in the human lung. Aerosol Sci
 8           Technol 38: 938-947. http://dx.doi.org/10.1080/027868290511236
 9    ASRM (American Society of Reproductive Medicine). (2008). The clinical utility of sperm  DNA integrity
10           testing [Review]. Fertil Steril 90: S178-S180. http://dx.doi.Org/10.1016/i.fertnstert.2008.08.054
11    ATS PR (Agency for Toxic Substances and Disease Registry). (1995). Toxicological profile for polycyclic
12           aromatic hydrocarbons [ATSDR Tox Profile]. Washington, DC: U.S. Department of Health and
13           Human Services, http://www.atsdr.cdc.gov/toxprofiles/tp69.pdf
14    ATSDR (Agency for Toxic Substances and Disease Registry). (2011). The priority list of hazardous
15           substances that will be the subject of toxicological profiles.
16           http://www.atsdr.cdc.gov/SPL/index.html
17    Baan, R; Grosse, Y; Straif, K; Secretan, B; El Ghissassi, F; Bouvard, V; Benbrahim-Tallaa, L; Guha, N;
18           Freeman, C; Galichet, L; Cogliano, V. (2009). A review of human carcinogens-part F: Chemical
19           agents and related occupations. Lancet Oncol  10: 1143-1144. http://dx.doi.org/10.1016/S1470-
20           2045(09)70358-4
21    Balu. N; Padgett. WT; Lambert. GR; Swank. AE; Richard. AM; Nesnow. S. (2004).  Identification and
22           characterization of novel stable deoxyguanosine and deoxyadenosine adducts of
23           benzo[a]pyrene-7,8-quinone from reactions at physiological pH. Chem Res Toxicol 17: 827-838.
24           http://dx.doi.org/10.1021/tx034207s
25    Barry, G; Cook, JW; Haslewood, GAD; Hewett, CL; Hieger, I; Kennaway,  EL. (1935). The Production of
26           Cancer by Pure Hydrocarbons.-Part III. Proc Biol Sci 117: 318-351.
27           http://dx.doi.org/10.1098/rspb.1935.0032
28    Bartsch, H; Nair, U; Risch, A; Rojas, M; Wikman, H; Alexandrov, K. (2000).  Genetic polymorphism of CYP
29           genes, alone or in combination, as a risk modifier of tobacco-related cancers [Review]. Cancer
30           Epidemiol Biomarkers Prev 9: 3-28.
31    Bartsch, H; Rojas, M; Alexandrov, K;  Risch, A. (1998). Impact of adduct determination  on the assessment
32           of cancer susceptibility. In M Schwab; HJ Senn; P Schlag; K Munk;  H Rabes (Eds.), Recent results
33           in cancer research, v!54: Genes and environment in cancer (pp. 86-96).  New York, NY: Springer.
34    Bartsch, H; Rojas, M; Nair, U; Nair, J; Alexandrov, K. (1999). Genetic cancer susceptibility and DNA
35           adducts: studies in smokers, tobacco chewers, and coke oven workers. Cancer Detect Prev 23:
36           445-453.
37    Beland, F; Gulp, S. (1998). Chronic bioassay of two composite samples from selected manufactured gas
38           plant waste sites [unpublished report]. (Technical Report 6722.02). Jefferson, AK: National
39           Center for Toxicological Research.
40    Benbrahim-Tallaa. L; Baan. RA; Grosse. Y; Lauby-Secretan. B; El ghissassi. F; Bouvard. V; Guha. N; Loomis.
41           D; Straif. K. (2012). Carcinogenicity of diesel-engine and gasoline-engine exhausts and some
42           nitroarenes.  Lancet Oncol 13: 663-664. http://dx.doi.org/10.1016/S1470-2045(12)70280-2
43    Benjamin, H; Storkson, J; Tallas, PG;  Pariza, MW. (1988). Reduction of benzo[a]pyrene-induced
44           forestomach neoplasms in mice given nitrite and dietary soy sauce. Food Chem Toxicol 26: 671-
45           678. http://dx.doi.org/10.1016/0278-6915(88)90066-X
46    Bennett, WP; Alavanja, MC; Blomeke, B; Vahakangas, KH; Castren, K; Welsh, JA;  Bowman,  ED; Khan,  MA;
47           Flieder, DB; Harris, CC. (1999). Environmental tobacco smoke, genetic susceptibility, and risk of
48           lung cancer in never-smoking women. J Natl Cancer Inst 91: 2009-2014.
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-2             DRAFT—DO NOT  CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Berenblum, I; Haran, N. (1955). The influence of croton oil and of polyethylene glycol-400 on
 2           carcinogenesis in theforestomach of the mouse. Cancer Res 15: 510-516.
 3    Berger, J; Manz, A. (1992). Cancer of the stomach and the colon-rectum among workers in a coke gas
 4           plant. Am J Ind Med 22: 825-834. http://dx.doi.org/10.1002/aiim.4700220605
 5    Biancifiori, C; Caschera, F; Giornelli-Santulli, F; Bucciarelli, E. (1967). The action of oestrone and four
 6           chemical carcinogens in intact and ovariectomized BALB/c/Cb/Se mice. Br J Cancer 21: 452-459.
 7    Bickers, PR; Mukhtar, H; Dutta-Choudhury, T; Marcelo, CL; Voorhees, JJ. (1984). Aryl hydrocarbon
 8           hydroxylase, epoxide hydrolase, and benzo[a]-pyrene metabolism in human epidermis:
 9           Comparative studies in normal subjects and patients with psoriasis. J Invest Dermatol 83: 51-56.
10           http://dx.doi.org/10.llll/1523-1747.epl2261680
11    Bingham, E; Falk, HL (1969). Environmental carcinogens. The modifying effect of cocarcinogens on the
12           threshold response. Arch Environ Health  19: 779-783.
13    Binkova, B; Chvatalova, I; Lnenickova, Z; Milcova, A; Tulupova, E; Farmer, PB; Sram, RJ. (2007). PAH-DNA
14           adducts in environmentally exposed population in relation to metabolic and DNA repair gene
15           polymorphisms. Mutat Res-Fundam  Mol  Mech Mutagen 620: 49-61.
16           http://dx.doi.0rg/10.1016/i.mrfmmm.2007.02.022
17    Blaha, L; Kapplova, P; Vondracek, J; Upham, B; Machala, M. (2002). Inhibition of gap-junctional
18           intercellular communication by environmentally occurring polycyclic aromatic hydrocarbons.
19           Toxicol Sci 65: 43-51.  http://dx.doi.Org/10.1093/toxsci/65.l.43
20    Blanton. RH; Lyte. M; Myers. MJ; Bick. PH. (1986). Immunomodulation by polyaromatic hydrocarbons in
21           mice and murine cells. Cancer Res 46: 2735-2739.
22    Boffetta. P; Burstyn. I; Partanen. T; Kromhout. H; Svane. O; Langard. S; Jarvholm. B; Frentzel-Beyme. R;
23           Kauppinen. T; Stucker. I; Shaham. J; Heederik. D; Ahrens. W; Bergdahl. IA; Cenee. S; Ferro. G;
24           Heikkila, P; Hooiveld,  M; Johansen, C; Randem, BG; Schill, W.  (2003). Cancer mortality among
25           European asphalt workers: an international epidemiological study II Exposure to bitumen fume
26           and other agents. Am J Ind Med 43: 28-39.  http://dx.doi.org/10.1002/aiim.10182
27    Boffetta, P; Jourenkova,  N; Gustavsson, P. (1997). Cancer risk from occupational and environmental
28           exposure to polycyclic aromatic hydrocarbons [Review]. Cancer Causes Control 8: 444-472.
29           http://dx.doi.Org/10.1023/A:1018465507029
30    Bolton, JL; Trush, MA; Penning, TM; Dryhurst, G;  Monks, TJ. (2000). Role of quinones in toxicology
31           [Review]. Chem  Res Toxicol 13: 135-160.  http://dx.doi.org/10.1021/tx9902082
32    Borini, A; Tarozzi, N; Bizzaro, D; Bonu, MA; Fava,  L;  Flamigni, C; Coticchio, G. (2006). Sperm DNA
33           fragmentation: paternal effect on early post-implantation embryo development in ART. Hum
34           Reprod 21: 2876-2881. http://dx.doi.org/10.1093/humrep/del251
35    Borman, SM; Christian, PJ; Sipes, IG; Hoyer, PB. (2000). Ovotoxicity in female Fischer rats and B6 mice
36           induced by low-dose exposure to three polycyclic aromatic hydrocarbons: comparison through
37           calculation of an ovotoxic index. Toxicol Appl Pharmacol 167: 191-198.
38           http://dx.doi.org/10.1006/taap.2000.9006
39    Bosetti. C; Boffetta. P; La Vecchia. C. (2007).  Occupational exposures to polycyclic aromatic
40           hydrocarbons, and respiratory and urinary tract cancers: a quantitative review to  2005 [Review].
41           Ann Oncol 18: 431-446. http://dx.doi.org/10.1093/annonc/mdll72
42    Bostrom. CE; Gerde. P; Hanberg. A; Jernstrom. B; Johansson. C; Kyrklund. T; Rannug. A; Tornqvist. M;
43           Victorin, K; Westerholm, R. (2002). Cancer risk assessment, indicators, and guidelines for
44           polycyclic aromatic hydrocarbons in  the ambient air [Review]. Environ Health Perspect 110
45           Suppl 3: 451-488.
46    Bouayed, J; Bohn, T; Tybl, E; Kiemer, AK; Soulimani, R. (2012). Benzo[a]pyrene-lnduced Anti-Depressive-
47           like Behaviour in Adult Female Mice: Role of Monoaminergic Systems. Basic & Clinical

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

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1           Pharmacology & Toxicology Online Pharmacology Online 110: 544-550.
 2           http://dx.doi.0rg/10.llll/i.1742-7843.2011.00853.x
 3    Bouayed, J; Desor, F; Rammal, H; Kiemer, AK; Tybl, E; Schroeder, H; Rychen, G; Soulimani, R. (2009a).
 4           Effects of lactational exposure to benzo[alpha]pyrene (B[alpha]P) on postnatal
 5           neurodevelopment, neuronal receptor gene expression and behaviour in mice. Toxicology 259:
 6           97-106. http://dx.doi.0rg/10.1016/i.tox.2009.02.010
 7    Bouayed, J; Desor, F; Soulimani, R. (2009b). Subacute oral exposure to benzo[alpha]pyrene (B[alpha]P)
 8           increases aggressiveness and affects consummatory aspects of sexual behaviour in male mice. J
 9           Hazard Mater 169: 581-585. http://dx.doi.Org/10.1016/i.ihazmat.2009.03.131
10    Broekman, BF; Chan, YH; Chong, YS; Quek, SC; Fung, D; Low, YL; Ooi, YP; Gluckman, PD; Meaney, MJ;
11           Wong, TY; Saw,  SM. (2009). The influence of birth size on intelligence in healthy children.
12           Pediatrics  123: e!011-el016. http://dx.doi.org/10.1542/peds.2008-3344
13    Brown, J. R.; Thornton, JL (1957). Percivall Pott (1714-1788) and chimney sweepers' cancer of the
14           scrotum. Br J Ind Med 14: 68-70.
15    Brown, LA; Khousbouei,  H; Goodwin, JS; Irvin-Wilson, CV; Ramesh, A; Sheng, L; Mcca I lister, MM; Jiang,
16           GC; Aschner, M; Hood, DB. (2007). Down-regulation of early ionotrophic glutamate receptor
17           subunit developmental expression as a mechanism for observed plasticity deficits following
18           gestational exposure to benzo(a)pyrene. Neurotoxicology 28: 965-978.
19           http://dx.doi.0rg/10.1016/i.neuro.2007.05.005
20    Brune. H; Deutsch-Wenzel.  RP; Habs. M; Ivankovic. S; Schmahl. D. (1981). Investigation of the
21           tumorigenic response to benzo(a)pyrene in aqueous caffeine solution applied orally to Sprague-
22           Dawley rats. J Cancer Res Clin Oncol 102: 153-157. http://dx.doi.org/10.1007/BF00410666
23    Bryan. WR; Shimkin. MB. (1943). Quantitative analysis of dose-response data obtained with three
24           carcinogenic hydrocarbons in strain C3H male mice. J Natl Cancer Inst 3: 503-531.
25    Buening,  MK; Wislocki, PG;  Levin, W; Yagi, H; Thakker, PR; Akagi, H; Koreeda, M; Jerina, DM; Conney,
26           AH. (1978). Tumorigenicity of the optical enantiomers of the diastereomeric benzo[a]pyrene
27           7,8-diol-9,10-epoxides in newborn mice: exceptional activity of (+)-7"beta", 8"alpha"-dihydroxy-
28           9"alpha",10"alpha"-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene. PNAS 75: 5358-5361.
29    Burdick. AD; Davis. JW; Liu.  KJ; Hudson. LG; Shi. H; Monske. ML; Burchiel. SW. (2003). Benzo(a)pvrene
30           quinones increase cell proliferation,  generate reactive oxygen species, and transactivate the
31           epidermal growth factor receptor in breast epithelial cells.  Cancer Res 63: 7825-7833.
32    Burns-Naas, LA; Hastings, KL; Ladies, GS; Makris, SL; Parker, GA; Holsapple,  MP. (2008). What's so special
33           about the  developing immune system? [Review]. Int J Toxicol 27: 223-254.
34           http://dx.doi.org/10.1080/10915810801978110
35    Burstyn, I; Kromhout, H; Johansen, C; Langard, S; Kauppinen, T; Shaham, J;  Ferro, G; Boffetta, P. (2007).
36           Bladder cancer incidence and exposure to polycyclic aromatic hydrocarbons among asphalt
37           pavers. Occup Environ Med 64: 520-526. http://dx.doi.org/10.1136/oem.2006.029801
38    Burstyn. I; Kromhout. H; Kauppinen. T; Heikkila. P; Boffetta. P. (2000). Statistical modelling of the
39           determinants of historical exposure to bitumen and polycyclic aromatic hydrocarbons among
40           paving workers. Ann Occup Hyg 44: 43-56.
41    Burstyn. I; Kromhout. H; Partanen. T; Svane. O; Langard. S; Ahrens.  W; Kauppinen. T; Stucker. I; Shaham.
42           J; Heederik. D; Ferro. G; Heikkila. P; Hooiveld. M; Johansen. C; Randem. BG; Boffetta. P. (2005).
43           Polycyclic aromatic hydrocarbons and fatal  ischemic heart disease.  Epidemiology 16: 744-750.
44           http://dx.doi.org/10.1097/01.ede.0000181310.65043.2f
45    Busby, WF, Jr; Goldman, ME; Newberne, PM; Wogan, GN. (1984). Tumorigenicity of fluoranthene in a
46           newborn mouse lung adenoma bioassay. Carcinogenesis 5: 1311-1316.
47           http://dx.doi.0rg/10.1093/carcin/5.10.1311

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

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Busby, WF, Jr; Stevens, EK; Martin, CM; Chow, FL; Garner, RC. (1989). Comparative lung tumorigenicity of
 2           parent and mononitro-polynuclear aromatic hydrocarbons in the BLU:Ha newborn mouse assay.
 3           Toxicol Appl  Pharmacol 99: 555-563. http://dx.doi.org/10.1016/0041-008X(89)90162-2
 4    Cal/EPA (California Environmental Protection Agency). (2010).  Public health goals for chemicals in
 5           drinking water. Benzo(a)pyrene. http://oehha.ca.gov/water/phg/pdf/091610Benzopyrene.pdf
 6    Car0e, TK; Ebbeh0j, NE; Wulf, HC; Agner, T. (2013). Occupational skin cancer may be underreported. 60:
 7           A4624.
 8    Cavalieri, E; Rogan, E; Cremonesi, P; Higginbotham, S; Salmasi, S. (1988a). Tumorigenicity of 6-
 9           halogenated derivatives of benzo[a]pyrene in mouse skin and rat mammary gland. J Cancer Res
10           Clin Oncol 114: 10-15. http://dx.doi.org/10.1007/BF00390479
11    Cavalieri, E; Rogan, E; Higginbotham, S; Cremonesi, P; Salmasi, S. (1988b). Tumor-initiating activity in
12           mouse skin and carcinogenicity in  rat mammary gland  of fluorinated derivatives of
13           benzo[a]pyrene and 3-methylcholanthrene. J Cancer Res Clin Oncol 114: 16-22.
14    Cavalieri, E; Rogan, E; Sinha, D. (1988c). Carcinogenicity of aromatic hydrocarbons directly applied to rat
15           mammary gland. J Cancer Res Clin Oncol 114: 3-9. http://dx.doi.org/10.1007/BF00390478
16    Cavalieri, E; Rogan, E; Toth, B; Munhall, A. (1981). Carcinogenicity of the environmental pollutants
17           cyclopenteno-[cd]pyrene and cyclopentano[cd]pyrene in mouse skin. Carcinogenesis 2: 277-281.
18    Cavalieri, EL; Devanesan,  PD; Rogan, EG. (1988d).  Radical cations in the horseradish peroxidase and
19           prostaglandin H synthase mediated metabolism and binding of benzo[a]pyrene to
20           deoxyribonucleic acid. Biochem Pharmacol 37: 2183-2187.  http://dx.doi.org/10.1016/0006-
21           2952(88)90579-5
22    Cavalieri. EL; Higginbotham. S; Ramakrishna. NV; Devanesan. PD; Todorovic. R; Rogan. EG; Salmasi. S.
23           (1991). Comparative dose-response tumorigenicity studies of dibenzo[alpha,l]pyrene versus
24           7,12-dimethylbenz[alpha]anthracene, benzo[alpha]pyrene and two dibenzo[alpha,l]pyrene
25           dihydrodiols in mouse skin and rat mammary gland. Carcinogenesis 12: 1939-1944.
26           http://dx.doi.org/10.1093/carcin/12.10.1939
27    Cavalieri, EL; Rogan,  EG. (1992). The approach to understanding aromatic hydrocarbon Carcinogenesis.
28           The central role of radical cations in metabolic activation [Review]. Pharmacol Ther 55: 183-199.
29    Cavalieri, EL; Rogan,  EG. (1995). Central role of radical cations in metabolic activation of polycyclic
30           aromatic hydrocarbons [Review]. Xenobiotica 25: 677-688.
31           http://dx.doi.org/10.3109/00498259509061885
32    Cavalieri, EL; Rogan,  EG; Cremonesi, P; Devanesan, PD. (1988e). Radical cations as precursors in the
33           metabolic formation of quinones from benzo[a]pyrene and 6-fluorobenzo[a]pyrene. Fluoro
34           substitution  as a probe for one-electron oxidation in aromatic substrates. Biochem Pharmacol
35           37: 2173-2182.
36    CDC (Centers for Disease Control and Prevention). (2004). The  health consequences of smoking: A report
37           of the Surgeon General. Washington, DC:  U.S. Department  of Health and Human Services.
38           http://www.surgeongeneral.gov/library/smokingconsequences/
39    Chakravarti.  D; Pelling. JC; Cavalieri. EL; Rogan. EG. (1995). Relating aromatic hydrocarbon-induced DNA
40           adducts and  c-H-ras mutations in mouse skin papillomas: The role of apurinic sites. PNAS 92:
41           10422-10426.
42    Charles. GD;  Bartels. MJ; Zacharewski. TR;  Gollapudi. BB; Freshour.  ML; Carney. EW. (2000). Activity  of
43           benzo[a]pyrene and its hydroxylated metabolites in an estrogen receptor-alpha reporter gene
44           assay. Toxicol Sci 55: 320-326. http://dx.doi.Org/10.1093/toxsci/55.2.320
45    Chau, N; Bertrand, JP; Mur, JM; Figueredo, A; Patris, A; Moulin, JJ; Pham, QT. (1993). Mortality in retired
46           coke oven plant workers. Br J Ind Med  50: 127-135.
47    Chen, C; Tang, Y; Jiang, X; Qi, Y; Cheng, S; Qiu, C; Peng, B; Tu, B. (2012).  Early postnatal benzo(a)pyrene
48           exposure in Sprague-Dawley rats causes persistent neurobehavioral impairments that emerge
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-5             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1           postnatally and continue into adolescence and adulthood. Toxicol Sci 125: 248-261.
 2           http://dx.doi.org/10.1093/toxsci/kfr265
 3    Chen, G; Gingerich, J; Soper, L; Douglas, GR; White, PA. (2010). Induction of lacZ mutations in
 4           MutaMouse primary hepatocytes. Environ Mol Mutagen 51: 330-337.
 5           http://dx.doi.org/10.1002/em.20540
 6    Chen, X; An, H; Ao, L; Sun, L; Liu, W; Zhou, Z; Wang, Y; Cao, J. (2011). The combined toxicity of dibutyl
 7           phthalate and benzo(a)pyrene on the reproductive system of male Sprague Dawley rats in vivo. J
 8           Hazard Mater 186: 835-841. http://dx.doi.Org/10.1016/i.ihazmat.2010.ll.078
 9    Cheng, L; Spitz, MR; Hong, WK; Wei, Q. (2000). Reduced expression levels of nucleotide excision repair
10           genes in lung cancer: A case-control  analysis. Carcinogenesis 21: 1527-1530.
11    Chengzhi, C; Yan, T; Shuqun, C; Xuejun, J; Youbin, Q; Yinyin, X; Qian, T; Baijie, T. (2011). New candidate
12           proteins for benzo(a)pyrene-induced spatial learning and memory deficits. J Toxicol Sci 36: 163-
13           171.
14    Chouroulinkov, I; Gentil, A; Guerin, M. (1967). [Study of the carcinogenic activity of 9,10-dimethyl-
15           benzanthracene and of 3,4-benzopyrene given orally]. Bull Cancer 54: 67-78.
16    Chung. JY; Kim. YJ; Kim. JY; Lee. SG; Park. JE;  Kim. WR; Yoon. YD; Yoo. KS; Yoo. YH;  Kim. JM. (2011).
17           Benzo[a]pyrene reduces testosterone production in rat Leydig cells via a direct disturbance of
18           testicular steroidogenic machinery. Environ Health Perspect 119: 1569-1574.
19           http://dx.doi.org/10.1289/ehp.1003391
20    Clavson. DB;  Iverson. F; Nera. EA; Lok. E. (1990). The significance of induced forestomach tumors. Annu
21           Rev Pharmacol Toxicol 30: 441-463.
22    Clement Associates.  Inc. . (1990). Comparative potency approach for estimation of the total cancer risk
23           associated with exposures to mixtures of polycyclic aromatic hydrocarbons in the environment:
24           Appendices.
25    Cooper, AR; Moley, KH. (2008). Maternal tobacco use and its preimplantation effects on fertility: More
26           reasons to stop smoking [Review]. Seminars in Reproductive Medicine 26: 204-212.
27           http://dx.doi.org/10.1055/s-2008-1042959
28    Costantino, JP; Redmond,  CK; Bearden, A. (1995). Occupationally related cancer risk among coke oven
29           workers:  30 years of follow-up. J Occup Environ Med 37: 597-604.
30    Cresteil, T. (1998). Onset of xenobiotic metabolism in children: Toxicological implications. Food Addit
31           Contam 15: 45-51. http://dx.doi.org/10.1080/02652039809374614
32    Gulp, SJ; Gaylor, DW; Sheldon, WG; Goldstein, LS; Beland, FA.  (1996). DNA Adduct Measurements in
33           Relation to Small Intestine and Forestomach Tumor Incidence during the Chronic Feeding of
34           Coal TAR or Benzo[ A ]Pyrene to Mice. Polycycl Aromat Compd 11: 161-168.
35           http://dx.doi.org/10.1080/10406639608544662
36    Gulp, SJ; Gaylor, DW; Sheldon, WG; Goldstein, LS; Beland, FA.  (1998). A comparison of the tumors
37           induced by coal tar and benzo[a]pyrene in a 2-year bioassay. Carcinogenesis 19: 117-124.
38    Gulp. SJ; Warbritton. AR; Smith. BA; Li. EE; Beland. FA. (2000). DNA adduct measurements, cell
39           proliferation and tumor mutation induction in relation to tumor formation in B6C3F1 mice fed
40           coal tar or benzo[a]pyrene. Carcinogenesis 21: 1433-1440.
41           http://dx.doi.0rg/10.1093/carcin/21.7.1433
42    Das. M; Asokan. P; Don. PS; Krueger. GG; Bickers. PR; Mukhtar. H. (1986). Carcinogen metabolism in
43           human skin grafted onto athymic nude mice: a model system for the study of human skin
44           Carcinogenesis. Biochem Biophys Res Commun 138: 33-39.
45    Davies, RF. (1969). Tumorigenicity of very fresh tobacco smoke condensate to mouse skin. Br J Cancer
46           23:858-860.
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-6             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    De Jong, WH; Kroese, ED; Vos, JG; Van Loveren, H. (1999). Detection of immunotoxicity of
 2           benzo[a]pyrene in a subacute toxicity study after oral exposure in rats. Toxicol Sci 50: 214-220.
 3           http://dx.doi.0rg/10.1093/toxsci/50.2.21
 4    De Waal. EJ; Schuurman. HJ; Van Loveren. H; Vos. JG. (1997). Differential effects of 2,3,7,8-
 5           tetrachlorodibenzo-p-dioxin, bis(tri-n-butyltin) oxide and cyclosporine on thymus
 6           histophysiology [Review]. Crit Rev Toxicol 27: 381-430.
 7           http://dx.doi.org/10.3109/10408449709089900
 8    Dean, JH; Luster, Ml; Boorman, GA; Lauer, LD; Leubke, RW; Lawson, L (1983). Selective
 9           immunosuppression resulting from exposure to the carcinogenic congener of benzopyrene in
10           B6C3F1 mice. Clin  Exp Immunol 52: 199-206.
11    DeMarini, DM; Landi, S; Tian,  D; Hanley, NM; Li, X; Hu, F; Roop, BC; Mass, MJ; Keohavong, P; Gao, W;
12           Olivier, M; Hainaut, P; Mumford, JL (2001). Lung tumor KRAS and TP53 mutations in
13           nonsmokers reflect exposure to PAH-rich coal combustion emissions. Cancer Res 61: 6679-6681.
14    Denissenko, MF; Pao, A; Tang, M; Pfeifer, GP. (1996). Preferential formation of benzo[a]pyrene adducts
15           at lung cancer mutational hotspots in P53. Science 274: 430-432.
16           http://dx.doi.org/10.1126/science.274.5286.430
17    Dietert, RR. (2008). Developmental immunotoxicology (DIT): windows of vulnerability, immune
18           dysfunction and safety assessment. J Immunotoxicol 5: 401-412.
19           http://dx.doi.org/10.1080/15476910802483324
20    Dietert. RR; Etzel. RA; Chen. D; Halonen. M; Holladay. SD; Jarabek. AM; Landreth. K; Peden. DB;
21           Pinkerton. K; Smialowicz. RJ; Zoetis. T. (2000). Workshop to identify critical windows of exposure
22           for  children's health: immune and respiratory systems work group summary [Review]. Environ
23           Health Perspect 108 Suppl 3: 483-490.
24    Dietert, RR; Piepenbrink, MS.  (2006). Perinatal immunotoxicity: Why adult exposure assessment fails to
25           predict risk [Review].  Environ Health Perspect 114: 477-483.
26           http://dx.doi.org/10.1289/ehp.8566
27    Dobbing, J;  Sands, J. (1973). Quantitative growth and development of human brain. Arch Dis Child 48:
28           757-767.
29    Dobbing, J;  Sands, J. (1979). Comparative aspects of the brain growth spurt. Early Hum Dev 3: 79-83.
30    Dong, MH; Redmond, CK; Mazumdar, S; Costantino, JP. (1988). A multistage approach to the cohort
31           analysis of lifetime lung cancer risk among steelworkers exposed to coke oven emissions. Am J
32           Epidemiol 128: 860-873.
33    Dontenwill, W; Chevalier, HJ;  Harke, HP; Lafrenz, U; Reckzeh, G; Schneider, B. (1973). Investigations on
34           the effects of chronic cigarette-smoke inhalation in Syrian golden  hamsters. J Natl Cancer Inst
35           51:  1781-1832. http://dx.doi.Org/10.1093/inci/51.6.1781
36    Driscoll, KE; Costa, PL; Hatch,  G; Henderson, R; Oberdorster, G; Salem, H; Schlesinger, RB. (2000).
37           Intratracheal instillation as an exposure technique for the evaluation of respiratory tract toxicity:
38           Uses and limitations. Toxicol Sci 55: 24-35.
39    EI-Bayoumy. K. (1985). Effects of organoselenium compounds on induction of mouse forestomach
40           tumors by benzo(a)pyrene. Cancer Res 45:  3631-3635.
41    EI-Bayoumy. K; Hecht. SS; Hoffmann. D. (1982). Comparative tumor initiating activity on mouse skin of 6-
42           nitrobenzo[a]pyrene,  6-nitrochrysene, 3-nitroperylene, 1-nitropyrene and their parent
43           hydrocarbons. Cancer Lett 16: 333-337. http://dx.doi.org/10.1016/0304-3835(82)90015-5
44    EI-Nemr,  A; Al-Shawaf, T; Sabatini, L; Wilson, C; Lower, AM; Grudzinskas, JG. (1998). Effect of smoking on
45           ovarian reserve and ovarian stimulation in in-vitro fertilization and embryo transfer. Hum
46           Reprod 13: 2192-2198.
47    Emmett,  EA; Bingham, EM; Barkley, W. (1981). A carcinogenic bioassay of certain roofing materials. Am J
48           Ind  Med 2: 59-64.  http://dx.doi.org/10.1002/aiim.4700020110
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-7             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Esser, C. (2009). The immune phenotype of AhR null mouse mutants: not a simple mirror of xenobiotic
 2           receptor over-activation [Review]. Biochem Pharmacol 77: 597-607.
 3           http://dx.doi.0rg/10.1016/i.bcp.2008.10.002
 4    Evanoff, BA; Gustavsson, P; Hogstedt, C. (1993). Mortality and incidence of cancer in a cohort of Swedish
 5           chimney sweeps: an extended follow up study. Occup Environ Med 50: 450-459.
 6    Everall.JD; Dowd,  PM. (1978). Influence of environmental factors excluding ultra violet radiation on the
 7           incidence of skin cancer. Bull Cancer 65: 241-247.
 8    Fedorenko, Z; Yansheva, N. (1967). [Experimental  reproduction of tumors of the antral part of the
 9           stomach in mice by administration of various dose of 3,4-benzpyrene]. Gig Sanit 32: 168-173.
10    Fernandez-Salguero, P; Pineau, T;  Hilbert, DM; McPhail, T; Lee, SS; Kimura, S; Nebert, DW; Rudikoff, S;
11           Ward, JM; Gonzalez, FJ. (1995). Immune system impairment and hepatic fibrosis in mice lacking
12           the dioxin-binding Ah receptor. Science 268: 722-726.
13    Feron, VJ; de Jong, D; Emmelot, P. (1973). Letter: Dose-response correlation for the induction of
14           respiratory-tract tumours  in Syrian golden hamsters by intratracheal instillations of
15           benzo(a)pyrene. Eur J Cancer 9: 387-390.
16    Feron, VJ; Kruysse, A. (1978). Effects of exposure to furfural vapour in hamsters simultaneously treated
17           with benzo[alpha] pyrene  or diethylnitrosamine. Toxicology 11: 127-144.
18    Perron, GA. (1994). Aerosol properties and lung deposition [Editorial]. Eur Respir J 7: 1392-1394.
19    Fertuck. KC; Matthews. JB; Zacharewski. TR. (2001). Hydroxylated benzo[a]pyrene metabolites are
20           responsible for in vitro estrogen receptor-mediated gene expression induced by
21           benzo[a]pyrene, but do not elicit uterotrophic effects  in vivo. Toxicol Sci 59: 231-240.
22           http://dx.doi.0rg/10.1093/toxsci/59.2.231
23    Fine. JS; Silverstone. AE; Gasiewicz. TA. (1990). Impairment of  prothymocyte activity by 2,3,7,8-
24           tetrachlorodibenzo-p-dioxin. J Immunol 144: 1169-1176.
25    Flowers, L; Bleczinski, WF; Burczynski, ME; Harvey, RG; Penning, TM. (1996). Disposition and biological
26           activity of benzo[a]pyrene-7,8-dione. A genotoxic metabolite generated by dihydrodiol
27           dehydrogenase. Biochemistry 35: 13664-13672. http://dx.doi.org/10.1021/bi961077w
28    Flowers, L; Ohnishi, ST; Penning, TM. (1997). DNA strand scission by polycyclic aromatic hydrocarbon o-
29           quinones:  role of reactive  oxygen species, Cu(ll)/Cu(l)  redox cycling, and o-semiquinone anion
30           radicals. Biochemistry 36:  8640-8648. http://dx.doi.org/10.1021/bi970367p
31    Francis, EZ; Kimmel, CA; Rees, DC.  (1990). Workshop on the qualitative and quantitative comparability of
32           human and animal developmental neurotoxicity: Summary and implications. Neurotoxicol
33           Teratol 12: 285-292. http://dx.doi.org/10.1016/0892-0362(90)90101-H
34    Friesen, MC; Benke, G; Del Monaco, A; Dennekamp, M; Fritschi, L; de Klerk, N; Moving, JL; Macfarlane, E;
35           Sim, MR. (2009). Relationship between cardiopulmonary mortality and cancer risk and
36           quantitative exposure to polycyclic aromatic hydrocarbons, fluorides, and dust in two prebake
37           aluminum smelters. Cancer Causes Control 20: 905-916. http://dx.doi.org/10.1007/sl0552-009-
38           9329-8
39    Friesen. MC; Demers. PA; Spinelli.  JJ; Eisen. EA; Lorenzi. MF; Le. ND. (2010). Chronic and acute effects of
40           coal tar pitch exposure and cardiopulmonary mortality among aluminum smelter workers. Am J
41           Epidemiol 172: 790-799. http://dx.doi.org/10.1093/aie/kwq208
42    Friesen. MC; Macnab. YC; Marion. SA; Demers. PA; Davies. HW; Teschke. K. (2006). Mixed models and
43           empirical bayes estimation for retrospective exposure assessment of dust exposures in Canadian
44           sawmills. Ann Occup Hyg 50: 281-288. http://dx.doi.org/10.1093/annhyg/mei076
45    Gale, CR; O'Callaghan, FJ; Bredow, M; Martyn, CM; Team, ALSoPaCS. (2006). The influence of head
46           growth in fetal life, infancy, and childhood on intelligence at the ages of 4 and 8 years. Pediatrics
47           118: 1486-1492. http://dx.doi.org/10.1542/peds.2005-2629

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

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Galvan, N; Page, TJ; Czuprynski, G; Jefcoate, CR. (2006). Benzo(a)pyrene and 7,12-
 2           dimethylbenz(a)anthrecene differentially affect bone marrow cells of the lymphoid and myeloid
 3           lineages. Toxicol Appl Pharmacol 213: 105-116. http://dx.doi.Org/10.1016/i.taap.2005.09.018
 4    Gao, M; Li, Y; Sun, Y; Shah, W; Yang, S; Wang, Y; Long, J. (2011). Benzo[a]pyrene exposure increases toxic
 5           biomarkers and morphological disorders in mouse cervix. Basic & Clinical  Pharmacology &
 6           Toxicology Online Pharmacology Online 109: 398-406. http://dx.doi.org/lO.llll/j. 1742-
 7           7843.2011.00755.x
 8    Garcon, G; Gosset, P; Garry, S;  Marez, T; Hannothiaux, MH; Shirali, P. (2001a). Pulmonary induction of
 9           proinflammatory mediators following the rat exposure to benzo(a)pyrene-coated onto Fe2O3
10           particles. Toxicol  Lett 121: 107-117.
11    Garcon, G; Zerimech, F; Hannothiaux, M; Gosset, P; Martin, A; Marez, T; Shirali, P. (2001b). Antioxidant
12           defense disruption by polycyclic aromatic hydrocarbons-coated onto Fe(2)O(3) particles in
13           human lung cells (A549). Toxicology 166: 129-137.
14    Gaspari, L; Chang, SS; Santella, RM; Garte, S; Pedotti, P; Taioli, E. (2003). Polycyclic aromatic
15           hydrocarbon-DNA adducts in human sperm as a marker of DNA damage and infertility. Mutat
16           Res 535: 155-160.
17    Geacintov, NE; Cosman, M; Hingerty, BE; Amin, S; Broyde, S; Patel, DJ. (1997). NMR solution structures
18           of stereoisometric covalent polycyclic aromatic carcinogen-DNA adduct: principles, patterns,
19           and diversity [Review]. Chem  Res Toxicol 10: 111-146. http://dx.doi.org/10.1021/tx9601418
20    Gehlhaus. MW; Gift. JS; Hogan. KA; Kopylev. L; Schlosser. PM; Kadry. AR. (2011). Approaches to cancer
21           assessment in EPA's Integrated Risk  Information System [Review]. Toxicol Appl Pharmacol 254:
22           170-180. http://dx.doi.0rg/10.1016/i.taap.2010.10.019
23    Centner. NJ; Weber. LP. (2011). Intranasal benzo[a]pyrene alters circadian  blood pressure patterns and
24           causes lung inflammation  in rats. Arch Toxicol 85: 337-346. http://dx.doi.org/10.1007/s00204-
25           010-0589-6
26    Gibbs, GW. (1985). Mortality of aluminum reduction plant workers, 1950 through 1977. J Occup Environ
27           Med 27: 761-770.
28    Gibbs, GW; Armstrong, B; Sevigny, M. (2007). Mortality and cancer experience of Quebec aluminum
29           reduction plant workers. Part 2: mortality of three cohorts hired on or before January  1, 1951. J
30           Occup Environ Med 49: 1105-1123. http://dx.doi.org/10.1097/JOM.Ob013e318157d34a
31    Gibbs, GW; Horowitz, I. (1979). Lung cancer mortality in aluminum reduction plant workers. J Occup
32           Med 21: 347-353.
33    Gibbs, GW; Sevigny, M. (2007a). Mortality and cancer experience of Quebec aluminum reduction  plant
34           workers, part 4: cancer incidence. J Occup Environ Med 49: 1351-1366.
35           http://dx.doi.org/10.1097/JOM.Ob013e318156ecbc
36    Gibbs, GW; Sevigny, M. (2007b). Mortality and cancer experience of Quebec aluminum reduction  plant
37           workers. Part 3: monitoring the mortality of workers first employed after January 1, 1950. J
38           Occup Environ Med 49: 1269-1287. http://dx.doi.org/10.1097/JOM.Ob013e3181593da8
39    Ginsberg. G; Hattis. D; Sonawane.  B. (2004).  Incorporating pharmacokinetic differences between
40           children and adults in assessing children's risks to environmental toxicants [Review]. Toxicol
41           Appl Pharmacol 198: 164-183. http://dx.doi.Org/10.1016/i.taap.2003.10.010
42    GLC (Great Lakes Commission). (2007). Assessment of benzo(a)pyrene air emissions in the Great Lakes
43           region: A report of the Great Lakes Regional Toxic Air Emissions Inventory Steering Committee.
44           Ann Arbor,  Ml.
45    Godschalk, R;  Curfs, D; Bartsch, H; Van Schooten, FJ; Nair, J. (2003). Benzo[a]pyrene enhances lipid
46           peroxidation induced DNA damage in aorta of apolipoprotein E knockout mice. Free Radic Res
47           37: 1299-1305.

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

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Godschalk, R; Nair, J; van Schooten, FJ; Risch, A; Drings, P; Kayser, K; Dienemann, H; Bartsch, H. (2002).
 2           Comparison of multiple DNA adduct types in tumor adjacent human lung tissue: effect of
 3           cigarette smoking. Carcinogenesis 23: 2081-2086.
 4    Godschalk, RW; Maas, LM; Kleinjans, JC; Van Schooten, FJ. (1998a). Influences of DNA isolation and RNA
 5           contamination on carcinogen-DNA adduct analysis by 32P-postlabeling. Environ Mol Mutagen
 6           32:344-350. http://dx.doi.org/10.1002/(SICI)1098-2280(1998)32:4<344::AID-EM8>3.0.CO;2-P
 7    Godschalk, RW; Ostertag, JU; Moonen, EJ; Neumann, HA; Kleinjans, JC; van Schooten, FJ. (1998b).
 8           Aromatic DNA adducts in human white blood cells and skin after dermal application of coal tar.
 9           Cancer Epidemiol Biomarkers Prev 7: 767-773.
10    Godschalk, RW; Ostertag, JU; Zandsteeg, AM; Van Agen, B; Neuman, HA; Van Straaten, H; Van Schooten,
11           Fl (2001). Impact of GSTM1 on aromatic-DNA adducts and p53 accumulation in human skin and
12           lymphocytes. Pharmacogenetics 11: 537-543.
13    Graem, N. (1986). Epidermal changes following application of 7,12-dimethylbenz(a)anthracene and 12-
14           O-tetradecanoylphorbol-13-acetate to human skin transplanted to nude mice studied with
15           histological species markers. Cancer Res 46: 278-284.
16    Grant, G; Roe, FJ. (1963). The effect of phenanthrene on tumour induction by 3,4-benzopyrene
17           administered to newly born mice. Br J Cancer 17: 261-265.
18    Grice, HC; Clayson, DB; Flamm, WG; Ito, N; Kroes, R; Newberne, PM; Scheuplein, R. (1986). Panel
19           discussion: Possible mechanisms of BHA carcinogenicity from a consideration of its chemical and
20           biological properties. Food Chem Toxicol 24: 1235-1242.
21    Grimmer. G; Brune. H; Deutsch-Wenzel. R; Dettbarn. G; Misfeld. J. (1984). Contribution of polycyclic
22           aromatic hydrocarbons to the carcinogenic impact of gasoline engine exhaust condensate
23           evaluated by implantation into the lungs of rats. J Natl Cancer Inst 72: 733-739.
24    Grimmer, G; Brune, H; Deutsch-Wenzel, R; Naujack, KW; Misfeld, J; Timm, J. (1983). On the  contribution
25           of polycyclic aromatic hydrocarbons to the carcinogenic impact of automobile exhaust
26           condensate evaluated by local application onto mouse skin. Cancer Lett 21: 105-113.
27    Grova, N; Schroeder, H; Farinelle, S; Prodhomme, E; Valley, A; Muller, CP. (2008). Sub-acute
28           administration of benzo[a]pyrene (B[a]P) reduces anxiety-related behaviour in adult mice and
29           modulates regional expression of N-methyl-D-aspartate (NMDA) receptors genes in relevant
30           brain regions. Chemosphere 73: S295-S302.
31           http://dx.doi.0rg/10.1016/i.chemosphere.2007.12.037
32    Grova, N; Valley, A; Turner, JD; Morel, A; Muller, CP;  Schroeder, H. (2007). Modulation of behavior and
33           NMDA-R1 gene mRNA expression in adult female mice after sub-acute administration of
34           benzo(a)pyrene. Neurotoxicology 28: 630-636. http://dx.doi.Org/10.1016/i.neuro.2007.01.010
35    Gustavsson, P; Jakobsson, R; Johansson, H; Lewin, F;  Norell, S; Rutkvist, LE. (1998). Occupational
36           exposures and squamous cell carcinoma of the oral cavity, pharynx, larynx, and oesophagus: A
37           case-control study in Sweden. Occup Environ Med 55: 393-400.
38    Gustavsson. P; Plato. N; Lidstrom. EB; Hogstedt. C. (1990). Lung cancer and exposure to diesel exhaust
39           among bus garage workers. Scand J Work Environ Health 16: 348-354.
40    Guvatt. GH; Oxman. AD; Kunz. R; Vist. GE; Falck-Ytter. Y; Schunemann. HJ. (2008a). GRADE: What is
41           "quality of evidence" and why is it important to clinicians? [Review]. BMJ 336: 995-998.
42           http://dx.doi.org/10.1136/bmi.39490.551019.BE
43    Guvatt, GH; Oxman, AD; Vist, GE; Kunz, R; Falck-Ytter, Y; Alonso-Coello, P; Schunemann, HJ. (2008b).
44           GRADE: An emerging consensus on  rating quality of evidence and strength of recommendations.
45           BMJ 336: 924-926. http://dx.doi.org/10.1136/bmi.39489.470347.AD
46    Habs, M; Jahn, SA; Schmahl, D. (1984). Carcinogenic activity of condensate from coloquint seeds
47           (Citrullus colocynthis) after chronic  epicutaneous administration to mice. J Cancer Res Clin Oncol
48           108: 154-156. http://dx.doi.org/10.1007/BF00390988
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-10             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Habs, M; Schmahl, D; Misfeld, J. (1980). Local carcinogenicity of some environmentally relevant
 2           polycyclic aromatic hydrocarbons after lifelong topical application to mouse skin. Arch
 3           Geschwulstforsch 50: 266-274.
 4    Hackman, P; Hou, SM; Nyberg, F; Pershagen, G; Lambert, B. (2000). Mutational spectra at the
 5           hypoxanthine-guanine phosphoribosyltransferase (HPRT) locus in T-lymphocytes of nonsmoking
 6           and smoking lung cancer patients. Mutat Res 468: 45-61.
 7    Hainaut, P; Pfeifer, GP. (2001). Patterns of p53 G->T transversions in lung cancers reflect the primary
 8           mutagenic signature of DNA-damage by tobacco smoke. Carcinogenesis 22: 367-374.
 9    Hakim, FT; Memon, SA; Cepeda, R; Jones, EC; Chow, CK; Kasten-Sportes, C; Odom, J; Vance, BA;
10           Christensen, BL; Mackall, CL; Gress, RE. (2005). Age-dependent incidence, time course, and
11           consequences of thymic renewal in adults. J Clin Invest 115: 930-939.
12           http://dx.doi.org/10.1172/JCI22492
13    Hales, DB. (2002). Testicular macrophage modulation of Leydig cell steroidogenesis [Review]. J Reprod
14           Immunol57:3-18.
15    Hammond, EC; Selikoff, IJ;  Lawther, PL; Seidman, H. (1976). Inhalation of benzpyrene and cancer in man.
16           Ann N YAcad Sci 271: 116-124. http://dx.doi.Org/10.llll/i.1749-6632.1976.tb23100.x
17    Hanahan, D; Weinberg, RA. (2011). Hallmarks of cancer: The next generation. Cell 144: 646-674.
18           http://dx.doi.0rg/10.1016/i.cell.2011.02.013
19    Hansen. ES. (1989). Cancer incidence in an occupational cohort exposed to bitumen fumes. Scand J Work
20           Environ Health 15: 101-105.
21    Hansen. ES. (1991). Mortality of mastic asphalt workers. Scand J Work Environ Health 17: 20-24.
22           http://dx.doi.org/10.5271/siweh.1739
23    Hardin. JA; Hinoshita. F; Sherr. PH. (1992). Mechanisms by which benzo[a]pyrene, an environmental
24           carcinogen, suppresses B cell lymphopoiesis. Toxicol Appl Pharmacol 117: 155-164.
25    Harvey, RG. (1996). Mechanisms of carcinogenesis of polycyclic aromatic hydrocarbons. Polycycl Aromat
26           Compd 9: 1-23. http://dx.doi.org/10.1080/10406639608031196
27    Hattis, D; Goble, R; Russ, A; Chu, M; Ericson, J. (2004). Age-related differences in susceptibility to
28           carcinogenesis: A quantitative analysis of empirical animal bioassay data. Environ Health
29           Perspect 112: 1152-1158. http://dx.doi.org/10.1289/ehp.6871
30    Henry,  MC; Port, CD; Bates, RR; Kaufman, DG. (1973). Respiratory tract tumors in hamsters induced by
31           benzo(a)pyrene. Cancer Res 33: 1585-1592.
32    HEW (U.S. Department of Health, Education and Welfare). (1964). Smoking and health: Report of the
33           advisory committee to the surgeon general of the public health service. Washington, DC: U.S.
34           Department of Health, Education, and Welfare.
35           http://profiles.nlm.nih.gov/ps/retrieve/ResourceMetadata/NNBBMQ
36    Higginbotham, S; Ramakrishna, NVS; Johansson, SL; Rogan, EG; Cavalieri, EL. (1993). Tumor-initiating
37           activity and carcinogenicity of dibenzo[a,l]pyrene versus 7,12-dimethylbenz[a]anthracene and
38           benzo[a]pyrene at low doses in mouse skin. Carcinogenesis 14: 875-878.
39           http://dx.doi.0rg/10.1093/carcin/14.5.875
40    Hill. AB. (1965). The environment and disease: Association or causation?  Proc R Soc Med 58: 295-300.
41    Hoffmann. D; Rathkamp. G; Nesnow. S; Wynder. EL. (1972). Fluoranthenes: Quantitative determination
42           in cigarette smoke, formation by pyrolysis, and tumor-initiating activity. J Natl Cancer Inst 49:
43           1165-1175.
44    Hogstedt, C; Jansson, C; Hugosson, M; Tinnerberg, H; Gustavsson, P. (2013). Cancer incidence in a cohort
45           of Swedish chimney sweeps, 1958-2006. Am J Public Health 103: 1708-1714.
46           http://dx.doi.org/10.2105/AJPH.2012.300860
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-ll             DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene


 1    Holladay, SD; Smialowicz, RJ. (2000). Development of the murine and human immune system:
 2           Differential effects of immunotoxicants depend on time of exposure [Review]. Environ Health
 3           Perspect 108: 463-473.
 4    Holladav, SD; Smith, BJ. (1994). Fetal hematopoietic alterations after maternal exposure to
 5           benzo[a]pyrene: A cytometric evaluation. J Toxicol Environ Health 42: 259-273.
 6           http://dx.doi.org/10.1080/15287399409531878
 7    Holladav, SD; Smith, BJ. (1995). Benzo[a]pyrene-induced alterations in total immune cell number and
 8           cell-surface antigen expression in the thymus, spleen and bone marrow of B6C3F1 mice. Vet
 9           Hum Toxicol 37: 99-104.
10    Homburger, F; Hsueh, SS; Kerr, CS; Russfield, AB. (1972). Inherited susceptibility of inbred strains of
11           Syrian hamsters to induction of subcutaneous sarcomas and mammary and gastrointestinal
12           carcinomas by subcutaneous and gastric administration of polynuclear hydrocarbons. Cancer
13           Res 32: 360-366.
14    Hood, DB; Nayyar, T;  Ramesh, A; Greenwood, M; Inyang, F. (2000). Modulation in the developmental
15           expression profile of Spl subsequent to transplacental exposure of fetal rats to desorbed
16           benzo[a]pyrene following maternal inhalation. Inhal Toxicol 12: 511-535.
17           http://dx.doi.org/10.1080/089583700402897
18    HSDB (Hazardous Substances Data Bank). (2012). Benzo(a)pyrene [Database]. Bethesda, MD: National
19           Library of Medicine. Retrieved from http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7HSDB
20    Hsu.  PC: Chen. IY: Pan. CH: Wu. KY: Pan. MH: Chen. JR: Chen. CJ: Chang-Chien. GP: Hsu. CH: Liu. CS: Wu.
21           MT. (2006). Sperm DNA damage correlates with polycyclic aromatic hydrocarbons biomarker in
22           coke-oven workers. Int Arch Occup Environ Health 79: 349-356.
23           http://dx.doi.org/10.1007/s00420-005-0066-3
24    IARC (International Agency for Research on Cancer). (1973).  Certain polycyclic aromatic hydrocarbons
25           and heterocyclic compounds. (HEEP/74/11596). Lyon,  France.
26    IARC (International Agency for Research on Cancer). (1983).  Polynuclear aromatic compounds, part 1:
27           Chemical, environmental and experimental data. Lyon, France: World Health Organization.
28           http://monographs.iarc.fr/ENG/Monographs/vol32/volume32.pdf
29    IARC (International Agency for Research on Cancer). (2002).  Section 3. Carcinogenicity studies in
30           experimental animals [IARC Monograph] (pp. 973-1003). Lyon, France.
31           http://monographs.iarc.fr/ENG/Monographs/vol83/mono83-6C.pdf
32    IARC (International Agency for Research on Cancer). (2003).  Predictive value of rodent forestomach and
33           gastric neuroendocrine tumours in evaluating carcinogenic risks to humans: Views and expert
34           opinions of an IARC working group, Lyon, 29 November -1 December 1999. (IARC Technical
35           Publication No. 39). Lyon, France.
36           http://monographs.iarc.fr/ENG/Publications/techrep39/IARCrep39.pdf
37    IARC (International Agency for Research on Cancer). (2004).  Tobacco smoke and involuntary smoking
38           [IARC Monograph] (pp. 1-1438). (RISKLINE/2004080004).  Lyon, France.
39           http://monographs.iarc.fr/ENG/Monographs/vol83/mono83.pdf
40    IARC (International Agency for Research on Cancer). (2006).  Preamble to the IARC monographs. Lyon,
41           France. http://monographs.iarc.fr/ENG/Preamble/
42    IARC (International Agency for Research on Cancer). (2010).  Some non-heterocyclic polycyclic aromatic
43           hydrocarbons and some related exposures [IARC Monograph] (pp. 1-853). Lyon, France.
44           http://monographs.iarc.fr/ENG/Monographs/vol92/mono92.pdf
45    IARC (International Agency for Research on Cancer). (2012).  Soot, as found in occupational exposure of
46           chimney sweeps [IARC Monograph]. In Chemical agents and related occupations A review of
47           human carcinogens (pp. 209-214). Lyon, France.
48           http://monographs.iarc.fr/ENG/Monographs/vollOOF/monolOOF-21.pdf
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-12             DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1    Ickstadt. K; Schafer. M; Fritsch. A; Schwender. H; Abel. J; Bolt. HM; Bruning. T; Ko. YD; Vetter. H; Harth.
 2           \A (2008). Statistical methods for detecting genetic interactions: a head and neck squamous-cell
 3           cancer study. J Toxicol Environ Health A 71: 803-815.
 4           http://dx.doi.org/10.1080/15287390801985745
 5    Inyang, F; Ramesh, A; Kopsombut, P; Niaz, MS; Hood, DB; Nvanda, AM; Archibong, AE. (2003). Disruption
 6           of testicular steroidogenesis and epididymal function by inhaled benzo(a)pyrene. Reprod Toxicol
 7           17:527-537.
 8    IOM (Institute of Medicine). (2008). Improving the presumptive disability decision-making process for
 9           veterans. In JM Samet; CC Bodurow (Eds.). Washington,  DC: National Academies Press.
10           http://www.nap.edu/openbook.php7record  id=11908
11    I PCS (International Programme on Chemical Safety). (1998). Selected non-heterocyclic polycyclic
12           aromatic hydrocarbons [WHO EHC]. In Environmental Health Criteria (pp. CHF 174).
13           (CIS/00/00657). Geneva, Switzerland: World Health Organization.
14           http://www.inchem.org/documents/ehc/ehc/ehc202.htm
15    Jayasekara, S; Sharma,  RP; Drown, DB. (1992). Effects of benzo[a]pyrene on steady-state levels of
16           biogenic amines and metabolizing enzymes in mouse brain regions. Ecotoxicol Environ Saf 24: 1-
17           12. http://dx.doi.org/10.1016/0147-6513(92)90029-3
18    Jeffrey, AM; Jennette, KW; Blobstein, SH; Weinstein, IB; Beland,  FA; Harvey, RG; Kasal, H; Miura, I;
19           Nakanishi. K. (1976). Letter: Benzo[a]pyrene-nucleic acid derivative found in vivo: structure of a
20           benzo[a]pyrenetetrahydrodiol epoxide-guanosine adduct. J Am Chem Soc 98: 5714-5715.
21    Jeng. HA; Yordt. D; Davis. S; Swanson. JR. (2013). Assessment of alteration of reproductive system in vivo
22           induced by subchronic exposure to benzo(a)pyrene via oral administration. Environ Toxicol.
23           http://dx.doi.org/10.1002/tox.21889
24    Jensen, TK; Henriksen, TB; Hjollund, NH; Scheike, T; Kolstad, H; Giwercman, A; Ernst, E; Bonde, JP;
25           Skakkebaek, NE; Olsen, J. (1998). Adult and prenatal  exposures to tobacco smoke as risk
26           indicators of fertility among 430 Danish couples. Am J Epidemiol 148: 992-997.
27    Jerina, DM; Chadha, A; Cheh, AM; Schurdak, ME; Wood, AW; Sayer, JM. (1991). Covalent bonding of
28           bay-region diol epoxides to nucleic acids [Review]. Adv Exp Med Biol 283: 533-553.
29    Jiang, H; Gelhaus, SL; Mangal, D;  Harvey, RG; Blair, IA; Penning, TM. (2007). Metabolism of
30           benzo[a]pyrene in human bronchoalveolar H358 cells using liquid chromatography-mass
31           spectrometry. Chem Res Toxicol 20: 1331-1341. http://dx.doi.org/10.1021/tx700107z
32    Jiang, H; Shen, YM; Quinn, AM; Penning, TM. (2005). Competing roles of cytochrome P450 1A1/1B1 and
33           aldo-keto reductase 1A1 in the metabolic activation of (+/-)-7,8-Dihydroxy-7,8-dihydro-
34           benzo[a]pyrene in human bronchoalveolar cell extracts.  Chem  Res Toxicol  18: 365-374.
35           http://dx.doi.org/10.1021/tx0497245
36    Jules, GE; Pratap, S; Ramesh, A; Hood, DB. (2012). In utero exposure to benzo(a)pyrene predisposes
37           offspring to cardiovascular dysfunction in later-life. Toxicology  295: 56-67.
38           http://dx.doi.0rg/10.1016/i.tox.2012.01.017
39    Kao. J; Patterson. FK; Hall. J. (1985). Skin penetration  and metabolism of topically applied chemicals in
40           six mammalian species, including man: an in vitro study with benzo[a]pyrene and testosterone.
41           Toxicol Appl Pharmacol 81: 502-516.
42    Kapitulnik. J; Wislocki. PG; Levin. W; Yagi. H; Jerina. DM; Conney. AH. (1978). Tumorigenicity studies with
43           diol-epoxides of benzo(a)pyrene which indicate that  "(+/-)"-trans-7"beta",8"alpha"-dihydroxy-
44           9"alpha" 10"alpha"-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene  is an ultimate carcinogen in
45           newborn mice. Cancer Res 38: 354-358.
46    Kappes, U; Schliemann-Willers, S; Bankova, L; Heinemann, C; Fischer, TW; Ziemer, M; Schubert, H;
47           Norgauer, J; Fluhr, JW; Eisner, P. (2004). The quality of human skin xenografts on SCID mice: A

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

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1           noninvasive bioengineering approach. Br J Dermatol 151: 971-976.
 2           http://dx.doi.0rg/10.llll/i.1365-2133.2004.06191.x
 3    Karakaya, A; Ates, I; Yucesoy, B. (2004). Effects of occupational polycyclic aromatic hydrocarbon
 4           exposure on T-lymphocyte functions and natural killer cell activity in asphalt and coke oven
 5           workers. Hum Exp Toxicol 23: 317-322.
 6    Karakaya, A; Yucesoy, B; Turhan, A; Erdem, O;  Burgaz, S; Karakaya, AE. (1999). Investigation of some
 7           immunological functions in a group of asphalt workers exposed to polycyclic aromatic
 8           hydrocarbons. Toxicology 135: 43-47.
 9    Karlehagen, S; Andersen, A; Ohlson, CG. (1992). Cancer incidence among creosote-exposed workers.
10           Scand J Work Environ Health 18: 26-29. http://dx.doi.org/10.5271/siweh.1612
11    Kee, K; Flores, M; Cedars, Ml; Reijo Pera, RA. (2010). Human primordial germ cell formation is
12           diminished by exposure to environmental toxicants acting through the AHR signaling pathway.
13           Toxicol Sci 117: 218-224. http://dx.doi.org/10.1093/toxsci/kfql79
14    Keohavong, P; Lan, Q; Gao, WM; DeMarini, DM;  Mass, MJ; Li, XM; Roop, BC; Weissfeld, J; Tian, D;
15           Mumford, JL. (2003). K-ras mutations in  lung carcinomas from nonsmoking women exposed to
16           unvented coal smoke in China. Lung Cancer 41: 21-27. http://dx.doi.org/10.1016/S0169-
17           5002(03)00125-9
18    Ketkar, M; Reznik, G; Schneider, P; Mohr, U. (1978). Investigations on the carcinogenic burden by air
19           pollution in man. Intratracheal instillation studies with benzo(a)pyrene in bovine serum albumin
20           in Syrian hamsters. Cancer Lett 4: 235-239. http://dx.doi.org/10.1016/S0304-3835(78)94787-0
21    Klemme. JC; Mukhtar. H; Elmets. CA. (1987). Induction of contact hypersensitivity to
22           dimethylbenz(a)anthracene and benzo(a)pyrene in C3H/HeN mice. Cancer Res 47: 6074-6078.
23    Knafla. A; Petrovic. S; Richardson. M; Campbell. J; Rowat. C. (2011). Development and application of a
24           skin cancer slope factor for exposures  to benzo[a]pyrene in soil. Regul Toxicol Pharmacol 59:
25           101-110. http://dx.doi.0rg/10.1016/i.vrtph.2010.09.011
26    Knuckles, ME; Inyang, F; Ramesh, A. (2001). Acute and subchronic oral toxicities of benzo[a]pyrene in F-
27           344 rats. Toxicol Sci 61: 382-388.
28    Kolb, B; Whishaw, IQ. (1989). Plasticity in the neocortex: mechanisms underlying recovery from early
29           brain damage. Prog Neurobiol 32: 235-276.
30    Kong, LY; Luster, Ml; Dixon, D; O'Grady, J; Rosenthal, GJ. (1994). Inhibition of lung immunity after
31           intratracheal instillation of benzo(a)pyrene. Am J Respir Crit Care Med  150: 1123-1129.
32    Koreeda, M; Moore, PD; Wislocki, PG; Levin, W; Yagi, H; Jerina, DM. (1978). Binding of benzo[a]pyrene
33           7,8-diol-9,10-epoxides to DNA,  RNA, and protein of mouse skin occurs with high
34           stereoselectivity. Science 199: 778-781.
35    Kristensen, P; Eilertsen, E; Einarsdottir,  E; Haugen, A; Skaug, V; Ovrebo, S. (1995). Fertility in mice after
36           prenatal exposure to benzo[a]pyrene and inorganic lead. Environ Health Perspect 103: 588-590.
37           http://dx.doi.org/10.1289/ehp.95103588
38    Kroese. ED; Muller. JJA; Mohn. GR; Dortant. PM; Wester. PW. (2001). Tumorigenic effects in Wistar rats
39           orally administered benzo[a]pyrene for two years (gavage studies):  Implications for human
40           cancer risks associated with oral exposure to polycyclic aromatic hydrocarbons. (658603 010).
41           Bilthoven, The Netherlands: National Institute for Public Health and the Environment (RIVM).
42           http://www.rivm.nl/bibliotheek/rapporten/658603010.pdf
43    Kummer, V; Maskova, J; Zraly, Z; Matiasovic, J; Faldyna, M. (2007). Effect of postnatal exposure to
44           benzo[a]pyrene on the uterus of immature rats. Exp Toxicol Pathol 59: 69-76.
45           http://dx.doi.0rg/10.1016/i.etp.2007.02.007
46    Kummer, V; Maskova, J; Zraly, Z; Neca, J; Simeckova, P; Vondracek, J; Machala, M. (2008). Estrogenic
47           activity of environmental polycyclic aromatic hydrocarbons in uterus of immature Wistar rats.
48           Toxicol Lett 180:  212-221. http://dx.doi.Org/10.1016/i.toxlet.2008.06.862
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-14             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Kuper, CF; Beems, RB; Bruijntjes, JP; Schuurman, HJ; Vos, JG. (1992). Normal development, growth, and
 2           aging of the thymus. In U Mohr; DL Dungworth; CC Capen (Eds.), Pathobiology of the aging rat,
 3           volume 1 (pp. 25-49). Washington, DC: ILSI press.
 4    Kuper, CF; De Heer, E; Van Loveren, H; Vos, JG. (2002). Immune system. In MA Wallig (Ed.), Handbook of
 5           toxicologic pathology (2nd ed., pp. 585-644). San Diego, CA: Academic Press.
 6    Landreth, KS. (2002). Critical windows in development of the rodent immune system [Review]. Hum Exp
 7           Toxicol 21: 493-498.
 8    Landreth, KS; Dodson, SVM. (2005). Development of the rodent immune system. In SD Holladay (Ed.),
 9           Developmental immunotoxicology (pp. 3-19). Boca Raton, FL: CRC Press.
10           http://dx.doi.org/10.1201/9781420024036.ptl
11    Lavoie, EJ; Amin, S; Hecht, SS; Furuya, K; Hoffmann, D. (1982). Tumour initiating activity of dihydrodiols
12           of benzo[b]fluoranthene, benzo[j]fluoranthene, and benzo[k]fluoranthene. Carcinogenesis 3:
13           49-52. http://dx.doi.0rg/10.1093/carcin/3.l.49
14    Lavoie, EJ; Braley, J; Rice, JE; Rivenson, A. (1987). Tumorigenic activity of non-alternant polynuclear
15           aromatic hydrocarbons in newborn mice. Cancer Lett 34: 15-20.
16    Lavoie, EJ; Cai, ZW; Meschter, CL; Weyand, EH. (1994). Tumorigenic activity of fluoranthene, 2-
17           methylfluoranthene and 3-methylfluoranthene in newborn CD-I mice. Carcinogenesis 15: 2131-
18           2135. http://dx.doi.org/10.1093/carcin/15.10.2131
19    Lavoue. J; Gerin. M; Cote. J; Lapointe. R. (2007). Mortality and cancer experience of Quebec aluminum
20           reduction  plant workers. Part I: The reduction plants and coal tar pitch volatile (CTPV) exposure
21           assessment. J Occup Environ Med 49: 997-1008.
22           http://dx.doi.org/10.1097/JOM.Ob013e3181484cf3
23    Letzel. S; Drexler. H. (1998). Occupationally related tumors in tar refinery workers. J Am Acad  Dermatol
24           39: 712-720. http://dx.doi.org/10.1016/S0190-9622(98)70043-X
25    Levin, W; Wood, AW; Wislocki, PG;  Kapitulnik, J; Yagi, H; Jerina, DM; Conney, AH. (1977). Carcinogenicity
26           of benzo-ring derivatives of benzo(a)pyrene on mouse skin. Cancer Res 37: 3356-3361.
27    Li, D; Firozi, PF; Wang, LE; Bosken, CH; Spitz, MR; Hong, WK; Wei, Q. (2001). Sensitivity to DNA damage
28           induced by benzo(a)pyrene diol epoxide and risk of lung cancer: a case-control analysis.  Cancer
29           Res 61: 1445-1450.
30    Li, Z; Chadalapaka, G; Ramesh, A; Khoshbouei, H; Maguire, M; Safe, S; Rhoades, RE; Clark, R; Jules, G;
31           Mccallister, M; Aschner, M; Hood, DB. (2012). PAH particles perturb prenatal processes and
32           phenotypes: protection from deficits in object discrimination afforded by dampening of brain
33           oxidoreductase following in utero exposure to inhaled benzo(a)pyrene. Toxicol Sci 125: 233-247.
34           http://dx.doi.org/10.1093/toxsci/kfr261
35    Liang, J; Zhu, H; Li, C; Ding, Y; Zhou, Z; Wu, Q. (2012). Neonatal exposure to benzo[a]pyrene decreases
36           the levels of serum testosterone and histone H3K14 acetylation of the StAR promoter in the
37           testes of SD rats. Toxicology 302: 285-291. http://dx.doi.Org/10.1016/i.tox.2012.08.010
38    Lindelof. B; Sigurgeirsson. B. (1993). PUVA and cancer: a case-control study. Br J Dermatol 129: 39-41.
39           http://dx.doi.0rg/10.llll/i.1365-2133.1993.tb03308.x
40    Liu. M; Grant. SG; Macina. OT; Klopman. G; Rosenkranz. HS. (1997). Structural and  mechanistic bases for
41           the induction of mitotic chromosomal loss and duplication (malsegregation1) in the yeast
42           Saccharomyces cerevisiae:  Relevance to human Carcinogenesis and developmental toxicology.
43           Mutat Res-Fundam Mol Mech Mutagen 374: 209-231. http://dx.doi.org/10.1016/S0027-
44           5107(96)00236-9
45    Liu, S; Abdelrahim, M; Khan, S; Ariazi, E; Jordan, VC; Safe, S. (2006). Aryl hydrocarbon receptor agonists
46           directly activate estrogen receptor alpha in MCF-7 breast cancer cells. Biol Chem 387: 1209-
47           1213. http://dx.doi.org/10.1515/BC.2006.149

                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-15            DRAFT—DO  NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Liu, SH; Wang, JH; Chuu, JJ; Lin-Shiau, SY. (2002). Alterations of motor nerve functions in animals
 2           exposed to motorcycle exhaust. J Toxicol Environ Health A 65: 803-812.
 3           http://dx.doi.org/10.1080/00984100290071144
 4    Liu, Z; Muehlbauer, KR; Schmeiser, HH; Hergenhahn, M; Belharazem, D; Hollstein, MC. (2005). p53
 5           mutations in benzo(a)pyrene-exposed human p53 knock-in murine fibroblasts correlate with
 6           p53 mutations in human lung tumors. Cancer Res 65: 2583-2587.
 7           http://dx.doi.org/10.1158/0008-5472.CAN-04-3675
 8    Luebke, RW; Chen, DH; Dietert, R; Yang, Y; King, M; Luster, Ml.  (2006). The comparative immunotoxicity
 9           of five selected compounds following developmental or adult exposure [Review]. J Toxicol
10           Environ Health B Crit Rev 9: 1-26. http://dx.doi.org/10.1080/15287390500194326
11    Lummus, ZL; Henningsen, G. (1995). Modulation of T-cell ontogeny by transplacental benzo(a)pyrene.
12           International Journal of Immunopharmacology 17: 339-350.
13    Luster, Ml; Portier, C; Pait, DG; White, KL, Jr; Gennings, C; Munson, AE; Rosenthal, GJ. (1992). Risk
14           assessment in immunotoxicology: I: Sensitivity and predictability of immune tests. Fundam Appl
15           Toxicol 18: 200-210. http://dx.doi.org/10.1016/0272-0590(92)90047-L
16    Lyte, M; Bick, PH. (1985). Differential immunotoxic effects of the environmental chemical
17           benzo[a]pyrene in young and aged mice. Mech Ageing  Dev 30: 333-341.
18           http://dx.doi.org/10.1016/0047-6374(85)90122-8
19    Ma. Q; Lu. AY. (2007). CYP1A induction and human risk assessment: an evolving tale of in vitro and  in
20           vivo studies [Review]. Drug Metab Dispos 35: 1009-1016.
21           http://dx.doi.org/10.1124/dmd.107.015826
22    Maciel. ES; Biasibetti. R; Costa. AP; Lunardi. P; Schunck. RV; Becker. GC; Arbo. MD; Dallegrave. E;
23           Goncalves. CA; Saldiva. PH; Garcia. SC; Leal. RB; Leal. MB. (2014). Subchronic oral administration
24           of Benzo[a]pyrene impairs motor and cognitive behavior and modulates S100B levels and
25           MAPKs in rats. Neurochem Res 39: 731-740. http://dx.doi.org/10.1007/sll064-014-1261-y
26    Mackall, CL; Fleisher, TA; Brown, MR; Andrich, MP; Chen, CC; Feuerstein, IM; Horowitz, ME; Magrath, IT;
27           Shad, AT; Steinberg, SM. (1995). Age, thymopoiesis, and CD4+ T-lymphocyte regeneration after
28           intensive chemotherapy. N Engl J Med 332: 143-149.
29           http://dx.doi.org/10.1056/NEJM199501193320303
30    Mackenzie,  KM; Angevine, DM. (1981). Infertility in mice exposed in utero to benzo(a)pyrene. Biol
31           Reprod 24: 183-192.
32    Mahler, BJ; Van Metre, PC; Bashara, TJ; Wilson, JT; Johns, DA. (2005). Parking lot sealcoat: an
33           unrecognized source of urban polycyclic aromatic hydrocarbons. Environ Sci Technol 39: 5560-
34           5566. http://dx.doi.org/10.1021/es0501565
35    Malaveille, C; Kuroki, T; Sims, P; Grover, PL; Bartsch, H. (1977).  Mutagenicity of isomeric diol-epoxides of
36           benzo[a]pyrene and benz[a]anthracene in S. typhimurium TA98 and TA100 and in V79 Chinese
37           hamster cells. Mutat Res 44: 313-326.
38    Mancini. R;  Romano. G; Sgambato. A; Flamini. G; Giovagnoli. MR; Boninsegna. A; Carraro. C; Vecchione.
39           A; Cittadini. A. (1999). Polycyclic aromatic hydrocarbon-DNA adducts in cervical smears of
40           smokers and nonsmokers. Gynecol Oncol 75: 68-71.  http://dx.doi.org/10.1006/gyno.1999.5525
41    Marshall. G; Vousden. KH; Phillips. DH. (1984). Activation of c-Ha-ras-1 proto-oncogene by in vitro
42           modification with a chemical carcinogen, benzo(a)pyrene diol-epoxide. Nature 310: 586-589.
43           http://dx.doi.org/10.1038/310586aO
44    Mass, MJ; Jeffers, AJ; Ross, JA; Nelson, G; Galati, AJ; Stoner, GD; Nesnow, S. (1993). Ki-ras oncogene
45           mutations in tumors and DNA adducts formed by benz[j]aceanthrylene and benzo[a]pyrene in
46           the  lungs of strain A/J mice. Mol Carcinog 8: 186-192.
47           http://dx.doi.org/10.1002/mc.2940080309

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

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Matiasovic, J; Leva, L; Maskova, J; Kummer, V; Faldyna, M. (2008). Effects of postnatal exposure to
 2           benzo[a]pyrene on the immunity of immature rats. Veterinarni Medicina 53: 93-100.
 3    Matikainen, T; Perez, Gl; Jurisicova, A; Pru, JK; Schlezinger, JJ; Ryu, HY; Laine, J; Sakai, T; Korsmeyer, SJ;
 4           Casper, RF; Sherr, DH; Tilly, JL. (2001). Aromatic hydrocarbon receptor-driven Bax gene
 5           expression is required for premature ovarian failure caused by biohazardous environmental
 6           chemicals. Nat Genet 28: 355-360. http://dx.doi.org/10.1038/ng575
 7    Matikainen, TM; Moriyama, T;  Morita, Y; Perez, Gl; Korsmeyer, SJ; Sherr, DH; Tilly, JL. (2002). Ligand
 8           activation of the aromatic hydrocarbon receptor transcription factor drives  Bax-dependent
 9           apoptosis in developing fetal ovarian germ cells. Endocrinology 143: 615-620.
10    Mattison, PR; Thorgeirsson, SS. (1977). Genetic differences in mouse ovarian metabolism of benzo
11           [a]pyrene and oocyte toxicity. Biochem Pharmacol 26: 909-912.
12    Mattison, PR; Thorgeirsson, SS. (1979). Ovarian aryl hydrocarbon hydroxylase  activity and primordial
13           oocyte toxicity of polycyclic aromatic hydrocarbons in  mice. Cancer Res 39:  3471-3475.
14    Mattison, PR; White, 1MB; Nightingale, MR. (1980). The effect of benzo(a)pyrene  on fertility, primordial
15           oocyte number, and ovarian response to pregnant mare's serum gonadotropin. Pediatr
16           Pharmacol 1: 143-151.
17    Matullo, G; Peluso, M; Polidoro, S; Guarrera, S; Munnia, A; Krogh, V; Masala, G; Berrino, F; Panico, S;
18           Tumino, R; Vineis, P; Palli, D. (2003). Combination of DNA repair gene  single nucleotide
19           polymorphisms and increased levels of DNA adducts in a population-based study. Cancer
20           Epidemiol Biomarkers Prev 12: 674-677.
21    Mauderly. JL. (1997). Relevance of particle-induced rat lung tumors for assessing lung carcinogenic
22           hazard and human lung cancer risk [Review]. Environ Health Perspect  105 (Suppl 5): 1337-1346.
23    Maughan. WZ; Muller. SA; Perry. HO; Pittelkow. MR; O'Brien. PC. (1980). Incidence of skin cancers in
24           patients with atopic dermatitis treated with ocal tar: A 25-year follow-up study. J Am Acad
25           Dermatol 3: 612-615.
26    McCallister, MM; Maguire, M; Ramesh, A; Qiao, AM; Sheng, L; Khoshbouei, H; Aschner, M; Ebner, FF;
27           Hood, DB. (2008). Prenatal  exposure to benzo(a)pyrene impairs later-life cortical neuronal
28           function. Neurotoxicology 29: 846-854. http://dx.doi.Org/10.1016/i.neuro.2008.07.008
29    McConnell, EE; Solleveld, HA; Swenberg, JA; Boorman, GA. (1986). Guidelines for combining neoplasms
30           for evaluation of rodent carcinogenesis studies. J Natl Cancer Inst 76: 283-289.
31    Mccoull, KD; Rindgen, D; Blair,  IA; Penning, TM. (1999). Synthesis and characterization of polycyclic
32           aromatic hydrocarbon  o-quinone depurinating N7-guanine adducts. Chem Res Toxicol 12: 237-
33           246. http://dx.doi.org/10.1021/tx980182z
34    Melikian, AA; Sun, P; Prokopczyk, B; EI-Bayoumy, K;  Hoffmann, D; Wang, X; Waggoner, S. (1999).
35           Identification of benzo[a]pyrene metabolites in cervical mucus and DNA adducts in cervical
36           tissues in humans by gas chromatography-mass spectrometry. Cancer Lett 146: 127-134.
37           http://dx.doi.org/10.1016/S0304-3835(99)00203-7
38    Midgette. AS; Baron. JA. (1990). Cigarette smoking and the risk of natural menopause [Review].
39           Epidemiology 1: 474-480.
40    Mielzynska. D; Siwinska. E; Kapka. L; Szyfter. K; Knudsen. LE; Merlo. DF. (2006). The  influence of
41           environmental exposure to complex mixtures including PAHs and lead on genotoxic effects in
42           children living in Upper Silesia, Poland. Mutagenesis 21: 295-304.
43           http://dx.doi.org/10.1093/mutage/gel037
44    Miller, KP; Ramos, KS. (2001). Impact of cellular metabolism on the biological effects of benzo[a]pyrene
45           and related hydrocarbons [Review]. Drug Metab Rev 33: 1-35. http://dx.doi.org/10.1081/DMR-
46           100000138
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-17            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Miller, MM; Plowchalk, PR; Weitzman, GA; London, SIM; Mattison, DR. (1992). The effect of
 2           benzo(a)pyrene on murine ovarian and corpora lutea volumes. Am J Obstet Gynecol 166: 1535-
 3           1541.
 4    Mitropoulos, P; Norman, R. (2005). Occupational nonsolar risk factors of squamous cell carcinoma of the
 5           skin: a population-based case-controlled study. Dermatol Online J 11: 5.
 6    Mohamed. E; Song. WH; Oh. SA; Park. YJ; You. YA; Lee. S; Choi. JY; Kim. YJ; Jo. I; Pang. MG. (2010). The
 7           transgenerational impact of benzo(a)pyrene on murine male fertility. Hum Reprod 25: 2427-
 8           2433. http://dx.doi.org/10.1093/humrep/deq205
 9    Moody, RP; Chu, I. (1995).  Dermal exposure to environmental contaminants in the Great Lakes [Review].
10           Environ Health Perspect 103 Suppl 9: 103-114.
11    Moody, RP; Joncas, J; Richardson, M; Chu, I. (2007). Contaminated soils (I): In vitro dermal absorption of
12           benzo[a]pyrene in  human skin. J Toxicol Environ Health A 70: 1858-1865.
13           http://dx.doi.org/10.1080/15287390701459296
14    Moolgavkar, SH; Knudson,  AG, Jr. (1981). Mutation and cancer: A model for human carcinogenesis. J
15           Natl Cancer Inst 66: 1037-1052.
16    Moolgavkar, SH; Venzon, DJ. (1979). Two-event models for carcinogenesis: Incidence curves for
17           childhood and adult tumors. Math Biosci 47: 55-77. http://dx.doi.org/10.1016/0025-
18           5564(79)90005-1
19    Moulin. JJ; Wild. P; Mur. JM; Lafontaine. M; Lefer. M; Mercier-Gallav. M; Villemot. P; Whebi. V; Coulon.
20           JP. (1989). Risk of lung, larynx, pharynx and buccal cavity cancers among carbon electrode
21           manufacturing workers. Scand J  Work Environ Health 15: 30-37.
22           http://dx.doi.org/10.5271/siweh.1885
23    Munson. AE; White. KL; Lvsy. HH. (1985). Effects of Subchronic Fourteen Day Exposure to Benzopyrene
24           in B6C3F1 Female Mice on Host  Resistance. Submitted under TSCA Section FYI.
25    Munson, AE; White, KU. (1983). Effects of subchronic  14-day exposure to benzopyrene in b6c3fl male
26           and female mice on specific immunological parameters. American Petroleum Institute.
27    N'Diaye, M; Le Ferrec, E; Lagadic-Gossmann, D; Corre, S; Gilot, D; Lecureur, V; Monteiro, P; Rauch, C;
28           Galibert, MD; Fardel, O. (2006). Aryl hydrocarbon receptor- and calcium-dependent induction of
29           the chemokine CCL1 by the environmental contaminant benzo[a]pyrene. J Biol Chem 281:
30           19906-19915. http://dx.doi.org/10.1074/ibc.M601192200
31    Nanez, A; Ramos, IN; Ramos, KS. (2011). A mutant Ahr allele protects the embryonic kidney from
32           hydrocarbon-induced deficits  in  fetal programming. Environ Health Perspect 119: 1745-1753.
33           http://dx.doi.org/10.1289/ehp.1103692
34    Naslund, I; Rubio, CA; Auer, GU. (1987). Nuclear DNA changes during pathogenesis of squamous
35           carcinoma of the cervix in 3,4-benzopyrene-treated mice. Anal Quant Cytol Histol 9: 411-418.
36    Naufal. Z; Zhiwen. L; Zhu. L; Zhou. GD; Mcdonald. T; He. LY; Mitchell. L; Ren. A; Zhu. H; Finnell. R;
37           Donnelly, KC. (2010). Biomarkers of exposure to combustion by-products in a human population
38           in Shanxi, China. J Expo Sci Environ Epidemiol 20: 310-319.
39           http://dx.doi.org/10.1038/ies.2009.19
40    Neal. J; Rigdon. RH. (1967). Gastric tumors in mice fed benzo(a)pyrene: A quantitative study. Tex Rep
41           Biol Med 25: 553-557.
42    Neal. MS; Mulligan Tuttle. AM; Casper. RF; Lagunov. A; Foster. WG. (2010). Aryl hydrocarbon receptor
43           antagonists attenuate the deleterious effects of benzo[a]pyrene on isolated rat follicle
44           development. Reprod Biomed Online 21: 100-108.
45           http://dx.doi.0rg/10.1016/i.rbmo.2010.03.025
46    Neal, MS; Zhu, J; Foster, WG. (2008). Quantification of benzo[a]pyrene and other PAHs in the serum and
47           follicular fluid of smokers versus non-smokers. Reprod Toxicol 25: 100-106.
48           http://dx.doi.0rg/10.1016/i.reprotox.2007.10.012
                This document is a draft for  review purposes only and does not constitute Agency policy.
                                                 R-18            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Neal, MS; Zhu, J; Holloway, AC; Foster, WG. (2007). Follicle growth is inhibited by benzo-[a]-pyrene, at
 2           concentrations representative of human exposure, in an isolated rat follicle culture assay. Hum
 3           Reprod 22: 961-967. http://dx.doi.org/10.1093/humrep/del487
 4    Nebert, DW; Puga, A; Vasiliou, V. (1993). Role of the Ah receptor and the dioxin-inducible [Ah] gene
 5           battery in toxicity, cancer, and signal transduction [Review]. Ann N Y Acad Sci 685: 624-640.
 6           http://dx.doi.0rg/10.llll/i.1749-6632.1993.tb35928.x
 7    Nesnow, S; Mass, MJ; Ross, JA; Galati, AJ; Lambert, GR; Gennings, C; Carter, WH; Stoner, GD. (1998a).
 8           Lung tumorigenic interactions in strain A/J mice of five environmental polycyclic aromatic
 9           hydrocarbons. Environ Health Perspect 106 Suppl 6: 1337-1346.
10    Nesnow, S; Ross, JA; Mass, MJ; Stoner, GD. (1998b). Mechanistic relationships between DNA adducts,
11           oncogene mutations, and lung tumorigenesis in strain A mice. Exp Lung Res 24: 395-405.
12    Nesnow, S; Ross, JA; Stoner, GD; Mass, MJ. (1995). Mechanistic linkage between DNA adducts,
13           mutations in oncogenes and tumorigenesis of carcinogenic environmental polycyclic aromatic
14           hydrocarbons in strain A/J mice. Toxicology 105: 403-413.  http://dx.doi.org/10.1016/0300-
15           483X(95)03238-B
16    Nesnow, S; Ross, JA; Stoner, GD; Mass, MJ. (1996). Tumorigenesis of Carcinogenic Environmental
17           Polycyclic Aromatic Hydrocarbons in Strain A/J Mice: Linkage to DNA Adducts and Mutations in
18           Oncogenes. Polycycl Aromat Compd 10: 259-266.
19           http://dx.doi.org/10.1080/10406639608034705
20    Nesnow. S; Triplett. LL;  Slaga. TJ. (1983). Mouse skin tumor initiation-promotion and complete
21           carcinogenesis  bioassays: mechanisms and biological activities of emission samples. Environ
22           Health Perspect 47: 255-268.
23    Nettesheim.  P; Griesemer. RA; Martin. DH; Caton. JE. Jr. (1977). Induction of preneoplastic and
24           neoplastic lesions in grafted rat tracheas continuously exposed to benzo(a)pyrene. Cancer Res
25           37: 1272-1278.
26    Newbold, RF; Brookes,  P. (1976). Exceptional mutagenicity of a benzo(a)pyrene diol epoxide in cultured
27           mammalian cells. Nature 261: 52-54. http://dx.doi.org/10.1038/261052aO
28    Nikonova, TV. (1977). [Transplacental  effect of benz(a)pyrene and  pyrene]. Biull Eksp Biol Med 84: 88-
29           91.
30    NIOSH (National Institute for Occupational Safety and Health). (1989). Assessment of the
31           cocarcinogenic/promoting activity of asphalt fumes.
32           http://www.ntis.gov/search/product.aspx?ABBR=PB91110213
33    Niu, Oj Zhang, H; Li, X; Li, M.  (2010). Benzo[a]pyrene-induced neurobehavioral function and
34           neurotransmitter alterations in coke oven  workers. Occup Environ Med 67: 444-448.
35           http://dx.doi.org/10.1136/oem.2009.047969
36    NRC (National Research Council). (1983). Risk assessment in the federal government: Managing the
37           process. Washington, DC: National Academies Press.
38           http://www.nap.edu/openbook.php7record  id=366&page=Rl
39    NRC (National Research Council). (2009). Science and decisions: Advancing risk assessment. Washington,
40           DC: National Academies Press, http://www.nap.edu/catalog/12209.html
41    NRC (National Research Council). (2011). Review of the Environmental Protection Agency's draft IRIS
42           assessment of formaldehyde.  Washington, DC: National Academies Press.
43           http://www.nap.edu/catalog/13142.html
44    OECD (Organisation for Economic Co-operation and Development). (2009). Series on testing and
45           assesment: Number 39. Guidance document on acute inhalation toxicity testing [OECD SIDS].
46           (ENV/JM/MONO(2009)28; JT03268144). Paris, Fr.
47           http://search.oecd. org/officialdocuments/displavdocumentpdf/?cote=env/im/mono%282009%
48           2928&doclanguage=en
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-19             DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1    Okey, AB; Riddick, PS; Harper, PA. (1994). Molecular biology of the aromatic hydrocarbon (dioxin)
 2           receptor [Review]. Trends Pharmacol Sci 15: 226-232. http://dx.doi.org/10.1016/0165-
 3           6147(94)90316-6
 4    Olsen, AK; Andreassen, A; Singh, R; Wiger, R; Duale, N; Farmer, PB; Brunborg, G. (2010). Environmental
 5           exposure of the mouse germ line: DNA adducts in spermatozoa and formation of de novo
 6           mutations during spermatogenesis.  PLoS ONE 5: el!349.
 7           http://dx.doi.org/10.1371/iournal.pone.0011349
 8    Olsson, A; Kromhout, H; Agostini, M; Hansen, J; Lassen, CF; Johansen, C; Kjaerheim, K; Langard, S;
 9           Stucker, I; Ahrens, W; Behrens, T; Lindbohm,  ML; Heikkila, P; Heederik, D; Portengen,  L; Shaham,
10           J; Ferro, G; de Vocht, F;  Burstyn,  I; Boffetta, P. (2010). A case-control study of lung cancer nested
11           in a cohort of European asphalt workers. Environ Health Perspect 118: 1418-1424.
12           http://dx.doi.org/10.1289/ehp.0901800
13    ONS (Office for National Statistics). (2003). Cancer statistics registrations: Registrations of cancer
14           diagnosed in 2003, England. London, UK: Palgrave Macmillan.
15           http://www.ons.gov.uk/ons/rel/vsobl/cancer-statistics-registrations--england--series-mbl-/no-
16           -34-2003/index.html
17    Park, JH; Gelhaus, S; Vedantam, S; Oliva,  AL; Batra,  A; Blair, IA; Troxel, AB; Field, J; Penning, TM.  (2008).
18           The pattern of p53 mutations caused by PAH o-quinones is driven by 8-oxo-dGuo formation
19           while the spectrum of mutations is determined by biological selection for dominance. Chem Res
20           Toxicol 21: 1039-1049. http://dx.doi.org/10.1021/tx700404a
21    Park. JH; Troxel. AB; Harvey. RG; Penning. TM. (2006). Polycyclic aromatic hydrocarbon (PAH) o-
22           quinones produced by the aldo-keto-reductases (AKRs) generate abasic sites, oxidized
23           pyrimidines, and 8-oxo-dGuo via reactive oxygen species. Chem Res Toxicol 19: 719-728.
24           http://dx.doi.org/10.1021/tx0600245
25    Pate Capps, N; Stewart, A; Burns, C. (2009). The interplay between secondhand cigarette smoke,
26           genetics, and cervical cancer: a review of the literature [Review]. Biol Res Nurs 10: 392-399.
27           http://dx.doi.org/10.1177/1099800408330849
28    Pavanello, S; Favretto, D; Brugnone, F; Mastrangelo, G; Dal Pra, G; Clonfero, E. (1999).
29           HPLC/fluorescence determination of anti-BPDE-DNA adducts in mononuclear white blood cells
30           from PAH-exposed humans. Carcinogenesis 20: 431-435.
31    Pavanello, S; Pulliero, A; Siwinska, E; Mielzynska, D; Clonfero, E. (2005). Reduced nucleotide excision
32           repair and GSTMl-null genotypes influence anti-B[a]PDE-DNA adduct levels in mononuclear
33           white blood cells of highly PAH-exposed coke oven workers. Carcinogenesis 26: 169-175.
34           http://dx.doi.org/10.1093/carcin/bgh303
35    Pavanello, S; Siwinska, E; Mielzynska, D; Clonfero, E. (2004). GSTM1 null genotype as a risk factor for
36           anti-BPDE-DNA adduct formation in mononuclear white blood cells of coke-oven workers. Mutat
37           Res 558: 53-62. http://dx.doi.Org/10.1016/i.mrgentox.2003.10.019
38    Penit. C; Vasseur. F. (1989). Cell  proliferation and differentiation in the fetal and early postnatal mouse
39           thymus. J Immunol 142: 3369-3377.
40    Penning. TM; Burczynski. ME; Hung. CF; Mccoull. KD;  Palackal. NT;Tsuruda. LS.  (1999). Dihydrodiol
41           dehydrogenases and polycyclic aromatic hydrocarbon activation: generation of reactive and
42           redox active o-quinones [Review]. Chem Res Toxicol 12:  1-18.
43           http://dx.doi.org/10.1021/tx980143n
44    Penning, TM; Ohnishi, ST; Ohnishi, T; Harvey, RG. (1996). Generation of reactive oxygen species during
45           the enzymatic oxidation of polycyclic aromatic hydrocarbon trans-dihydrodiols catalyzed by
46           dihydrodiol dehydrogenase. Chem Res Toxicol 9: 84-92. http://dx.doi.org/10.1021/tx950055s
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-20            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Perera, F; Li, TY; Lin, C; Tang, D. (2012a). Effects of prenatal polycyclic aromatic hydrocarbon exposure
 2           and environmental tobacco smoke on child IQ in a Chinese cohort. Environ Res 114: 40-46.
 3           http://dx.doi.0rg/10.1016/i.envres.2011.12.011
 4    Perera, FP; Li, Z; Whyatt, R; Hoepner, L; Wang, S; Camann, D; Rauh, V. (2009). Prenatal airborne
 5           polycyclic aromatic hydrocarbon exposure and child IQ at age 5 years. Pediatrics 124: e!95-
 6           e202. http://dx.doi.org/10.1542/peds.2008-3506
 7    Perera, FP; Rauh, V; Whyatt, RM; Tang, D; Tsai, WY; Bernert, JT; Tu, YH; Andrews, H; Barr, DB; Camann,
 8           DE; Diaz, D; Dietrich, J; Reyes, A; Kinney, PL. (2005a). A summary of recent findings on birth
 9           outcomes and developmental effects of prenatal ETS,  PAH, and pesticide exposures [Review].
10           Neurotoxicology 26: 573-587. http://dx.doi.Org/10.1016/i.neuro.2004.07.007
11    Perera, FP; Rauh, V; Whyatt, RM; Tsai, WY; Bernert, JT; Tu, YH; Andrews, H; Ramirez, J; Qu, L; Tang, D.
12           (2004). Molecular evidence of an interaction between prenatal environmental exposure and
13           birth outcomes in a multiethnic population. Environ Health Perspect 112: 626-630.
14    Perera. FP; Tang. D; Rauh. V; Lester. K; Tsai. WY; Tu. YH; Weiss. L; Hoepner. L; King. J; Del Priore. G;
15           Lederman,  SA. (2005b). Relationships among polycyclic aromatic hydrocarbon-DNA adducts,
16           proximity to the World Trade Center, and effects on fetal growth. Environ Health Perspect 113:
17           1062-1067.
18    Perera, FP; Tang, D; Wang, S; Vishnevetskv, J; Zhang, B; Diaz, D; Camann, D; Rauh, V. (2012b). Prenatal
19           Polycyclic Aromatic Hydrocarbon (PAH) Exposure and  Child Behavior at Age 6-7 Years. Environ
20           Health Perspect 120: 921-926. http://dx.doi.org/10.1289/ehp.1104315
21    Perera. FP; Weinstein. IB. (2000). Molecular epidemiology: Recent advances and future directions
22           [Review]. Carcinogenesis 21: 517-524. http://dx.doi.Org/10.1093/carcin/21.3.517
23    Perrin. J; Tassistro. V;  Mandon. M; Grillo. JM; Botta. A; Sari-Minodier. I. (2011a). Tobacco consumption
24           and benzo(a)pyrene-diol-epoxide-DNA adducts in spermatozoa: in smokers, swim-up procedure
25           selects spermatozoa with decreased DNA damage. Fertil Steril 95:  2013-2017.
26           http://dx.doi.0rg/10.1016/i.fertnstert.2011.02.021
27    Perrin, J; Tassistro, V;  Paulmyer-Lacroix, O; Courbiere,  B; Botta, A; Sari-Minodier, I. (2011b). In smokers,
28           swim-up and discontinuous gradient centrifugation recover spermatozoa with equally lower
29           amounts of DNA damage than spermatozoa obtained  from neat semen. Fertil Steril 95: 2680-
30           2682. http://dx.doi.0rg/10.1016/i.fertnstert.2011.04.092
31    Petrie, HT. (2002). Role of thymic organ structure and stromal  composition in steady-state postnatal T-
32           cell production [Review]. Immunol Rev 189: 8-19.
33    Pfeifer, GP; Denissenko, MF; Olivier, M; Tretvakova, N; Hecht,  SS; Hainaut, P. (2002). Tobacco smoke
34           carcinogens, DNA damage and p53 mutations  in smoking-associated cancers [Review].
35           Oncogene 21: 7435-7451. http://dx.doi.org/10.1038/si.onc.1205803
36    Pfeifer, GP; Hainaut, P. (2003). On the origin of G —> T transversions in lung cancer [Review]. Mutat Res
37           526: 39-43. http://dx.doi.org/10.1016/S0027-5107(03)00013-7
38    Pfeiffer. CA; Allen. E. (1948). Attempts to produce cancer in rhesus monkeys with carcinogenic
39           hydrocarbons and estrogens. Cancer Res 8: 97-127.
40    Pfeiffer. EH. (1977). Oncogenic interaction of carcinogenic and non-carcinogenic polycyclic aromatic
41           hydrocarbons in mice. In Air pollution and cancer in man. Lyon, France: IARC.
42    Phillips. PH. (2002). Smoking-related DNA and protein adducts in human tissues [Review].
43           Carcinogenesis 23: 1979-2004.
44    Pittelkow, MR; Perry,  HO; Muller, SA; Maughan, WZ; O'Brien, PC. (1981). Skin cancer in patients with
45           psoriasis treated with coal tar: A 25-year follow-up study. Arch Dermatol  117: 465-468.
46    Poel, WE. (1959). Effect of carcinogenic dosage and duration of exposure on skin-tumor induction in
47           mice.  J Natl Cancer Inst 22: 19-43.

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

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1    Poel, WE. (1963). Skin as a test site for the bioassay of carcinogens and carcinogen precursors [IARC
 2           Monograph]. In Biology of cutaneous cancers (pp. 611-631). Poel, WE.
 3    Postlind, H; Vu, TP; Tukey, RH; Quattrochi, LC. (1993). Response of human CYPl-luciferase plasmids to
 4           2,3,7,8-tetrachlorodibenzo-p-dioxin and polycyclic aromatic hydrocarbons. Toxicol Appl
 5           Pharmacol 118: 255-262. http://dx.doi.org/10.1006/taap.1993.1031
 6    Pru, JK; Kaneko-Tarui, T; Jurisicova, A; Kashiwagi, A; Selesniemi, K; Tilly, JL. (2009). Induction of
 7           proapoptotic gene expression and recruitment of p53 herald ovarian follicle loss caused by
 8           polycyclic aromatic hydrocarbons. Reproductive Sciences 16: 347-356.
 9           http://dx.doi.org/10.1177/1933719108327596
10    Puisieux, A; Lim, S; Groopman, J; Ozturk, M. (1991). Selective targeting of p53 gene mutational hotspots
11           in human cancers by etiologically defined carcinogens. Cancer Res 51: 6185-6189.
12    Pukkala, E. (1995). Cancer risk by social class and occupation : a survey of 109,000 cancer cases among
13           Finns of working age. In J Wahrendorf (Ed.). Basel, Switzerland: Karger.
14    Pukkala, E; Martinsen, J;  Lynge,  E; Gunnarsdottir, H; Sparen, P; Trvggvadottir, L; Weiderpass, E;
15           Kjaerheim, K. (2009). Occupation and cancer - follow-up of 15 million people in five Nordic
16           countries. Acta Oncol 48: 646-790. http://dx.doi.org/10.1080/02841860902913546
17    OJu, C; Cheng, S; Xia, Y; Peng, B; Tang, Q; Tu, B. (2011).  Effects of subchronic benzo(a)pyrene exposure
18           on neurotransmitter receptor gene  expression  in the rat hippocampus related with spatial
19           learning and memory change. Toxicology 289: 83-90.
20           http://dx.doi.0rg/10.1016/i.tox.2011.07.012
21    Oju. C; Peng. B; Cheng. S; Xia. Y; Tu. B. (2013). The effect of occupational  exposure to benzo[a]pyrene on
22           neurobehavioral function in coke oven workers. Am J Ind Med 56: 347-355.
23           http://dx.doi.org/10.1002/aiim.22119
24    Ramesh, A; Inyang, F; Lunstra, DP; Niaz, MS; Kopsombut, P; Jones, KM; Hood, DB; Hills, ER; Archibong,
25           AE^pOOS). Alteration of fertility endpoints in adult male F-344 rats by subchronic exposure to
26           inhaled benzo(a)pyrene. Exp Toxicol Pathol 60: 269-280.
27           http://dx.doi.0rg/10.1016/i.etp.2008.02.010
28    Ramesh, A; Walker, SA; Hood, DB; Guillen, MD; Schneider, K; Weyand, EH. (2004). Bioavailability and risk
29           assessment of orally ingested polycyclic aromatic hydrocarbons [Review]. Int J Toxicol 23: 301-
30           333. http://dx.doi.org/10.1080/10915810490517063
31    Ramos, KS;  Moorthy, B. (2005).  Bioactivation of polycyclic aromatic hydrocarbon carcinogens within the
32           vascular wall: implications for human atherogenesis [Review]. Drug Metab Rev 37: 595-610.
33           http://dx.doi.org/10.1080/03602530500251253
34    Rask-Nielsen, R. (1950). On the susceptibility of the thymus, lung, subcutaneous and mammary tissues in
35           strain street mice to direct application of small  doses of four different carcinogenic
36           hydrocarbons. Br J Cancer 4: 108-116.
37    Raveh, D; Slaga, TJ; Huberman, E.  (1982).  Cell-mediated mutagenesis and tumor-initiating activity of the
38           ubiquitous polycyclic hydrocarbon, cyclopenta[c,d]pyrene. Carcinogenesis 3: 763-766.
39           http://dx.doi.Org/10.1093/carcin/3.7.763
40    Raychoudhury. SS; Kubinski. D. (2003). Polycyclic aromatic hydrocarbon-induced  cytotoxicity in cultured
41           rat Sertoli cells involves differential  apoptotic response. Environ Health Perspect 111: 33-38.
42    Revel. A; Raanani. H; Younglai. E; Xu. J; Han. R; Savouret. JF; Casper. RF. (2001). Resveratrol, a natural
43           aryl hydrocarbon receptor antagonist, protects sperm from DNA  damage and apoptosis caused
44           by benzo(a)pyrene. Reprod Toxicol 15: 479-486.
45    Rice, JE; Makowski, GS; Hosted, TJ; Lavoie, EJ. (1985). Methylene-bridged bay region chrysene and
46           phenanthrene derivatives and their keto-analogs: mutagenicity in Salmonella typhimurium and
47           tumor-initiating activity on mouse skin. Cancer Lett 27: 199-206.
48           http://dx.doi.org/10.1016/0304-3835(84)90119-8
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-22             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Robinson, M; Laurie, RD; Bull, RJ; Stober, JA. (1987). Carcinogenic effects in A/J mice of particulate of a
 2           coal tar paint used in potable water systems. Cancer Lett 34: 49-54.
 3           http://dx.doi.org/10.1016/0304-3835(87)90072-3
 4    Robles. R; Morita. Y; Mann. KK; Perez. Gl; Yang. S; Matikainen. T; Sherr. DH; Tilly. JL (2000). The aryl
 5           hydrocarbon receptor, a basic helix-loop-helix transcription factor of the PAS gene family, is
 6           required for normal  ovarian germ cell dynamics in the mouse. Endocrinology 141: 450-453.
 7    Rodriguez, JW; Kirlin, WG; Wirsiy, YG; Matheravidathu, S; Hodge, TW;  Urso, P. (1999). Maternal
 8           exposure to benzo[a]pyrene alters development of T lymphocytes in offspring.
 9           Immunopharmacol Immunotoxicol 21: 379-396.  http://dx.doi.org/10.3109/08923979909052769
10    Roe, FJ; Peto, R; Kearns, F; Bishop, D.  (1970). The mechanism of carcinogenesis by the neutral fraction of
11           cigarette smoke condensate.  Br J Cancer 24: 788-806.
12    Roe, FJ; Waters, MA. (1967). Induction of hepatoma in mice by carcinogens of the polycyclic
13           hydrocarbon type. Nature 214: 299-300.
14    Roelofzen, JH; Aben, KK; Oldenhof, UT; Coenraads, PJ; Alkemade, HA; van de Kerkhof, PC; van der Valk,
15           PG; Kiemeney, LA. (2010). No increased risk of cancer after coal tar treatment in patients with
16           psoriasis or eczema. J Invest Dermatol 130: 953-961. http://dx.doi.org/10.1038/iid.2009.389
17    Rogan, EG; Devanesan, PD; Ramakrishna, NV; Higginbotham, S; Padmavathi,  NS; Chapman,  K; Cavalieri,
18           EL; Jeong, H; Jankowiak, R; Small, GJ. (1993). Identification and quantitation of benzo[a]pyrene-
19           DNA adducts formed in mouse skin. Chem Res Toxicol  6: 356-363.
20           http://dx.doi.org/10.1021/tx00033a017
21    Rojas. M; Alexandrov. K; Cascorbi. I; Brockmoller. J; Likhachev. A; Pozharisski. K; Bouvier. G; Auburtin. G;
22           Mayer. L; Kopp-Schneider. A; Roots. I; Bartsch. H. (1998). High benzo[a]pyrene diol-epoxide DNA
23           adduct levels in lung and blood cells from individuals with combined CYP1A1 Mspl/Msp-
24           GSTM1*0/*0 genotypes. Pharmacogenetics 8: 109-118.
25    Rojas, M; Cascorbi,  I; Alexandrov, K; Kriek, E; Auburtin, G; Mayer, L; Kopp-Schneider, A; Roots, I; Bartsch,
26           hL (2000). Modulation of benzo[a]pyrene diolepoxide-DNA adduct levels in human white  blood
27           cells by CYP1A1, GSTM1 and GSTT1 polymorphism. Carcinogenesis 21: 35-41.
28           http://dx.doi.0rg/10.1093/carcin/21.l.35
29    Rojas, M; Godschalk, R; Alexandrov, K; Cascorbi, I; Kriek,  E; Ostertag, J; Van Schooten, FJ; Bartsch, H.
30           (2001). Myeloperoxidase-463A variant reduces benzo[a]pyrene diol epoxide DNA adducts in
31           skin of coal tar treated patients. Carcinogenesis 22:  1015-1018.
32           http://dx.doi.0rg/10.1093/carcin/22.7.1015
33    Rojas, M; Marie, B; Vignaud, JM; Martinet, N; Siat, J; Grosdidier, G; Cascorbi, I; Alexandrov,  K. (2004).
34           High DNA damage by benzo[a]pyrene 7,8-diol-9,10-epoxide in bronchial epithelial cells from
35           patients with lung cancer: comparison with lung parenchyma. Cancer Lett 207: 157-163.
36           http://dx.doi.0rg/10.1016/i.canlet.2003.ll.016
37    Rothman, KJ; Greenland, S. (1998). Modern epidemiology (2nd ed.). Philadelphia, PA: Lippincott,
38           Williams, &Wilkins.
39    Rowe. AA; O'Connor. TP. (2011). Assessment of water quality of runoff from sealed asphalt surfaces.
40           (EPA/600/R-10/178). Cincinnati, OH: U.S. Environmental Protection Agency.
41    Rubes. J;  Rybar. R; Prinosilova. P; Veznik. Z; Chvatalova. I; Solansky. I; Sram. RJ. (2010). Genetic
42           polymorphisms influence the susceptibility of men to sperm DNA damage associated with
43           exposure to air pollution. Mutat Res 683: 9-15.
44           http://dx.doi.0rg/10.1016/i.mrfmmm.2009.09.010
45    Ruchirawat, M; Navasumrit,  P; Settachan, D. (2010). Exposure to benzene in various susceptible
46           populations: co-exposures to 1,3-butadiene and  PAHs  and implications for carcinogenic risk.
47           Chem Biol Interact 184: 67-76. http://dx.doi.Org/10.1016/i.cbi.2009.12.026

                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-23             DRAFT—DO NOT CITE OR  QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Ruggeri, B; Dirado, M; Zhang, SY; Bauer, B; Goodrow, T; Kleinszanto, AJP. (1993). Benzofalpyrene-
 2           induced murine skin tumors exhibit frequent and characteristic G to T mutations in the p53
 3           gene. PNAS90: 1013-1017. http://dx.doi.Org/10.1073/pnas.90.3.1013
 4    Sadeu, JC; Foster, WG. (2011). Cigarette smoke condensate exposure delays follicular development and
 5           function in a stage-dependent manner. Fertil Steril 95: 2410-2417.
 6           http://dx.doi.0rg/10.1016/i.fertnstert.2011.03.072
 7    Saffiotti, U; Montesano, R; Sellakumar, AR; Kaufman, DG. (1972). Respiratory tract carcinogenesis
 8           induced in hamsters by different dose levels of benzo[a]pyrene and ferric oxide. J Natl Cancer
 9           Inst 49: 1199-1204.
10    Sakkas, D; Alvarez, JG. (2010). Sperm DNA fragmentation: mechanisms of origin, impact on reproductive
11           outcome, and analysis [Review].  Fertil Steril 93: 1027-1036.
12           http://dx.doi.0rg/10.1016/i.fertnstert.2009.10.046
13    Saunders, CR; Das, SK; Ramesh, A; Shockley,  DC; Mukherjee, S. (2006). Benzo(a)pyrene-induced acute
14           neurotoxicity in the F-344 rat: role of oxidative stress. J Appl Toxicol 26: 427-438.
15           http://dx.doi.org/10.1002/iat.1157
16    Saunders, CR; Ramesh, A; Shockley, DC. (2002). Modulation of neurotoxic behavior in F-344 rats by
17           temporal disposition of benzo(a)pyrene. Toxicol  Lett 129: 33-45.
18    Saunders, CR; Shockley, DC; Knuckles, ME. (2001). Behavioral effects induced by acute exposure to
19           benzo(a)pyrene in F-344 rats. Neurotox Res 3:  557-579. http://dx.doi.org/10.1007/BF03033211
20    Schmahl. D; Schmidt. KG; Habs. M. (1977). Syncarcinogenic action of polycyclic hydrocarbons in
21           automobile exhaust gas condensates. In U Mohr; D Schmahl; L Tomatis; W Davis (Eds.), Air
22           pollution and cancer in man (pp.  53-59). Lyon,  France: IARC.
23    Schmidt. JV; Su. GH;  Reddy. JK; Simon. MC; Bradfield. CA. (1996). Characterization of a murine Ahr null
24           allele: involvement of the Ah receptor in hepatic growth and development. PNAS 93: 6731-6736.
25    Schmidt,  KG; Schmahl, D; Misfeld, J.  (1973). Investigations on the carcinogenic burden by air pollution in
26           man. VI. Experimental investigations to determine a dose-response relationship and to estimate
27           a threshold dose of benzo(a)pyrene in the skin of two different mouse strains. Zentralbl
28           Bakteriol, Parasitenkd, Infektionskrankh Hyg, Abt 1: Orig, Reihe B 158: 62-68.
29    Schnizlein, CT; Munson, AE; Rhoades, RA. (1987). Immunomodulation  of local and systemic immunity
30           after subchronic pulmonary exposure of mice to benzo(a)pyrene. International Journal of
31           Immunopharmacology 9: 99-106.
32    Schonland, SO; Zimmer, JK; Lopez-Benitez, CM; Widmann, T; Ramin, KD; Goronzy, JJ; Weyand, CM.
33           (2003). Homeostatic control of T-cell  generation in neonates. Blood 102: 1428-1434.
34           http://dx.doi.org/10.1182/blood-2002-ll-3591
35    Secretan, B; Straif, K; Baan, R; Grosse, Y;  El Ghissassi, F; Bouvard, V; Benbrahim-Tallaa, L; Guha, N;
36           Freeman, C;  Galichet, L; Cogliano, V. (2009). A  review of human carcinogens-Part E: tobacco,
37           areca nut, alcohol, coal smoke, and salted fish. Lancet Oncol 10: 1033-1034.
38    Seli. E; Gardner. DK;  Schoolcraft. WB; Moffatt. O; Sakkas. D. (2004). Extent of nuclear DNA damage in
39           ejaculated spermatozoa impacts  on blastocyst development after in vitro fertilization. Fertil
40           Steril 82: 378-383. http://dx.doi.Org/10.1016/i.fertnstert.2003.12.039
41    Shamsuddin. AK; Gan. R. (1988).  Immunocytochemical localization  of benzo(a)pyrene-DNA adducts in
42           human tissue. Hum Pathol 19: 309-315. http://dx.doi.org/10.1016/S0046-8177(88)80524-0
43    Sharovskava, J; Kobliakova, I; Solomatina, N; Kobliakov, V. (2006). Effect of some carcinogenic and non-
44           carcinogenic polycyclic aromatic  hydrocarbons on gap junction intercellular communication in
45           hepatoma cell cultures. Eur J Cell Biol 85: 387-397. http://dx.doi.Org/10.1016/i.eicb.2005.ll.006
46    Shen. H; Spitz. MR; Qiao. Y; Guo. Z; Wang. LE; Bosken. CH; Amos. Cl; Wei. Q. (2003). Smoking, DNA
47           repair capacity and risk of nonsmall cell lung cancer. Int J Cancer 107: 84-88.
48           http://dx.doi.org/10.1002/iic.11346
                 This document is a draft for review purposes only and does  not constitute Agency policy.
                                                  R-24            DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1    Shen, YM; Troxel, AB; Vedantam, S; Penning, TM; Field, J. (2006). Comparison of p53 mutations induced
 2           by PAH o-quinones with those caused by anti-benzo[a]pyrene diol epoxide in vitro: role of
 3           reactive oxygen and biological selection. Chem Res Toxicol 19: 1441-1450.
 4           http://dx.doi.org/10.1021/tx0601206
 5    Sheng, L, iu; Ding, X; Ferguson, M; Mccallister, M; Rhoades, R; Maguire, M; Ramesh, A; Aschner, M;
 6           Campbell, D; Levitt, P, at; Hood, DB. (2010). Prenatal Polycyclic Aromatic Hydrocarbon Exposure
 7           Leads to Behavioral Deficits and Downregulation of Receptor Tyrosine Kinase, MET. Toxicol Sci
 8           118: 625-634. http://dx.doi.org/10.1093/toxsci/kfq304
 9    Shimizu, Y;  Nakatsuru, Y; Ichinose, M; Takahashi, Y; Kume, H; Mimura, J; Fujii-Kuriyama, Y; Ishikawa, T.
10           (2000). Benzo[a]pyrene carcinogenicity is lost in mice lacking the aryl hydrocarbon  receptor.
11           PNAS97: 779-782. http://dx.doi.Org/10.1073/pnas.97.2.779
12    Silverstone, AE; Cantor, H; Goldstein, G; Baltimore, D. (1976). Terminal deoxynucleotidyl transferase is
13           found in prothymocytes. J Exp Med 144: 543-548.
14    Sivak, A; Niemeier, R; Lynch, D; Beltis, K; Simon, S; Salomon, R; Latta, R; Belinky, B; Menzies, K; Lunsford,
15           A; Cooper, C; Ross, A; Bruner, R. (1997). Skin carcinogenicity of condensed asphalt roofing fumes
16           and their fractions following dermal application to mice. Cancer Lett 117: 113-123.
17           http://dx.doi.org/10.1016/S0304-3835(97)00214-0
18    Slaga, TJ; Gleason, GL; Mills, G; Ewald,  L; Fu, PP; Lee,  HM; Harvey, RG. (1980). Comparison of the skin
19           tumor-initiating activities of dihydrodiols and diol-epoxides of various polycyclic aromatic
20           hydrocarbons. Cancer Res 40: 1981-1984.
21    Slaga. TJ; Huberman. E; Selkirk. JK; Harvey. RG; Bracken. WM. (1978). Carcinogenicity and mutagenicity
22           of benz(a)anthracene diols and diol-epoxides. Cancer Res 38: 1699-1704.
23    Scares. SR;  Melo. MA. (2008). Cigarette smoking and reproductive function [Review]. Curr Opin Obstet
24           Gynecol 20:  281-291. http://dx.doi.org/10.1097/GCO.Ob013e3282fc9cle
25    Soballe, PW; Montone, KT; Satyamoorthy, K; Nesbit,  M; Herlyn, M. (1996). Carcinogenesis in human skin
26           grafted to SCID mice. Cancer Res 56: 757-764.
27    Spinelli, JJ;  Band, PR; Svirchev, LM; Gallagher, RP. (1991). Mortality and cancer incidence in aluminum
28           reduction plant workers. J Occup Med 33: 1150-1155.
29    Spinelli, JJ;  Demers, PA; Le, IMP; Friesen, MD; Lorenzi, MF; Fang, R; Gallagher, RP. (2006). Cancer risk in
30           aluminum reduction plant workers (Canada). Cancer Causes Control 17: 939-948.
31           http://dx.doi.org/10.1007/sl0552-006-0031-9
32    Steiner, PE. (1955). Carcinogenicity of multiple chemicals simultaneously administered. Cancer Res 15:
33           632-635.
34    Stephanou, P; Konstandi, M; Pappas, P; Marselos, M. (1998). Alterations in central monoaminergic
35           neurotransmission induced by polycyclic aromatic hydrocarbons in rats. European Journal of
36           Drug Metabolism and Pharmacokinetics 23: 475-481. http://dx.doi.org/10.1007/BF03189998
37    Stern, RS; Laird, N. (1994). The carcinogenic risk of treatments for severe psoriasis. Photochemotherapy
38           Follow-up Study. Cancer 73: 2759-2764.
39    Stern. RS; Liebman. EJ; Vakeva. L. (1998). Oral psoralen and ultraviolet-A light (PUVA) treatment of
40           psoriasis and persistent risk of nonmelanoma skin cancer. PUVA Follow-up Study. J Natl Cancer
41           Inst 90: 1278-1284.
42    Stern. RS; Zierler. S; Parrish. JA. (1980). Skin carcinoma in patients with psoriasis treated with topical tar
43           and artificial ultraviolet radiation. Lancet 1: 732-735.
44    Straif, K; Baan, R; Grosse, Y; Secretan, B; El Ghissassi, F; Cogliano, V. (2005). Carcinogenicity of polycyclic
45           aromatic hydrocarbons. Lancet Oncol 6: 931-932. http://dx.doi.org/10.1016/S1470-
46           2045(05)70458-7
47    Sugiyama, T. (1973). Chromosomal aberrations and carcinogenesis by various benz[a]anthracene
48           derivatives. Cancer Sci 64: 637-639.
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-25             DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1    Suter, M; Abramovici, A; Showalter, L; Hu, M; Shope, CD; Varner, M; Aagaard-Tillery, K. (2010). In utero
 2           tobacco exposure epigenetically modifies placental CYP1A1 expression. Metabolism 59: 1481-
 3           1490. http://dx.doi.0rg/10.1016/i.metabol.2010.01.013
 4    Swartz, WJ; Mattison, DR. (1985). Benzo(a)pyrene inhibits ovulation in C57BL/6N mice. Anat Rec 212:
 5           268-276. http://dx.doi.org/10.1002/ar.1092120308
 6    Szczeklik, A; Szczeklik, J; Galuszka, Z; Musial, J; Kolarzyk,  E; Targosz, D. (1994). Humoral
 7           immunosuppression in men exposed to polycyclic aromatic hydrocarbons and related
 8           carcinogens in polluted environments. Environ Health Perspect 102: 302-304.
 9    Szymariska, K; Matos, E; Hung, RJ; Wunsch-Filho, V; Eluf-Neto, J; Menezes, A; Daudt, AW; Brennan, P;
10           Boffetta, P. (2010). Drinking of mate and the risk of cancers of the upper aerodigestive tract in
11           Latin America: a case-control study. Cancer Causes Control 21: 1799-1806.
12           http://dx.doi.org/10.1007/sl0552-010-9606-6
13    Takizawa, K; Yagi, H; Jerina, DM;  Mattison, DR. (1984). Murine strain differences in ovotoxicity following
14           intraovarian injection with benzo(a)pyrene, (+)-(7R,8S)-oxide, (-)-(7R,8R)-dihydrodiol, or (+)-
15           (7R,8S)-diol-(9S,10R)-epoxide-2. Cancer Res 44: 2571-2576.
16    Talaska, G; Ginsburg, D; Ladow, K; Puga, A; Dalton, T; Warshawsky, D. (2006). Impact of Cypla2 or Ahr
17           gene knockout in mice: implications for biomonitoring studies. Toxicol Lett 162: 246-249.
18           http://dx.doi.0rg/10.1016/i.toxlet.2005.09.020
19    Talaska. G; Jaeger. M; Reilman. R; Collins. T; Warshawsky. D.  (1996). Chronic, topical exposure to
20           benzo[a]pyrene induces  relatively high steady-state levels of DNA adducts in target tissues and
21           alters kinetics of adduct loss. PNAS 93: 7789-7793.
22    Tamaki. A; Havashi. H; Nakajima. H; Takii. T; Katagiri. D; Miyazawa. K;  Hirose. K; Onozaki. K. (2004).
23           Polycyclic aromatic hydrocarbon increases mRNA level for interleukin 1 beta in human
24           fibroblast-like synoviocyte line via aryl hydrocarbon receptor. Biol Pharm Bull 27: 407-410.
25    Tang, D; Li, TY; Liu, JJ; Chen, YH; Qu, L; Perera,  F. (2006).  PAH-DNA adducts in cord blood and fetal and
26           child development in a Chinese cohort. Environ Health Perspect 114: 1297-1300.
27           http://dx.doi.org/10.1289/ehp.8939
28    Tang, D; Warburton, D; Tannenbaum, SR; Skipper, P; Santella, RM; Cereijido, GS; Crawford, FG; Perera,
29           FP. (1999). Molecular and genetic damage from  environmental tobacco smoke in young
30           children. Cancer Epidemiol Biomarkers Prev 8: 427-431.
31    Tang. PL; Li. TY; Liu. JJ; Zhou. ZJ; Yuan. T; Chen. YH; Rauh. VA; Xie. J; Perera. F. (2008). Effects of prenatal
32           exposure to coal-burning pollutants on children's development in China. Environ Health
33           Perspect 116: 674-679. http://dx.doi.org/10.1289/ehp.10471
34    Tang, Cf Xia, Y; Cheng, S; Tu, B. (2011). Modulation of behavior and glutamate receptor mRNA
35           expression in  rats after sub-chronic administration of benzo(a)pyrene. Biomed Environ Sci 24:
36           408-414. http://dx.doi.org/10.3967/0895-3988.2011.04.012
37    Temple, L; Kawabata, TT; Munson, AE; White, KL. (1993). Comparison of ELISA and Plaque-Forming Cell
38           Assays for Measuring the Humoral Immune Response to SRBC in  Rats and Mice Treated with
39           Benzo[a]pyrene or Cyclophosphamide. Toxicol Sci 21: 412-419.
40           http://dx.doi.0rg/10.1093/toxsci/21.4.412
41    Thirman. MJ; Albrecht. JH; Krueger. MA; Erickson. RR; Cherwitz. PL; Park. SS; Gelboin. HV;  Holtzman. JL.
42           (1994). Induction of cytochrome CYPIA1 and formation of toxic metabolites of benzo[a]pyrene
43           by rat aorta: a possible role in atherogenesis. PNAS 91: 5397-5401.
44    Thyssen, J; Althoff, J; Kimmerle, G; Mohr, U. (1981). Inhalation studies with benzo[a]pyrene in Syrian
45           golden hamsters. J Natl Cancer Inst 66: 575-577.
46    Topping,  DC; Martin, DH; Nettesheim, P. (1981). Determination of cocarcinogenic activity of
47           benzo[e]pyrene for respiratory tract mucosa. Cancer Lett 11: 315-321.
48           http://dx.doi.org/10.1016/0304-3835(81)90097-5
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-26            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Torinuki, W; Tagami, H. (1988). Incidence of skin cancer in Japanese psoriatic patients treated with
 2           either methoxsalen phototherapy, Goeckerman regimen, or both therapies: A 10-year follow-up
 3           study. J Am Acad Dermatol 18: 1278-1281.
 4    Tornqvist, S; Norell, S; Ahlbom, A; Knave, B. (1986). Cancer in the electric power industry. Br J Ind Med
 5           43:  212-213.
 6    Triolo, AJ; Aponte, GE; Herr, PL (1977). Induction of aryl hydrocarbon hydroxylase and forestomach
 7           tumors ben benzo(a)pyrene. Cancer Res 37: 3018-3021.
 8    U.S. Congress. (2011). Consolidated Appropriations Act, 2012. (Pub. L No. 112-74; 125 STAT. 786). 112th
 9           U.S. Congress. http://www.gpo.gov/fdsys/pkg/PLAW-112publ74/pdf/PLAW-112publ74.pdf
10    U.S. EPA (U.S. Environmental Protection Agency). (1986a). Guidelines for mutagenicity risk assessment.
11           (EPA/630/R-98/003). Washington, DC: U.S.  Environmental Protection Agency, Risk Assessment
12           Forum, http://www.epa.gov/iris/backgrd.html
13    U.S. EPA (U.S. Environmental Protection Agency). (1986b). Guidelines for the health risk assessment of
14           chemical mixtures. Fed Reg 51: 34014-34025.
15    U.S. EPA (U.S. Environmental Protection Agency). (1986c). Guidelines for the health risk assessment of
16           chemical mixtures. (EPA/630/R-98/002). Washington, DC: U.S. Environmental Protection
17           Agency, Risk Assessment Forum. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=22567
18    U.S. EPA (U.S. Environmental Protection Agency). (1988). Recommendations for and documentation of
19           biological values for use in risk assessment. (EPA/600/6-87/008). Cincinnati, OH: U.S.
20           Environmental Protection Agency, National Center for Environmental Assessment.
21           http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=34855
22    U.S. EPA (U.S. Environmental Protection Agency). (1990). Development of a dose-response model for
23           inhaled B(a)P [EPA Report]. (EPA-600/R-92-135).
24           http://www.ntis.gov/search/product.aspx?ABBR=PB93161016
25    U.S. EPA (U.S. Environmental Protection Agency). (1991a). Dose-response analysis of ingested
26           Benzo(a)pyrene (CAS No. 50-32-8) [EPA Report]. (EPA600R92045).
27           http://www.ntis.gov/search/product.aspx?ABBR=PB93167484
28    U.S. EPA (U.S. Environmental Protection Agency). (1991b). Drinking water criteria document for
29           polycyclic aromatic hydrocarbons (PAHs). (ECAOCIND010).
30           http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=9100CSXZ.txt
31    U.S. EPA (U.S. Environmental Protection Agency). (1991c). Guidelines for developmental toxicity risk
32           assessment. (EPA/600/FR-91/001). Washington, DC: U.S. Environmental Protection Agency, Risk
33           Assessment Forum, http://www.epa.gov/raf/publications/guidelines-dev-toxicity-risk-
34           assessment.htm
35    U.S. EPA (U.S. Environmental Protection Agency). (1992). Draft report: A cross-species scaling factor for
36           carcinogen risk assessment based on equivalence of mg/kg(3/4)/day. Fed Reg 57: 24151-24173.
37    U.S. EPA (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation reference
38           concentrations and application of inhalation dosimetry.  (EPA/600/8-90/066F). Research Triangle
39           Park, NC: U.S. Environmental Protection Agency,  Environmental Criteria and Assessment Office.
40           http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993
41    U.S. EPA (U.S. Environmental Protection Agency). (1996). Guidelines for reproductive toxicity risk
42           assessment. (EPA/630/R-96/009). Washington, DC:  U.S.  Environmental Protection Agency, Risk
43           Assessment Forum. http://www.epa.gov/raf/publications/pdfs/REPRO51.PDF
44    U.S. EPA (U.S. Environmental Protection Agency). (1998). Guidelines for neurotoxicity risk assessment.
45           (EPA/630/R-95/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
46           Forum. http://www.epa.gov/raf/publications/pdfs/NEUROTOX.PDF
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-27             DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    U.S. EPA (U.S. Environmental Protection Agency). (2000a). Science policy council handbook: risk
 2           characterization. (EPA/100/B-00/002). Washington, D.C.: U.S. Environmental Protection Agency,
 3           Office of Science Policy, http://www.epa.gov/osa/spc/pdfs/rchandbk.pdf
 4    U.S. EPA (U.S. Environmental Protection Agency). (2000b). Supplementary guidance for conducting
 5           health risk assessment of chemical mixtures. (EPA/630/R-00/002). Washington, DC: U.S.
 6           Environmental Protection Agency, Risk Assessment Forum.
 7           http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=20533
 8    U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and reference
 9           concentration processes. (EPA/630/P-02/002F). Washington, DC: U.S. Environmental Protection
10           Agency, Risk Assessment Forum. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=51717
11    U.S. EPA (U.S. Environmental Protection Agency). (2004). Risk Assessment Guidance for Superfund
12           (RAGS), Volume I: Human health evaluation manual, (part E: Supplemental guidance for dermal
13           risk assessment): Final. (EPA/540/R/99/005). Washington, DC.
14           http://www.epa.gov/oswer/riskassessment/ragse/index.htm
15    U.S. EPA (U.S. Environmental Protection Agency). (2005a). Guidelines for carcinogen risk assessment.
16           (EPA/630/P-03/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
17           Forum, http://www.epa.gov/cancerguidelines/
18    U.S. EPA (U.S. Environmental Protection Agency). (2005b). Supplemental guidance for assessing
19           susceptibility from early-life exposure to carcinogens. In US Environmental Protection Agency,
20           Risk Assessment Forum (pp. 1125-1133). (EPA/630/R-03/003F). Washington, DC: U.S.
21           Environmental Protection Agency, Risk Assessment Forum.
22           http://www.epa.gov/cancerguidelines/guidelines-carcinogen-supplement.htm
23    U.S. EPA (U.S. Environmental Protection Agency). (2006a). Approaches for the application of
24           physiologically based pharmacokinetic (PBPK) models and supporting data  in risk assessment
25           (Final Report) [EPA Report]. (EPA/600/R-05/043F). Washington, DC: U.S. Environmental
26           Protection Agency, National Center for Environmental assessment.
27           http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=157668
28    U.S. EPA (U.S. Environmental Protection Agency). (2006b). A framework for assessing health risk of
29           environmental exposures to children. (EPA/600/R-05/093F). Washington, DC: U.S.
30           Environmental Protection Agency, National Center for Environmental Assessment.
31           http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=158363
32    U.S. EPA (U.S. Environmental Protection Agency). (2010a). Consumer factsheet on benzo(a)pyrene:
33           Ground water and drinking water [Fact Sheet].
34           http://water.epa.gov/drink/contaminants/basicinformation/historical/upload/Archived-
35           Consumer-Factsheet-on-Benzo.pdf
36    U.S. EPA (U.S. Environmental Protection Agency). (2010b). Integrated science assessment for carbon
37           monoxide [EPA Report]. (EPA/600/R-09/019F). Research Triangle Park, NC.
38           http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686
39    U.S. EPA (U.S. Environmental Protection Agency). (2010c). Revised draft: Multistage Weibull time-to-
40           tumor model in EPA's benchmark dose software (BMDS): Methodology description. Washington,
41           DC.
42    U.S. EPA (U.S. Environmental Protection Agency). (2011). Recommended use of body weight 3/4 as the
43           default method in derivation of the oral reference dose. (EPA/100/R11/0001). Washington, DC:
44           U.S. Environmental Protection Agency, Risk Assessment Forum.
45           http://www.epa.gov/raf/publications/interspecies-extrapolation.htm
46    U.S. EPA (U.S. Environmental Protection Agency). (2012a). Advances in inhalation gas dosimetry for
47           derivation of a reference concentration (RFC) and use in risk assessment. (EPA/600/R-12/044).
48           Washington, DC. http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=244650
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-28            DRAFT—DO NOT CITE OR QUOTE

-------
                                                           Toxicological Review ofBenzo[a]pyrene


 1    U.S. EPA (U.S. Environmental Protection Agency). (2012b). Benchmark Dose Software (BMDS). Version
 2           2.2. Retrieved from http://www.epa.gov/NCEA/bmds/index.html
 3    U.S. EPA (U.S. Environmental Protection Agency). (2012c). Benchmark dose technical guidance.
 4           (EPA/100/R-12/001). Washington, DC: Risk Assessment Forum.
 5           http://www.epa.gov/raf/publications/pdfs/benchmark dose  guidance.pdf
 6    U.S. EPA (U.S. Environmental Protection Agency). (2012d). EPA announces NAS' review of IRIS
 7           Assessment development process. Available online at
 8           http://vosemite.epa.gOV/opa/admpress.nsf/0/lce2a7875daf093485257aOOOQ54df547OpenDoc
 9           ument
10    U.S. EPA (U.S. Environmental Protection Agency). (2014). EPAs Integrated Risk Information System:
11           Assessment development process [EPA Report]. Washington, DC.
12           http://epa.gov/iris/process.htm
13    Urano, K; Katakai, Y; Tokuda, Y; Ueyama, Y; Nomura, T; Yamamoto, S. (1995). Failure of genotoxic
14           carcinogens to produce tumors in human skin xenografts transplanted to SCID mice.
15           Carcinogenesis 16: 2223-2226. http://dx.doi.Org/10.1093/carcin/16.9.2223
16    Urso, P; Gengozian, N. (1980). Depressed humoral immunity and increased tumor incidence in mice
17           following in utero exposure to benzo[alpha]pyrene. J Toxicol Environ Health 6: 569-576.
18           http://dx.doi.org/10.1080/15287398009529874
19    Urso. P; Gengozian. N. (1982). Alterations in the humoral immune response and tumor frequencies in
20           mice exposed to benzo[a]pyrene and X-rays before or  after birth. J Toxicol Environ Health 10:
21           817-835. http://dx.doi.org/10.1080/15287398209530297
22    Urso. P; Gengozian. N. (1984). Subnormal expression of cell-mediated  and humoral immune responses
23           in progeny disposed toward a high incidence of tumors after in utero exposure to
24           benzo[a]pyrene. J Toxicol Environ Health 14:  569-584.
25           http://dx.doi.org/10.1080/15287398409530606
26    Urso, P; Johnson, RA. (1987). Early changes in T lymphocytes and subsets of mouse progeny defective as
27           adults in controlling growth of a syngeneic tumor after in utero insult with benzo(a)pyrene.
28           Immunopharmacology 14: 1-10.
29    Urso, P; Johnson, RA. (1988). Quantitative and functional change in T cells of primiparous mice following
30           injection of benzo(a)pyrene at the second trimester of pregnancy. Immunopharmacol
31           Immunotoxicol 10: 195-217. http://dx.doi.org/10.3109/08923978809014333
32    Urso, P; Kramer, MK. (2008). The role of adherent cells in the immunosuppressed state of mouse
33           progeny transplacentally exposed to benzo(alpha)pyrene. In Vitro Cell Dev Biol Anim 44: 273-
34           279. http://dx.doi.org/10.1007/sll626-008-9112-2
35    Urso, P; Ryan, MC; Bennett, JS. (1988). Changes in peripheral blood cells  in mice after injection with
36           benzo-a-pyrene during pregnancy. Immunopharmacol Immunotoxicol 10: 179-194.
37           http://dx.doi.org/10.3109/08923978809014332
38    Urso. P; Zhang. W; Cobb. JR. (1992). Immunological consequences from exposure to benzo(a)pyrene
39           during pregnancy. Scand J Immunol  Suppl 11: 203-206.
40    van den Berg. FA; Baken. KA; Vermeulen. JP; Gremmer. ER; van Steeg.  H; van  Loveren. H. (2005). Use of
41           the local lymph node assay in assessment of immune function. Toxicology 211: 107-114.
42           http://dx.doi.0rg/10.1016/i.tox.2005.03.003
43    van Lipzig, MM; Vermeulen, IMP; Gusinu, R; Legler, J; Frank, H; Seidel, A; Meerman, JH. (2005). Formation
44           of estrogenic metabolites of benzo[a]pyrene and chrysene by cytochrome P450 activity and
45           their combined and supra-maximal estrogenic activity. Environ Toxicol Pharmacol 19: 41-55.
46           http://dx.doi.0rg/10.1016/i.etap.2004.03.010
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-29            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Van Metre, PC; Mahler, BJ. (2010). Contribution of PAHs from coal-tar pavement sealcoat and other
 2           sources to 40 U.S. lakes. Sci Total Environ 409: 334-344.
 3           http://dx.doi.0rg/10.1016/i.scitotenv.2010.08.014
 4    Varghese, SK; Gangamma, S. (2009). Particle deposition in human respiratory system: deposition of
 5           concentrated hygroscopic aerosols. Inhal Toxicol 21: 619-630.
 6           http://dx.doi.org/10.1080/08958370802380792
 7    Vesselinovitch, SD; Kyriazis, AP; Mihailovich, N; Rao, KV. (1975). Conditions modifying development of
 8           tumors in mice at various sites by benzo(a)pyrene. Cancer Res 35: 2948-2953.
 9    Vesselinovitch, SD; Rao, KV; Mihailovich, N. (1979). Neoplastic response of mouse tissues during
10           perinatal age periods and its significance in chemical carcinogenesis [Review].  Natl Cane Inst
11           Monogr239-250.
12    Vineis, P; Anttila, S; Benhamou, S; Spinola, M; Hirvonen, A; Kiyohara, C; Garte, SJ; Puntoni, R; Rannug, A;
13           Strange, RC; Taioli, E. (2007). Evidence of gene gene interactions in lung carcinogenesis in a large
14           pooled analysis [Review]. Carcinogenesis 28: 1902-1905.
15           http://dx.doi.org/10.1093/carcin/bgm039
16    Vineis, P; Perera, F. (2007). Molecular epidemiology and biomarkers in etiologic cancer research: the
17           new in light of the old [Review]. Cancer Epidemiol Biomarkers Prev 16: 1954-1965.
18           http://dx.doi.org/10.1158/1055-9965.EPI-07-0457
19    Voelter-Mahlknecht. S; Scheriau. R; Zwahr. G; Koch. B; Escobar Pinzon. LC; Drexler. H; Letzel. S. (2007).
20           Skin tumors among employees of a tar refinery: the current data and their implications. Int Arch
21           Occup Environ Health 80: 485-495. http://dx.doi.org/10.1007/s00420-006-0155-v
22    Vondracek. J; Kozubfk. A; Machala. M. (2002). Modulation of estrogen receptor-dependent reporter
23           construct activation and GO/Gl-S-phase transition by polycyclic aromatic hydrocarbons in
24           human breast carcinoma MCF-7 cells. Toxicol Sci 70: 193-201.
25    Wang, F;  He, Y; Guo, H; Li, J; Yang, Y; Wu, Z; Zheng, H; Wu, T. (2010). Genetic variants of nucleotide
26           excision repair genes are associated with DNA damage in coke oven workers. Cancer Epidemiol
27           Biomarkers Prev  19: 211-218. http://dx.doi.org/10.1158/1055-9965.EPI-09-0270
28    Wattenberg, LW. (1974).  Inhibition of carcinogenic and toxic effects of polycyclic hydrocarbons by
29           several sulfur-containing compounds. J Natl Cancer Inst 52: 1583-1587.
30    Waylen, AL; Metwally, M; Jones, GL; Wilkinson, AJ; Ledger, WL. (2009). Effects of cigarette smoking upon
31           clinical outcomes of assisted reproduction: a meta-analysis [Review]. Hum Reprod Update 15:
32           31-44. http://dx.doi.org/10.1093/humupd/dmn046
33    Wei, Q; Cheng, L; Amos, Cl; Wang, LE; Guo, Z; Hong, WK; Spitz, MR. (2000). Repair of tobacco
34           carcinogen-induced DNA adducts and lung cancer risk: a molecular epidemiologic study. J Natl
35           Cancer Inst 92: 1764-1772.
36    Wei. SJ; Chang. RL; Merkler. KA; Gwynne. M; Cui. XX; Murthy. B;  Huang. MT; Xie. JG; Lu. YP; Lou. YR;
37           Jerina, DM; Conney, AH. (1999). Dose-dependent mutation profile in the c-Ha-ras proto-
38           oncogene  of skin tumors in mice initiated with benzo[a]pyrene. Carcinogenesis 20: 1689-1696.
39    Weinberg. CR; Wilcox. AJ; Baird. DP. (1989). Reduced fecundability in women with prenatal exposure to
40           cigarette smoking. Am J Epidemiol 129:  1072-1078.
41    Wester. PW; Kroes. R. (1988). Forestomach carcinogens: pathology and relevance to man [Review].
42           Toxicol Pathol 16: 165-171.
43    Wester, RC; Maibach, HI; Bucks, DA; Sedik, L; Melendres, J; Liao,  C; Dizio, S. (1990). Percutaneous
44           absorption of [14C]DDT and [14C]benzo[a]pyrene from soil. Fundam Appl Toxicol 15: 510-516.
45    Weyand,  EH; Chen, YC; Wu, Y; Koganti, A; Dunsford, HA; Rodriguez, LV. (1995). Differences in the
46           tumorigenic activity of a pure hydrocarbon and a complex mixture following ingestion:
47           Benzo[a]pyrene vs manufactured gas plant residue. Chem Res Toxicol  8: 949-954.
48           http://dx.doi.org/10.1021/tx00049a008
                 This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-30            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Weyand, EH; He, ZM; Ghodrati, F; Wu, Y; Marshall, MV; Lavoie, EJ. (1992). Effect of fluorine substitution
 2           on benzo[j]fluoranthene genotoxicity. Chem Biol Interact 84: 37-53.
 3    Weyand, EH; Lavoie, EJ. (1988). Comparison of PAH DNA adduct formation and tumor initiating activity
 4           in newborn mice. Proc Am Assoc Cancer Res 29: 98.
 5    White, KL; Lysy, HH; Holsapple, MP. (1985). Immunosuppression by polycyclic aromatic hydrocarbons: a
 6           structure-activity relationship in B6C3F1 and DBA/2 mice. Immunopharmacology 9: 155-164.
 7    WHO (World Health Organization). (2012). Guidance for immunotoxicity risk assessment for chemicals.
 8           (Harmonization Project Document No. 10). Geneva, Switzerland.
 9           http://www.inchem.org/documents/harmproi/harmproi/harmproilO.pdf
10    Whyatt, RM; Santella, RM; Jedrychowski, W; Garte, SJ; Bell, DA; Ottman, R; Gladek-Yarborough, A;
11           Cosma, G; Young, TL; Cooper, TB; Randall, MC; Manchester,  DK; Perera, FP. (1998). Relationship
12           between ambient air pollution and DNA damage in Polish mothers and newborns. Environ
13           Health Perspect 106: 821-826.
14    Wilson, JS; Holland, LM. (1988). Periodic response difference in mouse epidermis chronically exposed to
15           crude-oils or BaP: Males vs. females. Toxicology 50: 83-94. http://dx.doi.org/10.1016/0300-
16           483X(88)90123-0
17    Winker, N; Tuschl, H; Kovac, R; Weber, E. (1997). Immunological investigations in a group of workers
18           exposed to various levels of polycyclic aromatic hydrocarbons. J Appl Toxicol  17: 23-29.
19    Wislocki. PG: Bagan. ES: Lu. AYH:  Doolev. KL: Fu. PP: Han-Hsu. H: Beland. FA: Kadlubar. FF. (1986).
20           Tumorigenicity of nitrated derivatives of pyrene, benz[a]anthracene, chrysene, and
21           benzo[a]pyrene in the newborn mouse assay. Carcinogenesis 7: 1317-1322.
22           http://dx.doi.0rg/10.1093/carcin/7.8.1317
23    Wolff. RK: Griffith. WC: Henderson. RF: Hahn. FF: Harkema. J. R.: Rebar. AH: Eidson. AF: Mcclellan. RO.
24           (1989). Effects of repeated inhalation exposures to 1-nitropyrene, benzo(a)pyrene, Ga2O3
25           particles, and SO2 alone and in combinations on particle clearance, bronchoalveolar lavage fluid
26           composition, and histopathology. J Toxicol Environ Health 27: 123-138.
27           http://dx.doi.org/10.1080/15287398909531283
28    Wood, AW; Levin, W; Chang, RL; Huang, MT; Ryan, DE; Thomas, PE; Lehr, RE; Kumar,  S; Koreeda, M;
29           Akagi, H; Ittah, Y; Dansette, P; Yagi, H; Jerina, DM; Conney, AH. (1980). Mutagenicity and tumor-
30           initiating activity of cyclopenta(c,d)pyrene and structurally related compounds. Cancer Res 40:
31           642-649.
32    Wormley, DP; Chirwa, S; Nayyar,  T; Wu, J; Johnson, S; Brown, LA; Harris, E; Hood, DB. (2004). Inhaled
33           benzo(a)pyrene impairs long-term potentiation in the Fl generation rat dentate gyrus. Cell Mol
34           Biol (Noisy-le-grand) 50: 715-721.
35    Wu, J; Hou, H; Ritz, B; Chen, Y. (2010). Exposure to polycyclic aromatic hydrocarbons  and missed
36           abortion in early pregnancy in a Chinese population. Sci Total Environ 408: 2312-2318.
37           http://dx.doi.0rg/10.1016/i.scitotenv.2010.02.028
38    Wu. J; Ramesh. A; Navyar. T; Hood. DB. (2003a). Assessment of metabolites and AhR  and CYP1A1 mRNA
39           expression subsequent to prenatal  exposure to inhaled benzo(a)pyrene. Int J  Dev Neurosci 21:
40           333-346.
41    Wu. MT: Pan. CH: Huang. YL: Tsai. PJ: Chen. CJ: Wu. TN. (2003b). Urinary excretion of 8-hydroxy-2-
42           deoxyguanosine and 1-hydroxypyrene in coke-oven workers. Environ Mol Mutagen 42: 98-105.
43           http://dx.doi.org/10.1002/em.10176
44    Wynder, EL; Fritz, L; Furth, N. (1957). Effect of concentration of benzopyrene in skin carcinogenesis. J
45           Natl Cancer Inst 19: 361-370.
46    Wynder, EL; Hoffmann, D. (1959). A study of tobacco carcinogenesis. VII. The role of higher polycyclic
47           hydrocarbons. Cancer 12: 1079-1086. http://dx.doi.org/10.1002/1097-
48           0142(195911/12)12:6<1079::AID-CNCR2820120604>3.0.CO;2-I
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-31            DRAFT—DO NOT CITE OR QUOTE

-------
                                                            Toxicological Review ofBenzo[a]pyrene


 1    Xia. Y; Cheng. S; He. J; Liu. X; Tang. Y; Yuan. H; He. L; Lu. T; Tu. B; Wang. Y. (2011). Effects of subchronic
 2           exposure to benzo[a]pyrene (B[a]P) on learning and memory, and neurotransmitters in male
 3           Sprague-Dawley rat. Neurotoxicology 32: 188-198.
 4           http://dx.doi.0rg/10.1016/i.neuro.2010.12.015
 5    Xia. Y; Zhu. P; Han. Y; Lu. C; Wang. S; Gu. A; Fu. G; Zhao. R; Song. L; Wang. X. (2009). Urinary metabolites
 6           of polycyclic aromatic hydrocarbons in relation to idiopathic male infertility. Hum Reprod 24:
 7           1067-1074. http://dx.doi.org/10.1093/humrep/dep006
 8    Xu. C; Chen. J; Qiu. Z; Zhao. Q; Luo. J; Yang. L; Zeng. H; Huang. Y; Zhang. L; Cao. J; Shu. W. (2010).
 9           Ovotoxicity and PPAR-mediated aromatase downregulation in female Sprague-Dawley rats
10           following combined oral exposure to benzo[a]pyrene and di-(2-ethylhexyl)  phthalate. Toxicol
11           Lett 199: 323-332. http://dx.doi.Org/10.1016/i.toxlet.2010.09.015
12    Xu, D; Penning, TM; Blair, IA; Harvey, RG.  (2009). Synthesis of phenol and quinone metabolites of
13           benzo[a]pyrene, a carcinogenic component of tobacco smoke implicated in lung cancer. J Org
14           Chem 74: 597-604. http://dx.doi.org/10.1021/io801864m
15    Xu, GB; Yu, CP. (1985). Theoretical lung deposition of hygroscopic NaCI aerosols. Aerosol Sci Technol 4:
16           455-461. http://dx.doi.org/10.1080/02786828508959070
17    Xu, Z; Brown, LM; Pan, GW; Liu, TF; Gao, GS; Stone, BJ; Cao, RM; Guan, DX; Sheng, JH; Yan, ZS; Dosemeci,
18           M; Fraumeni, JF; Blot, WJ. (1996). Cancer risks among iron and steel workers in Anshan, China,
19           Part  II: Case-control studies of  lung and stomach cancer. Am J Ind Med 30: 7-15.
20           http://dx.doi.org/10.1002/(SICI)1097-0274(199607)30:l<:7::AID-AJIM2>:3.0.CO:2-#
21    Xue. W; Warshawsky. D. (2005). Metabolic activation of polycyclic and heterocyclic aromatic
22           hydrocarbons and DNA damage: a review [Review]. Toxicol Appl Pharmacol 206: 73-93.
23           http://dx.doi.0rg/10.1016/i.taap.2004.ll.006
24    Ye, X; Skjaerven, R; Basso, O; Baird, DP; Eggesbo, M; Cupul Uicab, LA; Haug, K; Longnecker,  MP. (2010).
25           In utero exposure to tobacco smoke and subsequent  reduced fertility in females. Hum Reprod
26           25: 2901-2906. http://dx.doi.org/10.1093/humrep/deq235
27    Young, C; Cherrie, J; Van Tongeren, M;  Fortunato, L; Hutchings, S; Rushton, L. (2012). The burden of
28           occupational cancer in Great Britain: Non-melanoma  skin cancer. Sudbury,  Suffolk:  Health and
29           Safety Executive (HSE). http://www.hse.gov.uk/research/rrhtm/rr928.htm
30    Yu, D; Berlin, JA; Penning, TM; Field, J. (2002). Reactive oxygen species generated by PAH o-quinones
31           cause change-in-function mutations in p53. Chem Res Toxicol 15: 832-842.
32    Zaccaria,  KJ; McClure, PR. (2013). Using immunotoxicity information to improve cancer risk assessment
33           for polycyclic aromatic hydrocarbon mixtures [Review]. Int J Toxicol 32: 236-250.
34           http://dx.doi.org/10.1177/1091581813492829
35    Zandi-Nejad, K; Luyckx, VA; Brenner, BM.  (2006). Adult hypertension and kidney disease - The role of
36           fetal programming [Review]. Hypertension 47: 502-508.
37           http://dx.doi.org/10.1161/01.HYP.0000198544.09909.la
38    Zenzes. MT; Puy. LA; Bielecki. R. (1998). Immunodetection of  benzo[a]pyrene adducts in ovarian cells of
39           women exposed to cigarette smoke. Mol Hum  Reprod 4: 159-165.
40           http://dx.doi.0rg/10.1093/molehr/4.2.159
41    Zenzes. MT; Puv. LA; Bielecki. R; Reed. TE. (1999). Detection of benzo[a]pyrene diol epoxide-DNA
42           adducts in embryos from smoking couples:  evidence for transmission  by spermatozoa. Mol Hum
43           Reprod 5: 125-131.
44    Zhang, HM; Nie, JS; Li, X; Niu, Q. (2012). Characteristic analysis of peripheral blood mononuclear cell
45           apoptosis in coke oven workers. J Occup Health 54: 44-50. http://dx.doi.org/10.1539/ioh.ll-
46           0155-OA
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                  R-32             DRAFT—DO NOT CITE OR QUOTE

-------
                                                             Toxicological Review ofBenzo[a]pyrene


 1    Zhang, Y; Ramos, KS. (1997). The induction of proliferative vascular smooth muscle cell phenotypes by
 2           benzo[a]pyrene does not involve mutational activation of ras genes. Mutat Res 373: 285-292.
 3           http://dx.doi.org/10.1016/S0027-5107(96)00213-8
 4    Zhang, YJ; Li, Y; Deleo, VA; Santella, RM. (1990). Detection of DNA adducts in skin biopsies of coal tar-
 5           treated psoriasis patients: immunofluorescence and 32P postlabeling. Skin Pharmacol Physiol 3:
 6           171-179. http://dx.doi.org/10.1159/000210867
 7    Zheng, SJ; Tian, HJ; Cao, J; Gao, YQ. (2010). Exposure to di(n-butyl)phthalate and benzo(a)pyrene alters
 8           IL-1(3 secretion and subset expression of testicular macrophages, resulting in decreased
 9           testosterone production in rats. Toxicol Appl Pharmacol 248: 28-37.
10           http://dx.doi.0rg/10.1016/i.taap.2010.07.008
11
                This document is a draft for review purposes only and does not constitute Agency policy.
                                                   R-33             DRAFT—DO NOT CITE OR QUOTE

-------