DRAFT - DO NOT CITE OR QUOTE
EPA/635/R-07/002
www.epa.gov/iris
vvEPA
TOXICOLOGICAL REVIEW
OF
BROMOBENZENE
(CAS No. 108-86-1)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
June 2007
NOTICE
This document is an interagency review draft. It has not been formally released by the U.S.
Environmental Protection Agency and should not at this stage be construed to represent Agency
position on this chemical. It is being circulated for review of its technical accuracy and science
policy implications.
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document is a preliminary draft for review purposes only and does not constitute
U.S. Environmental Protection Agency policy. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
6/7/07 ii DRAFT - DO NOT CITE OR QUOTE
-------
CONTENTS-TOXICOLOGICAL REVIEW OF BROMOBENZENE (CAS No. 108-86-1)
LIST OF FIGURES vi
LIST OF TABLES vii
LIST OF ABBREVIATIONS AND ACRONYMS x
FOREWORD xi
AUTHORS, CONTRIBUTORS, AND REVIEWERS xii
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION RELEVANT TO ASSESSMENTS 3
3. TOXICOKINETICS RELEVANT TO ASSESSMENTS 6
3.1. ABSORPTION 6
3.2. DISTRIBUTION 6
3.3. METABOLISM 7
3.4. ELIMINATION 11
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS (PBTK) 12
4. HAZARD IDENTIFICATION 13
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL
CONTROLS 13
4.2. LESS-THAN-LIFETIME AND CHRONIC STUDIES AND CANCER
BIO ASSAYS IN ANIMALS—ORAL AND INHALATION 13
4.2.1. Oral Exposure 13
4.2.1.1. Subchronic Toxicity 13
4.2.1.2. Chronic Toxicity 23
4.2.2. Inhalation Exposure 24
4.2.2.1. Subchronic Toxicity 24
4.2.2.2. Chronic Toxicity 31
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND
INHALATION 31
4.3.1. Reproductive Toxicity Studies 31
4.3.2. Developmental Toxicity Studies 31
4.4. OTHER STUDIES 31
4.4.1. Acute Toxicity Studies 31
4.4.2. Genotoxicity Studies 32
4.4.3. Tumor Promotion Studies 35
4.5. MECHANISTIC STUDIES 35
4.5.1. Mechanistic Studies of Liver Effects 35
4.5.2. Mechanistic Studies of Kidney Effects 41
4.5.3. Genomic/Proteomic Responses of the Liver to Bromobenzene 42
4.5.4. Similarities Between Bromobenzene and Chlorobenzene 43
4.6. SYNTHESIS AND EVALUATION OF MAJOR NONCANCER EFFECTS 48
4.6.1. Oral 48
4.6.2. Inhalation 48
6/7/07 iii DRAFT - DO NOT CITE OR QUOTE
-------
4.6.3. Mode of Action Information 49
4.7. EVALUATION OF CARCINOGENICITY 51
4.8. SUSCEPTIBLE POPULATIONS 51
4.8.1. Possible Childhood Susceptibility 51
4.8.2. Possible Gender Differences 51
4.8.3. Other 52
5. DOSE-RESPONSE ASSESSMENTS 53
5.1. ORAL REFERENCE DOSE 53
5.1.1. Subchronic Oral RfD 53
5.1.1.1. Choice of Principal Study and Critical Effect - with Rationale and
Justification 53
5.1.1.2. Methods of Analysis - Including Models (PBPK, HMD, etc.) 54
5.1.1.3. Subchronic RfD Derivation - Including Application of
Uncertainty Factors (UFs) 59
5.1.2. Chronic Oral RfD 60
5.1.2.1. Choice of Principal Study and Critical Effect - with Rationale and
Justification 60
5.1.2.2. Methods of Analysis - Including Models (PBPK, BMD, etc.) 60
5.1.2.3. Chronic RfD Derivation - Including Application of Uncertainty
Factors (UFs) 60
5.1.3. Previous Oral Assessment 62
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 62
5.2.1. Subchronic Inhalation RfC 62
5.2.1.1. Choice of Principal Study and Critical Effect - with Rationale
and Justification 62
5.2.1.2. Methods of Analysis-Including Model (PBPK, BMD, etc.) 64
5.2.1.3. Subchronic RfC Derivation - Including Application of
Uncertainty Factors (UFs) 68
5.2.2. Chronic Inhalation RfC 69
5.2.2.1. Choice of Principal Study and Critical Effect - with Rationale
and Justification 69
5.2.2.2. Methods of Analysis - Including Model (PBPK, BMD, etc.) 70
5.2.2.3. Chronic RfC Derivation - Including Application of Uncertainty
Factors (UFs) 70
5.3. CANCER ASSESSMENT 71
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND
DOSE RESPONSE 72
6.1. HUMAN HAZARD POTENTIAL 72
6.2. DOSE RESPONSE 73
6.2.1. Noncancer/Oral 73
6.2.2. Noncancer/Inhalation 73
6.2.3. Cancer/Oral 74
6.2.4. Cancer/Inhalation 74
6/7/07 iv DRAFT - DO NOT CITE OR QUOTE
-------
7. REFERENCES 75
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION A-l
APPENDIX B: BENCHMARK DOSE CALCULATIONS FOR THE RfD B-l
APPENDIX C. BENCHMARK DOSE CALCULATIONS FOR THE RfC C-l
6/7/07 v DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
2-1 Chemical structure of bromobenzene 3
3-1 Proposed metabolic scheme for bromobenzene in mammals 8
4-1 Chemical structure of bromobenzene and chlorobenzene 43
6/7/07 vi DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF TABLES
4-1 Effects of bromobenzene on terminal body and liver weights and serum liver
enzymes of male and female Fischer 344/N rats exposed by oral gavage 5
days/week for 90 days in the basic study (mean +/- standard deviation) 15
4-2 Incidences of male and female Fischer 344/N rats with liver and kidney lesions
following administration of bromobenzene by gavage 5 days/week for 90 days
in the basic study 17
4-3 Effects of bromobenzene on terminal body and liver weights and levels of
selected serum liver enzymes of male and female B6C3F1 mice exposed by oral
gavage 5 days/week for 90 days in the basic study (mean +/- standard deviation) 21
4-4 Incidences of male and female B6C3F1 mice with liver and kidney lesions
following administration of bromobenzene by gavage 5 days/week for 90 days
in the basic study 22
4-5 Effects of bromobenzene on terminal body, liver, and kidney weights of male
and female rats exposed by inhalation 6 hours/day, 5 days/week for 13 weeks
(mean+/- standard deviation) 25
4-6 Incidences of male and female Fischer 344/N rats with liver and kidney lesions
following repeated exposure to bromobenzene vapors for 13 weeks 26
4-7 Effects of bromobenzene on terminal body, liver, and kidney weights of male
and female mice exposed by inhalation 6 hours/day, 5 days/week for 13 weeks
(mean+/- standard deviation) 29
4-8 Incidences of male and female B6C3F1 mice with liver and kidney lesions
following repeated exposure to bromobenzene vapors for 13 weeks 30
4-9 Results of bromobenzene genotoxicity testing 33
4-10 The effect of CYP inhibition on the hepatotoxicity and metabolism of single
intraperitoneal doses of bromobenzene 36
4-11 The influence of various treatments on the metabolism of bromobenzene (BB)
and severity of bromobenzene-induced hepatic necrosis in rats administered
single intraperitoneal doses of bromobenzene 39
4-12 Incidences of male and female Fischer 344/N rats with liver and kidney lesions
following administration of chlorobenzene by gavage 5 days/week for 13 weeks 46
6/7/07 vii DRAFT - DO NOT CITE OR QUOTE
-------
4-13 Incidences of male and female B6C3F1 mice with liver and kidney lesions
following administration of chlorobenzene by gavage 5 days/week for 13 weeks 46
5-1 Incidences of male and female Fischer 344/N rats and B6C3F1 mice with liver
lesions following administration of bromobenzene by gavage 5 days/week for
90 Days 55
5-2 Benchmark doses (BMDioS and BMDLios) from best fitting models predicting
combined incidences of Fischer 344/N rats or B6C3F1 mice with liver lesions 55
5-3 Weibull model-estimated BMDs and BMDLs (mg/kg-day) associated with
10, 5, and 1% extra risk for liver lesions in female B6C3F1 mice administered
bromobenzene by oral gavage 5 days/week for 90 days 55
5-4 Data for absolute liver weight and liver-to-body weight ratio for male and
female Fischer 344/N rats and male and female B6C3F1 mice following
administration of bromobenzene by gavage 5 days/week for 90 days (mean +/-
standard deviation) 56
5-5 Benchmark doses (BMDio and BMDLio) from best fitting models for increased
absolute liver weight and liver-to-body weight ratio in Fischer 344/N rats and
B6C3F1 mice administered bromobenzene by gavage 5 days/week for 90 days 57
5-6 The third-degree polynomial model-estimated BMDs and BMDLs (mg/kg-day)
associated with 1 and 0.5 standard deviation extra risk for increased absolute liver
weight in female B6C3F1 mice administered bromobenzene by oral gavage 5
days/week for 90 days 57
5-7 Data for SDH for male and female B6C3F1 mice following administration of
bromobenzene by gavage 5 days/week for 90 days (mean +/- standard deviation) 58
5-8 The power model estimated BMD and BMDLs associated with 10% extra risk
for increase SDH serum levels in B6C3F1 female mise exposed to bromobenzene
by gavage 5 days/week for 90 days 58
5-9 Incidences of female B6C3F1 mice with cytomegaly in the centrilobular region
of the liver following inhalation exposure to bromobenzene vapors for 6 hours/day,
5 days/week for 13 weeks 64
5-10 BMC modeling results for the incidence of liver cytomegaly in female B6C3F1
mice exposed to bromobenzene vapors 6 hours/day, 5 days/week for 13 weeks 66
5-11 BMCs and BMCLs predicted from the log-logistic and gamma models for 10, 5,
and 1% extra risk for hepatocellular cytomegaly in female B6C3F1 mice exposed
to bromobenzene vapors for 6 hours/day, 5 days/week for 13 weeks 66
5-12 Data for absolute liver weight and liver-to-body weight ratio for male and
female B6C3F1 mice following inhalation exposure to bromobenzene vapors
for 6 hours/day, 5 days/week for 13 weeks (mean +/- standard deviation) 67
6/7/07 viii DRAFT - DO NOT CITE OR QUOTE
-------
5-13 Model output for increased absolute liver weight and liver-to-body weight ratio
in female B6C3F1 mice following inhalation exposure to bromobenzene for
6 hours/day, 5 days/week for 13 weeks 67
5-14 The second-degree polynomial model-estimated BMCs and BMCLs associated
with 1 and 0.5 standard deviation extra risk for increased absolute liver weight
and liver-to-body weight ratio in female B6C3F1 mice exposed to bromobenzene
vapors for 6 hours/day, 5 days/week for 90 days 68
6/7/07 ix DRAFT - DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
AIC Akaike's Information Criteria
ALT Alanine aminotransferase
AST Aspartate aminotransferase
BB Bromobenzene
BCF Bioconcentration factor
BMC Benchmark concentration
BMD Benchmark dose
BMDS Benchmark Dose Software
BMR Benchmark response
BUN Blood urea nitrogen
CASRN Chemical Abstract Service Registry Number
DENA Diethylnitrosamine
EH Epoxide hydrolase
EPA Environmental Protection Agency
GC-MS Gas chromatography-mass spectrometry
GGT y-Glutamyltranspeptidase-positive
H&E Hematoxylin and eosin
HEC Human equivalent concentration
IRIS Integrated Risk Information System
LOAEL Lowest-observed-adverse-effect level
MCH Mean corpuscular hemoglobin
MCHC Mean corpuscular hemoglobin content
MCV Mean corpuscular volume
NOAEL No-observed-adverse-effect level
NTP National Toxicology Program
PAS Periodic acid-Schiff
PBPK Physiologially based pharmacokinetic
PBTK Physiologically based toxicokinetic
RfC Inhalation reference concentration
RfD Oral reference dose
SDH Sorbitol dehydrogenase
UF Uncertainty factor
VHC Volatile hydrocarbon
6/7/07
DRAFT - DO NOT CITE OR QUOTE
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to exposure to bromobenzene. It
is not intended to be a comprehensive treatise on the chemical or toxicological nature of
bromobenzene.
In Section 6, Major Conclusions in the Characterization of Hazard and Dose Response,
EPA has characterized its overall confidence in the quantitative and qualitative aspects of hazard
and dose response by addressing knowledge gaps, uncertainties, quality of data, and scientific
controversies. The discussion is intended to convey the limitations of the assessment and to aid
and guide the risk assessor in the ensuing steps of the risk assessment process.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address).
6/7/07 xi DRAFT - DO NOT CITE OR QUOTE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER
Carolyn L. Smallwood, Environmental Scientist
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Cincinnati, OH
AUTHORS
Carolyn L. Smallwood, Environmental Scientist
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Cincinnati, OH
David W. Wohlers, Ph.D.
Peter R. McClure, Ph.D., DABT
Daniel J. Plewak, B.S.
Environmental Science Center
Syracuse Research Corporation
Syracuse, NY
Under Contract No. 68-C-00-122 with Battelle Memorial Institute and EPA Contract No.
68-C-03-147 to Syracuse Research Corporation
REVIEWERS
This document and the accompanying IRIS Summary have been peer reviewed by EPA
scientists and independent scientists external to EPA. Comments from all peer reviewers were
evaluated carefully and considered by the Agency during the finalization of this assessment.
During the finalization process, the IRIS Program Director achieved common understanding of
the assessment among the Office of Research and Development; Office of Air and Radiation;
Office of Prevention, Pesticides, and Toxic Substances; Office of Solid Waste and Emergency
Response; Office of Water; Office of Policy, Economics, and Innovation; Office of Children's
Health Protection; Office of Environmental Information; and EPA's regional offices.
INTERNAL EPA REVIEWERS
Stiven Foster
National Center for Environmental Assessment
Washington, DC
6/7/07 xii DRAFT - DO NOT CITE OR QUOTE
-------
John Fox
National Center for Environmental Assessment
Washington, DC
Martin Gehlhaus
National Center for Environmental Assessment
Washington, DC
Steven S. Kueberuwa
Health and Ecological Criteria Division
Office of Science and Technology, Office of Water
Washington, DC
Allan Marcus
National Center for Environmental Assessment
Research Triangle Park, NC
Chandrika Moudgal
National Center for Environmental Assessment
Cincinnati, OH
Sederick Rice
National Center for Environmental Assessment
Washington, DC
Jeff Swartout
National Center for Environmental Assessment
Cincinnati, OH
6/7/07 xiii DRAFT - DO NOT CITE OR QUOTE
-------
1 1. INTRODUCTION
2
3
4 This document presents background information and justification for the Integrated Risk
5 Information System (IRIS) Summary of the hazard and dose-response assessment of
6 bromobenzene. IRIS Summaries may include oral reference dose (RfD) and inhalation reference
7 concentration (RfC) values for chronic and less-than-lifetime exposure durations, and a
8 carcinogenicity assessment.
9 The RfD and RfC provide quantitative information for use in risk assessments for health
10 effects known or assumed to be produced through a nonlinear (possibly threshold) mode of
11 action. The RfD (expressed in units of mg/kg-day) is defined as an estimate (with uncertainty
12 spanning perhaps an order of magnitude) of a daily exposure to the human population (including
13 sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a
14 lifetime. The inhalation RfC (expressed in units of mg/m3) is analogous to the oral RfD, but
15 provides a continuous inhalation exposure estimate. The inhalation RfC considers toxic effects
16 for both the respiratory system (portal-of-entry) and for effects peripheral to the respiratory
17 system (extrarespiratory or systemic effects). Reference values may also be derived for acute
18 (=24 hours), short-term (up to 30 days), and subchronic (up to 10% of average lifetime) exposure
19 durations, all of which are derived based on an assumption of continuous exposure throughout
20 the duration specified.
21 The carcinogenicity assessment provides information on the carcinogenic hazard
22 potential of the substance in question and quantitative estimates of risk from oral and inhalation
23 exposure. The information includes a weight-of-evidence judgment of the likelihood that the
24 agent is a human carcinogen and the conditions under which the carcinogenic effects may be
25 expressed. Quantitative risk estimates are derived from the application of a low-dose
26 extrapolation procedure, and are presented in two ways to better facilitate their use. First, route-
27 specific risk values are presented. The "oral slope factor" is an upper bound on the estimate of
28 risk per mg/kg-day of oral exposure. Similarly, a "unit risk" is an upper bound on the estimate of
29 risk per unit of concentration, either per |j,g/L drinking water or per ng/m3 air breathed. Second,
30 the estimated concentration of the chemical substance in drinking water or air when associated
31 with cancer risks of 1 in 10,000, 1 in 100,000, or 1 in 1,000,000 is also provided.
32 Development of these hazard identification and dose-response assessments for
33 bromobenzene has followed the general guidelines for risk assessment as set forth by the
34 National Research Council (1983). U.S. Environmental Protection Agency (EPA) guidelines and
35 Risk Assessment Forum Technical Panel Reports that were used in the development of this
36 assessment include the following: Guidelines for Developmental Toxicity Risk Assessment (U.S.
6/7/07 1 DRAFT - DO NOT CITE OR QUOTE
-------
1 EPA, 1991), Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines
2 for Neurotoxicity Risk Assessment (U.S. EPA, 1998a), Guidelines for Carcinogen Risk
3 Assessment (U. S. EPA, 2005a), Supplemental Guidance for Assessing Susceptibility from Early-
4 Life Exposure to Carcinogens (U.S. EPA, 2005b), Recommendations for and Documentation of
5 Biological Values for Use in Risk Assessment (U.S. EPA, 1988), (proposed) Interim Policy for
6 Particle Size and Limit Concentration Issues in Inhalation Toxicity (U.S. EPA, 1994a), Methods
7 for Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry
8 (U.S. EPA, 1994b), Use of the Benchmark Dose Approach in Health Risk Assessment (U.S. EPA,
9 1995), Science Policy Council Handbook: Peer Review (U. S. EPA, 1998b, 2000a, 2005c),
10 Science Policy Council Handbook: Risk Characterization (U.S. EPA, 2000b), Benchmark Dose
11 Technical Guidance Document (U. S. EPA, 2000c), and A Review of the Reference Dose and
12 Reference Concentration Processes (U.S. EPA, 2002).
13 The literature search strategy employed for this compound was based on the Chemical
14 Abstract Service Registry Number (CASRN) and at least one common name. Any pertinent
15 scientific information submitted by the public to the IRIS Submission Desk was also considered
16 in the development of this document. The relevant literature was reviewed through February,
17 2007.
18
6/7/07 2 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
7
8
9
10
11
12
13
14
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
2. CHEMICAL AND PHYSICAL INFORMATION RELEVANT TO ASSESSMENTS
Bromobenzene is a heavy, colorless liquid with a pungent odor (Lewis, 1997).
Synonyms include monobromobenzene and phenyl bromide (Budavari, 2001). Selected
chemical and physical properties of bromobenzene are listed below:
Figure 2-1. Chemical structure of bromobenzene
CASRN:
Molecular weight:
Chemical formula:
Boiling point:
Melting point:
Vapor pressure:
Density:
Vapor density:
Water solubility:
Other solubility:
Partition coefficient:
Flash point:
Heat of combustion:
Heat of vaporization:
Critical temperature:
Critical pressure:
Viscosity:
Vapor density (air=l):
Surface tension:
Soil sorption constant:
Air pollution factors:
Henry's Law constant:
OH reaction rate constant:
108-86-1 (Lide, 2000)
157.01 (Budavari, 2001)
C6H5Br (Budavari, 2001)
156.0°C (Lide, 2000)
-30.6°C (Lide, 2000)
4.18 mm Hg at 25°C (Riddick et al., 1986)
1.4950 g/mL at 20°C (Lide, 2000)
2.46 (air = 1) (Budavari, 2001)
4.46xl02 mg/L at 30°C (Chiou et al., 1977)
Miscible with chloroform, benzene, and petroleum
hydrocarbons. Solubility in alcohol (0.045 g/100 g at
25°C), in ether (71.3 g/100 g at 25°C) (Budavari, 2001)
log Kow = 2.99 (Hansch et al., 1995)
51°C (Budavari, 2001)
-1.98xl07 J/kg (HSDB, 2003)
44.54 kJ/mol at 25°C (Lide, 2000)
397°C (Budavari, 2001)
33,912 mm Hg (Budavari, 2001)
1.124 cp at 20°C (Budavari, 2001)
5.41 (Budavari, 2001)
0.036 N/m at 20°C (HSDB, 2003)
Koc= 150
1 mg/m3 = 0.15 ppm, 1 ppm = 6.53 mg/m3 (Verschueren,
2001)
2.47xlO"3 atm m3/mol at 25°C (Shiu and Mackay, 1997)
7.70xl013 cnrVmolecule sec at 25°C (Atkinson, 1989)
Bromobenzene is prepared commercially by the action of bromide on benzene in the
presence of iron powder (Budavari, 2001). An alternate procedure uses pyridine as a halogen
6/7/07
DRAFT - DO NOT CITE OR QUOTE
-------
1 carrier. Bromobenzene was produced in quantities less than 10,000 pounds (4.5xl03 kg) in 1986,
2 1990, 1994, 1998, and 2002 (U.S. EPA, 2002). U.S. imports of bromobenzene were 2.00xl03 kg
3 in 1984 (HSDB, 2003). Bromobenzene is used for organic synthesis, especially in the
4 production of the synthetic intermediate phenyl magnesium bromide (Budavari, 2001; Lewis,
5 1997). Bromobenzene is also used as an additive to motor oils and a crystallizing solvent.
6 Release of bromobenzene to the environment may occur during its production and the
7 production of phenyl magnesium bromide as well as in its use as a solvent and as an additive in
8 motor oil (HSDB, 2003). It has been detected at low frequencies and at low concentrations in
9 samples of food, ambient air, and finished water.
10 If released to air, bromobenzene will exist solely as a vapor in the ambient atmosphere,
11 based on its vapor pressure of 4.18 mm Hg at 25°C (Bidleman, 1988; Riddick et al., 1986).
12 Reaction of vapor-phase bromobenzene with photochemically-produced hydroxyl radicals will
13 result in degradation with an estimated half-life of 21 days (HSDB, 2003).
14 Bromobenzene is expected to have moderate to high mobility in soil, based on a soil
15 sorption constant (Koc) of 150 and an octanol/water partition coefficient (log Kow) of 2.99
16 (Hansch et al., 1995; U.S. EPA, 1987; Swann et al., 1983). Volatilization of bromobenzene from
17 moist soil surfaces may be significant, based on its Henry's Law constant of 2.47xlO"3 atm
18 nrVmol at 25°C (Shiu and Mackay, 1997; Lyman et al., 1990).
19 If released to water, bromobenzene is not expected to adsorb to suspended solids or
20 sediment, based on its Koc and water solubility (Swann et al., 1983). Bromobenzene will
21 volatilize from water surfaces, based on its Henry's Law constant (Lyman et al., 1990).
22 Hydrolysis of bromobenzene should be very slow because halogenated aromatics are generally
23 resistant to hydrolysis (Lyman et al., 1990). Experimental bioconcentration factor (BCF) values
24 ranging from 8.8 in carp to 190 in algae (Chlorellafused) suggest that bioconcentration in
25 aquatic organisms is low to moderately high (HSDB, 2003; CITI, 1992; Freitag et al., 1985).
26 Bromobenzene does not appear to be degraded rapidly by aquatic microorganisms (U.S.
27 EPA, 1987). It was not degraded at an initial concentration of 30 mg/L after 4 weeks of
28 inoculation in 100 mg/L activated sludge during a screening test (CITI, 1992).
29 Bromobenzene has been detected in water samples from the Delaware River basin, the
30 Mississippi River, the Hudson River, and Lake Michigan (U.S. EPA, 1987). The average
31 concentration of bromobenzene from eight observations in stream water reported in 1976 was
32 12.75 |ig/L, with a range of 3-38 ng/L, according to the STORET database (U.S. EPA, 1987).
33 Bromobenzene was identified with a maximum concentration of 10 ng/L in a contaminated
34 plume of groundwater near Falmouth, MA over 3500 meters long (Barber et al., 1988). The
35 plume resulted from the long-term disposal of secondary treated sewage effluent into a shallow,
6/7/07 4 DRAFT - DO NOT CITE OR QUOTE
-------
1 unconfined aquifer since 1936. The concentration of 10 ng/L was the lowest concentration
2 reported for approximately 50 volatile organic compounds that were detected.
3 Bromobenzene can be formed in small quantities during water chlorination (HSDB,
4 2003). For example, it has been detected (albeit infrequently) at low concentrations in finished
5 water in the lower Mississippi River area. During a groundwater supply survey (Westrick et al.,
6 1984), finished water samples were collected from public water systems located across the
7 United States that serve both greater than 10,000 persons and fewer than 10,000 persons.
8 Bromobenzene was detected above 0.5 |J,g/L (quantitation limit) in 3 out of 280 random sample
9 sites serving fewer than 10,000 persons with a median of positives of 1.9 ng/L and a maximum
10 value of 5.8 ng/L. It was also detected in 1 out of 186 random sample sites serving greater than
11 10,000 persons at 1.7 ng/L. In 2 of 321 nonrandom sample sites serving fewer than 10,000
12 persons, bromobenzene was detected with a median of positives of 0.97 ng/L and a maximum
13 value of 1.2 ng/L. Bromobenzene was not detected above the quantitation limit in 158
14 nonrandom sample sites serving more than 10,000 persons. In 0.13% of 24,125 public water
15 systems tested in a 20-state cross-section survey conducted for the U.S. EPA Office of Water
16 between 1993 and 1997 (U.S. EPA, 2003), bromobenzene was detected. The overall median
17 concentration of the detections was 0.5 ng/L. Detection frequency was higher in public water
18 systems using surface water (0.23% of 2664 surface water systems) than those using
19 groundwater (0.12% of 21,461 groundwater systems).
20 Bromobenzene has been detected at low concentrations in air samples collected near
21 unidentified emission sources (U.S. EPA, 1987; Brodzinsky and Singh, 1982). In 35 air samples
22 from El Dorado, AR collected from 1976 to 1978, bromobenzene concentrations ranged from
23 0.83 to 2100 ppt, with a mean concentration of 210 ppt. In 28 air samples from Magnolia, AR
24 collected in 1977, bromobenzene concentrations ranged from 0 to 8.3 ppt, with a mean
25 concentration of 1.5 ppt. Bromobenzene was not detected in seven air samples from Grand
26 Canyon, AZ or in one air sample from Edison, NJ.
27 Heikes et al. (1995) detected bromobenzene in 2 of 234 table foods above the limit of
28 quantitation (1.83 ppb) using EPA Method 524.2. Concentrations were 4.69 ppb in sandwich
29 cookies and 9.06 ppb in cake doughnuts. The authors stated that volatile halocarbons (VHCs)
30 are frequently encountered in table-ready foods as contaminant residues and that foods high in
31 fat had more elevated levels (>1000 ppb).
32
6/7/07 5 DRAFT - DO NOT CITE OR QUOTE
-------
1 3. TOXICOKINETICS RELEVANT TO ASSESSMENTS
2
3
4 3.1. ABSORPTION
5 Data on absorption of bromobenzene by the gastrointestinal tract, respiratory tract, or
6 skin in humans are not available. Findings of systemic effects following oral (Casini et al., 1984,
7 1985; Kluwe et al., 1984) or inhalation (Dahl et al., 1990; Brondeau et al., 1986) exposure of
8 animals serve as an indication that bromobenzene is absorbed through the gastrointestinal tract
9 and lungs. Quantitative data on absorption of orally-administered bromobenzene are limited.
10 However, bromobenzene is readily absorbed by the gastrointestinal tract, as evidenced by the
11 appearance of metabolites of bromobenzene (detected by gas chromatography-mass
12 spectrometry [GC-MS]) in the urine of rats, mice, and rabbits that had been administered single
13 oral doses (3-30 mg/kg-day) of bromobenzene (Ogino, 1984a). The urinary metabolites
14 accounted for 60-70% of the administered dose, most of which had been recovered in the first 8
15 hours following dosing. Absorption of bromobenzene across the lungs was demonstrated by the
16 appearance of parent compound (determined by head-space GC) in the blood of laboratory
17 animals immediately following a single 4-hour inhalation exposure to bromobenzene vapors
18 (Aarstad et al., 1990). At 1000 ppm, measured bromobenzene blood concentrations were 153,
19 102, and 47 mg/mL for rats, mice, and rabbits, respectively. In vitro experiments with rat blood
20 indicated a blood/air partition coefficient of approximately 200 (Aarstad et al., 1990). A
21 blood/air partition coefficient for bromobenzene in humans was not found.
22
23 3.2. DISTRIBUTION
24 Results of parenteral injection studies in animals indicate that, following absorption,
25 bromobenzene and its metabolites are widely distributed, with highest levels found in adipose
26 tissue (Ogino, 1984b; Zampaglione et al., 1973; Reid et al., 1971).
27 The distribution of bromobenzene following intraperitoneal injection of a 750 mg/kg-day
28 dose of bromobenzene (in sesame oil) was studied in male Sprague-Dawley rats (Reid et al.,
29 1971). Levels of bromobenzene in tissues obtained 4 and 24 hours after administration were
30 determined by gas-liquid chromatography of tissue extracts for all tissues except fat. Levels of
31 bromobenzene in fat were calculated from detected levels of 3H and the specific activity of the
32 applied 3H-bromobenzene. At 4 hours post-injection, the highest levels of bromobenzene were
33 found in fat (5600 |j,g/g tissue), followed by liver (282 ng/g), kidney (235 ng/g), brain (206
34 Hg/g), heart (146 |J,g/g), lung (142 |J,g/g), stomach (132 |J,g/g), and blood plasma (34 |J,g/g). After
35 24 hours, measured concentrations were: fat (400 ng/g), kidney (19 ng/g), stomach (17 ng/g),
36 liver (11 ng/g), brain (7.0 ng/g), lung (6.2 ng/g), heart (5.0 ng/g), and blood plasma (2 ng/g).
6/7/07 6 DRAFT - DO NOT CITE OR QUOTE
-------
1 In another study, concentrations of bromobenzene in tissues from rats 10 hours after
2 intraperitoneal injection of 5 mg of bromobenzene were highest in adipose tissue (3.38 ng/g),
3 followed by liver (0.18 ng/g), seminal fluid (0.15 ng/g), blood (0.12 ng/g), brain (0.08 ng/g), and
4 pectoral muscle (0.04 ng/g). Levels of bromobenzene in kidney, spleen, heart, and lung tissues
5 were below the detection limit of 0.01 ng/g. Levels of phenolic metabolites (m-bromophenol
6 and p-bromophenol) were highest in the kidney (0.43 ng/g), lungs (0.27 ng/g), and blood (0.19
7 Hg/g), with lesser amounts in seminal fluid, brain, heart, liver, and pectoral muscle; proportions
8 of the individual phenols (m-bromophenol and p-bromophenol) were approximately equal in
9 each of the tissues examined (Ogino, 1984b). The phenols were below the level of detection
10 (0.01 ng/g) in spleen and adipose tissues. Concentrations of bromobenzene were reported to
11 show a pattern of peaking within 10 hours after dosing, followed by rapidly decreasing
12 concentrations, but collected data to show this pattern were not reported (Ogino, 1984b).
13 In order to monitor tissue distribution immediately following exposure, male Sprague-
14 Dawley rats were administered 14C-bromobenzene intravenously at a dose of 10 |j,mol/kg and
15 plasma levels of radioactivity were monitored (Zampaglione et al., 1973). Plasma levels dropped
16 triphasically during 70 minutes following administration. During the first 5 minutes following
17 dosing, radioactivity in the liver increased to a peak, at which time measured radioactivity was
18 highest in the liver, followed by adipose tissue and plasma in decreasing order. Levels in the
19 liver subsequently dropped in a manner similar to that of plasma radioactivity, although
20 measured levels in the liver remained higher than those in the plasma. Adipose tissue levels
21 reached a peak within 20 minutes after dosing and remained high throughout the 70-minute
22 observation period.
23 Monks et al. (1982) assessed distribution by monitoring covalent binding to the protein
24 fraction in various tissues following intraperitoneal injection of 3 mmol/kg (471 mg/kg-day) of
25 14C-bromobenzene in male Sprague-Dawley rats. Covalent binding to proteins was most
26 prominent in the liver, followed by the kidney, small intestine, lung, and muscle.
27
28 3.3. METABOLISM
29 The metabolism of bromobenzene has been extensively studied in in vivo and in vitro
30 mammalian systems (see Lau and Monks, 1997a,b; Lertratanangkoon et al., 1993; Lau and
31 Monks, 1988). Based on available data, a proposed metabolic scheme for bromobenzene is
32 illustrated in Figure 3-1. There are two initial competing steps involving conversion of
33 bromobenzene to either the 3,4-oxide derivative catalyzed by phenobarbital-induced cytochrome
34 isozymes CYP 450 1A2, 2A6, 2B6, and 3 A4 or the 2,3-oxide derivative catalyzed by
35 3-methylcholanthrene and p-naphthoflavone-induced CYP isozymes, CYP 450 1A1, 1A2, and
36
6/7/07 7 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
o
5
4
BNF-, 3MC-, and PB-
induced CYP isozymes
PB-induced
CYP isozymes
Covalent binding
Hepatotoxicity?
Covalent
binding
Toxicity?
premercaturic
and mercapturic
acids in urine
Spontaneous
rearrangement
Covalent
binding
Nephrotoxicity?
H GSH
EH") / 3,4-oxide
4-glutathionyl 3-glutathionyl
conjugate conjugate
Rr
OH HO H
2,3-dihydrodiol 3,4-dihydrodiol
Oxidative
debromination
Spontaneous
rearrangement
OH Q
4-bromophenol 1,4-benzoquinone
OH
3-bromophenol
[4-bromophenol 5,6-oxide] '
I I
Toxicity?
[2-bromophenol oxides]
I
[3-bromophenol _ Covalent
oxides] binding
OH
OH
4-bromocatechol
OH
3-bromocatechol
Reactive
oxygen
species
Hepatotoxicity?
Nephrotoxicity?
Urinary metabolites
2-bromophenol
Covalent^
binding
Toxicity?
2-bromocatechol (2-bromohydroquinone)
2-bromo-3-(glutathion-S-yL)hydroquinone
2-bromo-5-(glutathon-S-yl_)hydroquinone
2-bromo-6-(glutathion-S-yL)hydroquinone
2-bromo-bis-(glutathion-S-yL)hydroquinone
4-bromoquinone Hepatotoxicity?
GSH-
[6-glutathion-S-yL-4-bromocatechol]
SMC = 3-methylcholanthrene; BNF = beta-naphthoflavone; CYP = cytochrome P-450; DDDH = dihydrodiol dehydrogenase;
EH = epoxide hydrolase; GSH = glutathione; GST = glutathione S-transferase; PB = phenobarbital
Figure 3-1. Proposed metabolic scheme for bromobenzene in mammals
(adapted from Lertratanangkoon et al., 1993; Lau and Monks, 1988)
6/7/07
DRAFT - DO NOT CITE OR QUOTE
-------
1 1B1, as well as phenobarbital-induced CYP isozymes (Girault et al., 2005; Krusekopf et al.,
2 2003; Lau and Zannoni, 1979, 1981a; Reid et al., 1971).
3 The predominant metabolic pathway in the rat liver leads to enzymatic (glutathione-S-
4 transferase) conjugation of the 3,4-oxide derivative with glutathione, followed by urinary
5 excretion as premercapturic and mercapturic acids, as evidenced by the recovery of
6 approximately 70% of the radioactivity as mercapturic acids in the urine of male Sprague-
7 Dawley rats that had been injected intravenously with 0.05 mmol/kg (7.9 mg/kg-day) of
8 14C-bromobenzene (Zampaglione et al., 1973). Glutathione conjugation is thought to be a
9 protective mechanism for acute bromobenzene hepatotoxicity (see Section 4.5.3). The 2,3-oxide
10 derivative has not been observed to undergo glutathione conjugation.
11 Both the 3,4- and 2,3-oxide derivatives may be converted to the corresponding
12 dihydrodiols by epoxide hydrolase (EH). The subsequent formation of bromophenols (2-, 3-,
13 and 4-bromophenol) from the oxide derivatives includes several proposed pathways
14 (Lertratanangkoon et al., 1993; Lau and Monks, 1988). The chemical instability of the 2,3-oxide
15 derivative and its relatively short biological half-life indicate that spontaneous rearrangement is
16 the predominant pathway to the formation of 2-bromophenol in the rat and guinea pig in vivo
17 (Lertratanangkoon et al., 1993), although it has been suggested that both 2- and 3-bromophenol
18 may also be formed by rearrangement of the 2,3-dihydrodiol (Lertratanangkoon et al., 1987,
19 1993; see also Figure 3-1). Other pathways to the formation of 3-bromophenol may include
20 rearrangement of the 3,4-dihydrodiol or the 4-S-glutathione conjugate of the 3,4-oxide derivative
21 (Lertratanangkoon et al., 1987, 1993). Spontaneous rearrangement of the 3,4-dihydrodiol is
22 thought to be the major pathway leading to the formation of 4-bromophenol in the rat, whereas
23 the pathway leading through the 3-S-glutathione conjugate of the 3,4-oxide derivative is thought
24 to predominate in the guinea pig (Lertratanangkoon et al., 1987, 1993).
25 The bromophenol metabolites may be subsequently oxidized by CYP to their respective
26 bromocatechols (2-, 3-, or 4-bromocatechol, Figure 3-1), likely involving bromophenol oxide
27 intermediates. The 4-bromocatechol may also be formed via dihydrodiol dehydrogenase
28 (DDDH)-catalyzed conversion of the 3,4-dihydrodiol, the pathway that appears to predominate
29 in the rat in vivo (Miller et al., 1990). The 4-bromophenol may undergo oxidative debromination
30 to form 1,4-benzoquinone (Slaughter and Hanzlik, 1991; Zheng and Hanzlik, 1992). Redox
31 cycling of 2- and 4-bromocatechol and conjugation by glutathione S-transferase (GT) produce
32 2-bromo-3-(glutathion-S-yL)hydroquinone and 6-glutathion-S-yL-4-bromocatechol, respectively
33 (Lau and Monks, 1988).
34 Mercapturic acids are the predominant urinary metabolites of bromobenzene in
35 laboratory animals, indicating that glutathione conjugation of the 3,4-epoxide is the primary
36 metabolic pathway for bromobenzene. Approximately 60-70% of the administered dose was
6/7/07 9 DRAFT - DO NOT CITE OR QUOTE
-------
1 detected (using GC-MS) as mercapturic acids, derived from the 3,4-oxide pathway, in the
2 24-hour urine of rats given bromobenzene parenterally at doses ranging from 7.9 to 158
3 mg/kg-day (Chakrabarti and Brodeur, 1984; Zampaglione et al., 1973). Following oral
4 administration of bromobenzene (10 mg/rat, 1 mg/mouse, 10 mg/rabbit), approximately 50-60%
5 of the 96-hour urinary recovery of bromobenzene metabolites was in the form of
6 4-bromophenylmercapturic acid (Ogino, 1984a). Other metabolites that have been measured in
7 the urine of rats include the phenolic compounds, dihydrodiols, catechols, and hydroquinones
8 (Miller et al., 1990; Lertratanangkoon and Horning, 1987; Chakrabarti and Brodeur, 1984; Lau et
9 al., 1984a; Monks et al., 1984a,b; Jollow et al., 1974; Zampaglione et al., 1973).
10 Animal studies have elucidated species-specific differences in urinary excretion of the
11 bromophenols (2-, 3-, and 4-bromophenol) following exposure to bromobenzene. For example,
12 in the 96-hour urine of mice that had been administered a nontoxic oral dose of bromobenzene (1
13 mg/mouse; approximately 33 mg/kg-day), 2-bromophenol accounted for 12.1% of the dose,
14 3-bromophenol accounted for 8.8%, and 4-bromophenol accounted for 3.1% (Ogino, 1984a). In
15 similarly-treated rats (10 mg/rat; approximately 56 mg/kg-day), however, 2-bromophenol
16 accounted for only 2.6% of the dose, while 3-bromophenol accounted for 19.2% and
17 4-bromophenol accounted for 13.1%. In the urine of the mice, 2-bromophenol was 4 times more
18 prevalent than 4-bromophenol, whereas 4-bromophenol was 5 times more prevalent than
19 2-bromophenol in the urine of the rats. This metabolic difference between rats and mice has
20 been associated with a difference in susceptibility to bromobenzene acute nephrotoxicity (Reid,
21 1973; see also Section 4.5.3).
22 Metabolism of bromobenzene in the liver appears to be capacity-limited. For example,
23 approximately 79% of the radioactivity from an intraperitoneally-injected nonhepatotoxic (130
24 mg/kg-day) dose of 14C-bromobenzene was recovered in the urine of rats within 24 hours
25 following administration, whereas only 47% was recovered following a hepatotoxic (1200
26 mg/kg-day) dose (Lertratanangkoon and Horning, 1987). Section 4.5.3 discusses relationships
27 between glutathione depletion and hepatotoxicity in more detail.
28 Although liver tissue has been shown to be capable of producing all of the major
29 metabolites depicted in Figure 3-1, as demonstrated by numerous in vivo and in vitro animal
30 studies, bromobenzene can be metabolized at sites other than the liver. In vitro studies in rats
31 and mice have demonstrated that lung (Monks et al., 1982; Reid et al., 1973) and kidney (Monks
32 et al., 1982) tissues are capable of metabolizing bromobenzene, although the extent to which
33 extrahepatic tissues metabolize bromobenzene in vivo is not known.
34 Following oral exposure, a first-pass metabolic effect is expected to occur due to the
35 extensive metabolic capacity of the liver; however, the extent of the first-pass effect as a function
6/7/07 10 DRAFT - DO NOT CITE OR QUOTE
-------
1 of administered dose has not been empirically characterized. Likewise, the extent of first-pass
2 metabolism in the lung has not been demonstrated following inhalation exposure.
3 Recent studies have noted that intraperitoneal injection of bromobenzene into rats can
4 induce many different types of enzymes. In a toxicogenomics approach, Heijne et al. (2005,
5 2004, 2003) noted induction of more than 20 liver proteins (including y-glutamylcysteine
6 synthetase, a key enzyme in glutathione biosynthesis) and transient changes in the transcriptional
7 expression of numerous genes involved in drug metabolism, oxidative stress, glutathione
8 depletion, the acute phase response, metabolism, and intracellular signaling following
9 intraperitoneal administration of bromobenzene to rats. Other studies (Minami et al., 2005;
10 Stierum et al., 2005; Waters et al., 2006) have utilized toxicogenomics to characterize the
11 relationship between bromobenzene hepatotoxicity and hepatic gene expression profiles.
12
13 3.4. ELIMINATION
14 Results of animal studies indicate that urinary excretion of metabolites is the principal
15 route of elimination of absorbed bromobenzene (Lertratanangkoon and Horning, 1987; Merrick
16 et al., 1986; Ogino, 1984a; Zampaglione et al., 1973; Reid et al., 1971), although biliary
17 excretion of the 3- and 4-glutathionyl conjugates formed from the 3,4-oxide derivative has been
18 demonstrated in bile-cannulated rats (Sipes et al., 1974).
19 In rats, mice, and rabbits given bromobenzene in single oral doses of approximately 3-30
20 mg/kg-day, detection of metabolites in urine collected for 4 days accounted for 60-70% of the
21 administered dose, most of which had been recovered within 8 hours following administration
22 (Ogino, 1984a). Small amounts of parent compound were detected in the urine and feces of all
23 three species. Approximately 85% of an intraperitoneally injected dose (250 mg/kg-day) of
24 14C-bromobenzene was excreted within 24 hours as metabolites in the urine of rats (Reid et al.,
25 1971). In other rat studies, metabolites detected in the urine collected for 48 hours accounted for
26 more than 90% of administered doses of 8 mg/kg-day (intravenous) or 1570 mg/kg-day
27 (intraperitoneal) (Zampaglione et al., 1973).
28 Biliary excretion of bromobenzene-glutathione conjugate has been demonstrated in rats;
29 the rate of biliary excretion can be used as an index of in vivo activation of bromobenzene
30 (Madhu and Klaassen, 1992). Additional information regarding biliary excretion of
31 bromobenzene metabolites was demonstrated in bile-cannulated rats that were administered a
32 non-hepatotoxic dose (20 mg/kg-day) of 14C-bromobenzene in the femoral vein (Sipes et al.,
33 1974). Cumulative excretion of radioactivity in the bile was 56% of administered radioactivity
34 during 3 hours after dosing. Combined with demonstrations that, in normal non-cannulated rats,
35 elimination of bromobenzene predominantly occurs via urinary excretion of metabolites (Ogino,
36 1984a; Zampaglione et al., 1973; Reid et al., 1971) and not via fecal excretion (Ogino, 1984a), it
6/7/07 11 DRAFT - DO NOT CITE OR QUOTE
-------
1 appears that most of the metabolites in the bile are reabsorbed from the intestine by enterohepatic
2 circulation and subsequently excreted by the kidneys.
3 The biological half-life of bromobenzene in laboratory animals is relatively short. Using
4 a two-phase model, Ogino (1984a) calculated a half-life of 4.65 hours for the first phase (0-16
5 hours) and 26.8 hours for the second phase (24-96 hours), based on total excretion of brominated
6 compounds in the urine of mice given a single oral dose of approximately 33 mg/kg-day. A first-
7 order elimination half-life of 5.87 hours was calculated for clearance of radioactivity from the
8 blood of rats given a relatively high (1178 mg/kg-day) dose of 14C-bromobenzene by
9 intraperitoneal injection (Merrick et al., 1986). A much shorter first-phase half-life
10 (approximately 10 minutes) was reported for the elimination of radioactivity from the whole
11 body of rats that had been injected intravenously with a nontoxic (8 mg/kg-day) dose of
12 radiolabeled bromobenzene (Zampaglione et al., 1973). In this study, a second-phase half-life
13 was not calculated.
14
15 3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS (PBTK)
16 No PBTK models have been developed for bromobenzene.
17
6/7/07 12 DRAFT - DO NOT CITE OR QUOTE
-------
1 4. HAZARD IDENTIFICATION
2
3
4 4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL
5 CONTROLS
6 Studies on health effects in humans exposed to bromobenzene were not identified in
7 literature searches.
8
9 4.2. LESS-THAN-LIFETIME AND CHRONIC STUDIES AND CANCER BIOASSAYS
10 IN ANIMALS—ORAL AND INHALATION
11 4.2.1. Oral Exposure
12 4.2.1.1. Subchronic Toxicity
13 The National Toxicology Program (NTP) conducted subchronic gavage studies of
14 bromobenzene in rats (NTP, 1985a) and mice (NTP, 1985b). These studies have not been
15 officially released by NTP, but unpublished reports, including the review comments and
16 conclusions of NTP's Pathology Working Group (NTP, 1986a), were obtained from NTP. The
17 unpublished NTP studies are available by calling EPA's IRIS Hotline at (202)566-1676, by fax
18 at (202)566-1749 or by email at iris@epa.gov.
19 Groups of 10 male and 10 female Fischer 344/N rats were given 0, 50, 100, 200, 400, or
20 600 mg/kg-day of bromobenzene (>99% purity) by gavage in corn oil 5 days/week for 90 days in
21 the basic study. In a supplementary study designed to evaluate clinical pathologic effects of
22 bromobenzene, groups of five rats/sex were similarly treated with 0, 50, 200, or 600 mg/kg-day
23 and housed individually in metabolism cages throughout the study; urine samples were collected
24 from these rats on days 1, 3, 23, and 94 for detailed urinalysis. Blood samples were collected on
25 days 2, 4, 24, and 95 for hematology and clinical chemistry. Rats from both the basic and
26 supplementary studies were observed twice daily for morbidity and mortality. Clinical
27 observations and body weight measurements were performed weekly. Blood samples for
28 hematologic and clinical pathologic examinations were collected from all surviving rats at
29 terminal sacrifice. Terminal body and organ (liver, brain, testis, kidney, lung, heart, and thymus)
30 weights were recorded; organ-to-body weight and organ-to-brain weight ratios were calculated
31 for each sex. Complete gross necropsy was performed on all rats. Complete histopathologic
32 examinations of all major tissues and organs (including liver, kidney, urinary bladder, spleen,
33 pancreas, brain, spinal cord, sciatic nerve [if neurologic signs were present], heart, lung, trachea,
34 nasal cavity, esophagus, stomach, small intestine, cecum, colon, uterus, ovaries, preputial or
35 clitoral glands, testes, prostate, seminal vesicles, sternebrae, adrenals, pituitary, thyroid,
36 parathyroids, salivary gland, mandibular and mesenteric lymph nodes, thymus, mammary gland,
6/7/07 13 DRAFT - DO NOT CITE OR QUOTE
-------
1 blood, gross lesions, and tissue masses) were performed on all control rats and all rats from the
2 400- and 600-mg/kg-day dose groups.
3 In the basic study, all rats of the 50- and 100-mg/kg-day groups were subjected to
4 histopathologic examination of liver and kidney. Furthermore, sections of livers from all control
5 and bromobenzene-treated rats were examined following hematoxylin and eosin (H&E) and
6 periodic acid-Schiff (PAS) staining for glycogen. In the supplementary study, liver and kidney
7 tissues from all rats and any gross lesions were examined histologically. Serum of rats in the
8 supplementary study was assessed for blood urea nitrogen (BUN), creatinine, alanine
9 aminotransferase (ALT), sorbitol dehydrogenase (SDH), glucose, and aspartate aminotransferase
10 (AST). Parameters assessed in urinalysis included volume, color, specific gravity, pH,
11 hemoglobin, glucose, creatinine, and protein. Hematologic evaluations of blood collected at
12 terminal sacrifice from all surviving rats included erythrocyte and leukocyte counts and
13 morphology; hemoglobin concentration; volume of packed cells; measures of mean corpuscular
14 volume (MCV), mean corpuscular hemoglobin (MCH), and mean corpuscular hemoglobin
15 content (MCHC); qualitative estimates of leukocyte differential count; and platelet and
16 reticulocyte counts. Serum was analyzed for BUN, creatinine, ALT, SDH, total protein,
17 albumin, albumin/globulin ratio, glucose, and AST.
18 In the basic study, treatment-related clinical signs were observed only at the 600 mg/kg-
19 day dose level and included ruffled fur (9/10 rats of each sex), emaciation (9/10 rats of each sex),
20 tremors (2/10 males and 1/10 females), ataxia (1/10 of each sex), hypoactivity (5/10 males and
21 7/10 females), and ocular discharge (2/10 of each sex). Observations of similar clinical signs
22 were made in rats of the supplementary study, but distinguishing between treatment-related
23 clinical signs and symptoms that may have resulted from repeated anesthesia, blood sample
24 collection, and prolonged housing in metabolism cages was difficult.
25 Treatment-related mortality was observed in male and female rats at 600 mg/kg-day (9/10
26 males and 8/10 females in the basic study and 3/5 males and 1/5 females in the supplementary
27 study). By the end of week 7, mortality rates in high-dose male and female rats were 7/10 and
28 3/10, respectively. Occasional deaths at lower doses were attributed to gavage error.
29 Statistically significantly reduced mean body weight (approximately 7-11% lower than controls)
30 was observed in 400-mg/kg-day male rats from week 5 until study end. At 600 mg/kg-day, both
31 male and female rats were visibly emaciated. Table 4-1 presents terminal body and liver weights
32 and serum levels of selected liver enzymes in male and female rats of the basic study. Dose-
33 related statistically significantly increased mean liver and kidney weights (absolute, relative-to-
34 body weight) were observed at doses >100 mg/kg-day in male rats and at all dose levels
35 (including 50 mg/kg-day) in female rats. Changes in the 600 mg/kg-day males were similar in
36 magnitude to changes in the 400 mg/kg-day males, but could not be assessed for statistical
6/7/07 14 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-1. Effects of bromobenzene on terminal body and liver weights and serum liver enzymes of male and female Fischer 344/N rats exposed by
oral gavage 5 days/week for 90 days in the basic study (mean +/- standard deviation)
Male rats
Dose (mg/kg-day)
Number of rats
Body weight (g)
Liver weight (g)
Difference (%)d
Ratio liver/body weight
Difference (%)d
Serum AST (IU/L)
Serum ALT (IU/L)
Serum SDH (IU/L)
Controls
10
343.0+12.9
9.16 + 0.66
26.72+1.88
83.70 + 10.97
41.90 + 9.33
3.90 + 2.59
50
10
330.3 + 12.2
Not available
Not available
93.40 + 18.39
41.30 + 6.66
3.68 + 1.85
100
9
342.3 + 18.5
10.64b + 0.76
+16.2
31.08b+1.18
+16.4
82.56 + 17.63
38.67 + 9.45
3.56 + 0.96
200
8
331.3+20.0
11.29b + 0.69
+23.3
34.10b + 0.68
+27.7
87.88 + 10.64
39.50 + 7.28
5.25 + 1.64
400
10
293.0b+11.9
11.87b + 0.80
+29.6
40.56b + 3.16
+51.9
820. 10b + 694.95
893.20b + 727.39
311.90b+ 228.19
600
1"
203. lc
10.50
+14.6
51.70C
+93.6
268.00
403.00
80.00
Female rats
Dose (mg/kg-day)
Number of rats
Body weight (g)
Liver weight (g)
Difference (%)d
Ratio liver/body weight
Difference (%)d
Serum AST (IU/L)
Serum ALT (IU/L)
Serum SDH (IU/L)
Controls
10
192.8 + 9.0
4.68 + 0.35
24.25 + 1.13
88.50 + 23.69
41.70 + 10.83
3.80 + 0.98
50
10
197.1 + 11.9
5.23b + 0.37
+11.6
26.55b+1.23
+9.5
83.50 + 5.35
37.50 + 5.16
4.00 + 1.26
100
10
193.5 + 9.1
5.55b + 0.36
+18.6
28.69b+1.20
+18.3
74.30+12.92
30.70 + 6.17
6.20b+1.47
200
10
187.6 + 8.2
6.28b + 0.40
+34.2
33.48b+1.37
+38.1
72.60 + 10.24
27.80 + 4.71
3.78 + 0.98
400
10
182.3b+10.5
7.85b + 0.49
+67.7
43.11b + 2.38
+77.8
215.20 + 339.55
265.38 + 596.73
61.60 + 143.07
600
3a
167.4b + 9.8
9.11b + 0.57
+94.7
54.78b + 6.64
+125.9
119.00 + 48.00
111.00 + 59.00
23.00 + 17.00
2 aHigh rates of early mortality at the 600 mg/kg-day dose level (9/10 males and 7/10 females) preclude meaningful statistical analysis of terminal body and organ
3 weight data or serum enzyme changes
4 bStatistically significantly increased from controls (p<0.05) based on student's two-tailed t-test
5 °Outside 3 standard deviations from the control mean
6 dChange relative to controls
7 Source: NTP (1985a)
6/7/07
15
DRAFT - DO NOT CITE OR QUOTE
-------
1 significance because only one survivor remained in this group at study termination. Significant
2 increases in serum enzymes indicative of hepatotoxicity (ALT, AST, SDH) were found in 400
3 mg/kg-day male rats, but not males of lower dose levels. Serum SDH was significantly
4 increased in 100 mg/kg-day female rats (approximately 60% greater than that of controls), but
5 was not increased at the next higher dose level (200 mg/kg-day). Female rats of the 400
6 mg/kg-day dose level exhibited mean serum levels of ALT, AST, and SDH that were markedly
7 increased over controls, but the large variance precluded using the t-test for statistical analysis
8 (see Table 4-1). Significant increases in serum creatinine (males and females) and BUN (males
9 only) were also observed at doses >400 mg/kg-day. Effects on the hematopoietic system were
10 generally unremarkable. Significantly increased mean relative (but not absolute) testis weight
11 was noted in male rats of the 400 and 600 mg/kg treatment groups (increased by 10 and 35%,
12 respectively, over controls). There were no indications of treatment-related effects on
13 reproductive organ weights in female rats.
14 As shown in Table 4-2, histopathologic examinations revealed treatment-related
15 significantly increased incidences of rats exhibiting cytomegaly (doses >200 mg/kg-day in males
16 and >400 mg/kg-day in females), inflammation (doses >200 mg/kg-day in males), and necrosis
17 (doses >400 mg/kg-day in males and females). Cytomegaly was characterized by an increase in
18 the size of the nucleus and cytoplasm of individual hepatocytes and was more common in the
19 central parts of the hepatic lobule. Liver necrosis was primarily coagulative in nature and
20 considered a direct result of bromobenzene treatment. Inflammation was principally
21 centrilobular and consisted of focal infiltrates of macrophages, lymphocytes, and occasional
22 neutrophils. The incidences and severity of each of these liver lesions generally increased with
23 increasing dose. Centrilobular mineralization was observed in 2/10 and 1/10 high-dose males
24 and females, respectively, and was considered to be the result of hepatocellular necrosis. Other
25 histological findings in the liver included cytoplasmic alterations, infiltration, and pigmentation,
26 which were generally of low incidence and did not exhibit consistent dose-response
27 characteristics.
28 There is some evidence to suggest a common mechanism of action for bromobenzene-
29 induced cytomegaly, necrosis, inflammation, and mineralization. All four lesions were
30 principally observed in the central part of the hepatic lobules. Significantly increased incidences
31 of hepatocellular necrosis or inflammation were observed only at doses equal to or greater than
32 those eliciting significantly increased incidences of cytomegaly. In the NTP report,
33 inflammation and mineralization were considered to be direct results of hepatocellular necrosis
34 (NTP, 1985a). Based on these observations, incidences of rats with one or more of these liver
35 lesions (cytomegaly, necrosis, inflammation, mineralization) were combined for each sex (as
36 shown in Table 4-2).
37
6/7/07 16 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-2. Incidences of male and female Fischer 344/N rats with liver and kidney lesions
following administration of bromobenzene by gavage 5 days/week for 90 days in the basic
study
Endpoint
Dose (mg/kg-day)a
0
Incidence
Severity
50
Incidence
Severity
100
Incidence
£
4*
4*
O5
200
Incidence
Severity
400
Incidence
£
4>
4>
O5
600b
Incidence
£
4>
4>
O5
Males
Liver, centrilobular
Inflammation
Cytomegaly
Necrosis
Mineralization
Combined0
Kidney, tubule
Necrosis
Degeneration
Casts
Mineralization
Pigment
2/10
0/10
0/10
0/10
2/10
0/10
2/10
0/10
0/10
0/10
1.0
1.0
2/10
0/10
0/10
0/10
2/10
0/10
1/10
0/10
0/10
0/10
1.0
1.0
2/10
0/10
0/10
0/10
2/10
0/10
2/10
0/10
0/10
0/10
1.0
2.0
7/1 Od
4/1 Od
3/10
0/10
7/1 Od
0/10
4/10
1/10
0/10
0/10
1.6
1.5
1.3
1.0
1.0
9/1 Od
10/10d
9/1 Od
0/10
10/10d
0/10
1/10
3/10
0/10
7/1 Od
2.1
2.0
2.0
2.0
2.0
1.9
7/1 Od
9/1 Od
9/1 Od
2/10
10/10d
6/1 Od
7/1 Od
7/1 Od
3/10
0/10
2.1
2.4
2.4
2.5
2.2
2.6
2.6
2.3
Females
Liver, centrilobular
Inflammation
Cytomegaly
Necrosis
Mineralization
Combined0
Kidney, tubule
Necrosis
Degeneration
Casts
Mineralization
Pigment
2/10
0/10
0/10
0/10
2/10
0/10
0/10
0/10
0/10
0/10
1.5
2/10
0/10
0/10
0/10
2/10
0/10
0/10
0/10
0/10
0/10
1.0
4/10
0/10
0/10
0/10
4/10
0/10
0/10
0/10
0/10
0/10
1.5
3/10
3/10
0/10
0/10
5/10
0/10
0/10
0/10
1/10
0/10
1.0
1.0
2.0
6/10
10/10d
7/1 Od
0/10
10/10d
0/10
1/10
0/10
0/10
8/1 Od
1.7
2.4
2.0
2.0
2.1
5/10
10/10d
9/1 Od
1/10
10/10d
6/1 Od
8/1 Od
6/1 Od
3/10
2/10
2.8
2.6
2.7
3.0
2.3
3.0
2.5
2.0
2.0
2 Incidence = number of animals in which lesion was found/number of animals in which organ
3 was examined.
4 bMost male and female rats of the 600 mg/kg-day dose level died during the study, which may
5 have affected incidences of selected lesions.
6 °Incidences of rats with one or more of the liver lesion types (cytomegaly, necrosis,
7 inflammation, mineralization), extracted from individual animal histopathologic results provided
8 to Syracuse Research Corporation by NTP.
9 dStatistically significantly different from control groups according to Fisher's exact test (p<0.05),
10 performed by Syracuse Research Corporation.
11 Severity: Average severity score: 1 = minimal, 2 = mild, 3 = moderate, 4 = severe.
12 Source: NTP (1985a)
13
6/7/07
17
DRAFT - DO NOT CITE OR QUOTE
-------
1 Observed kidney effects included a brown staining pigment of the cytoplasm (presumed
2 to be bile pigment) in the convoluted tubules of 400-mg/kg-day male and female rats and
3 degeneration of the convoluted tubules and necrosis in 600-mg/kg-day males and females, in the
4 absence of indications of tubular regeneration. It was noted that the reduced incidence of the
5 tubular (brown-staining) pigment in the 600-mg/kg-day rats (0/10 males and 2/10 females) might
6 be related to high rates of early mortality at this dose level, in which case there may not have
7 been enough time for this lesion to appear. Other histopathologic effects (hyperkeratosis,
8 ulceration, and hemorrhage in the stomach; brain mineralization and necrosis; thymus atrophy;
9 and bone marrow atrophy) were observed only in the high-dose groups of male and female rats.
10 The effects in the stomach were probably associated with bolus gavage dosing. Atrophy or
11 necrosis of the thymus was observed in six male and six female rats treated in the 600 mg/kg
12 dose group. These effects were only noted in rats that died or were euthanized while moribund
13 and were considered to be the result of stress. Testicular degeneration of moderate severity was
14 noted in a single high-dose male rat. Gross and histopathologic examinations of female
15 reproductive tissues did not reveal treatment-related effects.
16 The NTP Pathology Working Group (NTP, 1986a) reviewed the pathology results from
17 the subchronic gavage studies in rats and mice (NTP, 1985a). This group designated the brain as
18 an organ susceptible to chemically-related lesions based on cerebellar necrosis (granular layer)
19 in 1 of 10 males and 3 of 10 females in the 600 mg/kg dose group; however, some members of
20 the group (2 of 6 ) thought that degeneration, rather than necrosis, was a more appropriate
21 descriptor of the lesion in some animals. The Pathology Working Group (NTP, 1986a) noted
22 that bone marrow atrophy was either absent or only minimally present in the 400 mg/kg group,
23 but was recorded in 3 of 10 males and 6 of 10 females in the 600 mg/kg group. It was also noted
24 that most of the rats in this dose group died or were sacrificed in a moribund state and were
25 emaciated, raising the possibility of marrow atrophy as a secondary rather than a direct effect.
26 The Pathology Working Group (NTP, 1986a) indicated that testicular degeneration was apparent
27 in a number of high-dose male rats, but suggested that this effect may have been secondary to
28 emaciation. .
29 The most prominent toxicological effects observed in Fischer 344/N rats treated with
30 bromobenzene by oral gavage for 90 days (NTP, 1985a) were observed in the liver.
31 Significantly increased incidences of hepatocellular necrosis (a clear indicator of an adverse
32 effect) were observed at doses of 400 and 600 mg/kg-day in both male and female rats.
33 Significantly increased incidences of cytomegaly were noted at doses >200 mg/kg-day in male
34 rats and at doses >400 mg/kg-day in female rats. Statistically significant increases in mean liver
35 weight were observed at doses as low as 50 mg/kg-day in female rats and 100 mg/kg-day in male
36 rats.
6/7/07 18 DRAFT - DO NOT CITE OR QUOTE
-------
1 Treatment-related increased occurrence of cytomegaly and increased liver weight
2 represent an adaptive liver response to bromobenzene, a known enzyme-inducing agent, and may
3 provide an indication of liver toxicity from higher levels of exposure. By themselves, increased
4 liver weight and increased incidences of cytomegaly can be considered to be of questionable
5 toxicological significance.
6 The biological significance of the presence of pigments in the convoluted tubules of the
7 kidneys of 400 mg/kg-day male and female rats is unclear. Incidences of other renal tubular
8 effects (necrosis, degeneration, and casts) were statistically significantly increased only in high-
9 dose male and female rats.
10 In the NTP (1985a) study the LOAEL is considered to be 50 mg/kg-day in female rats for
11 statistically significant increased liver-to-body weight ratios and absolute liver weights. The
12 designation of increased liver weights as an adverse effect is supported by the presence of liver
13 lesions (including inflammation, cytomegaly, and necrosis) and elevated serum enzymes
14 indicative of liver damage at higher doses.
15 In the mouse study (NTP, 1985b), groups of 10 male and 10 female B6C3F1 mice were
16 administered 0, 50, 100, 200, 400, or 600 mg/kg-day of bromobenzene by gavage in corn oil 5
17 days/week for 90 days; supplementary groups of 10 mice/sex were similarly treated with 0, 50,
18 200, or 600 mg/kg-day and housed in pairs in metabolism cages throughout the study. Blood
19 samples were collected on days 2, 4, 24, and 95 for hematology and clinical chemistry. Urine
20 and clinical chemistry samples were collected from these mice on days 1, 3, 17, and 94. Other
21 details of study design were the same as those described for the rats (NTP, 1985a), with the
22 exception of histopathologic examination of kidney tissues, which was not performed in 50 or
23 100 mg/kg-day mice.
24 In the basic study of mice, clinical signs of treatment-related effects were minimal and
25 apparent mainly during the first week of treatment and included ruffled fur (8/10 of the 400
26 mg/kg-day males, 7/10 of the 600 mg/kg-day males, 8/10 of the 600 mg/kg-day females) and
27 hypoactivity (6/10 of the 600 mg/kg-day males). The only reported clinical sign in the
28 supplementary groups of treated mice was that of ruffled fur in 9/10 and 6/10 of the 600 mg/kg-
29 day males and females, respectively.
30 Deaths that could be attributed to bromobenzene included 5/10 and 2/10 of the 600
31 mg/kg-day males of the basic and supplementary studies, respectively. The original report
32 included 1/10 and 2/10 deaths in the 400 mg/kg-day males and females, respectively, from the
33 basic study. However, in these cases, results of histologic examinations indicated that gavage
34 error likely contributed to the deaths. Occasional other deaths among control and treated males
35 and females were likely the result of gavage error or anesthesia. At the end of the basic study,
36 body weight was significantly decreased (approximately 9% lower than controls) in 400
6/7/07 19 DRAFT - DO NOT CITE OR QUOTE
-------
1 mg/kg-day (but not 600 mg/kg-day) males. The 600 mg/kg-day males in the supplemental study
2 exhibited approximately 12% lower terminal body weight, relative to controls. Consistent
3 treatment-related effects on body weight were not seen in female mice. Table 4-3 presents
4 terminal body and liver weights and serum levels of selected liver enzymes in male and female
5 mice of the basic study. In male mice, absolute liver weight was significantly increased at dose
6 levels >200 mg/kg-day, while the liverbody weight ratio was significantly increased at dose
7 levels >100 mg/kg-day and the liverbrain weight ratio was significantly increased at dose levels
8 >400 mg/kg-day. In female mice, all three measures of liver weight were significantly increased
9 at all dose levels, relative to controls. The effect on absolute liver weight increased with dose,
10 ranging from approximately 12% in the 50 mg/kg-day group to greater than 50% in the 600
11 mg/kg-day group. Statistically significantly increased serum SDH activity (indicative of
12 hepatotoxicity) was observed in both sexes at dose levels >200 mg/kg-day, relative to sex-
13 matched controls, but the magnitude only approached a 2-fold increase (a biologically significant
14 level) at >200 mg/kg-day in males and >400 mg/kg-day in females. Activities of AST or ALT
15 were not elevated in any exposed mouse group, compared with control values. Results of
16 urinalysis and serum chemistry did not indicate clear evidence of bromobenzene-induced effects
17 on the renal system. Hematological results were generally unremarkable.
18 As shown in Table 4-4, histopathologic examination revealed statistically significant
19 effects on the liver that included cytomegaly in male and female mice at doses >200 mg/kg-day,
20 necrosis and mineralization in male mice at doses of 400 and 600 mg/kg-day, and necrosis and
21 inflammation in female mice at the 600 mg/kg-day dose level. The severity of these responses
22 was generally greater in males than females. Cytomegaly was the most common response seen
23 in the livers of treated mice and was characterized by an increase in the size of the nucleus and
24 cytoplasm of individual hepatocytes. Liver necrosis was primarily coagulative in nature and was
25 considered to be a direct result of bromobenzene treatment. Cytomegaly, inflammation, and
26 necrosis occurred primarily in the central part of the hepatic lobules. Significantly increased
27 incidences of hepatocellular necrosis or inflammation were observed only at doses equal to or
28 greater than those eliciting significantly increased incidences of cytomegaly. The study authors
29 considered inflammation and mineralization to be direct responses to hepatocellular necrosis.
30 Based on these observations, incidences of mice with one or more of these liver lesions
31 (cytomegaly, necrosis, inflammation, mineralization) were combined for each sex (as shown in
32 Table 4-4).
33 Treatment-related statistically significantly increased incidences of renal lesions (casts,
34 tubular degeneration without evidence of regeneration) were observed only in high-dose (600
6/7/07 20 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-3. Effects of bromobenzene on terminal body and liver weights and levels of selected serum liver enzymes of male
and female B6C3F1 mice exposed by oral gavage 5 days/week for 90 days in the basic study (mean +/- standard deviation)
Male mice
Dose (mg/kg-day)
Number of mice
Body weight (g)
Liver weight (g)
Difference (%)b
Ratio liver/body
weight
Difference (%)b
Serum AST (IU/L)
Serum ALT (IU/L)
Serum SDH (IU/L)
Controls
9
31.4 + 2.5
1.05 + 0.14
33.4 + 2.41
100 + 33.3
144 + 86.0
25 + 2.5
50
9
33.3 + 2.5
1.13+0.15
+7.6
33.9 + 3.52
+1.5
90 + 25.5
57a + 27.5
27 + 3.1
100
10
31.1 + 3.1
1.12 + 0.12
+6.7
36.0a+ 1.91
+7.8
80+11.6
80 + 43.0
27 + 3.2
200
10
33.4 + 3.5
1.25a + 0.22
+19.1
37.3a + 4.48
+11.7
88 + 23.2
102 + 61.5
41a+ 19.3
400
9
28.0a + 2.0
1.27a + 0.11
+21.0
45.3a+ 1.83
+35.6
99+17.2
132 + 41.0
89a + 28.3
600
5
30.5 + 2.5
1.56a + 0.16
+48.6
51.2a + 3.48
+53.3
70 + 8.8
115 + 35.8
101a + 29.0
Female mice
Dose (mg/kg-day)
Number of mice
Body weight (g)
Liver weight (g)
Difference (%)b
Ratio liver/body
weight
Difference (%)b
Serum AST (IU/L)
Serum ALT (IU/L)
Serum SDH (IU/L)
Controls
10
22.7+ 1.3
0.86 + 0.06
38.1 + 1.42
130 + 72.0
64 + 43.5
13 + 1.9
50
9
23.8+ 1.1
0.96a + 0.08
+11.6
40.2a + 2.02
+5.5
94 + 27.7
39+18.5
12+ 1.6
100
9
23.7+ 1.2
1.01a + 0.08
+17.4
42.5a+ 1.62
+11.6
101+21.4
51+28.9
14+ 1.8
200
10
24.3a+ 1.0
1.08a + 0.06
+25.6
44.4a + 2.12
+16.5
83 + 11.3
62 + 21.3
15a+1.7
400
8
23.4 + 0.6
1.12a + 0.07
+30.2
48.0a + 2.13
+26.0
91 + 18.4
73 + 31.2
23a + 4.6
600
10
23.6 + 0.8
1.30a + 0.06
+51.2
55.2a + 2.56
+44.9
123 + 55.4
126 + 79.0
43a+ 18.8
aStatistically significantly increased from controls (p<0.05) based on Student's two-tailed t-test
bChange relative to controls
2
o
5
4
5
Source: NTP (1985b)
6/7/07
21
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-4. Incidences of male and female B6C3F1 mice with liver and kidney lesions
following administration of bromobenzene by gavage 5 days/week for 90 days in the basic
study
Endpoint
Dose (mg/kg-day)a
0
Incidence
£
O)
4*
O5
50
Incidence
£
O)
4*
O5
100
Incidence
£
O)
4>
O5
200
Incidence
£
O)
4>
O5
400
Incidence
£
O)
4>
O5
600"
Incidence
£
O)
4>
O5
Males
Liver,
centrilobular
Inflammation
Cytomegaly
Necrosis
Mineralization
Combined0
Kidney, tubule
Degeneration
Casts
Mineralization
1/10
0/10
0/10
0/10
1/10
0/10
0/10
0/10
1.0
0/10
0/10
0/10
0/10
0/10
NE
NE
NE
1/10
1/10
0/10
0/10
2/10
NE
NE
NE
1.0
1.0
0/10
6/1 Od
1/10
0/10
6/1 Od
1/10
0/10
0/10
1.2
1.0
1.0
4/10
4/1 Od
4/1 Od
8/1 Od
10/10d
1/10
1/10
0/10
2.0
1.5
2.5
2.9
2.0
1.0
3/10
4/1 Od
8/1 Od
4/1 Od
10/10d
5/10d
5/10d
0/10
1.7
2.3
3.5
3.8
2.2
2.0
Females
Liver,
centrilobular
Inflammation
Cytomegaly
Necrosis
Mineralization
Combined0
Kidney, tubule
Degeneration
Casts
Mineralization
0/10
0/10
0/10
0/10
0/10
0/10
0/10
0/10
1/10
0/10
0/10
0/10
1/10
NE
NE
NE
1.0
0/10
1/10
1/10
0/10
2/10
NE
NE
NE
1.0
2.0
2/10
5/10d
0/10
0/10
6/1 Od
0/10
0/10
0/10
1.0
1.0
3/10
9/1 Od
1/10
0/10
9/1 Od
0/10
0/10
0/10
1.0
1.8
2.0
9/1 Od
10/10d
7/1 Od
2/10
10/10d
2/10
2/10
1/10
1.6
3.0
1.6
1.5
2 Incidence = number of animals in which lesion was found/number of animals in which organ
3 was examined.
4 bCytomegaly and mineralization were not diagnosed in 5 high-dose male mice that died on
5 treatment day 1
6 clncidences of mice with one or more of the liver lesion types (cytomegaly, necrosis,
7 inflammation, mineralization), extracted from individual animal histopathologic results provided
8 to Syracuse Research Corporation by NTP.
9 dStatistically significantly different from control groups according to Fisher's exact test (p<0.05),
10 performed by Syracuse Research Corporation.
11 Severity: Average severity score: 1 = minimal, 2 = mild, 3 = moderate, 4 = severe.
12 NE = Not examined.
13 Source: NTP (1985b)
14
6/7/07
22
DRAFT - DO NOT CITE OR QUOTE
-------
1 mg/kg-day) males. Sporadic lesions in other organs were not considered meaningful by the NTP
2 Pathology Working Group (NTP, 1986a). There was no report of bromobenzene-induced gross
3 or histopathological effects on reproductive tissues of male or female mice.
4 The most prominent toxicological effects observed in B6C3F1 mice treated with
5 bromobenzene by oral gavage for 90 days (NTP, 1985b) were observed in the liver.
6 Significantly increased incidences of hepatocellular necrosis (a clear indicator of an adverse
7 effect) were observed at doses of 400 and 600 mg/kg-day in male mice and the 600 mg/kg-day
8 dose level in female mice. Significantly increased incidences of cytomegaly were noted at doses
9 >200 mg/kg-day in male and female mice. Significant increases in mean liver weight were
10 observed at doses as low as 50 mg/kg-day in female mice and 100 mg/kg-day in male mice.
11 Treatment-related increased occurrence of cytomegaly (i.e., hypertrophy) and increased liver
12 weight may provide indication of liver toxicity from higher levels of exposure, but the
13 toxicological significance of these effects by themselves is questionable.
14 In the NTP (1985b) study the LOAEL is considered to be 50 mg/kg-day in female mice
15 for statistically significant increased absolute liver weight and increased liver-to-body weight
16 ratios. The designation of increased absolute liver weight and increased liver-to-body weight
17 ratios as an adverse effect is supported by the presence of liver lesions (including inflammation,
18 cytomegaly and necrosis) and statistically significantly increased SDH values at higher dose
19 levels. The increased serum enzyme (SDH) levels are indicative of liver damage.
20 Popper et al. (1952) investigated the hepatotoxic effects of subchronic dietary
21 bromobenzene exposure in rats. Control (n=9) and test (n=8) groups of female Wistar rats were
22 fed for 8 weeks on a synthetic diet that, in the test group, was supplemented with 5% (50,000
23 ppm) bromobenzene [corresponding to a dose of approximately 5130 mg/kg-day, calculated
24 using reference values for food consumption and body weight from U.S. EPA (1988)].
25 Histologic examination of the liver revealed mild changes, including distortion of the liver cell
26 plates and clumping and hydropic swelling in the cytoplasm of peripheral zone hepatocytes.
27 Alkaline phosphatase activity was markedly increased in both the liver and the serum. In
28 addition, liver and serum esterase levels were significantly decreased and serum xanthine
29 oxidase activity was increased (albeit not significantly). No other endpoints were monitored.
30
31 4.2.1.2. Chronic Toxicity
32 No studies were located on health effects in animals following chronic oral exposure to
33 bromobenzene.
34
6/7/07 23 DRAFT - DO NOT CITE OR QUOTE
-------
1 4.2.2. Inhalation Exposure
2 4.2.2.1. Subchronic Toxicity
3 NTP conducted subchronic inhalation studies of bromobenzene in rats (NTP, 1985c) and
4 mice (NTP, 1985d). These studies have not been officially released by NTP, but unpublished
5 reports, including the review comments and conclusions of NTP's Pathology Working Group
6 (NTP, 1986b), were obtained from NTP. The unpublished NTP studies are available by calling
7 EPA's IRIS Hotline at (202)566-1676, by fax at (202)566-1749 or by email at iris@epa.gov.
8 Groups of 10 male and 10 female Fischer 344/N rats were exposed to bromobenzene
9 vapors through whole body exposure at 0, 10, 30, 100, or 300 ppm (0, 64.2, 192.6, 642, or 1926
10 mg/m3) for 6 hours/day, 5 days/week for 13 weeks. Rats were observed twice daily for
11 morbidity and mortality. Clinical observations and body weight measurements were performed
12 weekly. Blood samples for hematologic examination (erythrocyte and leukocyte counts;
13 hemoglobin concentrations; red blood cell indices of MCV, MCH, and MCHC; leukocyte
14 differential counts) were collected from all surviving rats at terminal sacrifice. Terminal body
15 and organ (liver, brain, testis, kidney, lung, heart, and thymus) weights were recorded; organ-to-
16 body weight and organ-to-brain weight ratios were calculated for each sex. Complete gross
17 necropsy was performed on all rats. Complete histopathologic examinations of all major tissues
18 and organs (including liver, kidney, urinary bladder, spleen, pancreas, brain, spinal cord [if
19 neurologic signs were present], heart, lung, trachea, nasal cavity, larynx, esophagus, stomach,
20 small intestine, cecum, colon, skin, uterus, ovaries, preputial or clitoral glands, testes, prostate,
21 sternebrae, adrenals, pituitary, thyroid, parathyroids, salivary gland, mandibular lymph node,
22 thymus, mammary gland, blood, and gross lesions, and tissue masses) were performed on all
23 control rats and all rats from the 300-ppm groups. Kidney tissue was examined
24 histopathologically in all male rats of the lower exposure concentrations (10, 30, and 100 ppm).
25 No mortality was observed during the study. Clinical signs were unremarkable except for
26 tearing, facial grooming, and listlessness in 300-ppm rats on the first day of exposure. Terminal
27 body weights did not differ significantly from controls. Liver and kidney weights (absolute,
28 relative-to-body weight, and relative-to-brain weight) were significantly increased at
29 concentrations >100 ppm in both sexes. Liver and kidney weight data are reported in Table 4-5.
30 In males, absolute liver weights increased 13% at 100 ppm and 20% at 300 ppm. In females,
31 absolute liver weights increased 12% at 100 ppm and 22% at 300 ppm. MCH and MCV were
32 statistically significantly decreased in males at concentrations >10 ppm and in females at 300
33 ppm, but the changes were small and considered not to be biologically significant. There was no
34 histopathological evidence of bromobenzene-induced liver lesions, although livers were
35 examined only from control rats and rats of the highest exposure level (100 ppm in males and
36 300 ppm in females) (see Table 4-6).
6/7/07 24 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-5. Effects of bromobenzene on terminal body, liver, and kidney weights of male and female rats exposed by
inhalation 6 hours/day, 5 days/week for 13 weeks (mean +/- standard deviation)
Male rats
Exposure concentration (ppm)
Number of rats
Body weight (g)
Liver weight (g)
Difference (%)a
Ratio liver/body weight x 1000
Difference (%)a
Right kidney weight
Ratio right kidney /body weight x 1000
Difference (%)a
Controls
10
318 + 15.5
11.58+1.18
33.37 + 2.86
0.98 + 0.06
3.09+0.06
10
10
322.9 + 14.2
12.04 + 0.4
+ 4%
37.31 + 1.96
+ 10.5%
1.04 + 0.05
3.22 + 0.17
+ 4%
30
10
331.1 + 18.2
12.13 + 0.77
+ 5%
36.68 + 2.05
+ 9%
1.87 + 0.05
3.16 + 0.16
+ 2%
100
10
312.4 + 39.1
13.13b+1.59
+ 13%
42.11C + 2.09
+ 21%
1.07b + 0.11
3.43C + 0.19
+ 10%
300
10
309.4+ 18.3
14.33C+ 1.67
+ 20%
46.26c + 3.86
+ 28%
l.llc + 0.09
3.60C + 0.11
+ 14%
Female rats
Exposure concentration (ppm)
Number of rats
Body weight (g)
Liver weight (g)
Difference (%)a
Ratio liver/body weight x 1000
Difference (%)a
Right kidney weight
Ratio kidney /body weight x 1000
Difference (%)a
Controls
10
186.0+ 11.2
6.36 + 0.65
34.12+1.83
0.62 + 0.05
3.31+0.21
10
10
191.4 + 10.5
6.71 + 0.55
+ 7%
35.05 + 1.82
+ 3%
0.65 + 0.03
3.39 + 0.09
+ 2%
30
10
182.8 + 9.1
6.52 + 0.60
+ 4%
35.68 + 2.84
+ 4%
0.66 + 0.06
3.62b + 0.26
+ 9%
100
10
187.7 + 8.3
7.23C + 0.30
+ 12%
38.56C+ 1.62
12%
0.66b + 0.03
3.53b + 0.18
+ 6%
300
10
189.9+ 11.6
8.22C + 0.63
+ 23%
43.54c + 2.53
22%
0.70C + 0.05
3.73C + 0.16
+ 11%
2 aChange relative to controls
3 bStatistically significantly different from controls (p<0.05) based on student's two-tailed t-test
4 cOutside 3 standard deviations from the control mean
5 Source: NTP (1985d)
6/7/07
25
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-6. Incidences of male and female Fischer 344/N rats with liver and kidney
lesions following repeated exposure to bromobenzene vapors for 13 weeks
Endpoint
Exposure concentration*
0
Incidence
>>
+*
O)
4*
O5
10
Incidence
>>
+*
O)
4*
O5
30
Incidence
>>
+*
O)
4>
O5
100
Incidence
>>
+*
O)
4>
O5
300
Incidence
>>
+*
O)
O)
O5
Males
Liver
Necrosis
Inflammation
Kidney, tubule
Regeneration
1/10
0/10
10/10
1.0
1.0
NE
10/10
1.0
NE
9/10
1.0
NE
10/10
0.9
0/10
0/10
10/10
1.9
Females
Liver
Necrosis
Inflammation
Kidney, tubule
Regeneration
1/10
2/10
0/10
1
1
NE
NE
NE
NE
NE
NE
0/10
3/10
0/10
1
2
o
J
4
5
6
7
*Incidence = number of animals in which lesion was found/number of animals in which organ
was examined
Severity: 1 = minimal, 2 = mild, 3 = moderate, 4 = severe. NE = Not examined.
Source: NTP (1985c)
6/7/07
26
DRAFT - DO NOT CITE OR QUOTE
-------
1 Histopathologic examination of the kidneys revealed renal cortical tubular regeneration,
2 characterized by basophilic (regenerative) tubules scattered throughout the renal cortex of all
3 control and bromobenzene-exposed male rats (with the exception of a single male in the 30-ppm
4 exposure group; see Table 4-6). The renal tubular regeneration was observed in the absence of
5 convincing evidence of degeneration or necrosis. NTP (1985c) noted that the severity of
6 nephropathy in 300-ppm males could be distinguished from that of controls in blind evaluations.
7 These findings were confirmed upon re-examination of kidney tissues from control and 300-ppm
8 male mice by the Pathology Working Group (NTP, 1986b). The Working Group considered the
9 effect to be mild and not life threatening.
10 Gross and histopathologic examinations of reproductive tissues of male and female rats
11 did not reveal evidence of bromobenzene-induced effects. No significant treatment-related
12 lesions were found in gross or histopathologic examinations of other tissues in female rats.
13 Since increased liver weight at the 100 ppm and 300 ppm dose groups in the NTP
14 (1985c) study were not accompanied by bromobenzene induced liver lesions these effects were
15 considered to be of questionable toxicological significance and not considered to be a LOAEL;
16 therefore, the highest dose level tested (300 ppm) is considered to be a NOAEL in this study.
17 In the mouse study, groups of 10 male and 10 female B6C3F1 mice were exposed to 0,
18 10, 30, 100, or 300 ppm (females only) and (0, 64.2, 192.6, 642, or 1926 mg/m3) of
19 bromobenzene 6 hours/day, 5 days/week for 13 weeks (NTP, 1985d). No rationale for excluding
20 a 300-ppm exposure level for the male mice was included in the available study report. Clinical
21 observations and body weight measurements were performed weekly. Blood samples for
22 hematologic examination (erythrocyte and leukocyte counts; hemoglobin concentrations; red
23 blood cell indices of MCV, MCH, and MCHC; leukocyte differential counts) were collected
24 from all surviving mice at terminal sacrifice. Terminal body and organ (liver, brain, testis,
25 kidney, lung, heart, and thymus) weights were recorded; organ-to-body weight and organ-to-
26 brain weight ratios were calculated for each sex. Complete gross necropsy was performed on all
27 mice. Histopathologic examinations of all major tissues and organs (including liver, kidney,
28 urinary bladder, spleen, pancreas, gall bladder, brain, spinal cord [if neurologic signs were
29 present], heart, lung, trachea, nasal cavity, larynx, esophagus, stomach, small intestine, cecum,
30 colon, skin, uterus, ovaries, preputial or clitoral glands, testes, prostate, sternebrae, adrenals,
31 pituitary, thyroid, parathyroids, salivary gland, mandibular lymph node, thymus, mammary
32 gland, blood, gross lesions, and tissue masses) were performed on all control, 100-ppm male and
33 300-ppm female mice. Liver and kidney tissues were examined histopathologically in all other
34 groups of mice.
35 There were no deaths during this study and no clinical signs of toxicity were observed.
36 Terminal body weights of treated groups did not differ significantly from controls. In female
6/7/07 27 DRAFT - DO NOT CITE OR QUOTE
-------
1 mice, liver weights (absolute, relative to body weight, relative to brain weight) were statistically
2 significantly increased in an exposure concentration-related manner. Absolute liver weights
3 were increased approximately 8, 17, and 66% at 30, 100, and 300 ppm, respectively. Liver-to-
4 body weight ratios were increased approximately 6, 5, 14, and 53% at 10, 30, 100, and 300 ppm,
5 respectively. Smaller increases in these parameters were also seen in 100-ppm males. Liver and
6 kidney weight data are reported in Table 4-7. Sporadic changes in hematology parameters,
7 observed in male and female mice of most exposure groups, were not considered to be
8 biologically significant. Females of the 300 ppm exposure level exhibited enlarged, diffusely
9 mottled livers.
10 Incidences of histopathologic liver lesions are summarized in Table 4-8. In the original
11 study report, histopathologic evidence of hepatic effects was presented. Cytomegaly was
12 diagnosed in the liver of 4/10 and 2/10 male mice of the 30- and 100-ppm exposure groups,
13 respectively, as well as 2/10 and 10/10 female mice of the respective 100- and 300-ppm exposure
14 groups. The Pathology Working Group agreed with the diagnoses of cytomegaly, hepatic
15 necrosis, and mineralization in the 300-ppm female mice, but did not consider observed liver
16 effects to be adverse in female mice at lower exposure levels (NTP, 1986b). Furthermore, the
17 Pathology Working Group considered the reported cytomegaly in 100-ppm male mice to be
18 more appropriately described as centrilobular hepatocellular hypertrophy or enlargement and to
19 be less severe than cytomegaly observed in the female mice (NTP, 1986b). The associated effect
20 in 30-ppm males was not considered by the Pathology Working Group to be indicative of
21 centrilobular hypertrophy, but it was noted that some increased eosinophilic staining of
22 centrilobular hepatocytes suggested an effect typical of hepatocellular enzyme induction.
23 The NTP study report (NTP, 1985d) also presented histopathological evidence for renal
24 lesions (see Table 4-8). The kidneys of 2/10 and 3/10 of the 30- and 100-ppm male mice
25 exhibited evidence of minimal tubular degeneration, but the Pathology Working Group did not
26 consider this finding to represent an adverse effect since it may have been the result of artifacts
27 of fixation and staining procedures (NTP, 1986b). Gross and histopathologic examinations of
28 reproductive tissues of male and female mice did not reveal evidence of bromobenzene-induced
29 effects.
30 In the NTP (1985d) inhalation study in mice, the highest dose tested, 300 ppm, is
31 considered to be a LOAEL (Lowest Observed Adverse Effect Level). The 100 ppm dose is
32 considered to be a NOAEL because the increases in absolute liver weight and increases in
33 cytomegally were not considered to be adverse by the Pathology Working Group at exposure
34 levels below 300 ppm. Treatment-related significantly increased liver weights were seen in all
35 exposure groups of female mice, and a significantly increased incidence of cytomegaly was
36 observed in the 300 ppm female mice. Treatment-related increased occurrence of cytomegaly
6/7/07 28 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-7. Effects of bromobenzene on terminal body, liver, and kidney weights of male and female mice exposed by
inhalation 6 hours/day, 5 days/week for 13 weeks (mean +/- standard deviation)
Male mice
Exposure concentration (ppm)
Number of mice
Body weight (g)
Liver weight (g)
Difference (%)a
Ratio liver/body weight x 1000
Difference (%)a
Right kidney weight
Ratio right kidney /body weight x 1000
Difference (%)a
Controls
10
36.3 + 3.6
1.84 + 0.21
50.71 + 3.66
0.29 + 0.02
8.13 + 0.66
10
10
33.4 + 2.0
1.73 + 0.14
-6.0
51.86 + 3.57
+2.2
0.30 + 0.03
8.84 + 0.86
+8.7
30
10
33.6 + 3.0
1.73+0.18
-6.0
51.57 + 2.78
+1.7
0.30 + 0.02
8.88b + 0.75
+9.2
100
10
34.4 + 3.2
1.87 + 0.21
+1.6
54.28b + 2.42
+7.0
0.30 + 0.02
8.78 + 0.90
+8.0
300
Female mice
Exposure concentration (ppm)
Number of mice
Body weight (g)
Liver weight (g)
Difference (%)a
Ratio liver/body weight x 1000
Difference (%)a
Right kidney weight
Ratio kidney /body weight x 1000
Difference (%)a
Controls
10
27.4+ 1.4
1.43 + 0.15
52.0 + 3.22
0.19 + 0.01
6.80 + 0.28
10
10
27.5 + 1.3
1.52 + 0.09
+6.3
55.25b + 3.49
+6.3
0.20C + 0.01
7.38C + 0.25
+8.5
30
10
28.3 + 1.7
1.54b + 0.07
+7.7
54.66b + 1.80
+5.1
0.20 + 0.02
7.04 + 0.51
+3.5
100
10
28.3 + 0.9
1.68C + 0.10
+17.5
59.37c + 3.43
+14.2
0.20C + 0.01
7.14 + 0.32
+5.0
300
10
29.7C+ 1.7
2.37C + 0.21
+65.7
79.73c + 5.27
+53.3
0.23C + 0.02
7.64 + 0.45
+12.4
2
3
4
5
6
aChange relative to controls
bStatistically significantly different from controls (p<0.05) based on Student's two-tailed t-test
°Outside 3 standard deviations from the control mean
Source: NTP (1985d)
6/7/07
29
DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-8. Incidences of male and female B6C3F1 mice with liver and kidney lesions
following repeated exposure to bromobenzene vapors for 13 weeks
Endpoint
Exposure concentration3
0
Incidence
>*
•^
'•Z
4*
4*
O5
10
Incidence
>*
+rf
'•Z
4>
4>
O5
30
Incidence
>*
•^
'Z
4>
4>
O5
100
Incidence
>*
•^
'Z
4>
4>
O5
300
Incidence
>*
•^
'Z
4>
4>
O5
Males
Liver
Cytomegalyb
Necrosis
Inflammation
Kidney, tubule
Degenerationd
0/10
0/10
1/10
0/10
3.0
0/10
0/10
0/10
0/10
4/1 Oc
0/10
0/10
2/10
2.0
1.5
2/10
2/10
4/10
3/10
1.5
1.0
1.8
2.0
NG
NG
Females
Liver
Cytomegaly
Necrosis
Inflammation
Mineralization6
Kidney*
0/10
2/10
4/10
0/10
1.0
1.5
0/10
1/10
3/10
0/10
1.0
1.3
0/10
0/10
2/10
0/10
1.0
2/10
2/10
2/10
0/10
1.0
1.0
1.5
10/10C
5/10
2/10
2/10
3.2
1.3
1.3
2.0
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
Incidence = number of animals in which lesion was found/number of animals in which organ
was examined. Severity: 1 = minimal, 2 = mild, 3 = moderate, 4 = severe. NG = No group (the
study did not include a 300 ppm exposure group of male mice)
b The Pathology Working Group (NTP, 1986b) considered this diagnosis in 100-ppm male mice
to be more appropriately described as centrilobular hepatocellular hypertrophy or enlargement
and the results in 30-ppm male mice to be suggestive of hepatocellular enzyme induction, rather
than cytomegaly as noted in female mice.
GStatistically significantly different from control groups according to Fisher's exact test (p<0.05),
performed by Syracuse Research Corporation.
dKidney tubular degeneration could not be distinguished from artifacts of fixation or staining.
eMineralization was not reported in male mice.
fNo histopathologic renal lesions were identified in any group of female mice.
Source: NTP (1985d)
6/7/07
30
DRAFT - DO NOT CITE OR QUOTE
-------
1 and increased liver weight may provide an early indication of liver toxicity from higher level
2 exposure. Hepatocyte necrosis was noted in 5/10 of the 300-ppm female mice, but the incidence
3 of this lesion was not significantly greater than the incidence in controls (2/10). The 300-ppm
4 exposure level may represent an effect level in female mice that is near the threshold for
5 bromobenzene hepatotoxicity.
6 Shamilov (1969) exposed rats to 3 or 20 |J,g/m3 of bromobenzene 4 hours daily for 140
7 days. At 20 ng/m3, bromobenzene gradually accumulated in the tissues, producing decreases in
8 body growth, liver sulfhydryl concentration, serum protein levels and leukocyte, platelet, and
9 reticulocyte counts as well as neurological disorders. No effects were seen at 3 ng/rn3. More
10 detailed study information was not presented in the available abstract thus precluding critical
11 assessment of the study.
12
13 4.2.2.2. Chronic Toxicity
14 No studies were located on health effects in animals following chronic inhalation
15 exposure to bromobenzene.
16
17 4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES—ORAL AND INHALATION
18 4.3.1. Reproductive Toxicity Studies
19 No reproductive toxicity studies were located for bromobenzene.
20
21 4.3.2. Developmental Toxicity Studies
22 No developmental toxicity studies were located for bromobenzene.
23
24 4.4. OTHER STUDIES
25 4.4.1. Acute Toxicity Studies
26 The toxic effects of bromobenzene following acute exposure have been extensively
27 studied. Liver, kidney, and lung have been identified as the target organs for this chemical by a
28 variety of routes. Histopathologic examinations have revealed necrotic changes in all of these
29 organs following short-term bromobenzene exposure (Szymahska and Piotrowski, 2000;
30 Szymahska, 1998; Becher et al., 1989; Casini et al., 1986; Forkert, 1985; Rush et al., 1984;
31 Kluwe et al., 1984; Roth, 1981; Reid et al., 1973; Patrick and Kennedy, 1964).
32 The liver is the most sensitive target following acute oral exposure. In rats given single
33 oral doses of bromobenzene by gavage, a dose of 39 mg/kg resulted in reduced hepatic
34 glutathione; a higher dose (157 mg/kg-day) resulted in moderate periportal and midzonal
35 hydropic changes, while increased serum liver enzyme levels and hepatic centrilobular necrosis
36 were observed following dosing at 314 mg/kg-day (Kluwe et al., 1984). In the same study, renal
6/7/07 31 DRAFT - DO NOT CITE OR QUOTE
-------
1 glutathione was reduced at a dose of 157 mg/kg-day, but no other renal effects were noted at
2 doses up to 628 mg/kg-day. Other acute oral studies reported hepatic necrosis in rats (Heijne et
3 al., 2004) or mice (Patrick and Kennedy, 1964) administered bromobenzene at doses in the range
4 of 500-700 mg/kg; reduced renal glutathione levels, increased BUN levels, and severe tubular
5 necrosis in mice given 2355 mg/kg-day (Casini et al., 1986); extensive vacuolization and
6 necrosis in Clara cells in the lungs of mice given 785 mg/kg-day (Forkert, 1985); and increased
7 LDH levels in lung lavage fluid accompanied by bronchiolar damage in the lungs of mice given
8 2355 mg/kg-day (Casini et al., 1986).
9 When rats were exposed to a bromobenzene vapor concentration of 107 ppm for 4 hours,
10 serum liver enzyme changes were noted (Brondeau et al., 1983). Extrahepatic effects observed
11 in other acute inhalation studies included pulmonary effects, seen as moderate vacuolization of
12 pulmonary Clara cells in mice exposed to 250 ppm for 4 hours (Becher et al., 1989) and
13 pulmonary necrosis in mice exposed to 1000 ppm for 4 hours (Becher et al., 1989).
14
15 4.4.2. Genotoxicity Studies
16 Table 4-9 summarizes available results of genotoxicity tests for bromobenzene. Results
17 of gene mutation assay systems did not indicate a mutagenic response in several strains of
18 Salmonella typhimurium at bromobenzene concentrations as high as 500 |j,g/plate with or without
19 S-9 activation (Nakamura et al., 1987; Rosenkranz and Poirier, 1979; Simmon, 1979; Simmon et
20 al., 1979; McCann et al., 1975). Bromobenzene was not mutagenic in an in vivo test for
21 nondisjunction in Drosophila (Ramel and Magnusson, 1979). Bromobenzene did not induce
22 sister chromatid exchanges in Chinese hamster ovary cells (Galloway et al., 1987) or cell
23 transformation in Syrian hamster embryo cells (Pienta et al., 1977). A weakly positive result was
24 reported for bromobenzene-induced chromosomal aberrations in Chinese hamster ovary cells in
25 the absence, but not the presence, of metabolic S-9 activation (Galloway et al., 1987).
26 Bromobenzene was observed to increase formation of micronucleated erythrocytes, in
27 femoral polychromatic mouse bone marrow cells in vivo (Mohtashamipur et al., 1987) and
28 actively bind to rat and mouse DNA, RNA, and proteins both in vivo and in vitro (Prodi et al.,
29 1986; Colacci et al., 1985). Following intraperitoneal injection of 14C-bromobenzene (6.35
30 |imol/kg; lower than a minimally hepatotoxic dose) in rats and mice, the degree of binding in
31 liver, kidney, and lung tissues of both species was RNA > proteins > DNA (Colacci et al., 1985).
32 Mouse kidney exhibited a much greater degree of binding to macromolecules than rat kidney. In
33 both rats and mice, the relative order of binding to DNA in the various organs was liver > kidney
34 > lung. Bromobenzene was second only to 1,2-dibromoethane in its relative in vivo reactivity
35 with rat liver DNA, exhibiting higher reactivity than 1,2-dichloroethane, chlorobenzene,
36 epichlorohydrin, and benzene (Prodi et al., 1986). Microsomal enzyme-catalyzed the in vitro
6/7/07 32 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-9. Results of bromobenzene genotoxicity testing
Assay and test system
Reverse mutation in S. typhimurium
strains TA1535, TA1537, TA98,
TA100
Reverse mutation in S. typhimurium
strains TA1535, TA1538
Reverse mutation in S. typhimurium
strains TA1535, TA1536, TA1537,
TA1538, TA98, TA100
Reverse mutation in S. typhimurium
strains TA1530, TA1538 (host-
mediated assay using mice)
Reverse mutation in S. typhimurium
strains TA1535, TA1538 (host-
mediated assay using mice)
SOS-response in S. typhimurium
strain TA1535/pSK1002
Nondisj unction in Drosophila
Sister chromatid exchanges in
Chinese hamster ovary cells (CHO-
W-B1)
Cell transformation in Syrian
hamster embryo cells
Chromosomal aberrations in Chinese
hamster ovary cells (CHO-W-B1)
Micronuclei in mouse (NMRI)
bone marrow cells
DNA binding in rat and mouse (in
vivo)
RNA binding in rat and mouse (in
vivo)
Dose/
concentration
NS
+ S9 activation
10 |o,g/plate
+ S9 activation
250 jog/plate
+ S9 activation
600 mg/kg-day
lOOOmg/kg-day
Up to 500 ng/mL
+ S9 activation
lOOOppm
50-500 ng/mL
+ S9 activation
0.0001-0.5 ng/mL
50-500 ng/mL
+ S9 activation
125-500 mg/kg-day
(2x62.5-2x250
doses 24 hours apart)
6.35 |o,mol/kg
(intraperitoneal)
6.35 |o,mol/kg
(intraperitoneal)
HID or
LED*
NS
10
250
600
1000
500
1000
500
0.5
500
125
6.35
6.35
Result
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Negative
Weakly
positive
-S9,
negative
+S9
Positive
Positive,
rat and
mouse
liver,
mouse
kidney
Positive,
rat and
mouse
liver,
kidney,
and lung
Reference
McCann et al.,
1975
Rosenkranz and
Poirier, 1979
Simmon, 1979
Simmon et al.,
1979
Simmon et al.,
1979
Nakamura et al.,
1987
Ramel and
Magnusson, 1979
Galloway et al.,
1987
Pientaetal., 1977
Galloway et al.,
1987
Mohtashamipur et
al., 1987
Colacci et al.,
1985; Prodi etal.,
1986
Colacci et al.,
1985; Prodi etal.,
1986
2
o
J
4
*HID, highest ineffective dose/concentration for negative tests; LED, lowest effective dose/concentration
for positive tests; NS, not stated
6/7/07
33
DRAFT - DO NOT CITE OR QUOTE
-------
1 binding of 14C-bromobenzene to rat and liver DNA; liver microsomes of mice appeared to be
2 slightly more efficient than those of rats (Colacci et al., 1985). The degree of in vitro binding in
3 liver, kidney, and lung tissues of both species was RNA > proteins > DNA. In both rat and
4 mouse microsomal preparations, the relative order of binding to macromolecules was liver >
5 lung > kidney.
6 Reactive metabolites of bromobenzene are produced in vivo as discussed in Section 3.3
7 and could be expected to interact with DNA. The central pathway for the mammalian
8 metabolism of bromobenzene appears to be the production of bromocatechols via bromophenols,
9 as depicted in Figure 3-1 (Lertratanangkoon et al., 1993; Lau and Monks, 1988). Although
10 reactive metabolites, 2,3-oxide and 3,4-oxide, are formed as precursors in the predominant
11 pathway in bromobenzene's metabolism, 2,3-oxide has a very short biological half-life,
12 indicating spontaneous rearrangement to the formation of 2-bromophenol in the rat and pig
13 (Lertratanangkoon et al., 1993). Another reactive intermediate, 2,3-dihydrodiol, also rapidly
14 rearranges to form both 2-bromophenol and 3-bromophenol in the detoxification bromocatechol
15 pathway (Lertratanangkoon et al., 1987). Furthermore, spontaneous rearrangement of the
16 3,4-dihydrodiol is considered to be the major pathway in bromobenzene's metabolism, leading to
17 the formation of 4-bromophenol in the rat, while a pathway leading through an S-glutathione
18 conjugate to 4-bromophenol is predominant in the guinea pig (Lertratanangkoon et al., 1987,
19 1993). The bromophenols are subsequently oxidized by CYP to their respective bromocatechols
20 in a detoxification pathway (Miller et al., 1990; Lau and Monks, 1988). While these
21 toxicokinetic events are expected to elicit a toxicity response from liver tissue, the reactive
22 metabolites generated in the process may be too transient and reactive to elicit measurable
23 responses in Salmonella mutagenicity assays and other genotoxicity assays involving external rat
24 liver S-9 metabolic activation.
25 In conclusion, the available data from bacterial mutagenicity assays were predominately
26 negative however, clastogenic and mutagenic results in mammalian cell cultures and whole
27 animals studies were positive. Bromobenzene was not mutagenic in the Ames assay and did not
28 consistently produce marked cytogenic effects in vitro with mammalian cells, even in the
29 presence of rat liver S-9 preparations. Bromobenzene increased formation of micronucleated
30 polychromatic erythrocytes in bone marrow of mice given acute oral doses of 125 mg/kg and
31 was bound to DNA and RNA following intraperitoneal injection. Results of in vivo testing of
32 DNA binding in rat and mouse liver indicate that bromobenzene is greater than 20-fold more
33 reactive to rat liver DNA than benzene (Prodi et al., 1986), the nonhalogenated parental
34 compound known to be carcinogenic and considered a weak tumor initiator. Whereas the extent
35 of DNA binding was similar in other tissues examined such as lung and kidney. However,
6/7/07 34 DRAFT - DO NOT CITE OR QUOTE
-------
1 bromobenzene has not been tested in tumor initiation assays or long-term carcinogenicity
2 bioassays.
o
J
4 4.4.3. Tumor Promotion Studies
5 The potential for bromobenzene to promote diethylnitrosamine (DENA)-initiated rat liver
6 foci was investigated in two rat liver assays. Herren-Freund and Pereira (1986) dosed male and
7 female Sprague-Dawley rats by gavage (0.5 mmol/kg of DENA), followed by intraperitoneal
8 injection of bromobenzene (1.0 mmol/kg), 1 and 5 weeks after DENA administration. The rats
9 were sacrificed 2 weeks after the last injection of bromobenzene. Treatment with bromobenzene
10 did not enhance the occurrence of y-glutamyltranspeptidase-positive (GGT) foci in the liver. Ito
11 et al. (1988) administered a single intraperitoneal injection of DENA to male Fischer rats to
12 initiate hepatocarcinogenesis. Some of these rats were administered bromobenzene (15
13 mg/kg-day) by intraperitoneal injections (eight injections, initiated 2 weeks following DENA
14 treatment and ending before sacrifice at 8 weeks post-DENA administration). All rats were
15 subjected to 2/3 partial hepatectomy at 3 weeks to maximize any interaction between
16 proliferation and effects of test compound. The number and area per cm2 of induced glutathione
17 S-transferase placental form-positive (GST-P+) foci in the liver of bromobenzene-treated rats
18 was assessed and compared with those receiving DENA only. Bromobenzene treatment did not
19 result in statistically significant increases in the number or area per cm2 of DENA-induced GST-
20 P+ foci.
21
22 4.5. MECHANISTIC STUDIES
23 4.5.1. Mechanistic Studies of Liver Effects
24 As discussed in Sections 4.2 and 4.4, animal studies identify the liver as the most
25 sensitive toxicity target of oral or inhalation exposure to bromobenzene. As discussed in detail
26 below, the results of numerous mechanistic studies in animals collectively demonstrate that
27 bromobenzene hepatotoxicity is associated with metabolism of parent compound, cytotoxicity
28 may result from modifications of hepatocellular macromolecules by one or more reactive
29 metabolites, and that these reactive metabolites are formed primarily via the metabolic pathway
30 that involves the 3,4-oxide (rather than the 2,3-oxide) derivative of bromobenzene (see Slaughter
31 and Hanzlik, 1991; Monks et al., 1984a; Jollow et al., 1974; Mitchell et al., 1971).
32 Nephrotoxicity has also been observed in animals following acute-duration exposure to
33 bromobenzene, albeit at higher doses than the lowest hepatotoxic doses. Repeated-dose oral and
34 inhalation studies in rats and mice provide evidence for kidney effects, but only at the highest
35 exposure levels tested, which also resulted in lethality. Nephrotoxicity also appears to result
6/7/07 3 5 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
from modification of macromolecules in cells of the proximal convoluted tubule by one or more
reactive metabolites (Reid, 1973).
To demonstrate that hepatotoxic effects are elicited by metabolites of bromobenzene and
not bromobenzene itself, one group of rats was administered single intraperitoneal doses (1500
mg/kg-day) of bromobenzene, while another group was administered p-diethylaminoethyl
diphenylpropyl acetate (SKF 525 A, a CYP inhibitor) before and after administration of the same
intraperitoneal dose (1.5 mg/kg-day) of bromobenzene (Mitchell et al., 1971). As shown in
Table 4-10, extensive centrilobular necrosis was observed in the group of bromobenzene-treated
rats examined 24 hours following dosing. However, the CYP-inhibited rats exhibited no clear
signs of the liver lesion, although concentrations of parent compound in plasma and liver of the
CYP-inhibited rats were five to six times higher than those in the group not treated with the
CYP-inhibitor.
Table 4-10. The effect of CYP inhibition on the hepatotoxicity and metabolism of single
intraperitoneal doses of bromobenzene
Treatment
Bromobenzene
(1500 mg/kg-day)
Bromobenzene (1500
mg/kg-day) + SKF 525A
Severity of hepatic
centrilobular
necrosis
Extensive
No specific lesions
24-Hour bromobenzene concentration
Plasma (ng/mL)
2.8 + 0.3*
14.4 + 0.5
Liver (ng/g)
26 + 3
149 + 8
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
*Mean + standard error from 5-7 rats/group; CYP = cytochrome P-450 isozymes; SKF 525
P-diethylaminoethyl diphenylpropyl acetate
Source: Mitchell et al. (1971)
Chemically reactive metabolites of bromobenzene may damage cellular macromolecules,
leading to cytotoxicity. These metabolites include the 2,3- and 3,4-oxides of bromobenzene, the
oxides of the bromophenols, the 1,4-benzoquinone, and the radicals and quinones derived from
redox cycling of the 2- and 4-bromocatechols (Slaughter and Hanzlik, 1991; Lau and Monks,
1988). The 3,4-epoxide binds covalently to microsomal protein at the site of synthesis while the
2,3- epoxide binds to the soluble protein, i.e., hemoglobin P chain (Lau and Zannoni, 1981b).
The bromobenzene 3,4-oxide alkylates histidine and lysine side chains in rat liver proteins in
vivo (Bambal and Hanzlik, 1995). Phenolic metabolites of bromobenzene are activated to toxic
metabolites, which deplete cellular glutathione and have caused cell death in isolated hepatocytes
(Lau and Monks, 1997a). Hydroquinone metabolites of bromobenzene have been indicated as
subcellular targets of nephrotoxicity in the rat, causing changes in proximal tubular brush border,
6/7/07
36
DRAFT - DO NOT CITE OR QUOTE
-------
1 nuclei, and endoplasmic reticulum (Lau and Monks, 1997b). Slaughter et al. (1993)
2 demonstrated that bromobenzene-derived oxides, quinones, and bromoquinones are capable of
3 alkylating protein sulfhydryl groups, the major adduct arising from the 1,4-benzoquinone
4 electrophilic metabolite. Quinone-derived protein adducts appear to be formed to a greater
5 extent than those derived from the epoxides (Bambal and Hanzlik, 1995; Slaughter and Hanzlik,
6 1991). Several liver proteins have been identified as targets for reactive metabolites of
7 bromobenzene (Koen and Hanzlik, 2002; Koen et al., 2000; Rombach and Hanzlik, 1997, 1998,
8 1999; Aniya et al., 1988). While electrophilic metabolites of bromobenzene have the ability to
9 interact with tissue macromolecules, a causal role for this binding in hepatotoxicity has yet to be
10 demonstrated (Koen and Hanzlik, 2002; Lau and Monks, 1997a).
11 Results of mechanistic studies further indicate that hepatotoxicity is primarily elicited via
12 the metabolic pathway that involves the 3,4-oxide derivative of bromobenzene, and that the toxic
13 effect is likely mediated via covalent binding of one or more reactive metabolites with
14 hepatocellular macromolecules (Monks et al., 1984a; Jollow et al., 1974; Reid and Krishna,
15 1973; Zampaglione et al., 1973; Brodie et al., 1971). Supporting evidence includes the findings
16 that: (1) induction of p-naphthoflavone- or 3-methylcholanthrene-induced CYP isozymes
17 (possibly cytochrome P-488) results in increased urinary excretion of 2-bromophenol (formed
18 via the 2,3-oxide pathway) and decreased hepatotoxicity (Lau et al., 1980; Lau and Zannoni,
19 1979; Jollow et al., 1974; Zampaglione et al., 1973), whereas (2) induction of phenobarbital-
20 induced CYP isozymes results in increased urinary excretion of 4-bromophenol (formed via the
21 3,4-oxide pathway), as well as increases in both severity of hepatocellular necrosis and the extent
22 of covalent binding of radioactivity from 14C-bromobenzene to hepatocellular macromolecules in
23 the region of observed hepatocellular necrosis (Brodie et al., 1971).
24 The importance of glutathione conjugation as a protective mechanism for bromobenzene
25 acute hepatotoxicity was demonstrated in male Sprague-Dawley rats that were administered a
26 single intraperitoneal dose of 14C-bromobenzene (1570 mg/kg; 236 mg/kg in phenobarbital-
27 pretreated rats) (Jollow et al., 1974). Selected groups of these rats were additionally treated with
28 either phenobarbital (a known CYP inducer), SKF 525A (a known CYP inhibitor), diethyl
29 maleate (which depletes glutathione), or cysteine (a precursor of glutathione). Selected rats from
30 each group were periodically sacrificed during 48 hours following bromobenzene treatment in
31 order to determine rates of liver glutathione depletion. Bromobenzene metabolism was
32 associated with clearance of radioactivity from the whole body over time. All groups of rats
33 were assessed for the severity of centrilobular necrosis. Results are summarized in Table 4-11.
34 Bromobenzene treatment alone resulted in minimal signs of necrosis. In contrast, rats that had
35 been pretreated with phenobarbital exhibited massive necrotic areas, as well as statistically
36 significant (p<0.05) increases in bromobenzene metabolism and rate of glutathione depletion
6/7/07 37 DRAFT - DO NOT CITE OR QUOTE
-------
1 from the liver. CYP-inhibition (by SKF 525A) significantly retarded bromobenzene metabolism
2 and reduced the rate of glutathione depletion; these rats exhibited no histopathologic signs of
3
6/7/07 3 8 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-11. The influence of various treatments on the metabolism of bromobenzene (BB)
and severity of bromobenzene-induced hepatic necrosis in rats administered a single
intraperitoneal dose of bromobenzene
Treatment
BB(1570mg/kg)
BB (236 mg/kg) + Phenobarbital
BB (1570 mg/kg) + SKF 525 A
BB (1570 mg/kg) + Diethyl maleate
BB (1570 mg/kg) + Cysteine
Severity3 of
centrilobular
liver necrosis
Minimal
Massive
None
Extensive
None
Metabolism of
bromobenzene
(tl/2 in minutes)b
10.0 + 0.8
5.5 + 0.5c
15.5 + 1.8C
10.2 + 0.7
9.8 + 0.8
Rate of glutathione
depletion
(tl/2 in minutes)
66 + 8
20 + 3C
230+ 15C
17 + 3C
68 + 6
2 aCriteria of Brodie et al. (1971) (minimal = a few degenerating parenchymal cells; extensive =
3 central veins surrounded by several rows of dead or degenerating cells; massive = necrosis of
4 extensive liver areas).
5 bHalf-time of clearance of radioactivity from the whole body of rats administered
6 14C-bromobenzene.
7 cSignificantly different from the values of rats treated with bromobenzene only;/><0.05.
8 Source: Jollow et al. (1974)
6/7/07
39
DRAFT - DO NOT CITE OR QUOTE
-------
1 hepatocellular necrosis. The experimental depletion of liver glutathione in the diethyl maleate-
2 treated rats resulted in increased severity of necrosis even though the rate of bromobenzene
3 metabolism was not significantly different from that of rats that were not depleted of glutathione
4 experimentally. Conversely, addition of the glutathione precursor (cysteine) was protective of
5 liver necrosis. Not only do the results demonstrate that metabolism of bromobenzene is
6 correlated with hepatotoxicity, since CYP-induction (phenobarbital-treated rats) increased
7 hepatotoxicity and CYP-inhibition (SKF 525A-treated rats) decreased hepatotoxicity, but they
8 further indicate that acute hepatotoxicity is related to depletion of glutathione.
9 The liver appears to develop a tolerance to acute bromobenzene insult after repeated
10 exposure. Kluwe et al. (1984) assessed bromobenzene-induced effects on liver glutathione
11 levels, serum ALT and SDH levels, and histopathologic liver lesions in male Fischer 344/N rats
12 following single or repeated oral dosing (1 time/day for 10 days). Nonprotein sulfhydryl group
13 concentrations were used as a measure of glutathione levels. A single oral dose of 628 mg/kg
14 resulted in >50% decrease in liver glutathione between 3 and 12 hours posttreatment, partial
15 recovery by 24 hours, and marked increase above control levels at 48 hours. Differences in
16 minimum glutathione levels between treated animals and controls became less pronounced
17 during repeated oral treatment until, following the tenth treatment, there was no significant
18 difference from controls. Within 24 hours posttreatment, the single 628 mg/kg dose of
19 bromobenzene produced moderate focal centrilobular and midzonal hepatocellular necrosis, as
20 well as an inflammatory response. Although these liver lesions were somewhat more severe 24
21 hours following the second treatment, only minimal necrosis was noted following the fourth
22 treatment and was not detected following the tenth treatment. Serum ALT activity was increased
23 following the first, second, and fourth treatments, but not after the tenth treatment.
24 In a similarly-designed dose-response study (0, 9.8, 78.5, or 315 mg/kg-day), a single 315
25 mg/kg dose resulted in glutathione depletion, liver lesions, and increased ALT and SDH (Kluwe
26 et al., 1984). Following the tenth dose, glutathione depletion was less pronounced, ALT and
27 SDH were no longer increased, and liver lesions were not seen. NTP (1985a,b) assessed serum
28 ALT, AST, and SDH levels in male and female Fischer 344/N rats and B6C3F1 mice
29 administered bromobenzene by oral gavage at doses of 0, 50, 200, or 600 mg/kg-day, 5
30 days/week for 90 days. Significantly increased mean serum ALT, AST, and SDH levels
31 (approximately 30- to 100-fold) were noted after the first treatment. After the third treatment,
32 levels of all three enzymes remained significantly elevated on day 3, but the magnitude
33 decreased to approximately 2- to 6-fold above control levels. Serum ALT, AST, and SDH levels
34 were no longer significantly different from controls at terminal sacrifice on day 94. Collectively,
35 these results indicate that acute hepatotoxic levels of bromobenzene may be tolerated upon
6/7/07 40 DRAFT - DO NOT CITE OR QUOTE
-------
1 repeated exposure and that such an adaptive effect may be due to chemically-induced increased
2 production of liver glutathione.
3 As noted in the proposed metabolic scheme for bromobenzene (Section 3.3, Figure 3-1),
4 candidates for reactive metabolites of the 3,4-oxide pathway that may elicit hepatotoxicity
5 include the 3,4-oxide itself, the oxide derivative of 4-bromophenol, the quinone
6 (4-bromoquinone) formed from 4-bromocatechol, and reactive oxygen species formed via redox
7 cycling of 4-bromoquinone. The relative importance of these metabolites to bromobenzene
8 hepatotoxicity is uncertain. There is some evidence that 4-bromophenol and its further
9 metabolites may not be involved in hepatotoxicity since centrilobular hepatic necrosis was
10 observed in rats that were administered bromobenzene (400 mg/kg-day) intraperitoneally, but not
11 in other rats administered 4-bromophenol (up to 440 mg/kg-day) or 4-bromocatechol (up to 485
12 mg/kg-day) (Monks et al., 1984a).
13
14 4.5.2. Mechanistic Studies of Kidney Effects
15 Nephrotoxicity also has been associated with acute exposure to bromobenzene in mice
16 and rats, albeit at doses much higher than those eliciting hepatotoxicity. Mice appear to be more
17 sensitive to the nephrotoxic effects than rats. For example, extensive renal necrosis was
18 observed in male C57 Black/61 mice following a single intraperitoneal injection of a 760 mg/kg-
19 day dose of 14C-bromobenzene, whereas a 1460 mg/kg-day dose to male Sprague-Dawley rats
20 resulted in less severe effects (ranging from swollen and vacuolated tubular cells to dilated
21 convoluted tubules filled with protein casts) (Reid, 1973).
22 The nephrotoxic effects appear to be associated with covalent binding of reactive
23 metabolites to cellular macromolecules in cells of the proximal convoluted tubules, as evidenced
24 by findings that (1) covalent binding of 14C-compounds to kidney proteins in the convoluted
25 tubules peaked several hours prior to the appearance of histopathologic lesions and (2)
26 pretreatment with piperonyl butoxide (a CYP inhibitor) decreased both the rate of metabolism of
27 bromobenzene and the severity of kidney lesions (Reid, 1973). These results, together with
28 demonstrations that intraperitoneal administration of either 2-bromophenol or 2-bromoquinone
29 in rats resulted in histopathologic kidney lesions similar to those induced by bromobenzene,
30 implicate reactive metabolites formed via the 2,3-oxide pathway (see Section 3.3, Figure 3-1) as
31 the most likely source(s) of covalent binding and associated nephrotoxicity, at least in the rat.
32 Lau et al. (1984b) suggested that bromobenzene nephrotoxicity in rats is caused by a
33 metabolite that is produced in the liver and transported to the kidney. In rats, intraperitoneally-
34 injected 2-bromophenol (a metabolite of bromobenzene) resulted in renal necrosis similar to that
35 observed following bromobenzene administration, but at a dose about one-fifth as large as the
36 dose of bromobenzene required to produce lesions of similar severity. Renal glutathi one levels
6/7/07 41 DRAFT - DO NOT CITE OR QUOTE
-------
1 were rapidly and significantly decreased within 90 minutes following administration of
2 2-bromophenol, whereas hepatic glutathione levels were not decreased in the same time period.
3 Conversion of 2-bromophenol to covalently bound material in the kidney was 4-fold greater than
4 that observed in the liver. Furthermore, intraperitoneal administration of another major
5 metabolite of bromobenzene, namely 2-bromohydroquinone, caused renal lesions that were
6 indistinguishable from those induced following bromobenzene treatment (Lauetal., 1984a). In
7 the presence of glutathione, 2-bromohydroquinone gave rise to several hydroquinone-glutathione
8 conjugates, including the very potent nephrotoxicant (2-bromo-bis[glutathione-S-
9 yl]hydroquinone), which is the most likely candidate for a bromobenzene metabolite produced in
10 the liver and transported to the kidney to ultimately exert its toxic effect (Lau and Monks, 1997b;
11 Monks etal., 1985).
12 The 3,4-oxide pathway may also be involved in the nephrotoxic effects observed in mice.
13 Histopathologic lesions of the convoluted tubules were demonstrated in male ICR mice
14 following single parenteral administration of any of the bromophenols (2-, 3-, or 4-bromophenol)
15 or 4-bromocatechol (Rush et al., 1984).
16
17 4.5.3. Genomic/Proteomic Responses of the Liver to Bromobenzene
18 Toxicogenomics involves the application of functional genomics technologies to
19 conventional toxicology. The development of recent analytical techniques allows for the
20 simultaneous detection of numerous biomolecules, thus facilitating complete description of the
21 genome for a particular organism (genomics). These techniques can be applied to analysis of
22 multiple gene transcripts (transcriptomics), proteins (proteomics), and metabolites
23 (metabolomics) as well.
24 Heijne and coworkers (Stierum et al., 2005; Heijne et al., 2004, 2003) used these
25 techniques to identify changes in gene expression in the rodent liver in response to
26 bromobenzene. As previously discussed, bromobenzene undergoes CYP-mediated epoxidation
27 to form the electrophilic 3,4-epoxide, which has been demonstrated to irreversibly bind to
28 proteins such as glutathione S-transferase, liver fatty acid binding protein, and carbonic
29 anhydrase (Koen et al., 2000). Heijne et al. (2003) administered acute intraperitoneal
30 hepatotoxic doses of bromobenzene (0.5-5 mM/kg) to rats and assessed liver tissue for
31 physiological signs of toxicity and changes in protein and gene expression 24 hours
32 posttreatment. Vehicle controls were included in the study. Bromobenzene treatment resulted in
33 glutathione depletion (primarily due to conjugation) within 24 hours, which coincided with the
34 induction of more than 20 liver proteins, including y-glutamylcysteine synthetase (a key enzyme
35 in glutathione biosynthesis). Bromobenzene-induced oxidative stress was indicated by the strong
36 upregulation of a number of genes, including heme oxygenase and peroxiredoxin 1. Transient
6/7/07 42 DRAFT - DO NOT CITE OR QUOTE
-------
1 changes were also noted in the transcript!onal expression of numerous other genes, including
2 ones involved in drug metabolism, intracellular signaling, metabolism, and the acute phase
3 response.
4 Heijne et al. (2004) demonstrated dose-and time-related changes in bromobenzene-
5 induced liver gene expression profiles by administering bromobenzene to groups of rats by oral
6 gavage at doses of 0.5, 2.0, or 4.0 mM/kg and assessing changes in the liver transcriptome at 6,
7 24, and 48 hours posttreatment. Dose- and time-related changes were observed in the
8 transcript!onal expression of numerous genes involved in GSH depletion, drug metabolism,
9 intracellular signaling, metabolism (cholesterol, fatty acid, and protein metabolism), and the
10 acute phase response. At the highest dose, the time-course of altered gene expression coincided
11 with that of histopathological evidence of bromobenzene-induced liver lesions, with few signs of
12 adverse effects at 6 hours and increased evidence of histopathologic liver lesions and altered
13 transcriptional expression at 24 and 48 hours. Although histopathologic liver lesions were not
14 observed at the two lower doses, dose-related altered transcriptional expression was noted and
15 recovery was apparent in the mid-dose group at 48 hours posttreatment. Results of available
16 toxicogenomics assessments provide suggestive evidence for the involvement of some genes in
17 particular aspects of bromobenzene hepatotoxicity. However, the toxicogenomics studies
18 available do not establish key events in the mode of action for bromobenzene-induced
19 hepatotoxicity.
20
21 4.5.4. Similarities Between Bromobenzene and Chlorobenzene
22 Bromobenzene and chlorobenzene (structures shown below) are monohalogenated
23 benzene compounds that are distinguished from one another structurally by the particular
24 halogen, bromine in the case of bromobenzene, and chlorine in the case of chlorobenzene. The
25 two chemicals are structurally similar, with similar Pauling electronegativites of 3.16 and 2.96
26 for chlorine and bromine (Loudon, 1988), respectively. In addition, neither the chlorine nor the
27 bromine atoms are removed from the benzene ring through metabolism.
28
29
30 Figure 4-1. Chemical structure of bromobenzene and chlorobenzene
31
32 Independent in vivo and in vitro studies indicate that bromobenzene and chlorobenzene
33 have similar toxicokinetic properties and share the same critical target of toxicity (liver).
6/7/07 43 DRAFT - DO NOT CITE OR QUOTE
-------
1 Bromobenzene and chlorobenzene each exhibit the ability to enter the systemic circulation of
2 laboratory animals following inhalation or oral exposure (see Section 3.1 for a detailed
3 discussion of the toxicokinetics of bromobenzene and Hellman (1993) for a summary of
4 toxicokinetic information for chlorobenzene). Results of parenteral injection studies in animals
5 indicate that, following absorption, bromobenzene and its metabolites are widely distributed,
6 with highest levels found in adipose tissue (Ogino, 1984b; Zampaglione et al., 1973; Reid et al.,
7 1971). Similar distribution of chlorobenzene has been observed in rats following inhalation
8 exposure to radiolabeled chlorobenzene (Sullivan et al., 1983). Metabolic schemes for both
9 bromobenzene and chlorobenzene include initial CYP-catalyzed epoxidation to reactive epoxide
10 intermediates and subsequent formation of corresponding dihydrodiol derivatives, phenols,
11 glutathione conjugates, catechols, and quinones. Elimination is mainly accomplished via the
12 urinary excretion of bromobenzene- and chlorobenzene-derived metabolites.
13 In a recent study, Chan et al. (2007) demonstrated the usefulness of isolated normal and
14 phenobarbital induced rat hepatocytes for predicting in vivo toxicity caused by a series of
15 halobenzene congeners, including bromobenzene and chlorobenzene. The underlying molecular
16 mechanism of halobenzene hepatotoxicity was elucidated using Quantitative structure-activity
17 relationships (QSARs) and accelerated cytotoxicity mechanism screening (ACMS) techniques in
18 rat and human hepatocytes. The in vivo and in silico studies suggest that halobenzene interaction
19 with cytochrome P450 for oxidation is the rate limiting step for toxicity and is similar in both
20 species.
21 The subchronic oral toxicity studies of bromobenzene in Fischer 344/N rats (NTP, 1985a)
22 and B6C3F1 mice (NTP, 1985b) and chlorobenzene in Fischer 344/N rats and B6C3F1 mice
23 (NTP, 1985e) are the best available data from which to compare the toxicities of repeated
24 exposure to bromobenzene and chlorobenzene. These studies identified the liver and kidney as
25 the most sensitive targets of bromobenzene and chlorobenzene toxicity. Tables 4-12 and 4-13
26 summarize the liver and kidney effects observed for chlorobenzene.
27
28 The database for bromobenzene does not include reproductive or developmental toxicity
29 studies. However, chlorobenzene was assessed for reproductive toxicity in a two-generation
30 study of rats exposed to chlorobenzene vapor concentrations of 0, 50, 150, or 450 ppm daily, 6
31 hours/day for 10 or 11 weeks prior to mating and throughout mating, gestation, and lactation
32 (Nairetal., 1987). Statistically significantly increased incidences of rats with histopathologic
33 liver and kidney lesions were observed in FO and FI male rats at exposure levels >150 ppm. The
34 NOAEL for hepatic effects in this study was 50 ppm. The highest exposure level (450 ppm) did
35 not elicit any clear signs of reproductive toxicity in either generation. Furthermore,
6/7/07 44 DRAFT - DO NOT CITE OR QUOTE
-------
1 chlorobenzene did not induce developmental effects in the fetuses of pregnant rats exposed to
2 vapor concentrations as high as 590 ppm for 6 hours/day on gestation days 6-15 (John et al.,
3 1984).
4 The oral database for chlorobenzene includes one developmental study in which Charles
5 River albino rat dams were administered chlorobenzene at oral dose levels of 100 or 300 mg/kg-
6 day on gestation days 6-15 (IBT, 1977). Although no developmental toxicity was elicited, it is
7
6/7/07 45 DRAFT - DO NOT CITE OR QUOTE
-------
Table 4-12. Incidences of male and female Fischer 344/N rats with liver and kidney lesions
following administration of chlorobenzene by gavage 5 days/week for 13 weeks
Endpoint
Dose (mg/kg-day)
0
125
250
500
750a
Males
Liver
Necrosis
Degeneration
Kidney
Nephropathy
0/10
0/10
0/10
0/10
0/10
0/10
2/10
0/10
1/10
3/10
2/10
2/10
7/1 Ob
1/10
2/10
Females
Liver
Necrosis
Degeneration
Kidney
Nephropathy
0/10
0/10
0/10
0/10
0/10
0/10
1/10
0/10
0/10
1/10
0/10
0/10
6/1 Ob
4/1 Ob
7/1 Ob
2
o
J
4
5
6
7
aSignificantly decreased survival in the 750 mg/kg-day dose groups may have influenced
observed incidences of animals with liver and/or kidney lesions.
bStatistically significantly different from control groups according to Fisher's exact test (p<0.05),
performed by Syracuse Research Corporation.
Source: NTP (1985e)
Table 4-13. Incidences of male and female B6C3F1 mice with liver and kidney lesions
following administration of chlorobenzene by gavage 5 days/week for 13 weeks
Endpoint
Dose (mg/kg-day)
0
60
125
250
500a
750a
Males
Liver
Necrosis
Degeneration
Kidney
Nephropathy
0/10
0/10
0/10
1/10
0/10
NE
1/10
0/10
0/10
7/1 Ob
2/10
4/1 Ob
10/10b
0/10
9/1 Ob
10/10b
0/10
8/1 Ob
Females
Liver
Necrosis
Degeneration
Kidney
Nephropathy
0/10
0/10
0/10
0/10
0/10
NE
0/10
0/10
0/10
10/10b
0/10
4/1 Ob
8/1 Ob
9/1 Ob
0/10
1/10
4/1 Ob
0/10
9
10
11
12
13
14
aSignificantly decreased survival in the 500 and 750 mg/kg-day dose groups may have
influenced observed incidences of animals with liver and/or kidney lesions.
bStatistically significantly different from control groups according to Fisher's exact test (p<0.05),
performed by Syracuse Research Corporation.
NE = not examined, due to the absence of lesions at the next higher dose
Source: NTP (1985e)
6/7/07
46
DRAFT - DO NOT CITE OR QUOTE
-------
1 uncertain whether repeated oral doses of chlorobenzene as high as those known to induce
2 histopathologic liver lesions in rats (750 mg/kg-day) might also cause developmental effects.
3 Significantly increased mean relative (but not absolute) testis weight was noted in 400
4 and 600 mg/kg treatment groups of male rats administered bromobenzene via oral gavage
5 5days/week for 13 weeks (NTP, 1985a). However, gross and histopathologic examinations of
6 these dose groups did not reveal other significant treatment-related testicular effects. No
7 treatment-related effects were observed at any exposure level among female rats or male or
8 female mice in the oral study (NTP, 1985a,b). There were no indications of significant
9 exposure-related effects on reproductive organs or tissues in male or female rats or mice exposed
10 to bromobenzene at any of the vapor concentrations used in the 13-week inhalation study of NTP
11 (NTP, 1985c,d). Taken together, these results indicate that reproductive and developmental
12 endpoints do not appear to be more sensitive targets of chlorobenzene or bromobenzene toxicity
13 than the liver.
14 Although no chronic-duration oral or inhalation animal studies are available for
15 bromobenzene, a 2-year toxicity and carcinogenicity study is available for chlorobenzene (NTP,
16 1985e). Groups of male and female F344/N rats and B6C3F1 mice (50/sex/species) were
17 administered chlorobenzene by oral gavage at doses of 0, 60, or 120 mg/kg-day (0, 30, or 60
18 mg/kg-day for male mice), 5 days/week for 2 years. There was no evidence of treatment-related
19 increased incidences of nonneoplastic liver lesions in female rats or male or female mice,
20 including the highest dose level tested (120 mg/kg-day for female rats and mice, 60 mg/kg-day
21 for male mice). There was equivocal evidence for treatment-related increased incidence of
22 hepatocellular necrosis in high-dose (120 mg/kg-day) male rats. The original pathology report
23 noted necrosis in 7/50 high-dose males (0/50 in vehicle controls). However, an independent
24 pathological review resulted in a diagnosis of hepatocellular necrosis in one vehicle control male
25 rat (1/50) and a single high-dose male rat (1/50). The NTP 2-year oral study of chlorobenzene
26 identified a free-standing no-observed-adverse-effect level (NOAEL) of 120 mg/kg-day in
27 female rats and equivocal evidence of a lowest-observed-adverse-effect level (LOAEL) of 120
28 mg/kg-day for hepatocellular necrosis in male rats. In male and female mice, free-standing
29 NOAELs were 60 and 120 mg/kg-day, respectively, for nonneoplastic liver effects. In a
30 similarly-designed subchronic (90-day) oral toxicity study in mice, a NOAEL of 125 mg/kg-day
31 was identified in both males and females; the LOAEL was 250 for chlorobenzene-induced liver
32 lesions (NTP, 1985e). These results suggest the development of some degree of tolerance to
33 chlorobenzene during chronic exposure (i.e., dose-response relationships for subchronic and
34 chronic exposure appear to be similar). It is reasonable to expect such similarities in dose-
35 response relationships for subchronic and chronic exposure to bromobenzene as well because
6/7/07 47 DRAFT - DO NOT CITE OR QUOTE
-------
1 mechanistic studies have demonstrated the development of some degree of tolerance upon
2 repeated exposure to bromobenzene (Kluwe et al., 1984).
3
4 4.6. SYNTHESIS AND EVALUATION OF MAJOR NONCANCER EFFECTS
5 4.6.1. Oral
6 No data are available on health effects in humans following oral exposure to
7 bromobenzene. No chronic-duration toxicity, reproductive toxicity, or developmental toxicity
8 studies are available in animals following oral exposure to bromobenzene. Pertinent information
9 on health effects in animals is restricted to results from studies of rats and mice administered
10 bromobenzene by oral gavage at doses of 0, 50, 100, 200, 400, or 600 mg/kg-day, 5 days/week
11 for 90 days (NTP, 1985a,b, 1986a). The liver was the most sensitive toxicity target in these NTP
12 studies. Results of mechanistic studies involving acute oral exposures support this finding (e.g.,
13 Heijne et al., 2004; Bambal and Hanzlik, 1995; Kluwe et al., 1984). Dose-related significantly
14 increased liver weights were observed in all treated groups of female rats and mice (50-600
15 mg/kg-day) and all but the 50 mg/kg-day groups of male rats and mice (liver weights were not
16 available for the 50 mg/kg-day group of male rats). Oral doses >200 mg/kg-day resulted in
17 significantly increased incidences of histopathologic liver lesions in male and female rats and
18 male mice (>400 mg/kg-day in female mice).
19 Subchronic-duration oral exposure to bromobenzene also resulted in statistically
20 significantly increased incidences of renal lesions such as necrosis and degeneration (without
21 observable regeneration) in the proximal convoluted tubules in male and female rats and male
22 mice, but only at the highest (600 mg/kg-day) dose level (NTP, 1985a).
23 The Pathology Working Group (NTP, 1986a) reported that lesions in the brain, stomach,
24 thymus and bone marrow of the rats were present primarily or solely at the 600 mg/kg-day level.
25 Liver and kidney lesions persisted through the 400 mg/kg-day dosed rats, but were essentially
26 absent or present to a minimal degree in the rats at the 200 mg/kg-day dose level. In the NTP
27 study in mice (NTP, 1985b), bromobenzene lesions were limited to the liver and were of less
28 severity at 400 and 200 mg/kg-day and were essentially absent at 100 and 50 mg/kg-day.
29 Relatively high single oral doses (>785 mg/kg-day) have been shown to elicit hepatic,
30 renal, and pulmonary effects in laboratory animals (Casini et al., 1986; Forkert, 1985; Kluwe et
31 al., 1984; Patrick and Kennedy, 1964). However, pulmonary effects were not observed in the
32 subchronic oral studies of NTP (1985a,b).
33
34
35 4.6.2. Inhalation
6/7/07 48 DRAFT - DO NOT CITE OR QUOTE
-------
1 No data are available on health effects in humans following inhalation exposure to
2 bromobenzene. No chronic-duration toxicity, reproductive toxicity, or developmental toxicity
3 studies are available in animals following inhalation exposure to bromobenzene. Pertinent
4 information on health effects in animals is restricted to results from studies in rats and mice
5 exposed to bromobenzene at vapor concentrations of 0, 10, 30, 100, or 300 ppm, 6 hours/day, 5
6 days/week for 13 weeks (NTP, 1985c,d). The liver appeared to be the most sensitive toxicity
7 target in these studies. Liver weights (absolute and relative-to-body weight) were significantly
8 increased at exposure concentrations >100 ppm in both sexes of rats. The liver-to body weight
9 ratio was significantly increased in 100-ppm male mice (the study did not include a 300-ppm
10 male group). Statistically significantly increased liver-to-body weight ratios occurred in female
11 mice at all bromobenzene exposure concentrations (including 10 ppm). Statistically significantly
12 increased absolute liver weights occurred at all exposure concentrations >30 ppm.
13 A statistically significantly increased incidence of cytomegaly was observed only in
14 female mice of the highest exposure level (300 ppm; male mice were not exposed at this
15 concentration). The Pathology Working Group (NTP, 1986b) agreed with the diagnosis of
16 cytomegaly, hepatic necrosis, and mineralization in the 300 ppm group, but considered necrosis
17 and inflammation in the liver of female mice to be minimal or not present in the 100 ppm or
18 lower exposure groups. There was no clear evidence of renal toxicity in mice repeatedly
19 exposed to bromobenzene vapor concentrations up to and including the highest concentration
20 tested (100 ppm in males and 300 ppm in females) (NTP, 1985d).
21 The liver was shown to be a target of bromobenzene toxicity in mice following a single
22 4-hour exposure to bromobenzene vapor concentrations as low as 250 ppm (Becher et al., 1989;
23 Brondeau et al., 1983). Necrosis was also noted in the lungs of mice following a single 4-hour
24 exposure to bromobenzene at a vapor concentration of 1000 ppm (Becher et al., 1989).
25 However, lung lesions were not seen in rats or mice repeatedly exposed to bromobenzene vapors
26 at concentrations up to 300 ppm (NTP, 1985c,d).
27
28
29 4.6.3. Mode of Action Information
30 No human data are available for health effects following exposure to bromobenzene by
31 any exposure route for any duration. Animal studies demonstrate that relatively high single oral
32 doses (>785 mg/kg) of bromobenzene elicit lesions in the liver, kidney, and lung. Parenteral
33 injection studies support these findings. Hepatic effects have also been elicited in mice
34 following a single 4-hour exposure to bromobenzene vapors at a concentration of 250 ppm; a
35 higher concentration (1000 ppm) resulted in lung lesions. Sub chronic-duration (90-day) oral and
6/7/07 49 DRAFT - DO NOT CITE OR QUOTE
-------
1 inhalation studies in rats and mice have identified the liver as the most sensitive target of
2 repeated exposure to bromobenzene.
3 The results of several mechanistic studies in animals demonstrate that bromobenzene
4 hepatotoxicity is associated with metabolism of the parent compound and that cytotoxicity likely
5 results from modifications of hepatocellular macromolecules by one or more reactive
6 metabolites.
7 Available data further indicate that reactive metabolites are formed via the metabolic
8 pathway that involves the 3,4-oxide (rather than the 2,3-oxide) derivative of bromobenzene.
9 Supporting evidence includes the findings that:
10 • Induction of p-naphthoflavone- or 3-methylcholanthrene-induced CYP isozymes
11 results in increased urinary excretion of 2-bromophenol (formed via the 2,3-oxide
12 pathway) and decreased hepatotoxicity (Jollow et al., 1974; Lau and Zannoni, 1979;
13 Lau et al., 1980; Zampaglione et al., 1973), whereas
14 • Induction of phenobarbital-induced CYP isozymes results in increased urinary
15 excretion of 4-bromophenol (formed via the 3,4-oxide pathway) as well as increases
16 in severity of hepatocellular necrosis and increases in the extent of covalent binding
17 of radioactivity from 14C-bromobenzene to hepatocellular macromolecules in the
18 region of observed hepatocellular necrosis (Brodie et al., 1971).
19 Candidates for reactive metabolites of the 3,4-oxide pathway that may elicit
20 hepatotoxicity include the 3,4-oxide itself, the oxide derivative of 4-bromophenol, the quinone
21 (4-bromoquinone) formed from 4-bromocatechol, and reactive oxygen species formed via redox
22 cycling of 4-bromoquinone. The relative importance of these metabolites to bromobenzene
23 hepatotoxicity is uncertain. There is some evidence that 4-bromophenol and its further
24 metabolites may not be involved in hepatotoxicity since centrilobular hepatic necrosis was
25 observed in rats that were administered bromobenzene (400 mg/kg) intraperitoneally but not in
26 other rats administered 4-bromophenol (up to 440 mg/kg) or 4-bromocatechol (up to 485 mg/kg)
27 (Monks et al., 1984a).
28 Molecular mechanisms responsible for bromobenzene hepatotoxicity may include
29 bromobenzene-induced alterations in liver proteins and gene expression. Heijne and coworkers
30 used a toxicogenomics approach to study molecular mechanisms of bromobenzene
31 hepatotoxicity (Heijne et al., 2003, 2004). Rats were administered bromobenzene
32 intraperitoneally (0.5-5 mM/kg), and liver tissue was assessed for physiological signs of toxicity
33 and changes in protein and gene expression for up to 48 hours posttreatment. Bromobenzene
34 treatment resulted in glutathione depletion (primarily due to conjugation) within 24 hours, which
35 coincided with induction of more than 20 liver proteins, including y-glutamylcysteine synthetase
36 (a key enzyme in glutathione biosynthesis). Transient changes were also noted in the
6/7/07 50 DRAFT - DO NOT CITE OR QUOTE
-------
1 transcript!onal expression of numerous genes involved in drug metabolism, oxidative stress,
2 glutathione depletion, the acute phase response, metabolism, and intracellular signaling.
3 Nephrotoxicity has also been observed in animals following acute-duration exposure to
4 bromobenzene, albeit at higher doses than the lowest hepatotoxic doses. Repeated-dose oral and
5 inhalation studies in rats and mice provide evidence for kidney effects but only at the highest
6 exposure levels tested, which also resulted in lethality. Nephrotoxicity also appears to result
7 from modification of macromolecules in cells of the proximal convoluted tubule by one or more
8 reactive metabolites.
9
10 4.7. EVALUATION OF CARCINOGINICITY
11 Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), there is
12 inadequate information available for an assessment of the human carcinogenic potential of
13 bromobenzene. Cancer studies in humans and cancer bioassays in animals exposed to
14 bromobenzene were not found. As discussed in Section 4.4.2, bromobenzene was not mutagenic
15 in the Ames assay and did not consistently produce marked cytogenetic effects in vitro with
16 mammalian cells, even in the presence of rat liver S-9 preparations. Bromobenzene increased
17 formation of micronucleated polychromatic erythrocytes in bone marrow of mice given acute
18 oral doses of 125 mg/kg and was bound to DNA and RNA following intraperitoneal injection.
19 The available genotoxicity data, therefore, is inadequate to assess bromobenzene genotoxicity.
20
21 4.8. SUSCEPTIBLE POPULATIONS
22 4.8.1. Possible Childhood Susceptibility
23 Limited data were located regarding age-related susceptibility to bromobenzene. Single
24 intraperitoneal injection of bromobenzene at concentrations that produced extensive centrilobular
25 necrosis in the liver of adult rats failed to produce similar lesions in neonatal rats (Green et al.,
26 1984; Mitchell et al., 1971). The lack of hepatotoxicity in the neonatal rats was presumably the
27 result of a generally low level of hepatic microsomal enzymes observed in early neonatal stages
28 of development (Kato et al., 1964).
29
30 4.8.2. Possible Gender Differences
6/7/07 51 DRAFT - DO NOT CITE OR QUOTE
-------
1 Available information regarding gender-related susceptibility to bromobenzene is
2 restricted to animal studies. In rats (NTP, 1985a), results of subchronic-duration oral exposure to
3 bromobenzene indicate that males are somewhat more susceptible than females to hepatocellular
4 effects such as centrilobular necrosis and cytomegaly (see Table 4-2). Male-female differences
5 were not as apparent following subchronic-duration oral exposure in mice (see Table 4-4) (NTP,
6 1985b).
7
8 4.8.3. Other
9 No data are available regarding the effects of bromobenzene on other potentially
10 susceptible populations. However, since the experimental depletion of glutathione in
11 bromobenzene-treated animals has been demonstrated to potentiate bromobenzene hepatotoxicity
12 (Jollow et al., 1974), individuals with abnormally low levels of glutathione, such as those with
13 GSH synthetase deficiency (Meister, 1982), could potentially be at increased risk for
14 bromobenzene hepatotoxicity. The importance of glutathione conjugation as a protective
15 mechanism for bromobenzene hepatotoxicity may also make individuals exposed to other
16 glutathione depleting chemicals more susceptible to bromobenzene hepatotoxicity.
17
6/7/07 52 DRAFT - DO NOT CITE OR QUOTE
-------
5. DOSE-RESPONSE ASSESSMENTS
4 5.1. ORAL REFERENCE DOSE
5 5.1.1. Subchronic Oral RfD
6 5.1.1.1. Choice of Principal Study and Critical Effect - with Rationale and Justification
7 As discussed in Section 4.1, there are no human studies available for development of a
8 subchronic RfD. The toxicity database for repeated oral exposure in laboratory animals that are
9 available for selection of a subchronic RfD consists of two 90-day gavage studies: one in rats
10 (NTP, 1985a) and one in mice (NTP, 1985b). No reproductive or developmental toxicity studies
11 are available.
12 The liver appears to be the principal target organ for bromobenzene toxicity in rodents.
13 Significantly increased incidences of hepatocellular necrosis (a clear indicator of an adverse
14 effect) were observed at doses of 400 and 600 mg/kg-day in male and female B6C3Flmice and
15 male Fischer 344/N rats (600 mg/kg-day in female Fischer 344/N rats) (NTP, 1985a,b). These
16 dose levels also resulted in greater than 3-fold increases (statistically and biologically significant)
17 in serum concentrations of SDH, an enzyme indicative of liver damage. Significantly increased
18 incidences of cytomegaly were observed at doses >200 mg/kg-day in male and female mice and
19 male rats (>400 mg/kg-day in female rats). Significantly increased mean liver weights were
20 observed at doses as low as 50 mg/kg-day in female rats and mice and 100 mg/kg-day in male
21 rats and mice.
22 Kidney lesions were associated with the proximal convoluted tubule and consisted of
23 degeneration, casts, necrosis (rats only), and mineralization in male and female rats and male
24 mice. The incidence of kidney lesions was not considered for the development of the subchronic
25 RfD because the lowest dose associated with a statistically significant increase in the incidence
26 of renal lesions (600 mg/kg-day in rats and mice) was higher than the lowest dose (400
27 mg/kg-day rats and mice) resulting in a clear treatment-related adverse liver effect
28 (hepatocellular necrosis), indicating that the liver effects are a more sensitive indicator of
29 bromobenzene toxicity.
30 Comprehensive histopathologic examinations of all major tissues and organs in the
31 subchronic studies of rats and mice revealed no significantly increased incidences of exposure-
32 related lesions at sites other than liver and kidney.
33 The increase in the incidence of liver lesions and the increase in absolute and relative
34 liver weight in rats and mice, and the increase in serum concentrations of SDH in male and
35 female mice, were considered in the selection of the critical effect for the development of the
36 subchronic RfD. The increase in liver weight and enzyme levels may be considered to be on a
6/7/07 53 DRAFT - DO NOT CITE OR QUOTE
-------
1 continuum leading to the observed liver toxicity. It is difficult to ascertain which liver lesions
2 are most important or occur first in the development of liver toxicity. Therefore, liver lesions
3 were combined so that an animal with any of the four observed lesions (centrilobular
4 cytomegaly, necrosis, inflammation, or mineralization) was counted as having a liver lesion.
5 The rationale for combining the liver lesions in this manner includes findings that: (1) all four
6 lesions were principally observed in the centrilobular region of the liver; (2) statistically
7 significantly increased incidences of hepatocellular necrosis or inflammation were observed and
8 associated only with doses equal to or greater than those eliciting statistically significantly
9 increased incidences of cytomegaly; and (3) inflammation and mineralization were considered,
10 by the NTP study authors, to be direct results of hepatocellular necrosis (NTP, 1985a,b).
11
12 5.1.1.2. Methods of Analysis - Including Models (PBPK, BMD, etc.)
13 All available models in the EPA Benchmark Dose Software (BMDS version 1.3.2) were
14 fit to the incidence data for the combined incidence of animals with one or more of the
15 histopathologic liver lesions (centrilobular cytomegaly, necrosis, inflammation, mineralization).
16 All models were also fit to the increases in absolute liver weight and liver-to-body weight ratios
17 in male and female rats and mice and the increases in SDH levels in male and female mice from
18 the subchronic oral gavage studies (NTP, 1985a,b). Modeling results are presented in
19 Appendix B.
20 The modeled liver lesion data are shown in Table 5-1. Results of the best fitting models
21 (lowest Akaike Information Criterion [AIC]) for incidences of male and female rats and mice
22 with liver lesions are presented in Table 5-2. The female mouse liver lesion data produced the
23 lowest BMDLio (24.8 mg/kg-day), indicating that female mice have a lower point of departure
24 for bromobenzene hepatotoxicity (BMDs and BMDLs for 10, 5, and 1% extra risk are shown in
25 Table 5-3). The conventional BMR of 10% extra risk (U.S. EPA, 2000c) was selected because
26 the small group sizes (n=10) in the principal study preclude selecting a lower benchmark risk
27 level.
28 The modeled data for absolute liver weight and liver-to-body weight ratio (relative liver
29 weight) for rats and mice are shown in Table 5-4. Dose-related statistically significantly
30 increased mean liver weights (absolute, relative-to-body weight) were observed in male rats at
31 doses of 100-400 mg/kg-day and at all dose levels in female rats. Changes in the 600 mg/kg-day
32 males were similar in magnitude to changes in the 400 mg/kg-day males, but could not be
33 assessed for statistical significance because only one survivor remained in this group at study
34 termination. In male mice, absolute liver weight was significantly increased at dose levels >200
35 mg/kg-day, while the liver-to-body weight ratio was significantly increased at dose levels >100
36 mg/kg-day. In female mice, both measures of liver weight were significantly increased in a
37
6/7/07 54 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-1. Incidences of male and female Fischer 344/N rats and B6C3F1 mice with liver
lesions3 following administration of bromobenzene by gavage 5 days/week for 90 days
Dose (mg/kg-day)
Male rats
Female rats
Male mice
Female mice
0
2/10
2/10
1/10
0/10
50
2/10
2/10
0/10
1/10
100
2/10
4/10
2/10
2/10
200
7/1 Ob
5/10
6/1 Ob
6/1 Ob
400
10/10b
10/10b
10/10b
9/1 Ob
600
10/10b
10/10b
10/10b
10/10b
2
o
J
4
5
6
7
8
9
10
Incidences of rats with one or more of the liver lesion types (cytomegaly, necrosis,
inflammation, mineralization), extracted from individual animal histopathologic results provided
to Syracuse Research Corporation by NTP. Liver lesions were not seen in 2/10 male rats of the
200 mg/kg-day dose level that died early due to gavage error.
bStatistically different from control groups according to Fisher's exact test (p<0.05), performed
by Syracuse Research Corporation.
Source: NTP (1985a,b)
Table 5-2. Benchmark doses (BMDioS and BMDLios) from best fitting models predicting
combined incidences of Fischer 344/N rats or B6C3F1 mice with liver lesions (see
Appendix B)
Data set
Male rats
Female rats
Male mice
Female mice
Model
Log-logistic
Log-logistic
Multi-stage
Weibull
BMDioS and BMDLios
(mg/kg-day)
BMDio
172.1
184.7
98.0
56.1
BMDLio
69.2
66.1
38.8
24.8
Fit statistics
x2 p-value
1.00
0.85
0.87
0.99
AIC
46.2
52.7
35.9
40.8
11
12
Table 5-3. Weibull model-estimated BMDs and BMDLs (mg/kg-day) associated with 10,
5, and 1% extra risk for liver lesions in female B6C3F1 mice administered bromobenzene
by oral gavage 5 days/week for 90 days
BMDs and BMDLs (mg/kg-day)
10% Extra risk
BMDio
56.1
BMDLio
24.8
5% Extra risk
BMD05
36.0
BMDL05
12.7
1% Extra risk
BMDoi
13.2
BMDLoi
2.8
13
14
15
Source: NTP (1985b)
6/7/07
55
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-4. Data for absolute liver weight and liver-to-body weight ratio for male and
female Fischer 344/N rats and male and female B6C3F1 mice following administration of
bromobenzene by gavage 5 days/week for 90 days (mean +/- standard deviation)
Dose (mg/kg-day)
0
50
100
200
400
600
Absolute liver weight (grams)
Rats
Male
Female
Mice
Male
Female
9.16 + 0.66
4.68 + 0.35
1.05 + 0.14
0.86 + 0.06
NA
5.23* + 0.37
1.13 + 0.15
0.96* + 0.08
10.64* + 0.76
5. 55* + 0.36
1.12 + 0.12
1.01* + 0.08
11.29* + 0.69
6.28* + 0.40
1.25* + 0.22
1.08* + 0.06
11. 87* + 0.80
7.85* + 0.49
1.27* + 0.11
1.12* + 0.07
—
—
1.56* + 0.16
1.30* + 0.06
Liver-to-Body Weight Ratio (relative liver weight)
Rats
Male
Female
Mice
Male
Female
26.72 + 1.88
24.25 + 1.13
33.4 + 2.41
38.1 + 1.42
NA
26.55* + 1.23
33.9 + 3.52
40.2* + 2.02
31.08* + 1.18
28.69* + 1.20
36.0* + 1.91
42.5* + 1.62
34.10* + 0.68
33.48* + 1.37
37.3* + 4.48
44.4* + 2.12
40.56* + 3. 16
43.11*+2.38
45.3* + 1.83
48.0* + 2.13
—
—
51.2* + 3.48
55.2* + 2.56
2 * Statistically significantly different from controls (p<0.05) based on Student's two-tailed t-test.
3 Source: NTP (1985a,b)
4
6/7/07
56
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
dose-related manner in all bromobenzene treatment groups. Results for the best fitting models
(lowest AIC) for absolute liver weight and liver-to-body weight ratio in male and female rats and
mice are presented in Table 5-5. The lowest BMDLisd value for liver weight effects was 25.8
mg/kg-day for absolute liver weight in female mice. A 0.5 standard deviation (O.Ssd) change
from the control mean was also considered as a potential benchmark response (BMR) for
absolute liver weight in female mice (see Table 5-6).
Table 5-5. Benchmark doses (BMDio and BMDLio) from best fitting models for increased
absolute liver weight and liver-to-body weight ratio in Fischer 344/N rats and B6C3F1
mice administered bromobenzene by gavage 5 days/week for 90 days
Data set
Model
BMDlsd and BMDLlsd
(mg/kg-day)
BMDlsd
BMDLisd
Fit-statistics
X2p-value
AIC
Absolute liver weight (grams)
Male rats
Female rats
Male mice
Female mice
Polynomial (2°)
Linear
Linear
Polynomial (3°)
48.82
49.18
215.16
34.78
35.4
41.44
164.36
25.79
0.47
0.80
0.29
0.90
16.10
-42.58
-139.46
-242.57
Liver-to-body weight ratio (relative liver weight)
Male rats
Female rats
Male mice
Female mice
Linear
Linear
Linear
Polynomial (3°)
41.29
30.90
97.91
40.61
31.15
26.27
81.36
29.32
0.52
0.96
0.49
0.79
89.98
91.83
169.81
136.58
9
10
8 Source: NTP (1985a,b)
Table 5-6. The third-degree polynomial model-estimated BMDs and BMDLs (mg/kg-day)
associated with 1 and 0.5 standard deviation extra risk for increased absolute liver weight
in female B6C3F1 mice administered bromobenzene by oral gavage 5 days/week for 90
days
BMDs and BMDLs (mg/kg-day)
BMDisd
34.78
BMDLisd
25.79
BMDo.ssd
16.43
BMDLo.ssd
12.34
11
12
Source: NTP (1985b)
6/7/07
57
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
ALT, AST, and SDH serum levels in F-344/N rats generally showed increases over
controls. ALT and AST serum levels in B6CF1 mice did not demonstrate a clear dose response,
had a large variance and, as such, were not used for benchmark dose modeling. Statistically
increased serum SDH values were observed at dose levels >200 mg/kg-day relative to sex
matched controls in male and female mice.
The linear, polynomial, power, and Hill models were used to model the SDH serum
levels for male and female mice data shown in Table 5-7. The power model for female mice
data provided the best fit for SDH modeling. The results for the power model are shown in
Table 5-8.
Table 5-7. Data for SDH for male and female B6C3F1 mice following administration of
bromobenzene by gavage 5 days/week for 90 days (mean +/- standard deviation)
Sex
Male
Female
Dose mg/kg-day
0
25 + 2.5
13 + 1.9
50
27 + 3.1
12+ 1.6
100
27 + 3.2
14+ 1.8
200
41* + 19.3
15* + 1.7
400
89* + 28.3
23* + 4.6
600
101* + 29.0
43* + 18.8
11
12
* Statistically significantly increased from controls (p<0.05) based on students two tailed t-test.
Table 5-8. The power model estimated BMD and BMDLs associated with 10% extra risk
for increased SDH serum levels in B6C3F1 female mice exposed to bromobenzene by
gavage 5 days/week for 90 days
Data Set
Female mice
BMD
(mg/kg-day)
196.47
BMDL
(mg/kg-day)
145.79
Fit-statistics
x p-value
1.33
AIC
192.63
13
14
15
16
17
18
19
20
21
22
23
24
25
In summary, female mice have a lower point of departure for hepatotoxicity of
bromobenzene than male mice or male or female rats as indicated by the BMDLs in Tables 5-2
(liver lesions) and 5-5 (absolute liver weight and liver-to-body weight ratio). The increase in
SDH levels in male and female mice was a less sensitive effect and was highly variable. The
lowest BMDLisd from the best fitting model for liver weight changes was 25.8 mg/kg-day, which
was very similar to the lowest BMDLio from the best fitting model for combined liver lesions of
24.8 mg/kg-day. For this reason, liver toxicity in female mice, as defined by an increase in liver
weight and liver lesions was selected as the critical effect for deriving the subchronic RfD. The
average BMDL of 25 mg/kg-day was selected as the point of departure to derive the subchronic
RfD for bromobenzene.
6/7/07
58
DRAFT - DO NOT CITE OR QUOTE
-------
1 5.1.1.3. Subchronic RfD Derivation - Including Application of Uncertainty Factors (UFs)
2 Benchmark dose (BMD) analysis of liver toxicity data for female mice yielded an
3 average BMDL of 25 mg/kg-day, which was selected as the point of departure for deriving a
4 subchronic RfD for bromobenzene (see Section 5.1.2). The point of departure (25 mg/kg-day for
5 mice that were administered bromobenzene by gavage 5 days/week for 90 days) was adjusted to
6 account for daily exposure (25 mg/kg-day x 5 days/7 days =17.8 mg/kg-day) and divided by a
7 total UF of 1000. The UF consists of three areas of uncertainty: (1) interspecies extrapolation,
8 (2) interindividual human variability, and (3) database deficiencies.
9 A 10-fold UF was used to account for laboratory animal-to-human interspecies
10 differences (UFA). No information is available on toxicokinetic or toxicodynamic differences or
11 similarities for bromobenzene in animals and humans.
12 A 10-fold UF for intraspecies differences (UFH) was used to account for variability in
13 susceptibility in human populations. The default value of 10 was selected in the absence of
14 information indicating the degree to which humans may vary in susceptibility to bromobenzene
15 hepatotoxicity.
16 A 10-fold UF was used to account for database deficiencies (UFo). Subchronic studies in
17 rats and mice are available. Well-designed developmental toxicity and multi-generation
18 reproductive toxicity studies are lacking. Therefore, an uncertainty factor if 10 was applied.
19 Bromobenzene and chlorobenzene exhibit striking similarities in structure, toxicokinetic
20 properties, and critical target of toxicity (liver) in rats and mice (see Section 4.5.4 for a detailed
21 discussion). Therefore, the toxicity database for chlorobenzene was assessed for its potential to
22 address database deficiencies for bromobenzene. For example, in a 2-generation reproductive
23 toxicity study in rats, chlorobenzene did not elicit any clear signs of reproductive toxicity in
24 either generation at an exposure level of 450 ppm (Nair et al., 1987). In the same study, both FO
25 and FI male rats exhibited chlorobenzene-induced hepatotoxicity from inhalation exposure at
26 concentrations as low as 150 ppm. Chlorobenzene did not induce developmental effects in the
27 fetuses of pregnant rats exposed to vapor concentrations as high as 590 ppm for 6 hours/day on
28 gestation days 6-15 (John et al., 1984) or in fetuses of rat dams administered chlorobenzene at
29 oral dose levels of 100 or 300 mg/kg-day on gestation days 6-15 (IBT, 1977). In addition to the
30 chlorobenzene data, the subchronic oral gavage studies of bromobenzene in rats and mice did not
31 reveal evidence of significant treatment-related effects on reproductive organs or tissues at dose
32 levels that were clearly hepatotoxic (NTP, 1985a,b). Collectively, these results indicate that
33 reproductive and developmental endpoints may not be particularly sensitive targets of
34 bromobenzene or chlorobenzene toxicity. However, because database deficiencies for
35 chlorobenzene include the lack of a developmental toxicity study in a second animal species, the
6/7/07 59 DRAFT - DO NOT CITE OR QUOTE
-------
1 default value of 10 for deficiencies in the bromobenzene database was not reduced.
2 The subchronic RfD for bromobenzene was calculated as follows:
3
4 Subchronic RfD = (average BMDL x 5/7) H- UF
5 = (25 mg/kg-day x 5/7) - 1000
6 =17.8 mg/kg-day - 1000
7 = 0.02 mg/kg-day (rounded to one significant figure)
8
9 5.1.2. Chronic Oral RfD
10 5.1.2.1. Choice of Principal Study and Critical Effect - with Rationale and Justification
11 As discussed in Section 4.1, there are no human studies available for development of a
12 chronic RfD. The toxicity database for repeated oral exposure in laboratory animals that are
13 available for selection of a chronic RfD consists of two 90-day gavage studies: one in rats (NTP,
14 1985a) and one in mice (NTP, 1985b). No chronic-duration, reproductive toxicity, or
15 developmental toxicity studies are available.
16 The choices of principal study and critical effect for development of a chronic RfD for
17 bromobenzene are the same as those described for the development of a subchronic RfD (see
18 Section 5.1.1.1). The increase in the incidence of liver lesions and the increase in absolute and
19 relative liver weight in rats and mice, and the increase in serum concentrations of SDH in male
20 and female mice were considered in the selection of the critical effect for the development of the
21 chronic RfD. Liver toxicity in female mice, as defined by an increase in liver weight and liver
22 lesions was selected as the critical effect for deriving the chronic RfD.
23
24 5.1.2.2. Methods of Analysis - Including Models (PBPK, BMD, etc.)
25 The methods of analysis used to derive the subchronic RfD for bromobenzene apply to
26 the derivation of the chronic RfD as well (see Section 5.1.1.2).
27
28 5.1.2.3. Chronic RfD Derivation - Including Application of Uncertainty Factors (UFs)
29 The lowest BMDLisd from the best fitting model for liver weight changes was 25.8
30 mg/kg-day, which was very similar to the lowest BMDLio from the best fitting model for
31 combined liver lesions of 24.8 mg/kg-day. An average dose of 25 mg/kg-day was selected as the
32 point of departure for deriving a chronic RfD for bromobenzene (see Section 5.1.2). The point
33 of departure (25 mg/kg-day for female mice administered bromobenzene by gavage 5 days/week
34 for 90 days) was adjusted to account for daily exposure (25 mg/kg-day x 5 days/7 days =17.8
35 mg/kg-day) and divided by a total UF of 3000. The UF consists of four areas of uncertainty: (1)
6/7/07 60 DRAFT - DO NOT CITE OR QUOTE
-------
1 interspecies extrapolation, (2) interindividual human variability, (3) subchronic to chronic
2 duration extrapolation, and (4) database deficiencies.
3 A 10-fold UF was used to account for laboratory animal-to-human interspecies
4 differences (UFA). No information is available on toxicokinetic or toxicodynamic differences or
5 similarities for bromobenzene in animals and humans.
6 A 10-fold UF for intraspecies differences (UFH) was used to account for variability in
7 susceptibility in human populations. The default value of 10 was selected in the absence of
8 information indicating the degree to which humans may vary in susceptibility to bromobenzene
9 hepatotoxicity.
10 A 3-fold UF was used to account for extrapolating from a subchronic study to chronic
11 exposure scenarios (UFs). Subchronic oral studies in both male and female rats and mice
12 identify the liver as a critical target of bromobenzene toxicity. As discussed in Section 4.5, the
13 liver appears to develop a tolerance to bromobenzene insult during repeated exposure. For
14 example, a single 315 mg/kg oral dose of bromobenzene administered to male rats resulted in
15 marked glutathione depletion, increased serum ALT and SDH concentrations, and observed
16 histopathologic liver lesions (Kluwe et al., 1984). Following 10 days of dosing at 315 mg/kg-
17 day, glutathione depletion was less pronounced, serum ALT and SDH concentrations were no
18 longer increased, and histopathologic liver lesions were no longer detected. NTP (1985a,b) also
19 found increased serum levels of ALT, AST, and SDH were not significantly different from the
20 controls after 90 days of bromobenzene exposure.
21 Although chronic oral or inhalation animal studies are not available for bromobenzene, a
22 chronic oral toxicity study is available for chlorobenzene. As discussed in detail in Section
23 4.5.4, bromobenzene and chlorobenzene exhibit striking similarities in structure, toxicokinetic
24 properties, and critical target of toxicity (liver) in rats and mice. Mice appear to be more
25 sensitive than rats to nonneoplastic hepatotoxicity induced by either bromobenzene or
26 chlorobenzene. The NTP 2-year oral study of chlorobenzene concluded that, nonneoplastic
27 lesions clearly attributable to chlorobenzene were not observed, and identified free-standing
28 NOAELs of 60 and 120 mg/kg-day in male and female mice, respectively (NTP, 1985e). In a
29 similarly-designed subchronic (90-day) oral toxicity study in mice, a NOAEL of 125 and a
30 LOAEL of 250 mg/kg-day were identified in both males and females for chlorobenzene-induced
31 liver lesions (NTP, 1985e). These results suggest that the dose-response relationships for liver
32 effects from subchronic and chronic exposure are similar. It is reasonable to expect such
33 similarities in dose-response relationships for subchronic and chronic exposure to bromobenzene
34 as well, due to the similarity between the two chemicals with respect to chemical reactivity and
35 structure, including similar Pauling electronegativities of chlorine (3.16) and bromine (2.96)
36 (Loudon, 1988). In addition, a study by Chan et al. 2007 suggests that halobenzene congeners
6/7/07 61 DRAFT - DO NOT CITE OR QUOTE
-------
1 interact with cytochrome P450 for oxidation as the primary metabolic activating pathway for
2 toxicity. Mechanistic studies, demonstrating possible hepatic tolerance to repeated
3 bromobenzene exposure (NTP, 1985a,b; Kluwe et al., 1984), further support the similarity
4 between the two compounds. The available data for chronic exposure to chlorobenzene lend
5 support to the database for bromobenzene. Therefore, a UF of 3 was selected to account for
6 extrapolation from subchronic to chronic exposure to bromobenzene.
7 A 10-fold UF was used to account for database deficiencies (UFD). As discussed
8 previously (Section 5.1.1.3), the oral database for bromobenzene lacks well-designed
9 developmental toxicity and multi-generation reproductive toxicity studies. Therefore, the default
10 value of 10 for database deficiencies was not reduced.
11 The chronic RfD for bromobenzene was calculated as follows:
12
13 Chronic RfD = (average BMDL x 5/7) + UF
14 = (25 mg/kg-day x 5/7) - 3000
15 =17.8 mg/kg-day - 3000
16 = 0.006 mg/kg-day (rounded to one significant figure)
17
18 5.1.3. Previous Oral Assessment
19 An RfD was not previously available on IRIS.
20
21 5.2. INHALATION REFERENCE CONCENTRATION (RfC)
22 5.2.1. Subchronic Inhalation RfC
23 5.2.1.1. Choice of Principal Study and Critical Effect - with Rationale and Justification
24 As discussed in Section 4.6.2, there are no available reports of health effects in humans
25 following inhalation exposure to bromobenzene. The toxicity database for repeated inhalation
26 exposure in laboratory animals consists of two 13-week studies, one in rats (NTP, 1985c) and
27 one in mice (NTP, 1985d). No chronic-duration toxicity, reproductive toxicity, or developmental
28 toxicity studies are available.
29 An adverse effect level was not identified in the 13-week inhalation study in male and
30 female Fischer 344/N rats repeatedly exposed to bromobenzene vapor concentrations as high as
31 300 ppm (NTP, 1985c). Significantly increased mean liver weights in 100- and 300-ppm male
32 and female rats may be indicators of an adaptive liver effect of questionable toxicological
33 significance in the absence of more overt toxicity, e.g., liver lesions or necrosis. It should be
34 noted also that this finding is in general agreement with the available oral studies in rats (NTP,
35 1985a) indicating that, unlike mice, this species does not exhibit overt liver toxicity following
36 bromobenzene exposure. Cortical tubular regeneration in the kidney of male rats appeared to be
6/7/07 62 DRAFT - DO NOT CITE OR QUOTE
-------
1 slightly more pronounced in severity in 300-ppm male rats, compared to controls. However, a
2 statistically significant effect on incidence or severity of this kidney lesion could not be
3 discerned. Therefore, this study is not selected for deriving the subchronic RfC.
4 The liver was the most sensitive toxicity target in female B6C3F1 mice exposed to
5 bromobenzene vapors for 6 hours/day, 5 days/week for 13 weeks. Treatment-related
6 significantly increased liver weights were seen in male mice at exposure concentrations >100
7 ppm and in all exposure groups of female mice (including the 50 ppm level). A significantly
8 increased incidence of cytomegaly was observed in 300-ppm female mice (10/10 versus 0/10
9 controls). Necrosis was noted in 5/10 of the 300-ppm female mice, but the incidence of this
10 lesion was not significantly greater than the incidence in controls (2/10). In the 90-day oral
11 studies of rats and mice discussed earlier (NTP, 1985a,b), significantly increased incidences of
12 cytomegaly were observed at doses equal to or slightly lower than those eliciting significantly
13 increased incidences of necrosis. Therefore, it is reasonable to expect that somewhat higher
14 exposure levels in the 90-day inhalation studies (NTP, 1985c,d) would have also resulted in
15 hepatocellular necrosis in the female mice. The 300-ppm exposure level may represent an effect
16 level in female mice that is near the threshold for bromobenzene hepatotoxicity. Therefore, the
17 treatment-related increased occurrence of cytomegaly and increased liver weight may provide
18 early indication of liver toxicity that could occur at higher levels of exposure. For these reasons,
19 the subchronic inhalation study in mice (NTP, 1985d) was selected as the principal study and the
20 increased occurrence of cytomegaly and increased absolute and relative liver weight in female
21 mice was selected as potential critical effects for deriving the subchronic RfC.
22 Other effects in rats and mice were considered for the critical effect but were discounted.
23 In rats, renal histopathology was associated with bromobenzene only at the highest exposure
24 level tested (300 ppm) (NTP, 1985c). Although this lesion was observed in all male rats of the
25 highest exposure group, it was also noted (albeit in slightly lesser severity) in all control males.
26 The increased severity of the renal lesion (cortical tubular regeneration without observable
27 degeneration or necrosis) at the highest exposure level (300 ppm) may represent a treatment-
28 related renal effect in the male rats. However, the Pathology Working Group considered this
29 effect to be mild in all rats in the high-exposure group (NTP, 1986b). Exposure of female rats at
30 levels up to and including 300 ppm did not result in adverse renal effects. Evidence of renal
31 effects was not detected in male or female mice at exposure concentrations up to and including
32 the highest level tested (300 ppm for females; 100 ppm for males). Comprehensive
33 histopathologic examinations of all major tissues and organs in the subchronic inhalation studies
34 of rats and mice revealed no clear evidence of exposure-related lesions at sites other than the
35 kidney (rats) and liver (mice).
36
6/7/07 63 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
5.2.1.2. Methods of Analysis - Including Models (PBPK, BMD, etc.)
Available models in U.S. EPA BMDS version 1.3.2 were fit to the liver lesion
(cytomegaly) data for female B6C3F1 mice and to absolute liver weight and liver-to-body
weight data for male and female B6C3F1 mice from the 90-day inhalation studies (NTP,
1985c,d). Modeling results are presented in Appendix C.
Table 5-9 presents incidence data for microscopically detected cytomegaly and necrosis
in the centrilobular region of the liver in female mice exposed to bromobenzene vapors for 6
hours/day, 5 days/week for 13 weeks (NTP, 1985d). Cytomegaly was the lesion used for BMD
analysis because the Pathology Working Group (NTP, 1986b) agreed with the diagnosis of
cytomegaly, hepatic necrosis, and mineralization in the 300-ppm group, but considered necrosis
and inflammation in the liver of the female mice to be minimal or not present in the 100-ppm or
lower exposure groups. Based on statements of the original study pathologist, quality assurance
pathologist, and the Pathology Working Group, hepatic necrosis and associated effects observed
in the 300-ppm female mice were apparently distinguishable from the necrosis, inflammation,
and mineralization observed in some of the control, 10-, 30-, and 100-ppm female mice. In a
summary statement, the Pathology Working Group (NTP, 1986b) considered the 100-ppm
exposure level to represent a NOAEL for liver effects in the female mice. Regardless, the
statistically significant increase in liver weight at lower doses may be indicative of liver toxicity
in this study. Given the available data sets, it is difficult to determine the region of the dose-
response curve where precursor effects for liver toxicity might occur.
Table 5-9. Incidences of female B6C3F1 mice with cytomegaly in the centrilobular region
of the liver following inhalation exposure to bromobenzene vapors for 6 hours/day, 5
days/week for 13 weeks
Lesion
Cytomegaly
Necrosis
Inflammation
Mineralization
Exposure concentration (ppm)
0
0/10
2/10
4/10
0/10
10
0/10
1/10
3/10
0/10
30
0/10
0/10
2/10
0/10
100
2/10
2/10
2/10
0/10
300
10/10*
5/10
2/10
2/10
22
23
24
25
26
27
28
29
* Statistically significantly different from control incidences according to Fisher's exact test
(p<0.05), performed by Syracuse Research Corporation
Source: NTP (1985d)
Consideration was given to using a NOAEL/LOAEL approach for the cytomegaly data
set since there is little change in effect until a dose of 100 ppm. However, it was decided that the
use of the entire dataset in a BMD modeling approach would be a more sound method since the
6/7/07
64
DRAFT - DO NOT CITE OR QUOTE
-------
1 curve was sigmoidal in shape. It was expected that a number of sigmoidal models would fit such
2 data adequately and equivalently (e.g., gamma, probit, logistic, higher degree multistage). As a
3 consequence, considerable uncertainty about the 'best' model among sigmoidal models is
4 expected.
5 Sigmoidal models and two non-sigmoidal models (quantal quadratic and quantal linear)
6 in the U.S. EPA BMDS (version 1.3.2.) were fit to the data in Table 5-9. Modeling results,
7 presented in Table 5-10, show that: (1) all sigmoidal models provided excellent fit to the data (as
8 expected due to the nature of the data); (2) the non-sigmoidal models provided poorer fits to the
9 data; and (3) all sigmoidal models provided similar estimates of BMCios (ranging from about 77
10 ppm to 97 ppm, a 1.3-fold range) and BMCLioS (ranging from about 40 ppm to 60 ppm, a
11 1.5-fold range). The conventional BMR of 10% extra risk (U.S. EPA, 2000c) was selected
12 because the small group sizes (n=10) in the principal study preclude selecting a lower benchmark
13 risk level. Following U.S. EPA (2000c) guidance for selecting models for point of departure
14 computation, the model with the best fit and the lowest AIC is selected to calculate the BMCL
15 which in this case corresponds to the log-logistic and gamma models (Table 5-10). The
16 BMCLioS from these best-fitting models (from the log-logistic and gamma models) were
17 averaged (55 ppm) to arrive at the point of departure for deriving the RfC, as per U. S. EPA
18 (2000c) guidance. Table 5-11 shows BMCs and BMCLs associated with 10, 5, and 1% extra
19 risk levels.
20 The data for absolute liver weight and liver-to-body weight ratios (relative liver weight)
21 for male and female mice are shown in Table 5-12. Although a significantly increased liver-to-
22 body weight ratio was observed in 100-ppm male mice, there was no evidence of bromobenzene-
23 induced histopathologic liver lesions. Therefore, the male mouse liver weight data were not
24 modeled.
25 All continuous variable models in the U.S. EPA BMDS (version 1.3.2.) were fit to the
26 absolute and relative liver weight data for female mice. As shown in Table 5-13, all models
27 provided adequate fits to the data for absolute liver weight and liver-to-body weight ratio in
28 female B6C3F1 mice as assessed by a chi-square goodness-of-fit test. Second-degree
29 polynomial models provided the best fits for both variables as determined by the AIC (Table
30 5-13). One standard deviation change from the control mean corresponds to an excess risk of
31 approximately 10% for the proportion of individuals above the 98th percentile (or below the 2nd
32 percentile) of the control distribution for normally distributed effects (see Appendix C).
33 Predicted BMCisd values were 52.38 ppm for absolute liver weight and 52.42 ppm for relative
34 liver weight; associated 95% lower confidence limits (BMCLisds) were 33.51 ppm for absolute
35 liver weight and 33.90 ppm for relative liver weight (see Table 5-13). A 0.5 standard deviation
6/7/07 65 DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-10. BMC modeling results for the incidence of liver cytomegaly in female B6C3F1
mice exposed to bromobenzene vapors 6 hours/day, 5 days/week for 13 weeks
Model
Log-logistic21
Gammab
Multi-stage0
Weibullb
Log-probita
Logistic
Probit
Quantal quadratic
Quantal linear
BMCio
(ppm)
95.59
89.24
77.09
92.34
92.95
96.75
93.71
55.15
21.38
BMCL10
(ppm)
58.73
51.42
40.33
47.08
57.45
59.75
54.94
40.15
13.18
x2/>-value
1.00
1.00
0.999
1.00
1.00
1.00
1.00
0.87
0.16
AIC
12.01
12.01
12.17
14.01
14.01
14.01
14.01
14.05
22.78
J
4
5
6
7
aSlope restricted to >1
bRestrict power > = 1
GRestrict betas > = 0; degree of polynomial = 3 (maximum degree restricted to #dose groups
minus 2)
Source: NTP (1985d)
Table 5-11. BMCs and BMCLs predicted from the log-logistic and gamma models for 10,
5, and 1% extra risk for hepatocellular cytomegaly in female B6C3F1 mice exposed to
bromobenzene vapors for 6 hours/day, 5 days/week for 13 weeks
10% Extra risk
BMCio
BMCLio
5% Extra risk
BMCos
BMCLos
1% Extra risk
BMCoi
BMCLoi
Log-logistic model
95.59
58.73
91.71
46.09
83.67
26.47
Gamma model
89.24
51.42
80.98
38.52
66.93
20.53
9
10
Source: NTP (1985d)
6/7/07
66
DRAFT - DO NOT CITE OR QUOTE
-------
Table 5-12. Data for absolute liver weight and liver-to-body weight ratio for male and
female B6C3F1 mice following inhalation exposure to bromobenzene vapors for 6
hours/day, 5 days/week for 13 weeks (mean +/- standard deviation)
Exposure concentration (ppm)
0
10
30
100
300
Absolute liver weight (grams)
Male
Female
1.84 + 0.21
1.43+0.15
1.73 + 0.14
1.52 + 0.09
1.73+0.18
1.54a + 0.07
1.87 + 0.21
1.68a + 0.10
2.37b + 0.21
Liver-to-body weight ratio (relative liver weight)
Male
Female
50.71 + 3.66
52.00 + 3.22
51.86 + 3.57
55.25a + 3.49
51.57 + 2.78
54.66a+1.80
54.28a + 2.42
59.37b + 3.43
79.73b + 5.27
J
4
5
6
aStatistically significantly different from controls (p<0.05) based on Student's two-tailed t-test
bOutside 3 standard deviations from the control mean
Source: NTP (1985d)
Table 5-13. Model output for increased absolute liver weight and liver-to-body weight
ratio in female B6C3F1 mice following inhalation exposure to bromobenzene for 6
hours/day, 5 days/week for 13 weeks
Model3
BMClsd (ppm)
BMCLlsd (ppm)
x2 p-value
AIC
Absolute liver weightb
Linear
Polynomial (2°)
Polynomial (3°)
Power
35.24
52.38
32.67
56.82
28.39
33.51
14.45
32.56
0.1838
0.3922
0.2891
0.2901
-150.18
-151.16
-149.91
-150.55
Liver-to-body weight ratiob
Linear
Polynomial (2°)
Polynomial (3°)
Power
41.03
52.42
45.52
57.55
34.52
33.90
18.56
34.12
0.08619
0.09284
0.09301
0.07211
183.82
182.19
184.05
182.77
9
10
11
12
aStatistical tests indicated that variances were not constant across exposure groups. Model
results are for non-homogeneous variance, with the exception of the linear and 3-degree
polynomial models for liver-to-body weight ratio.
bModeled as a continuous variable using one standard deviation as the BMR.
Source: NTP (1985d)
6/7/07
67
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
(O.Ssd) change from the control mean was also considered as a potential BMR for absolute liver
weight and liver-to-body weight ratio (see Table 5-14).
Table 5-14. The second-degree polynomial model-estimated BMCs and BMCLs associated
with 1 and 0.5 standard deviation extra risk for increased absolute liver weight and liver-
to-body weight ratio in female B6C3F1 mice exposed to bromobenzene vapors for 6
hours/day, 5 days/week for 90 days
Endpoint
Absolute liver weight
Liver-to-body weight ratio
BMCs and BMCLs (ppm)
BMClsd
52.38
52.42
BMCLisd
33.51
33.90
BMCossd
27.65
27.76
BMCLo5sd
16.83
17.08
5
6
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Source: NTP (1985d)
The BMDLio for absolute and relative liver weight changes in female mice was 34 ppm.
The BMDLio for the incidence of cytomegaly was 55 ppm derived from an average of the
BMDLioS from the two best-fitting models. There is some uncertainty associated with the choice
of the critical effect and the point of departure. Although cytomegaly in the absence of necrosis
or other indicators of degenerative effects may represent an adaptive hepatic effect rather than an
adverse effect, necrosis and mineralization observed in livers of some of the 300-ppm female
mice was considered by the Pathology Working Group (NTP, 1986b) to be an exposure-related
effect. Therefore, the 300-ppm exposure level may represent an effect level in female mice that
is near the threshold of significantly detectable bromobenzene hepatotoxicity. For this reason,
the average BMCLio of 55 ppm (from the log-logistic and gamma models) for cytomegaly in
female mice was selected as the point of departure to derive the subchronic RfC for
bromobenzene. There is less uncertainty in choosing this endpoint over the increase in liver
weight due to the lack of directly observable statistically significant toxicity at higher doses.
5.2.1.3. Subchronic RfC Derivation - Including Application of Uncertainty Factors (UFs)
Following U.S. EPA (1994b) methodology, the human equivalent concentration (FtEC)
for an extra respiratory effect produced by a category 3 gas, such as bromobenzene (not highly
water soluble or reactive in the respiratory tract, the liver as the critical extrarespiratory target), is
calculated by multiplying the duration-adjusted BMCL or NOAEL by the ratio of the blood:gas
partition coefficients in animals and humans [(Hb/g)A / Hb/g)n]. Because bromobenzene blood:gas
partition coefficients are not available for humans or mice, a default value of 1 is used for this
ratio. The BMCLio of 55 ppm for hepatocellular cytomegaly in female mice was converted to
353.2 mg/m3 (55 ppm x MW[157] / 24.45 = 353.2 mg/m3), which was then converted to a
6/7/07
68
DRAFT - DO NOT CITE OR QUOTE
-------
3
1 continuous exposure basis (353.2 mg/m x 6/24 hr x 5/7 days = 63 mg/m ) and multiplied by a
3
2 default blood:gas partition coefficient ratio of 1 to obtain the BMCLionEc of 63 mg/m. The
3 BMCLioHEc of 63 mg/m3 was divided by a total UF of 300. The UF consists of three areas of
4 uncertainty: (1) interspecies extrapolation, (2) interindividual human variability, and (3) database
5 deficiencies.
6 A factor of 3 was selected to account for uncertainties in extrapolating from mice to
7 humans (UFA). Although no human data are available, it appears reasonable to assume that
8 hepatic effects observed in female mice would be relevant to humans. The default value of 10
9 was reduced to 3 because dosimetric adjustment methodology (U.S. EPA, 1994b) for a category
10 gas 3, with a default value of 1 for the ratio of the blood:gas partition coefficients in animals and
1 1 humans [(Hb/g)A / Hb/g)n]), was applied to derive the BMCLioHEc point of departure for the
12 sub chronic RfC.
13 A default 10-fold UF was selected to account for interindividual toxicokinetic and
14 toxicodynamic variability in humans (UFH). Although hepatotoxicity was observed only in
15 female mice, a 300-ppm (1926 mg/m3) group of male mice was not included in the study. Due to
16 the lack of conclusive information concerning gender-specific differences in bromobenzene
17 hepatotoxicity following inhalation exposure, as well as the lack of data concerning the extent of
18 variation in sensitivity to bromobenzene within the human population, the default value of 10
19 was not reduced.
20 A 10-fold UF was used to account for database deficiencies (UFD). Subchronic studies in
21 rats and mice are available. Developmental toxicity and multi -generation reproductive toxicity
22 studies are lacking. Therefore, the default value of 10 was not reduced.
23 The subchronic RfC for bromobenzene was calculated as follows:
24
25 Subchronic RfC = BMCLiOHEc - UF
26 = 63 mg/m3 H- 300
27 = 0.2 mg/m3 (rounded to one significant figure)
28
29 5.2.2. Chronic Inhalation RfC
30 5.2.2.1. Choice of Principal Study and Critical Effect - with Rationale and Justification
31 As discussed in Section 4.6.2, there are no available reports of health effects in humans
32 following inhalation exposure to bromobenzene. The toxicity database for repeated inhalation
33 exposure in laboratory animals consists of two 13-week studies, one in rats (NTP, 1985c) and
34 one in mice (NTP, 1985d). No chronic-duration toxicity, reproductive toxicity, or developmental
35 toxicity studies are available.
6/7/07 69 DRAFT - DO NOT CITE OR QUOTE
-------
1 The choices of principal study and critical effect for development of the chronic RfC for
2 bromobenzene are the same as those described for the development of a subchronic RfC (see
3 Section 5.2.1.1). The increase in incidence of cytomegaly and the increase in absolute and
4 relative liver weight in female mice (NTP, 1985d) were considered in the selection of the critical
5 effect for development of the subchronic RfC for bromobenzene.
6
7 5.2.2.2. Methods of Analysis - Including Models (PBPK, BMD, etc.)
8 The methods of analysis used to derive the subchronic RfC for bromobenzene apply to
9 the derivation of the chronic RfC as well (see Section 5.2.1.2).
10
11 5.2.2.3. Chronic RfC Derivation - Including Application of Uncertainty Factors (UFs)
12 As described in detail in Section 5.2.1.3, the average BMCLio of 55 ppm for cytomegaly
13 in female mice was selected as the point of departure to derive the subchronic RfC for
14 bromobenzene. The same point of departure was used to derive the chronic RfC. The BMCLio
15 of 55 ppm for hepatocellular cytomegaly in female mice was converted to a BMCLionEc of 63
16 mg/m3 (see Section 5.2.1.3 for details regarding conversion to the HEC). The BMCLionEc of 63
17 mg/m3 was divided by a total UF of 1000. The UF consists of four areas of uncertainty: (1)
18 interspecies extrapolation, (2) interindividual human variability, (3) extrapolation from
19 subchronic- to chronic-duration exposure, and (4) database deficiencies.
20 A factor of 3 was selected to account for uncertainties in extrapolating from mice to
21 humans (UFA). Although no human data are available, it appears reasonable to assume that
22 hepatic effects observed in female mice would be relevant to humans. The default value of 10
23 was reduced to 3 because dosimetric adjustment methodology (U.S. EPA, 1994b) for a category
24 gas 3, with a default value of 1 for the ratio of the blood:gas partition coefficients in animals and
25 humans [(Hb/g)A / Hb/g)n]), was applied to derive the BMCLionEc point of departure for the
26 chronic RfC.
27 A 10-fold UF was selected to account for interindividual toxicokinetic and
28 toxicodynamic variability in humans (UFH). Although hepatotoxicity was observed only in
29 female mice, a 300-ppm (1926 mg/m3) group of male mice was not included in the study. Due to
30 the lack of conclusive information concerning gender-specific differences in bromobenzene
31 hepatotoxicity following inhalation exposure, as well as the lack of data concerning the extent of
32 variation in sensitivity to bromobenzene within the human population, the default value of 10
33 was not reduced.
34 A 3-fold UF was used to account for extrapolating from a subchronic study to chronic
35 exposure scenarios (UFs). Subchronic oral studies in both male and female rats and mice
36 identify the liver as a critical target of bromobenzene toxicity. A subchronic inhalation study in
6/7/07 70 DRAFT - DO NOT CITE OR QUOTE
-------
1 mice provides supporting evidence for the hepatotoxicity of bromobenzene. There are no
2 chronic exposure studies for bromobenzene, but results of chronic exposure to chlorobenzene
3 indicate the subchronic and chronic dose-responses are similar (see Section 5.1.2.3). It is
4 reasonable to expect the subchronic and chronic dose-responses from exposure to bromobenzene
5 to be similar as well. Therefore, a UF of 3 was selected to account for extrapolation from
6 subchronic to chronic exposure to bromobenzene.
7 A 10-fold UF was used to account for database deficiencies (UFD). Subchronic studies in
8 rats and mice are available. Developmental toxicity and multi-generation reproductive toxicity
9 studies are lacking. Therefore, the default value of 10 was not reduced.
10 The chronic RfC for bromobenzene was calculated as follows:
11
12 Chronic RfC = BMCLiOHEc - UF
13 =63 mg/m3 + 1000
14 = 0.06 mg/m3 (rounded to one significant figure)
15
16 5.3. CANCER ASSESSMENT
17 No studies of cancer risks of humans or cancer bioassays in animals exposed to
18 bromobenzene were located. Bromobenzene was not mutagenic in the Ames assay and did not
19 consistently produce marked cytogenetic effects in vitro with mammalian cells, even in the
20 presence of rat liver S-9 preparations. Bromobenzene induced micronuclei in bone marrow of
21 mice given acute oral doses of 125 mg/kg and was bound to DNA and RNA following
22 intraperitoneal injection. Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA,
23 2005a), there is "inadequate information to assess the carcinogenic potential" of brombenzene.
24
6/7/07 71 DRAFT - DO NOT CITE OR QUOTE
-------
1 6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD
2 AND DOSE RESPONSE
3
4
5 6.1. HUMAN HAZARD POTENTIAL
6 No human data are available for health effects following exposure to bromobenzene by
7 any exposure route for any duration. Animal studies demonstrate that relatively high single oral
8 doses (>785 mg/kg-day) of bromobenzene elicit hepatic, renal, and pulmonary effects (Becher et
9 al., 1989; Casini et al., 1986; Forkert, 1985; Kluwe et al., 1984; Rush et al., 1984; Roth, 1981;
10 Reid et al., 1973; Patrick and Kennedy, 1964). Hepatic effects have been elicited in mice
11 following a single 4-hour exposure to bromobenzene vapors at a concentration of 250 ppm; a
12 higher concentration (1000 ppm) resulted in lung lesions (Becher et al., 1989). Subchronic-
13 duration (90-day) oral and inhalation studies in rats and mice identify the liver as the most
14 sensitive target of bromobenzene toxicity (NTP, 1985a,b,c,d). The threshold for renal effects
15 appears to be somewhat higher than that for hepatic effects. Bromobenzene has not been
16 assessed for reproductive or developmental toxicity or for carcinogenicity in animals. It is
17 reasonable to assume that bromobenzene-induced human health effects would be similar to those
18 demonstrated in laboratory animals.
19 Results of additional well-designed studies of bromobenzene toxicity in animals would
20 be helpful in assessing the hazards associated with exposure to bromobenzene. The chronic oral
21 and inhalation toxicity of bromobenzene should be assessed in two animal species at exposure
22 concentrations that include a clear advese effect level. In addition, as discussed in Section 4.6.1,
23 a well-designed developmental toxicity study and a multi-generation reproductive toxicity study
24 should be performed using the oral and/or inhalation exposure route.
25 Following EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), there is
26 "inadequate information to assess the carcinogenic potential" of bromobenzene due to the lack of
27 data on the possible carcinogenicity of bromobenzene in humans or animals. Bromobenzene was
28 not mutagenic in bacterial assays and did not consistently produce marked cytogenetic effects in
29 vitro with mammalian cells, even in the presence of rat liver metabolizing preparations.
30 Bromobenzene increased formation of micronucleated polychromatic erythrocytes in bone
31 marrow of mice given acute oral doses of 125 mg/kg and was bound to DNA and RNA
32 following intraperitoneal injection. The available genotoxicity data, therefore, provide only
33 limited evidence of bromobenzene genotoxicity.
34
6/7/07 72 DRAFT - DO NOT CITE OR QUOTE
-------
1 6.2. DOSE RESPONSE
2 6.2.1. Noncancer/Oral
3 The liver was selected as the critical target of bromobenzene toxicity because it is the
4 most sensitive indicator of bromobenzene toxicity. BMD analysis of the incidence data for
5 combined liver lesions (centrilobular inflammation, cytomegaly, mineralization or necrosis),
6 absolute liver weight, liver-to-body weight ratio, and SDH levels in rats and mice (NTP,
7 1985a,b) indicated that female mice have a lower point of departure than male mice or male or
8 female rats. Liver toxicity defined as the combined incidence of hepatic lesions and liver weight
9 changes in female mice was selected as the critical effect for deriving the chronic and
10 sub chronic RfD.
11 The average of the lower 95% confidence limit for a BMD of 10% extra risk for liver
12 weight changes (BMDLio = 25.8 mg/kg-day) and combined liver lesions (24.8 mg/kg-day) was
13 used as the point of departure. The average BMDL of 25 mg/kg-day was adjusted to account for
14 daily exposure (25 mg/kg-day x 5 days/7 days = 17.8 mg/kg-day). The subchronic RfD was
15 derived by dividing the average BMDLAoj of 17.8 mg/kg-day by a composite UF of 1000 to
16 account for three areas of uncertainty (10 for interspecies extrapolation, 10 for interindividual
17 human variability, and 10 for database deficiencies). The resulting RfD is 17.8 mg/kg-day +
18 1000 = 0.02 mg/kg-day. The derivation of the chronic RfD included an additional UF of 3 to
19 account for extrapolation from a subchronic study to chronic exposure scenarios for a composite
20 UF of 3000. The resulting chronic RfD is 17.8 mg/kg-day - 3000 = 0.006 mg/kg-day.
21
22 6.2.2. Noncancer/Inhalation
23 The NTP 90-day inhalation studies in rats and mice provided adequate exposure-response
24 data for bromobenzene (NTP, 1985a,b). The liver was selected as the critical target of
25 bromobenzene toxicity because the liver was the only target that provided clear evidence of
26 bromobenzene toxicity. Significantly increased incidences of cytomegaly were observed in
27 female mice of the highest exposure level (300 ppm). Incidences of histopathologic liver lesions
28 in bromobenzene-exposed groups of male and female rats and male mice were not significantly
29 different from controls up to and including the highest exposure level tested (300 ppm in male
30 and female rats, 100 ppm in male mice). Significantly increased liver weights were noted in
31 100- and 300-ppm male and female rats, 100-ppm male mice, and all bromobenzene-exposed
32 groups (50-300 ppm) of female mice. The incidence of hepatocellular cytomegaly in female
33 mice was selected as the critical effect for deriving the chronic and subchronic RfC.
34 The average BMCLio of 55 ppm (from the log-logistic and gamma models) for
35 cytomegaly in female mice was selected as the point of departure. The BMCLio was converted
36 to 353.2 mg/m3 (55 ppm x MW[157] / 24.45 = 353.2 mg/m3), which was then converted to a
6/7/07 73 DRAFT - DO NOT CITE OR QUOTE
-------
3
I continuous exposure basis (353.2 mg/m x 6/24 hours x 5/7 days = 63 mg/m ) and multiplied by
3
2 a default blood:gas partition coefficient ratio of 1 to obtain the BMCLionEc of 63 mg/m. The
3 subchronic RfC was derived by dividing the BMCLionEc of 63 mg/m3 by a composite UF of 300
4 to account for three areas of uncertainty (3 for interspecies extrapolation using dosimetric
5 conversion, 10 for interindividual human variability, and 10 for database deficiencies). The
6 resulting subchronic RfC is 63 mg/m3 + 300 = 0.2 mg/m3. The derivation of the chronic RfC
7 included an additional UF of 3 to account for extrapolation from a subchronic study to chronic
8 exposure scenarios. The resulting chronic RfC is 63 mg/m3 + 1000 = 0.06 mg/m3.
9 6.2.3. Cancer/Oral
10 The lack of cancer studies in humans and cancer bioassays in animals precludes a cancer
1 1 dose-response assessment for bromobenzene.
12
13 6.2.4. Cancer/Inhalation
14 The lack of cancer studies in humans and cancer bioassays in animals precludes a cancer
15 dose-response assessment for bromobenzene.
16
6/7/07 74 DRAFT - DO NOT CITE OR QUOTE
-------
1 7. REFERENCES
2
3
4 Aarstad, K; Becker, R; Dahl, J; Dybing, E; Nilsen, OG. (1990) Short term inhalation of
5 bromobenzene: Methodology and absorption characteristics in mouse, rat, and rabbit. Pharmacol
6 Toxicol 67:284-287.
7 Aniya, Y; McLenithan, JC; Anders, MW. (1988) Isozyme selective arylation of cytosolic
8 glutathion S-transferase by [14C]bromobenzene metabolites. Biochem Pharmacol
9 37(2):251-257.
10 Atkinson, R. (1989) Kinetics and mechanisms of the gas-phase reactions of the hydroxyl radical
11 with organic compounds. J Phys Chem Ref Data. Monograph 1.
12 Bambal, RB; Hanzlik, RP. (1995) Bromobenzene 3,4-oxide alkylates histidine and lysine side
13 chains of rat liver proteins in vivo. Chem Res Toxicol 8(5):729-735.
14 Barber, LB; Thurman, EM; Schroeder, MP. (1988) Long-term fate of organic micropollutants in
15 sewage-contaminated groundwater. Environ Sci Technol 22:205-211.
16 Becher, R; Dahl, JE; Aarstad, K; Nilsen, OG; Dybing, E. (1989) Lung and liver damage in mice
17 after bromobenzene inhalation: effects of enzyme inducers. Inhal Toxicol 1:181-196.
18 Bidleman, TF. (1988) Atmospheric processes. Wet and dry deposition of organic compounds are
19 controlled by their vapor-particle partitioning. Environ Sci Technol 22(4):361-367.
20 Brodie, BB; Reid, WD; Cho, AK; Sipes, G;Krishna, G; Gillette, JR. (1971) Possible mechanism
21 of liver necrosis caused by aromatic organic compounds. Proc Natl Acad Sci 68:160-164.
22 Brodzinsky, R; Singh, HB. (1982) Volatile Organic Chemicals in the Atmosphere: An
23 Assessment of Available Data. Prepared by the Environmental Sciences Research Laboratory,
24 U.S. Environmental Protection Agency, Research Triangle Park, NC. EPA/600/S3-83-027.
25 Brondeau, MT; Bonnet, P; Guenier, JP; de Ceaurriz, J. (1983) Short-term inhalation test for
26 evaluating industrial hepatotoxicants in rats. Toxicol Lett 19:139-146.
27 Brondeau, MT; Ban, M; Bonnet, P; Guenier, JP; de Ceaurriz, J. (1986) Concentration-related
28 changes in blood and tissue parameters of hepatotoxicity and their interdependence in rats
29 exposed to bromobenzene and 1,2-dichlorobenzene. Toxicol Lett 31:159-166.
30 Budavari, S. (2001) Bromobenzene. In: O'Neil, MJ; Smith, A; Heckelman, PE; et al., Eds. The
31 Merck Index, 13th ed. Rahway, NJ: Merck & Co., Inc. pp. 234-235.
32 Casini, AF; Pompella, A; Comporti, M. (1984) Glutathione depletion, lipid peroxidation, and
33 liver necrosis following bromobenzene and iodobenzene intoxication. Toxicol Pathol
34 12:295-299.
6/8/07 75 DRAFT - DO NOT CITE OR QUOTE
-------
1 Casini, AF; Pompella, A; Comporti, M. (1985) Liver glutathione depletion induced by
2 bromobenzene, iodobenzene, and diethylmaleate poisoning and its relation to lipid peroxidation
3 and necrosis. Am J Pathol 118:225-237.
4 Casini, AF; Ferrali, M; Pompella, A; Maellaro, E; Comparti, M. (1986) Lipid peroxidation and
5 cellular damage in extrahepatic tissues of bromobenzene-intoxicated mice. Am J Pathol
6 123(3):520-531.
7 Chakrabarti, S; Brodeur, J. (1984) Dose-dependent metabolic excretion of bromobenzene and its
8 possible relationship to hepatotoxicity in rats. J Toxicol Environ Health 14:379-391.
9 Chan, K; Jensen, NS; Silber, PM; O'Brien, PJ. (2007) Structure-activity relationships for
10 halobenzene induced cytotoxicity in rat and human hepatocytes. Chem Biol Interact
11 165(3):165-174.
12 Chiou, CT; Freed, VH; Schmedding, DW; Kohnert, RL. (1977) Partition coefficient and
13 bioaccumulation of selected organic chemicals. Environ Sci Technol ll(5):475-478.
14 CITI (Chemicals Inspection and Testing Institute). (1992) Biodegradation and bioaccumulation.
15 Data of existing chemicals based on the CSCL Japan. Japan Chemical Industry Ecological-
16 Toxicology & Information Center.
17 Colacci, A; Arfellini, G; Mazzullo, M; Prodi, G;Grilli, S. (1985) The covalent binding of
18 bromobenzene with nucleic acids. Toxicol Pathol 13:276-282.
19 Dahl, JE; Becher, R; Aarstad, K; Nilson, OG; Dybing, E. (1990) Species differences in short
20 term toxicity from inhalation exposure to bromobenzene. Arch Toxicol 64:370-376.
21 Forkert, PG. (1985) Bromobenzene causes Clara cell damage in mice. Can J Physiol Pharmacol
22 63:1480-1484.
23 Freitag, D; Ballhorn, L; Geyer, H; Korte, F. (1985) Environmental hazard profile of organic
24 chemicals. An experimental method for the assessment of the behavior of organic chemicals in
25 the ecosphere by means of simple laboratory tests with 14C labelled chemicals. Chemosphere
26 14(10):1589-1616.
27 Galloway, SM; Armstrong, MJ; Reuben, C; et al. (1987) Chromosome aberrations and sister
28 chromatid exchanges in Chinese hamster ovary cells: evaluations of 108 chemicals. Environ Mol
29 Mutagen 10:1-175.
30 Girault, I; Rougier, N; Chesne, C; et al. (2005) Simultaneous measurement of 23 isoforms from
31 the human cytochrome P450 families 1-2 by quantitative reverse transcriptase-polymerase chain
32 reaction. DrugMetab Dispos 33:1803-1810.
33 Green, MD; Shires, TK; Fischer, LJ. (1984) Hepatotoxicity of acetaminophen in neonatal and
34 young rats. 1. Age-related changes in susceptibility. Toxicol Appl Pharmacol 74:116-124.
6/7/07 76 DRAFT - DO NOT CITE OR QUOTE
-------
1 Hansch, C; Leo, A; Hoekman, D. (1995) Exploring QSAR. Hydrophobic, electronic, and steric
2 constants. In: Heller, SR, Ed. ACS professional reference book. Washington, DC: American
3 Chemical Society.
4 Heijne, WHM; Stierum, RH; Slijper, M; van Bladeren, PJ; van Ommen, B. (2003)
5 Toxicogenomics of bromobenzene hepatotoxicity: A combined transcriptomics and proteomics
6 approach. Biochem Pharmacol 65:857-875.
7 Heijne, WHM; Slitt, AL; van Bladeren, PJ; et al. (2004) Bromobenzene-induced hepatotoxicity
8 at the transcriptome level. Toxicol Sci 79(2):411-422.
9 Heijne, WHM; Lamers, RJAN; van Bladeren, PJ; Groten, JP; van Nesselrooij, JHJ; van Ommen,
10 B. (2005) Profiles of metaobolites and gene expression in rats with chemically induced hepatic
11 necrosis. Toxicol Pathol 33:425-433.
12 Heikes, DL; Jensen, SR; Fleming-Jones, ME. (1995) Purge and trap extraction with GC-MS
13 determination of volatile organic compounds in table-ready foods. J Agric Food Chem
14 43:2869-2875.
15 Hellman, B. (1993) NIOH and NIOSH basis for an occupational health standard: Chlorobenzene.
16 Arbete och Hal sa 31:1-73 (also published by the U.S. Department of Health and Human Services
17 under a joint agreement with NIOH and NIOSH. Available at
18 http://www.cdc.gov/niosh/pdfs/93-102a.pdf).
19 Herren-Freund, SL; Pereira, MA. (1986) Carcinogenicity of by-products of disinfection in mouse
20 and rat liver. Environ Health Perspect 69:59-65.
21 HSDB (Hazardous Substances Data Bank). (2003) Bromobenzene. Last review dated September
22 19, 1996. National Library of Medicine, National Toxicology Program, Bethesda, MD. Available
23 at http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7HSDB.
24 IBT (Industrial Bio-Test Laboratories). (1977) Teratogenic study with monochlorobenzene in
25 albino rats. Report to Monsanto Company, St. Louis, MO. BTL No. 76-89; IBT No.
26 8533-09115. May 18, 1977.
27 Ito, N; Tsuda, H; Tatematsu, M; et al. (1988) Enhancing effect of various hepatocarcinogens on
28 induction of preneoplastic glutathione S-transferase placental form positive foci in rats - an
29 approach for a new medium-term bioassay system. Carcinogenesis 9:387-394.
30 John, JA; Hayes, WC; Hanley, TR, Jr.; Johnson, KA; Gushow, TS; Rao, KS. (1984) Inhalation
31 teratology study on monochlorobenzene in rats and rabbits. Toxicol Appl Pharmacol
32 76:365-373.
33 Jollow, DJ; Mitchell, JR; Zampaglione, N; Gillette, JR. (1974) Bromobenzene-induced liver
34 necrosis. Protective role of glutathione and evidence for 3,4-bromobenzene oxide as the
35 hepatotoxic metabolite. Pharmacology 11:151-169.
6/7/07 77 DRAFT - DO NOT CITE OR QUOTE
-------
1 Kato, R; Vassanelli, P; Frontino, G; Chiesara, E. (1964) Variation in the activity of microsomal
2 drug-metabolizing enzymes in rats in relation to the age. Biochem Pharmacol 13:1037-1051.
3 Kluwe, WM; Maronpot, RR; Greenwell, A; Harrington, F. (1984) Interactions between
4 bromobenzene dose, glutathione concentrations, and organ toxicities in single- and multiple-
5 treatment studies. Fundam Appl Toxicol 4:1019-1028.
6 Koen, YM; Hanzilk, RP. (2002) Identification of seven proteins in the endoplasmic reticulum as
7 targets for reactive metabolites of bromobenzene. Chem Res Toxicol 15(5):699-706.
8 Koen, YM; Williams, TD; Hanzlik, RP. (2000) Identification of three protein targets for reactive
9 metabolites of bromobenzene in rat liver cytosol. Chem Res Toxicol 13(12): 1326-1335.
10 Krusekopf, S; Roots, I; Hildebrandt, AG; Kleeberg, U. (2003) Time-dependent transcriptional
11 induction of CYP1A1, CYP1A2 and CYP1B1 mRNAs by H+/K+-ATPase inhibitors and other
12 xenobiotics. Xenobiotica 33(2):107-118.
13 Lau, SS; Monks, TJ. (1988) The contribution of bromobenzene to our current understanding of
14 chemically-induced toxicities. Life Sci 42:1259-1269.
15 Lau, SS; Monks, TJ. (1997a) Bromobenzene hepatotoxicity: A paradigm of reactive electrophilic
16 metabolites binding covalently to tissue macromolecules. Is there light at the end of the tunnel?
17 In: McCuskey, R, Ed. Comprehensive Toxicology, Vol. 9: Hepatic and Gastrointestinal Toxicity.
18 Austin, TX: Elsevier-Science, University of Texas, pp. 465-473.
19 Lau, SS; Monks, TJ. (1997b) Bromobenzene nephrotoxicity: A model of metabolism-dependent
20 toxicity. In: Goldstein, RS, Ed. Comprehensive toxicology, Vol 7: Renal Toxicology. Austin,
21 TX: Elsevier-Science, University of Texas, pp. 617-632.
22 Lau, SS; Zannoni, VG. (1979) Hepatic microsomal epoxidation of bromobenzene to phenols and
23 its toxicological implication. Toxicol Appl Pharmacol 50:309-318.
24 Lau, SS; Zannoni, VG. (1981a) Bromobenzene metabolism in the rabbit. Specific forms of
25 cytochrome P-450 involved in 2,3- and 3,4-epoxidation. Mol Pharmacol 20:234-235.
26 Lau, SS; Zannoni, VG. (1981b) Bromobenzene epoxidation leading to binding on
27 marcromolecular protein sites. J Pharmacol Exp Ther 219(2):563-572.
28 Lau, SS; Abrams, GD; Zannoni, VG. (1980) Metabolic activation and detoxification of
29 bromobenzene leading to cytotoxicity. J Pharmacol Exp Ther 214(3):703-214.
30 Lau, SS; Monks, TJ; Gillette, JR. (1984a) Identification of 2-bromohydroquinone as a metabolite
31 of bromobenzene and o-bromophenol: implications for bromobenzene-induced nephrotoxicity. J
32 Pharmacol Exp Ther 230(2):360-366.
33 Lau, SS; Monks, TJ; Greene, KE; Gillette, JR. (1984b) The role of ort/zo-bromophenol in the
34 nephrotoxicity of bromobenzene in rats. Toxicol Appl Pharmacol 72:539-549.
6/7/07 78 DRAFT - DO NOT CITE OR QUOTE
-------
1 Lertratanangkoon, K; Horning, MG. (1987) Bromobenzene metabolism in the rat and guinea pig.
2 Drug Metab Dispos 15(1):1-11.
3 Lertratanangkoon, K; Horning, EC; Horning, MG. (1987) Conversion of bromobenzene to 3-
4 bromophenol: a route to 3- and 4-bromophenol through sulfur-series intermediates derived from
5 the 3,4-oxide. Drug Metab Dispos 15(5):857-867.
6 Lertratanangkoon, K; Horning, EC; Horning, MG. (1993) Pathways of formation of 2-, 3- and
7 4-bromophenol from bromobenzene. Proposed mechanism for C-S lyase reactions of cysteine
8 conjugates. Res Commun Chem Pathol Pharmacol 80(3):259-282.
9 Lewis, RJ. (1997) Bromobenzene. In: Hawley's Condensed Chemical Dictionary, 13th ed. New
10 York, NY: John Wiley & Sons, Inc. pp. 164.
11 Lide, DR. (2000) CRC Handbook of Chemistry and Physics, 81st ed. Washington, DC: CRC
12 Press, pp. 3-30.
13 Loudon, GM. (1988) Organic chemistry, 2nd ed. Menlo Park, California: Benjamin/Cummings
14 Publishing Company, Inc. p. 10.
15 Lyman, WJ; Reehl, WF; Rosenblatt, DH. (1990) Handbook of chemical property estimation
16 methods. Environmental behavior of organic compounds. Washington, DC: American Chemical
17 Society.
18 Madhu, C; Klaassen, CD. (1992) Bromobenzene-glutathione excretion into bile reflects toxic
19 activation of bromobenzene in rats. Toxicol Lett 60(2):224-236.
20 McCann, J; Choi, E; Yamaskai, E; Ames, BN. (1975) Detection of carcinogens as mutagens in
21 the Salmonella/microsome test: assay of 300 chemicals. Proc Nat Acad Sci USA 12:5135-5139.
22 Meister, A. (1982) 5-Oxoprolinuria (pyroglutamic aciduria) and other disorders of the y-glutamyl
23 cycle. In: Stanbury, JB; Wyngaarder, JB; Frederickson, JL; et al., Eds. Metabolic Basis of
24 Inherited Diseases, 5th ed. New York, NY: McGraw-Hill.
25 Merrick, AB; Davies, MH; Schnell, RC. (1986) Effect of sodium selenite upon bromobenzene
26 toxicity in rats. II. Metabolism. Toxicol Appl Pharmacol 83:279-286.
27 Miller, NE; Thomas, D; Billings, RE. (1990) Bromobenzene metabolism in vivo and in vitro.
28 The mechanism of 4-bromocatechol formation. Drug Metab Dispos 18(3):304-308.
29 Minami, K; Saito, T; Narahara, M; Tomita, H; Kato, H; Sugiyama, H; Katoh, M; Nakajima, M;
30 Yokoi, T. (2005) Relationship between hepatic gene expression profiles and hepatotoxicity in
31 five typical hepatotoxicant-administered rats. Toxicol Sci 87(1):296-305.
32 Mitchell, JR; Reid, WD; Christie, B; Moskowitz, J; Krishma, G; Brodie, BB. (1971)
33 Bromobenzene-induced hepatic necrosis: species differences and protection by SKF 525-A. Res
34 Commun Chem Pathol Pharmacol 2(6):877-888.
6/7/07 79 DRAFT - DO NOT CITE OR QUOTE
-------
1 Mohtashamipur, E; Triebel, R; Straeter, H; Norpoth, K. (1987) The bone marrow clastogenicity
2 of eight halogenated benzenes in male NMRI mice. Mutagenesis 2(2): 111-113.
3 Monks, TJ; Hinson, JA; Gillette, JR. (1982) Bromobenzene and p-bromophenol toxicity and
4 covalent binding in vivo. Life Sci 30:841-848.
5 Monks, TJ; Lau, SS; Highet, RJ. (1984a) Formation of nontoxic reactive metabolites of
6 /7-bromophenol: identification of a new glutathione conjugate. Drug Metab Dispos
7 12(4):432-437.
8 Monks, TJ; Lau, SS; Pohl, LR; Gillette, JR. (1984b) The mechanism of formation of o-
9 bromophenol from bromobenzene. Drug Metab Dispos 12(2): 193-198.
10 Monks, TJ; Lau, SS; Highet, RJ; Gillette, JR. (1985) Glutathione conjugates of 2-
11 bromohydroquinone are nephrotoxic. Drug Metab Dispos 13(5):553-559.
12 Nair, RS; Barter, JA; Schroeder, RE; et al. (1987) A two-generation reproduction study with
13 monochlorobenzene vapor in rats. Fund Appl Toxicol 9(4):678-686.
14 Nakamura, S-I; Oda, Y; Shimada, T; Oki, I; Sugimoto, K. (1987) SOS-inducing activity of
15 chemical carcinogens and mutagens in Salmonella typhimurium TA1535/pSK1002: examination
16 with 151 chemicals. MutatRes 192:239-246.
17 National Research Council. (1983) Risk assessment in the federal government: managing the
18 process. Committee on the Institutional Means for Assessment of Risks to Public Health,
19 Commission on Life Sciences, NRC. Washington, DC: National Academy Press.
20 NTP (National Toxicology Program). (1985a) Subchronic gavage study of bromobenzene in rats.
21 National Institutes of Health, National Toxicology Program, Research Triangle Park, NC.
22 [Unpublished study]. April, 1985. (Available by calling EPA's IRIS Hotline at (202)566-1676,
23 by fax at (202)566-1749 or by email at iris@epa.gov).
24 NTP (National Toxicology Program). (1985b) Subchronic gavage study of bromobenzene in
25 mice. National Institutes of Health, National Toxicology Program, Research Triangle Park, NC.
26 [Unpublished study]. May, 1985. (Available by calling EPA's IRIS Hotline at (202)566-1676, by
27 fax at (202)566-1749 or by email at iris@epa.gov).
28 NTP (National Toxicology Program). (1985c) Subchronic inhalation study of bromobenzene in
29 rats. National Institutes of Health, National Toxicology Program, Research Triangle Park, NC.
30 [Unpublished study]. October, 1985. (Available by calling EPA's IRIS Hotline at (202)566-
31 1676, by fax at (202)566-1749 or by email at iris@epa.gov).
32 NTP (National Toxicology Program). (1985d) Subchronic inhalation study of bromobenzene in
33 mice. National Institutes of Health, National Toxicology Program, Research Triangle Park, NC.
34 [Unpublished study]. November, 1985. (Available by calling EPA's IRIS Hotline at (202)566-
35 1676, by fax at (202)566-1749 or by email at iris@epa.gov).
6/7/07 80 DRAFT - DO NOT CITE OR QUOTE
-------
1 NTP (National Toxicology Program). (1985e) NTP Technical report on the toxicology and
2 carcinogenesis studies of chlorobenzene (CAS No. 108-90-7) in F344/N rats and B6C3F1 mice
3 (gavage studies). National Institutes of Health, National Toxicology Program, Research Triangle
4 Park, NC. NTP TR 261. NIH Publication No. 86-2517. NTIS PB86-144714.
5 NTP (National Toxicology Program). (1986a) Pathology Working Group summary report on
6 bromobenzene (C55492) subchronic study (basic and special) F344 rats and B6C3F1 mice.
7 National Institutes of Health, National Toxicology Program, Research Triangle Park, NC.
8 (Available by calling EPA's IRIS Hotline at (202)566-1676, by fax at (202)566-1749 or by email
9 at iris@epa.gov).
10 NTP (National Toxicology Program). (1986b) Pathology Working Group (PWG) review of
11 bromobenzene (C55492) by inhalation in F344 rats and B6C3F1 mice 90-day study. National
12 Institutes of Health, National Toxicology Program, Research Triangle Park, NC. (Available by
13 calling EPA's IRIS Hotline at (202)566-1676, by fax at (202)566-1749 or by email at
14 iris@epa.gov).
15 Ogino, Y. (1984a) Studies on the biological toxicity of several hydrocarbon pollutants of the
16 environment. III. Metabolites and excretion rates of brominated benzenes. Okayama Igakkai
17 Zasshi 96(5/6):553-567.
18 Ogino, Y. (1984b) Studies on the biological toxicity of several brominated benzene pollutants of
19 the environment. IV. Biological fate of brominated benzenes in animals. Okayama Igakkai
20 Zasshi 96(5/6):569-578.
21 Patrick, RS; Kennedy, JS. (1964) S35-labelled amino acids in experimental liver disease. J Pathol
22 Bacteriol88:107-114.
23 Pienta, RJ; Poiley, JA; Lebherz, WB. (1977) Morphological transformation of early passage
24 golden Syrian hamster embryo cells derived from cryopreserved primary cultures as a reliable in
25 vitro bioassay for identifying diverse carcinogens. Int J Cancer 19:642-655.
26 Popper, H; Koch-Weser, D; De La Huerga, J. (1952) Serum and hepatic enzymes in
27 experimental liver damage. J Mt Sinai Hosp 19:256-265.
28 Prodi, G; Arfellini, G; Colacci, A; Grilli, S; Mazzullo, M. (1986) Interaction of halocompounds
29 with nucleic acids. Toxicol Pathol 14(4):438-444.
30 Ramel, C; Magnusson, J. (1979) Chemical induction of nondisjunction in Drosophila. Environ
31 Health Perspect 1:59-66.
32 Reid, WD. (1973) Mechanism of renal necrosis induced by bromobenzene or chlorobenzene.
33 Exp Mol Pathol 19:197-214.
34 Reid, W; Krishna, G. (1973) Centrilobular hepatic necrosis related to covalent binding of
35 metabolites of halogenated aromatic hydrocarbons. Exp Mol Pathol 18:80-99.
6/7/07 81 DRAFT - DO NOT CITE OR QUOTE
-------
1 Reid, WD; Christie, B; Krishna, G; Michell, JR; Moskowitz, J; Brokie, BB. (1971)
2 Bromobenzene metabolism and hepatic necrosis. Pharmacology 6:41-55.
3 Reid, WD; Ilett, KF; Glick, JM; Krishna, G. (1973) Metabolism and binding of aromatic
4 hydrocarbons in the lung. Relationships to expermimental bronchiolar necrosis. Am Rev Respir
5 Dis 107:539-551.
6 Riddick, JA; Bunger, WB; Sakano, TK. (1986) Bromobenzene. In: Techniques of Chemistry.
7 Organic solvents: Physical properties and methods of purification, 4th ed., Vol. 2. New York,
8 NY: John Wiley & Sons. pp. 534.
9 Rombach, EM; Hanzlik, RP. (1997) Detection of benzoquinone adducts to rat liver protein
10 sulfhydryl groups using specific antibodies. Chem Res Toxicol 10(12): 1407-1411.
11 Rombach, EM; Hanzlik, RP. (1998) Identification of a rat liver microsomal esterase as a target
12 protein for bromobenzene metabolites. Chem Res Toxicol 11(3): 178-184.
13 Rombach, EM; Hanzlik, RP. (1999) Detection of adducts of bromobenzene 3,4-oxide with rat
14 liver microsomal protein sulfhydryl groups using specific antibodies. Chem Res Toxicol
15 12(2):159-163.
16 Rosenkranz, S; Poirier, LA. (1979) Evaluation of the mutagenicity and DNA-modifying activity
17 of carcinogens and noncarcinogens in microbial systems. J Natl Cancer Inst 62(4):873-892.
18 Roth, RA. (1981) Effect of pneumotoxicants on lactate dehydrogenase activity in airways of rats.
19 Toxicol Appl Pharmacol 57:69-78.
20 Rush, CF; Newton, JF; Maita, K; Kuo, CH; Hook, JB. (1984) Nephrotoxicity of phenolic
21 bromobenzene metabolites in the mouse. Toxicology 30:259-272.
22 Shamilov, TA. (1969) Toxicity characteristics and hazards of bromobenzene. Gig Tr Prof Zabol
23 13(9):56-58.
24 Shiu, W-Y; Mackay, D. (1997) Henry's law constants of selected aromatic hydrocarbons,
25 alcohols, and ketones. J Chem Eng Data 42:27-30.
26 Simmon, VF. (1979) In vitro mutagenicity assays of chemical carcinogens and related
27 compounds with Salmonella typhimurium. J Natl Cancer Inst 62(4):893-899.
28 Simmon, VF; Rosenkranz, HS; Zeiger, E; Poirier, LA. (1979) Mutagenic activity of chemical
29 carcinogens and related compounds in the intraperitoneal host-mediated assay. J Natl Cancer Inst
30 62(4):911-918.
31 Sipes, IG; Gigon, PL; Krishna, G. (1974) Biliary excretion of metabolites of bromobenzene.
32 Biochem Pharmacol 23:451-455.
33 Slaughter, DE; Hanzlik, RP. (1991) Identification of epoxide- and quinone-derived
34 bromobenzene adducts to protein sulfur nucleophiles. Chem Res Toxicol 4(3):349-359.
6/7/07 82 DRAFT - DO NOT CITE OR QUOTE
-------
1 Slaughter, DE; Zheng, J; Harriman, S; Hanzlik, RP. (1993) Identification of covalent adducts to
2 protein sulfur nucleophiles by alkaline permethylation. Anal Biochem 208(2):288-295.
3 Stierum, R; Heijne, W; Kienhuis, A; van Ommen, B; Groten, J. (2005) Toxicogenomics concepts
4 and applications to study hepatic effects of food additives and chemicals. Toxicol Appl
5 Pharmacol 207(Suppl. 2): 179-188.
6 Sullivan, TM; Born, GS; Carlson, GP; Kessler, WV. (1983) The pharmacokinetics of inhaled
7 chlorobenzene in the rat. Toxicol Appl Pharmacol 71:194-203.
8 Swann, RL; Laskowski, DA; McCall, PJ; Vander Kuy, K; Dishburger, HJ. (1983) A rapid
9 method for the estimation of the environmental parameters octanol/water partition coefficient,
10 soil sorption constant, water to air ratio, and water solubility. Res Rev 85:18-28.
11 Szymahska, JA. (1998) Hepatotoxicity of brominated benzenes: Relationship between chemical
12 structure and hepatotoxic effects in acute intoxication of mice. Arch Toxicol 72(2):97-103.
13 Szymahska, JA; Piotrowski, JK. (2000) Hepatotoxicity of monobromobenzene and
14 hexabromobenzene: Effects of reported dosage in rats. Chemosphere 41(10): 1689-1696.
15 U.S. EPA. (1987) Drinking water health advisory for bromobenzene. External Review Draft.
16 Prepared by the Environmental Criteria and Assessment Office, Office of Health and
17 Environmental Assessment, U.S. Environmental Protection Agency, Cincinnati, OH for the
18 Office of Drinking Water, Washington, DC. ECAO-CIN-W001.
19 U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk
20 assessment. EPA/600/6-87/008. NTIS PB88-179874/AS. February 1988.
21 U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Federal Register
22 56(234):63798-63826. Available at http://www.epa.gov/iris/backgr-d.htm.
23 U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation
24 toxicity: Notice of availability. Federal Register 59(206):53799. Available at
25 http://www.epa.gov/iris/backgr-d.htm.
26 U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application
27 of inhalation dosimetry. EPA/600/8-90/066F. Available at http://www.epa.gov/iris/backgr-d.htm.
28 U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment.
29 EPA/630/R-94/007.
30 U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register
31 61(212):56274-56322. Available at http://www.epa.gov/iris/backgr-d.htm.
32 U.S. EPA. (1998a) Guidelines for neurotoxicity risk assessment. Federal Register
33 63(93):26926-26954. Available at http://www.epa.gov/iris/backgr-d.htm.
6/7/07 83 DRAFT - DO NOT CITE OR QUOTE
-------
1 U.S. EPA. (1998b) Science policy council handbook: Peer review. Prepared by the Office of
2 Science Policy, Office of Research and Development, Washington, DC. EPA/100/B-98/001.
3 U.S. EPA. (2000a) Science policy council handbook: Peer review, 2nd ed. Prepared by the Office
4 of Science Policy, Office of Research and Development, Washington, DC. EPA/100/B-OO/OOl.
5 Available at http://www.epa.gov/iris/backgr-d.htm.
6 U.S. EPA. (2000b) Science policy council handbook: Risk characterization. Prepared by the
7 Office of Science Policy, Office of Research and Development, Washington, DC.
8 EPA/100/B-00/002.
9 U.S. EPA. (2000c) Benchmark dose technical support document. External Review Draft, Office
10 of Research and Development, Risk Assessment Forum, Washington, DC. EPA/630/R-00/001.
11 October. Available at http://www.epa.gov/iris/backgr-d.htm.
12 U.S. EPA. (2002) A review of the reference dose and reference concentration processes. Risk
13 Assessment Forum, Washington, DC. EPA/630/P-02/0002F. Available at
14 http://www.epa.gov/iris/backgr-d.htm.
15 U.S. EPA. (2003) Analysis of national occurrence of the 1998 Contaminant Candidate List
16 (CCL) regulatory determination priotity contaminants in public water systems. U.S.
17 Environmental Protection Agency, Office of Water. EP A/815/D-01/002. Available at
18 http://www.epa.gov/OGWDW/ccl/pdfs/reg determine I/support ccl nation-occur analysis.pdf.
19 U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. U.S. Environmental Protection
20 Agency, Washington, DC. EPA/630/P-03/001B. Available at
21 http://www.epa.gov/iris/backgr-d.htm.
22 U.S. EPA (2005b) Supplemental guidance for assessing susceptibility from early-life exposure to
23 carcinogens. Risk Assessment Forum, Washington, DC. EPA/630/R-03/003F. Available at
24 http://www.epa.gov/iris/backgr-d.htm. November 3.
25 U.S. EPA. (2005c) Peer review handbook, 3rd ed. Review draft. Science Policy Council,
26 Washington, DC. Available at http://intranet.epa.gov/ospintra/scipol/prhndbk05.doc.
27 Verschueren, K. (2001) Bromobenzene. In: Handbook of Environmental Data on Organic
28 Chemicals, Vol 1. New York, NY: John Wiley & Sons. pp. 333.
29 Waters, NJ; Waterfield, CJ; Farrant, RD; Holmes, E; Nicholson, JK. (2006) Intergrated
30 metabonomic analysis of bromobenzene-induced hepatotoxicity: novel induction of
31 5-oxoprolinosis. J Proteome Res 5(6): 1448-1459.
32 Westrick, JJ; Mello, JW; Thomas, RF. (1984) The groundwater supply survey. JAWWA
33 76:52-59.
34 Zampaglione, N; Jollow, DJ; Stripp, MB; Mitchell, JR; Hamrick, M; Gillette, JR. (1973) Role of
35 detoxifying enzymes in bromobenzene-induced liver necrosis. J Pharmacol Exp Ther
36 187:218-227.
6/7/07 84 DRAFT - DO NOT CITE OR QUOTE
-------
1 Zheng, J; Hanzlik, RP. (1992) Dihydroxylated mercapturic acid metabolites of bromobenzene.
2 Chem Res Toxicol 5(4):561-567.
6/7/07 85 DRAFT - DO NOT CITE OR QUOTE
-------
1 APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
2 COMMENTS AND DISPOSITION
3
4 [to be provided]
6/7/07 A-1 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
APPENDIX B. BENCHMARK DOSE CALCULATIONS FOR THE RfD
All available models in the EPA BMDS (version 1.3.2) were fit to incidence data for
histopathologic liver lesions in male and female Fischer 344/N rats and male and female B6C3F1
mice from the 90-day oral gavage studies (NTP, 1985a,b). The data that were modeled are
shown in Table 5-1.
All models provided adequate fits to the data for histopathologic liver lesions
(centrilobular inflammation, cytomegaly, mineralization, or necrosis; combined) in the NTP
(1985a,b) studies, as assessed by a chi-square goodness-of-fit test (see Tables B-l, B-2, B-3, and
B-4 below and respective plots of observed and predicted values from the various models
[Figures B-l, B-2, B-3, and B-4]).
Table B-l. BMD mod
Fischer 344/N rats ex
Model
Log-logistic21
gammab
Multi-stage0
Quantal quadratic
Log-probita
Weibullb
Probit
Logistic
Quantal linear
leling results for the incidence of combined liver effects in male
posed to bromobenzene by gavage 5 days/week for 90 days
BMD10s and BMDL10s (mg/kg-day)
BMDio
172.07
134.60
127.91
65.62
160.78
156.79
45.50
49.24
20.13
BMDLio
69.23
54.59
27.49
49.47
67.44
47.09
31.74
33.29
13.61
x2 p-value
1.00
1.00
1.00
0.88
1.00
1.00
0.73
0.73
0.20
AIC
46.24
46.25
46.27
47.67
48.24
48.24
48.37
48.45
53.93
14
15
16
17
18
aSlope restricted to >1
bRestrict power >=1
cRestrict betas >=0; Degree of polynomial = 5
6/7/07
B-l
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-2. BMD modeling results for the incidence of combined liver effects in female
Fischer 344/N rats exposed to bromobenzene by gavage 5 days/week for 90 days
Model
Log-logistic11
gamma
Quantal quadratic
Multi-stage0
Probit
Logistic
Weibullb
Log-probita
Quantal linear
BMD10s and BMDL10s (mg/kg-day)
BMDio
184.67
161.04
73.60
56.75
46.29
49.08
126.69
181.98
21.40
BMDLio
66.05
37.75
54.85
21.35
32.82
33.73
35.45
59.88
14.41
x2 p-value
0.85
0.85
0.86
0.92
0.81
0.77
0.79
0.71
0.34
AIC
52.66
52.69
53.01
53.83
53.31
53.68
54.40
54.66
57.45
3
4
5
6
7
aSlope restricted to >1
bRestrict power >=1
cRestrict betas >=0; Degree of polynomial = 5
Table B-3. BMD modeling results for the incidence of combined liver effects in male
B6C3F1 mice exposed to bromobenzene by gavage 5 days/week for 90 days
Model
Multi-stage21
Logistic
Probit
Quantal quadratic
Weibullb
Gammab
Log-probitc
Log-logistic0
Quantal linear
BMDioS and BMDLi0s (mg/kg-day)
BMD10
97.99
77.20
69.43
68.85
98.67
99.40
100.10
107.28
22.64
BMDL10
38.82
50.47
46.08
53.53
53.72
57.87
63.56
64.0
15.65
x2 p-value
0.87
0.65
0.60
0.72
0.74
0.71
0.66
0.62
0.09
AIC
35.86
36.89
37.07
37.16
37.86
37.97
38.25
38.61
46.35
9
10
11
12
aRestrict power >=1
bRestrict betas >=0; Degree of polynomial = 5
°Slope restricted to >1
6/7/07
B-2
DRAFT - DO NOT CITE OR QUOTE
-------
Table B-4. BMD modeling results for the incidence of combined liver effects in female
B6C3F1 mice exposed to bromobenzene by gavage 5 days/week for 90 days
Model
Weibuir
Gammaa
Quantal quadratic
Log-probitb
Log-logisticb
Quantal linear
Probit
Logistic
Multi-stage0
BMDioS and BMDLi0s (mg/kg-day)
BMDio
56.08
59.27
74.86
63.34
65.47
23.08
74.52
78.28
50.55
BMDLio
24.81
24.92
59.49
35.33
34.62
16.27
50.54
52.22
20.62
x2 p-value
0.99
0.98
0.87
0.91
0.92
0.73
0.84
0.83
0.95
AIC
40.84
40.98
41.65
41.68
41.70
42.22
42.30
42.38
42.83
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
aRestrict power >=1
bSlope restricted to >1
GRestrict betas >=0; Degree of polynomial = 5
The log-logistic model provided the best fit to the male rat data (see Table B-l),
and was
thus selected to estimate a BMD for the male rats from the NTP (1985a) data. The BMDio
associated with a 10% extra risk for histopathologic liver lesions in male rats was 172.1
day and its lower 95% confidence limit (BMDLio) was 69.2 mg/kg-day (see Figure B-l
mg/kg-
for a plot
of observed and predicted values). The form and parameters of the log-logistic model for male
rat liver effects (NTP, 1985a) are:
P(d) = Bo+(l-B0)/[l+exp(-intercept-slope*log(d))]
d = exposure dose
Bo =0. 199998 (se = 0.0730298)
intercept = -94.8589 (se = 0.786551)
slope = 18; no standard error because this parameter hit a bound
(Eq. B-l)
6/7/07
DRAFT - DO NOT CITE OR QUOTE
-------
Log-Logistic Model with 0.95 Confidence Level
1
2
3
4
5
6
9
10
11
12
13
14
15
16
17
0.8
0.6
0.2
Log-Logistic
BM.D
100
200
300
dose
400
500
600
13:1311/102003
Figure B-l. Observed and predicted incidences of male Fischer 344/N rats
exhibiting bromobenzene-induced combined liver lesions following gavage
treatment 5 days/week for 90 days. BMD=EDi0; BMDL=LEDi0
The log-logistic model provided the best fit to the female rat data (see Table B-2 and
Figure B-2) and was thus selected to estimate a BMD for the female rats from the NTP (1985a)
data. The BMDio associated with a 10% extra risk for histopathologic liver lesions in female rats
was 184.7 mg/kg-day and its lower 95% confidence limit (BMDLio) was 66.1 mg/kg-day (see
Figure B-2 for a plot of observed and predicted values). The form and parameters of the log-
logistic model for female rat liver effects are as follows:
P(d) = Bo+(l-B0)/[l+exp(-intercept-slope*log(d))] (Eq. B-2)
d = exposure dose
Bo = 0.266665 (se = 0.0807368)
intercept = -96.1318 (se = 1.05229)
slope = 18; no standard error because this parameter hit a bound
6/7/07
B-4
DRAFT - DO NOT CITE OR QUOTE
-------
Log-Logistic Model with 0.95 Confidence Level
1
2
o
6
4
5
6
9
10
11
12
13
14
15
16
17
18
19
T3
03
c
g
LL
0.8
0.6
0.2
Log-Logistic
BMDL
IBMD
100
200
300
dose
400
500
600
14:29 11/102003
Figure B-2. Observed and predicted incidences of female Fischer 344/N rats
exhibiting bromobenzene-induced combined liver lesions following gavage
treatment 5 days/week for 90 days. BMD=EDi0; BMDL=LEDi0
The multi-stage model provided the best fit to the male mouse liver lesion data (see Table
5-1), and was thus selected to estimate a BMD for the male mice from the NTP (1985a) data.
The BMDio associated with a 10% extra risk for histopathologic liver lesions in male mice was
97.99 mg/kg-day and its lower 95% confidence limit (BMDLio) was 38.82 mg/kg-day (see
Figure B-3 for a plot of observed and predicted values). The form of the multi-stage model for
male mouse liver effects are as follows:
P(d) = background + (l-background)*[l-EXP(-pl*dose-p2*d2-p3d3- P4d4)] (Eq. B-3)
background = 0
pi = 1.94919e+017
P2 =1.63151e+013
P3 =0
P4 =0.
6/7/07
B-5
DRAFT - DO NOT CITE OR QUOTE
-------
Multistage Model with 0.95 Confidence Level
j
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
••8
0.8
0.6
0.4
0.2
Multistage
BMDL| ,BMD
100
200
300
dose
400
500
600
12:20 02/01 2006
Figure B-3. Observed and predicted incidences of male B6C3F1 mice
exhibiting bromobenzene-induced combined liver lesions following following
gavage treatment 5 days/week for 90 days. BMD=ED10; BMDL=LED10
The Weibull model provided the best fit to the female mouse liver lesion data (see Table
B-4), and was thus selected to estimate a BMD for liver lesions in female mice from the NTP
(1985b) data. The BMDio associated with a 10% extra risk for histopathologic liver lesions in
female mice was 56.1 mg/kg-day and its lower 95% confidence limit (BMDLio) was 24.8
mg/kg-day (see Figure B-4 for a plot of observed and predicted values). Estimated BMDs and
BMDLs associated with 5% and 1% extra risk are presented in Table 5-3 (see Figures B-5 and
B-6 for a plot of observed and predicted values associated with 5% and 1% extra risk,
respectively). The form and parameters of the Weibull model for female mouse liver effects are
as follows:
P(d) = B0+(l-B0)*[l-exp(-slope*d power)] (Eq. B-4)
= exposure dose
= 0
= 0.00152103 (se = 0.000322079)
= 1.62425 (se = 0.383589)
d
Bo
slope
power
6/7/07
B-6
DRAFT - DO NOT CITE OR QUOTE
-------
Weibull Model with 0.95 Confidence Level
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
T3
0)
c
g
'
0.8
0.6
0.4
0.2
Weibull
, .Biyipy,, ,|BMP,,
o
11:41 11/102003
100
200
300
dose
400
500
600
16
17
18
19
Figure B-4. Observed and predicted incidences of female B6C3F1 mice
exhibiting bromobenzene-induced 10% extra risk for liver lesions following
gavage treatment 5 days/week for 90 days. BMD=ED10; BMDL=LED10
6/7/07
B-7
DRAFT - DO NOT CITE OR QUOTE
-------
0.8
0.6
1 0.4
0.2
Weibull Model with 0.95 Confidence Level
Weibull
BMDL
[BMP
100
15:1402/01 2006
200
300
dose
400
500
600
2
3
4
5
6
7
9
10
11
12
13
Figure B-5. Observed and predicted incidences of female B6C3F1 mice
exhibiting bromobenzene-induced 5% extra risk for liver lesions following
gavage treatment 5 days/week for 90 days
The form and parameters for the Weibull model for female mouse liver effects are as
8 follows:
P[response] = background = (l-background)*[l-EXP(-slope*doseApower)] (Eq. B-5)
background = 0
slope = 0.000152103 (se = 0.000322079)
power = 1.62425 (se = 0.383589)
6/7/07
DRAFT - DO NOT CITE OR QUOTE
-------
Weibull Model with 0.95 Confidence Level
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0.8
1 0.6
g
'"§ 0.4
0.2
Weibull
BMDLJ JBMD : : : : : : :
100
200
300
dose
400
500
600
15:1502/01 2006
Figure B-6. Observed and predicted incidences of female B6C3F1 mice
exhibiting bromobenzene-induced 1% extra risk for liver lesions following
gavage treatment 5 days/week for 90 days
The form and parameters for the Weibull model for female mouse liver effects are as
follows:
P[response] = background + (l-background)*[l-EXP(-slope*doseApower)] (Eq. B-6)
background = 0
slope = 0.000152103 (se = 0.000322079)
power = 1.62425 (se = 0.383589)
All available models in the EPA BMDS (version 1.3.2) were fit to absolute liver weight
and liver-to-body weight data in male and female Fischer 344/N rats and male and female
B6C3F1 mice from the 90-day oral gavage studies (NTP, 1985a,b). The data that were modeled
are shown in Table 5-4.
6/7/07
B-9
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Results from the best fitting models for absolute liver weight and liver-to-body weight
ratio in male and female rats and mice are presented in Table 5-5. The BMDLisd of 25.8
mg/kg-day for increased absolute liver weight in female mice represents the lowest BMDLisd
among the male and female rat and mouse data (see Figure B-7 for a plot of observed and
predicted values). The BMDo.ssd and BMDLo.ssd are presented in Table 5-6 (see Figure B-8 for a
plot of observed and predicted values).
The 3-degree polynomial model form of the response function for the female mice
absolute liver weight ratio data is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2
A constant variance was assumed.
(Eq. B-7)
14
15
16
17
18
CD
tf>
C
o
a
1.3
1.2
1.1
0.9
0.8
Polynomial Model with 0.95 Confidence Level
Polynomial
BMDLJ BMD
100
200
300
dose
400
500
600
12:19 02/01 2005
Figure B-7. Observed and predicted 1 standard deviation extra risk for
absolute liver weight changes in female B6C3F1 mice administered
bromobenzene by gavage 5 days/week for 90 days. BMD=EDio;
BMDL=LEDio
6/7/07
B-10
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Polynomial Model with 0.95 Confidence Level
1.3
1.2
0.9
0.8 :
Polynomial
BMDLBMD
100
200
300
dose
400
500
600
Figure B-8. Observed and predicted 0.5 standard deviation extra risk for
absolute liver weight changes in female B6C3F1 mice administered
bromobenzene by gavage 5 days/week for 90 days
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ...
(Eq. B-8)
Third degree parameter estimates for 0.5 standard deviation for absolute liver weight data
in the female mice are presented in Table B-5.
Table B-5. Third degree polynomial estimates for 0.5 standard deviation for absolute liver
weight data
Variable
beta 0
beta 1
beta 2
beta 3
Estimate
0.863152
0.002071
-6.25619e-006
6.69735e-009
Standard Error
0.0184339
0.000348955
1.51029e-006
1.68304e-009
31
6/7/07
B-ll
DRAFT - DO NOT CITE OR QUOTE
-------
The third-degree polynomial model parameter estimates for 1 standard deviation for
absolute liver weight data in female mice are presented in Table B-6.
Table B-6. Third-degree polynomial model parameter estimates for the female mice
absolute liver weight data
Variable
betaO
beta 1
beta 2
beta3
alpha
Estimate
0.863152
0.002071
-6.25619e-006
6.69735e-009
0.00419238
Standard error
0.0184339
0.000348955
1.51029e-006
1.68304e-009
0.000792286
4
5
6
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Serum levels for sorbital dehydrogenase (SDH) for male and female mice were modeled
using the linear, polynomial, power and hill models. The power model results for female mice
provided the best fit and the results of that model follow.
POWER MODEL FOR SDH FEMALE MICE
The form of the response function is:
Y[dose] = control + slope * doseApower
Dependent variable = MEAN
Independent variable = Dose
The power is restricted to be greater than or equal to 1
The variance is to be modeled as Var(i) = alpha*mean(i)Arho
Total number of dose groups = 6
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
(Eq. B-9)
alpha =
rho =
control =
slope =
power =
68.6796
0
12
0.00131174
1.53281
6/7/07
B-12
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
alpha
rho
control
slope
power
Variable
alpha
rho
control
slope
power
Asymptotic Cor
alpha rho
1 -0.99
-0.99 1
-0.07 0.037
-0.3 0.31
0.33 -0.34
Estimate
0.000143716
0.85033
12.9423
1.91405e-006
2.5891
relation Matrix of Parameter Estimates
control slope
-0.07 -0.3
0.037 0.31
1 -0.54
-0.54 1
0.53 -1
Parameter Estimates
95.0% Wald Confidence Interval
Std. Err. Lower Conf. Limit
0.000212936 -0.000273631
0.525075 2.8212
0.351905 12.2526
4.65239e-006 -7.20447e-006
0.392941 1.81895
power
0.33
-0.34
0.53
-1
1
Upper Conf. Limit
0.000561064
4.87945
13.6321
1.10326e-005
3.35925
Table of Data and Estimated Values of Interest
Dose
0
50
100
200
400
600
Model
Model
N Obs Mean
10 13
9 12
9 14
10 15
8 23
10 43
Est Mean Obs Std Dev Est Std Dev Scaled Res.
12.9 1.9 1.66
13 1.6 1.67
13.2 1.8 1.73
14.7 1.7 2.11
23.4 4.6 5.18
42.8 18.8 16.6
0.11
-1.78
1.33
0.481
-0.212
0.0405
Descriptions for likelihoods calculated
Al:
Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model
A2:
Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model
A3:
Yij = Mu(i) + e(ij)
Var{e(ij)} = alpha* (Mu(i))Arho
Model
R:
Yi = Mu + e(i)
Var{e(i)} = SigmaA2
6/7/07
B-13
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Model
Al
A2
A3
fitted
R
Log(likelihood)
-143.251459
-87.617373
-88.708442
-91.313743
-174.876017
Likelihoods of Interest
# Param's AIC
7 300.502917
12 199.234745
8 193.416884
5 192.627486
2 353.752034
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test -2*log(Likelihood Ratio)
Testl 174.517
Test 2 111.268
Test3 2.18214
Test 4 5.2106
Test df
10
5
4
3
p-value
O.OOOl
O.OOOl
0.7023
0.157
The p-value for Test 1 is less than 0.05. There appears to be a difference between response
and/or variances among the dose levels. It seems appropriate to model the data.
The p-value for Test 2 is less than 0.1. A non-homogeneous variance model appears to be
appropriate.
The p-value for Test 3 is greater than 0.1. The modeled variance appears to be appropriate here.
The p-value for Test 4 is greater than 0.1. The model chosen seems to adequately describe the
data.
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
6/7/07
B-14
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Confidence level = 0.95
BMD = 196.474
BMDL=145.789
The lowest BMDLisd from the best fitting model for liver weight changes was 25.8
mg/kg-day, which was very similar to the lowest BMDLio from the best fitting model for
combined liver lesions of 24.8 mg/kg-day. For this reason, liver toxicity in female mice, as
defined by an increase in liver weight and liver lesions was selected as the critical effect for
deriving the subchronic RfD. The average BMDLio of 25 mg/kg-day was selected as the point
of departure to derive the chronic and subchronic RfD for bromobenzene. Full modeling results
for 10% extra risk for combined liver lesions in the Weibull model in female B6C3F1 mice are
presented after Figure B-9.
Weibull Model with 0.95 Confidence Level
0.8
0.6
0.4
0.2
Weibull
BMDL
BMD
100
200
300
dose
400
500
600
13:4805/31 2006
Figure B-9. Full modeling results for 10% extra risk for combined liver
lesions in the Weibull model in female B6C3F1 mice treated by oral gavage
that were used to estimate the RfD
The form of the probability function is:
P[response] = background + (l-background)*[l-EXP(-slope*doseApower)]
background = 0
slope = 0.000152103 (se = 0.000322079)
power = 1.62425 (se = 0.383589)
(Eq. B-10)
6/7/07
B-15
DRAFT - DO NOT CITE OR QUOTE
-------
1 APPENDIX C. BENCHMARK DOSE CALCULATIONS FOR THE RfC
2
3
4 Liver Lesion Data
5 Incidence data for centrilobular cytomegaly in the liver of female B6C3F1 mice were
6 considered as a potential basis of the RfC, based on the results from the 13-week NTP inhalation
7 study indicating that female mice have a lower point of departure for bromobenzene
8 hepatotoxicity than male mice or male or female rats. The data considered for BMD modeling
9 are shown in Table 5-7. Based on the lack of data points from which to readily characterize
10 exposure-response relationships between no-effect and effect levels (i.e., 100 and 300 ppm), it is
11 expected that a number of sigmoidal models will fit such data adequately and equivalently (e.g.,
12 gamma, probit, logistic, higher degree multistage). As a consequence, considerable uncertainty
13 about the 'best' model among sigmoidal models is expected.
14 Sigmoidal models and two non-sigmoidal models (quantal quadratic and quantal linear)
15 in the U.S. EPA BMDS (version 1.3.2.) were fit to the data in Table 5-7. Modeling results are
16 presented in Table C-l showing that (1) all sigmoidal models provided excellent fit to the data
17 (as expected due to the nature of the data) (2) the non-sigmoidal models provided poorer fits to
18 the data, and (3) all sigmoidal models provided similar estimates of BMCio values (ranging from
19 about 77 to 97 ppm, a 1.3-fold range) and BMCLio values (ranging from about 40 to 60 ppm, a
20 1.5-fold range). Following U.S. EPA (2000c) guidance for selecting models for point of
21 departure computation, the model with the best fit and the lowest AIC is selected to calculate the
22 BMCL. The log-logistic and gamma models both have the best fit and the lowest AIC value
23 (Table C-l). The BMCLi0s from these best-fitting models (log-logistic and gamma models)
24 were averaged (55 ppm) to arrive at the point of departure for deriving the RfC, as per U. S. EPA
25 (2000c) guidance. Estimated BMCs and BMCLs associated with 5 and 1% extra risk are
26 presented in Table 5-9. Figures C-l, C-2 and C-3 are plots of the log-logistic models for 10%,
27 5% and 1% extra risk, respectively. Figures C-4, C-5 and C-6 are plots of the gamma models for
28 10%, 5% and 1% extra risk, respectively. Figures C-l and C-4 are plots of observed and
29 predicted values for 10% extra risk from the log-logistic and gamma models, respectively, which
30 were used for the RfC determination. Full modeling details for the 10% log-logistic and gamma
31 models appear at the end of Appendix C.
32
6/7/07 C-1 DRAFT - DO NOT CITE OR QUOTE
-------
Table C-l. BMC modeling results for the incidence of liver cytomegaly in female B6C3F1
mice exposed to bromobenzene vapors 6 hours/day, 5 days/week for 13 weeks
Model
Log-logistic11
Gammab
Multi-stage0
Weibullb
Log-probita
Logistic
Probit
Quantal quadratic
Quantal linear
BMCio
(ppm)
95.59
89.24
77.09
92.34
92.95
96.75
93.71
55.15
21.38
BMCLio
(ppm)
58.73
51.42
40.33
47.08
57.45
59.75
54.94
40.15
13.18
x2 p-value
1.00
1.00
0.999
1.00
1.00
1.00
1.00
0.87
0.16
AIC
12.01
12.01
12.17
14.01
14.01
14.01
14.01
14.05
22.78
2
o
J
4
5
6
aSlope restricted to >1
bRestrict power > = 1
cRestrict betas > = 0; degree of polynomial = 3 (maximum degree restricted to #dose groups
minus 2)
6/7/07
C-2
DRAFT - DO NOT CITE OR QUOTE
-------
Log-Logistic Model with 0.95 Confidence Level
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0.8
0.6
0.4
0.2
Log-Logistic
BMDL
BMD
50
100
150
dose
200
250
300
15:3003/032006
Figure C-l. Observed and predicted incidences of female B6C3F1 mice
exhibiting 10% extra risk of bromobenzene-induced hepatocellular
cytomegaly following inhalation exposure for 6 hours/day, 5 days/week for 13
weeks. Log-logistic model predictions. dose=concentration in ppm.
The form and parameters of the log-logistic model for the incidence of female mouse
cytomegaly are as follows:
P[response] = background+(l-background)/[l+EXP(-intercept-slope*Log(dose))] (Eq. C-l)
background = 0
intercept = -84.2793 (se = 0.790565)
slope =18
6/7/07
C-2
DRAFT - DO NOT CITE OR QUOTE
-------
Log-Logistic Model with 0.95 Confidence Level
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0.8
1 0.6
I
c
o
1 0.4
0.2
Log-Logistic
BMD
BMD
50
100
150
dose
200
250
300
14:1202/022006
Figure C-2. Observed and predicted incidences of B6C3F1 mice exhibiting
5% extra risk of bromobenzene-induced hepatocellular cytomegaly following
inhalation exposure for 6 hours/day, 5 days/week for 13 weeks. Log-logistic
model predictions. dose=concentration in ppm.
The form and parameters of the log-logistic model for incidence of female mice
cytomegaly are as follows:
P[response] = background+(l-background)/[l+EXP(-intercept-slope*Log(dose))] (Eq. C-2)
background = 0
intercept = -84.2793 (se = 0.790565)
slope =18
6/7/07
C-4
DRAFT - DO NOT CITE OR QUOTE
-------
Log-Logistic Model with 0.95 Confidence Level
1
2
3
4
5
6
7
9
10
11
12
13
0.8
1 0.6
I
c
o
'"§ 0.4
0.2
Log-Logistic
BMD
BMD
50
100
150
dose
200
250
300
14:1302/022006
Figure C-3. Observed and predicted incidences of female B6C3F1 mice
exhibiting 1% extra risk for bromobenzene-induced hepatocellular
cytomegaly following inhalation exposure for 6 hours/day, 5 days/week for 13
weeks. Log-logistic model predictions. dose=concentration in ppm.
The form and parameter estimates of the log-logistic model for incidence of female mice
cytomegaly are as follows:
P[response] = background+(l-background)/[l+EXP(-intercept-slope*Log(dose))] (Eq. C-3)
background = 0
intercept = -84.2793 (se = 0.790565)
slope =18
6/7/07
C-5
DRAFT - DO NOT CITE OR QUOTE
-------
Gamma Multi-Hit Model with 0.95 Confidence Level
2
o
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0.8
0.6
c
p
ro 0.4
0.2
Gamma Multi-Hit
BMD
BMD
50
100
150
dose
200
250
300
14:4303/032006
Figure C-4. Observed and predicted incidences of female B6C3F1 mice
exhibiting 10% extra risk of bromobenzene-induced hepatocellular
cytomegaly following inhalation exposure for 6 hours/day, 5 days/week for 13
weeks. Gamma model predictions. dose=concentration in ppm.
The form and parameters of the gamma model for the incidence of female mouse
cytomegaly are as follows:
P[response]= background+(l-background)*CumGamma[slope*dose,power] (Eq. C-4)
where CumGammaQ is the cumulative Gamma distribution function]
background = 0
slope = 0.143677 (se = 0.0164918)
power =18
6/7/07
C-6
DRAFT - DO NOT CITE OR QUOTE
-------
Gamma Multi-Hit Model with 0.95 Confidence Level
1
2
3
4
5
6
7
9
10
11
12
13
14
0.8
1 0.6
£
0.4
0.2
Gamma Multi-Hit
BMDL
BMD
50
100
150
dose
200
250
300
14:0802/022006
Figure C-5. Observed and predicted incidences of female B6C3F1 mice
exhibiting 5% extra risk of bromobenzene-induced hepatocellular
cytomegaly following inhalation exposure for 6 hours/day, 5 days/week for 13
weeks. Gamma model predictions. dose=concentration in ppm.
The form and parameters of the gamma model for 5% extra risk for female mouse
cytomegaly are as follows:
P[response]= background+(l-background)*CumGamma[slope*dose,power] (Eq. C-5)
where CumGammaQ is the cumulative Gamma distribution function
background = 0
slope = 0.0143677 (se = 0.0164918)
power =18
6/7/07
C-7
DRAFT - DO NOT CITE OR QUOTE
-------
Gamma Multi-Hit Model with 0.95 Confidence Level
1
2
3
4
5
6
7
9
10
11
12
13
14
0.8
1 0.6
I
c
o
'"§ 0.4
0.2
Gamma Multi-Hit
BMD
BMD
50
100
150
dose
200
250
300
14:1002/022006
Figure C-6. Observed and predicted incidences of female B6C3F1 mice
exhibiting 1% extra risk of bromobenzene-induced cytomegaly following
inhalation exposure for 6 hours/day, 5 days/week for 13 weeks. Gamma
model predictions. dose=concentration in ppm.
The form and parameters of the gamma model for 1% extra risk for female mice
cytomegaly are as follows:
P[response]= background+(l-background)*CumGamma[slope*dose,power] (Eq. C-6)
where CumGammaQ is the cumulative Gamma distribution function
background = 0
slope = 0.143677 (se = 0.0164918)
power =18
6/7/07
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Liver Weight Data
Absolute liver weight and liver-to-body weight ratio (relative liver weight) in female
mice were also considered as potential bases of the RfC. All available models in the EPA
BMDS (version 1.3.2) were fit to absolute liver weight data and liver-to-body weight ratio for
female B6C3F1 mice from the 13-week inhalation study (NTP, 1985d). The data that were
modeled are shown in Table 5-12. The model outputs are displayed in Table C-2. Second-
degree polynomial models provided the best fit for the absolute and relative liver weight data for
female mice, as determined by the AIC, and yielded BMCLisdS of 33.51 ppm and 33.90 ppm,
respectively (Table C-2). See Figures C-7 and C-8 for a plot of observed and predicted values
Isd and O.Ssd, respectively, for absolute liver weight. See Tables C-3 and C-4 for model outputs
for the second-degree polynomial models for Isd and O.Ssd extra risk for absolute liver weight.
See Figures C-9 and C-10 for a plot of the observed and predicted values for the second-degree
polynomial model for Isd and O.Ssd extra risk for relative liver weight. Model outputs are
displayed in Tables C-5 and C-6. The BMCo.ssd and BMCL0.5Sd are presented in Table 5-14.
Table C-2. Model output for increased absolute liver weight and liver-to-
body weight ratio in female B6C3F1 mice following inhalation exposure to
bromobenzene for 6 hours/day, 5 days/week for 13 weeks
Model3
BMC (ppm)
BMCLlsd (ppm)
x2 p-value
AIC
Absolute liver weightb
Linear
Polynomial (2°)
Polynomial (3°)
Power
35.24
52.38
32.67
56.82
28.39
33.51
14.45
32.56
0.1838
0.3922
0.2891
0.2901
-150.18
-151.16
-149.91
-150.55
Liver-to-body weight ratiob
Linear
Polynomial (2°)
Polynomial (3°)
Power
41.03
52.42
45.52
57.55
34.52
33.90
18.56
34.12
0.08619
0.09284
0.09301
0.07211
183.82
182.19
184.05
182.77
16
17
18
19
20
aStatistical tests indicated that variances were not constant across exposure groups. Model
results are for non-homogeneous variance, with the exception of the linear and third-degree
polynomial models for liver-to-body weight ratio.
Modeled as a continuous variable using one standard deviation as the BMR.
6/7/07
C-9
DRAFT - DO NOT CITE OR QUOTE
-------
Polynomial Model with 0.95 Confidence Level
i
2
3
4
5
6
9
10
11
12
2.6 r
2.4
2.2
5T 2
a)
1.8
1.6
1.4
Polynomial
BMD
BMD
50
100
150
dose
200
250
300
09:10 01/282005
Figure C-7. The second-degree polynomial model prediction for changes 1
standard deviation extra risk in absolute liver weight in female B6C3F1 mice
exposed to bromobenzene vapors for 6 hours/day, 5 days/week for 13 weeks
The second-degree polynomial model form of the response function for the female mouse
absolute liver weight data is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ... (Eq. C-l)
The variance was modeled as:
Var(i) = alpha*mean(i)Arho
(Eq. C-2)
6/7/07
C-10 DRAFT - DO NOT CITE OR QUOTE
-------
Polynomial Mode! with 0.95 Confidence Level
2
3
4
5
6
7
8
9
10
11
12
13
09:55 02/06 2006
Figure C-8. The second-degree polynomial model prediction for changes 0.5
standard deviation extra risk in absolute liver weight in female B6C3F1 mice
exposed to bromobenzene vapors for 6 hours/day, 5 days/week for 13 weeks
The second-degree polynomial model parameter estimates for the absolute liver weight
data in the female mice are presented in Table C-2.
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2
(Eq. C-7)
Table C-3. Second-degree polynomial model parameter estimates for 1 standard
deviation extra risk in absolute liver weight for the female B6C3F1 mice with variance
as a power function of dose
Variable
betaO
beta 1
beta 2
alpha
rho
Estimate
1.47
0.002
0.000004
0.004
2.45
Standard error
0.02
0.0007
0.000002
0.002
1.03
14
15
6/7/07
C-ll
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-4. Second-degree polynomial model parameter estimates for 0.5 standard
deviation extra risk in absolute liver weight for female B6C3F1 mice with variance as a
power function of dose
Variable
betaO
beta 1
beta 2
alpha
rho
Estimate
1.46979
0.00174434
4.19465e-006
0.00412809
2.44506
Standard error
0.0233122
0.000695665
2.343 12e-006
0.00234934
1.02732
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
Polynomial Model with 0.95 Confidence Level
85
80
75
70
65
60
55
50
Polynomial
BMDI
BMD
50
100
150
dose
200
250
300
16:0805/252006
Figure C-9. The second-degree polynomial prediction for 1 standard
deviation extra risk in relative liver weight in female B6C3F1 mice exposed
to bromobenzene vapors for 6 hours/day, 5 days/week for 13 weeks
The second degree polynomial model form of the response function for the female mice
relative liver weight is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2
The variance was modeled as:
Var(i) = alpha*mean(i)Arho
(Eq. C-8)
(Eq. C-9)
6/7/07
C-12
DRAFT - DO NOT CITE OR QUOTE
-------
Polynomial Model with 0.95 Confidence Level
1
2
3
4
5
6
7
8
9
10
11
12
85 r
80
75
2 70
o
CL
CO
0
* 65
ro
0
60
55
50
Polynomial
BMD
BMD
50
100
150
dose
200
250
300
10:0002/062006
Figure C-10. The second-degree polynomial prediction for 0.5 standard
deviation changes in relative liver weight in female B6C3F1 mice exposed to
bromobenzene vapors for 6 hours/day, 5 days/week for 13 weeks
The second degree polynomial model form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2
(Eq. C-10)
Table C-5. Second-degree polynomial model parameter estimates for 1 standard
deviation extra risk in relative liver weight for female B6C3F1 mice with variance as a
power function of dose
Variable
betaO
beta 1
beta 2
alpha
rho
Estimate
53.2265
0.0498228
0.000128264
0.000538819
2.44035
Standard error
0.668405
0.0195324
6.5069e-005
0.00280452
1.27305
13
6/7/07
C-13
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-6. Second-degree polynomial model parameter estimates for 0.5 standard
deviation extra risk in relative liver weight for female B6C3F1 mice with variance as a
function of dose
Variable
betaO
beta 1
beta 2
alpha
rho
Estimate
53.2265
0.0498228
0.000128264
0.000538819
2.44035
Standard error
0.668405
0.0195324
6.5069e-005
0.00280452
1.27305
6/7/07
C-14
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
The average BMCLio from the log-logistic and gamma models for cytomegaly in female
mice was selected as the point of departure to derive the subchronic and chronic RfC for
bromobenzene. Full modeling results for the log-logistic model appear after Figure C-l 1 and
full modeling results for the gamma model appear after Figure C-l2.
Log-Logistic Model with 0.95 Confidence Level
0.8
3 0.6
c
o
0.4
0.2
Log-Logistic
BMDL| |BMD
50
100
150
dose
200
250
300
15:3003/032006
Figure C-ll. Full modeling results for 10% extra risk for cytomegaly in the
log-logistic model in female B6C3F1 mice treated by inhalation that were
used to estimate the RfC
The form of the probability function is:
P[response] =background+(l-background)/[l+EXP(-intercept-slope*Log(dose))] (Eq. C-ll)
Dependent variable = response
Independent variable = dose
Slope parameter is restricted as slope >= 1
6/7/07
C-15
DRAFT - DO NOT CITE OR QUOTE
-------
1 Total number of observations = 5
2 Total number of records with missing values = 0
3 Maximum number of iterations = 250
4 Relative Function Convergence has been set to: le-008
5 Parameter Convergence has been set to: le-008
6
7 User has chosen the log transformed model
8
9 Default Initial Parameter Values
10 background = 0
11 intercept =-8.09038
12 slope =1.74428
13
14
15 Asymptotic Correlation Matrix of Parameter Estimates
16
17 (*** The model parameter(s) -background -slope have been estimated at a boundary
18 point, or have been specified by the user, and do not appear in the correlation matrix )
19
20 intercept
21
22 intercept 1
23
24
25 Parameter Estimates
26
27 Variable Estimate Std. Err.
28 background 0 NA
29 intercept -84.2793 0.790565
30 slope 18 NA
31
32 NA - Indicates that this parameter has hit a bound implied by some inequality constraint and thus
33 has no standard error.
34
35
36 Analysis of Deviance Table
37
38 Model Log(likelihood) Deviance Test DF P-value
39 Full model -5.00402
40 Fitted model -5.00402 2.0891 le-007 4 1
41 Reduced model -27.554 45.0999 4 <.0001
42
43 AIC: 12.008
44
6/7/07 C-16 DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Goodness of Fit
Scaled
Dose Est._Prob. Expected Observed Size
Residual
0.0000
10.0000
30.0000
100.0000
300.0000
0.0000
0.0000
0.0000
0.2000
1.0000
0.000
0.000
0.000
2.000
10.000
0
0
0
2
10
10
10
10
10
10
0
-1.581e-009
-3.112e-005
-2.199e-005
0.0003213
Chi-square = 0.00 DF = 4 P-value = 1.0000
Benchmark Dose Computation
Specified effect =0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD
BMDL
= 95.5947
= 58.7312
6/7/07
C-17
DRAFT - DO NOT CITE OR QUOTE
-------
Gamma Multi-Hit Model with 0.95 Confidence Level
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0.8
1 0.6
I
c
o
'"§ 0.4
0.2
Gamma Multi-Hit
BMD
BMD
50
100
150
dose
200
250
300
14:4303/032006
Figure C-12. Full modeling results for 10% extra risk for cytomegaly in the
gamma model in female B6C3F1 mice treated by inhalation that were used to
estimate the RfC
The form of the probability function is:
P[response]=background+(l-background)*CumGamma[slope*dose,power]
where CumGammaQ is the cumulative Gamma distribution function
Dependent variable = response
Independent variable = dose
Power parameter is restricted as power >=1
Total number of observations = 5
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial (and Specified) Parameter Values
background = 0.0454545
(Eq. C-12)
6/7/07
C-18
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
slope = 0.
power = 1.
00531194
O
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Power have been estimated at a boundary
point, or have been specified by the user, and do not appear in the correlation matrix )
Slope
Slope 1
Parameter Estimates
Variable Estimate
Background 0
Slope 0.143677
Power 18
NA - Indicates that this parameter
has no standard error.
Std. Err.
NA
0.0164918
NA
has hit a bound implied by some inequality constraint and thus
Analysis of Deviance Table
Model Log(likelihood) Deviance Test DF P-value
Full model -5.00402
Fitted model -5.00408
Reduced model -27.554
AIC: 12.0082
Goodness of Fit
Dose Est. Prob. Expected
0.0000 0.0000 0.000
10.0000 0.0000 0.000
30.0000 0.0000 0.000
100.0000 0.2000 2.000
300.0000 1.0000 10.000
Chi-square = 0.00 DF = 4
0.000120288 4 1
45.0999 4 <0001
Scaled
Observed Size Residual
0 10 0
0 10 -5.228e-007
0 10 -0.00267
2 10 -0.000151
10 10 0.007281
P-value = 1.0000
6/7/07
C-19
DRAFT - DO NOT CITE OR QUOTE
-------
1
2
3 Benchmark Dose Computation
4
5 Specified effect =0.1
6
7 Risk Type = Extra risk
8
9 Confidence level = 0.95
10
11 BMD =89.2392
12
13 BMDL =51.4215
14
6/7/07 C-20 DRAFT - DO NOT CITE OR QUOTE
------- |