EPA 635/R-04/067
www.epa.gov/iris
SEPA
TOXICOLOGICAL REVIEW
OF
1,2-DIBROMOETHANE
(CAS No. 106-93-4)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
June 2004
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
11
-------
CONTENTS —TOXICOLOGICAL REVIEW for 1,2-DIBROMOETHANE
(CAS No. 106-93-4)
LIST OF TABLES v
LIST OF FIGURES vii
FOREWORD viii
AUTHORS, CONTRIBUTORS, AND REVIEWERS ix
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL PROPERTIES RELEVANT TO ASSESSMENT 3
3. TOXICOKINETICS RELEVANT TO ASSESSMENTS 5
3.1. ABSORPTION 5
3.2. METABOLISM 5
3.3. DISTRIBUTION 10
3.4. EXCRETION 11
3.5. PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODELS .... 11
4. HAZARD IDENTIFICATION 13
4.1. STUDIES IN HUMANS 13
4.1.1. LETHALITY 13
4.1.2. REPRODUCTIVE TOXICITY 14
4.1.3. CANCER 17
4.2. PRECHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS 20
4.2.1. CARCINOGENICITY BIOASSAYS AND CHRONIC ORAL
STUDIES 20
4.2.2. CARCINOGENICITY BIOASSAY AND CHRONIC INHALATION
STUDIES 25
4.2.3. SUBCHRONIC INHALATION STUDIES 28
4.2.4. OTHER STUDIES 32
4.3. REPRODUCTIVE AND DEVELOPMENTAL STUDIES IN ANIMALS - ORAL
AND INHALATION 37
4.3.1. INHALATION STUDIES 37
4.3.2 ORAL STUDIES 39
4.4. OTHER STUDIES 43
4.4.1. REPRODUCTIVE/DEVELOPMENTAL 43
4.4.2. DEVELOPMENTAL NEUROTOXICITY 45
4.4.3. GENOTOXICITY 47
4.4.4. ACUTE TOXICITY 49
4.5. SYNTHESIS AND EVALUATION OF MAJORNONCANCER EFFECTS AND
MODE OF ACTION - ORAL AND INHALATION 50
in
-------
4.6. WEIGHT-OF-EVIDENCE EVALUATION AND CANCER
CHARACTERIZATION 53
4.6.1. HUMAN 53
4.6.2. ANIMAL 53
4.6.3. MODE OF ACTION 55
4.6.4. WEIGHT-OF-EVIDENCE CHARACTERIZATION 56
4.7. SUSCEPTIBLE POPULATIONS 56
4.7.1. POSSIBLE CHILDHOOD SUSCEPTIBILITY 58
4.7.2. POSSIBLE GENDER DIFFERENCES 58
5. DOSE-RESPONSE ASSESSMENTS 59
5.1. ORAL REFERENCE DOSE 59
5.1.1. METHODS OF ANALYSIS 61
5.1.2. RfD DERIVATION 62
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 63
5.2.1. METHODS OF ANALYSIS 65
5.2.2. RfC DERIVATION 65
5.3. CANCER ASSESSMENT 68
5.3.1. ORAL CARCINOGENICITY 70
5.3.2. INHALATION CARCINOGENICITY 81
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE 93
6.1. HUMAN HAZARD POTENTIAL 93
6.2. DOSE RESPONSE 97
7. REFERENCES 99
APPENDIX A. External Peer Review-Summary of Comments and Disposition A-l
APPENDIX B. BMDS Analyses of Noncancer Endpoints B-l
APPENDIX C. Dose-Response Analyses of Cancer Endpoints C-l
C.I. Analyses in support of slope factor for oral exposure C-l
C.2. Analyses in support of inhalation unit risk C-34
IV
-------
LIST OF TABLES
Table 2-1. Chemical and physical properties of 1,2-dibromoethane 4
Table 4-1. Incidence of tumors in Osborne-Mendel rats in 1,2-dibromoethane oral gavage
bioassay 24
Table 4-2. Incidence of tumors in B6C3F1 mice in 1,2-dibromoethane oral gavage bioassay . . 25
Table 4-3. Nasal cavity tumor types in rats following chronic inhalation of
1,2-dibromoethane 30
Table 4-4. Lung tumor types in mice following chronic inhalation of
1,2-dibromoethane 30
Table 4-5. Enhancement of 1,2-dibromoethane-induced tumor with disulfiram coadministration
in rats 34
Table 4-6. Lung tumor incidence in A/J mice following exposure to 1,2-dibromoethane via
several routes of exposure 36
Table 5-1. Oral subchronic and chronic studies in laboratory animals 61
Table 5-2 Application of uncertainty factors (UFs) for RfD calculation 62
Table 5-3. Inhalation subchronic and chronic studies in laboratory animals6 64
Table 5-4. Reproductive and developmental inhalation studies in laboratory animals 65
Table 5-5. HEC estimates from BMDLs derived from NTP (1982); Table B-2, Appendix B . . 67
Table 5-6. Observed and adjusted tumor incidences (and percentages) in rats and mice exposed
by oral gavage to 1,2-dibromoethane 78
Table 5-7. Human equivalent exposures and adjustment factors for extrapolating from less-than-
chronic exposure to lifetime exposure for the exposure periods in the NCI gavage study of 1,2-
dibromoethane 79
Table 5-8. Estimation of benchmark doses (BMD), lower 95% confidence limits (BMDL),
estimated model parameter q and standard error (se), and goodness-of-fit p-value; using Poly-3
adjusted tumor incidence rates (Table 5-1), for animals exposed orally to 1,2-dibromoethane . 80
Table 5-9. Observed and adjusted tumor incidence rates in rats and mice exposed to 1,2-
-------
dibromoethane by inhalation 88
Table 5-10: Summary of regional gas dose ratios for estimating human equivalent exposures
corresponding to respiratory tumors observed in NTP inhalation bioassay 89
Table 5-11. Estimates of benchmark concentration (BMC) associated with an extra risk of 10%,
lower 95% confidence limits (BMCL), estimated model parameter (and standard error), and
goodness-of-fit p-value; using Poly-3 incidence rates listed in Table B-3 (Appendix B), for
animals exposed by inhalation to 1,2-dibromoethane 90
Table 5-12. Estimation of benchmark concentration (BMC) associated with an extra risk of
10%, lower 95% confidence limits (BMCL), estimated model parameter (and standard error),
and goodness-of-fit p-value, using multistage-Weibull time-to-tumor modeling for animals
exposed by inhalation to 1,2-dibromoethane 91
Table 5-13. Estimation of benchmark concentration (BMC) associated with an extra risk of
10%, lower 95% confidence limits (BMCL), estimated model parameter (and standard error),
and goodness-of-fit p-value; using multistage-Weibull time-to-tumor modeling, for animals
exposed by inhalation to 1,2-dibromoethane 92
VI
-------
LIST OF FIGURES
Figure 3-1. Metabolism of dibromoethane by the oxidative route 8
Figure 3-2. Metabolism of dibromoethane by the conjugative route 9
Figure 5-1. Log-probit model of nasal inflammation in female mice 68
Figure C-l. Kaplan-Meier hazard curves for the incidence of forestomach squamous cell
carcinomas in male rats in the oral gavage study C-7
Figure C-2. Cumulative incidence curves for hemangiosarcomas in male rats in the oral gavage
study C-10
Figure C-3. Cumulative incidence curves for thyroid follicular cell adenomas or carcinomas in
male rats in the oral gavage study C-13
Figure C-4. Kaplan-Meier hazard curves for the incidence of forestomach squamous cell
carcinomas in female rats in the oral gavage study C-16
Figure C-5. Cumulative incidence curves for hemangiosarcomas in female rats in the oral
gavage study C-17
Figure C-6. Cumulative incidence curves for hepatocellular carcinomas and neoplastic nodules
in female rats in the oral gavage study C-20
Figure C-7. Cumulative incidence curves for adrenocortical carcinomas and neoplastic nodules
in female rats in the oral gavage study C-23
Figure C-8. Kaplan-Meier hazard curves for the incidence of forestomach squamous cell
carcinomas in male mice in the oral gavage study C-26
Figure C-9: Cumulative incidence curves for lung adenomas in male mice in the oral gavage
study C-27
Figure C-10. Kaplan-Meier hazard curves for the incidence of forestomach squamous cell
carcinomas in female mice in the oral gavage study C-30
Figure C-l 1. Cumulative incidence curves for lung adenomas in female mice in the oral gavage
study C-31
vn
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to 1,2-
dibromoethane. It is not intended to be a comprehensive treatise on the chemical or
toxicological nature of 1,2-dibromoethane.
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.
Vlll
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER AND AUTHORS
Jeff Gift (chemical manager)
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Research Triangle Park, NC
Mark Greenberg (chemical manager)
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Research Triangle Park, NC.
Karen Hogan
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Audrey Cummings, Ph.D
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Research Triangle Park, NC
John J. Liccione, Ph.D
Sciences International, Inc.
Alexandria, VA
Joseph A. Spinnato III, Ph.D
Sciences International, Inc.
Alexandria, VA
David Gaylor, Ph.D
Sciences International, Inc.
Alexandria, VA
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;
ix
-------
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
Esther Rinde, Ph.D.
Office of Prevention, Pesticides and Toxic Substances
U.S. Environmental Protection Agency
Washington, DC
Sanjivani Diwan, Ph.D.
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
EXTERNAL PEER REVIEWERS
Michael J. DiBartolomeis, Ph.D., D.A.B.T.
California Environmental Protection Agency
Office of Environmental Health Hazard Assessment
Oakland, CA
Lynne T. Haber, Ph.D.
Toxicology Excellence for Risk Assessment
Cincinnati, OH
Cynthia Van Landingham, M.S.
Environ International Corporation
Ruston, LA
Summaries of the external peer reviewers' comments and the disposition of their
recommendations are in Appendix A.
-------
1. INTRODUCTION
This document presents background information and justification for the Integrated Risk
Information System (IRIS) Summary of the hazard and dose-response assessment of 1,2-
dibromoethane. IRIS Summaries may include an oral reference dose (RfD), inhalation reference
concentration (RfC) and a carcinogenicity assessment.
The RfD and RfC provide quantitative information for noncancer dose-response
assessments. The toxicity values are based on the assumption that thresholds exist for certain
toxic effects such as cellular necrosis but may not exist for other toxic effects such as some
carcinogenic responses. In general, the RfD is an estimate (with uncertainty spanning perhaps an
order of magnitude) of a daily exposure to the human population (including sensitive subgroups)
that is likely to be without an appreciable risk of deleterious noncancer effects during a lifetime.
It is expressed in units of mg/kg-day. The inhalation RfC is analogous to the oral RfD, but
provides a continuous inhalation exposure estimate. The inhalation RfC considers toxic effects
for both the respiratory system (portal-of-entry) and for effects peripheral to the respiratory
system (extrarespiratory or systemic effects). It is generally expressed in units of mg/m3.
The carcinogenicity assessment provides information on the carcinogenic hazard potential
of the substance in question and quantitative estimates of risk from oral and inhalation exposure.
The information includes a weight-of-evidence judgment of the likelihood that the agent is a
human carcinogen and the conditions under which the carcinogenic effects may be expressed.
Quantitative risk estimates are presented in three ways to better facilitate their use: (1) generally,
the slope factor is the result of application of a low-dose extrapolation procedure and is presented
as the risk per mg/kg-day of oral exposure; (2) the unit risk is the quantitative estimate in terms of
either risk per |ig/L drinking water or risk per |ig/m3 air breathed; and (3) the 95% lower bound
and central estimate on the estimated concentration of the chemical substance in drinking water or
air that presents cancer risks of 1 in 10,000, 1 in 100,000, or 1 in 1,000,000.
Development of these hazard identification and dose-response assessments for 1,2-
dibromoethane has followed the general guidelines for risk assessment as set forth by the National
Research Council (1983). EPA guidelines that were used in the development of this assessment
may include the following: Guidelines for the Health Risk Assessment of Chemical Mixtures (U.S.
1
-------
EPA, 1986a), Guidelines for Mutagenicity Risk Assessment (U.S. EPA, 1986b), Guidelines for
Developmental Toxicity Risk Assessment (U.S. EPA, 1991), Guidelines for Reproductive Toxicity
Risk Assessment (U.S. EPA, 1996b), Guidelines for Neurotoxicity Risk Assessment (U.S. EPA,
1998a), Draft Revised Guidelines for Carcinogen Assessment (U.S. EPA, 1999),
Recommendations for and Documentation of Biological Values for Use in Risk Assessment (U.S.
EPA, 1988), (proposed) Interim Policy for Particle Size and Limit Concentration Issues in
Inhalation Toxicity (U.S. EPA, 1994a), Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994b), Use of the
Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995), Science Policy Council
Handbook: Peer Review (U.S. EPA, 1998b, 2000a), Science Policy Council Handbook: Risk
Characterization (U.S. EPA, 2000b), Benchmark Dose Technical Guidance Document (U.S.
EPA, 2000c), Supplementary Guidance for Conducting Health Risk Assessment of Chemical
Mixtures (U.S. EPA, 2000d), and^4 Review of the Reference Dose and Reference Concentration
Processes (U.S. EPA, 2002).
The literature search strategy employed for this compound was based on the CASRN and
at least one common name. Any pertinent scientific information submitted by the public to the
IRIS Submission Desk was also considered in the development of this document. The relevant
literature was reviewed through May 2003.
-------
2. CHEMICAL AND PHYSICAL PROPERTIES RELEVANT TO ASSESSMENT
1,2-Dibromoethane, also known as ethylene dibromide, is a colorless, heavy liquid that
has a mildly sweet, chloroform-like odor (HSDB, 1999). It is sparingly soluble in water and
miscible with most organic solvents. It is soluble in alcohol, ether, acetate, and benzene. The
physical properties of 1,2-dibromoethane are listed in Table 2-1.
Prior to the phaseout of leaded gasoline in the United States, 1,2-dibromoethane was
primarily used as an anti-knock compound. Sources of 1,2-dibromoethane included emissions
and exhaust from vehicles using leaded gasoline. Its use as a fumigant for citrus, grain, and soil
was discontinued in 1984 (U.S. Environmental Protection Agency [EPA], 2000e). It is currently
used as a solvent for resins, gums, and waxes; as a chemical intermediate in the synthesis of dyes
and pharmaceuticals; and as a precursor in the synthesis of vinyl chloride. Also, 1,2-
dibromoethane appears to be formed naturally by microalgae growth and has been detected in
ocean waters and air.
1,2-Dibromoethane exhibits low-to-moderate soil adsorption, with experimental Koc
values ranging from 14 to 160, indicating that 1,2-dibromoethane will leach quickly into
groundwater. 1,2-Dibromoethane volatilizes readily from surface soil as predicted by its
relatively high vapor pressure (11.2 mm Hg at 25°C). 1,2-Dibromoethane is very stable towards
hydrolysis (half-life, T1/2 =13.2 years at pH 7 and 20°C) and is more likely to undergo aerobic
biodegradation in the soil rather than abiotic degradation. After 8 weeks under anaerobic
conditions in the presence of denitrifying bacteria, no biodegradation was observed compared
with 97% degradation to ethylene under aerobic conditions.
In aquatic environments, the primary removal process for 1,2-dibromoethane is
evaporation: the volatilization half-life from a typical lake or river is 1-5 days. Biodegradation of
1,2-dibromoethane in groundwater can be slow (with half-lives in months), and biotic hydrolysis
only readily occurs in the presence of a natural catalyst such as hydrogen sulfide. The presence
of hydrogen sulfide increases the rate of hydrolysis from several years to approximately 2
months.
In the ambient atmosphere, 1,2-dibromoethane exists as a vapor, and, although it
-------
undergoes degradation in a reaction with photochemically produced hydroxyl radicals, it is likely
to be persistent. Direct photolysis is not likely to occur.
Table 2-1. Chemical and physical properties of 1,2-dibromoethane
Property Value Reference
Molecular weight 187.88 HSDB, 1999
Density at 25°C 2.172 g/mL
Melting point 9.8°C
Boiling point 131-132°C
Vapor pressure at 25°C 11.2mmHg
Henry's Law constant 8.2 x 10~4 atm-m3/mol
Water solubility at 25°C 4150 mg/L
_K« 66
1 ppm = 7.68 mg/m3 at 25°C and 760 mm Hg.
-------
3. TOXICOKINETICS RELEVANT TO ASSESSMENTS
3.1. ABSORPTION
1,2-Dibromoethane absorption via the gastrointestinal tract was described by Hissink et
al. (2000) using data from Roth et al. (1993). Hissink et al. (2000) found that orally exposed rats
appeared to absorb 1,2-dibromoethane via two first-order processes presumed to be from the
stomach to the intestines and from the intestines to the liver. For the purposes of their models
(see section 3.5), Hissink et al. (2000) estimated rat absorption parameters from in vivo oral and
intravenous (i.v.) exposures and assumed human absorption was not different from that of rats.
For rats dosed orally (gavage) with 50 and 150 mg 1,2-dibromoethane/kg, absorption was
reported to be fast, with blood concentrations nearing estimated Cmax levels within 30 minutes.
Oral exposure to 1 gram of 1,2-dibromoethane resulted in serum 1,2-dibromoethane
concentrations of approximately 1 ppm (0.1% of the dose based on blood volume) in one minipig
15 minutes after compound administration (Kirby et al., 1980). Very rapid 1,2-dibromoethane
metabolism was considered as causing relatively low serum 1,2-dibromoethane concentrations.
Inhalation studies (National Toxicology Program [NTP], 1982; Stinson et al., 1981; Nitschke et
al., 1981; Reznik et al., 1980; Short et al., 1978; Smith and Goldman, 1983) show that 1,2-
dibromoethane is absorbed via the inhalation route of exposure and distributed systemically.
Stott and McKenna (1984) showed that 1,2-dibromoethane is about 50% absorbed when
presented to either the upper or lower respiratory tract of Fisher 344 rats at a flow rate equivalent
to the animals' respiratory minute volume (53 mL/min). Following dermal exposure to 1 mL
1,2-dibromoethane in guinea pigs, 1,2-dibromoethane blood levels increased rapidly to 2.1 |ag/mL
after 1 hour and decreased slightly to 1.8 |ag/mL after 6 hours (Jakobson et al., 1982).
3.2. METABOLISM
1,2-Dibromoethane is metabolized by two major pathways, cytochrome-P450-
monooxygenase and glutathione (GSH) conjugation via glutathione-S-transferase (GST). These
pathways are depicted in Figures 3-1 and 3-2. Oxidative metabolism by cytochrome-P450 leads
to the formation of the reactive metabolite 2-bromoacetaldehyde (Hill et al., 1978) via
dehydrohalogenation of a gew-halohydrin. This route has been demonstrated to account for 80%
-------
of the metabolism of 1,2-dibromoethane in the rat (van Bladeren et al., 1981). 2-
Bromoacetaldehyde can then be converted to s-(2-hydroxyethyl) glutathione (HEG) and S-
carboxymethylglutathione (CMG) through several different pathways involving either direct
interaction with GSH or oxidative metabolism to 2-bromoethanol and 2-bromoacetic acid
followed by conjugation with GSH (Jean and Reed, 1992). However, only the former pathway
involving direct interaction of 2-bromoacetaldehyde with GSH is believed to be catalyzed by
GSTs (Jean and Reed, 1992). Under hypoxic conditions, free radical intermediates have been
generated in microsomal and hepatocyte assays (Tomasi et al., 1983).
1,2-Dibromoethane can also conjugate directly with GSH by a GST-mediated reaction to
form S-(2-bromoethyl)GSH (Jean and Reed, 1992), a half-mustard that can spontaneously
rearrange to an episulfonium ion, thiiranium, and is further hydrolyzed to S-(P-
hydroxyethyl)GSH; or it can bind to DNA (Hodgson and Levi, 1994; Jean and Reed, 1992;
Peterson et al., 1988). S-(2-bromoethyl)GSH can also undergo further GSH conjugation to form
S,S-l,2-ethanediyl-bis-GSH (Jean and Reed, 1992). Formation of these GSH-containing
metabolites correlated with a 71% depletion of intracellular GSH. In vitro studies have shown
that approximately 60% of the episulfonium ion is trapped as S,S-ethanediyl-bis-GSH with the
remainder reacting with water to form S-hydroxyethylGSH (Cmarik et al., 1990). The
episulfonium ion is believed to be responsible for the genotoxicity of 1,2-dibromoethane. The
major adduct derived from the episulfonium pathway is S-[2-7V7-guanyl)ethyl]GSH (Koga et al.,
1986). However, N2- and O6-guanyl adducts may also contribute (Cmarik et al., 1992; Kim and
Guengerich, 1997; Kim and Guengerich, 1998). These may explain the predominance of GC:AT
transitions in Escherichia coli (Foster et al., 1988) and Drosophila melanogaster (Ballering et al.,
1994).
The half-mustard (Ozawa and Guengerich, 1983) and 2-bromoacetaldehyde (Guengerich
and Persmark, 1994) can react with DNA to generate miscoding adducts, although the rate of
reaction of 2-bromoacetaldehyde with DNA has been shown to be rather slow (Guengerich et al.,
1981). The in vitro rate of hydrolysis of S-(2-bromoethyl)GSH is l.e/min'1 with a T1/2 of 0.44
min (Wheeler et al., 2001).
In vivo metabolic studies have identified a number of urinary metabolites following 1,2-
dibromoethane exposure. Nachtomi et al. (1966) identified S-(hydroxyethyl)mercapturic acid as
a major urinary metabolite and S-(p-hydroxyethyl)cysteine as a minor metabolite following oral
-------
administration of 100 mg/kg 1,2-dibromoethane in albino rats. S-(hydroxyethyl)mercapturic acid,
thiodiacetic acid, and thiodiacetic acid sulfoxide have been identified as major urinary
metabolites following oral administration of 1,2-dibromoethane in Wistar rats (Wormhoudt et al.
1998). Thiodiacetic acid formation is dependent on cytochrome-P450 oxidation for formation
and not GSH conjugation (Wormhoudt et al., 1997). S-[2-(N7-guanyl)ethyl]-N-acetylcysteine,
which is derived from the nucleic acid adduct S-[2-(N7-guanyl)ethyl]GSH, has also been
identified as a urinary metabolite in rats following intraperitoneal injection (Kim and Guengerich,
1989).
Van Bladeren et al. (1981) reported that the oxidative route compared to the conjugative
route occurred in a ratio of about 4:1 in rats. This would mean that 1,2-dibromoethane is
preferentially metabolized to 2-bromoacetaldehyde and then conjugated with GSH. The
cytochrome-P450 isozyme responsible for the oxidation of 1,2-dibromoethane to 2-
bromoacetaldehyde appears to be CYP2E1. Wormhoudt et al. (1996a) found that microsomes
from Wistar rats pretreated with pyrazole, an inducer of CYP2E1, had a turnover (Vmax/Km) of
1,2-dibromoethane to 2-bromoacetaldehyde 74 times greater than microsomes from rats
pretreated with p-naphthoflavone, an inducer of CYP2A1. Similarly, heterologously expressed
CYP2E1 had a catalytic efficiency in terms of Vmax/Km 100 times greater than CYP2B6 and 600
times greater than CYP2A6 (Wormhoudt et al., 1996b). The same study found that
heterologously expressed CYP1A1, CYP1A2, CYP2C8, CYP2C9, CYP2C18, CYP3A4, and
CYP3A5 had no ability to metabolize 1,2-dibromoethane to 2-bromoacetaldehyde. Microsomes
from 21 different human livers were able to catalyze the oxidation of 1,2-dibromoethane to 2-
bromoacetaldehyde with activities that ranged from 22.2 to!027.6 pmol/min-mg of protein
(Wormhoudt et al., 1996b). This oxidation was significantly inhibited by specific CYP2E1
inhibitors, disulfiram and diethyldithiocarbamate.
The results of several experiments suggest that of the several mammalian GSH
transferases that can catalyze conjugation with GSH (Cmarik et al., 1990), theta-class GSH
transferase (GSTT) may be most important for conjugation of 1,2-dibromoethane. Investigation
of human erythrocyte cytosol from 12 people not exposed to 1,2-dibromoethane revealed that two
of the cytosols did not catalyze GSH conjugation with 1,2-dibromoethane (Ploemen et al., 1995).
Every cytosol had similar activity toward the classic GSH substrate, l-chloro-2,4-dinitrobenzene.
However, the two cytosol enzymes incapable of catalyzing 1,2-dibromoethane-GSH conjugation
were also incapable of catalyzing GSH conjugation with l,2-epoxy-3-(p-nitrophenoxy)-propane
7
-------
(EPNP). EPNP is a highly selective substrate for GSTT, and this suggests that GSTT is specific
for GSH conjugation of 1,2-dibromoethane in human erythrocyte cytosol. Human GSTT is
polymorphic in humans (Pemble et al., 1994; Warholm et al., 1995; Nelson et al., 1995;
Kempkes et al., 1996). If the carcinogenicity of 1,2-dibromoethane is attributable only to a
conjugate formed only by GSTT, people with a null genotype leading to no activity for this
enzyme may not be as susceptible to cancer from 1,2-dibromoethane (see discussion in section
4.7).
In vitro experiments found Salmonella typhimurium expressing human GST-0 had greater
genotoxicity following 1,2-dibromoethane exposure than strains that did not express this enzyme
(Thier et al., 1996). Simula et al. (1993) reported that GST-a expression increased the
mutagenicity of 1,2-dibromoethane in an S. typhimurium assay but GST-Ti did not. The authors
did not investigate the role of GST-0.
Metabolism of Dibromoethane
I. Oxidative Route
EDB
cytochrame P-450
oxidative metabolism
2- bromoacetaldehyde (binds to microsomai
Br - C - C - BR
H H
+GSH
proteins)
-K3SH
S- (2 hydroxyethyl)
glutathiooe
(HEG)
further
oxidative
metabolism
2- bromoethanol and 2- bromoaeetic acid
+GSH
'+GSH
S- carboxyrnethylglutathione
+GSH
Figure 3-1. Metabolism of dibromoethane by the oxidative route.
-------
Metabolism of DBE HE. Conjugative Route
EDB + GSH
glutathione-
S - transferase
+GSH
S, S'-(1,2-ethanediyl) bis
(glutathione)
(GEG)
Binds to DNA
S - (2 - bromoethyl) glutathione
f (spontaneously)
Thiiranium
(episulfoniumion)
hydrolysis
S - (B-hydroxyethyl) glutathione
(HEG)
S - (hydroxyethyl) mercapturic acid
excreted in urine
Figure 3-2. Metabolism of dibromoethane by the conjugative route.
-------
3.3 DISTRIBUTION
Following oral administration of 15 mg/kg 14C-l,2-dibromoethane in rats, retention of the
compound 24 hours after administration was limited to 3% of administered label in organs with
the majority of the dose retained in the kidney, liver, and spleen (Plotnick et al., 1979). Forty-
eight hours after exposure, 1.1% of administered label was present in the liver. Disulfiram, a
compound known to enhance the toxicity of 1,2-dibromoethane in animals, significantly
increased the tissue concentration of 14C-1,2-dibromoethane in liver, kidneys, spleen, testes, and
brain of male Sprague-Dawley rats that consumed a diet containing 0.05% disulfiram for 12 days
prior to oral intubation with 15 mg/kg 14C-1,2-dibromoethane (Plotnick et al., 1979). Also, label
associated with liver nuclei was significantly increased in disulfiram-treated rats. As would be
expected, urinary elimination of 14C-1,2-dibromoethane was decreased significantly with
increased tissue binding.
One hundred sixty-eight hours after i.v. (10 and 50 mg/kg) or oral (50 and 150 mg/kg)
administration of 14C-1,2-dibromoethane, <1% of administered label was found in the liver, lungs,
and kidneys and 0.3% was present in erythrocytes (Wormhoudt et al., 1998). Short et al. (1979),
who monitored total radioactivity in whole organs (kidney, stomach, liver, and testes) and tissue
fractions 4 hours after oral administration of 10 and 100 mg/kg 14C-1,2-dibromoethane, found that
label was greater in the kidney, liver, and stomach than in testes. A portion of the label in these
tissues was covalently bound to DNA, RNA, and protein.
In mice, irreversibly bound metabolites of 14C-1,2-dibromoethane (2.6 mg/kg) were
reported to be highest in the nasal mucosa, followed by the liver, lung, and kidney, 3 hours after
intraperitoneal (i.p.) injection (Brittebo et al., 1989). Similarly, Hill et al. (1978) reported
significant macromolecular binding to epithelial tissues of the respiratory tract, liver, kidney, and
small intestine of rats following i.p. injection of 0.8 mg/kg 14C-1,2-dibromoethane. In one male
cynomolgous monkey administered 4 |_imol/kg 14C-1,2-dibromoethane intraperitoneally, the liver
and kidney proximal tubules had the highest amount of administered radioactivity per mg of
tissue, but very little binding was observed in the lung and other organs of the respiratory tract
(Brandt et al., 1987). When tissues from a female monkey were incubated with 20 |J,M 14C-1,2-
dibromoethane, the distribution of bound metabolites in the kidney was identical to that observed
in vivo. The authors also described a distinct zone of binding in the adrenal zona reticularis, both
in vitro and in vivo. Significant levels of radiolabeled metabolites irreversibly bound to the
10
-------
adrenal cortex, nasal cavity, lung, and other tissues of rats (Sprague-Dawley and Fischer) and
C57BL mice were also observed by Kowalski et al. (1985) after i.v. and i.p. injection of 14C-1,2-
dibromoethane. For results in fetal tissues, see section 4.7.1.
3.4. EXCRETION
1,2-Dibromoethane is eliminated mainly in the urine. In vivo metabolic studies have
identified a number of urinary metabolites, including S-(hydroxyethyl)mercapturic acid and S-(P-
hydroxyethyl)cysteine following oral administration of 100 mg/kg 1,2-dibromoethane in albino
rats (Nachtomi et al., 1966) ). S-(hydroxyethyl)mercapturic acid, thiodiacetic acid, and
thiodiacetic acid sulfoxide have been identified as major urinary metabolites following oral
administration of 1,2-dibromoethane in Wistar rats (Wormhoudt et al., 1998). S-[2-(N7-
guanyl)ethyl]-N-acetylcysteine, which is derived from the nucleic acid adduct S-[2-(N7-
guanyl)ethyl]GSH, has also been identified as a urinary metabolite in rats following
intraperitoneal injection (Kim and Guengerich, 1989).
Following oral administration of 15 mg/kg 14C-1,2-dibromoethane in rats, 72% of the dose
was excreted in the urine and 1.65% in the feces (Plotnick et al., 1979). Forty-eight hours later,
73% of the dose had been accounted for in the urine and 3% in the feces. Similar results were
reported by Wormhoudt et al. (1998), who found that following i.v. (10 and 50 mg/kg) or oral (50
and 150 mg/kg) administration of 14C-1,2-dibromoethane, 75 - 82% of the radioactivity was
excreted in the urine within 48 hours, 3.2 - 4% was eliminated in the feces within 48 hours, and
0.53 - 7.2% was eliminated in the expired air within 2 hours. The only major difference between
the two routes of administration was that a much higher percentage of the dose was eliminated in
the expired air following i.v. administration.
3.5. PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODELS
Hissink et al. (2000) have developed a PBPK model for 1,2-dibromoethane for rats and
humans that is largely based on in vitro data published by Ploemen et al. (1997). Due primarily to
the higher relative ventilation rate, cardiac output, and metabolic rate of rats, the Hissink et al.
(2000) model predicts that blood concentrations of 1,2-dibromoethane and the metabolites of both
11
-------
pathways (P450 oxidative and GST conjugation) described above would be higher in rats than in
humans. Their model predicts that this would be the case for average humans and for humans
with extremely high contributions from one or the other pathway (i.e., high GST/P450 or
P450/GST contribution ratios). Assuming that a GSH conjugate is the toxicant and that total
GSH conjugate level correlates to the specific toxic episulfonium ion would suggest that humans
may be less sensitive than rats to the genotoxic and carcinogenic effects of 1,2-dibromoethane.
The model has not been fully validated, however, and makes several assumptions that affect the
conclusions, including the following: (1) there is no P450 activity in human kidneys but
significant P450 activity in rat kidneys; (2) there is considerably lower GSH activity in human
skeletal muscle, lung, and stomach compared with rat; and (3) that steady state levels were
reached in blood after an 8-hour exposure. The model only simulated an 8-hour exposure, and
longer simulation runs are needed to evaluate steady-state levels and subsequent differences in
GSH conjugate levels. The model also does not include terms for GSH synthesis and
degradation rates (also needed for chronic exposures), factors that can greatly affect GSH
conjugate levels. It also does not attempt to distinguish (or correlate) the episulfonium conjugate
from all other GSH conjugates or the P450 metabolite conjugates from the parent compound GSH
conjugates. Given these limitations, it would not be appropriate to use the Hissink et al. (2000)
model at this time for quantitative (route-to-route or animal-to-human) extrapolations. However,
the Hissink et al. (2000) report and model provide useful information regarding the mode of
action of 1,2-dibromoethane, particularly with respect to the cancer effects of 1,2-dibromoethane.
12
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS
4.1.1. Lethality
Olmstead (1960) described a case report of a 43-year old woman who ingested capsules
containing 1,2-dibromoethane. Symptoms included abdominal pain, nausea, vomiting, diarrhea,
darkening of the urine, tachypnea, and marked agitation; she died 54 hours after ingesting of the
capsules. Histological examination revealed massive central lobular necrosis of the liver with
focal proximal tubular epithelial damage in the kidney. Olmstead (1960) commented that these
changes were similar to those observed in various experimental animals following oral
administration of 1,2-dibromoethane.
Six cases of suicidal poisoning via ingestion of 1,2-dibromoethane were reported by
Saraswat et al. (1986). Two of the six individuals died. In one of these individuals, postmortem
examination revealed massive liver necrosis; ulceration of the oral cavity; congestion and erosion
of the stomach; congestion of the lungs, spleen, brain, and kidneys; pulmonary edema; and cloudy
swelling of the kidney with occasional tubular necrosis. The other individual showed
centrilobular necrosis of the liver and congestion of the lungs and stomach. Clinical signs
observed in those who survived ingestion of 1,2-dibromoethane included nausea, vomiting, and
burning of the throat.
1,2-Dibromoethane was lethal to two workers following acute exposure to 1,2-
dibromoethane in a tank (Letz et al., 1984). One worker exhibited metabolic acidosis, central
nervous system depression, and liver damage. The other worker displayed metabolic acidosis and
hepatic and renal failure. Elevated serum bromide concentrations were detected in both cases
prior to death.
13
-------
4.1.2. Reproductive Toxicity
Semen quality has been studied in papaya workers with long-term exposure to 1,2-
dibromoethane used in the fumigation of papaya (Ratcliffe et al., 1987). Workers at six papaya
fumigation plants located near Hilo, Hawaii, were assessed for potential 1,2-dibromoethane-
related reproductive toxicity.
Exposure of the papaya workers to 1,2-dibromoethane was characterized in detail by
Clapp (1986). The workers were exposed to 1,2-dibromoethane for an average of 5 years.
Samples were reported to be collected on at least two separate days to give an estimate of
exposure during the workweek. Three to four packers/sorters and one to two forklift operators
were sampled per plant for an 8-hour work shift; in three of the plants, air samples were taken
both in 1982 and 1983. In these three plants, there is a considerable amount of variation between
the 1982 and 1983 mean air sample concentrations measured for the same plant. Full-shift
exposures of 16 to 175 ppb (0.1 to 1.4 mg/m3) among the six plants investigated were reported
with a geometric mean of 88 ppb (0.68 mg/m3) and peak exposures as high as 262 ppb (2.0
mg/m3). Significant plant-to-plant variation was reported: full-shift exposures for papaya
packers/sorters ranged from 36 to 148 ppb (0.3 - 1.1 mg/m3) and those for forklift operators
ranged from 16 to 175 ppb (0.1 - 1.4 mg/m3). A sugar processing plant adjacent to one of the
fumigation plants served as a control group for exposed workers.
Semen was collected and delivered for analysis within 1 hour of collection (Ratcliffe et
al., 1987). The following potentially confounding variables were considered: tobacco smoking
(current, ex-smokers, and nonsmokers), marijuana use (times per week), alcohol consumption
(drinks per week), caffeine consumption (cups of coffee and/or tea per day), history of urogenital
disorders, prescription medication taken in the past year, history of fever in the past three months,
abstinence time, subject age, sample age, and racial background. The only significant difference
between control and exposed workers was marijuana use, which was higher in exposed workers
(41.3%) than in controls (20.9%).
Ratcliffe et al. (1987) reported summary air exposure data, but there was moderate dermal
exposure that could not be quantified (Schrader et al., 1988). Semen of exposed workers
exhibited significantly decreased average sperm count per ejaculate and percentage of viable and
motile sperm. There were statistical increases in certain types of morphological abnormalities
14
-------
(tapered heads, absent heads, and abnormal tails) in exposed workers. There was also a
significant increase in percentage of subjects with sperm counts fewer than 20 million in exposed
workers (21.7% compared to 4.7% in controls). The highly variable inhalation exposures and the
confounding dermal exposures preclude the use of this population for the development of an RfC.
A comparison of 1,2-dibromoethane-exposed cohorts of forestry workers in Colorado
(short-term exposure) and papaya workers from the Ratcliffe et al. (1987) study (long-term
exposure) was performed by Schrader et al. (1988). Ten 1,2-dibromoethane-exposed and six
unexposed forestry workers were identified for the study. Schrader et al. (1988) reported that 8-
hour air samples were collected for three days during the six-week fumigation season and several
short-term (15-minute and 1-hour) samples were also taken. Short-term exposure concentrations
ranged from non-detectable levels to 2165 ppb (16.6 mg/m3) for workers filling truck tanks with
1,2-dibromoethane, 57 to 525 ppb (0.4 to 4.0 mg/m3) for workers spraying log piles, and 8 to 184
ppb (0.1 to 1.4 mg/m3) for workers pouring an 1,2-dibromoethane emulsion on log piles. The
results of the 8-hour samples were not reported, but a time-weighted average exposure of 60 ppb
(0.5 mg/m3) was reported. It should be noted that dermal exposure was recognized as a major
potential route of exposure. Schrader et al. (1988) stated that dermal exposure in forestry workers
was excessive, and the papaya workers described in Ratcliffe et al. (1987) were subject to
moderate dermal exposure. Attempts to quantify the dermal route of exposure were unsuccessful.
Exposed and control worker ejaculates were collected 1-2 weeks before exposure and during the
last week of exposure.
When postexposure semen samples of forestry workers were compared to those collected
pre-exposure, sperm velocity was significantly decreased in all 10 exposed workers compared to
controls, and nine exposed workers had significantly decreased semen volume compared to
controls. Sperm viability, sperm concentration, semen pH, sperm morphology, and sperm
morphometry of exposed workers were not significantly different from controls. Sperm motility
was a measured parameter, but there was no explicit indication of the percent of motile sperm in
the report. It is assumed that there was no effect of 1,2-dibromoethane on motility.
These results contrast with those of papaya workers described in Ratcliffe et al. (1987).
Schrader et al. (1988) suggest that the differences in toxic responses between the two studies were
due to the duration of exposure. Markers of mature sperm function, such as sperm velocity, might
be expected to be altered during short-term exposure, but morphogenic effects, as seen in
15
-------
Ratcliffe et al. (1987), would not be observed unless exposure occurred for at least the duration of
one spermatogenic cycle (74 days). Therefore, Schrader et al. (1988) conclude that the two
studies complement each other when the duration of exposure (6 weeks vs. 5 years) is considered.
Wong et al. (1979) conducted a retrospective evaluation of reproductive performance of
male workers in four chemical plants that produced 1,2-dibromoethane during the 1958-1977
time period. This was done by comparing actual numbers of live births delivered by their wives
compared to the expected numbers of live births derived from national fertility tables published
by the National Center for Health Statistics. In addition to 1,2-dibromoethane, one plant
produced dibromochloropropane (a known male reproductive toxicant), another produced "other
brominated compounds," workers in another plant were also probably exposed to ethylene
dichloride (EDC), and the remaining plant was limited to 1,2-dibromoethane production. 1,2-
Dibromoethane was monitored through industrial hygiene studies, except at one plant in which
1,2-dibromoethane levels were not determined. Observed and expected births were adjusted for
maternal age, parity, race, and calendar year for all 1,2-dibromoethane-exposed workers at all
four plants. Workers were classified as being exposed to either < 0.5 ppm (< 3.8 mg/m3) or
0.5-5.0 ppm (3.8 - 38 mg/m3). There was no difference between observed and expected births in
wives of workers for three of the plants. There was, however, a significant difference (< 0.05)
pertaining to the plant producing 1,2-dibromoethane and also using EDC: the 1,2-dibromoethane
levels at this plant ranged from 0.1 to 4 ppm (0.8 to 30 mg/m3). Considering the limited
exposure data and co-exposure to other chemicals, no conclusions can be drawn concerning the
potential antifertility effect of 1,2-dibromoethane from this study.
Sperm counts have been assessed in agricultural workers with known exposure to 1,2-
dibromoethane (Takahashi et al., 1981). Volunteers were asked to complete a medical history
questionnaire related to possible factors that could affect spermatogenesis. Agricultural workers
and unexposed controls were asked to provide semen samples. Sperm count, morphology, and
motility were determined 0.5-1 hour after collection. Co-exposure to dibromochloropropane
(DBCP) also occurred in agricultural workers, but levels of DBCP and 1,2-dibromoethane
exposure could not be precisely determined. The only observed adverse effect in the workers was
significantly lower sperm counts compared to controls. Conclusions about the effect of 1,2-
dibromoethane on male fertility cannot be determined from this study because of a lack of
exposure data and co-exposure to DBCP, a well-documented gonadotoxin.
16
-------
Fertility history was evaluated for male workers employed in a 1,2-dibromoethane
production facility (Turner and Barry, 1979) by examining family size. A total of 82 married
workers employed at the time of the study (1977) was identified, 41 of whom were determined to
have very low or no exposure to 1,2-dibromoethane, while the remaining 41 had some exposure
to 1,2-dibromoethane. Of those thought to have some exposure, 13 were regular process workers,
17 were process operators with variable exposure, and 11 were maintenance workers with
sporadic potential exposure. In addition, 75 process workers were identified from company
records of all married employees who had left employment by the time of the survey.
Presumably, this earlier employed group would have had higher exposure to 1,2-dibromoethane
than in the first set identified, but no information was mentioned concerning exposure categories.
No exposure measurements or frequency information was available for any of the workers,
however. A comparison group of 80 men was selected for this historical group from the patients
of a local general practitioner, excluding any ever employed at the facility. For all groups of men,
the investigators compared average number of children per family enjoining the company with
the average number of children at the time of the survey and concluded that there were no
statistical differences among any of the groups. Among those employed at the time of the survey,
average family sizes were slightly higher among those thought to be exposed to 1,2-
dibromoethane (2.18 to 2.85 children per family) than for those thought not to be exposed (1.95).
Ages of the men were similar, and average number of years employed when the children were
born varied from 1.5 to 4.5 among those with exposure and was 5 years for those without
exposure. For the historical group and its control, family sizes were very similar at 1.97 and 1.98
children/family, respectively; however, no information was provided regarding comparability of
ages of the men or of the amount of time over which the fertility rates were evaluated. Due to
several limitations of the study design- lack of exposure data, small sample size (especially
within exposure categories), and apparent lack of control regarding whether any of the families
were trying to have more children- this study is inconclusive regarding the effect of 1,2-
dibromoethane on male fertility.
4.1.3. Cancer
Mortality of workers occupationally exposed to 1,2-dibromoethane has been investigated in
several studies (Ott et al., 1980; Sweeney et al., 1986). Ott et al. (1980) investigated the cause of
death of 161 workers occupationally exposed to 1,2-dibromoethane in two production facilities,
17
-------
one in operation from 1942 to 1969 and the other from the mid-1920s to 1976. The study
primarily focused on cancer mortality and respiratory disease. Quantitative data from which to
calculate an 8-hour time-weighted average (TWA) were not available for the first facility.
However, area sampling data from 1950, 1952, and 1971 - 72 and personal air monitoring in 1975
allowed for estimation of 1,2-dibromoethane exposure in the second facility. Area samples from
1950 ranged from 1 - 10.6 ppm (7.7 - 81.4 mg/m3), 19-31 ppm (146 - 238 mg/m3) in 1952, 0 -
110 ppm (0 - 845 mg/m3) from 1971 - 1972, and 1.8-96 ppm (14 - 737 mg/m3) in 1975. An
estimated TWA of 3.5 ppm (26.9 mg/m3) was calculated for 1971 - 1972 and 5 ppm (38.4 mg/m3)
for 1975.
In the first facility, two deaths from malignant neoplasms were observed compared to 3.6
expected. No other organic bromide compounds were manufactured at this facility and exposure
was primarily limited to 1,2-dibromoethane, bromine, ethylene, sulfur dioxide, and chlorine. In
the second facility, 5 deaths were attributed to malignant neoplasms compared to 2.2 expected.
However, this facility manufactured other organic bromide chemicals, such as vinyl bromide,
trimethylene chlorobromide, propylene chlorobromide, ethyl bromoacetate, isobutyl bromide, and
acetylene tetrabromide, to which workers were potentially exposed. In addition to organic
bromide chemicals, workers in the second plant were also indirectly exposed to allyl chloride,
benzene, bromochloromethane, carbon tetrachloride, chloroform, ethyl bromide, hydrogen
bromide, methylene chloride, methylene dibromide, tert-bromobutyl phenol, and tert-butyl
phenol.
The neoplasms were not tissue-specific as can be seen from the following incidence: three
lung, two stomach, one prostate, one reticulum-cell sarcoma, one pancreas, and one unknown. In
addition, two of the lung neoplasms were associated with workers who had prior history of
working with arsenicals which can also cause this type of cancer. The authors did not count these
two workers as contributors to the total neoplasm population from this facility because of their
prior arsenical exposure. Also, the two workers who succumbed to stomach cancer were father
and son and were reported to have a family history of cancer. Smoking history for neoplasm-
related deaths was available for one of nine workers. The study is inconclusive in regards to 1,2-
dibromoethane as a potential human carcinogen because of co-exposure to other potential
carcinogens.
Sweeney et al. (1986) studied the cause-specific mortality of 156 male workers in a
18
-------
chemical plant that manufactured tetraethyl lead. The study was undertaken to investigate an
apparent cluster of three cases of multiple myeloma and four cases of brain cancer. In this
historical prospective study, employees who worked at least 1 day between 1952 and 1977 were
eligible for the study. Other chemicals used in the tetraethyl lead manufacturing processes were
ethylene dibromide, ethylene dichloride, chloroethane, ethylene, inorganic lead, and dyes.
Environmental exposure data were available for some chemicals, but the work history records
were not sufficiently detailed to construct exposure indices. Consequently, the exposure levels of
the various chemicals and particular combinations of chemicals could not be determined.
The standardized mortality ratios (SMRs) for carcinomas at several sites were slightly
elevated: colon and rectum (5 observed vs. 3.7 expected); trachea, bronchus, and lung (14
observed vs. 12.5 expected); and brain (4 observed vs 1.88 expected). No SMRs were
statistically significantly elevated, however. The investigators reported that there was low power
for detecting excess risk of mortality from multiple myeloma (27% at a 5% significance level),
brain cancer (31%), or other rare cancers. Concerning the apparent clusters, it was determined
that there had been some misclassification of original diagnoses, as one brain cancer was actually
a metastatic carcinoma of the lung. One of the three multiple myeloma cases died subsequent to
the end date of the study, however, and was not included in the analysis. It is not clear why the
study period ended in 1977, given that the study's purpose was to investigate these specific cases.
Mortality in the Sweeney et al. (1986) study from all causes was lower than that predicted
based on the death rates of United States males (156 observed vs. 211 expected). This is common
in occupational studies and is frequently characterized as the "healthy worker effect." Ideally, the
most appropriate comparison group would have been a group of workers exposed to similar levels
of the same chemicals except ethylene dibromide. Overall, the study is inconclusive, given the
small study size and co-exposure to other chemicals. However, it does not rule out an association
of increased cancer incidence with exposure to 1,2-dibromoethane.
Turner and Barry (1979) evaluated mortality due to cancer and other causes in two 1,2-
dibromoethane production facilities operated by the same company. One facility (Factory A)
operated from 1940 to 1970, while the second (Factory B) started operating in 1952 and was still
in operation at the time of the study (1977). Records for men who had been employed at least 4
years during the operation of the facility, with potential exposure to 1,2-dibromoethane, were
examined. No exposure measurements or frequency information was available for any of the
19
-------
workers, however. From Factory A, 117 men were identified with potential exposure to 1,2-
dibromoethane for at least 4 years, 34 of whom had died at the time of the study. From Factory
B, 274 men were identified with potential exposure to 1,2-dibromoethane for at least 4 years, 26
of whom had died by the time of the study. Death rate per 1000 man-years was calculated within
broad age-at-death categories (25-44, 45-64, 65-74, and 75 years and over) due to the small
number of subjects and compared with analogous death rates in the local population in 1961 and
1970 to help consider possible differences associated with date of death. Death rates for all
causes and for cancer only were similar or lower in the exposed groups than in the local
population for all age groups, except in the 75 years and over group (Factory A only; there were
none in this age group in the Factory B set). In the oldest age group, the death rate among
exposed workers was 149.9 per 1000 man-years for all causes, while the analogous death rate for
the local population was 136.5 in 1961 and 135.8 in 1970. For cancer, the death rate was 21.3,
while the analogous rate in the local population was 16.2 in 1961 and 17.0 in 1970; however,
there was only one cancer case in the exposed group. Given the extremely small study size, lack
of exposure information, and insufficient allowance for cancer latency, especially for Factory B,
this study is inconclusive regarding the effect of 1,2-dibromoethane on carcinogenicity.
4.2. PRECHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS
4.2.1. Carcinogenicity Bioassays and Chronic Oral Studies
The National Cancer Institute (NCI) (1978) examined the potential carcinogenicity of 1,2-
dibromoethane in rats and mice. Male and female Osborne-Mendel rats and B6C3Fj mice
(50/sex/species/exposure group; n = 20 in untreated control group and n = 20 in vehicle control
group) were administered 1,2-dibromoethane in corn oil by gastric intubation for 5 days/week
until sacrificed.
Rats: The initial doses utilized for male and female rats were 40 and 80 mg/kg-day.
However, high treatment-related mortality (18/50 males and 20/50 females) caused a
discontinuation in the intubation of the high-dose group after treatment in week 16. Intubation of
this group was suspended for 13 weeks and then restarted at week 30. At this time the surviving
20
-------
rats received the low-dose regimen. All surviving male and female rats in both dosage groups
were sacrificed at weeks 49 and 61, respectively. Time-weighted average low- and high-doses
were 38 and 41 mg/kg-day for male rats, and 37 and 39 mg/kg-day for female rats.
The treated and vehicle control rats were placed on test simultaneously at the age of 8
weeks with intubation of vehicle controls suspended after 49 weeks, followed by a 14-day
observation period. All male and female vehicle controls were sacrificed in week 63 at the age of
71 weeks. Gavage of female vehicle controls was suspended after 61 weeks and was followed by
a 2-week observation period. Untreated controls (5 weeks of age) were placed on test 15 weeks
after the start of the treated rats and vehicle controls. Thus, all untreated (without vehicle)
controls were on test for 107 weeks.
Treatment-related effects included squamous cell carcinoma of the forestomach in rats of
both sexes, judged by the investigators to be a possible cause of the high mortality. There was a
positive association (p < 0.001) in both groups between increasing dosage and accelerated
mortality. Because 40% of high-dose females died in week 15 either during intubation or shortly
thereafter, it was suggested that acute toxic reactions were the likely cause of those deaths.
However, squamous cell carcinomas (in all low-dose females surviving beyond week 15) were
also considered by the investigators to be associated with mortality in this group. There was no
mention in the report if gavage error contributed to the incidence of early mortality.
There were statistically significant increases in tumor incidence for both male and female
rats, both at the point of contact (forestomach) and systemically. Crude incidences of
forestomach tumors were 90% in low-dose males and 80% in low-dose females, but 66% and
58% in high dose animals. Lower incidence in the high-dose group may have been due to the
higher rate of early deaths. Metastases of the forestomach tumors to multiple organs were
common. There was an elevated incidence of hepatocellular carcinoma in high-dose female rats
only. Hemangiosarcoma, particularly of the spleen, was observed in male and female animals.
Males also had a small incidence of hemangiosarcoma in other organs: liver, kidney, pancreas,
and abdominal cavity. The incidence of hemangiosarcomas was statistically significant for low-
dose animals only. The incidence of these selected neoplastic lesions in the rat is summarized in
Table 4-1.
Non-neoplastic lesions in the high-dose animals of both sexes included hyperkeratosis and
21
-------
acanthosis of the forestomach (12/50 males and 18/50 females). The incidence of hyperkeratosis
and acanthosis was lower in low-dose animals with only 4/50 females (compared with 1/20
untreated controls) and no males presenting this adverse effect. Peliosis (mottled-blue liver)
hepatitis was observed in 0/40 controls, 10/50 low-dose, and 9/50 high-dose males and in 0/40
controls, 1/47 low-dose, and 5/48 high-dose females. Inflammation of the liver was also present
in 1/40 controls, 4/50 low-dose, and 5/50 high-dose males and in 0/40 controls, 4/47 low-dose,
and 2/48 high-dose females. Degeneration of the adrenal cortex was noted in 0/40 controls and
13/48 and 9/47 low- and high-dose males. The incidence of adrenocortical degeneration was
slightly lower in females (1/40 controls, 3/44 low-dose, and 8/45 high-dose animals).
Testicular atrophy was observed in 11/20 untreated controls, 0/20 vehicle controls, and
14/49 and 18/50 low- and high-dose males, respectively. Because the untreated controls were on
test for a far longer period (107 weeks) than the treated or vehicle controls, the vehicle controls
are the appropriate reference group. It was stated that necropsy was performed on each animal
regardless of whether it died, was sacrificed while moribund, or was at the end of the study. Thus,
the incidence data for testicular atrophy for only those male rats that survived until the end of
study (35 weeks) were likely higher than the data given for the original 50 rats (18 of which died
early). The results suggest that the low dose (38 mg/kg-day) was a lowest-observed adverse-effect
level (LOAEL), indicating that 1,2-dibromoethane induced early development of testicular
lesions.
Mice: Male and female mice were initially treated with low and high doses of 60 and 120
mg/kg/day, respectively. At week 11, low- and high-dose concentrations were increased to 100
and 200 mg/kg-day, respectively. Treatment with increased doses continued for 2 weeks and then
returned to initial concentrations at week 13. The high-dose treatments were further decreased to
60 mg/kg/day in week 40. At week 54, compound treatment was ceased in both low- and high-
dose groups. All surviving male mice and high-dose female mice were sacrificed by week 78,
while low-dose females were sacrificed at week 90. Time-weighted average low- and high-doses
were 62 and 107 mg/kg/day, respectively, for mice of both sexes.
Mortality was high in all treated groups (p < 0.001). Only 20/50 low- and 10/50 high-
dose males survived for at least 58 weeks. Due to excessive mortality, males were sacrificed at
week 78. Of the males that survived to at least week 26, 44/46 low-dose and 29/40 high-dose
males exhibited squamous-cell carcinoma of the forestomach. In females, 28/50 low-dose and
22
-------
8/50 high-dose animals survived at least 70 weeks. Low- and high-dose females were terminated
at weeks 90 and 78, respectively. As in males, early development of squamous-cell carcinoma of
the forestomach was observed, with the first occurrence in the low-dose group at week 40 and the
first occurrence in the high-dose group at week 34. Of the high-dose females that survived to at
least week 34, 28/35 exhibited squamous-cell carcinoma of the forestomach.
Squamous-cell carcinoma of the forestomach was the primary tumor observed in both
sexes. Similar to rats, this neoplasm metastasized throughout the abdominal cavity and lung. In
addition, lung adenomas were observed in all mouse exposure groups of both sexes and were
considered compound-related. There was a significant positive association between dosage and
incidence for both sexes. Although there was little evidence of lung carcinomas (one in the low-
dose females), the adenomas nevertheless are considered evidence of a carcinogenic response and
are relevant for human risk assessment. The incidence of these selected neoplastic lesions in the
mouse is summarized in Table 4-2.
Non-neoplastic lesions were observed primarily in the forestomach of treated mice.
Acanthosis was observed in 0/40 controls, 1/50 low-dose, and 5/49 high-dose males and in 0/40
controls and 9/50 high-dose females. Hyperkeratosis of the forestomach was also observed in
13/49 high-dose males and in 0/40 controls, 1/49 low-dose, and 12/50 high-dose females.
Testicular atrophy was also observed in 10/47 high-dose male mice (none in 39 control animals or
in 45 low-dose animals). Thus, 1,2-dibromoethane appears to have caused early development of
testicular atrophy as it had in male rats.
The carcinogenicity of 1,2-dibromoethane and its potential metabolites was the focus of a
study by Van Duuren et al. (1985). Groups of 30 male and 30 female B6C3FJ mice were
administered bromoacetaldehyde, bromoethanol, or 1,2-dibromoethane in drinking water at a
concentration of 4 mM for each compound for 15-18 months. Untreated control groups,
consisting of 50 mice of each sex, were given distilled drinking water. Based on measured
drinking water consumption, the authors estimated that 1,2-dibromoethane-treated male mice
received 116 mg/kg/day, and 1,2-dibromoethane-treated female mice received 103 mg/kg/day on
average.
Van Duuren et al. reported that 1,2-dibromoethane-treated animals had significant
mortality and weight loss that were not observed in the other treatment groups. Survival in males
23
-------
was approximately 25% at week 65 when the group was terminated, while survival in females
dropped to this level by about week 74. 1,2-Dibromoethane-induced squamous carcinoma of the
forestomach was observed in 26 of 28 male mice necropsied and in 22 of 29 female mice. In
addition, five forestomach papillomas were observed in these 29 females. Males also exhibited
eight squamous carcinomas of the glandular stomach. As in the NCI study, there were metastases
of the squamous carcinomas to other organs. Bromoethanol induced leukemia (2/29 in both
males and females compared with 2/50 for female controls and none in males) and had a high
incidence (10/29 males and 9/29 females compared with 1/50 in each of the control groups) of
forestomach papillomas. The incidence of tumor-bearing mice treated with bromoacetaldehyde
was lower than untreated controls. The authors concluded that it was unlikely that
bromoacetaldehyde or bromoethanol were the activated carcinogenic intermediates of 1,2-
dibromoethane.
Table 4-1. Incidence of tumors in Osborne-Mendel rats in 1,2-
dibromoethane oral gavage bioassay
Organ/tissue
Tumor
Control
0 mg/kg-day
incid.
%
Vehicle control
0 mg/kg-day
incid.
%
38 mg/kg-day
incid.
%
41 mg/kg-day
incid.
%
Males
Stomach
Forestomach papilloma or
tumor
Circulatory system
Hemangiosarcoma
Thyroid gland
Follicular cell adenoma or
carcinoma
0/20
0/20
1/20
0
0
5
0/20
0/20
0/20
0
0
0
45/50
11/50
5/50
90
22
10
33/50
4/50
8/49
66
8
16
Females
Forestomach
Tumors
Liver
Hepatocellular carcinoma
Circulatory system
Hemangiosarcoma
Adrenals
Adrenocortical carcinoma
0/20
0/20
0/20
0/20
0
0
0
0
0/20
0/20
0/20
0/20
0
0
0
0
40/50
1/47
1/49
0/44
80
2
2
0
29/50
5/48
3/48
4/45
58
10
6
9
24
-------
Source: NCI, 1978.
Table 4-2. Incidence of tumors in B6C3F1 mice in 1,2-dibromoethane oral
gavage bioassay
Organ/tissue
Tumor
Control
0 mg/kg-day
incid.
%
Vehicle control
0 mg/kg-day
incid.
%
62 mg/kg-day
incid.
%
107 mg/kg-day
incid.
%
Males
Forestomach
Squamous cell papilloma or
carcinoma
Lung
Adenoma
0/20
0/20
0
0
0/20
0/20
0
0
45/50
4/45
90
8
31/49
10/47
63
21
Females
Forestomach
Papillomas or tumors
Lung
Adenoma
0/20
0/20
0
0
0/20
0/20
0
0
47/48
10/48
92
20
28/50
5/50
56
10
Source: NCI, 1978.
4.2.2. Carcinogenicity Bioassay and Chronic Inhalation Studies
The National Toxicology Program (NTP, 1982) performed an inhalation carcinogenicity
bioassay in rats and mice. Male and female Fischer 344 rats and B6C3FJ mice (n = 50 per sex,
species, and exposure group) were exposed to 0, 10, or 40 ppm (0, 77, or 307 mg/m3) 1,2-
dibromoethane for 6 hr/day, 5 days/week. The study was designed to assess potential adverse
effects of 1,2-dibromoethane following 103 weeks of exposure. However, high mortality in both
species prompted early termination in some of the exposure groups.
Rats: High-exposure male rats exhibited high mortality (90%) resulting in termination of
that exposure group at week 88. Similarly, high-exposure female rats exhibited high mortality
(84%) and were terminated at week 91. Mortality in control and low-exposure rats of both sexes
were comparable.
Mean body weights of high-exposure rats of either sex were decreased compared to
25
-------
controls throughout the study. However, body weight data were reported graphically in the text,
and determination of the week that the decrease in body weight became significant (> 10% of
controls) is difficult to determine. Nasal cavity tumors were the primary neoplastic lesion
observed in rats. The total numbers of animals with primary nasal cavity tumors were
significantly increased (p < 0.001) in 1,2-dibromoethane-exposed male rats (0/50 control, 39/50
low-exposure, and 41/50 high-exposure) and female rats (1/50 control, 34/49 low-exposure, and
43/50 high-exposure). These summary statistics were presented in Tables 5 and 6 of the NTP
report. Table 4-3 is a summary of the incidences of nasal cavity tumor types, including those
from animals that died on study.
Other neoplastic lesions were also observed. Hemangiosarcoma of the spleen was
observed in 1 low-exposure and 15 high-exposure male rats, and this tumor type was also
observed in 5 high-exposure females. Male rats had a dose-dependent increase in mesothelioma
of the tunica vaginalis (1/50 control, 13/50 low-exposure, and 26/50 high-exposure). Female rats
had a statistically significant increase in mammary fibroadenomas in both exposure groups (4/50
control, 29/49 low-exposure, and 24/50 high-exposure). In addition, 4/50 high-exposure females
were observed to have adenocarcinoma of the mammary glands compared to 1/50 for controls.
Lung carcinoma (4/50) and adenoma (1/50) were also reported in high-exposure females.
Several non-neoplastic lesions were considered related to treatment. A dose-dependent
increase in hepatic necrosis was observed in 2/50 control, 6/50 low-exposure, and 19/50 high-
exposure males and 2/50 control, 3/49 low-exposure, and 13/48 high-exposure females. Toxic
nephropathy was observed in 4/50 low-exposure and 28/50 high-exposure males and 8/50 high-
exposure females. This lesion was not observed in any of the controls. Dose-dependent testicular
degeneration (1/50 control, 10/50 low-exposure, and 18/49 high-exposure) and atrophy (1/50
control, 2/50 low-exposure, and 5/49 high-exposure) was observed in male rats, and spermatic
granulomas were observed in high-exposure males (2/49). Female rats were observed to have a
dose-dependent degeneration of the adrenal cortex (4/50 control, 7/49 low-exposure, and 13/47
high-exposure). Adrenocortical degeneration was only observed in one male from each exposure
group. Retinal atrophy was also observed in 1/50 low-exposure male, 10/50 low-exposure
females, and 5/50 high-exposure females.
Mice: As mentioned previously, high mortality prompted early termination of some of the
mouse exposure groups. Low-exposure female mice displayed moderate mortality (62%) with
26
-------
controls displaying 20% mortality. Exposures were not terminated until the end of the
experiment (104-106 weeks). High-exposure female mice exhibited excessive mortality (86%)
with treatment terminated at week 90. Mortality was high in exposure and control groups of male
mice, and all male mice were sacrificed at 78 weeks. The principal cause of mortality in male
mice was ascending, suppurative urinary tract infection, progressing to necrotic and ulcerative
lesions around the urethral opening; chronic or suppurative cystitis; and ascending, suppurative
pyelonephritis. These effects in the male mice were not related to exposure.
Mean body weights of high-exposure mice were decreased compared to controls
throughout the study. However, body weight data were reported graphically in the text, and
determination of the week that decrease in body weight became significant (> 10% of controls) is
difficult to determine. As in rats, the principal neoplastic lesions observed were manifest in the
respiratory tract. However, tumors were primarily found in the lung of mice and not the nasal
cavities. The total numbers of animals with lung tumors were significantly increased (p < 0.001)
in high-exposure male mice (0/41 control, 3/48 low-exposure, and 25/46 high-exposure) and
female mice (4/49 control, 11/49 low-exposure, and 42/50 high-exposure) (p < 0.045 and <
0.001, respectively). These summary statistics were presented in Tables 8 and 9 of the NTP
report. Table 4-4 is a summary of the incidence of lung tumor types, including those from animals
that died on study.
In addition to the above-mentioned lung neoplasms, high-exposure females had an
increased incidence of carcinoma (6/50), adenoma (2/50), and adenomatous polyp (3/50) of the
nasal cavity. Hemangiosarcomas were observed in 2 high-exposure males, 11 low-exposure
females, and 23 high-exposure females, and hemangiomas occurred in 2 high-exposure males, 1
low-exposure female, and 4 high-exposure females. These lesions were primarily observed in the
retroperitoneal cavity in areas adjacent to the adrenal glands, kidneys, ovaries, and uteri of female
mice. Occasionally the hemangiosarcoma and hemangioma invaded the adjacent organ.
Fibrosarcoma was observed in 2 high-exposure males, 5 low-exposure females, and 15 high-
exposure females, with the majority of fibrosarcomas located in subcutaneous tissues. Malignant
mammary tumors (adenocarcinoma, adenocarcinoma with squamous metaplasia, or
adenosquamous carcinoma) were reported in 2/50 controls, 18/50 low-exposure, and 10/50 high-
exposure females.
Non-neoplastic lesions were generally not observed in male mice probably due to the high
27
-------
mortality. The majority of non-neoplastic lesions observed in female mice were located in the
respiratory tract. Dose-dependent epithelial hyperplasia of the lung alveoli, bronchioles, and
bronchi was observed. Hyperplasia was also observed in the nasal cavities of high-exposure
females (13/50). Suppurative inflammation of the nasal cavity was also observed in low-exposure
(4/50) and high-exposure (20/50) females. In addition, adenomatous hyperplasia of the lung was
reported in high-exposure females (37/50). In other tissues of female mice, dose-dependent
hematopoiesis in the spleen was observed (0/50 control, 8/49 low-exposure, and 16/49 high-
exposure), and a low, dose-dependent incidence of hepatic necrosis was also reported (0/50
control, 2/50 low-exposure, and 5/50 high-exposure).
Proliferative lesions in the nasal epithelium have been reported in mice following long-
term inhalation exposure to 1,2-dibromoethane (Stinson et al., 1981). Groups of 50 male and 50
female B6C3FJ mice were exposed to 10 or 40 ppm (77 or 307 mg/m3) 1,2-dibromoethane 6
hours/day, 5 days/week for 103 (10 ppm) or 90 (40 ppm) weeks. At 10 ppm (77 mg/m3), 1 male
and 3 female mice developed focal epithelial hyperplasia, which increased in incidence to 10
male and 11 female mice at 40 ppm (307 mg/m3) exposure. Benign neoplasms were observed in
the 40 ppm (307 mg/m3) exposure groups of male and female mice and identified as squamous
papilloma (three in males and seven in females) and adenomas (two females); the latter were also
observed to have squamous papilloma. Benign neoplasms first appeared in male mice at 59
weeks and in female mice at 79 weeks. Carcinomas were present in seven female mice exposed
to 40 ppm (307 mg/m3) 1,2-dibromoethane. The first carcinoma appeared at 45 weeks, and
carcinomas were identified as squamous carcinoma (n = 2), adenocarcinoma (n = 2), and mixed
carcinoma (n = 3). Sarcomas were observed in one 10 ppm (77 mg/m3) female (characterized as
poorly differentiated) and two 40 ppm (307 mg/m3) females (hemangiosarcoma). Mortality was
not reported, and no other toxicological parameters were monitored.
4.2.3. Subchronic Inhalation Studies
Preliminary to chronic bioassays, the NTP (1982) conducted subchronic inhalation studies
in rats and mice. Male and female Fischer 344 rats (n = 4 to 6 per sex and exposure group) and
B6C3Flmice (n = 10 per sex and exposure group) were exposed to 0, 3, 15, or 75 ppm (0, 23,
115, or 576 mg/m3) 1,2-dibromoethane, 6 hours/day, 5 days/week for 13 weeks. No deaths were
reported for rats at any exposure group. High-exposure female rats had depressed weight gain,
and male rats exhibited a dose-dependent depression of weight gain for all exposures. In rats of
28
-------
both sexes exposed to 75 ppm (576 mg/m3) 1,2-dibromoethane, swelling and/or vacuolation of
adrenal cortical cells and decreases in thyroid follicular size were observed. In mice, 4/10 males
in the 3 ppm (23 mg/m3) exposure group and 1 female in the 75 ppm (576 mg/m3) exposure group
died prior to the termination of the study. A dose-dependent decrease in body weight was
observed for both sexes. Eye irritation was observed in both sexes in the 75 ppm (576 mg/m3)
exposure group at weeks 12 and 13. In the high-exposure mice, 3 males and 9 females exhibited
megalocytic cells in the lining of the bronchioles. Based on the frank effects observed at 75 ppm
(576 mg/m3), 10 and 40 ppm (77and 307 mg/m3) exposures to 1,2-dibromoethane were chosen for
chronic toxicity and cancer studies described above.
Nitschke et al. (1981) described the results of a study designed principally to evaluate the
role of 1,2-dibromoethane in inducing nasal lesions. Male and female F344 rats were exposed by
inhalation to 1,2-dibromoethane at 0, 3, 10, or 40 ppm (0, 23, 77, or 307 mg/m3) for 6 hours/day,
5 days/week for 13 weeks. Forty male and 20 female rats were used per exposure group, and
serial sacrifices of 10 males per exposure group were conducted at 1, 6, and 13 weeks; 10 females
per exposure group were sacrificed at 13 weeks. The remaining male and female animals were
sacrificed 88-89 days postexposure. No treatment-related effects were observed in the urinalysis
of male rats. However, the female rats in the 40 ppm (307 mg/m3) exposure group exhibited a
slight decrease in specific gravity of the urine compared to controls; this parameter returned to
normal in the postexposure period. No hematological effects of toxicological significance were
observed in any of the exposure groups.
29
-------
Table 4-3. Nasal cavity tumor types in rats following chronic inhalation of
1,2-dibromoethane
Tumor type
Adenomatous polyp
Adenoma
Adenocarcinoma
Carcinoma
Squamous cell
carcinoma
Control
male
0/50
0/50
0/50
0/50
0/50
10 ppm
male
18/50
11/50
20/50
0/50
3/50
40 ppm
male
5/50
0/50
28/50
21/50
3/50
Control
female
0/50
0/50
0/50
0/50
1/50
10 ppm
female
5/50
11/50
20/50
0/50
1/50
40 ppm
female
5/50
3/50
29/50
25/50
5/50
Source: NTP, 1982.
Table 4-4. Lung tumor types in mice following chronic inhalation of
1,2-dibromoethane
Tumor type
Carcinoma-bronchus
Adenoma-bronchus
Adenomatous polyp-
bronchus
Adenomatous polyp-
bronchiole
Alveolar/bronchiolar
adenoma
Alveolar/bronchiolar
carcinoma
Control
male
0/41
0/41
0/41
0/41
0/41
0/41
10 ppm
male
0/48
0/48
0/48
0/48
0/48
3/48
40 ppm
male
0/46
2/46
3/46
2/46
11/46
19/46
Control
female
0/49
0/49
0/49
0/49
3/49
1/49
10 ppm
female
1/49
0/49
0/49
1/49
7/49
5/49
40 ppm
female
4/50
5/50
1/50
2/50
13/50
37/50
Source: NTP, 1982.
Males exposed to 40 ppm (307 mg/m3) showed significantly decreased body weights
throughout most of the 13-week exposure period, which returned to control levels during the
recovery period. No significant differences in body weight were observed for any other exposure
groups. Relative liver and kidney weights were significantly elevated in the 40 ppm (307 mg/m3)
males at 6 and 13 weeks. Females in the 40 ppm (307 mg/m3) exposure group had elevated liver
weights only. Organ weights returned to control levels during the recovery period. The principal
histopathological finding was scattered-to-diffuse nasal epithelial hyperplasia with focal necrosis
30
-------
at all sacrifice periods for animals exposed to 40 ppm (307 mg/m3). This increased in severity at
week 6 and progressed to diffuse or focal nonkeratinizing squamous metaplasia of the respiratory
epithelium by week 13. Males in the 10 ppm exposure group exhibited single or multiple nasal
epithelial hyperplasia foci at all three sacrifice intervals; this effect was also present in 10 ppm
females at 13 weeks. The respiratory epithelial effects in the 10 and 40 ppm (77 and 307 mg/m3)
groups had completely resolved, except for one animal, during the recovery period. No testicular,
kidney, liver, or lung effects were observed. Thyroid and adrenals, a target of 1,2-dibromoethane
in the NTP (1982) subchronic study, were not examined by histopathology. The study authors
identified 3 ppm (23 mg/m3) as the no-observed adverse effect level (NOAEL) from these results.
Reznik et al. (1980) also examined the respiratory system in both the rat and mouse
exposed subchronically to 1,2-dibromoethane. Male and female F344 rats (5
animals/sex/exposure group) and B6C3F1mice (10 animals/sex/exposure group) were exposed by
inhalation to 0, 3, 15, or 75 ppm (0, 23, 115, or 576 mg/m3) 1,2-dibromoethane for 6 hr/day, 5
days/week, for 13 weeks. Apparently, there was no mortality as none was reported.
Histomorphological changes were observed in the nasal cavity of both species exposed to 75 ppm
(576 mg/m3) 1,2-dibromoethane with a much lower incidence in rats at 15 ppm (115 mg/m3). The
concentration-dependent changes included cytomegaly, focal hyperplasia, squamous metaplasia,
and loss of cilia. Rats and mice exposed to 75 ppm (576 mg/m3) showed severe necrosis and
atrophy of the olfactory epithelium. No lesions were noted in any other tissue (e.g., liver, kidney,
and testis). NOAELs identified in this study were 3 ppm (23 mg/m3) in rats and 15 ppm (115
mg/m3) in mice. LOAELs were 15 ppm (115 mg/m3) in rats and 75 ppm (576 mg/m3) in mice.
Rowe et al. (1952) reported the results of a study that examined the toxicity of 1,2-
dibromoethane (99% pure) to rats, rabbits, guinea pigs, and monkeys following inhalation
exposure. Chamber concentrations were within 10% of desired concentrations. Examination of
tissues (lung, heart, liver, kidney, spleen, testis, pancreas, and adrenal gland) was performed with
light microscopy. The strains of the animals were not specified. Rats were exposed to 50 ppm
(384 mg/m3) 1,2-dibromoethane for 7 hours/day, 5 days/week, for 91 days. High mortality (50%)
in male rats was attributed to pneumonia and infections of the upper respiratory tract. Male rats
also exhibited statistically significant increases in relative lung, kidney, and liver weights and
decreased testicular weight. In females, 4/20 animals died before study termination, and relative
liver and kidney weights were increased while spleen weight was decreased. Histological
examination did not reveal any significant changes except patches of pneumonic consolidation in
31
-------
male rats. Guinea pigs were exposed in the same manner as rats, except that exposure was
terminated after 80 days. Body weight gain was depressed in both sexes, but mortality was not
different than in controls. Microscopic examination of tissues revealed a slight central fatty
degeneration of the liver and slight internal congestion and edema of the kidney tubular
epithelium. No other changes were observed. There were no apparent effects of toxicological
significance in rabbits exposed to 50 ppm (384 mg/m3) 1,2-dibromoethane for 7 hr/day, 5
days/week, for 84 days. There were slight, but not significant, increases in liver and kidney
weights.
Monkeys exposed to 50 ppm (384 mg/m3) 1,2-dibromoethane for 7 hr/day, 5 days/week, for
70 days exhibited fatty degeneration of the liver and increased relative kidney weight. Animals
were also described as appearing ill, nervous, and unkempt throughout the experimental period.
In addition, the same species were exposed to 25 ppm (192 mg/m3) 1,2-dibromoethane 7 hr/day, 5
days/week, for 213, 205, 214, and 200 days for rats, guinea pigs, rabbits, and monkeys,
respectively. No adverse effects were noted in any species except for high mortality in male rats
(50%) and male and female guinea pigs (50 and 25%, respectively).
The investigators stated that rabbits and monkeys and probably rats and guinea pigs can
tolerate daily repeated 7-hour exposures to 25 ppm (192 mg/m3) 1,2-dibromoethane without
adverse effects. However, this statement should be regarded with extreme caution due to the
limitations of this study. Evidence from the previously mentioned subchronic studies (Nitschke et
al., 1981; Reznik et al., 1980) suggests that the nasal cavity in rats is the major target organ
following inhalation exposure to 1,2-dibromoethane. This tissue was not examined by Rowe et
al. (1952). In addition, the study is limited because of the number of guinea pigs (7 per sex in 50
ppm exposure group, 8 per sex in 25 ppm exposure group), rabbits (3 male and 1 female for both
exposure groups), and monkeys (1 per sex for both exposure groups) used in this study. The use
of a larger number of animals might have detected significant adverse effects. Also, the study
provided limited details of methods used.
4.2.4. Other Studies
Male Fischer 344 rats (eight animals per dose group) were given 40 or 80 mg/kg 1,2-
dibromoethane in corn oil by gavage 5 days per week for 2 weeks (Ghanayem et al., 1986). Fifty
32
-------
percent of high-dose rats exhibited an increased incidence of forestomach cellular proliferation
compared to none in the low-dose group, vehicle controls, and two negative control groups.
Hyperkeratosis was also significantly increased in high-dose animals. The study authors
concluded that induced forestomach cellular proliferation provides a "favorable environment" for
neoplastic development in the forestomach.
The potential influence of disulfiram on the metabolism and carcinogenicity of 1,2-
dibromoethane has been studied (Wong et al., 1982). During the oxidative metabolism of 1,2-
dibromoethane, 2-bromoacetaldehyde is formed, which then undergoes further oxidative
metabolism or conjugation with GSH. Disulfiram inhibits aldehyde dehydrogenase and may alter
the metabolism of 1,2-dibromoethane by preventing further metabolism of 2-bromoacetaldehyde.
The study authors hypothesized that because disulfiram is used to treat alcoholics, alcoholics
might be at a greater risk to 1,2-dibromoethane toxicity due to modified metabolism. To test this
hypothesis, four groups of 48 male and 48 female Sprague-Dawley rats received either control air,
control air and 0.05% disulfiram in the diet, 20 ppm (154 mg/m3) 1,2-dibromoethane and control
diet, or 20 ppm (154 mg/m3) 1,2-dibromoethane and 0.05% disulfiram in the diet for 18 months.
1,2-Dibromoethane air concentrations were maintained for 7 hr/day, 5 days/week.
Rats in the control air/0.05% disulfiram group showed decreased body weight gain
throughout the experimental period. They also had consistently lower body weight gains than rats
exposed to 1,2-dibromoethane alone. Rats in the 20 ppm (154 mg/m3) l,2-dibromoethane/0.05%
disulfiram group displayed lower body weight gains compared to all other exposure groups
throughout the experiment. Weight gain reduction appeared to be correlated with decreased food
consumption except in animals exposed to 1,2-dibromoethane alone. Excessive mortality was
observed in animals receiving 1,2-dibromoethane/control diet and 1,2-dibromoethane/disulfiram
compared to controls and disulfiram-diet rats. 1,2-Dibromoethane/control rats had normal
hematological parameters, but 1,2-dibromoethane/disulfiram animals showed decreased
hematocrit, hemoglobin, and red blood cell (RBC) counts. Control air/disulfiram animals had an
increased incidence of hemosiderosis in the spleen, while females displayed an increase in
mammary tumors.
Tumor incidence is listed in Table 4-5. Rats exposed to 1,2-dibromoethane alone had an
increased incidence of splenic hemangiosarcoma and adrenal tumors with males exhibiting an
increase in subcutaneous mesenchymal tumors and females having a high incidence of mammary
33
-------
tumors. 1,2-Dibromoethane/disulfiram rats exhibited an increase in liver, kidney, and thyroid
tumors compared to rats receiving 1,2-dibromoethane or disulfiram alone. Hemangiosarcoma
was present in the liver, spleen, and mesentery, and males had an increase in lung tumors and
testicular atrophy when compared to animals receiving 1,2-dibromoethane alone. From this
study, it appears that disulfiram increases the toxicity and carcinogenicity of 1,2-dibromoethane.
Table 4-5. Enhancement of 1,2-dibromoethane-induced tumor with disulfiram
coadministration in rats
Tumor type
Liver
Kidney
Adrenal
Subcutaneous
Thyroid
Lung
Hemangiosarcoma
Spleen
Mesentery
Control/
control diet
Male3
0
0
2
3
4
0
Female3
0
0
1
0
5
0
Control/
disulfiram diet
Male3
1
0
1
1
1
0
Female3
0
0
0
0
1
0
1,2-
Dibromoethane/
control diet
Male"
2
o o
JJ
11
11
o
J
3
Female3
3
1
6
1
1
0
1,2-
Dibromoethane/
disulfiram diet
Male3
36
17
6
4
18
9
Female0
32
7
8
4
18
2
0
1
0
8
0
4
0
18
6
5
0
9
30
15
19
11
a Groups consisted of 48 rats at the initiation of the study; all were examined histopathologically.
b Group consisted of 48 rats at theinitiation of the study, but two rats were not examined due to autolysis.
0 Group consisted of 48 rats at the initiation of the study, but three rats were not examined due to autolysis.
Source: Wong et al. (1982).
1,2-Dibromoethane has been evaluated in the A/J mouse lung tumor bioassay by several
routes of exposure (Stoner et al, 1986; Adkins et al, 1986). Stoner et al. (1986) exposed groups of
approximately 16 male and 16 female A/J mice to 1,2-dibromoethane either by gavage at 840
mg/kg or intraperitoneally at 168, 420, or 840 mg/kg. Control groups received a comparable
injection of vehicle, trycaprylin-2 for the gavage study and tricaprylin-1 for the intraperitoneal
study. Injections were administered 3 times per week for a total of 24 injections. The mice were
observed until 24 weeks after the first administration, then sacrificed and examined for lung
34
-------
tumors. The results are summarized in Table 4-6. Both male and female mice showed increased
tumor incidence at 840 mg/kg, both by oral and intraperitoneal administration, of at least twice
the corresponding control response. The response in the high-dose female mice by the
intraperitoneal route, at 14/16 (88%), was markedly higher than that in the high-dose male mice
(44%). The authors concluded that 1,2-dibromoethane was active in only the females at only the
highest dose when given intraperitoneally and inactive in both sexes when given orally. Note that
these studies have limited power to detect increases in tumor incidence, relative to chronic
studies; that is, if each treatment group had started with 50 mice, all of the high-dose responses,
except in the male mice treated intraperitoneally, would have been statistically significantly
increased relative to vehicle control. While this outcome is driven in part by the lower than
average response in the vehicle control mice, at about 14 - 20% compared with about 30% across
all of the untreated and vehicle controls, this experiment does not clearly rule out the possibility
that oral administration of 1,2-dibromoethane is associated with induction of lung adenomas in
A/J mice.
In a second set of studies, Adkins et al. (1986) exposed groups of 30 and 60 female A/J
mice by inhalation to 0, 20 or 50 ppm 1,2-dibromoethane for 6 hr/day, 5 days/week, for 6 months
(24 weeks) and then sacrificed and examined for lung tumors. In the first study, which started
with 30 mice per group, 51% of the control animals and 100% of all surviving dibromoethane-
exposed animals developed lung adenomas. A dose-related increase in pulmonary adenoma
formation was observed in the second study with 60 mice per group: 26%, 68%, and 100% of
the surviving control, low-dose, and high-dose mice, respectively, developed lung adenomas.
The authors concluded that the two studies together indicated a concentration-related increase in
the frequency and incidence of adenoma formation, significant and reproducible at 50 ppm.
These data are also summarized in Table 4-6.
These studies demonstrate qualitative evidence of lung tumor induction by several routes of
exposure. Because only results from animals who survived the 24-week experimental period are
considered in the A/J mouse lung tumor assay, it is possible that tumor incidences are
underreported and would therefore be inadequate for a quantitative dose-response assessment.
The evidence does not clarify whether lung tumors are solely the result of portal of entry effects
in the case of inhalation exposure to dibromoethane but does demonstrate that lung tumors can
result from systemic exposure to dibromoethane.
35
-------
Table 4-6. Lung tumor incidence in A/J mice following exposure to 1,2-
dibromoethane via several routes of exposure
Sex
Exposure level
Number of
mice placed
on test
Number of
mice
surviving at
24 weeks
Number (%) of surviving
mice with lung tumors
Oral exposure"
Male
Female
840 mg/kg, 3 times/week
0 mg/kg, 3 times/week
840 mg/kg, 3 times/week
0 mg/kg, 3 times/week
16
16
16
16
16
15
16
14
7 (44)
(20)
5 (31)
(14)
Intraperitoneal exposure3
Male
Female
840 mg/kg, 3 times/week
420 mg/kg, 3 times/week
168 mg/kg, 3 times/week
0 mg/kg, 3 times/week
840 mg/kg, 3 times/week
420 mg/kg, 3 times/week
168 mg/kg, 3 times/week
0 mg/kg, 3 times/week
16
16
15
16
16
16
17
16
16
15
14
15
16
16
17
15
7 (44)
3 (20)
2 (14)
(30)
14 (88)
9 (56)
4 (24)
(30)
Inhalation15
Female
Female
50 ppm, 6h/d, 5d/wk
20 ppm, 6h/d, 5d/wk
0 ppm, 6h/d, 5d/wk
50 ppm, 6h/d, 5d/wk
20 ppm, 6h/d, 5d/wk
0 ppm, 6h/d, 5d/wk
30
30
30
60
60
60
11
11
30
57
57
58
(100)
(100)
(51)
(100)
(68)
(26)
a Source: Stoneret al. (1986).
b Source: Adkins et al. (1986).
0 Only the percentage was reported.
36
-------
4.3. REPRODUCTIVE AND DEVELOPMENTAL STUDIES IN ANIMALS - ORAL AND
INHALATION
4.3.1. Inhalation Studies
The potential effects of 1,2-dibromoethane on reproduction in male and female rats have
been reported (Short et al., 1979). Male Charles River CD rats (9-10/group) were exposed whole-
body to 0, 19, 39, and 89 ppm (0, 146, 300, and 684 mg/m3) 1,2-dibromoethane for 7 hours/day,
5 days/week, for 10 weeks, and females (20/group) were exposed to 0, 20, 39, and 80 ppm (0,
154, 300, and 614 mg/m3) 1,2-dibromoethane for 7 hours/day, 7 days/week, for 3 weeks. Males
in the 89 ppm (684 mg/m3) exposure group and females in the 80 ppm (614 mg/m3) exposure
group gained less weight, consumed less food, and had higher mortality rates (21% males, 20%
females) than controls. In the 89 ppm (684 mg/m3) group, testicular weight and testosterone
concentrations were significantly reduced, and none of the males in this group were able to
impregnate non-exposed females after the 10-week exposure period, compared to a 90%
impregnation rate for the other exposed males. Also, atrophy of the testis, epididymis, prostate,
and seminal vesicle was observed in the 89 ppm (684 mg/m3) group. No treatment-related
reproductive effects were observed in males exposed to 19 or 39 ppm (146 or 300 mg/m3) 1,2-
dibromoethane. The litters from these males were normal with respect to total implants, viable
implants, and resorptions. The fact that measures of the quality and count of sperm were not
taken is considered a significant study limitation, however, given the effects observed in other
studies of human and bull sperm following 1,2-dibromoethane exposure.
After females were exposed for 3 weeks, they were mated with non-exposed males and
vaginal smears were taken. Vaginal smears were normal in the 20 and 39 ppm (154 and 300
mg/m3) exposure groups, but the 80 ppm (614 mg/m3) group was in constant diestrus and did not
begin a normal cycle until 3 or 4 days postexposure. This resulted in fewer females in this group
mating during a 10-day mating period with non-exposed males. For all 1,2-dibromoethane
exposure groups, all mated females were pregnant when sacrificed at mid-gestation and had
normal uterine contents for total implants, viable implants, and resorptions. Hormone levels
were not measured. In the high exposure group, there was an incidence of 6/20 animals with mild
vacuolated degeneration of the epithelium of the uterus (3/20 in controls) and 3/20 incidence of
ovarian cysts (0/20 in controls). The study authors concluded that, although adverse reproductive
37
-------
effects were observed in both sexes, these effects only occurred at concentrations that were
associated with significant morbidity and mortality.
The effects of 1,2-dibromoethane administered by inhalation to rats and mice during
gestation were also reported by Short et al.(1978). Pregnant Charles River CD rats (15-17/group)
and CD-I mice (18-22/group) were exposed whole-body at concentrations of 0, 20, 38, and 80
ppm (0, 154, 292, and 614 mg/m3) for 23 hr/day. Exposure began on day 6 of gestation and lasted
for 10 days. High mortality was observed in rats exposed to 80 ppm (614 mg/m3) 1,2-
dibromoethane, and weight loss was apparent in the 38 and 80 ppm (292 and 614 mg/m3)
exposure groups. Feed consumption was decreased in all exposure groups and failed to recover in
the high-exposure group after exposure. Total number of implants was decreased, and the
number of resorptions increased in the rat high-exposure group. However, there was no
examination of the uterus at necropsy to definitively determine the number of implantation sites.
Inasmuch as there was no effect of exposure in the Short et al. (1979) intermittent-exposure study,
the effects seen here are likely due to the continuous exposure. Decreased fetal weight was noted
in the 38 ppm (292 mg/m3) exposure group. No viable fetuses were observed in the high-
exposure group. None of the external or soft tissue anomalies (occluded nasal passage,
hydronephrosis, solidified kidney cortex, distended urinary bladder, inferior vena cava
hemorrhage, and blunt snout) noted in rat fetuses were dose-dependent.
Mice exhibited high mortality in the 38 ppm (292 mg/m3) group and complete mortality in
the 80 ppm (614 mg/m3) group. Weight gain was reduced in mice that were exposed to 20 and 38
ppm (154 and 292 mg/m3) 1,2-dibromoethane but returned to normal in all but one animal in the
38 ppm (292 mg/m3) group. From this study, it appears that pregnant mice may be more
sensitive to 1,2-dibromoethane than pregnant rats. Also, the CD-I may be more sensitive than the
B6C3Fj as mortality was not seen (Reznik et al., 1980) in the latter strain after 13 weeks of
exposure to levels higher than 38 ppm (292 mg/m3). The 20 ppm (154 mg/m3) mice had an
increase in late resorptions and decreased fetal weight compared to controls, and mice exposed to
38 ppm (292 mg/m3) 1,2-dibromoethane had a decrease in viable fetuses, increased resorptions,
and reduced fetal body weights. Although the observed fetotoxic and teratogenic effects
occurred at exposures that also caused maternal toxicity, it can be difficult to distinguish direct
effects of 1,2-dibromoethane on the fetus from secondary effects resulting from maternal toxicity.
Smith and Goldman (1983) examined potential behavioral effects in offspring of rats
38
-------
exposed to 1,2-dibromoethane. Pregnant female Long-Evans hooded rats (12/group) were
exposed to 0, 0.43, 6.67, or 66.67 ppm (0, 3, 51, or 512 mg/m3) 1,2-dibromoethane by inhalation
for 4 hr/day, 3 days/week from day 3 to 20 of gestation. Defecation during exposure was directly
related to the concentration of 1,2-dibromoethane (p < 0.001). Mid- and high-exposure animals
defecated significantly more (p < 0.01 and p < 0.05, respectively) than controls. There was also
an inverse relationship between exposure and weight gain that was statistically significant. There
was no significant difference in the number of delivering females or litter size, but low- and mid-
exposure pups weighed significantly more than controls while high-exposure pups weighed
significantly less than controls. These weight effects in pups were not observed by day 66 post-
gestation.
In newborn animals, rotorod performance on days 30 and 63 post-gestation was
significantly (p < 0.005) increased in the mid- and high-exposure groups (6 and 6 per group). The
toxicological significance of an increase in rotorod performance is not clear since there was a
differential effect at certain rod speeds as well as a sex times speed interaction. The terminal
performances of the mid- and high-exposure offspring in T-maze discrimination were also
significantly (p < 0.001) different, with the high-dose group performing better than control. No
other behavioral parameters in offspring were influenced by exposure. Measures of maternal
behavior (nest rating and pup retrieval) were not influenced by exposure to 1,2-dibromoethane.
The authors concluded that 1,2-dibromoethane produced long-term and possibly permanent
alterations in the behavior of exposed offspring. The increased defecation and decreased weight
gain (p < 0.05) in dams suggest maternal stress could have influenced behavior in the offspring,
but the number of neonates tested was too small to make a definitive conclusion.
4.3.2. Oral Studies
Several studies have examined the potential reproductive toxicity of 1,2-dibromoethane in
bulls after oral treatment. Amir and Ben-David (1973) examined the effects of 1,2-
dibromoethane on bull spermatozoa. Three bulls (15 - 20 months) were administered (by gelatin
capsules) 10 doses of 1,2-dibromoethane (4 mg/kg) on alternate days. Semen was collected 2-3
times per week before, during, and for 2 - 3 months after treatment. During the third week after
the start of treatment, sperm abnormalities became evident. 1,2-Dibromoethane-induced effects
in the sperm included coiled tails, acrosomic effects, acrosome loss, decreased motility,
39
-------
degeneration, and disintegration. From the beginning of the fourth week and until day 40 after
the first treatment, approximately 90 - 100% of the sperm were observed to be abnormal.
Gradually, sperm abnormalities decreased, and, approximately 60 days after the first treatment
(one month post-treatment), the percent of sperm abnormalities had returned to control levels.
Amir and Volcani (1967) provided further evidence of the effect of 1,2-dibromoethane on
sperm by examining bull testes histologically. Three bull calves (4 days old) were orally treated
with 2 mg/kg 1,2-dibromoethane for 17.5 - 22.5 months. Semen was collected with the aid of an
artificial vagina for 4 - 6 six months prior to castration of one testis of each bull. Histological
examination of the castrated testis revealed a depopulation in the majority of seminiferous
tubules, the lumens of which were either empty or filled with cell debris. The caput and corpus
epididymis were empty of spermatozoa and connective-tissue cells were thickened around the
ductus. A high pseudostratified epithelium was also observed in both the caput and corpus. 1,2-
Dibromoethane administration was discontinued after castration, and recovery of semen
properties was monitored. In two bulls, recovery of semen properties was complete within 3-4
months. However, the third animal had decreased sperm density and motility for several more
months although no sperm morphological abnormalities were observed. Seven months after
castration, the bulls were slaughtered and the remaining testis was examined histologically.
Seminiferous tubules of two of the bulls were normal, but in the third animal the majority of the
seminiferous tubules remained inactive, showing hyalinization and hyperplasia of the interstitial
tissue. The caput and corpus epididymis were normal in the two bulls but were still abnormal in
the third animal.
The effects of 1,2-dibromoethane on sperm production in bulls have also been studied
(Amir and Volcani, 1965). Four bull calves (4 days old) were fed 1,2-dibromoethane for 14-16
months. During the first 3 months of the calves' life, 2 mg/kg-day of 1,2-dibromoethane was fed
to calves in milk. 1,2-Dibromoethane was then administered in feed (2 mg/kg-day) for an
additional 9 months. After the first year of life, the method of administration was again altered by
administering 4 mg/kg of 1,2-dibromoethane to calves in gelatin capsules on alternate days. 1,2-
Dibromoethane treatment did not appear to affect the growth or health of the animals compared to
controls. The libido of the treated bulls was similar to that of untreated animals. However,
effects on the sperm were observed when semen was collected at termination of 1,2-
dibromoethane administration. These consisted of abnormalities (tailless, coiled tails, pyriform
heads), low sperm density, and poor sperm motility. Recovery after discontinuation of the
40
-------
treatment varied from 10 days to approximately 3 months in different animals.
Amir (1973) provided evidence that 1,2-dibromoethane affected the shape of the
spermatozoa during maturation in the epididymis and during spermogenesis. In this experiment,
two bulls (15-20 months old) were orally administered 4 mg/kg 1,2-dibromoethane on alternate
days for 12 and 21 days (7 and 10 doses, respectively). Smears of testicular spermatozoa,
different parts of the epididymis, and the ductus deferens were analyzed for morphological
abnormalities. In the animal that received seven doses of 1,2-dibromoethane, approximately 50%
of spermatozoa from the testis and 10% of the spermatozoa from the first two segments of the
caput epididymis had misshapen heads. Various segments of the epididymis also contained
sperm with tail and acrosomal defects. In the animals that received 10 doses, almost the entire
spermatozoa population of both the testis and caput epididymis had misshapen heads. The
number of spermatozoa in the corpus and caput epididymides with tail and acrosomal defects also
increased.
Amir and Lavon (1976) examined the protein changes in the epididymal and ejaculated
spermatozoa of young (15-18 months) and old (4.5 - 5.5 years) bulls following 1,2-
dibromoethane treatment. Bulls were administered 10 doses of 4 mg/kg 1,2-dibromoethane orally
on alternate days. Treatment with 1,2-dibromoethane did not significantly change total nitrogen,
amino acid, or lipoprotein content of epididymal and ejaculated spermatozoa. However, the
amino acid composition of spermatozoan proteins did reveal an increase in the percent isoleucine
and tyrosine of the caput epididymis, arginine, and glycine in the cauda epididymis, and proline in
ejaculate. Ejaculate lipoproteins had an increase in percent half-cystine and tyrosine and a
decrease in percent threonine, serine, glutamic acid, and isoleucine.
In addition to reporting on abnormal sperm, Amir et al. (1977) reported a reduction in
DNA and protein content of sperm in three bulls treated with 10 oral doses of 4 mg/kg 1,2-
dibromoethane on alternate days. Semen was collected twice weekly during treatment and for 5
weeks posttreatment and analyzed for morphological changes and DNA and protein content.
DNA and protein content decreased during the first 13 days postexposure but returned to normal
by days 20 and 27 postexposure for DNA and protein, respectively. A statistically significant
decrease in DNA content was observed in the corpus and cauda epididymis. Protein content was
only significantly reduced in the cauda epididymides. The study authors hypothesized that 1,2-
dibromoethane-induced sperm abnormalities might have been due to an alkylating effect of the
41
-------
chemical on the normal amino acid sequence of the sperm proteins during the replacement of the
somatic histones by sperm histones in late spermiogenesis.
Amir (1975) studied individual differences in the response of bulls to 1,2-dibromoethane
treatment. Thirteen young (15 - 24 months) and two adult (4.5-5 years) bulls received 10 oral
doses (by gelatin capsules) of 4 mg/kg 1,2-dibromoethane on alternate days. In untreated
animals, only 2 - 4% of spermatozoa had misshapen heads. However, in treated young animals, a
"large proportion" of spermatozoa had misshapen heads at 1 day postexposure. The area of the
genital tract with the greatest percentage of abnormal sperm varied among individual animals, but
generally, the ductus efferentes had the highest proportion (50 - 90%). In young bulls, the
maximum percentage (80 - 100%) of spermatozoa with misshapen heads that appeared in the
ejaculate varied from 2 to 10 days after end of treatment. Reasons for this variation may be
differences in the sperm transit time through the epididymis as well as the variation in the release
time of the affected spermatozoa from the testis. The study author states that the effects of 1,2-
dibromoethane treatment was more acute in adult bulls. Sperm concentration was only slightly
decreased in young bulls but was significantly decreased in adults. Also, the percentage of sperm
with misshapen heads returned to control levels (~ 5%) by week 5 postexposure in young bulls;
but, in adults, the percentage of abnormal spermatozoa at week 5 postexposure was
approximately 45% and remained elevated at 16 weeks postexposure (25%).
In all of these studies on bulls, the experimental design employed sample sizes that were
too small to indicate statistically significant increases in effects observed. However, the power of
all of these studies has been examined and found to be low for a number of reasons (Dobbins,
1987). Despite statistical problems with these studies, they provide a substantial qualitative
evidence that bull sperm is sensitive to 1,2-dibromoethane exposure.
42
-------
4.4. OTHER STUDIES
4.4.1. Reproductive/Developmental
The effects of 1,2-dibromoethane on semen quality and fertility in the rabbit have been
evaluated (Williams et al., 1991). Male New Zealand white rabbits (8-10 per group) were given
1,2-dibromoethane in corn oil at 0, 15, 30, or 45 mg/kg-day for 5 days by subcutaneous
administration. Weekly semen samples were collected for 6 weeks pre-exposure, during
exposure, and for 12 weeks postexposure. Semen samples were analyzed for sperm
concentration, number, morphology, viability, motion parameters, pH, osmolality, volume,
fructose, citrate, carnitine, protein, and acid phosphatase. Male fertility was assessed at 4 and 12
weeks by artificial insemination of three females/male with 1 million motile sperm.
Significant mortality (30%) was observed in the 45 mg/kg dose group. Food consumption
was decreased in dosed animals. There was no change in group mean body weight for any dose
group compared to controls. Liver function enzymes (SDH and ALT) were significantly elevated
in three surviving animals from the 45 mg/kg dose group. In the 45 mg/kg dose group,
curvilinear velocity, straight-line velocity, percent motility, and amplitude of lateral head
displacement were the only sperm characteristics statistically decreased by 1,2-dibromoethane
exposure. Motility was also decreased in the 30 mg/kg dose group. The low- and mid-dose
groups had significant decreases in ejaculate volume, and there was a dose-dependent decrease in
pH. Male fertility, fetal structural development, litter size, and mean fetal weight were not
affected by 1,2-dibromoethane exposure. There was no discussion about fetal survival. The
toxicological implications of this study are limited considering the route of administration.
The binding of 1,2-dibromoethane to fetal epithelial tissue in mice has been reported
(Kowalski et al., 1985, 1986). Pregnant C57BL mice in different stages of gestation (days 13-17)
were injected i.v. with 14C-1,2-dibromoethane (1.2-1.6 mg/kg) and then sacrificed. In addition,
three fetuses were injected in utero on days 17 or 18 of gestation to circumvent the effect of
maternal disposition on the fate of 1,2-dibromoethane in fetal tissues. Autoradiography and
computer-assisted image analysis were utilized to investigate the binding of 1,2-dibromoethane to
fetal epithelial tissue. Metabolites were not identified. At day 13 of gestation, radioactivity was
uniformly distributed in fetuses. Average radioactivity amounted to 6 and 55% of maternal blood
and liver levels, respectively, and was completely extracted by organic solvents. The fetuses
43
-------
from dams treated on days 16 or 17 of gestation had a high concentration of radioactivity in the
epithelia of the oral cavity, esophagus, forestomach, nasal mucosa, trachea, bronchi, and liver and
in the thymus, lens, and choroid plexus. The radioactivity of the last three tissues was completely
extractable by organic solvents, but the epithelial- and liver-associated radioactivity was only
partially extractable. The distribution of radioactivity in fetuses treated in utero was similar to
that of fetuses from dams treated on days 16 and 17. In vitro studies with excised fetal tissues
were performed to assess possible fetal epithelial metabolism of 1,2-dibromoethane. The results
indicated that fetal epithelia can produce 1,2-dibromoethane metabolites that bind to tissue.
Abnormal sperm have been observed in rams following subcutaneous treatment with 1,2-
dibromoethane (Eljack and Hrudka, 1979). Rams were administered 1,2-dibromoethane (7.8-
13.5 mg/kg-day) subcutaneously for 12 days. Sperm motility began to decline during week 5
after start of compound administration and declined maximally between weeks 9 and 10.
Acrosomal and nuclear abnormalities were also observed. The acrosomal abnormalities began to
appear in week 5 and were characterized by enlarged and misshapen apical segments. Nuclear
abnormalities appeared later and were characterized by a misshapen nuclei and the formation of
nuclear cristae.
The spermicidal effect of 1,2-dibromoethane in bulls and rams was reviewed by Amir
(1991). Differences in the pathology of spermatozoic effects between the two species was
discussed. While in the bulls, the abnormal spermatozoa issued from the affected spermatids
were also collected in the ejaculates; this was not the case with treated rams. In the rams, the
abnormal spermatids seem to be phagocytised in the epididymis before their arrival in the
ejaculate. In addition, whereas the alkylating effect of 1,2-dibromoethane occurred also in the
upper parts of the epididymis of the bulls, causing tail and acrosome defects to the spermatozoa,
in the rams such an effect seems to occur all along the epididymal duct. These differences
between bulls and rams in the sites of the genital tract where the chemical takes effect, and in the
mechanism of this effect, are an indication of probable differences in the physiology of the
reproductive tract between these species.
A model to assess potential developmental toxicity of dihaloalkanes in humans has been
described (Mitra et al., 1992). Glutathione S-transferase isozymes from human fetal liver were
purified and used to investigate the potential embryotoxicity of 1,2-dibromoethane in rat embryos
in culture. Five isozymes were detected in the human fetal liver. All enzymes were capable of
44
-------
metabolizing 1,2-dibromoethane. 1,2-Dibromoethane activation by one of the isozymes,
designated P-3, resulted in toxicity to cultured rat embryos. 1,2-Dibromoethane metabolism by
GST caused significant decreases in crown-rump length, yolk sac diameter, somite number, and
composite score for different morphological features. Central nervous and olfactory structures
were most severely affected. Yolk sack circulation and allantois were also affected. The
conclusion drawn was that 1,2-dibromoethane is a suspected developmental toxicant in humans.
In a study by Brown-Woodman et al. (1998), three solvents—chloroform,
dichloromethane, and dibromoethane—were examined for embryotoxic/teratogenic potential using
rat embryo culture. The results showed that each of the solvents had a concentration-dependent
embryotoxic effect on the developing rat embryo in vitro. The effect and no-effect concentrations
(expressed in |imol/mL culture medium), respectively, for each of the halogenated hydrocarbons
tested were dibromoethane—0.33, < 0.18; chloroform—2.06, 1.05; dichloromethane—6.54, 3.46.
Histological studies were performed after exposure of rat embryos to an embryotoxic level of
each of the halogenated hydrocarbons studied for increasing time periods up to the standard 40-
hour culture. Marked cell death in the neuroepithelium of the developing neural tube was a
prominent feature in all embryos exposed to an embryotoxic level of these solvents for periods of
16 hours or longer.
Bishop et al. (1997) examined alterations in the reproductive patterns of female mice
exposed to 1,2-dibromoethane. Female mice were given a single i.p. injection of 100 or 150
mg/kg 1,2-dibromoethane and allowed to mate with untreated males for the duration of the female
reproductive life span. 1,2-Dibromoethane treatment had no effect on total pups born, number of
litters per female, or first and second litter size.
4.4.2. Developmental Neurotoxicity
Fanini et al. (1984) studied the effects of paternal exposure of rats to 1,2-dibromoethane
on behavior in developing offspring. Male Fischer 344 rats were treated i.p. with a daily dose of
1.25, 2.5, 5.0, or 10 mg/kg 1,2-dibromoethane for 5 successive days and then mated with
untreated females at 4 or 9 weeks after treatment. A comprehensive behavioral assessment of
motor reflexes and motor coordination was conducted up to 21 days of age.
45
-------
Paternal 1,2-dibromoethane exposure did not alter the development of surface righting
ability in 3 - 6-day-old neonates or the ability of 10- and 15-day-old neonates to demonstrate
negative geotaxis. The acquisition of cliff avoidance in the Ft progeny bred at week 4 was
suppressed in progeny of 5 mg/kg males on days 4 and 5, but cliff avoidance was not significantly
different than controls after day 5. Cliff avoidance was not significantly different than in control
Fj progeny bred at week 9.
Swimming direction in week 4 progeny was significantly different (swimming in circles
rather than straight) than controls on days 6 and 8, but this was not observed on days 10 or 12.
No effect on swimming direction was observed in week 9 progeny. However, the ability of week
9 progeny to raise their heads higher with age when forced to swim was slower compared to
controls. Week 4 progeny did not exhibit this trait. Rats typically swim with hindlimbs while
keeping the forelimbs stationary, but significantly fewer animals swam in this manner on day 16
in week 4 progeny compared to controls. Open-field ambulation was significantly repressed in
weeks 4 and 9 Fx progeny on days 14 and 21. While the results are suggestive of developmental
neurotoxicity, none of these effects exhibited a linear dose-response relationship. The study
authors state that this is typical of behavioral teratology studies. The authors also hypothesize
that the premeiotic stages of spermatogenesis are sensitive to the genotoxic effects of 1,2-
dibromoethane.
Hsu et al. (1985) examined the activities of various neurotransmitter enzymes in the
developing brain of Ft progeny of 1,2-dibromoethane-treated males. Male Fisher 344 rats were
treated intraperitoneally with 5 daily doses of 1 mg/kg-day 1,2-dibromoethane. The activities of
choline acetyltransferase, acetylcholinesterase, and glutamic acid decarboxylase were examined
in various brain regions of the Fx progeny from 7 to 90 days of age. Selected brain regions
included the cerebellum, corpus striatum, frontal cortex, hippocampus, and hypothalamus.
Choline acetyltransferase was significantly decreased in the hypothalamus of 7-day-old
rats but was significantly increased in the cerebellum, corpus striatum, hippocampus, and
hypothalamus of 21-day-old rats. Acetylcholinesterase was significantly increased in the corpus
striatum and hippocampus of 7-day-old rats, decreased in the cerebellum, corpus striatum, and
hippocampus of 14-day-old rats, and increased in the hypothalamus and hippocampus of 21-day-
old rats. However, acetylcholinesterase was decreased in the cerebellum of 21-day-old rats.
Glutamic acid decarboxylase was increased in the corpus striatum of 21-day-old rats, but
46
-------
decreased in the frontal cortex of 21- and 90-day-old rats. Although this study did not examine
behavior, the study authors considered that these alterations might have been associated with
behavioral abnormalities observed in developing offspring by Fanini et al. (1984).
4.4.3. Genotoxicity
The evidence for 1,2-dibromoethane's potential genotoxicity is strong. 1,2-
Dibromoethane is a direct-acting mutagen in bacteria. 1,2-Dibromoethane was positive for S.
typhimurium revertant strains TA1535, TA100, and TA98 (Barber et al., 1981). Metabolic
activation was not necessary for the mutagenic effects. 1,2-Dibromoethane induced point
mutations in S. typhimurium strains TA1535 and TA100, S. coelicolor, and A nidulans (Carere
and Morpurgo, 1981). Wheeler et al. (2001) has shown that expression of a variety of GSTs
within S. typhimurium 1535 caused a dose-dependent increase in the number of revertants upon
incubation with 1,2-dibromoethane. The half-mustard, S-(2-bromoethyl)GSH, resulted in a
significant increase in the number of revertants with S. typhimurium (without expressed GSTs)
with the response dependent on whether the leaving group was Br-, C1-, or F1-. The first Br-
leaving group of 1,2-dibromoethane appeared to be associated with the highest reversion rate
fastest decomposition rate for the half-mustard. It was hypothesized that the stability of the half-
mustard likely plays an important role in its ability to enter cells and cause mutations.
1,2-Dibromoethane has also been shown to induce reproducible positive responses in
chromosomal aberrations and sister chromatid exchanges in Chinese hamster ovary cells (Ivett et
al., 1989; Tan and Hsie, 1981; Brimer et al., 1982; Ballering et al., 1998; Graves et al., 1996).
1,2-Dibromoethane enhanced the production of micronuclei in tetrads of microsporogenesis of
Tradescantia (Ma et al., 1978).
1,2-Dibromoethane has been demonstrated to cause hepatic DNA damage in rats
following oral administration (Kitchin and Brown, 1986, 1987). 1,2-Dibromoethane also induced
DNA damage in the stomach, kidney, liver, lung, and bladder in mice when administered
intraperitoneally (Sasaki et al., 1998). The 1,2-dibromoethane-induced DNA damage could not
be attributed to toxic cell death. DiRenzo et al. (1982) demonstrated that 1,2-dibromoethane can
bind covalently to calf thymus DNA in vitro. Following i.p. administration, label was shown to
bind DNA in the liver, kidney, stomach, and lung of rats and mice (Arfellini et al., 1984); binding
47
-------
was highest in the liver and kidney. S-[2-(N7-guanyl)ethyl]glutathione has been identified as the
major DNA adduct in rats following treatment with 1,2-dibromoethane (Kim et al., 1990; Koga et
al., 1986).
1,2-Dibromoethane was negative when tested for dominant lethal and electrophoretically-
detectable specific-locus mutations in germ cells of male DBA/2J mice following i.p. injection of
100 mg/kg 1,2-dibromoethane (Barnett et al., 1992). 1,2-Dibromoethane was also negative when
tested for micronucleated reticulocyte induction in mice (Asita et al., 1992). A slight but
significant increase in sister chromatid exchange was noted in mice administered 1,2-
dibromoethane by the i.p. route; however, the increase was not dose-related (Krishna et al., 1985).
Kale and Baum (1979) reported sex-linked recessive lethal mutations in spermatozoa ofD.
melanogaster. In a later study, Kale and Baum (1983) provided evidence that D. melanogaster
embryonic spermatogonia are particularly sensitive to 1,2-dibromoethane exposure.
Dusek et al. (2003) used chick embryo in ovo to investigate the effects of 1,2-
dibromoethane on hematopoiesis at a developmental stage where the primitive erythroid cells
divide and differentiate in circulation. Early after 1,2-dibromoethane treatment on embryonic day
3, annexin V/propidium iodide labelling showed acute cell death of erythroid elements, which
was subsequently compensated for by the release of immature cells into the circulation. At the
same time, the comet assay indicated increased DNA damage in 1,2-dibromoethane-exposed
blood cells when compared with controls. After embryonic day 5, there was no indication for
ongoing prominent cell death in the 1,2-dibromoethane-treated group. However, the DNA
damage assessed by the comet assay persisted until embryonic day 10 in the peripheral blood
cells and for even longer in cells from thymus and bursa. The kinetics of DNA fragmentation in
both erythroid and lymphoid cells implied genotoxic damage by 1,2-dibromoethane to the stem
cells of the definitive elements and transmission of this damage through the successive cell generations.
Bjorge et al. (1996) paper tested testicular cells, prepared from human organ transplant
donors and from Wistar rats, for DNA damage caused by 15 known reproductive toxicants. Four
chemicals induced significant levels of single-stranded DNA breaks in testicular cells from both
species: styrene oxide (> or = 100 microM, rat and human), 1,2-dibromoethane (> or = 100
microM, rat; 1000 microM human), thiram (> or = 30 microM, rat; > or = 100 microM, human),
and chlordecone (300 microM, rat; > or = 300 microM, human).
48
-------
1,2-Dibromoethane was positive when tested for gene mutations in two human
lymphoblastoid cell lines, AHH-1 and TK6 (Crespi et al., 1985). For the AHH-1 line, gene
mutations were measured at the hypoxanthine guanine phosphoribosyl transferase locus, while,
for the TK6 line, gene mutations were measured at the thymidine kinase locus. 1,2-
Dibromoethane-induced sister chromatid exchanges were detected in human peripheral
lymphocyte cultures (Tucker et al., 1984). 1,2-Dibromoethane induced both dose- and time-
dependent increases in micronuclei in both mononucleated and binucleated human peripheral
lymphocytes (Channarayappa et al., 1992).
1,2-Dibromoethane exposed papaya workers (n = 60) described by the Ratcliffe et al.
study (1987) were assessed for genotoxic damage by monitoring sister chromatid exchanges and
chromosomal aberrations (Steenland et al., 1986). As previously described (section 4.1.2.),
workers from a nearby sugar plant served as controls (n = 42). Exposure to 1,2-dibromoethane
was not associated with sister chromatid exchanges. There was a statistically significant increase
in chromosomal exchanges, but this was the least frequent chromosomal aberration observed in
this study. The study authors considered it possible that this finding was due to chance, given that
multiple comparisons were made and that no significance was found when combining both types
of chromosomal aberrations (exchanges and deletions). Therefore, the study authors concluded
that exposure to low concentrations of 1,2-dibromoethane was not associated with chromosomal
aberrations or sister chromatid exchanges.
Steenland et al. (1985) performed a cytogenetic examination of forestry workers and
controls described in the Schrader et al. (1988) study (section 4.1.2.). Sister chromatid exchanges
and chromosomal aberrations were assessed in peripheral lymphocytes. There was no significant
increase in either sister chromatid exchanges or chromosomal aberrations in 1,2-dibromoethane-
exposed forestry workers.
4.4.4. Acute Toxicity
Rowe et al. (1952) performed a comprehensive acute oral and inhalation toxicity study in
the rat, mouse, guinea pig, chicks, and rabbits. The acute oral LD50 s (mg/kg) were 146 (male
rats), 117 (female rats), 420 (female mice), 55 (female rabbits), 79 (male and female chicks), and
110 (male and female guinea pigs). For the acute inhalation study, rats were exposed to
49
-------
100-10,000 ppm 1,2-dibromoethane for up to 16 hours while guinea pigs were exposed to 200 or
400 ppm for up to 7 hours. The 9-hour LC50 for rats was approximately 200 ppm. All guinea
pigs exposed to 200 or 400 ppm for 2-7 hours died.
Centrilobular necrosis and sinusoidal dilations in the liver were observed in adult male
albino rats treated with 110 mg/kg 1,2-dibromoethane by oral intubation (Broda et al., 1976). The
centrilobular necrotic changes were reported by the authors to be similar to those caused by
carbon tetrachloride.
Storer and Conolly (1983) conducted a comparative genotoxicity and acute hepatoxicity
study of 1,2-dibromoethane in mice. Non-necrogenic doses of 1,2-dibromoethane (0.25 or 0.5
mmol/kg) were administered to male B6C3F1 mice by i.p. injection. DNA damage was then
assessed by an alkaline DNA unwinding assay, which assessed single strand breaks or alkali-
labile sites in hepatic DNA. The high dose caused a significant decrease in percent double-
stranded DNA recovered. Liver and kidney damage were assessed following single i.p. injections
of 0, 0.5, 0.75, 1.0, or 1.5 mmol/kg 1,2-dibromoethane. 1,2-Dibromoethane was observed to be
hepatotoxic as measured by a dose-dependent increase in liver weight, serum IDH, and serum
AAT. Renal toxicity was also observed as a dose-dependent increase in kidney weight and BUN.
Four of five animals died following i.p injection of 1.5 mmol/kg 1,2-dibromoethane.
4.5. SYNTHESIS AND EVALUATION OF MAJOR NONCANCER EFFECTS AND
MODE OF ACTION-ORAL AND INHALATION
The epidemiological studies pertaining to subchronic and chronic effects focused
primarily on reproductive and cytogenetic endpoints. The human data suggest that 1,2-
dibromoethane is a male reproductive toxin. Decreased sperm count, ejaculate volume, and
motility, and abnormal sperm morphology, have been reported in workers following long-term
inhalation exposure to 1,2-dibromoethane (Ratcliffe et al., 1987). Workers exposed to 1,2-
dibromoethane for 6 weeks displayed decreased sperm velocity and volume (Schrader et al.,
1988). Also, male workers in a plant manufacturing 1,2-dibromoethane exhibited a significant
decrease in fertility (Wong, 1979). However, these studies have at least one limitation (e.g.,
inadequate exposure data, potential exposure to other reproductive toxins, moderate-to-extensive
dermal exposure potential, or other confounding factors) that renders any interpretation,
50
-------
particularly a quantitative assessment, of the evidence regarding the potential of inhaled 1,2-
dibromoethane to induce reproductive effects in humans inconclusive.
The evidence that inhaled 1,2-dibromoethane is associated with reproductive and
developmental effects in laboratory animals is incontrovertible. Reproductive and developmental
effects have been reported in rats and mice following inhalation exposure (Short et al., 1978,
1979). Effects in male rats included decreased testicular weight, decreased serum testosterone
levels, testicular atrophy, and impairment of reproductive performance. Testicular atrophy has
also been observed in male mice. Reported developmental effects in rats and mice consisted of
decreased fetal body weight, increased resorptions, decreased fetal survival, and/or skeletal
anomalies; however, these reproductive and developmental effects occurred at doses associated
with significant toxicity and/or mortality in parental/maternal animals. There are also indications
that there may be species and strain differences in response among pregnant animals. Oral
administration of 1,2-dibromoethane also has been shown to adversely affect male reproductive
endpoints. When 1,2-dibromoethane was administered orally to bulls at doses that did not affect
the growth or health of the animals, it was shown to adversely affect various sperm parameters.
Adverse effects included altered sperm morphology, decreased motility, and depleted sperm from
seminiferous tubules. In addition, the spermicidal effect of 1,2-dibromoethane occurs during
spermatogenesis, indicating that the effect is not direct. Overall, the data indicate that 1,2-
dibromoethane is a male reproductive toxin in animals.
Cytogenetic studies in humans are quite limited. Steenland et al. (1985, 1986) found no
evidence of genotoxicity following inhalation exposure to 1,2-dibromoethane. However,
concentrations to which workers were exposed were quite low and may have been below levels
that would induce DNA damage in humans. Genetic mutations have been observed in human
peripheral lymphocyte cultures (Channarayappa et al., 1992; Tucker et al., 1984), and DNA
damage has been noted in rats and mice (Arfellini et al., 1984; Kitchin and Brown, 1986, 1987).
Animal studies have demonstrated noncancer effects in rats and mice after subchronic-
and chronic-duration inhalation or oral exposure to 1,2-dibromoethane. Early mortality,
depression of body weight gain, and nonneoplastic lesions of the respiratory system, liver, kidney,
testis, eye, and adrenal cortex in rats and mice were reported in a chronic inhalation study (NTP,
1982). Proliferative lesions of the nasal epithelium in mice have also been reported after chronic
inhalation exposure to 1,2-dibromoethane (Stinson et al., 1981). It appears the respiratory
51
-------
system, particularly the nasal epithelium, is the target tissue following inhalation exposure in both
species. Excessive mortality, weight gain depression, testicular atrophy, forestomach lesions, and
liver and adrenocortical degeneration were reported in rats after long-term oral exposure to 1,2-
dibromoethane (NCI, 1978). In mice, long-term oral exposure to 1,2-dibromoethane was
associated with body weight depression, high mortality, and testicular atrophy. The systemic
effects associated with long-term oral exposure are generally consistent with those resulting from
inhalation dosing. The results of the NCI study (1978) suggest that the forestomach is the target
organ following oral exposure in rats and mice.
The results of a subchronic inhalation study in rats and mice revealed weight gain
depression, swelling of adrenocortical cells, decreases in thyroid follicle size, and formation of
megalocytic cells of the lining of bronchioles in rats and mice (NTP, 1982). In addition, Nitschke
et al. (1981) reported elevated relative liver and kidney weights, focal epithelial hyperplasia of the
nares, and diffuse respiratory hyperplasia. Similar respiratory effects were reported by Reznik et
al. (1980), and Rowe et al. (1952) reported adverse liver and kidney effects in rats, guinea pigs,
and monkeys. The subchronic toxicity data are in general consistent with reported chronic
effects.
Some of the effects of 1,2-dibromoethane, particularly effects in the liver and nasal tract,
are clearly related to its cytotoxicity. The mechanism of 1,2-dibromoethane-mediated
cytotoxicity has been studied in isolated rat hepatocytes (Khan et al., 1993). It was demonstrated
that microsomal cytochrome P-450-dependent oxidative metabolism of 1,2-dibromoethane
produces the metabolite 2-bromoacetaldehyde. The results suggest that the cytotoxic mechanisms
for 1,2-dibromoethane may possibly be attributed to lipid peroxidation and/or protein binding
induced by 2-bromoacetaldehyde. In addition, the study authors considered that the conjugation
of 1,2-dibromoethane with GSH may also contribute to cytotoxicity. Botti et al. (1982, 1986,
1989a, 1989b) and Masini et al. (1986) provided evidence that 1,2-dibromoethane-induced
depletion of hepatic mitochondrial GSH correlated with hepatotoxicity and perturbations in
mitochondrial Ca2+ homeostasis.
The results of in vitro and in vivo experiments suggest that the renal toxicity of 1,2-
dibromoethane may be due to its biotransformation by GSH conjugation followed by further
conversion in the kidney to highly reactive metabolites (Novotna et al., 1994). Repeated
administration of 1,2-dibromoethane to rats has been shown to enhance the content of GSH in the
liver and kidney (Mann and Darby, 1985). It has been suggested that lipid peroxidation may play
a role in the 1,2-dibromoethane-induced pathogenesis of liver cell necrosis (Albano et al., 1984).
52
-------
4.6. WEIGHT-OF-EVIDENCE EVALUATION AND CANCER CHARACTERIZATION
-SYNTHESIS OF HUMAN, ANIMAL, AND OTHER SUPPORTING EVIDENCE,
CONCLUSIONS ABOUT HUMAN CARCINOGENICITY, AND LIKELY MODE OF
ACTION
4.6.1. Human
In a cancer study of 161 workers at two 1,2-dibromoethane manufacturing plants, Ott et
al. (1980) reported that total cancer deaths of workers exposed to 1,2-dibromoethane did not
exceed those expected based on national rates. There was an increase in deaths due to malignant
neoplasms at one plant while the population at the second plant actually had fewer deaths due to
malignant neoplasms than expected. Similarly, Sweeney et al. (1986) studied the cause-specific
mortality of 156 male workers in a chemical plant that manufactured tetraethyl lead. The findings
of both studies are inconclusive due to the small sample size, lack of control group, poorly
characterized exposure assessment, and exposure to other potential or known carcinogens.
Therefore, human data are inconclusive regarding the potential carcinogenicity of 1,2-
dibromoethane to humans.
4.6.2. Animal
The results of animal oral and inhalation bioassays have demonstrated that 1,2-
dibromoethane is a carcinogen in rats and mice of both sexes at multiple sites. Long-term oral
administration of 1,2-dibromoethane was associated with forestomach carcinomas,
hemangiosarcomas, and lung adenomas or carcinomas in rats and mice as well as hepatocellular
and adrenocortical carcinomas (female rats) and thyroid follicular cell adenomas (male rats)
(NCI, 1978; Van Duuren et al., 1985). The majority of the cancers associated with oral dosing
were forestomach carcinomas. The relevance of forestomach effects to humans has been
questioned, particularly for nongenotoxic chemicals whose mode of action is believed to involve
irritation and cell proliferation from long-term exposure (Poet et al., 2003). It is true that the
forestomach is not present in humans and contains features, such as minimal vascularization and
stratified squamous cells, that result in a longer residence time of food-borne agents than is
received by comparable human organs such as the oesophagus and the glandular stomach (Grice,
53
-------
1988; Poet et al., 2003). In this case, however, effects in this organ are believed to be of potential
relevance to humans because 1,2-dibromoethane and other genotoxic chemicals do not appear to
require precursor events (e.g., irritation) associated with long residence time to induce these kinds
of tumors. 1,2-Dibromoethane does not appear to cause significant irritation in the forestomach
and was reported to induce forestomach tumors after just 168 days of exposure (NCI, 1978).
Long-term inhalation exposure of rats and mice to 1,2-dibromoethane resulted in nasal
cavity carcinomas and adenocarcinomas, alveolar/bronchiolar carcinomas, splenic
hemangiosarcomas, mammary gland adenocarcinomas, subcutaneous fibrosarcomas, and tunica
vaginalis mesotheliomas in rats and mice (NTP, 1982; Stinson et al., 1981; Wong et al., 1982).
The majority of the cancers were in the lungs of mice and in the nasal cavities of rats. The NTP
(1982) inhalation study was well designed, using an adequate number of animals of both sexes,
but was limited because of excessive mortality in the high-dose groups of both species, moderate
mortality in low-dose female mice, and excessive mortality in male mice not related to 1,2-
dibromoethane exposure. The Wong et al. (1982) study, which reported that disulfiram enhanced
the toxicity and carcinogenicity of 1,2-dibromoethane, is limited because there was only one
exposure dose.
1,2-Dibromoethane has been reported to be a direct acting mutagen in S. typhimurium
assays (Barber et al., 1981). 1,2-Dibromoethane has also been shown to induce point mutations
in S. typhimurium strains TA 1535 and TA 100, S. coelicolor, and A nidulans (Carere and
Morpungo, 1981). 1,2-Dibromoethane-induced chromosomal aberrations and sister chromatid
exchanges have been demonstrated in Chinese hamster ovary cells (Ballering et al., 1998; Brimer
et al., 1982; Graves et al., 1996; Ivett et al., 1989; Tan and Hsie, 1981;). 1,2-Dibromoethane has
also been shown to bind to DNA in vivo and in vitro (Arfellini et al., 1984; Kim et al., 1990; Koga
et al., 1986), and DNA damage has been reported in rats and mice following oral and i.p.
administration (Kitchin and Brown, 1986, 1987; Sasaki et al., 1998). In/), melanogaster, 1,2-
dibromoethane induced sex-linked recessive lethal mutations in spermatozoa and mutations in
embryonic spermatogonia (Kale and Baum, 1979, 1983). 1,2-Dibromoethane has also been
shown to produce mutations in human cells lines AHH-1 and TK-6 (Crespi et al., 1985), and
sister-chromatid exchanges and increases in micronuclei have been demonstrated in human
peripheral lymphocyte cultures (Channarayappa et al., 1992; Tucker et al., 1984). The above
studies indicate that 1,2-dibromoethane is genotoxic in a variety of test systems.
54
-------
4.6.3. Mode of Action
The genotoxicity of 1,2-dibromoethane is thought to be related to its conjugation with
GSH as catalyzed by glutathione S-transferase (Cmarik et al., 1990; Inskeep and Guengerich,
1984; Sundheimer et al., 1982; Van Bladeren et al., 1980, 1982). The conjugation results in the
formation of an episulfonium ion that can react with DNA (Peterson et al., 1988). As mentioned
previously, the major DNA adduct formed is S-[2-(N7-guanyl)ethyl]glutathione. Findings from a
forward mutation assay utilizing the bacteriophage Ml3 lacZ gene and mutation spectra have
established the importance of this adduct in 1,2-dibromoethane-mediated mutagenicity (Cmarik et
al., 1992).
The results of a study designed to examine the ability of purified rat and human
glutathione S-transferases to conjugate 1,2-dibromoethane with glutathione revealed that the
metabolism of 1,2-dibromoethane by glutathione-S-transferase and the genotoxic effects of 1,2-
dibromoethane are similar for rats and humans (Cmarik et al., 1990). Additional evidence that
1,2-dibromoethane is genotoxic via modification at ring nitrogens in DNA, primarily at the N7-
guanine site, was obtained from the mutation spectra of 1,2-dibromoethane in excision repair-
proficient and repair-deficient strains ofD. melanogaster (Ballering et al., 1994).
Working et al. (1986) also provided some evidence that the conjugation of 1,2-
dibromoethane with GSH and its subsequent metabolism may be involved in its genotoxic
properties. The ability of 1,2-dibromoethane to cause DNA damage was associated with
unscheduled DNA synthesis (UDS) in rat hepatocytes and spermatocytes exposed both in vitro
and in vivo. Inhibition of cytochrome P-450-mediated oxidation in vitro did not affect 1,2-
dibromoethane-induced UDS in either cell type; however, depletion of cellular GSH inhibited the
induction of UDS in both cell types. Inhibition of hepatic mixed-function oxidases in vivo was
associated with positive UDS response to 1,2-dibromoethane in spermatocytes, but there was no
effect on 1,2-dibromoethane-induced UDS in hepatocytes.
1,2-Dibromoethane has been demonstrated to act as an initiator of cell transformation in
the two-stage BALB/c3T3 cell transformation test (Colacci et al., 1995, 1996). The cell
transformation test is regarded as a model system for carcinogenesis in vivo. Utilizing y-
glutamyl-transpeptidase-positive foci as an early histochemical marker for hepatocarcinogenesis,
1,2-dibromoethane has been shown to possess promoter activity in the rat liver (Milks et al.,
1982). 1,2-Dibromoethane has been demonstrated to induce a mitogenic response in the rat liver
55
-------
(Ledda-Columbano et al., 1987a; Nachtomi and Farber, 1978; Nachtomi and Sarma, 1977). A
single oral administration of 1,2-dibromoethane to male rats was reported to induce cell
proliferation in the kidney, as monitored by increased thymidine incorporation into DNA and by
mitotic index (Ledda-Columbano et al., 1987b). It has been suggested that minor perturbations in
Ca2+ levels might play a role in triggering cell proliferation, while a more severe interference with
the homeostatic control of Ca2+ may lead to cell death (Ledda-Columbano et al., 1987b;
Nachtomi and Farber, 1978).
4.6.4. Weight-of-Evidence Characterization
Under the Draft Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1999), 1,2-
dibromoethane is considered "likely to be carcinogenic to humans" based on strong evidence of
carcinogenicity in animals and inconclusive evidence of carcinogenicity in an exposed human
population. This weight-of-evidence carcinogenicity characterization replaces the previous
classification of "B2; probable human carcinogen," entered on IRIS on September 7, 1988. The
new classification and slope factor estimates are based on a review of newer data and a reanalysis
of the data used in the earlier assessment. Based on the consistent findings of several studies
reporting increased incidences of a variety of tumors in rats and mice of both sexes by different
routes of administration at both the site of application and at distant sites, it can be concluded that
there is strong evidence of the carcinogenicity of 1,2-dibromoethane in animals. The available
evidence further supports a conclusion that 1,2-dibromoethane is a genotoxic carcinogen based on
evidence from a variety of in vitro and in vivo test systems.
4.7. SUSCEPTIBLE POPULATIONS
As has been described, GSTs are believed to play an important role in the mode of action
for 1,2-dibromoethane carcinogenicity. Human GSTs comprise several subfamilies of
isoenzymes: principally GSTM, GSTP, and GSTT. A polymorphism in the GSH conjugation of
1,2-dibromoethane by GSTT was demonstrated in the cytosol derived from human erythrocytes
(Guengerich et al., 1995; Ploemen et al., 1995). Erythrocyte cytosols from 2 out of 12 subjects
were unable to catalyze the conjugation of 1,2-dibromoethane with GSH due to a mutation of
GSTT. The relevance of this genetic polymorphism to interindividual differences in response to
1,2-dibromoethane and in response to other carcinogens is not clear. However, deletions in
56
-------
subunit 1 of the GSTT gene (GSTT1) are known to produce null genotypes that lead to absence of
activity of these enzymes (Seow et al., 2002). Further, the expression of the rat theta class
glutathione-S-transferase 5-5 (which is structurally similar to the human theta class GSTT) in S.
typhimurium TA1535 increased the mutagenicity of 1,2-dibromoethane (Thier et al., 1996).
Moreover, the mutagenicity of 1,2-dibromoethane was enhanced (compared to controls) in S.
typhimurium TA1535 cells expressing the human theta ortholog GSTT1-1 (the homodimer of
GSTT1) (Thier et al., 1996). GSTT1-1 polymorphism is apparently responsible for the bimodal
distribution of sensitivity to sister chromatid exchange induction observed after in vitro exposure
to butadiene diepoxide and other chemicals (Thier et al., 1996). While this polymorphism may
reduce human susceptibility to 1,2-dibromoethane induced cancers, it should be noted that GST
polymorphisms resulting in null or low activity of this genotype are generally thought to increase
overall cancer risk (Seow et al., 2002).
DeLeve (1997) addressed the issue of whether variations in endogenous GSH in human
cells could modify the genotoxicity of 1,2-dibromoethane. The incidence of sister chromatid
exchanges in normal fibroblasts and in fibroblasts obtained from two human individuals with
greatly reduced intracellular GSH levels due to hereditary generalized GSH synthetase
deficiency, an inborn error of GSH metabolism, was studied. The induction of sister chromatid
exchanges was significantly lower in the fibroblasts with GSH synthetase deficiency compared to
control cells. DeLeve (1997) concluded that low endogenous GSH levels may protect against
1,2-dibromoethane-induced genotoxicity in human fibroblasts. In addition, DeLeve (1997) noted
that little is known about the range of normal intracellular GSH in the population.
As mentioned previously, 2-bromoacetaldehyde is formed during the oxidative
metabolism of 1,2-dibromoethane and undergoes further metabolism or conjugation with GSH
(Wong et al., 1982). There may be considerable interhuman variability in their ability to catalyze
the oxidation of 1,2-dibromoethane to 2-bromoacetaldehyde. Microsomes from 21 different
human livers were able to catalyze the oxidation of 1,2-dibromoethane to 2-bromoacetaldehyde
with activities that ranged from 22.2 to!027.6 pmol/min-mg of protein (Wormhoudt et al.,
1996b). It has also been suggested that since disulfiram is used to treat alcoholics, these
individuals may be at a greater risk to 1,2-dibromoethane toxicity (Wong et al., 1982).
Disulfiram inhibits aldehyde dehydrogenase and may alter the metabolism of 1,2-dibromoethane
by preventing the further metabolism of 2-bromoacetaldehyde. Disulfiram has been reported to
enhance the carcinogenicity of 1,2-dibromoethane (Elliott and Ashby, 1980).
57
-------
4.7.1. Possible Childhood Susceptibility
There are no human studies indicating that children are more susceptible to the toxic
effects of 1,2-dibromoethane. However, there is evidence in mice that fetal epithelia can bind
14C-1,2-dibromoethane nonvolatile metabolites after i.v. injection to pregnant animals in different
stages of gestation (Kowalski et al., 1986). High-level binding was observed in the oral
epithelium, nasal mucosa, and forestomach. These results suggest that fetuses are likely to be
exposed to 1,2-dibromoethane from maternal circulation. The embryotoxic potential of 1,2-
dibromoethane to humans has been suggested by the results of an in vitro study with cultured rat
embryos in which it was shown that bioactivation of 1,2-dibromoethane by GST induced
manifestations of embryotoxicity (Mitra et al., 1992).
It has been concluded that children's respiratory vulnerability is in part due to the fact that
they have narrower airways than those of adults, and thus irritation that would produce only a
slight response in an adult can result in potentially significant obstruction in the airways of a
young child.1 As such, the nasal inflamation effects of 1,2-dibromoethane may have a more
significant health impact for infants and small children.
4.7.2 Possible Gender Differences
There are no human or animal data that suggest that gender differences in toxicity or
carcinogenicity might occur as a result of exposure to 1,2-dibromoethane.
1 Ambient Air Pollution: Respiratory Hazards to Children statement by the American Academy of
Pediatrics, online at http://www.aap.org/policY/04408.html - also the topic of the journal Pediatrics, Volume 92, No.
3, 1993 (which could not be retrieved prior to submitting these comments).
58
-------
5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE
An RfD can be derived based upon results of a rat chronic oral exposure study reported by
NCI (1978).2 In this study 50 Osborne-Mendel rats/sex/group were administered 1,2-
dibromoethane in corn oil, by gastric intubation. The initial doses utilized for male and female
rats were 40 and 80 mg/kg-day. High treatment-related mortality (18/50 males and 20/50
females) caused a discontinuation in the intubation of the high-dose group after treatment in week
16. Intubation of this group was suspended for 13 weeks and then restarted at week 30. At this
time the surviving rats received the low-dose regimen. All surviving male and female rats in both
dosage groups were sacrificed at weeks 49 and 61, respectively. The authors calculated time-
weighted average low- and high-doses of 38 and 41 mg/kg-day for male rats, and 37 and 39
mg/kg-day for female rats.
Peliosis was chosen as one of three co-critical endpoints for use in RfD derivation.
Among male rats, liver peliosis was observed in 0/40 control, 10/50 low-dose, and 9/50 high-dose
groups.3 Among female rats, peliosis was observed in 0/40 control, 4/47 low-dose, and 2/48 high-
dose groups. Peliosis is marked by engorgement of the liver with blood due to blockage of the
lumen of the sinus or destruction of the epithelial wall of the sinusoid. Peliosis is not considered
to be a precursor to liver cancer. Although hepatocellular carcinomas were detected in 5/48
female rats, the increase was not statistically significant. No increases in hepatocellular tumors
were detected in males. Squamous cell tumors found in the liver were the result of metastases
from forestomach tumors.
Another co-critical endpoint in this study is the induction of testicular atrophy. The
incidence of atrophy was 0/20 in vehicle controls, 14/49 in the low-dose group, and 18/50 in the
2NCI also studied mice, but high mortality in both dose groups of male and female mice precludes the use
of these bioassays for derivation of an RfD.
3Mortality adjusted incidence used for the BMD analysis of this data was 4/42 and 9/25 in the low and high
dose groups, respectively. (See Table B-l and output files in Appendix B.)
59
-------
high dose group.4 Testicular atrophy was also noted in 11/20 untreated control rats sacrificed at
107 weeks. Testicular atrophy in the latter group are likely to be age related, since they were not
seen in 61 week controls and because the duration was much longer than for exposed rats. The 61
week vehicle controls are therefore considered to be the appropriate control group
Although tumors of the tunica vaginalis, the serous covering of the testes, were reported in
the NCI (1978) study, testicular atrophy is not considered a precursor to this effect. The tunica
albuginea, a dense, fibrous membrane, lies between the testis and the tunica vaginalis. Thus, the
tunica vaginalis is neither in direct contact with the testes nor is it composed of structurally or
functionally related tissue. The reported testicular effects can, therefore, be considered a
noncarcinogenic endpoint separated from nearby tunica vaginalis tumors.
Testicular effects were also reported in bulls administered oral doses of 2-4 mg/kg 1,2-
dibromoethane (Amir and Ben-David, 1973; Amir and Volcani, 1965, 1967). Although doses
were lower, these studies were not selected for RfD development because of the small numbers of
animals and the use of one exposure level and because they are ruminants with a significantly
different physiology. Moreover, an allometric adjustment would result in a dose quite similar to
those used in the NCI (1978) study.5 These bull studies, however, provide supporting evidence
for testicular effects of low-dose 1,2-dibromoethane.
A third co-critical endpoint, was adrenal cortical degeneration (0/40 controls, 13/48 low-
dose, and 9/47 high-dose) in the male Osborne-Mendel rats of the NCI (1978) study.6 A similar
effect was observed in the female rats of this study (1/40 controls, 3/44 low-dose, and 8/45 high-
dose animals) and female F344 rats of the NTP (1982) chronic inhalation study. NCI (1978)
reported an increased incidence of adrenal tumors in female Osborne-Mendel rats following oral
exposure, and Wong et al. (1982) identified an increased incidence of adrenal tumors in both male
"Mortality adjusted rates used for the BMD analysis of this data were 14/43 and 18/36 in the low and high
dose groups, respectively. (See Table B-l and output files in Appendix B.)
5The allometric adjustment refers to the scaling of doses between species according to body mass raised to
the 3/4 power which the Agency has endorsed for carcinogens. The presumption of this adjustment is that equal
doses in these units (i.e., mg/kg^/4/day) when administered daily over a lifetime, will result in equal risk across
mammalian species (USEPA, 2002a). The bulls were assumed to weigh 1000 kg (Amir, 1975). After allometric
adjustment, the doses given to the NCI (1978) male rats were less than 2 times, and therefore similar to, the doses
given to the Amir and Volcani (1965; 1967) bulls (11 vs 21 mg/kg^/day).
"Mortality adjusted rates used for the BMD analysis of this data were 13/48 and 9/26 in the low and high
dose groups, respectively. (See Table B-l and output files in Appendix B.)
60
-------
and female Sprague-Dawley rats following inhalation exposure. However, adrenal tumors were
not observed in male Osborne-Mendel rats of the NCI (1978) oral study and were not observed in
the F344 rats of either sex in the NTP (1982) inhalation study. Thus, a clear association between
this degenerative effect and adrenal tumors cannot be established, and use of this endpoint in
support of a noncancer RfD is deemed appropriate.
Table 5-1. Oral subchronic and chronic studies in laboratory animals
Reference
NCI, 1978
Amir and
Volcani,
1967
Amir and
Volcani,
1965
Species (strain)
Rat
(Osborne-Mendel)
Bull
(Calves)
Bull
(Calves)
Sex
M
F
M
M
Animals/
dose
50
50
3
4
Exposure NOAEL
Regimen (mg/kg-day)
49 weeks,
5d/wk
61 weeks,
5d/wk
17.5-22.5b
Months
14-16b
Months
LOAEL
(mg/kg-day)
38a
37a
2C
2C
a Critical effects were peliosis and adrenal cortical degeneration in males and females and testicular atrophy in males.
b Animals were dosed orally via milk (1-3 mo), feed (4-9 mo) and via gelatin capsules (> 1 yr).
0 Critical effects were adverse alterations in various sperm parameters and testicular histology.
5.1.1. Methods of Analysis
RfDs can be derived by either development of a benchmark dose or through determination
of a NOAEL or a LOAEL. TWA doses for male and female rats of the NCI (1978) study were
very similar, but a higher incidence and severity of effects were observed in male rats at the low
dose. Adjustment of the lower TWA low dose for intermittent exposure of 5 days/week (38
mg/kg-day x 5/7) results in a LOAEL of 27 mg/kg-day. A benchmark dose analysis was
performed for all three of the co-critical endpoints (Appendix B, Table B-l). This analysis was
done using both the initial, unadjusted doses of 40 and 80 mg/kg-day and TWA doses reported by
the authors. Because the peliosis and adrenal cortical degeneration effects occurred at a much
greater incidence towards the end of the study, the BMD assessments that used TWA doses may
be more appropriate for these endpoints. However, because the incidence of testicular atrophy
was similar at the beginning, middle, and end of the study (6/20 at 9-24 wks; 4/10 at 25-39 wks;
4/9 at 40-44 wks and 4/11 at 45-49 wks) the dose rate may be more critical to the occurrence of
this effect. However, there is not enough data to make a definitive determination in this regard.
61
-------
Depending on whether unadjusted or TWA doses were used, BMDL10 estimates for these three
endpoints ranged from approximately 7-10 mg/kg-day. Because of high mortality and unusual
dosing in the high dose group, there is a great deal of uncertainty associated with these BMD
results, and they are only provided here in support of the NOAEL/LOAEL approach. Adjustment
of the lower TWA low dose for intermittent exposure of 5 days/week (38 mg/kg-day x 5/7)
results in a LOAEL of 27 mg/kg-day.
Table 5-2 Application of uncertainty factors (UFs) for RfD calculation
POD and UF Factors
Point of Departure (POD)-based on testicular, liver, and adrenal
effects in male rats and adjusted for intermittent exposure.
UFH = Variation from average humans to sensitive humans
UFA = Uncertainty in extrapolating from rodents to humans
UFS = Uncertainty in extrapolating from subchronic to chronic effect
levels
UFL = Uncertainty in extrapolating from LOAELs to NOAELs
(NOAEL/LOAEL approach)
UFn = Uncertainty in database
UF(TolaI)
RfD (mg/kg-day)
NOAEL/
LOAEL
LOAEL = 27
(ms/ks-dav)
10
10
1
10
10
10,000a
9E-3
'The UF(Total) is reduced from 10,000 to recommended maximum uncertainty factor of 3000 for the purposes of
calculating the final RfD due to recognized overlap in UFs (U.S.EPA, 2002a, page 4-41; U.S.EPA, 1994; Dourson
and Stara, 1983).
5.1.2. RfD Derivation
The adjusted daily dose of 27 mg/kg can be considered a LOAEL for peliosis (20%
response), testicular atrophy (29% response), and adrenal cortical degeneration (27% response).
Because testicular atrophy occurs with high incidence in aged untreated rats, the lack of testicular
atrophy in the vehicle control group at 61 weeks suggests that 1,2-dibromoethane hastens the
onset of testicular atrophy. Uncertainty factors of 10 for interspecies variability, 10 for
intraspecies variability in sensitivity, and 10 for adjustment from a LOAEL to a NOAEL were
assigned, as there was no information available that suggested other values were appropriate.
A 10-fold uncertainty factor accounting for the extent and quality of the database was
62
-------
deemed necessary due primarily to the poor quality of the principal study, the lack of high quality
developmental and reproductive studies by the oral route of exposure, and limited studies in bulls
that suggest adverse effects on sperm at low doses. High mortality in the principal study (NCI,
1978) causes considerable uncertainty with respect to the exposures that the animals received and
with respect to the responses that might have been observed had the animals survived to term.
The lack of a multigeneration study is also of concern in light of the genotoxicity of 1,2-
dibromoethane, because any genetic damage to the germ cells of the Fl generation would not be
detected until the F2 generation. Developmental toxicity studies covering major organogenesis
(but not studies covering the entire period of gestation) are available in two species via the
inhalation route, and inhalation systemic toxicity studies that evaluated the respiratory tract are
available in two species. There is also some limited evidence for neurobehavioral
developmental effects caused by 1,2-dibromoethane as well as endocrine disruption (based on
effects on other endocrine organs as well as changes in hormone levels).
A subchronic to chronic uncertainty factor was not considered necessary because all
animals of the low dose group were continuously exposed for approximately one year. This
results in a overall uncertainty factor, UF(Total), of 10,000. In general, the individual uncertainty
factors that comprise the UF(Total) are expected to be conservative with respect to the behavior of
the average chemical (Dourson and Stara, 1983). For this reason, the Agency has recommended
the application of a maximum uncertainty factor of 3000 for the purposes of calculating RfDs and
RfCs (U.S.EPA, 2002a, pages 4-41; U.S.EPA, 1994b). Application of a total uncertainty factor of
3,000 to the LOAEL of 27 mg/kg-day yields an RfD of 9 E-3 mg/kg-day.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
An RfC can be derived based upon the NTP (1982) bioassay of 1,2-dibromoethane. In
this study the noncarcinogenic effects observed are hepatic necrosis (male and female rats),
testicular degeneration (male rats), retinal atrophy (female rats), adrenal cortical degeneration
(female rats), splenic hematopoiesis (female mice), and inflammation of the nasal cavity (female
mice).7 The critical liver endpoint differs somewhat from that which was used to develop the oral
7The NTP (1982) study of male mice was not considered as relevant for derivation of an RfC because of
high mortality in control and exposed groups due to complications from urinary tract infections that were not
exposure related.
63
-------
RfD (peliosis). They are, however, both measures of liver toxicity and are likely to be closely
related. Because liver cancer induction was not statistically significantly increased in either sex
following inhalation exposure to 1,2-dibromoethane and because necrosis is not considered a
necessary precursor to induction of liver cancer, necrosis is considered to be a suitable endpoint
for quantifying the noncancer effects of 1,2-dibromoethane.
Table 5-3. Inhalation subchronic and chronic studies in laboratory animals '
Reference
NTP, 1982
NTP, 1982
Stinson et al., 1981
Nitschkeetal., 1981
Rezniketal.,
1980C
Rezniketal.,
1980C
Species (strain)
Rat (F344)
Mouse (B6C3F1)
Mouse (B6C3F1)
Rat (F344)
Rat (F344)
Mouse (B6C3F1)
Sex
M
F
M
F
M
F
M
F
M
F
M
F
No./
dose
group
50
50
48
50
50
50
10
10
5
5
10
10
Exposure
duration
103 weeks
103 weeks
78 weeks b
106 weeks
103 weeks
103 weeks
13 weeks d
13 weeks e
13 weeks d
13 weeks e
13 weeks d
13 weeks e
NOAEL
(mg/m3)
~
~
-
~
-
~
23
23
23
23
115
115
LOAEL
(mg/m3)
76.8
76.8
76.8
76.8
76.8
76.8
76.8
76.8
115
115
576
576
a Results of Rowe et al. (1952) not included because of the many limitations of that investigation.
bMortality was high in exposure and control groups of male mice.
0 Study was designed principally to evaluate the role of 1,2-dibromoethane in promoting nasal lesions.
d Forty males per exposure group; serial sacrifices of 10 males per exposure group were conducted at 1, 6, and 13
weeks; remaining animals were sacrificed 88-89 days postexposure.
e Twenty females per exposure group; 10 females per exposure group were sacrificed at 13 weeks; remaining animals
were sacrificed 88-89 days postexposure.
64
-------
Table 5-4. Reproductive and developmental inhalation studies in laboratory
animals
Reference
Short etal., 1978
Short etal., 1978
Short et al.,
1979
Smith and Goldman,
1983
Species (strain)
Rat (Charles River
CD)
Mouse (CD-I)
Rat (Charles River
CD)
Rat (Long-Evans)
Sex
M&F
M&F
M
F
F
#/Dose
group
15-17
18-22
9-10
20
12
Exposure
Duration
10 days a
10 days a
10 weeks c
3 weeks °
Days 3 to 20 of
gestation e
NOAEL
(mg/m3)
154
~
300
300
3.3
LOAEL
(mg/m3)
292 b
154 b
684 d
614 d
51 f
a For 23 hr/day beginning on day 6 of gestation.
b Developmental effects plus some maternal toxicity.
0 For 7 hr/day, 5 days/week.
d Reproductive effects plus significant morbidity and mortality.
e For 4 hr/day, 3 days/week.
f Based on behavioral effects in offspring.
5.2.1. Methods of Analysis
Because the NTP study demonstrated adequate spacing of exposure levels with increasing
response levels with increasing exposure levels, the inhalation toxicity of 1,2-dibromoethane was
evaluated using benchmark dose (BMD) analysis. A summary of the dose-response data and
results and the actual EPA BMDS model runs are provided in Table B-2 of Appendix B. The
benchmark doses were estimated for both 10% and 5% extra risk. An extra risk of 10% has
generally been the default benchmark response (BMR) level for quantal data because it is at or
near the limit of sensitivity in most bioassays. The Agency's benchmark dose technical guidance
(U.S.EPA, 2000c) does indicate that a lower BMR can be used if a study has "greater than usual
sensitivity." However, this study cannot be said to have greater than usual sensitivity because it
involved only two dose groups and high mortality in all dose groups. In addition, responses in the
low dose group were in the range of or greater than 10% for virtually all endpoints under
consideration. BMD estimates for lower response levels are more dependent on model choice
(i.e., the range of BMD and BMDL estimates are more variant across models). Therefore, the
BMDL for an extra risk of 10% is used for the RfC derivation.
5.2.2. RfC Derivation
65
-------
To derive an RfC, the BMDL10 values presented in Table B-2 of Appendix B were
adjusted to equivalent continuous exposures, converted to human equivalent concentrations
(HECs), and then divided by uncertainty factors. The lowest of the resulting values was
determined to be the RfC.
Human equivalent exposures were estimated following the EPA RfC methodology
(U.S.EPA, 1994b). 1,2-Dibromoethane is relatively insoluble in water and demonstrates systemic
toxicity, although there was some respiratory tract toxicity that may or may not be considered
portal of entry effects. Therefore, 1,2-dibromoethane is considered a Category 2 gas. For
Category 2 gases, HEC values are calculated using methods for category 1 gases for portal-of-
entry effects and category 3 methods for systemic effects (U.S.EPA, 1994b). Thus, in Table 5-5,
the EPA RfC method for Category 3 gases was used to derive BMDL10 (HEC)s for the liver,
testicular, retinal, adrenal, and splenic effects, and the method for Category 1 gases was used to
derive BMDL10(HEC) for nasal effects. The NTP (1982) data for nasal inflammation in female
mice resulted in the lowest BMDL10(HEC) of 2.8 mg/m3. Figure 5-1 is a plot showing the
selected, BMDS log-probit model results for this endpoint. The model runs that resulted in the
BMDL10 estimates for nasal inflammation and other effects are summarized in Appendix B, Table
B-2.
An uncertainty factor of 3 was applied for interspecies pharmacodynamics as a
consequence of considering human equivalent dosimetry above and due to the lack of data
suggesting a more or less divergent response in humans. An uncertainty factor of 10 for
intraspecies variability in sensitivity results was applied, as well as a default value due to the lack
of data indicating a different degree of variability in humans.
An uncertainty factor for less than lifetime exposure is considered to be unnecessary
because the principal study was carried out for at least 88 weeks. A database uncertainty factor of
10 is applied. High mortality in the principal study (NTP, 1982) causes considerable uncertainty
with respect to the exposures that the animals received and with respect to the responses that
might have been observed had the animals survived to term. A one-generation inhalation
reproductive toxicity study is available but no multigeneration study. The lack of the
multigeneration study is of particular concern in light of the genotoxicity of 1,2-dibromoethane,
Table 5-5. HEC estimates from BMDLs derived from NTP (1982); Table B-2,
Appendix B
66
-------
NOAEL
LOAEL
BMDS model0
BMD10
BMDL10
BMDL10 (ADJ)d
BMDL10 (HEC)e
Male Rats
Hepatic
necrosis
(mg/m3)"
NAb
76.8
Probit
131.774
105.343
18.8
18.8
Testicular
degeneration
(mg/m3)
NA
76.8
LogLogistic
53.2579
35.0725
6.3
6.3
Female Rats
Hepatic
necrosis
(mg/m3)
NA
76.8
LogProbit
172.374
122.02
21.8
21.8
Adrenal
cortical
degeneration
(mg/m3)
NA
76.8
LogLogistic
124.823
66.495
12.0
12.0
Retinal
atrophy
(mg/m3)
NA
76.8
LogLog
125.806
105.41
18.8
18.8
Female Mice
Splenic
hematopoiesis
(mg/m3)
NA
76.8
LogLogistic
59.554
40.2456
7.2
7.2
Nasal
inflammation
(mg/m3)
NA
76.8
LogProbit
102.192
80.1088
14.3
2.8
a Assuming a temperature of 25°C, a barometric pressure of 760 mm Hg, and a molecular weight for 1,2-
dibromoethane of 187.88, 1 ppm = 7.68 mg/m3 (187.88/24.45).
bNA=not applicable.
°In accordance with EPA draft guidance (U. S.EPA, 2000d), the selected models were chosen on the basis of goodness
of fit criteria (AIC and chi-square residual values) and visual inspection.
Adjustment to continuous exposure involved multiplying the BMDL by 5 All d x 6 hr/24 hr.
eHEC values were calculated in accordance with U.S. EPA (1994b) RfC methods. For extrarespiratory effects, a
default adjustment factor of 1.0 was used for adjusting from ADJ to HEC values because 1,2-dibromoethane blood:air
partition coefficients are not known for the experimental species and humans. For the respiratory effect (nasal
inflammation), the HEC was calculated for an effect in the extrathoracic (ET) region. Minute volumemouse = 0.041
L/min, minute volumehumm = 13.8 L/min, surface area(ET)mouse = 3 cm3, surface area(ET)humm = 200 cm2. Regional gas
dose ratio(ET) = [minute volumemouse/surface area(ET)mouse]/[minute volumehumm/surface area(ET)hl]man] = 0.198.
BMDL(HEC) = BMDL(ADJ) x regional gas dose ratio(ET) = 14.3 mg/m3 x 0.198 = 2.8 mg/m3.
because any genetic damage to the germ cells of the Fl generation would not be detected until the
F2 generation. Furthermore, the absence of an evaluation of sperm in reproductive toxicity study
is of concern in light of the effects observed in humans and bulls. Developmental toxicity studies
covering major organogenesis (but not studies covering the entire period of gestation) are
available in two species via the inhalation route, and inhalation systemic toxicity studies that
evaluated the respiratory tract are also available in two species. There is also some limited
evidence for neurobehavioral developmental effects caused by 1,2-dibromoethane as well as
endocrine disruption (based on effects on other endocrine organs as well as changes in hormone
levels).
The composite uncertainty factor is 300 (10 for UFH, 3 for UFA, and 10 for UFD).
Application of this composite factor to the BMDL10(F£EC) of 2.8 mg/m3 yields an RfC of 9 E-3
mg/m3.
67
-------
Probit Model with 0.95 Confidence Level
0.6
0.5
"o
CD
0.3
c
o
"o n
co 0.
0.1
0
Probit
BMDL BMD
0 50 100 150 200 250 300
dose
15:2909/032003
Figure 5-1: Log-Probit Model of Nasal Inflammation in Female Mice (NTP (1982)
5.3 CANCER ASSESSMENT
There are no definitive reports of cancer in humans associated with exposure to 1,2-
dibromoethane; the available epidemiological studies have numerous limitations (section 4.1.3.)
and are inconclusive. Chronic bioassays in rats and mice by both oral and inhalation routes
provide evidence of tumors in multiple organ systems and also at direct points of contact (e.g.,
nasal cavity and lung tumors following inhalation and forestomach tumors following ingestion).
NCI (1978) provided evidence for the induction of forestomach squamous cell carcinoma
68
-------
and hemangiosarcoma in male and female rats, thyroid follicular cell adenomas in male rats, and
hepatocellular carcinomas and adrenocortical carcinomas in female rats following gavage
administration of 1,2-dibromoethane. In addition, there were increases in forestomach tumors and
lung adenomas in mice of both sexes. The effects in rats and mice are supported by carcinogenic
findings (forestomach tumors) in a drinking water study in mice (Van Duuren et al., 1985),
conducted at a single dose comparable to the high dose in the NCI mouse study. NTP (1982)
provided evidence of 1,2-dibromoethane-induced nasal cavity tumor and other benign and
malignant tumors in male and female Fischer 344 rats and in female B6C3FJ mice in a 2-year
inhalation cancer bioassay. Further, screening level evidence from A/J mouse lung tumor assays
supports tumor induction by the oral, inhalation, dermal, and intraperitoneal routes of exposure
(Adkins et al., 1986; Stoner et al., 1986; Van Duuren et al., 1979). In addition, 1,2-dibromoethane
has been reported to be a direct acting mutagen in S. typhimurium assays and has also been shown
to induce point mutations in S. typhimurium. 1,2-Dibromoethane has also been demonstrated to
induce chromosomal aberrations and sister chromatid exchanges in Chinese hamster ovary cells
and bind to DNA in vivo and in vitro.
This weight-of-evidence carcinogenicity characterization and quantitative estimate of
carcinogenicity from oral exposure replace the previous classification of "B2; probable human
carcinogen," and oral slope factor of 85 mg/kg/day entered on IRIS on September 7, 1988. The
new classification and slope factor estimate are based on a review of newer data and a reanalysis
of the data used in the earlier assessment. This is based on the consistent findings of several
studies reporting increased incidences of a variety of tumors in rats and mice of both sexes by
different routes of administration at both the site of application and at distant sites. It can be
concluded that there is strong evidence of the carcinogenicity of 1,2-dibromoethane in animals.
The available evidence further supports a conclusion that 1,2-dibromoethane is a genotoxic
carcinogen based on evidence from a variety of in vitro and in vivo test systems. Under the Draft
Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1999), 1,2-dibromoethane is considered
"likely to be carcinogenic to humans" based on strong evidence of carcinogenicity in animals and
inconclusive evidence of carcinogenicity in an exposed human population.
There is some indication that a glutathione-dependent metabolite may be implicated in 1,2-
dibromoethane's carcinogenic activity (as reported by Hissink et al., 2000), but the available
evidence does not make any quantitative estimates possible. The carcinogenicity of 1,2-
dibromoethane is evaluated according to the linear approach as described in U.S. EPA (1999),
because 1,2-dibromoethane has demonstrated genotoxicity and not enough in addition is known
69
-------
about the mode of action to support a nonlinear low-dose extrapolation (see section 4.6.3.).
5.3.1. Oral Carcinogenicity
5.3.1.1. Choice of Oral Study/Data with Rationale and Justification
The study by NCI (1978) was used for development of an oral slope factor. Although the
drinking water study by Van Duuren et al. (1985) used an adequate number of animals and
examined proper endpoints, the study is limited for development of an oral slope factor because
only one dose level was examined. The NCI (1978) study used an adequate number of test
animals and two dose levels plus untreated and vehicle control groups, and examined appropriate
toxicological endpoints.
5.3.1.2. Oral Dose-Response Data
In the NCI (1978) study, male and female Osborne-Mendel rats and B6C3FJ mice were
administered 1,2-dibromoethane by gavage 5 days/week. High mortality in both species in all
exposure groups necessitated reduced dosage levels, reduced frequency of dosing (see section
4.2.1..), and early termination of the studies. Time-weighted average low- and high-doses were 38
and 41 mg/kg/day for male rats, 37 and 39 mg/kg/day for female rats, and 62 and 107 mg/kg/day
for mice of both sexes. Male and female rats were sacrificed at weeks 49 and 61, respectively. All
surviving male mice and high-dose female mice were sacrificed at week 78, while low-dose
females were sacrificed at week 90.
Tumor incidences were elevated with increasing exposure level at several sites:
forestomach squamous cell carcinoma, hemangiosarcoma, thyroid follicular cell adenomas in male
and female rats, hepatocellular carcinomas in female rats, and forestomach squamous cell
carcinoma and lung adenoma in both sexes of mice. Benign and malignant tumors were combined
for sites where an eventual progression from the benign to the malignant form was plausible (for
example, lung alveolar/bronchiolar adenomas or carcinomas, and forestomach squamous cell
papillomas or carcinomas). These data are summarized in Table 5-6.
Because there was high mortality in both species that resulted in early termination of the
study, statistical procedures that can reflect the influence of intercurrent mortality on site-specific
70
-------
tumor incidence rates were used to evaluate the tumor incidence levels. The individual animal
data were not included in the bioassay report and were requested from NTP/NCI for this purpose
(see Appendix C for a listing of the tumor incidence data by time of death). Subsequent to the
completion of the NCI study of 1,2-dibromoethane, the NTP has adopted the Poly-3 procedure
(Bailer and Portier, 1988) for adjusting tumor incidence rates for intercurrent mortality. The
procedure is based on the observation that the cumulative incidence of tumors tends to increase
with time raised to the second through the fourth powers for a large proportion of cases. In the
Poly-3 procedure, for a study of T weeks duration, an animal that is removed from the study after t
weeks (t < T) without a specified type of tumor of interest is given a weight of (t/T)3. An animal
that survives until the terminal sacrifice at T weeks is assigned a weight of (T/T)3 = 1. An animal
that develops the specific type of tumor of interest obviously lived long enough to develop the
tumor and is also assigned a weight of 1. The Poly-3 tumor incidence, adjusted for intercurrent
mortality up to time T, is the number of animals in a dose group with the specified type of tumor
divided by the sum of the weights (the effective number of animals at risk). The tumor incidences
adjusted using this procedure are also provided in Table 5-6 with the results of applying the
Cochran-Armitage test for trend to the adjusted incidences.
Note that the low-dose female rats and both sexes of mice all had adjusted incidences of
100% for forestomach tumors, while the low-dose male rats had an adjusted incidence of 95%.
Considering only the administered dose rate and adjusted tumor incidence at the low dose, female
rats were slightly more sensitive than male rats to 1,2-dibromoethane exposure. In addition,
forestomach tumors appeared the earliest in the female rat study, at week 12 in both dose groups,
contrasted with week 15 in the high-dose male rats and week 31 in the low-dose male rats.
5.3.1.3. Oral Dose Adjustments and Extrapolation Methods
In preparation for carrying out dose-response analyses of the tumor data, adjustments to the
administered doses for approximating human equivalent, continuous exposure levels were
considered. Following EPA default procedures, the time-weighted daily average doses were
converted to human equivalent doses on the basis of (body weight)374 (U.S.EPA, 1992). This
adjustment is summarized in Table 5-7.
The dose levels as reported by NCI had been averaged over the period of observation in
terms of weeks, with time-weighted average low and high doses of 38 and 41 mg/kg/day for male
rats, and 37 and 39 mg/kg/day for female rats. Although some of the survival-adjusted incidence
71
-------
of lesions and tumors in the high-dose rats were no higher than in the low dose groups, there is
some evidence that there was a dose rate effect - that is, that the rate at which the 1,2-
dibromoethane was administered was more directly related to the effects seen than was the time-
weighted average. The tumor incidence by time curves shown in Figures C-l through C-l 1
(Appendix C) illustrates that, for many of the significant tumor sites, the incidence in the high-
dose group started earlier than in the corresponding low-dose group. One exception might be the
forestomach tumors in female rats (Figure C-4, Appendix C), for whom the incidence in the two
groups was very similar. However, since the incidence in the low-dose female rats was effectively
100% after adjusting for competing causes of mortality, the observation that the high dose female
rats did not reach 100% much sooner does not really inform a characterization of a dose rate
effect. In addition, the mice in the van Duuren et al. study (1985) received a daily oral exposure
that was slightly higher than the high dose in the NCI mouse study. Their response was slightly
less than the NCI mice, most likely attributable to drinking water administration rather than the
oral gavage in the NCI study - another example of dose rate differences. Most notably, 18 high
dose male rats died in week 15, three of which had forestomach tumors, while in the low-dose
group the first forestomach tumor was not observed until week 31. For this reason dose averaging
over the period of the study may not be the most representative measure of exposure.
Another adjustment typically made to extrapolate to continuous daily exposure, that is
lowering each dose by multiplying by (5 days)/(7 days)=0.71, was not incorporated prior to dose-
response modeling, also in recognition of the possible dose rate effect. However, the assumption
that a cumulative amount of exposure delivered over different time periods producing a similar
effect may be more likely to hold at very low exposures. Since this is a linear adjustment, it can be
incorporated after dose-response modeling as an alternative characterization to compare results.
Because the studies all ended before the usual 104 weeks, an adjustment to extrapolate to
lifetime human exposure was considered. EPA typically adjusts each dose level by a factor related
to the Poly-3 adjustment. That is, each dose is adjusted by multiplying by (T/104)3, where T is the
time of final sacrifice in a given dose group. When all groups have been terminated at the same
time, this adjustment is linear, and can be applied before or after modeling is completed. The rat
studies were terminated at approximately 50% of their intended length, which results in lowering
the male and female doses by factors of 10 and 5, respectively. It was decided to consider the
adjustment as an alternative characterization after the modeling is complete. Although the female
mouse groups were terminated at different times, at 90 (low-dose) and 78 (high-dose) weeks, it
was decided to apply a composite lifetime equivalent adjustment after modeling, to parallel the
72
-------
risk estimates derived from the rat study. Calculation of the lifetime exposure adjustments is
summarized in Table 5-7.
The results of the oral bioassays were characterized by fitting dose-response models to the
tumor results showing significant elevations with increasing exposure. Because there was high
mortality in both species that also resulted in early termination of each study, models which can
reflect the influence of intercurrent mortality on site-specific tumor incidence rates were
considered. EPA has generally used two approaches. Modest effects on survival can be addressed
by omitting the animals in each treatment group who died before the first occurrence of the tumors
being analyzed. In these bioassays, however, effects on survival were not modest. Consequently,
for the quantal model analyses, intercurrent mortality was addressed using the Poly-3 adjusted data
discussed in section 5.3.1.2. The poly-3 adjustment is an approximate adjustment that may not
characterize the extremes of tumor incidence (Bailer and Portier, 1988), however. When tumors
appear earlier with increasing exposure levels, the multistage-Weibull model is the preferred
model because it incorporates the time at which death-with-tumor occurred (e.g., U.S.EPA: 1988b
- 1,2-dibromoethane; 2001 - bromate; 2002b - 1,3-butadiene). However, the multistage-Weibull
model cannot accommodate variable dosing schedules as in this study. Consequently, the more
approximate approach using the Poly-3 adjusted data was the most suitable dose-response method
for these data.
Dose-response analyses were conducted from the individual animal data for sites
demonstrating an increased cancer incidence summarized in Table 5-6. Etiologically different
tumor types were not combined across sites prior to modeling because the numbers of animals at
risk differed by tumor site, depending on the patterns of intercurrent mortality.
EPA generally uses the multistage model with quantal cancer data, to estimate a 95% upper
confidence limit (UCL) on cancer risk (extra risk) for humans. The multistage model has the form
P(d) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)],
where P(d) represents the lifetime risk (probability) of cancer at dose (i.e., human equivalent
exposure concentration in this case) d, and parameters q; > 0, for i = 0, 1, ..., k. The model
parameters and 95% UCL were calculated using the multistage model in the EPA BMDS software.
Extra risk over the background tumor rate is defined as [P(d) - P(0)]/[l - P(0)]. Point estimates of
the dose coefficients (q;s), and consequently the extra risk function, at any dose d are calculated by
73
-------
maximizing the likelihood function with respect to the tumor incidence data. The incremental
lifetime unit cancer risk for humans (q^) is defined as the 95% UCL on the parameter ql3 which is
the linear dose coefficient. This 95% UCL represents a plausible upper bound on the true risk.
5.3.1.4. Oral Slope Factor
The results of applying quantal models to the Poly-3 adjusted tumor incidence data are
provided in Table 5-8. The data for forestomach tumors for female rats and both sets of mice
could not be adequately fit by the available models. For the remaining tumor data, a point of
departure (PoD) near the lower end of the observed data was selected for linear extrapolation to
low doses, consistent with the Draft Guidelines for Carcinogen Risk Assessment (U.S.EPA, 1999).
An oral slope factor for each of these tumor sites was calculated by dividing the BMR level by the
corresponding BMDL for these points of departure. A slope factor for each data set not fit by a
model (e.g., forestomach tumors) was approximated by dividing the response at the low dose -
100% in each case - by the low dose.
Under the assumption that dose rate is the most important dose metric, estimated human
equivalent oral slope factors ranged from 0.013 to 0.12 (mg/kg/day)"1, for less than lifetime
exposure. The highest slope factor corresponded to forestomach tumors in male rats. Adjustment
for lifetime exposure, using the default adjustments listed in Table 5-7, leads to a range of 0.065 to
1.0 (mg/kg/day)"1 for oral slope factors, as presented in the last column of Table 5-8.
The approximate slope factors for the other forestomach tumor (female rats; male and
female mice) sets were at least as high, which is notable because these were central tendency
estimates. No useful upper bound estimate was possible with the response already at 100%.
These estimates and the estimate from the male rats are only useful if it can be assumed that
carcinogenicity associated with 1,2-dibromoethane follows a linear relationship throughout the
range of doses below 93%-100% response.
There is some concern that the forestomach tumors may result primarily as a portal of entry
effect and may not have a dose-response pattern that extends linearly from the observed responses.
The hemangiosarcomas and other tumors demonstrate absorption of 1,2-dibromoethane, although
it is not clear whether absorption was impacted by adverse effects in the forestomach. In addition,
there is considerable uncertainty in extrapolating from the period of the experiment, 49 weeks for
the male rats, to full lifetime exposure. Given the multiplicity of tumor sites, however, basing the
74
-------
slope factor on one tumor site underestimates oral carcinogenic potential of 1,2-dibromoethane, all
else being equal.
In order to gain some understanding of the total risk from multiple tumor sites in male and
female rats, it was assumed that the more significant tumor types observed were mechanistically
independent - that is, that the occurrence of hemangiosarcomas, say, was not dependent upon
whether there were forestomach tumors. Accordingly, a statistically appropriate upper bound risk
was estimated using the following steps: (1) the central tendency, or maximum likelihood
estimates (MLE) of unit potency were summed across forestomach tumors, hemangiosarcomas,
and thyroid follicular cell adenomas for male rats and across hemangiosarcomas, hepatocellular
carcinomas, adrenocortical carcinomas in female rats; (2) an estimate of the 95% upper bound on
the summed unit risk was calculated by assuming a normal distribution for the individual risk
estimates, and deriving the variance of the risk estimate for each tumor site from its 95% upper
confidence limit (UCL) according to the formula
95% UCL = MLE + 1.645 • s.d.,
where 1.645 is the t-statistic corresponding to a one-sided 95% confidence interval and >120
degrees of freedom, and the standard deviation (s.d.) is the square root of the variance of the MLE.
The variances were summed across tumor sites to obtain the variance of the sum of the MLE. The
95% UCL on the sum of the individual MLEs was calculated from the variance of the sum using
the same formula.
Summing the cancer risks in this manner, after the dose-response for each tumor site has
been evaluated, is superior to EPA's previous practice of carrying out one dose-response analysis
of tumor-bearing animals. The primary reason is that the biological relevance of the multistage
model is maximized by allowing different multistage models to be fit to qualitatively different
tumor types that might not be expected to develop through exactly the same modes of action.
Time courses in the tumor types evaluated here did vary, for example. In this case, however, these
summed estimates are approximate because the risk estimates could not be taken from analogous
segments of the respective dose-response curves. For example, the risk for the male forestomach
tumors comes from the portion of the dose-response near 90% response, while the risks for the
other two sites are from a much lower portion (<20%) of their respective dose-response curves.
The resulting 95% UCL on the slope factor for the summed risk of forestomach tumors,
75
-------
hemangiosarcomas, and thyroid follicular cell adenomas for male rats was 0.13 (mg/kg/day)"1. The
slope factor for forestomach tumors was nearly an order of magnitude higher than for the other two
tumor sites, and the variability in the slope factor based on forestomach tumors alone was the
greatest of the three sites, so the other sites end up having very little effect on the upper bound of
the summed risks. Since risk values are rounded to one significant figure, the summed cancer
slope factor is equivalent to that for forestomach tumors alone.
For comparison, the sum of the three significant tumor sites for females rats, excluding the
forestomach tumors, has a 95% UCL of 0.044 (mg/kg/day)"1. If the summed risk estimate is
combined with the approximate slope factor estimated for female rat forestomach tumors of 0.1
(mg/kg/day)"1, this results in an estimate of 0.14 (mg/kg/day)"1. Although this risk estimate is
higher than that obtained from the male rats, the contribution of risk from the forestomach tumors
for female rats is relatively uncertain-there was no dose-response information between 0% and
100% response levels (and the modeling approach could not provide a confidence interval). The
lowest response level for the males was still relatively high, close to 90%, but a confidence limit
was estimable. The slope factor should be based on the sum of the male rat tumors slope factors
rather than the sum of the female rat slope factors.
At low doses, it is not known whether or not there would be a dose rate effect. Without
information to the contrary, the estimated cancer slope factor should be adjusted to daily exposure
by multiplying by (7 days)/(5 days). The recommended oral slope factor for approximately half a
lifetime of exposure is 0.13 (mg/kg/day)"1 x 7/5 = 0.18 (mg/kg/day)"1.
For application to lifetime exposures, the default lifetime exposure adjustment factor for male rats
of 0.1, described in Table 5-7, leads to a slope factor of 1.8 (mg/kg/day)"1. This slope factor should
not be used with exposures greater than approximately 0.5 mg/kg/day (i.e, the human equivalent
lifetime exposure level corresponding to 90% risk of forestomach tumors), since the observed
dose-response would not be expected to continue linearly above this estimated lifetime-equivalent
exposure level.
An oral slope factor of 85 (mg/kg/day)"1 was listed previously on IRIS (U.S. EPA, 1988b).
The earlier assessment used the multistage-Weibull model, which has a form similar to the
multistage model (Equation 5-1):
P(d,t) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)*(t - t0)z], (5-2)
76
-------
with the addition of the time t that a tumor was observed, raised to the z power. The model fit to
the NCI (1978) data fit z = 7.6 to the time that tumors were observed, which contrasts with the
power of 3 used in the current analysis, through the Poly-3 adjustment for deaths occurring before
the end of the study, and through the lifetime exposure adjustment for extrapolating from the end
of the study to full lifetime. These modeling differences resulted in a less extreme extrapolation to
lifetime exposure from the truncated bioassay that was available. Several quantitative
considerations and assumptions were also updated in the reanalysis, including the use of (body
weight)374 scaling rather than surface area scaling, (body weight)273, to estimate human equivalent
doses; estimation of risk close to the range of the observed data rather than extrapolated to the
exposure expected to be associated with 1 x 10"6 extra risk; and extrapolation to a human lifetime
of 70 rather than 76.2 years.
77
-------
Table 5-6. Observed and adjusted tumor incidences (and percentages) in
rats and mice exposed by oral gavage to 1,2-dibromoethane
Tumor site
Incidence
adjustment
Male rats
Forestomach papillomas
or tumors
Hemangiosarcoma
Thyroid follicular cell
adenoma or carcinoma
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
Female Rats
Forestomach papillomas
or tumors
Hemangiosarcoma
Hepatocellular carcinoma
or neoplastic nodule
Adrenocortical
carcinoma
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
Male mice
Forestomach papillomas
or tumors
Alveolar/bronchiolar
adenoma
Observed
Poly-3
Observed
Poly-3
Female mice
Forestomach papillomas
or tumors
Alveolar/bronchiolar
adenoma
Observed
Poly-3
Observed
Poly-3
Dose groups3
Tumor incidence1"
0 mg/kg/day
0/20
0/13
0/20
0/13
0/20
0/13
0%
0%
0%
0%
0%
0%
Low
45/50
45/50
11/50
11/42
5/50
5/39
dose
90.0%
90.0%
22.0%
26.2%
10.0%
12.8%
High
33/50
33/34
4/50
4/23
8/50
8/23
dose
66.0%
97.1%
8.0%
17.4%
16.0%
36.4%
Trend
test
p-valuec
0.001
0.12d
0.001
0/20
0/19
0/20
0/19
0/20
0/19
0/20
0/19
0%
0%
0%
0%
0%
0%
0%
0%
40/50
40/40
0/50
0/20
1/50
1/20
1/50
1/21
80%
100%
0%
0%
2.0%
5.0%
2.0%
4.8%
29/50
29/29
3/50
3/17
6/50
6/18
4/50
4/17
58%
100%
6.0%
17.6%
12.0%
33.3%
8.0%
23.5%
0.001
0.026
0.006
O.001
0/20
0/18
0/20
0/18
0%
0%
0%
0%
45/50
45/45
4/50
4/20
90%
100%
8%
20%
31/50
31/32
10/50
10/18
64.0%
96.9%
20.0%
55.6%
0.001
O.001
0/20
0/18
0/20
0/18
0%
0%
0%
0%
47/48
47/47
10/48
10/30
97.9%
100%
20.0%
33.3%
28/50
28/31
5/50
5/18
56.0%
93.3%
10.0%
27.8%
0.001
0.027
a The NCI bioassay doses were administered 5 days/week. The doses reported by NCI (1978) reflect averaging over
the total number of weeks of the study, but not over 7 days/week.
b Numbers of animals at risk for observed incidences do not always agree with those given in Table 4-1 because the
individual animal data tables containing the times of observation did not include notations of tissues missing due to
autolysis, etc.
0 Cochran-Armitage test for trend, using the Poly-3 adjusted incidence data.
d Trend test for male rat hemangiosarcomas was not statistically significant (the dose-response was not monotonically
increasing), the low-dose response (26%) was statistically significantly elevated relative to control (Fisher's exact test,
p = 0.04).
Source: NCI, 1978.
78
-------
Table 5-7. Human equivalent exposures and adjustment factors for
extrapolating from less-than-chronic exposure to lifetime exposure for the
exposure periods in the NCI gavage study of 1,2-dibromoethane (NCI,
1978)
Sex, species
Male rats
Female rats
Male mice
Female mice
Bioassay doses
(mg/kg/day)
Low Weeks l-49d
High Weeks 1-16
17-28
29-49d
Low Weeks l-61d
High Weeks 1-16
17-28
29-6 ld
Low Weeks 1-10
11-12
13-53
High Weeks 1-10
11-12
13-39
40-53
Low Weeks 1-10
11-12
13-53
High Weeks 1-10
11-12
13-39
40-53
40
80
0
40
40
80
0
40
60
100
60
120
200
120
60
60
100
60
120
200
120
60
Body
weight3
(kg)
0.45
0.30
0.030
0.025
Human
equivalent
exposuresb
(mg/kg/day)
11
22
10
20
8.
6
17
8.
2
16
Time of Lifetime
terminal exposure
sacrifice adjustment
(weeks) factor0
49 0.10
61 0.20
78 0.42
90 0.64
78 0.42
a Body weights are lifetime averages, rounded to the nearest 0.05 kg (rats) or 0.005 kg (mice).
b The doses initially administered by NCI (1978) were adjusted for human equivalence by multiplying by [animal
body weight (kg)/70 kg]0 25.
0 Lifetime exposure adjustment factor = (t/104)3, where t is the time of final sacrifice in a particular dose group.
d Male rats were purposely not dosed during weeks 42 and 46, female rats not during weeks 42, 47, 52, and 56.
79
-------
Table 5-8. Estimation of benchmark doses (BMD), lower 95% confidence
limits (BMDL), and oral slope factors; using Poly-3 adjusted tumor
incidence rates (see Table 5-1), for animals exposed orally to 1,2-
dibromoethane (NCI, 1978)
Sex,
species
Male
rats
Female
rats
Male
mice
Female
mice
Tumor type
Forestomach
Hemangiosarcoma
Thyroid follicular
cell adenoma
Forestomach0
Hemangiosarcoma
Hepatocellular
carcinoma
Adrenocortical
carcinoma
Forestomach0
Lung adenoma
Forestomach0
Lung adenoma
Point of departure
Benchmark
response
90%
20%
12.5%
100%
10%
15%
10%
100%
30%
100%
25%
BMD,
mg/kg/day
11.9
12.8
10.8
10
18
11
10
8.6
9.3
8.2
8.6
BMDL,
mg/kg/day
9.1
8.6
5.7
-
7.9
6.5
5.4
-
6.1
-
5.8
Partial
lifetime
exposure
oral slope
factor3
(mg/kg/day)1
0.10
0.023
0.022
(0.1)
0.013
0.023
0.018
0.12
0.049
0.12
0.043
Lifetime-
adjusted oral
slope factorb
(mg/kg/day)1
1.0
0.23
0.22
(0.50)
0.065
0.12
0.090
0.29
0.12
0.23
0.082
a Slope factors were estimated by dividing the BMR (expressed as a proportion) by the BMDL. The slope factors
should not be used with exposures higher than the corresponding BMDL, because the dose-response relationships tend
to be nonlinear above the point of departure. They also only describe risk for exposure lasting the specified percentage
of a lifetime.
b Slope factors were adjusted for lifetime exposure equivalence using the sex- and species-specific adjustments from
Table 5-7, under the assumption that tumor incidence over time increases with the third power of time.
0 Because there were no data available between 0% and 100% responses, curve fitting with meaningful confidence
bounds is not possible. An approximate slope factor is derived by dividing the proportion responding at the lowest
dose by that dose, expressed as the human equivalent dose, given in the BMD column.
80
-------
5.3.2. Inhalation Carcinogenicity
5.3.2.1. Choice of Inhalation Study/Data with Rationale and Justification
Three studies were identified that reported 1,2-dibromoethane-induced tumors following
inhalation. Stinson et al. (1981) reported benign neoplasms and carcinomas of the nasal cavity in
male and female B6C3FJ mice. Although the study was well designed and an adequate number of
test animals were used, this study is limited for the development of an inhalation slope factor
because only the nasal cavities of the animals were examined. NTP (1982) demonstrated that the
nasal cavity is not necessarily the most sensitive site of tumor formation in mice. Therefore the
Stinson et al. (1981) study was not used because similarly sensitive sites (lung, circulatory system)
were not examined. The study by Wong et al. (1982) is also not suitable for the development of an
inhalation slope factor because only one dose group was examined. The NTP (1982) study was
well-conducted, used an adequate number of test animals and dose levels, and examined
appropriate toxicological endpoints. Therefore, the study by NTP (1982) was used for
development of an inhalation unit risk.
5.3.2.2. Inhalation Dose-Response Data
In the NTP (1982) study, groups of 50 F344 rats and B6C3FJ mice of each sex were
exposed by inhalation to concentrations of 10 or 40 ppm 1,2-dibromoethane, 6 hours per day for 5
days per week. Untreated controls consisted of 50 rats and 50 mice of each sex exposed in
chambers to ambient air. Terminal sacrifices were conducted at 106 weeks in control animals and
at 104 weeks in low-dose animals. Survival in low-dose and control rats was similar for both
sexes. Terminal sacrifices were conducted at 79 weeks in the male mice, 89 weeks in the high-
dose male rats and 91 weeks in the high-dose female rats and mice. Although the treated male
mice demonstrated histopathology similar to that seen in the female mice, high mortality in all
groups that was not related to treatment and that started by week 10 of the study made these data
unsuitable for quantitative assessment.
Because there was increased mortality in both species that resulted in early termination of
the study, statistical procedures that can reflect the influence of intercurrent mortality on site-
specific tumor incidence rates were used to evaluate the tumor incidence levels. The Poly-3
procedure, described in section 5.3.3., was used for these data. Tumor incidences were statistically
significantly elevated with increasing exposure level at several sites: nasal cavity and circulatory
81
-------
system (hemangiosarcoma) in all three data sets; mesothelioma in male rats; mammary
fibroadenoma or adenocarcinoma in female rats; alveolar/bronchiolar adenoma or carcinoma in
female rats and mice; and adenomas or carcinomas of the bronchus, subcutaneous fibrosarcomas,
or mammary adenocarcinomas in female mice. These data are summarized in Table 5-9. The
individual times of tumor observations for each tumor type analyzed are listed in Appendix C2.
5.3.2.3. Inhalation Dose Adjustments and Extrapolation Methods
In order to extrapolate to low environmental exposure, the results of the inhalation bioassay
were characterized by fitting dose-response models to the tumor results showing significant
elevations with increasing exposure. Differences in survival and time course of tumor observation
among dose groups were addressed using both quantal dose-response models (with Poly-3
adjusted data) and the multistage-Weibull model.
Human equivalent exposures were estimated following the EPA RfC methodology
(U.S.EPA, 1994b). 1,2-Dibromoethane is relatively insoluble in water and demonstrates systemic
toxicity although there was some respiratory tract toxicity which may or may not be considered
portal of entry effects. Therefore, 1,2-dibromoethane is considered a Category 2 gas. Following
the RfC Methodology, the respiratory tract effects were assessed alternatively as portal of entry
effects or systemic effects while the rest of the effects were assessed as systemic effects.
For systemic effects, conversion to human equivalent concentrations involves comparison
of the blood:air partition coefficients for humans, mice, and rats (U.S.EPA, 1994b). Gargas et al.
(1989) reported a blood:air partition coefficient for rats of 119 for 1,2-dibromoethane. For their
pharmacokinetic model, Hissink et al. (2000) assumed that the value for humans would be the
same as for rats (see section 3). Without further information, the same assumption is made for this
assessment. Therefore, no adjustment for interspecies pharmacokinetics is assumed to be
necessary to estimate human equivalent concentrations for systemic effects. The equivalent
average continuous concentrations for systemic effects are respectively, (10 ppm) x (30 hrs per
wk)/(24 hrs per day x 7 days per wk) =1.8 ppm and (40 x 30)/(24 x 7) = 7.1 ppm.
For portal of entry effects, the human equivalent concentration is derived by multiplying
the duration-adjusted concentrations by an interspecies dosimetric adjustment for gas:respiratory
effects in affected regions of the lung. For example, for effects seen in male rats in the extra-
thoracic region, the adjustment uses the following calculation (U.S. EPA, 1994b):
82
-------
RGDR(ET) = (MVa/Sa)/(MVh/Sh), (Equation 5-3)
where
RGDR(ET) = regional gas dose ratio for the nasal cavity (extra-thoracic area of the
respiratory system)
MVa = animal minute volume ( male rat: 0.21 L/min)
MVh = human minute volume (13.8 L/min)
Sa = surface area of the extra-thoracic region of the animal respiratory
system (rat: 15.0 cm2)
Sh = surface area of the extra-thoracic region of the human respiratory
system (200 cm2)
Using these default values, the RGDR(ET)rat male = (0.21/15)7(13.8/200) = 0.20. The resulting
adjustments for rats and female mice are summarized in Table 5-10. In addition, female rats and
mice were also observed to have tumors in other regions of the respiratory tract. The
corresponding RGDRs are also summarized in Table 5-10.
Because the high-dose groups in all three studies were terminated slightly earlier than the
remaining groups, an adjustment to extrapolate to lifetime human exposure was applied to these
exposure levels prior to quantal dose-response modeling. Unlike the oral studies (see section
5.3.1.3.), the early termination was only approximately 10% of the intended length and only
affected one group in each set. The high-dose male rat exposure level was multiplied by (t/104)3 =
(89/104)3 = 0.63, and the female rat and mouse exposure levels were multiplied by (91/104)3 =
0.67.
In addition to a quantal analysis (as described in section 5.3.1.3), the characteristics of the
dose-response relationships for different tumor sites were assessed through time-to-tumor analyses
in order to adjust for competing mortality from cancer at other sites and differing time courses of
tumor incidence with increasing dose. The general model used for the time-to-tumor (or time-to-
response) analyses was the multistage-Weibull model, which has the form
P(d,t) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)*(t - t0)z]
where P(d,t) represents the probability of a tumor (or other response) by age t (in bioassay weeks)
for dose d (i.e., human equivalent exposure), and parameters z>l, t0>0, and q;>0 for i = 0, 1, ..., k,
83
-------
where k = the number of dose groups - 1. The parameter t0 represents the time between when a
potentially fatal tumor becomes observable and when it causes death (see below). The analyses
were conducted using the computer software TOX_RISK version 5.2 (Crump, 2000), which is
based on Weibull models drawn from Krewski et al. (1983). Parameters are estimated using the
method of maximum likelihood. Note that it was not necessary to adjust the administered dose for
lifetime exposure prior to modeling, because the software program characterizes the tumor
incidence over time from which it provides an extrapolation to lifetime exposure.
Tumor types were categorized by tumor context as either fatal or incidental tumors, in
order to adjust appropriately for competing risks. Incidental tumors are those tumors thought not
to have caused the death of an animal, while fatal tumors are thought to have resulted in animal
death. For the rats, nasal cavity tumors were treated as fatal tumors unless observed at the terminal
sacrifice in which case they were considered incidental. Furthermore, these tumors were
considered rapidly fatal, and t0 was set equal to 0 as there were insufficient data to reliably
estimate t0 in any event. Tumors at all other sites were treated as incidental for the rats. These
determinations are consistent with the determination made by EPA for 1,3-butadiene (U.S. EPA,
2002b). The work of Portier et al. (1986) in analyzing tumor types in NTP historical controls
lends support to these tumor context assumptions. For the mice, there were more tumor types
observed than for the rats, and it was not clear which tumor type may have been most uniformly
the cause of death. It was not possible to carry out an animal by animal determination of cause of
death. Consequently, all of the significant female mouse tumors were considered incidental.
Specific n-stage Weibull models were selected for the individual tumor types for each sex
based on the values of the log-likelihoods according to the strategy used by EPA (U.S.EPA,
2002b). If twice the difference in log-likelihoods was less than a chi-square with degrees of
freedom equal to the difference in the number of stages included in the models being compared,
the models were considered comparable and the most parsimonious model (i.e., the lowest-stage
model) was selected.
5.3.2.4. Inhalation Unit Risk
The results of applying quantal models to the Poly-3 adjusted tumor incidence data are
provided in Table 5-11. In several cases, the multistage model could not provide an adequate fit (p
> 0.1). The log-logistic model was applied in these cases. The model output is provided in
Appendix C2. Consistent with the Draft Guidelines for Carcinogen Risk Assessment (U.S.EPA,
84
-------
1999), a point of departure near the lower end of the observed data was selected for linear
extrapolation to low doses. An inhalation unit risk for each tumor site was calculated by dividing
the BMR level by the corresponding BMCL. Estimated unit risks ranged from 0.039 (ppm)"1 to
4.8 (ppm)"1. The highest slope factor for each species/sex combination resulting from quantal
dose-response modeling corresponded to nasal cavity tumors.
The results of applying the multistage-Weibull data to the time-to-tumor data are
summarized in Table 5-12, and the model output and dose-response curves are provided in
Appendix C2. The maximum likelihood estimates of the BMDs and 95% lower bounds (BMDLs)
for 10% extra risk for each tumor are provided, consistent with the BMDS technical guidance
(U.S.EPA, 2000d). Currently, the TOX_RISK software does not provide estimates corresponding
to risks higher than 10%, so the unit risks were based on the BMDL10s. The highest slope factor
for each species/sex combination resulting from time-to-tumor modeling corresponded to nasal
cavity tumors. However, approximate points of departure can be estimated from the central
tendency risk estimate at the lower end of the observed data range for each tumor site and are
provided in Table 5-13. These are provided only for comparison, since an estimate of the upper
bound on risk is not available; these approximate risks follow the same patterns as the other unit
risks, with the nasal cavity tumors showing the highest risk.
Unit risks calculated using the time-to-tumor approach were similar to those calculated
using the quantal approach, within a factor of 3. Goodness of fit varied between the two
approaches, with fits from one model not uniformly better than the other. The highest unit risks
within each sex and species corresponded to nasal cavity tumors in all cases, with nasal cavity
tumors an order of magnitude higher in male and female rats than in female mice. Because there
did not appear to be substantial model dependence in the results, the highest unit risks were used to
develop the overall unit risk. The quantal dose-response curves for these two data sets are
provided in Appendix C2, Model Outputs C-9 and C-15. An inhalation unit risk for exposure to
1,2-dibromoethane is calculated by dividing the response rate of 0.87 (87%) by the BMCL87 for
nasal cavity tumors in male rats, or 0.87/ 0.18 ppm = 4.8 (ppm)"1. At 25°C and standard pressure
(760 mm Hg), 1 ppm x (molecular wt )/24.5 mg/m3. The molecular weight of 1,2-dibromoethane
is 188. Thus, one ppm is equivalent to 188/24.5 = 7.7 mg/m3 of 1,2-dibromoethane. Hence, the
inhalation unit risk can be expressed as 4.8 (ppm)"V7.7 [(mg/m3)/ppm] ~ 0.6 (mg/m3)"1.
This unit risk should not be used with exposures greater than 0.18 ppm, or 0.023 mg/m3. Above
this level, the dose-response is not linear, and the modeled dose-response pattern should be used to
85
-------
estimate risk.
This inhalation unit risk is approximately 3-fold higher than the inhalation unit risk of 0.22
(mg/m3)-1 listed previously on IRIS (U.S. EPA, 1988b). The use of the RfC methodology to
characterize human equivalent exposures in the nasal region contributes a 5-fold increase over the
previous unit risk, all else being equal. The changes in cancer risk assessment since that time
account for the remaining difference, primarily extrapolating linearly from a point within the
observed data rather than from a modeled estimate of 1 x 10"6 risk.
There are several areas of uncertainty. First, there is uncertainty associated with
extrapolating to low doses from relatively high responses in the rats, greater than 70% in the low
dose groups. These studies did continue for close to the full lifetime, however, unlike the oral
study. Second, there is some uncertainty involving the appropriate dose metric for some of the
respiratory tumors. For the results from both modeling approaches, the unit risks for the
respiratory sites are also expressed in terms of systemic toxicity rather than portal of entry effects,
that is, as if the observed toxicity occurred after absorption and systemic distribution of the
carcinogenic moiety. In the case of the alveolar/bronchi olar adenomas in female rats and mice,
there is some indication from the oral study that these tumors can occur after ingestion of 1,2-
dibromoethane, at least in the female mice. Following the RfC methodology, if systemic toxicity
were solely responsible for these tumors, then the unit risks for these sites should be about 2- to 3-
fold higher than if the tumors were due solely to portal of entry effects. There is no evidence that
the nasal cavity tumors would result from systemic toxicity.
Last, reliance on single tumor sites probably somewhat underestimates the carcinogenic
potential of 1,2-dibromoethane by the inhalation route. Using the same analysis that was applied
to the oral slope factors to consider total risk from multiple tumor sites (see section 5.3.1.4), the
combined risk of the significant tumors for male and female rats and for female mice was
considered. As with those analyses, these sums are approximate (see section 5.3.1.4 for details).
For the rats, the risks of nasal cavity tumors were clearly much larger than for the other sites, so it
is not surprising that the final outcome only increases to 5.0 (ppm)"1, which when converted to
units of mg/m3 still rounds to 0.6 (mg/m3)"1.
For the female rats, the sum of the five significant sites increases the unit risk for nasal
cavity tumors more than for the males, with a sum of 3.9 (ppm)"1 compared with 3.5 (ppm)"1 for
nasal cavity tumors alone. For the female mice, however, four of the six sites were very similar,
86
-------
with respiratory (lung/bronchial or alveolar/bronchiolar) adenomas or carcinomas somewhat
lower. The resulting 95% UCL on the slope factor for all six sites was 0.86 (ppm)"1, almost three
times higher than the risk for nasal cavity tumors alone, but roughly a factor of six lower than the
unit risk based on male rat nasal cavity tumors.
87
-------
Table 5-9. Observed and adjusted tumor incidence rates in rats and mice
exposed to 1,2-dibromoethane by inhalation
Tumor site
Incidence
adjustment
Administered concentration
0
ppm
10
ppm
40
ppm
Male rats
Nasal cavity tumorsa
Hemangiosarcoma
Mesothelioma, tunica
vaginalis
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
1/50
1/46
0/50
0/46
1/50
1/46
2.0%
2.2%
0%
0%
2.0%
2.1%
39/50
39/45
1/50
1/43
8/50
8/43
78.0%
86.6%
2.0%
2.2%
16.3%
18.6%
41/50
41/43
15/50
15/28
25/50
25/35
82.0%
95.3%
30.0%
53.6%
50.0%
71.4%
Female Rats
Nasal cavity tumorsb
Hemangiosarcoma
Mammary
fibroadenoma
Mammary
adenocarcinoma
Alveolar/bronchiolar
adenoma/carcinoma
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
1/50
1/47
0/50
0/47
4/50
4/47
1/50
1/47
0/50
0/47
2.0%
2.1%
0%
0%
8.0%
8.5%
2.0%
2.1%
0%
0%
34/49
34/46
0/49
0/42
29/49
29/46
0/49
0/42
0/49
0/42
70.8%
73.9%
0%
0%
59.2%
63.0%
0%
0%
0%
0%
43/50
43/46
5/50
5/26
24/50
24/34
4/50
4/26
5/50
5/25
86.0%
93.5%
10.0%
19.2%
48.0%
70.6%
8.0%
15.4%
10.0%
20.0%
Female Mice
Nasal cavity tumors0
Hemangiosarcoma
Alveolar/bronchiolar
adenoma/carcinoma
Lung/bronchial
adenoma/carcinoma
Fibrosarcoma
Mammary
adenocarcinoma
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
None (observed)
Poly-3
0/50
0/46
0/50
0/46
4/50
4/46
0/50
0/46
0/50
0/46
2/50
2/46
0%
0%
0%
0%
8.0%
8.7%
0%
0%
0%
0%
4.0%
4.3%
0/50
0/38
12/50
12/40
11/50
11/40
1/50
1/38
5/50
5/39
14/50
14/40
0%
0%
24.0%
30.0%
22.0%
27.5%
2.0%
2.6%
10.0%
12.8%
28.0%
35.0%
8/50
8/27
25/50
25/35
41/50
41/44
8/50
8/26
11/50
11/27
9/50
9/27
16.0%
29.6%
50.0%
71.4%
82.0%
93.2%
16.0%
30.8%
22.0%
40.7%
18.0%
33.3%
Trend test
p-valued
0.001
O.001
O.001
0.001
O.001
0.001
0.023
O.001
0.001
0.001
O.001
O.001
0.001
0.001
a Adenoma, carcinoma, adenocarcinoma, adenomatous polyp, papillary adenoma, squamous cell carcinoma, or
squamous cell papilloma.
b Adenoma, carcinoma, adenocarcinoma, adenomatous polyp, papillary adenoma, papillary polyp, or squamous cell
carcinoma.
0 Adenoma or carcinoma. d Cochran-Armitage test for trend, using the Poly-3 adjusted incidence data.
Source: NTP 1982.
88
-------
Table 5-10: Summary of regional gas dose ratios for estimating human
equivalent exposures corresponding to respiratory tumors observed in
NTP (1982) inhalation bioassay
Sex,
species
Male
rats
Female
rats
Female
mice
Body
weight,
kg
0.30
0.20
0.035
Minute
volume3,
L/min
0.21
0.15
0.041
Affected respiratory
regions
Extra-thoracic (nasal cavity)
Extra-thoracic (nasal cavity)
Pulmonary
Extra-thoracic (nasal cavity)
Tracheobronchial
Pulmonary
Surface
areab
15.0 cm2
15.0 cm2
0.34 m2
3.0 cm2
3.5 cm2
0.05 m2
Regional
gas dose
ratio0
0.20
0.14
1.73
0.20
2.7
3.2
Human equivalent
continuous11
concentrations
Low- High-
dose, dose,
ppm ppm
0.36 1.42
0.25 0.99
3.1 12.3
0.36 1.42
4.86 19.2
5.8 22.7
a Minute volumes were estimated using the allometric equations provided in the RfC methodology document (U.S.
EPA, 1994b):
MVmouse: ln(MV) = 0.326 + 1.05 ln(BW)
MVrat: ln(MV) = -0.578 + 0.821 ln(BW)
b Surface areas provided in the RfC methodology document (U.S. EPA, 1994b). In addition, the corresponding surface
areas for humans:
Extra-thoracic (nasal cavity): 200 cm2
Tracheobronchial: 3,200 cm2
Pulmonary: 54 m2
RGDR = (MVa/Sa)/(MVh/Sh), where
RGDR = regional gas dose ratio for a specific area of the respiratory system
Mva = animal minute volume
Mvh = human minute volume (13.8 L/min)
Sa= surface area of the specified region of the animal respiratory system
Sh= surface area of the specified region of the human respiratory system
d Administered concentrations were low dose, 10 ppm, and high dose, 40 ppm. A continuous exposure adjustment was
included by multiplying by (5 days/7 days) x (6 hours/24 hours) = 0.178. Human equivalent
continuous concentrations were estimated by multiplying equivalent continuous exposures (not shown) by the regional
gas dose ratio.
89
-------
Table 5-11. Estimates of benchmark concentration (BMC), lower 95%
confidence limits (BMCL), and unit risks; using Poly-3 incidence rates",
for animals exposed by inhalation to 1,2-dibromoethane (NTP, 1982)
Sex,
species
Male
rats
Female
rats
Female
mice
Tumor type
Nasal tumors
Hemangiosarcoma
Mesothelioma
Nasal tumors
Alveolar/bronchiolar
adenoma/carcinoma
Hemangiosarcoma
Mammary
fibroadenomab
Mammary
adenocarcinoma
Nasal tumors
Lung/bronchial
adenoma/carcinoma
Alveolar/bronchiolar
adenoma/carcinoma
Hemangiosarcoma
Fibrosarcoma
Mammary
adenocarcinomab
Point of departure
BMR
87%
10%
20%
70%
10%
10%
55%
10%
25%
10%
30%
35%
15%
25%
BMC, BMCL,
ppm ppm
0.38
1.9
1.9
0.26
6.6
3.9
1.9
5.7
0.96
5.4
6.0
1.8
1.7
2.1
0.18
1.5
1.3
0.20
3.4
2.0
1.2
2.6
0.73
3.3
4.8
1.4
1.1
1.3
Unit risk
Regional
respiratory
dosimetry0
4.8
NR
NR
3.5
0.029
NR
NR
NR
0.34
0.03
0.062
NR
NR
NR
(ppm)1
Systemic
dosimetryd
0.96
0.067
0.15
0.49
0.051
0.050
0.46
0.039
0.068
0.082
0.20
0.25
0.14
0.19
a Tumor incidence data (Poly-3 incidence rates in Table 5-9) were fit using the general form of the multistage
model: P(d) = 1 - exp(-q0 - q{d -... -q6d6), except where p-values are footnoted (see footnote b).
b In cases where the multistage model did not provide an adequate fit (p<0.1), the log-logistic model was used:
P(d) = q0 + (1- q0 )/[l+exp(-a-b*ln(d)].
0 Reported modeling results reflect RGDR dosimetry adjustments as summarized in Table 5-10. NR indicates that
regional respiratory dosimetry was not relevant for particular tumor sites.
d Reported unit risks reflect use of human equivalent lifetime concentrations. For respiratory tract endpoints, unit
risks corresponding to systemic toxicity rather than portal of entry effects were estimated by multiplying the
RGDR-based unit risks by the corresponding RGDR adjustments.
90
-------
Table 5-12. Estimates of benchmark concentration (BMC) associated with
an extra risk of 10%, lower 95% confidence limits (BMCL), and unit
risks; using multistage-Weibull time-to-tumor modeling", for animals
exposed by inhalation to 1,2-dibromoethane (NTP, 1982)
Sex, species
Male rats
Female rats
Female mice
Tumor type
Nasal tumors
Hemangiosarcoma
Mesothelioma
Nasal tumors
Alveolar/ bronchial
adenoma/ carcinoma
Hemangiosarcoma
Mammary fibroadenoma
Mammary
adenocarcinoma
Nasal tumors
Alveolar/ bronchiolar
adenoma/
carcinoma
Lung/bronchial
adenoma/ carcinoma
Hemangiosarcoma
Fibrosarcoma
Mammary
adenocarcinoma
Point of departure
BMC10,
ppm
0.065
3.1
0.92
0.057
1.7
7.5
0.042
1.7
0.86
3.7
9.3
0.69
1.9
2.2
BMCL10,
ppm
0.053
2.3
0.69
0.046
0.64
3.8
0.030
0.71
0.84
2.2
4.4
0.052
1.1
1.2
Unit riskc
Regional
respiratory
dosimetryb
1.9
NR
NR
2.2
0.16
NR
NR
NR
0.12
0.045
0.022
NR
NR
NR
(ppm)1
Systemic
dosimetry0
0.38
0.040
0.14
0.31
0.27
0.026
3.3
0.14
0.024
0.14
0.059
0.19
0.090
0.083
a Individual tumor incidence data (Appendix C-2) were fit using the multistage-Weibull model:
P(d)=l-exp[(-q0-qid-...-q6d6)tz].
b Reported modeling results reflect RGDR dosimetry adjustments as summarized in Table 5-10. NR indicates
that regional respiratory dosimetry was not relevant for particular tumor sites.
0 Reported unit risks reflect use of human equivalent lifetime concentrations.
91
-------
Table 5-13. Approximate unit risks, using central tendency dose-response
estimates from multistage-Weibull time-to-tumor modeling, for animals
exposed by inhalation to 1,2-dibromoethane (NTP, 1982)
Sex, species
Male Rats
Female rats
Female mice
Tumor type
Nasal tumors
Hemangiosarcoma
Mesothelioma
Nasal tumors
Alveolar/ bronchiolar
adenoma/ carcinoma
Hemangiosarcoma
Mammary fibroadenoma
Mammary adenocarcinoma
Nasal tumors
Alveolar/ bronchiolar
adenoma/ carcinoma
Lung/bronchial
adenoma/ carcinoma
Hemangiosarcoma
Fibrosarcoma
Mammary adenocarcinoma
Point of departure
BMRa
0.92
0.03
0.21
0.82
0.033
0.026
0.73
0.048
0.043
0.084
0.049
0.31
0.088
0.16
Low doseb, ppm
0.36
1.8
1.8
0.36
3.1
1.8
1.8
1.8
0.36
5.8
4.9
1.8
1.8
1.8
Central tendency
unit risk,
(ppm)1
2.6
0.017
0.12
2.3
0.011
0.014
0.41
0.027
0.12
0.014
0.010
0.17
0.049
0.089
a Central tendency BMR estimated using time-to-tumor models referenced in Table 5-12.
b Low dose (100 ppm) adjusted by site for sex- and species-specific RGDR dosimetry, as summarized in Table 5-
10.
92
-------
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF
HAZARD AND DOSE RESPONSE
6.1. HUMAN HAZARD POTENTIAL
1,2-Dibromoethane is a colorless liquid that has a chloroform-like odor and evaporates
easily. It has limited solubility in water but is miscible and soluble in many organic solvents. 1,2-
Dibromoethane was used primarily as an anti-knock compound in leaded gasoline and also has
been used as a fumigant for grain, fruits, and soil. Today, it is more likely to be used as a solvent
in resins, gums, and waxes, and as a intermediate in dye and pharmaceutical manufacturing. Past
exposure from 1,2-dibromoethane was primarily from emissions and exhaust from vehicles using
leaded gasoline and from its use as a fumigant. However, with restrictions on the use of leaded
gasoline and limited fumigation practices, exposure to 1,2-dibromoethane has decreased.
It appears from both human and animal studies that 1,2-dibromoethane is rapidly absorbed
by the oral, inhalation, and dermal routes. Animal studies indicate that 1,2-dibromoethane is
metabolized by two pathways: oxidation by P450-monooxygenases and GSH conjugation
mediated by glutathione-S-transferase. 1,2-Dibromoethane is eliminated mainly in the urine as
mercapturic acid derivatives. However, a small amount of 1,2-dibromoethane is eliminated in the
expired air and feces.
No subchronic- or chronic-duration human studies were located concerning ingestion of
1,2-dibromoethane. Noncancer effects reported in experimental animals orally exposed to 1,2-
dibromoethane include weight gain depression, high mortality, hyperkeratosis and acanthosis of
the forestomach, liver and adrenal cortex degeneration, and testicular atrophy (NCI, 1978). In
mice, weight gain depression, high mortality, and testicular atrophy have been reported (NCI,
1978). All non-neoplastic changes reported following oral exposure to 1,2-dibromoethane have
occurred at doses that also produced cancer. In bulls, oral administration of 1,2-dibromoethane
produced adverse alterations in various sperm parameters and testicular histology (Amir and
Volcani, 1965, 1967). Although doses were lower, these studies were not selected for RfD
development because of the small numbers of animals and use of one exposure level and because
bulls are ruminants with a significantly different physiology. Moreover, an allometric adjustment
93
-------
would result in a dose quite similar to those used in the NCI (1978) study.8 These bull studies,
however, provide supporting evidence for testicular effects of low-dose 1,2-dibromoethane.
In humans, inhalation exposure to 1,2-dibromoethane in the workplace has been reported to
cause adverse reproductive and fertility effects (Ratcliffe et al., 1987; Schrader et al., 1988; Wong
et al., 1979). Specifically, decreases in average sperm count, percentage of viable sperm, and
percentage of motile sperm have been documented following long-term exposure. Morphological
abnormalities in sperm, such as tapered heads, absent heads, and abnormal tails, have also been
reported following chronic exposure. Short-term exposures have been associated with sperm
abnormalities in workers exposed to 1,2-dibromoethane. After six weeks of exposure to 1,2-
dibromoethane, sperm velocity and semen volume were reported to decrease. Decreased
reproductive performance of occupationally exposed workers was assessed by the number of live
births to exposed workers' wives. Poor exposure data and moderate-to-extensive exposure by the
dermal route limit the value of these findings for risk assessment purposes. Reproductive studies
in laboratory animals (Short et al., 1978) provide useful quantitative information for risk
assessment purposes but suffer from a lack of sperm quality and count measures.
In chronic animal studies, weight gain depression, high mortality, hepatic necrosis,
nephropathy, testicular atrophy, and degeneration of the adrenal cortex have only been reported in
rats and mice at exposures that also cause cancer. The results of subchronic inhalation studies
revealed weight gain depression, swelling of adrenocortical cells, decreases in thyroid follicle size,
and formation of megalocytic cells of the lining of bronchioles in rats and mice (NTP, 1982;
Nitschke et al., 1981; Reznik et al., 1980). In rats, relative liver and kidney weights, focal
epithelial hyperplasia of the nares, and diffuse respiratory hyperplasia have also been reported.
Cancers have not been reported following subchronic-duration inhalation exposure to 1,2-
dibromoethane.
There are no reports of cancer in humans associated with exposure to 1,2-dibromoethane.
However, the human studies have serious limitations, such as poor exposure assessment and co-
exposure to other potential carcinogens. Oral exposure to 1,2-dibromoethane induces cancer of the
forestomach in rats and mice (NCI, 1978). The relevance of forestomach effects to humans has
been questioned, particularly for nongenotoxic chemicals whose mode of action is believed to
involve irritation and cell proliferation from long-term exposure ( Poet et al., 2003). While the
8After allometric adjustment, the doses given to the NCI (1978) male rats were less than 2 times, and
therefore similar to, the doses given to the Amir and Volcani (1965, 1967) bulls (11 vs 21 mg/kg3/4/day).
94
-------
forestomach contains features, such as minimal vascularization and stratified squamous cells,
which result in a longer residence time of food-borne agents than is received by the oesophageal
tissue or the glandular stomach (Grice, 1988; Poet et al., 2003), effects in this organ are believed to
be relevant to 1,2-dibromoethane and other genotoxic chemicals that do not appear to require
precursor events (e.g., irritation) associated with long residence time to induce tumors.
Hemangiosarcoma in male rats, hepatocellular carcinoma in female rats, and lung cancer in mice
were also reported following oral exposure to 1,2-dibromoethane (NCI, 1978). Chronic-duration
inhalation exposure produces tumors of the nasal cavity in rats and lung tumors in mice (NTP,
1982). Hemangiosarcoma, mesothelioma of the tunica vaginalis, mammary carcinoma, lung
adenoma, and fibrosarcoma have also been reported following long-term exposure to 1,2-
dibromoethane (NTP, 1982).
There is also no evidence from human studies that 1,2-dibromoethane is genotoxic.
Steenland et al. (1985, 1986) performed a cytogenetic examination on exposed workers to
investigate induction of sister chromatid exchanges and chromosomal aberrations in peripheral
lymphocytes and found no significant increase in either parameter. However, exposure
concentrations were low and may not have been capable of inducing chromosomal damage. In
contrast to human studies, there is strong evidence from in vitro and in vivo animal studies that
1,2-dibromoethane is genotoxic. 1,2-Dibromoethane is a direct-acting mutagen in bacteria. 1,2-
Dibromoethane was positive for S. typhimurium revertant strains TA1535, TA100, and TA98
(Barber et al., 1981) and induced point mutations in S. typhimurium strains TA1535 and TA100, S.
coelicolor, and A nidulans (Carere and Morpurgo, 1981). 1,2-Dibromoethane has been shown to
induce chromosomal aberrations and sister chromatid exchanges in Chinese hamster ovary cells
(Ivett et al., 1989; Tan and Hsie, 1981; Brimer et al., 1982; Ballering et al., 1998; Graves et al.,
1996). 1,2-Dibromoethane has also been reported to induce gene mutations in two human
lymphoblastoid cell lines, AHH-1 and TK6 (Crespi et al., 1985), and sister chromatid exchanges
in human peripheral lymphocyte cultures (Tucker et al., 1984). In vivo, 1,2-dibromoethane
induced DNA damage in rats following oral administration (Kitchin and Brown, 1986, 1987;
Sasaki et al., 1998). Intraperitoneal administration of radiolabeled 1,2-dibromoethane has been
shown to bind DNA in the liver, kidney, stomach, and lung of rats and mice (Arfellini et al., 1984),
and ,S'-[2-(N7-guanyl)ethyl]glutathione has been identified as the major DNA adduct formed in rats
following treatment with 1,2-dibromoethane (Kim et al., 1990; Koga et al., 1986).
The genotoxicity of 1,2-dibromoethane is thought to be related to its conjugation with GSH
as catalyzed by glutathione-S-transferase, which results in the formation of an episulfonium ion
that can react with DNA to form ,S'-[2-(N7-guanyl)ethyl]glutathione DNA. 1,2-Dibromoethane has
95
-------
been reported to cause DNA damage and UDS in rat hepatocytes and spermatocytes exposed both
in vitro and in vivo, and inhibition of cytochrome P-450-mediated oxidation in either cell type did
not inhibit UDS. However, depletion of cellular GSH inhibited the induction of UDS in both cell
types, which suggests that GSH conjugation and not P450 oxidation is responsible for the
genotoxic effects of 1,2-dibromoethane. Inhibition of hepatic mixed-function oxidases in vivo was
associated with positive UDS response to 1,2-dibromoethane in spermatocytes, but there was no
effect on 1,2-dibromoethane-induced UDS in hepatocytes.
1,2-Dibromoethane-induced cytotoxicity may be dependent on its metabolism by
cytochrome P450 and/or conjugation with GSH. Microsomal cytochrome P-450-dependent
oxidative metabolism of 1,2-dibromoethane produces the metabolite 2-bromoacetaldehyde. This
metabolite has been reported to cause lipid peroxidation and protein binding. Glutathione
conjugation may also contribute to cytotoxicity. Depletion of hepatic mitochondrial GSH by 1,2-
dibromoethane has been correlated with hepatotoxicity and perturbations in mitochondrial Ca2+
homeostasis. The results of in vitro and in vivo experiments suggest that the renal toxicity of 1,2-
dibromoethane may be due to its biotransformation by GSH conjugation followed by further
conversion in the kidney to highly reactive metabolites. It has been suggested that lipid
peroxidation may play a role in the 1,2-dibromoethane-induced pathogenesis of liver cell necrosis.
The available evidence further supports a conclusion that 1,2-dibromoethane is a genotoxic
carcinogen. Under the Draft Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1999), 1,2-
dibromoethane is classified as "likely to be carcinogenic to humans" via both the oral and
inhalation routes of exposure based on strong evidence of carcinogenicity in animals and
inconclusive evidence of carcinogenicity in an exposed human population. Under the Guidelines
for Carcinogen Risk Assessment (U.S. EPA, 1986a), 1,2-dibromoethane would be classified as
Group B2 -Probable Human Carcinogen.
In conclusion, while the potential for human exposure to 1,2-dibromoethane is limited,
studies suggest that even low levels of 1,2-dibromoethane exposure can pose considerable health
risks. 1,2-Dibromoethane is rapidly absorbed by all potential routes of human exposure and is a
likely human carcinogen and reproductive toxicant. There is also some limited evidence for
neurobehavioral developmental effects caused by 1,2-dibromoethane as well as endocrine
disruption (based on effects on other endocrine organs as well as changes in hormone levels). As
is discussed in Chapter 5, the database associated with 1,2-dibromoethane's potential to cause
effects on the developing fetus and newborn is limited. Additional research in this area could
increase the certainty of this assessment.
96
-------
6.2. DOSE RESPONSE
An RfD of 9E-3 mg/kg-day was derived. It is based on liver, testicular, and adrenal effects
observed in male rats following chronic oral exposure. The RfD is supported by a benchmark dose
analysis of these effects and is based on a LOAEL of 27 mg/kg-day and application of an
uncertainty factor of 3000, which was reduced from an overall uncertainty factor of 10,000 (10 for
interspecies, 10 for intraspecies, 10 for LOAEL to NOAEL extrapolation, 10 for database
deficiency) in recognition of the lack of independence of the individual factors and that the
multiplication of four or five values of 10 is likely to yield unrealistically conservative RfCs
(Dourson and Stara, 1983; U.S. EPA, 1994, 2002).
An RfC of 9E-3 mg/m3 was derived based on induction of liver pathology in both male and
female rats using benchmark dose methodology in a lifetime inhalation bioassay (NTP, 1982).
The RfC was calculated by application of uncertainty factors 3 for interspecies, 10 for intraspecies,
and 10 for database uncertainty to the BMDL(HEC) of 2.8 mg/m3.
A cancer oral slope factor of 2 (mg/kg-day)"1 was calculated from the carcinogenicity
bioassay in male rats (NCI, 1978) using benchmark dose methodology. An inhalation cancer slope
factor estimate of 0.6 (mg/m3)"1 was calculated from an inhalation carcinogenicity bioassay in male
rats (NTP 1982).
Confidence in the study utilized to derive the RfD and oral cancer slope factor is low to
medium. Although the critical study was of chronic duration and involved a large number of
animals, high mortality, close dose spacing, and the absence of a NOAEL made this study difficult
to assess. Confidence in the oral database is also considered to be medium. Although oral data on
reproductive/developmental effects are very limited, some indication of doses that might cause
these effects can be obtained from gavage studies in bulls and inhalation studies in rats and mice.
The overall confidence in these oral benchmarks is considered to be low to medium.
Confidence in the study used to derive the RfC and inhalation cancer slope factor is
medium. The NTP (1982) inhalation study was well designed, using an adequate number of
animals of both sexes, but was limited because of excessive mortality in the high-dose groups of
both species, moderate mortality in low-dose female mice, and excessive mortality in male mice
not related to 1,2-dibromoethane exposure. Confidence in the inhalation database is medium.
97
-------
There are systemic inhalation toxicity studies in two species, one-generation (but no
multigeneration) reproductive study, and developmental toxicity studies in two species that did not
cover the full period of gestation. Although animal studies have shown that
reproductive/developmental effects in females are likely to occur only at doses inducing maternal
toxicity, the possibility remains that sperm quality may be adversely affected at lower doses. The
overall confidence in these inhalation benchmarks is medium.
98
-------
7. REFERENCES
Adkins, B, Jr.; Van Stee, EW; Simmons, JE; et al. (1986) Oncogenic response of strain A/J mice to
inhaled chemicals. J Toxicol Environ Health 17:311-322.
Albano, E; Poli, G; Tomasi, A; et al. (1984) Toxicity of 1,2-dibromoethane in isolated
hepatocytes: role of lipid peroxidation. Chem Biol Interact 50:255-265.
Amir, D. (1973) The sites of the spermicidal action of ethylene dibromide in bulls. J Reprod Fertil
35:519-525.
Amir, D. (1975) Individual and age differences in the spermicidal effect of ethylene dibromide in
bulls. J Reprod Fertil 44:561-565.
Amir D. (1991) The spermicidal effect of ethylene dibromide in bulls and rams. Mol Reprod Dev
Jan;28(l):99-109.
Amir, D; Ben-David, E. (1973) The pattern of structural changes induced in bull spermatozoa by
oral or injected ethylene dibromide (EDB). Ann Biol Anim Biochem Biophys 13:165-170.
Amir, D; Lavon, U. (1976) Changes in total nitrogen, lipoproteins and amino acids in epididymal
and ejaculated spermatozoa of bulls treated orally with ethylene dibromide. J Reprod Fertil
47:73-76.
Amir, D; Volcani, R. (1965) Effect of dietary ethylene dibromide on bull semen. Nature
206:99-100.
Amir, D; Volcani, R. (1967) The effect of dietary ethylene dibromide (1,2-dibromoethane) on the
testes of bulls: a preliminary report. Fertil Steril 18:144-148.
Amir, D; Esnault, C; Nicolle, JC; et al. (1977) DNA and protein changes in the spermatozoa of
bulls treated orally with ethylene dibromide. J Reprod Fertil 51:453-456.
Anderson, EL; Albert, RE; McGaughy, R; et al. (1983) Quantitative approaches in use to assess
cancer risk. Risk Anal 3:277-295.
Arfellini, G; Bartoli, S; Colacci, A; et al. (1984) In vivo and in vitro binding of 1,2-dibromoethane
and 1,2-dichloroethane to macromolecules in rat and mouse organs. J Cancer Res Clin Oncol
108:204-213.
Asita, AO; Hayashi, M; Kodama, Y; et al. (1992) Micronucleated reticulocyte induction by
ethylating agents in mice. Mutat Res 271:29-37.
Bailer, AJ; Portier, CJ. (1988) Effects of treatment-induced mortality and tumor-induced mortality
on tests for carcinogenicity in small samples. Biometrics 44:417-431.
99
-------
Ballering, LAP; Nivard, MJM; Vogel, EW. (1994) Mutation spectra of 1,2-dibromoethane,
1,2-dichloroethane and l-bromo-2-chloroethane in excision repair proficient and repair deficient
strains of Drosophila melanogaster. Carcinogenesis 15:869-875.
Ballering, LA; Vogel, EW; Vrieling, H; et al. (1998) Strand-specific mutation induction by 1,2-
dibromoethane at the hypoxanthine-guanine phosphoribosyltransferase locus of Chinese hamster
ovary cells. Mutagenesis 13:61-65.
Barber, ED; Donish, WH; Mueller, KR. (1981) A procedure for the quantitative measurement of
volatile liquids in the Ames Salmonella/microsome assay. Mutat Res 90:31-48.
Barnett, LB; Lovell, DP; Felton, CF; et al. (1992) Ethylene dibromide: negative results with the
mouse dominant lethal assay and the electrophoretic specific-locus test. Mutat Res 282:127-133.
Bishop, JB; Morris, RW; Seely, JC; et al. (1997) Alterations in the reproductive patterns of female
mice exposed to xenobiotics. Fundam Appl Toxicol 40:191-204.
Bjorge C; Brunborg G; Wiger R; et al. (1996) A comparative study of chemically induced DNA
damage in isolated human and rat testicular cells. Reprod Toxicol 10(6):509-19.
Botti, B; Moslen, MT; Trieff, NM; et al. (1982) Transient decrease of liver cytosolic glutathione S-
transferase activities in rats given 1,2-dibromoethane or CC14. Chem Biol Interact 42:259-270.
Botti, B; Bini, A; Calligaro, A; et al. (1986) Decrease of hepatic mitochondrial glutathione and
mitochondrial injury induced by 1,2-dibromoethane in the rat in vivo: effect of diethylmaleate
pretreatment. Toxicol Appl Pharmacol 83:494-505.
Botti, B; Ceccarelli, D; Tomasi, A; et al. (1989a) Biochemical mechanism of GSH depletion
induced by 1,2-dibromoethane in isolated rat liver mitochondria. Evidence of a GSH conjugation
process. Biochim Biophys Acta 992:327-332.
Botti, B; Franceschini, V; Tomasi, A; et al. (1989b) Mechanistic aspects of
1,2-dibromoethane-induced mitochondrial GSH depletion in vitro. Adv Biosci 76:99-105.
Brandt, I; Brittebo, EB; Kowalski, B; et al. (1987) Tissue binding of 1,2-dibromoethane in the
cynomolgus monkey (Macacafascicularis). Carcinogenesis 8:1359-1361.
Brimer, PA; Tan, EL; Hsie, AW. (1982) Effect of metabolic activation on the cytotoxicity and
mutagenicity of 1,2-dibromoethane in the CHO/HGPRT system. Mutat Res 95:377-388.
Brittebo, EB; Kowalski, B; Brandt, I. (1989) Epithelial binding of 1,2-dichloroethane in mice.
Toxicology 56:35-45.
Broda, C; Nachtomi, E; Alumot, E. (1976) Differences in liver morphology between rats and
chicks treated with ethylene dibromide. Gen Pharmacol 7:345-348.
100
-------
Brown-Woodman PD; Hayes LC; Huq F; et al. (1998) In vitro assessment of the effect of
halogenated hydrocarbons: chloroform, dichloromethane, and dibromoethane on embryonic
development of the rat. Teratology 57(6):321-33.
Carere, A; Morpurgo, G. (1981) Comparison of the mutagenic activity of pesticides in vitro in
various short term assays. Prog Mutat Res 2:87-104.
Channarayappa; Ong, T; Nath, J. (1992) Cytogenetic effects of vincristine sulfate and ethylene
dibromide in human peripheral lymphocytes: micronucleus analysis. Environ Mol Mutagen
20:117-126.
Chem Eng News (1983) 61:4.
Clapp, DE. (1986) Industrial hygiene in-depth survey report, papaya packing/shipping facilities,
Hilo, Hawaii. NIOSH 64.17 (unpublished).
Cmarik, JL; Inskeep, PB; Meyer, DJ; et al. (1990) Selectivity of rat and human glutathione S-
transferases in activation of ethylene dibromide by glutathione conjugation and DNA binding and
induction of unscheduled DNA synthesis in human hepatocytes. Cancer Res 50:2747-2752.
Cmarik, JL; Humphreys, WG; Bruner, KL; et al. (1992) Mutation spectrum and sequence
alkylation selectivity resulting from modification of bacteriophage M13mpl8 DNA with S-(2-
chloroethyl)glutathione. Evidence for a role of S-(2-N7-guanyl)ethyl)glutathione as a mutagenic
lesion formed from ethylene dibromide. J Biol Chem 267:6672-6679.
Colacci, A; Perocco, P; Vaccari, M; et al. (1995) 1,2-Dibromoethane as an initiating agent for cell
transformation. Jpn J Cancer Res 86:168-173.
Colacci, A; Vaccari, M; Perocco, P; et al. (1996) Enhancement of BALB/c 3T3 cells
transformation by 1,2-dibromoethane promoting effect. Carcinogenesis 17:225-231.
Crespi, CL; Seixas, GM; Turner, TR; et al. (1985) Mutagenicity of 1,2-dichloroethane and 1,2-
dibromoethane in two human lymphoblastoid cell lines. Mutat Res 142:133-140.
Crump, K.S. (2001) TOX_RISK dose-response software, version 5.2 (2000-2001) KS Crump
Group.
DeLeve, LD. (1997) Effect of decreased glutathione levels in hereditary glutathione synthetase
deficiency on dibromoethane-induced genotoxicity in human fibroblasts. Mutat Res 389:291-297.
DiRenzo, AB; Gandolfi, AJ; Sipes, IG. (1982) Microsomal bioactivation and covalent binding of
aliphatic halides to DNA. Toxicol Lett 11:243-252.
Dourson, ML; Knauf, L; Swartout, J. (1992) On reference dose (RfD) and its underlying toxicity
database. Toxicol Ind Health 8:171-189.
101
-------
Dourson, M.L; Stara J.F. (1983) Regulatory history and experimental support of uncertainty
(safety) factors. Regul Toxicol Pharmacol 3:224-238.
Dourson, ML; Felter, SP; Robinson, D. (1996) Evolution of science-based uncertainty factors in
noncancer risk assessment. Regul Toxicol Pharmacol 24:108-120.
Dusek, Z; Novotna, B; Vodickova, L; et al. (2003) Effects of 1,2-dibromoethane on hematopoiesis
in the chick embryo. Xenobiotica 33(4):443-58.
El-Masri, HA; Bell, DA; Portier, CJ (1999) Effects of glutathione transferase theta polymorphism
on the risk estimates of dichloromethane to humans. Toxicol Appl Pharmacol 158:221-230.
Eljack, AH; Hrudka F. (1979) Pattern and dynamics of teratospermia induced in rams by
parenteral treatment with ethylene dibromide. J Ultrastruct Res 67:124-134.
Elliott, BM; Ashby, J. (1980) Ethylene dibromide and disulfiram: Studies in vivo and in vitro on
the mechanism of the observed synergistic carcinogenic response. Carcinogenesis 1:1049-1058.
Fanini, D; Legator, MS; Adams, PM. (1984) Effects of paternal ethylene dibromide exposure on
F1 generation behavior in the rat. Mutat Res 139:133-138.
Foster, PL; Wilkinson, WG; Miller, JK; et al. (1988) An analysis of the mutagenicity of
1,2-dibromoethane to Escherichia coli: influence of DNA repair activities and metabolic pathways.
Mutat Res 194:171-181.
Gargas, ML; Burgess, RJ; Voisard, DE; et al. (1989) Partition coefficients of low-molecular-
weight volatile chemicals in various liquids and tissues. Toxicol Appl Pharmacol 98:87-99.
Ghanayem, BI, Maronpot, RR, Matthews, HB. (1986) Association of chemically induced
forestomach cell proliferation and carcinogenesis. Cancer Lett 32:271-278.
Graves, RJ; Trueman, P; Jones, S; et al. (1996) DNA sequence analysis of methylene chloride-
induced HPRT mutations in Chinese hamster ovary cells: comparison with the mutation spectrum
obtained for 1,2-dibromoethane and formaldehyde. Mutagenesis 11:229-233.
Grice, HC. (1988) Safety evaluation of butylated hydroxyanisole from the perspective of effects on
forestomach and oesophageal squamous epithelium. Food Chem Toxicol 26:717-723.
Guengerich, FP; Persmark, M. (1994) Mechanism of formation of ethenoguanine adducts from 2-
haloacetaldehydes: 13C labeling patterns of 2-bromoacetaldehyde. Chem Res Toxicol 7:205-208.
Guengerich, FP; Mason, PS; Stott, WT; et al. (1981) Roles of 2-haloethylene oxides and
2-haloacetaldehydes derived from vinyl bromide and vinyl chloride in irreversible binding to
protein and DNA. Cancer Res 41:4391-4398.
102
-------
Guengerich, FP; Thier, R; Persmark, M; et al. (1995) Conjugation of carcinogens by theta class
glutathione s-transferases: mechanisms and relevance to variations in human risk.
Pharmacogenetics 5:S103-S107.
Hill, DL; Shih, T.-W; Johnston, TP; et al. (1978) Macromolecular binding and metabolism of the
carcinogen 1,2-dibromoethane. Cancer Res 38:2438-2442.
Hissink, AM; Wormhoudt, LW; Sherratt, PJ; et al. (2000) A physiologically-based
pharmacokinetic(PB-PK) model for ethylene dibromide: relevance of extrahepatic metabolism.
Food Chem Toxicol 38:707-716.
Hodgson, E; Levi, PE, eds. (1994) Introduction to biochemical toxicology. 2nd ed. Norwalk, CT:
Appleton & Lange; p. 228.
HSDB (Hazardous Substances Data Bank). (1999) Ethylene dibromide. National Library of
Medicine, Bethesda, MD. Available from: http://toxnet.nlm.nih.gov. [20 October, 2003].
Hsu, LL; Adams, PM; Fanini, D; et al. (1985) Ethylene dibromide: effects of paternal exposure on
the neurotransmitter enzymes in the developing brain of Fl progeny. Mutat Res 147:197-203.
Inskeep, PG; Guengerich, FP. (1984) Glutathione-mediated binding of dibromoalkanes to DNA:
specificity of rat glutathione-S-transferases and dibromoalkane structure. Carcinogenesis
5:805-808.
Inskeep, PB; Koga, N; Cmarik, JL; et al. (1986) Covalent binding of 1,2-dihaloalkanes to DNA
and stability of the major DNA adduct, S-[2-(N7-Guanyl)ethyl]glutathionel. Cancer Res
46:2839-2844.
Ivett, JL; Brown, BM; Rodgers, C; et al. (1989) Chromosomal aberrations and sister chromatid
exchange tests in Chinese hamster ovary cells in vitro. IV. Results with 15 chemicals. Environ
MolMutagen 14:165-187.
Jakobson, I; Wahlberg, JE; Holmberg, B; et al. (1982) Uptake via the blood and elimination of 10
organic solvents following epicutaneous exposure of anesthetized guinea pigs. Toxicol Appl
Pharmacol 63:181-187.
Jean, PA; Reed, DJ.. (1992) Utilization of glutathione during 1,2-dihaloethane metabolism in rat
hepatocytes. Chem Res Toxicol 5:386-391.
Kale, P; Baum, JW. (1979) Sensitivity of Drosophila melanogaster to low concentrations of the
gaseous 1,2-dibromoethane: I. Acute exposures. Environ Mutagen 1:15-18.
Kale, PG; Baum, JW. (1983) Sensitivity of drosophila-melanogaster to low concentrations of
gaseous mutagens. 4. Mutations in embryonic spermatogonia. Mutat Res 113:135-144.
103
-------
Kempkes, M; Wiebel, FA; Golka, K; et al. (1996) Comparative genotyping and phenotyping of
glutathione S-transferase GSTT1. Arch Toxicol 70:306-309.
Khan, S; Sood, C; O'Brien, PJ. (1993) Molecular mechanisms of dibromoalkane cytotoxicity in
isolated rat hepatocytes. BioChem Pharmacol 45:439-447.
Kim, DH; Guengerich, FP. (1989) Excretion of the mercapturic acid
S-[2-(N7-guanyl)ethyl]-N-acetylcysteine in urine following administration of ethylene dibromide
to rats. Cancer Res 49:5843-5847.
Kim, M.-S; Guengerich, FP. (1997) Synthesis of oligonucleotides containing the ethylene
dibromide-derived DNA adducts S-[2-(N7-guanyl)ethyl]glutathione,
S-[2-(N2-guanyl)ethyl]glutathione, S-[2-(O6-guanyl)ethyl]glutathione at a single site. Chem Res
Toxicol 10:1133-1143.
Kim, M.-S; Guengerich, FP. (1998) Polymerase blockage and misincorporation of dNTPs opposite
the ethylene dibromide-derived DNA adducts S-[2-(N7-guanyl)ethyl]glutathione,
S-[2-(N2-guanyl)ethyl]glutathione, S-[2-(O6-guanyl)ethyl]glutathione at a single site. Chem Res
Toxicol 11:311-316.
Kim, DH; Humphreys, WG; Guengerich, FP. (1990) Characterization of
S-[2-(Nl-adenyl)ethyl]glutathione as an adduct formed in RNA and DNA from
1,2-dibromoethane. Chem Res Toxicol 3:587-594.
Kirby, KW; Tremmel, H; Keiser, JE. (1980) Determination and metabolism of ethylene dibromide
in minipigs. Bull Environ Contam Toxicol 24:774-777.
Kitchin, KT; Brown, JL. (1986) 1,2-Dibromoethane causes rat hepatic DNA damage at low doses.
BioChem Biophys Res Commun 141:723-727.
Kitchin, KT and Brown, JL. (1987) Biochemical effects of two promoters of hepatocarcinogenesis
in rats. Food Chem Toxicol. 8:603-607.
Koga, N; Inskeep, PB; Harris, TM; et al. (1986) S-[2-(N7-guanyl)ethyl]glutathione, the major
DNA adduct formed from 1,2-dibromoethane. Biochemistry 25:2192-2198.
Kowalski, B; Brittebo, EB; Brandt, I. (1985) Epithelial binding of 1,2-dibromoethane in the
respiratory and upper alimentary tracts of mice and rats. Cancer Res 45:2616-2625.
Kowalski, B; Brittebo, EV; D'Argy, R; et al. (1986) Fetal epithelial binding of 1,2-dibromoethane
in mice. Carcinogenesis 7:1709-1714.
Krewski, D; Crump, KS; Farmer, J; et al. (1983) A comparison of statistical methods for low dose
extrapolation utilizing time-to-tumour data. Fundam Appl Toxicol 3:140-160.
104
-------
Krishna, G; Xu, J; Nath, J; et al. (1985) In vivo cytogenetic studies on mice exposed to ethylene
dibromide. Mutat Res 158:81-87.
Ledda-Columbano, GM; Columbano, A; Coni, P; et al. (1987a) Liver cell proliferation induced by
the mitogen ethylene dibromide, unlike compensatory cell proliferation, does not achieve initiation
of rat liver carcinogenesis by diethylnitrosamine. Cancer Lett 36:247-252.
Ledda-Columbano, GM; Columbano, A; Coni, P; et al. (1987b) Cell proliferation in rat kidney
induced by 1,2-dibromoethane. Toxicol Lett 37:85-90.
Letz, GA; Pond, SM; Osterloh, JD; et al. (1984) Two fatalities after acute occupational exposure to
ethylene dibromide. JAMA 252:2428-2431.
Ma, T.-H; Sparrow, AH; Schairer, LA; et al. (1978) Effect of 1,2-dibromoethane (DBE) on meiotic
chromosomes of tradescantia. Mutat Res 58:251-258.
Mann, AM; Darby, FJ. (1985) Effects of 1,2-dibromoethane on glutathione metabolism in rat liver
and kidney. BioChem Pharmacol 34:2827-2830.
Maronpot, RR; Shimkin, MB; Witschi, HP; et al. (1986) Strain A mouse pulmonary tumor test
results for chemicals previously tested in the National Cancer Institute carcinogenicity tests. J
Natl Cancer Inst 76:1101-1112.
Masini, A; Botti, B; Ceccarelli, D; et al. (1986) Induction of calcium efflux from isolated rat-liver
mitochondria by 1,2-dibromoethane. Biochim Biophys Acta 852:19-24.
Milks, MM; Wilt, SR; Ali, I; et al. (1982) Dibromoethane effects on the induction of
y-glutamyl-transpeptidase positive foci in rat liver. Arch Toxicol 51:27-36.
Mitra, A; Hilbelink, DR; Dwornick, JJ; et al. (1992) Rat hepatic glutathione
S-transferase-mediated embryotoxic bioactivation of ethylene dibromide. Teratology 46:439-446.
Nachtomi, E; Farber, E. (1978) Ethylene dibromide as a mitogen for liver. Lab Invest 38:279-283.
Nachtomi, E; Sarma, DS. (1977) Repair of rat liver DNA in vivo damaged by ethylene dibromide.
BioChem Pharmacol 26:1941-1945.
Nachtomi, E; Alumot, E; Bondi, A. (1966) The metabolism of ethylene dibromide in the rat. I.
Identification of detoxification products in urine. Isr J Chem 4:239-246.
NCI (National Cancer Institute). (1978) Bioassay of 1,2-dibromoethane for possible
carcinogenicity. National Institutes of Health, U.S. Department of Health, Education, and
Welfare, Bethesda, MD; DHEW Publication No. (NIH) 78-1336. Available from: National
Technical Information Service, Springfield, VA; PB-288428.
105
-------
Nelson, HH; Wiencke, JK; Christiani, DC; et al. (1995) Ethnic differences in the prevalence of the
homozygous deleted genotype of glutathione S-transferase theta. Carcinogenesis 16:1243-1245.
Nitschke, KD; Kociba, RJ; Keyes, DG; et al. (1981) A thirteen week repeated inhalation study of
ethylene dibromide in rats. Fundam Appl Toxicol 1:437-442.
Novotna, B; Duverger-Van Bogaert, M. (1994) Role of kidney S9 in the mutagenic properties of
1,2-dibromoethane. Toxicol Lett 74:255-263.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing
the process. Washington, DC: National Academy Press. Available from:
http://books.nap.edu/books/POD115/html/index.html.
NTP (National Toxicology Program). (1982) Carcinogenesis bioassay of 1,2-dibromoethane (CAS
no. 106-93-4) in F344 rats and B6C3F1 mice (inhalation study). National Institutes of Health,
U.S. Department of Health and Human Services, Research Triangle Park, NC; NTP-80-28, NTP
210, NIH Publication No. 82-1766. Available from: National Technical Information Service,
Springfield, VA; PB82-181710.
Olmstead, EV. (1960) Pathological changes in ethylene dibromide poisoning. AMA Arch Ind
Health 21:525-529.
Ott, MG; Scharnweber, HC; Langner, RR. (1980) Mortality experience of 161 employees exposed
to ethylene dibromide in two production units. Br J Ind Med 37:163-168.
Ozawa, N; Guengerich, FP. (1983) Evidence for formation of an
S-[2-(N7-guanyl)ethyl]glutathione adduct in glutathione-mediated binding of the carcinogen
1,2-dibromoethane to DNA. Proc Natl Acad Sci 80:5266-5270.
Pemble, S; Schroeder, KR; Spencer, SR; et al. (1994) Human glutathione S-transferase theta
(GSTT1): cDNA cloning and the characterization of a genetic polymorphism. BioChem J
300:271-276.
Peterson, LA; Harris, TM; Guengerich, FP. (1988) Evidence for an episulfonium ion intermediate
in the formation of S-[2-(N7-guanyl)ethyl]glutathione in DNA. J Am Chem Soc 110:3284-3291.
Ploemen, JHTM; Wormhoudt, LW; Van Ommen, B; et al. (1995) Polymorphism in the glutathione
conjugation activity of human erythrocytes towards ethylene dibromide and
l,2-epoxy-3-(p-nitrophenoxy)-propane. Biochim Biophys Acta 1243:469-476.
Ploemen, J.-PHTM; Wormhoudt, LW; Haenen, GRMM; et al. (1997) The use of human in vitro
metabolic parameters to explore the risk assessment of hazardous compounds: the case of ethylene
dibromide. Toxicol Appl Pharmacol 143:56-69.
106
-------
Plotnick, HB; Weigel, WW; Richards, DE; et al. (1979) The effect of dietary disulfiram upon the
tissue distribution and excretion of 14C-l,2-dibromoethane in the rat. Res Commun Chem Pathol
Pharmacol 26:535-545.
Poet, TS; Soelberg, JJ; Weitz, KK; et al. (2003) Mode of action and pharmacokinetic studies of
2-butoxyethanol in the mouse with an emphasis on forestomach dosimetry. Toxicol Sci
71:176-189.
Portier, C; Hedges, JC; Hoel, DG. (1986) Age-specific models of mortality and tumor onset for
historical control animals in the National Toxicology Program's carcinogenicity experiments.
Cancer Res 46:4372-4378.
Ratcliffe, JM; Schrader, SM; Steenland, K; et al. (1987) Semen quality in papaya workers with
long term exposure to ethylene dibromide. Br J Ind Med 44:317-326.
Reznik, G; Stinson, SF; Ward, JM. (1980) Respiratory pathology in rats and mice after inhalation
of l,2-dibromo-3-chloropropane or 1,2 dibromoethane for 13 weeks. Arch Toxicol 46:233-240.
Roth, WL; Freeman, RA; Wilson, AGE. (1993) A physiologically based model for gastrointestinal
absorption and excretion of chemicals carried by lipids. Risk Anal 13:531-543.
Rowe, VK; Spencer, HC; McCollister, DD; et al. (1952) Toxicity of ethylene dibromide
determined on experimental animals. AMA Arch Ind Hyg Occup Med 6:158-173.
Saraswat, PK; Kandara, M; Dhruva, AK; et al. (1986) Poisoning by ethylene di-bromide - six
cases: a clinicopathological and toxicological study. Indian J Med Sci 40:121-123.
Sasaki, YF; Saga, A; Akasaka, M; et al. (1998) Detection of in vivo genotoxicity of haloalkanes
and haloalkenes carcinogenic to rodents by the alkaline single cell gel electrophoresis (comet)
assay in multiple mouse organs. Mutat Res 419:13-20.
Schrader, SM; Turner, TW; Ratcliffe, JM. (1988) The effects of ethylene dibromide on semen
quality: a comparison of short-term and chronic exposure. Reprod Toxicol 2:191-198.
Seow, A; Yuan, J.-M; Sun, C.-L; et al. (2002) Dietary isothiocyanates, glutathione s-transferase
polymorphisms and colorectal cancer risk in the Singapore Chinese Health Study. Carcinogenesis
23:2055-2061.
Short, RD; Minor, JL; Winston, JM; et al. (1978) Inhalation of ethylene dibromide during
gestation by rats and mice. Toxicol Appl Pharmacol 46:173-182.
Short, RD; Winston, JM; Hong, C.-B; et al. (1979) Effects of ethylene dibromide on reproduction
in male and female rats. Toxicol Appl Pharmacol 49:97-105.
107
-------
Simula, TP; Glancey, MJ; Wolf, CR. (1993) Human glutathione S-transferase-expressing
Salmonella typhimurium tester strains to study the activation/detoxification of mutagenic
compounds: studies with halogenated compounds, aromatic amines and aflatoxin Bl.
Carcinogenesis 14:1371-1376.
Smith, RF; Goldman, L. (1983) Behavioral effects of prenatal exposure to ethylene dibromide.
Neurobehav Toxicol Teratol 5:579-585.
Steenland, K; Carrano, A; Clapp, D; et al. (1985) Cytogenetic studies in humans after short-term
exposure to ethylene dibromide. J Occup Med 27:729-732.
Steenland, K; Carrano, A; Ratcliffe, J; et al. (1986) A cytogenetic study of papaya workers
exposed to ethylene dibromide. MutatRes 170:151-160.
Stinson, SF; Reznik, G; Ward, JM. (1981) Characteristics of proliferative lesions in the nasal
epithelium of mice following chronic inhalation of 1,2-dibromoethane. Cancer Lett 12:121-129.
Stoner, GD; Conran, PB; Greisiger, EA; et al. (1986). Comparison of two routes of chemical
administration on the lung adenoma response in strain A/J mice. Toxicol Appl Pharmacol
82:19-31.
Storer, RD; Conolly, RB. (1983) Comparative in vivo genotoxicity and acute hepatotoxicity of 1,2-
dihaloethanes. Carcinogenesis 4:1491-1494.
Stott, WT; McKenna, MJ. (1984) The comparative absorption and excretion of chemical vapors by
the upper, lower, and intact respiratory tract of rats. Fundam Appl Toxicol 4:594-602.
Sundheimer, DW; White, RD; Brendel, K; et al. (1982) The bioactivation of 1,2-dibromoethane in
rat hepatocytes: covalent binding to nucleic acids. Carcinogenesis 3:1129-1133.
Sweeney, MH; Beaumont, JJ; Waxweiler, RJ; et al. (1986). An investigation of mortality from
cancer and other causes of death among workers employed at an east Texas chemical plant. Arch
Environ. Health 41:23-28.
Takahashi, W; Wong, L; Rogers, BJ; et al. (1981) Depression of sperm counts among agricultural
workers exposed to dibromochloropropane and ethylene dibromide. Bull Environ Contam Toxicol
27:551-558.
Tan, E.-L; Hsie, AW. (1981) Mutagenicity and cytotoxicity of haloethanes as studied in the
CHO/HGPRT system. MutatRes 90:183-191.
Thier, R; Pemble, SE; Kramer, H; et al. (1996) Human glutathione S-tranferase Tl-1 enhances
mutagenicity of 1,2-dibromethane, dibromomethane and 1,2,3,4-diepoxybutane in Salmonella
typhimurium. Carcinogenesis 17:163-166.
108
-------
Tomasi, A; Albano, E; Dianzani, MU; et al. (1983) Metabolic activation of 1,2-dibromoethane to a
free radical intermediate by rat liver microsomes and isolated hepatocytes. FEES Lett 160:191-
194.
Tucker, JD; Xu, J; Stewart, J; et al. (1984) Detection of sister-chromatid exchanges in human
peripheral lymphocytes indue 1,2-dibromomethane ethylene dibromide vapor. MutatRes 138:93-
98.
Turner, D; Barry, PSI. (1979) An epidemiological study of workers in plants manufacturing
ethylene dibromide. Arh hig rada toksikol 30(suppl.): 621-626.
U.S. EPA (Environmental Protection Agency). (1986) Guidelines for carcinogen risk assessment.
Fed Regist 51(185):33992-34003.
U.S. EPA. (1986) Guidelines for mutagenicity risk assessment. Fed Regist 51(185):34006-34012
and http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (1986) Guidelines for the health risk assessment of chemical mixtures. Fed Regist
51(185):34014-34025 and http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (1988a) Recommendations for and documentation of biological values for use in risk
assessment. Environmental Criteria and Assessment Office, Office of Health and Environmental
Assessment, Office of Research and Development, Cincinnati, OH; EPA/600/6-87/008. Available
from: National Technical Information Service, Springfield, VA; PB-88179874.
U.S. EPA. (1988b) IRIS Summary of 1,2-dibromomethane. Integrated Risk Information System
(IRIS). National Center for Environmental Assessment, Washington, DC. Available from:
http ://www. epa.gov/iris.
U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Fed Regist
56(234):63798-63826 and http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (1992) Draft report: a cross-species scaling factor for carcinogen risk assessment based
on equivalence of mg/kg3/4/day. Fed Reg 57(109):24152-24174.
U.S. EPA. (1994a) Methods for derivation of inhalation reference concentrations and application
of inhalation dosimetry. Office of Environmental Health Assessment, Washington, DC;
EPA/600/8-90/066F. Available from: http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (1994b) Interim policy for particle size and limit concentration issues in inhalation
toxicity: notice of availability. Fed Regist 59(206):53799 and
http://www.epa.gov/iris/backgrd.htm.
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. Risk
Assessment Forum, Washington, DC; EPA/630/R-94/007. Available from:
http://www.epa.gov/iris/backgr-d.htm.
109
-------
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Fed Regist
61(212):56274-56322 and http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (1998a) Guidelines for neurotoxicity risk assessment. Fed Regist 63(93):26926-26954
and http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (1998b) Science policy council handbook: peer review. Office of Science Policy for
the Office of Research and Development, Washington, DC; EPA/100/B-98/001. Available from:
http: //www. epa. gov/cl ariton/clhtml/pubtitl eOther. html.
U.S. EPA. (1999) Guidelines for carcinogen risk assessment [review draft]. Washington, DC:
Risk Assessment Forum. NCEA-F-0644. Available from: http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (2000a) Science policy council handbook: peer review [second edition]. Office of
Science Policy for the Office of Research and Development, Washington, DC; EPA/1 OO/B-00/001.
Available from: http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (2000b) Science policy council handbook: risk characterization. Office of Science
Policy for the Office of Research and Development, Washington, DC; EPA/100/B-00/002.
Available from: http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (2000c) Benchmark dose technical support document [external review draft]. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/001. Available from:
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=42601.
U.S. EPA. (2000d) Supplemental guidance for conducting for health risk assessment of chemical
mixtures. EPA/630/R-00/002. Available from: http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (2000e) Status of Chemicals in Special Review. Washington, DC: Office of
Prevention, Pesticides and Toxic Substances, Washington, DC; EPA/738/R-00/001. Available
from: http://www.epa.gov/oppsrrdl/docs/srOOstatus.pdf
U.S. EPA. (2001) IRIS Summary of bromate. Integrated Risk Information System (IRIS).
National Center for Environmental Assessment, Washington, DC. Available from:
http ://www. epa. gov/iris.
U.S. EPA. (2002a) A review of the reference dose concentration and reference concentration
processes [final report]. Risk Assessment Forum, Washington, DC; EPA/630/P-02/0002F.
Available from: http://www.epa.gov/iris/backgr-d.htm.
U.S. EPA. (2002b) Health assessment of 1,3-butadiene. National Center for Environmental
Assessment, Washington, DC; EPA/600/P-98/001F. Available from:
http://www.epa.gov/iris/supdocs/buta-sup.pdf.
Van Bladeren, PJ; Breimer, DD; Rotteveel-Smijs, GM; et al. (1980) The role of glutathione
conjugation in the mutagenicity of 1,2-dibromoethane. BioChem Pharmacol 29:2975-2982.
110
-------
Van Bladeren, PJ; Hoogeterp, JJ; Breimer, DD; et al. (1981) The influence of disulfiram and other
inhibitors of oxidative metabolism on the formation of 2-hydroxyethyl-mercapturic acid from
1,2-dibromoethane by the rat. BioChem Pharmacol 30:2983-2987.
Van Bladeren, PJ; Armstrong, RN; Cobb, D; et al. (1982) Stereo selective formation of
benz(a)anthracene dextro-(5S,6R)-oxide and benz(a)anthracene dextro-(8R,9S)-oxide by a highly
purified and reconstituted system containing cytochrome P-450c. BioChem Biophys Res Commun
106:602-609.
Van Duuren, BL; Goldschmidt, BM; Loewengart, G; et al. (1979) Carcinogenicity of halogenated
olefinic and aliphatic hydrocarbons in mice. J Natl Cancer Inst 63:1433-1439.
Van Duuren, BL; Seidman, I; Melchionne, S; et al. (1985) Carcinogenicity bioassays of
bromoacetaldehyde and bromoethanol—potential metabolites of dibromoethane. Teratog Carcinog
Mutagen 5:393-403.
Warholm, M; Rane, A; Alexandrie, A-K; et al. (1995) Genotypic and phenotypic determination of
polymorphic glutathione transferase Tl in a Swedish population. Pharmacogenetics 5:252-254.
Wheeler, JB; Stourman, NV; Armstrong, RN; et al. (2001) Conjugation of haloalkanes by bacterial
and mammalian glutathione transferases: mono- and vicinal dihaloethanes. Chem Res Toxicol
14:1107-1127.
Williams, J; Gladen, BC; Turner, TW; et al. (1991) The effects of ethylene dibromide on semen
quality and fertility in the rabbit: evaluation of a model for human seminal characteristics.
Fundam Appl Toxicol 16:687-700.
Wong, O; Utidjian, HMD; Karten, VS. (1979) Retrospective evaluation of reproductive
performance of workers exposed to ethylene dibromide (EDB). J Occup Med 21:98-102.
Wong, LCK; Winston, JM; Hong, CB; et al. (1982) Carcinogenicity and toxicity of
1,2-dibromoethane in the rat. Toxicol Appl Pharmacol 63:155-165.
Working, PK; Smith-Oliver, T; White, RD; et al. (1986) Induction of DNA repair in rat
spermatocytes and hepatocytes by 1,2-dibromoethane: the role of glutathione conjugation.
Carcinogenesis 7:467-72.
Wormhoudt, LW; Ploemen, JH; Commandeur, JN; et al. (1996a) Cytochrome P450 catalyzed
metabolism of 1,2-dibromoethane in liver microsomes of differentially induced rats. Chem Biol
Interact 99:41-53.
Wormhoudt, LW; Ploemen, JH; De Waziers, I; et al. (1996b) Inter-individual variability in the
oxidation of 1,2-dibromoethane: use of heterologously expressed human cytochrome P450 and
human liver microsomes. Chem Biol Interact 101:175-192.
Ill
-------
Wormhoudt, LW; Commandeur, JN; Ploemen, JH; et al. (1997) Urinary thiodiacetic acid. A
selective biomarker for the cytochrome P450-catalyzed oxidation of 1,2-dibromoethane in the rat.
Drug Metab Dispos 25:508-515.
Wormhoudt, LW; Hissink, AM; Commandeur, JN; et al. (1998) Disposition of 1,2-
[14C]Dibromoethane in male Wistar rats. Drug Metab Dispos 26:437-447.
112
-------
APPENDIX A. External Peer Review-Summary of Comments and Disposition
The Toxicological Review that supports the IRIS file for 1,2-dibromoethane has
undergone both internal peer review performed by scientists within EPA or other federal
agencies and a more formal external peer review performed by scientists chosen by EPA's
contractor in accordance with U.S. EPA (1994c). Comments made by the internal reviewers
were addressed prior to submitting the documents for external peer review and are not part of
this appendix. Public comments also were read and carefully considered. The three external
peer reviewers were tasked with providing written answers to general questions on the overall
assessment and on chemical-specific questions in areas of scientific controversy or uncertainty.
A summary of comments made by the external reviewers and EPA's response to these comments
follows.
General Charge Question 1. Are you aware of any other data/studies that are relevant (i.e.,
useful for the hazard identification or dose-response assessment) for the assessment of the
adverse health effects, both cancer andnoncancer, of this chemical?
Comment: Reviewers identified specific studies (see below) addressing carcinogenicity that
they felt had not been adequately considered in the assessment.
A/J mice studies and other secondary oncogenicity studies: Stoner et al. (1986), Adkins et al
(1986), and Van Duuren et al. (1979) studies should be considered.
Response: We agree that the studies identified [Stoner et al. (1986), Adkins et al (1986), Van
Duuren et al. (1979)] are relevant. They have been summarized and considered in the weight of
evidence.
Chronic bioassay: While the Van Duuren (1985) study was described in the assessment, one
reviewer felt it had not been adequately considered in the dose-response assessment.
Response: We agree with the comment, and have more thoroughly characterized the consistency
of this drinking water study with the oral gavage study. In the case of the mouse oral gavage
study, the change in dosing regimen did not result in two similar dose levels, however, as with
the rats. Drinking water is a more realistic exposure medium, though, so, although this study
used much higher dose levels and reported less complete tumor incidence data (no times of
observation and no information concerning which animals had multiple tumors), it should be
considered more thoroughly.
Epidemiology studies: Sweeney et al. (1986) and Turner and Barry (1979) should be
considered.
Response: EPA agrees that Sweeney et al. (1986) should be included in the document and it is
included in Section 4 with a qualitative discussion of the limitations of the study and its use in
the quantitative assessment of the carcinogenicity of this chemical. There are several limitations
A-l
-------
to this study. Levels of 1,2-dibromoethane were not reported, except that they were below
OSHA standards (20 ppm) and NIOSH recommended standards. 1,2-Dibromoethane was
apparently added to tetraethyl lead to reduce the deposition of inorganic lead on engine parts and
was used continuously because of the chemical plant's start-up, so workers were exposed from
the beginning of the study period, however. Another limitation of the study was lack of a more
comparable control group, unexposed workers. This study neither establishes nor dismisses an
association between occupational 1,2-dibromoethane exposure and carcinogenicity.
Turner and Barry (1979) came to the attention of one reviewer as a secondary citation. As
indicated by the same reviewer, this study had major limitations, including lack of information
concerning duration and magnitude of exposure, number of workers studied, and length of time
elapsing since first exposure, which make the study inconclusive and of little or no value to this
assessment. However, its contributions and limitations have been summarized.
PBPK studies: Hissink et al. (2000) and Ploemen et al. (1997) should be considered.
Response: EPA agrees and the studies identified [Hissink et al. (2000) and Ploemen et al.
(1997)] have been summarized and considered in the document for their relevance in
characterizing human relevance and potential use for route-to-route extrapolation and for the
estimation of human equivalent concentrations from laboratory animal studies. Given the
limitations of these studies, however (see Section 3.5), it would not be appropriate to use the
Hissink et al. (2000) model at this time for quantitative (route-to-route or animal-to-human)
extrapolations. However, the Hissink et al. (2000) report and model provide useful information
regarding the mode of action of 1,2-dibromoethane, particularly with respect to the cancer
effects of 1,2-dibromoethane.
Comment: One reviewer suggested that EPA should consider requiring additional studies of
(presumably pesticide) registrants to characterize the dose-response for developmental
neurotoxicity (adverse behavioral effects) and the endocrine disruptive potential of 1,2-
dibromoethane.
Response: Additional studies in these areas may help to improve the certainty of the RfC and
RfD. The extent to which additional studies in these areas would benefit the 1,2-dibromoethane
risk assessment is addressed in Section 6 of the EDB IRIS toxicological review document.
General Charge Question 2. For the RfD and RfC, has the most appropriate critical effect been
chosen (i.e., that adverse effect appear ing first in a dose-response continuum)? For the cancer
assessment, are the tumors observed biologically significant? relevant to human health? Points
relevant to this determination include whether or not the choice follows from the dose-response
assessment, whether the effect is considered adverse, and if the effect (including tumors observed
in the cancer assessment) and the species in which it is observed is a valid model for humans.
Comment: With respect to the RfD, one reviewer questioned whether all co-critical effects were
identified and discussed. Another reviewer did not feel that the Agency had adequately justified
A-2
-------
not using a benchmark dose approach on the oral data.
Response: A discussion of co-critical endpoints, including liver, testicular, and adrenal cortical
effects has been added to Section 5.1. A benchmark dose analysis of these data was performed
and is discussed in Section 5.1.1. Details are included in Appendix B (Table B-l and associate
output files) of the toxicological review. The RfD estimates that result from the BMD (5E-3
mg/kg/day) and NOAEL/LOAEL (9E-3 mg/kg/day) assessment approaches are very similar.
Because of high mortality and unusual dosing in the high dose group, there is a great deal of
uncertainty associated with these BMD results and they are only provided in support of the
NOAEL/LOAEL approach.
Comment: With respect to the RfC, one reviewer suggested that all co-critical effects be
considered and analyzed via the benchmark dose approach.
Response: A benchmark dose analysis of all co-critical endpoints, including liver, testicular,
adrenal cortical, splenic, and nasal effects, has been added to Section 5.2. The details of this
analysis are provided in Appendix B of the toxicological review.
Comment: One reviewer pointed out that 1,2-dibromoethane is most likely a Category 2 gas,
and the portal of entry effects should be reevaluated accordingly.
Response: 1,2-Dibromoethane is relatively insoluble in water and demonstrates systemic
toxicity, although there was some respiratory tract toxicity that may or may not be considered
portal of entry effects. EPA agrees that dosimetry for 1,2-dibromoethane's effects at points of
contact should take into account EPA's RfC Methodology (U.S.EPA, 1994b) for Category 2
gases. Following the RfC Methodology, the respiratory tract effects were assessed alternatively
as portal of entry effects or systemic effects while the rest of the effects were assessed as
systemic effects. This approach has been incorporated into the assessment (e.g., Table 5-4 of the
toxicological review and Table 1 of the summary document and associate descriptive text), and
the new human equivalent concentration (HEC) estimates have resulted in the critical effect for
the RfC changing from liver necrosis to nasal inflammation.
Comment: Two reviewers requested a better rationale for the use of forestomach tumors in the
cancer risk assessment.
Response: The Agency believes that forestomach tumors caused by genotoxic carcinogens such
as 1,2-dibromoethane are more applicable to humans than forestomach tumors caused by
nongenotoxic mechanisms that require cellular damage, turnover, and a long residence time of
the chemical in the forestomach. Additional rationale/explanation of the use of forestomach
tumors for the 1,2-dibromoethane assessment and their applicability to humans has been added
to Sections 4.6, 6, and other sections of the toxicological review.
General Charge Question 3. Have the noncancer and cancer assessments been based on the
most appropriate studies? These studies should present the critical effect/cancer (tumors or
A-3
-------
appropriate precursor) in the clearest dose-response relationship. If not, what other study (or
studies) should be chosen and why?
Comment: All reviewers felt that the appropriate studies were chosen.
General Charge Question 4. In the IRIS Summary document, studies included in the RfD and
RfC under the heading "Supporting/Additional Studies " are meant to lend scientific justification
for the designation of critical effect by including any relevant pathogenesis in humans, any
applicable mechanistic information, any evidence corroborative of the critical effect, or to
establish the comprehensiveness of the database with respect to various endpoints (such as
reproductive/developmental toxicity studies). Should other studies be included under the
"Supporting/Additional" category? Should some studies be removed?
Comment: One reviewer felt that the summary document was too short and needed more detail
to support the choice of study, species, sex, and critical effect for both the RfC and RfD.
Response: Additional detail and supporting documentation has been added to the summary
document.
General Charge Question 5. For the noncancer assessments, are there other data that should
be considered in developing the uncertainty factors or the modifying factor? Do the data
support the use of different values than those proposed?
Comment: For the RfD, the reviewers agreed with the uncertainty factors that were applied, but
one reviewer disagreed with the rational for the 3-fold database uncertainty factor. This
reviewer pointed out that the lack of oral reproductive and developmental studies should be
stressed as a primary basis for this UF.
Response: The Agency agrees that the rationale for the RfD database UF should include the lack
of oral reproductive and developmental toxicity studies, and it has been revised in both Chapter 5
of the toxicological review and Section I. A.3. of the IRIS summary document to take into
account the reviewer's comments. In fact, the RfD database UF has been increased to 10 to
account for this uncertainty and the problems with the critical study associated with the high
mortality and odd dosing regimen of the high-dose group.
Comment: For the RfC, two reviewers questioned the lack of a database UF and one reviewer
questioned the rationale used for the application of an effect level extrapolation factor (ELF,
comparable to a LOAEL to NOAEL UF) to the BMDL10 point of departure.
Response: In the external review draft of this assessment, the Agency did not apply a database
UF, but did apply an ELF. After considering the reviewer comments, the Agency agrees that a
database uncertainty factor is warranted due to the lack of any multigenerational reproductive
toxicity studies given the genotoxicity of this compound and the spermatotoxicity concerns.
Given that the merits and approach for the application of an ELF are still under discussion within
A-4
-------
the Agency and the point of departure in this case is the lower bound 95% confidence limit on
the dose that causes a 10% increase in a minimal effect, the Agency has determined that the
application of an ELF is not warranted in this case. The overall UF remains 300 but is now
based on 10 for UFH, 3 for UFA and 10 for UFD.
General Charge Question 6. Do the confidence statements andweight-of-evidence statements
present a clear rationale and accurately reflect the utility of the studies chosen, the relevancy of
the effects (cancer and noncancer) to humans, and the comprehensiveness of the database? Do
these statements make sufficiently apparent all the underlying assumptions and limitations of
these assessments? If not, what needs to be added?
Comment: Two reviewers felt that the confidence statements should be improved for both the
RfD and RfC and that lower confidence should be placed on the RfD. One reviewer felt that the
confidence discussions should be included in Section 6 of the toxicological review.
Response: The Agency generally agreed with the reviewer comments, and the confidence
statements have been revised and expanded accordingly.
Specific Charge Question 1. Regarding the RfD and RfC
(a) Is the Agency justified in not making use of the bull studies of sperm effects to
quantify noncancer risk?
Comment: All reviewers agreed that the bull studies should not be used to quantify human risk,
citing the small sample size and differences in physiology (including absorption and metabolism)
between bulls and humans. One reviewer felt that the latter reason was more valid than the
former because of the consistent adverse response observed in these studies.
Response: The text in Sections 4.3.2 and 6 has been modified to emphasize the physiological
differences between bulls and humans as the primary reason these studies were not used as the
basis for the RfD.
(b) Is liver necrosis an appropriate noncancer endpointfor derivation of an RfC?
Comment: All reviewers felt that liver necrosis was an appropriate endpoint, but one reviewer
suggested that the nasal inflammation endpoint in mice was also appropriate and might result in
a lower value if the RfC were derived using the method described in EPA (1994b) for calculating
FIEC values for category 1 gases.
Response: The current EPA practice is to perform "endpoint specific" HEC calculations. In
other words, HEC values are calculated using EPA (1994b) methods for category 1 gases for
portal-of-entry effects and category 3 methods for systemic effects. The HEC for nasal
inflammation effects was originally calculated using the method described for category 3 gases.
When the HEC was recalculated using the method for category 1 gases, the resultant HEC value
A-5
-------
was lower than the liver necrosis HEC value and was thereby used to derive the revised RfC of
9E-3 mg/m3.
Specific Charge Question 2. Regarding the cancer assessment
(a) Were the adjustments made to account for early tumor formation and mortality
adequate and clear?
Comment: Two reviewers noted that the cancer dose-response modeling was carried out using a
nonstandard EPA method.
Response: An interim approach that approximated the multistage-Weibull model has been
deleted. The cancer dose-response modeling was carried out using the multistage-Weibull model
for time-to-tumor. These results are presented in parallel with the results from the survival-
adjusted tumor incidences. Upon further consideration, the multistage-Weibull model appears
not to be suitable for use with the oral data because of the unusual dosing schedule that was
used; the multistage-Weibull model currently available can only handle a single level of
exposure per group when dose rate is an issue. The multistage-Weibull is still superior to the
multistage model when survival has been differentially impacted, leading to competing causes of
death confounding the response patterns, and when tumors occur earlier with increasing
exposure, as was stated in the document. A more transparent explanation supporting the use of
the multistage-Weibull model has been added. The output from the multistage model for all
cancer dose-response analyses has also been added.
(b) Was it appropriate to attempt a low dose extrapolation from the high tumor incidence
estimates obtained after the lifetime extrapolation approach?
Comment: All reviewers agreed that low-dose extrapolation was appropriate and that the linear
approach should be used.
(c) Was the use offorestomach tumors adequately justified given the absence of this
organ in humans?
Comment: Two reviewers agreed that the use offorestomach tumors was appropriate. One
reviewer indicated that use offorestomach tumors was questionable because the NCI study was a
gavage study and the tumors could have been by [long-term contact and] irritation that would not
be relevant to humans.
Response: The rationale for using the forestomach tumors has been enhanced in the text of the
toxicological review in Sections 4.6, 5, and 6. Because EDB is genotoxic, the occurrence of
these tumors may not depend on irritation and cell damage caused by long retention times in the
forestomach compartment of rodents (an argument that has been made for lack of relevancy to
humans). Thus, it is felt that these tumors must be considered. Some expert consultation had led
to the conclusion that the forestomach tumors should be treated as incidental since it was not
A-6
-------
clear that death had been caused by the tumor and not some other more acute lesion. After more
review, it appears that the forestomach tumors should be treated as the most likely cause of death
for all animals who had them in the oral study, except those whose tumors were only detected at
the terminal sacrifice. The other tumor types in the oral study have now been taken to be
incidental, although in the few early deaths where there were no forestomach tumors some of
these may have been the cause of death.
The discussion of forestomach tumors in the weight-of-evidence section (Section 4.6) has been
expanded to include the following discussion:
The relevance of forestomach effects to humans has been questioned, particularly
for nongenotoxic chemicals whose mode of action is believed to involve irritation
and cell proliferation from long-term exposure (Poet et al., 2003). It is true that
the forestomach is not present in humans and contains features, such as minimal
vascularization and stratified squamous cells, that result in a longer residence time
of food-borne agents than is received by comparable human organs such as the
oesophagus and the glandular stomach (Grice, 1988; Poet et al., 2003). In this
case, however, effects in this organ are believed to be of potential relevance to
humans because 1,2-dibromoethane and other genotoxic chemicals do not appear
to require precursor events (e.g., irritation) associated with long residence time to
induce these kinds of tumors. 1,2-Dibromoethane does not appear to cause
significant irritation in the forestomach and was reported to induce forestomach
tumors after just 168 days of exposure (NCI, 1978).
Other
Comment: In several areas of the noncancer and cancer portions of the assessment, reviewers
noted instances of tables being mislabeled or missing. They suggested adding tables or adding
information to existing tables in order to summarize study information (study characteristics,
results, statistical significance). Inconsistent interpretation of observed effects was also noted.
Numerous instances were identified where additional study information would be useful.
Response: Inconsistencies in the document have been corrected, and the requested information
has been added to the text and in additional tables.
A-7
-------
APPENDIX B. BMDS Analyses of Noncancer Endpoints
Table B-l. BMDS analysis of NCI (1978) in support of RFD
Model
Multistage Power
BMD10
BMDL10
AIC
p-value
Wabutt BMD10
BMDL10
AIC
p-value
Gamma BMD10
BMDL10
AIC
p-value
Logistic BMDi0
BMDL10
AIC
p-value
LogLogist BMDi0
BMDL10
AIC
p-value
Probit BMD10
BMDL10
AIC
p-value
LogProbit BMD10
BMDL10
AIC
p-value
First occurrence
of effect (weeks)"
BMDL10(mg/m3)
Peliosis in male rats
(see pp. B-2 to B-4)a
No dose
adjustment
1
17.0924
11.9529
80.9598
0.9119
17.0924
11.9529
80.9598
0.9119
17.0924
11.9529
80.9598
0.9119
34.7812
26.6526
86.6093
0.1038
14.89
9.59883
80.8128
0.9820
32.6319
25.0429
86.1626
0.1193
26.0742
19.937
81.8126
0.5923
TWA dose
adjustment
6
32.3459
9.12312
80.7837
0.9964
32.8565
9.12815
82.7764
1.0000
32.6612
9.1141
80.7968
0.9898
32.8668
21.2288
82.7831
0.9521
33.2451
7.65899
82.7764
1.0000
33.1771
19.6937
82.7764
0.9989
33.6301
15.3227
82.7764
1.0000
Low dose - 39
High dose - 36
9.59883
9.12312
Testicular atrophy in male
rats
(see pp. B-5 to B-8)a
No dose
adjustment
1
11.5101
8.68059
106.298
0.9389
11.5101
8.68059
106.298
0.9389
11.5101
8.68059
106.298
0.9389
26.2594
20.6013
113.248
0.0614
9.88164
6.19692
108.173
1.0000
24.7944
19.5476
112.678
0.0727
19.5799
15.4431
106.87
0.7020
TWA dose
adjustment
7
31.3928
6.77996
106.182
0.9958
31.8256
6.78526
108.173
0.9999
30.2683
6.65175
106.403
0.8918
31.9776
18.675
108.175
0.9754
32.5999
5.29211
108.173
1.0000
32.4969
17.0531
108.173
0.9999
33.0559
12.0592
108.173
1.0000
Low dose - 38
High dose -15
8.68059
6.77996
Adrenal cortical
degeneration in male rat
(see pp. B-9 to B-ll)a
No dose
adjustment
1
15.9817
11.4453
92.4547
0.6547
15.9817
11.4453
92.4547
0.6547
15.9817
11.4453
92.4547
0.6547
33.2869
25.5276
98.8978
0.0540
13.6897
9.05438
92.0322
0.8114
31.1452
23.9958
98.4693
0.0604
24.5429
19.063
93.8791
0.3186
TWA dose
adjustment
3
26.1145
8.46243
91.6386
0.9878
28.6819
8.47521
93.614
1.0000
29.5879
8.47521
93.614
1.0000
30.205
20.0266
93.6949
0.8137
29.3368
6.96311
93.614
1.0000
29.4539
18.6165
93.6245
0.9338
29.9603
14.5032
93.614
1.0000
Low dose -41
High dose - 35
9.05438 | 8.46243
z The output file for the selected model (usually that which resulted in the lowest AIC) is included in this Appendix.
b An attempt was made account for early mortality. In general, rats that died before the first occurrence of the effect
(see Appendix C) were excluded from the analysis of that effect. However, 13 rats that died at 15 weeks without
testicular atrophy were excluded from the analysis even though this effect was observed in 5 rats at that time.
B-l
-------
Output B-l. Peliosis in male rats
Multistage Model with 0.95 Confidence Level
Multistage
0.5
?0.4
o
&
^0.3
c
o
=80.2
0.1
0
BMD.L.
BMP
0 5
14:2602/122004
10 15 20 25
dose
30 35 40 45
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: F:\BMDS\DATA\EDB\MALERATORAL\PMULT.(d)
Gnuplot Plotting File: F:\BMDS\DATA\EDB\MALERATORAL\PMULT.plt
Thu Feb 12 09:44:29 2004
Observation # < parameter # for Multistage model.
The form of the probability function is:
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 7
Total number of specified parameters = 0
Degree of polynomial = 6
B-2
-------
Parameter Estimates
Variable
Background
Beta (1)
Beta(2)
Beta(3)
Beta (4)
Beta(5)
Beta(6)
Estimate
0
0
0
0
0
o
9.19938e-011
Std. Err
NA
NA
NA
NA
NA
NA
3. 90849e-011
Analysis of Deviance Table
Model Log(likelihood) Deviance Test DF P-value
Full model -39.3882
Fitted model -39.3918 0.00729548 2
Reduced model -45.6631 12.5499 2
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 32.3459
BMDL = 9.12312
-------
Output B-2. Testicular atrophy in male rats
Multistage Model with 0.95 Confidence Level
0.7
0.6
Multistage
I0'5
0.4
cO.3
^0.2
0.1
0
BMDL
BMP
0 5 10
17:0502/122004
15 20 25
dose
30 35 40 45
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: F:\BMDS\DATA\EDB\MALERATORAL\TDMULT.(d)
Gnuplot Plotting File: F:\BMDS\DATA\EDB\MALERATORAL\TDMULT.plt
Thu Feb 12 09:32:35 2004
Observation # < parameter # for Multistage model.
The form of the probability function is:
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 8
Total number of specified parameters = 0
Degree of polynomial = 7
B-4
-------
Default Initial Parameter Values
Background = 2.04281e-014
Beta(l) = 0
Beta(2) = 0
Beta(3) = 0
Beta(4) = 0
Beta(5) = 0
Beta(6) = 0
Beta(7) = 0
WARNING: Completion code = -4. Optimum not found. Trying new starting pont****
WARNING 0: Completion code = -4 trying new start****
WARNING 1: Completion code = -4 trying new start****
WARNING 2: Completion code = -4 trying new start****
WARNING 3: Completion code = -4 trying new start****
WARNING 4: Completion code = -4 trying new start****
WARNING 5: Completion code = -4 trying new start****
WARNING 6: Completion code = -4 trying new start****
WARNING 7: Completion code = -4 trying new start****
WARNING 8: Completion code = -4 trying new start****
WARNING 9: Completion code = -4 trying new start****
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background -Beta(l)
-Beta(5) -Beta(6)
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Estimate
Std. Err.
NA
NA
NA
NA
NA
NA
NA
9.6754e-013
Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
AIC:
P-value
B-5
-------
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 31.3928
BMDL = 6.77996
B-6
-------
Output B-3. Adrenal cortical degeneration in male rats
Multistage Model with 0.95 Confidence Level
Multistage
0.5
"80.4
t5
0)
^0.3
••§0.2
co
0
BMDL
BMP
0
17:1802/122004
10 15 20 25
dose
30 35 40 45
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: F:\BMDS\DATA\EDB\MALERATORAL\TDMULT.(d)
Gnuplot Plotting File: F:\BMDS\DATA\EDB\MALERATORAL\TDMULT.plt
Thu Feb 12 09:08:13 2004
BMDS MODEL RUN
P[response] = background + (1-background)*[1-EXP(
-betal*dose"l-beta2*dose'"2-beta3*dose'"3)]
The
parameter betas are restricted to be positive
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 4
Total number of specified parameters = 0
B-7
-------
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0
Beta(3) = 6.04978e-006
Beta(3)
1
Parameter Estimates
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model Log(likelihood) Deviance Test DF P-value
Full model -44.807
Fitted model -44.8193 0.0246029 2
Reduced model -51.1468 12.6797 2
0
13
9
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 26.1145
BMDL = 8.46243
-------
Table B-2. BMDS analysis of NTP (1982) study in support of RfC
development
Model
Multistage Degree
AIC
p-value
BMDju/BMDjs
BMDIVBMDL,;
Weibull AIC
p-value
BMDju/BMDjs
BMDIVBMDL.,5
Gamma AIC
p-value
BMDju/BMDjs
BMDIVBMDL.,5
Logistic AIC
p-value
BMDj
-------
Output B-4. Hepatic necrosis in male rats of NTP (1982)
Probit Model with 0.95 Confidence Level
o
0
o
"o
CD
0.5
0.4
0.3
0.1
Probit
BMDL BMQ
50
100
150
dose
200
250
300
13:0402/162004
Probit Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:53 $
Input Data File: C:\BMDS\DATA\MTBE\CHUNFEMALES\UNSAVED1.(d)
Gnuplot Plotting File: C:\BMDS\DATA\MTBE\CHUNFEMALES\UNSAVEDl.plt
Mon Feb 16 13:05:37 2004
Hepatic Necrosis in Male Rats of NTP (1982)
The form of the probability function is:
P[response] = CumNorm(Intercept+Slope^Dose),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = COLUMNS
Independent variable = COLUMN1
Slope parameter is not restricted
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
B-10
-------
Default Initial (and Specified) Parameter Values
background = 0 Specified
intercept = -1.6355
slope = 0.0044451
intercept
slope
intercept
1
-0.81
Parameter Estimates
Analysis of Deviance Table
Log(likelihood) Deviance Test DF
-59.9467
-60.1512 0.409113 1
-70.709 21.5247 2
Est. Prob.
DF = 1
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 131.774
BMDL = 105.343
B-ll
-------
Output B-5. Testicular degeneration in male rats of NTP (1982)
Log-Logistic Model with 0.95 Confidence Level
T3
0>
0.5
0.4
0.3
I 0.2
0.1
Log-Logistic
BMDL BMD
0
17:4908/052003
50
100
150
dose
200
250
300
Logistic Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:20 $
Input Data File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,2-
dibromoethane\MR_TD_LOGLOGISTIC.(d)
Gnuplot Plotting File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,
dibromoethane\MR_TD_LOGLOGISTIC.plt
Thu Jul 31 08:55:46 2003
Testicular Degeneration in Male Rats of NTP (1982)
The form of the probability function is:
P[response] = background+(1-background)/[l+EXP(-intercept-slope^Log(dose))
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-00<:
Parameter Convergence has been set to: le-008
B-12
-------
Default Initial Parameter Values
background = 0.02
intercept = -6.08038
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -slope
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
background intercept
background 1 -0.39
intercept -0.39 1
Variable
background
intercept
slope
Analysis of Deviance Table
Log(likelihood) Deviance Test DF
-62.141
-62.5988 0.915693 1
-73.4374 22.5929 2
Est. Prob.
DF = 1
1
10
18
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 53.2579
BMDL = 35.0725
B-13
-------
Output B-6. Hepatic necrosis in female rats of NTP (1982)
Probit Model with 0.95 Confidence Level
T3
0
t3
£
<
c
o
t5
LL
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
r Pr/-ihiit ~
: rlODIl :
? ^
? ^
^ ^
: ^-^ -
: ^^-^ :
r ^^ \
~ ^^^ ~
r ^^ ~
: ^^ \
r /^^^ i
; ^^ \
— _— -— ~~~~~^~~ —
1 RMDL BMD !
0 50 100 150 200 250 300
dose
18:0008/052003
Probit Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:53 $
Input Data File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,2-
dibromoethane\FR_HN_LOGPROBIT.(d)
Gnuplot Plotting File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,2-
dibromoethane\FR_HN_LOGPROBIT.pit
Thu Jul 31 10:39:58 2003
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept + Slope*Log(Dose) ) ,
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
B-14
-------
User has chosen the log transformed model
Default Initial (and Specified) Parameter Values
background = 0.04
intercept = -6.39367
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
background intercept
background 1 -0.41
intercept -0.41 1
Variable
background
intercept
slope
Analysis of Deviance Table
Model Log(likelihood) Deviance Test DF
Full model -47.7192
Fitted model -47.7226 0.0069185 1
Reduced model -54.6511 13.8638 2
Dose
Chi-sguare = 0.01 DF = 1
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 172.374
BMDL = 122.02
B-15
-------
Output B-7. Adrenal cortical degeneration in female rats of NTP (1982)
Log-Logistic Model with 0.95 Confidence Level
T3
0
t3
0
§
c
o
CO
^_
LL
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
I rto I rtoiotio
i_og-i_ogisiic
^ ^
? ^
_ _
: _— -~~^~^^ :
: ^^^~~~^~^ '-
r ^^^-^^^^ ;
: ^ !
r ^^^~~~~^^ ~
: ^^~^~^^ :
r ^-^^^^ \
\ ^^^ \
'-- '-_
\ BMDL BMP j
0 50 100 150 200 250 300
dose
15:3408/062003
Logistic Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:20 $
Input Data File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,2-
dibromoethaneX FR_ACD_LOGLOGISTIC. (d)
Gnuplot Plotting File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,2-
dibromoethane\FR_ACD_LOGLOGISTIC.pit
~ ~ Thu Jul 31 10:58:48 2003
The form of the probability function is:
P[response] = background+(1-background)/[l+EXP(-intercept-slope*Log(dose))
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
B-16
-------
Default Initial Parameter Values
background = 0.08
intercept = -6.99302
slope = 1
background intercept
background 1 -0.56
intercept -0.56 1
Parameter Estimates
Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
Analysis of Deviance Table
Log(likelihood) Deviance Test DF
-61.7505
-61.7529 0.00471493 1
-65.2427 6.9844 2
P-value
Chi-sguare =
DF = 1
13
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 124.823
BMDL = 66.495
B-17
-------
Output B-8. Splenic hematopoiesis in female mice of NTP (1982)
0.5
0.4
T3
CD
t> 0.3
I
c
•- 02
t5 U'^
03
0.1
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
BMDL BMD
0
15:5608/062003
50
100
150
dose
200
250
300
Splenic Hematopoiesis in Female Mice of NTP (1982)
The form of the probability function is:
P [response] = background+ (1-background )/ [l + EXP(-intercept-slope*Log(dose) ) '_
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
B-18
-------
Default Initial Parameter Values
background = 0
intercept = -6.2138
slope = 1
intercept
intercept 1
Variable Estimate Std. Err.
background 0 NA
intercept -6.28411 0.244499
slope 1 NA
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model Log(likelihood) Deviance Test DF
Full model -52.7602
Fitted model -53.2053 0.890189 2
Reduced model -65.5992 25.678 2
Scaled
Est._Prob. Expected Observed Size Residual
Chi-sguare = 0.93 DF = 2
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 59.554
BMDL = 40.2456
B-19
-------
Output B-9. Nasal inflammation in female mice of NTP (1982)
Probit Model with 0.95 Confidence Level
T3
0>
0.6
0.5
0.4
0>
§ 0.3
o
'"§ 0.2
0.1
Probit
BMDL
BMD
0
15:2909/032003
50
100
150
dose
200
250
300
Probit Model $Revision: 2.1 $ $Date: 2000/02/26 03:38:53 $
Input Data File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,2-
dibromoethane\FM_NI_LOGPROBIT.(d)
Gnuplot Plotting File: F:\BMDS\DATA\1,2-dibromoethane\INHALATION\l,2-
dibromoethaneXFM_NI_LOGPROBIT.pit
Fri Aug 29 14:46:40 2003
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept + Slope*Log(Dose) ) ,
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
B-20
-------
User has chosen the log transformed model
Default Initial (and Specified) Parameter Values
background = 0
intercept = -5.86426
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
Variable
background
intercept
slope
Estimate
Analysis of Deviance Table
Log(likelihood) Deviance Test DF
-47.5891
-47.8522 0.526361 2
-65.9505 36.7229 2
Est. Prob.
DF = 2
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 102.192
BMDL = 80.1088
B-21
-------
APPENDIX C. Dose-Response Analyses of Cancer Endpoints
C-l. Analyses in support of slope factor for oral exposure
Table C-l. Number of animals with and without specified tumor types at
time of death, male rats orally exposed to 1,2-dibromoethane
Week of
Group death
Control 28
29
49
63
Low-dose 3 1
32
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
High-dose 9
15
24
26
30
31
34
35
36
38
39
40
41
42
44
45
47
49
Forestomach tumors
Death with
tumor
0
0
0
0
1
1
2
2
5
3
0
2
2
5
15
0
o
5
i
0
2
2
1
4
1
3
2
1
8
Death without
tumor
1
1
9
9
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
4
1
15
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hemansiosarcomas
Death with
tumor
0
0
0
0
1
0
2
0
0
0
1
0
0
0
0
0
0
0
0
0
1
6
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
1
0
1
Death without
tumor
1
1
9
9
0
1
0
1
1
1
0
2
1
1
1
6
3
0
2
2
4
13
1
18
1
0
2
2
1
4
0
3
1
1
7
Thyroid
follicular cell
adenoma/carcinoma
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
4
0
1
0
0
0
0
0
1
0
0
0
1
0
0
1
1
0
o
J
Death without
tumor
1
1
9
9
2
2
6
2
0
2
2
5
15
1
17
0
1
2
2
0
4
1
2
1
1
6
Source: NCI, 1978.
C-l
-------
Table C-2. Number of animals with and without specified tumor types
at time of death, female rats orally exposed to 1,2-dibromoethane
Group
Control
Low-dose
High-dose
Week
49
63
1
3
4
5
11
12
24
25
28
34
35
38
39
42
43
44
45
46
48
49
50
51
52
53
54
56
57
59
60
61
1
12
14
15
33
39
40
41
42
45
47
48
50
51
53
54
55
56
58
59
61
Forestomach tumors
Death
with
tumor
0
0
0
0
0
0
0
1
1
1
1
1
1
2
1
1
1
1
2
4
2
4
2
1
2
1
2
2
2
1
1
2
0
1
0
2
1
1
1
1
2
2
1
1
1
1
4
1
1
3
3
1
1
Death
without
tumor
1
19
5
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
1
18
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hemansioarcoma
Death
with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
1
0
Death
without
tumor
1
19
5
1
1
1
1
2
1
1
1
1
1
2
1
1
1
1
2
4
2
4
2
1
2
1
2
2
2
1
1
2
2
1
1
20
1
1
1
1
1
2
1
1
1
1
4
1
1
2
3
0
1
Hepatocellular
carcinomas
Death
with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
1
1
0
0
1
0
0
1
Death
without
tumor
1
19
5
1
1
1
1
2
1
1
1
1
1
2
1
1
1
1
2
4
2
4
2
1
2
1
2
2
2
1
1
1
2
1
1
20
0
1
1
1
2
2
1
1
1
0
3
1
1
2
3
1
0
Adrenocortical
carcinomas
Death
with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
1
0
1
0
Death
without
tumor
1
19
5
1
1
1
1
2
1
1
1
1
1
2
1
1
1
1
2
4
2
4
2
1
2
1
2
2
2
1
1
1
2
1
1
20
1
1
1
0
2
2
1
1
1
0
4
1
1
2
3
0
1
Source: NCI, 1978.
C-2
-------
Table C-3. Number of animals with and without
specified tumor types at time of death, male mice
orally exposed to 1,2-dibromoethane
Group Week
Control 11
36
59
Low-dose 12
19
21
24
26
29
33
36
40
42
43
45
46
48
49
50
52
53
54
56
58
59
60
64
65
66
68
69
72
73
74
78
Forestomach tumors
Death with
tumor
0
0
0
0
0
0
1
0
1
1
2
1
2
2
2
2
1
1
2
3
2
1
1
2
1
1
3
1
2
1
2
3
1
1
2
Death without
tumor
1
1
18
1
1
1
0
1
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Lung
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
1
1
0
0
0
1
0
adenomas
Death without
tumor
18
2
1
2
3
2
2
1
1
2
3
2
:
0
2
3
1
0
2
C-3
-------
Table C-3. Number of animals with and without
specified tumor types at time of death, male mice
orally exposed to 1,2-dibromoethane (continued)
Group Week
High-dose 12
13
14
21
26
27
32
34
38
39
41
42
43
44
45
46
47
48
49
52
53
54
55
57
59
60
61
65
66
68
74
77
Forestomach tumors
Death with
tumor
0
0
0
0
1
0
1
0
0
0
1
0
1
2
2
1
1
2
2
3
2
:
Death without
tumor
4
4
1
1
2
1
0
1
1
2
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
Lung
Death with
tumor
0
0
0
0
1
0
0
0
0
0
1
0
1
0
1
0
0
0
0
2
0
0
0
0
1
1
0
0
1
0
0
1
adenomas
Death without
tumor
4
4
1
1
2
1
1
1
1
2
0
1
0
2
1
1
1
2
2
1
2
1
1
1
2
0
1
1
0
1
1
0
Source: NCI, 1978.
C-4
-------
Table C-4. Number of animals with and without specified
tumor types at time of death, female mice orally exposed to
1,2-dibromoethane
Groun
Control
Low-dose
Week
15
51
59
60
18
40
43
50
52
54
55
57
58
59
61
63
65
66
67
68
69
70
73
75
76
77
78
79
82
83
85
86
90
Forestomach tumors
Death with
tumor
0
0
0
0
0
2
1
1
1
1
2
0
1
1
1
1
2
2
1
1
1
2
2
3
3
2
1
1
2
1
2
2
7
Death without
tumor
1
1
9
9
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Luns
Death with
tumor
0
0
0
0
0
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
2
1
0
1
0
0
0
3
adenomas
Death without
tumor
1
1
9
9
1
2
1
0
1
1
2
0
0
1
1
1
2
2
1
1
1
1
2
3
3
0
0
1
1
1
2
2
4
C-5
-------
Table C-4. Number of animals with and without specified
tumor types at time of death, female mice orally exposed to
1,2-dibromoethane
Groun Week
High-dose 12
13
27
28
34
37
39
40
41
42
43
44
49
50
51
52
53
54
55
57
58
61
62
65
66
67
69
70
71
73
76
78
Forestomach tumors
Death with
tumor
0
0
0
0
1
1
1
1
0
0
1
1
0
1
1
1
2
0
2
1
1
1
1
1
1
1
1
1
1
1
2
2
Death without
tumor
9
4
1
1
0
0
0
0
1
1
0
1
1
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
Luns
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
adenomas
Death without
tumor
9
4
1
1
1
1
1
1
1
1
1
2
0
0
1
0
3
1
2
1
1
1
1
1
1
1
1
1
1
1
2
1
Source: NCI, 1978.
C-6
-------
mr_Forst3_f.ttd - male rats - forestomach squamous cell carcinomas
cc
1
0.8
0.6
0.4
0.2
0
Model: Multistage Weib
Dose (mg/kg/day)=11.2
Dose (mg/kg/day)=22.4
Kaplan Meier (11.2)
Kaplan Meier (22.4)
10
20
30
Time (wks)
40
50
Figure C-l. Kaplan-Meier hazard curves for the incidence of forestomach
squamous cell carcinomas in male rats in the oral gavage study (NCI, 1978).
C-7
-------
Output C-l. Multistage analysis of forestomach tumors in male rats, 1,2-dibromoethane
oral gavage exposure
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose"l-beta2*dose~2)]
The parameter betas are restricted to be positive
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Asymptotic Correlation Matrix of Parameter Estimates
Beta(1)
1
Analysis of Deviance Table
AIC:
Goodness of Fit
Est._Prob. Expected Observed
34
-------
Benchmark Dose Computation
Specified effect =
Risk Type
BMDL =
0.1
Extra risk
Specified effect
Risk Type
Confidence level
BMD
BMDL
Multistage Model with 0.95 Confidence Level
Multistage
0.8
0.6
0.4
0.2
BMDL
BMD
08:2802/122004
10
15
20
dose
Multistage Model with 0.95 Confidence Level
0.8
0.6
0.4
0.2
Multistage
BMD Lower Bound
BMDLBMD
0
11:1002/122004
10
15
20
dose
C-9
-------
10:5902/16/2004
CD
cr
0,8
0.6
0.4
0.2
0
Incidental Graph
mr_hemang3_mc.ttd - male rats - hemangiosarcomas
Model: Multistage Weib
Dose (mg/kg/day)=11.2
Dose (mg/kg/day)=22.4
• • Hoel Walburg (11.2)
- Hoel Walburg (22.4)
10
20
30
Time (wks)
40
50
Figure C-2. Cumulative incidence curves for hemangiosarcomas in male rats
in the oral gavage study (NCI, 1978.).
C-10
-------
Output C-2. Multistage analysis of hemangiosarcomas in male rats, 1,2-dibromoethane
oral gavage exposure
The form of the probability function is:
The parameter betas are restricted to be positive
Dependent variable = hemang
Independent variable = hec feb04
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
The model parameter ( s ) -Background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(1)
Beta(l) 1
Parameter Estimates
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
P-value
AIC:
0
11
4
C-ll
-------
DF = 2
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Specified effect
Risk Type
Confidence level
BMD
BMDL
Multistage Model with 0,95 Confidence Level
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Multistage
BMD Lower Bound
BMDL
BMD
10
15
20
dose
11:1702/122004
Multistage Model with 0.95 Confidence Level
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Multistage
BMD Lower Bound
BMDL
BMD
10
15
20
dose
14:1602/132004
-------
Incidental Graph
mr_thyr3.ttd . ma|e rats . thyroid foiiicular cell adenomas
Model: Multistage Weib
0.6
0.4
0.2
Dose (mg/kg/day)=22.4
Hoel Walburg (22.4)
-\—h
10
20
30
Time (wks)
40
50
Figure C-3. Cumulative incidence curves for thyroid follicular cell adenomas
or carcinomas in male rats in the oral gavage study (NCI, 1978).
C-13
-------
Output C-3. Multistage analysis of thyroid follicular cell adenomas in male rats, 1,2-
dibromoethane oral gavage exposure
The form of the probability function is:
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(l) = 0.0194293
Parameter Estimates
Variable Estimate
Background 0
Beta(l) 0.0159794
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
13
C-14
-------
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Multistage Model with 0.95 Confidence Level
0.5
0.4
0.3
0.2
0.1
Multistage
BMD Lower Bound
BMDL
BMD
10
15
20
dose
11:2002/122004
Multistage Model with 0.95 Confidence Level
0.5
0.4
T3
CD
"u
f 0.3
d
O
1 0.2
u_
0.1
o
Multistage
BMD Lower Bound
BMDL
BMD
10
15
20
dose
09:0602/122004
C-15
-------
11:06 02/16/2004_forst3.ttd - EDB, oral gavage - Female Rats: forestomach tumors
Model: Multistage Weib
(fl
cE
0.6
0.4
0.2
Dose (mg/kg/day)=20
Kaplan Meier (20)
10
20
30
Time (wks)
40
50
Figure C-4. Kaplan-Meier hazard curves for the incidence of forestomach
squamous cell carcinomas in female rats in the oral gavage study (NCI, 1978).
C-16
-------
11 -09 CP/16/2004 Incidental Graph
"" Tt_hemang3.ttd - EDB, oral gavage - Female Rats: hemangiosarcomas
Model: Multistage Weib
0.8
0.6
0.4
0.2
Dose (mg/kg/day)=20
Hoel Walburg (20)
10
20
30
Time (wks)
40
50
Figure C-5. Cumulative incidence curves for hemangiosarcomas in female
rats in the oral gavage study (NCI, 1978).
C-17
-------
Output C-4. Multistage analysis of hemangiosarcomas in female rats, 1,2-dibromoethane
oral gavage exposure
The form of the probability function is:
Dependent variable = hemang
Independent variable = mg kg day
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(l) = 0.0097078
Variable
Background
Beta(1)
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
0.0572 1.144
C-18
-------
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 17 . 8906
BMDL = 7.87117
Multistage Model with 0,95 Confidence Level
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Multistage
BMDL
BMD
11:2902/122004
10
dose
15
20
C-19
-------
09/1R/2004 Incidental Graph
*" fr_hep3.ttd - EDB, oral gavage - Female Rats: hepato. carcinomas
Model: Multistage Weib
0.6
0.4
0.2
Dose (mg/kg/day)=20
Hoel Walburg (20)
10
20
30
Time (wks)
40
50
Figure C-6. Cumulative incidence curves for hepatocellular carcinomas and
neoplastic nodules in female rats in the oral gavage study (NCI, 1978).
C-20
-------
Output C-5. Multistage analysis of hepatocellular carcinomas in female rats, 1,2-
dibromoethane oral gavage exposure
The form of the probability function is:
Dependent variable = hepcarc
Independent variable = mg kg day
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta (1) = 0.0202733
Parameter Estimates
Variable Estimate Std. Err.
Background 0 NA
Beta(1) 0.0142265 0.0112183
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
C-21
-------
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.6
0.5
0.4
0.3
Multistage Model with 0.95 Confidence Levei
0.1
Multistage
BMDL
BMD
11:3502/122004
10
dose
15
20
0.6
0.5
0.4
0.3
0.2
0.1
0
Multistage Model with 0.95 Confidence Level
Multistage
BMDL
BMD
14:3602/132004
10
dose
15
20
C-22
-------
11-1502/16P004. Incidental Graph
"" Tr_adren3.ttd - EDB, oral gavage - Female Rats: adren. carcinomas
Model: Multistage Weib
0.6
0.4
0.2
Dose (mg/kg/day)=20
Hoel Walburg (20)
10
20
30
Time (wks)
40
50
Figure C-7. Cumulative incidence curves for adrenocortical carcinomas and
neoplastic nodules in female rats in the oral gavage study (NCI, 1978).
C-23
-------
Output C-6. Multistage analysis of adrenocortical carcinomas in female rats, 1,2-
dibromoethane oral gavage exposure
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose~l)]
Dependent variable = adren
Independent variable = mg kg day
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Parameter Estimates
Variable Estimate Std. Err.
Background 0 NA
Beta(l) 0.0101304 0.0111373
Model
Full model
Fitted model
Reduced model
Goodness of Fit
Dose Est._Prob. Expected Observed Size Chi~2 Res.
i: 1
C-24
-------
Chi-square =
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 10. 4005
BMDL =
0
1
4
P-value
o
I
u
CO
ul
0.5
0.4
0.3
0.2
0.1
Multistage Model with 0.95 Confidence Level
Multistage
11:3902/122004
BMDL
BMD
10
dose
15
20
C-25
-------
11:2402/16/2004
05
cr
0.6
0.4
0.2
Fatal Graph
mm_forst3.ttd - Male mice, oral gavage, forestornach tumors
Model: Multistage Weib
Dose (mg/kg/day)=17
Kaplan Meier (17)
10
20
30
40 50
Time (wks)
60
70
80
Figure C-8. Kaplan-Meier hazard curves for the incidence of forestomach
squamous cell carcinomas in male mice in the oral gavage study (NCI, 1978).
C-26
-------
11:2702/16/2004
CO
ce
0.6
0.4
0.2
0
Incidental Graph
mmjung.ttd - Male mice, oral gavage, lung adenomas
Model: Multistage Weib
Dose (mg/kg/day)=17
HoelWalburg(17)
-t-
H
0 10 20 30 40 50 60 70 80 90
Time (wks)
Figure C-9: Cumulative incidence curves for lung adenomas in male mice in
the oral gavage study (NCI, 1978).
C-27
-------
Output C-7: Multistage analysis of alveolar/bronchiolar adenomas in male mice, 1,2-
dibromoethane oral gavage exposure
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose~l)]
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -Background
have been estimated at a boundary point, or have been specified by the user,
and do not appear in the correlation matrix )
Beta(1)
1
Parameter Estimates
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
P-value
Fitted model
Reduced model
AIC:
C-28
-------
DF = 2
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Specified effect
Risk Type
Confidence level
BMD
BMDL
Multistage Model with 0.95 Confidence Level
0.7
0.6
T3 0.5
ED
Tj
| °'4
I 0.3
O
O5
£ 0.2
0.1
0
Multistage
BMDL BMD
11:4502/122004
0.7
O.B
0.5
0.3
U
05
£ 0.2
0.1
0
Multistage
14:4702/132004
10
15
dose
Multistage Model with 0.95 Confidence Level
BMDL
BMD
10
15
dose
C-29
-------
11-3002/16/2004 Fatal Graph
fm_forst3.ttd - Female mice, oral gavage, forestomach tumors
Model: Multistage Weib
to
ce
0.6
0.4
0.2
0
Dose (mg/kg/day)=16
Kaplan Meier (16)
20
40 60
Time (wks)
80
100
Figure C-10. Kaplan-Meier hazard curves for the incidence of forestomach
squamous cell carcinomas in female mice in the oral gavage study (NCI,
1978).
C-30
-------
11:31 02/16/2004
CT
1
0.8
0.6
0.4
0.2
0
Incidental Graph
fm_lung3.ttd - Female mice, oral gavage, lung tumors
Model: Multistage Weib
Dose (m g/k g/d a y)=16
HoelWalburg (16)
H
20
40
60
Time (wks)
80
100
Figure C-ll. Cumulative incidence curves for lung adenomas in female mice
in the oral gavage study (NCI, 1978).
C-31
-------
Output C-8. Multistage analysis of alveolar/bronchiolar adenomas in female mice, 1,2-
dibromoethane oral gavage exposure
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose~l-beta2*dose~2)]
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Asymptotic Correlation Matrix of Parameter Estimates
Beta(1)
1
Analysis of Deviance Table
P-value
Goodness of Fit
Dose Est._Prob. Expected Observed
o noon o nnn n
C-32
-------
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.5
0,4
0.3
t3 0,2
05
£
0.1
0
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Multistage Model with 0,95 Confidence Level
Multistage
0 2
11:4702/122004
10
12
dose
14
16
0.5
0.4
0.3
ti 0.2
0
Multistage Model with 0.95 Confidence Level
Multistage
0 2
15:0702/132004
BMDL
BMD
10
12
14
dose
16
C-33
-------
C-2. Analyses in support of inhalation unit risk
Table C-5. Number of animals with and without specified tumor types at
time of death, male rats exposed by inhalation to 1,2-dibromoethane
Group
Control
Low-dose
High-dose
Week
77
88
89
90
93
97
99
102
103
104
106
7
38
56
85
86
93
95
96
97
98
99
102
104
43
50
53
56
62
63
64
67
68
69
70
71
74
76
77
78
80
81
82
83
84
85
86
87
88
89
All nasal
Death with
tumor
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
1
1
1
3
32
1
0
0
1
0
0
2
1
2
0
2
0
1
2
1
1
5
2
3
1
3
2
2
2
1
5
tumors
Death without
tumor
1
1
1
1
2
1
1
1
3
18
19
1
1
1
1
1
1
1
1
0
0
0
0
3
0
1
1
0
1
1
0
1
0
1
2
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
Hemangiosarcomas
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
1
1
0
1
1
0
2
0
0
0
0
1
0
2
1
3
Death without
tumor
1
1
1
1
2
1
1
1
3
19
19
1
1
1
1
1
1
1
1
1
1
1
3
35
1
0
1
1
1
0
2
2
2
1
3
0
1
1
0
1
3
2
3
1
3
1
2
0
0
3
Mesothelioma
Death with
tumor
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
7
0
1
1
1
0
1
1
1
1
0
3
0
0
1
1
0
0
1
1
0
1
0
1
2
1
6
Death
tumor
:
3
18
19
3
28
1
0
0
0
1
0
0
1
5
1
2
1
2
2
1
0
0
0
Source: NTP, 1982.
C-34
-------
Table C-6. Number of animals with and without specified tumor types at
time of death, female rats exposed by inhalation to 1,2-dibromoethane
Groun Week
Control 59
88
93
97
98
101
103
104
106
Low-dose 8
13
52
79
82
98
99
104
High-dose 54
55
60
63
67
69
71
72
73
76
78
79
80
82
83
84
85
86
87
88
89
90
91
Nasal
Death with
tumors
0
0
0
1
0
0
0
0
0
0
0
1
0
0
o
6
I
29
0
0
1
2
0
:
2
o
5
i
o
3
2
1
5
2
11
tumor
Death without
tumors
1
1
1
0
1
2
1
21
21
1
1
0
1
1
2
0
9
1
1
0
0
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
Fibroadenoma
Death with
tumors
0
0
0
0
0
0
0
3
1
0
0
1
0
1
5
1
21
0
0
0
1
0
0
0
1
0
0
0
0
0
1
2
1
0
1
2
1
5
1
8
Death without
tumors
1
1
1
1
1
2
1
18
20
1
1
0
1
0
0
0
17
:
:
0
0
2
1
2
0
0
1
1
3
Mammary adenocarcinoma
Death with
tumors
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
Death without
tumors
1
1
1
1
1
2
1
21
20
1
1
1
1
1
5
1
38
1
1
1
2
0
2
1
2
1
1
1
1
1
1
2
2
1
3
2
1
6
2
10
C-35
-------
Table C-6. Number of animals with and without
specified tumor types at time of death, female rats
exposed by inhalation to 1,2-dibromoethane
(continued)
Group
Control
Low-dose
High-dose
Week
59
88
93
97
98
101
103
104
106
8
13
52
79
82
98
99
104
54
55
60
63
67
69
71
72
73
76
78
79
80
82
83
84
85
86
87
88
89
90
91
Hemansiosarcoma
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
1
0
0
2
Death without
tumor
1
1
1
1
1
2
1
21
21
1
1
1
1
1
5
1
38
1
1
1
2
1
2
1
2
1
1
1
1
1
1
2
3
1
3
0
1
6
2
9
Alveolar/bronchiolar
adenoma/carcinoma
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
1
0
2
Death without
tumor
1
1
1
1
1
2
1
21
21
1
1
1
1
1
5
1
38
1
1
1
2
1
2
1
2
2
1
1
1
1
1
2
3
0
3
1
1
5
2
9
Source: NTP, 1982.
C-36
-------
Table C-7. Number of animals with and without specified tumor types at
time of death, female mice exposed by inhalation to 1,2-dibromoethane
Group Week
Control 24
85
86
94
95
96
98
100
101
104
106
Low-dose 4
54
56
62
66
73
82
84
90
91
94
95
96
97
98
100
101
102
104
High-dose 19
45
50
63
65
66
68
70
71
73
74
75
77
78
79
81
82
83
84
85
86
87
89
90
91
Nasal
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
1
0
1
0
0
1
2
0
0
0
0
1
0
0
0
0
tumors
Death without
tumor
22
19
3
1
3
o
6
2
o
5
2
o
6
2
19
1
0
1
o
5
0
i
i
2
2
0
1
0
2
1
3
0
3
2
1
o
5
2
1
3
1
8
Hemansiosarcoma
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
0
0
2
1
1
3
0
0
0
1
0
1
1
1
0
0
0
0
1
0
3
1
2
2
0
3
1
1
2
1
4
Death without
tumor
22
19
0
2
0
2
2
2
o
3
0
2
1
16
1
1
1
2
1
0
0
1
2
0
1
0
2
0
1
0
4
Alveolar/bronchiolar
adenoma/carcinoma
Death with
tumor
0
0
0
0
0
0
0
0
0
3
1
0
0
0
0
0
0
0
1
0
0
0
1
3
1
0
0
1
0
4
0
0
1
2
1
0
1
2
1
1
1
1
1
1
4
2
2
2
1
3
3
0
2
1
8
Death without
tumor
1
1
1
1
1
1
1
1
1
19
18
1
1
1
1
1
1
1
0
1
3
1
2
0
1
3
2
2
2
15
1
1
0
1
0
1
0
0
1
0
0
0
1
0
0
0
1
0
0
0
0
1
1
0
0
C-37
-------
Table C-7. Number of animals with and without specified tumor types at time of death,
female mice exposed by inhalation to 1,2-dibromoethane (continued)
Group
Control
Low-dose
High-dose
Week
24
85
86
94
95
96
98
100
101
104
106
4
54
56
62
66
73
82
84
90
91
94
95
96
97
98
100
101
102
104
19
45
50
63
65
66
68
70
71
73
74
75
77
78
79
81
82
83
84
85
86
87
89
90
91
Lung/bronchial
adenoma/carcinoma
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
1
0
1
0
0
0
1
1
0
0
1
0
1
Death without
tumor
22
19
3
1
3
3
1
3
2
3
2
19
:
:
0
3
2
3
2
0
2
3
1
2
1
7
Fibrosarcoma
Death with
tumor
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
0
2
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
1
0
1
1
1
0
0
1
1
0
3
1
0
Death without
tumor
22
19
0
0
1
3
1
3
3
2
1
2
3
2
18
1
1
0
3
1
1
1
2
2
1
1
1
1
1
3
1
2
2
1
2
2
1
0
0
8
Mammary
adenocarcinoma
Death with Death without
tumor
0
0
0
0
0
0
0
0
1
1
0
1
0
0
0
0
0
0
0
0
1
0
1
0
0
2
0
2
2
5
0
0
1
1
0
0
0
1
0
0
0
1
0
0
2
1
0
0
0
2
0
0
0
0
0
tumor
1
1
1
1
1
1
1
1
0
21
19
0
1
1
1
1
1
1
1
1
2
1
2
3
2
1
2
1
0
14
1
1
0
2
1
1
1
1
2
1
1
0
2
1
2
1
3
2
1
1
3
1
3
1
8
Source: NTP, 1982.
C-38
-------
Output C-9. Multistage analysis of nasal cavity tumors in male rats, 1,2-dibromoethane
inhalation exposure
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.0217391
intercept = 3.14555
slope = 1.27162
Asymptotic Correlation Matrix of Parameter Estimate
background intercept slope
background 1 -0.028 0.0028
intercept -0.028 1 0.89
slope 0.0028 0.89 1
Parameter Estimates
Variable
background
intercept
slope
Analysis of Deviance Table
P-value
Est. Prob.
C-39
-------
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
0.0149722
0.0030344
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.8
0.6
'r, 0.4
0.2
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
BMD Lower Bound
BMDL BMD
0
13:0202/162004
0.2
0.4
0.6
0.8
dose
Log-Logistic Model with 0.95 Confidence Level
0.6
ti 0.4
Log-Logistic
BMDL
BMD
0.2
0.4
0.6
0.8
dose
13:0302/162004
C-40
-------
Output C-10. Multistage-Weibull analysis of nasal cavity tumors in male rats, 1,2-
dibromoethane inhalation exposure
12:43 02/16/2004
.
en
ir
0.6
0.4
0.2
Fatal Graph
mr_Nasai_f.ttd - Nasal cavity tumors, male rats
Model: One Stage Weib
Dose (ppm)=1.42
Kaplan Meier (1.42)
0 10 20 30 40 50 60 70 80 90
Time (wks)
12:4202/16/2004
r
1
0.8
0.6
0.4
0.2
0
Incidental Graph
mr_Nasal_f.ttd - Nasal cavity tumors, male rats
Model: One Stage Weib
Dose (ppm)=1.42
Hoel Walburg (1.42)
20
40
60
Time (wks)
100
C-41
-------
Generating Model Fit Table
TITLE: Nasal cavity tumors, male rats
Model: One Stage Weib
Files\TOX_RISK\edb_inh\mr_Nasal_f.ttd
Functional form: 1 - EXP [ ( -QO - Ql * D ) * (T -
Maximum Log-Likelihood = -1.785287e+002
of animals
— Number --
with fatal
tumors
1
with incidental
tumors
Model: One Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Induction Time (TO) Set by User to 0
Incid Extra Risk
1.OOOOE-006
1.OOOOE-005
0.0001
0.0010
0.01
0.10
Dose Estimates (ppb)
95.00 %
MLE
6.2145E-004
6.2145E-003
6.2148E-002
6.2176E-001
6.2458E+000
6.5476E+001
C-42
-------
Output C-ll. Multistage analysis of hemangiosarcomas in male rats, 1,2-dibromoethane
inhalation exposure
The form of the probability function is:
Dependent variable = hemang
Independent variable = rgdr ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 1.01882
Asymptotic Correlation Matrix of Parameter Estimates
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Analysis of Deviance Table
P-value
Goodness of Fit
Dose Est._Prob. Expected Observed Size Chi^2 Res.
i: 1
0.000 0 46
C-43
-------
0.3600 0.0934 4.015
i: 3
0.8900 0.4507 12.619
Chi-square = 3.31 DF = 2
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.373241
BMDL = 0.291394
1 43 -0.828
15 28 0.343
P-value = 0.1906
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Multistage Model with 0.95 Confidence Level
Multistage
13:1402/162004
BMDL
BMD
0.2
0.4
0.6
0.8
dose
C-44
-------
Output C-12. Multistage-Weibull analysis of hemangiosarcomas in male rats, 1,2-
dibromoethane inhalation exposure
13:2302/16/2004
en
tr
0.6
0.4
0.2
Incidental Graph
mr_hernang_i.ttd - hemangiosarcomas, male rats
Model: Two Stage Weib
Dose (ppm)=1.42
HoelWalburg(1.42)
M
20
40
60
Time (wks)
80
100
Generating Model Fit Table
TITLE: hemangiosarcomas, male rats
Dataset: C:\Program
Model: Two Stage Weib
Files\TOX_RISK\edb_inh\mr_hemang_i.ttd
Functional form: 1~- EXP[T -QO -~Q1 * D - Q2 * D"2) * (T - TO)'
Maximum Log-Likelihood = -3.508137e+001
Dataset: C:\Program Files\TOX_RISK\edb_inh\mr_hemang_i.ttd
Exposure Pattern
Model: Two Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Unit Potency [ per mg/kg/day ] (computed for Risk of l.OE-6)
Lower Bound = Not Regstd MLE = 2.3189E-004 Upper Bound(gl*) = 7.0002E-002
C-45
-------
Induction Time (TO) Set by User to 0
Dose Estimates (ppb)
95.00 % 95.00 %
Incid Extra Risk Time Lower Bound MLE Upper Bound
l.OOOOE-006 70.00 6.3887E-003 1.9286E+000 Not Reqstd
l.OOOOE-005 70.00 6.3884E-002 6.0988E+000 Not Reqstd
0.0001 70.00 6.3857E-001 1.9287E+001 Not Reqstd
0.0010 70.00 6.3583E+000 6.1003E+001 Not Reqstd
0.01 70.00 6.1072E+001 1.9335E+002 Not Reqstd
0.10 70.00 4.6604E+002 6.2601E+002 Not Reqstd
C-46
-------
Output C-13. Multistage analysis of mesotheliomas in male rats, 1,2-dibromoethane
inhalation exposure
The form of the probability function is:
Dependent variable = meso
Independent variable = rgdr ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Default Initial Parameter Values
Background = 0.0134084
Beta(l) = 0
Beta(2) = 1.56244
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(2)
Background 1 -0.49
Beta(2) -0.49 1
Variable Estimate Std. Err.
Background 0.0208905 0.11861
Beta(l) 0 NA
Beta(2) 1.52272 0.433051
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Dose Est. Prob. Expected Observed Size Chi^2 Res.
C-47
-------
i: 1
1 46
8 43
25 35
P-value = 0.8431
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
0.263045
0.138834
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.6
0.7
0.6
0,5
0.4
0,3
0.2
0.1
0
Multistage Model with 0.95 Confidence Level
BMDL
BMD
0.2
0.4
0.6
0.8
dose
13-2702/162004
Multistage Model with 0.95 Confidence Level
0.8
0,7
0,6
O)
S 0,5
^ 0,4
O
I 0,3
0,2
0.1
0
Multistage
BMDL
BMD
13:2802/162004
0.2
C-48
0.4
0.6
O.B
dose
-------
Output C-14. Multistage-Weibull analysis of mesotheliomas in male rats, 1,2-
dibromoethane inhalation exposure
13:3302/16/2004
Incidental Graph
mrjneso.ttd - mesotheliomas, male rats
Model: One Stage Weib
0.6
0.4
0.2
0
Dose (ppm)=1.42
HoelWalburg(142)
20
40 60
Time (wks)
100
Generating Model Fit Table
TITLE: mesotheliomas, male rats
of animals
with incidental
tumors
1
8
Generating Extrapolated Doses Table
TITLE: mesotheliomas, male rats
Dataset: C:\Program Files\TOX_RISK\edb_inh\mr_meso.ttd
Exposure Pattern
Model: One Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Induction Time (TO) Set by User to 0
C-49
-------
Dose Estimates
95.00 %
Incid Extra Risk Time Lower Bound
l.OOOOE-006 70.00 1.3014E-003
l.OOOOE-005 70.00 1.3014E-002
0.0001 70.00 1.3015E-001
0.0010 70.00 1.3021E+000
0.01 70.00 1.3080E+001
0.10 70.00 1.3712E+002
C-50
-------
Output C-15. Multistage analysis of nasal cavity tumors in female rats, 1,2-dibromoethane
inhalation exposure
The form of the probability function is:
Dependent variable = all nasal
Independent variable = rgdr ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Default Initial Parameter Values
Background = 0.132247
Beta(l) = 4.03267
Beta(2) = 0
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background 1 -0.48
Beta(l) -0.48 1
Variable
Background
Beta(l)
Beta(2)
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
C-51
-------
i: 1
Chi-square =
1
34
43
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
0.0224109
0.0177547
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0 . 7
Extra risk
0. 95
0.256093
0.202887
0.8
I 0.6
•fi 0.4
0.2
1 . Multistage
y, o.e
0.2
Multistage Model with 0.95 Confidence Level
Multistage
BMDLBMD
0 0.1 0.2 0.3 0,4 0.5 0.6
dose
Multistage Model with 0.95 Confidence Level
0,7
BMDL BMD
0 0
13:3802/162004
C-52
0.2
0.3 0,4
dose
0.5
0.6
0,7
-------
Output C-16. Multistage-Weibull analysis of nasal cavity tumors in female rats, 1,2-
dibromoethane inhalation exposure
13:4302/16/2004
CE
0.6
0,4
0.2
0
Fatal Graph
fr_Nasal_f.ttd - Nasal cavity tumors - all, female rats
Model: One Stage Weib
Dose (ppm)=0.99
Kaplan Meier (0.99)
20
40 60
Time (wks)
80
100
C-53
-------
Generating Model Fit Table
TITLE: Nasal cavity tumors - all, female rats
of animals
-- Number --
with fatal
tumors
1
with incidental
tumors
Generating Extrapolated Doses Table
TITLE: Nasal cavity tumors - all, female rats
Dataset: C:\Program Files\TOX_RISK\edb_inh\fr_Nasal_f.ttd
Exposure Pattern
Model: One Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Unit Potency [ per mg/kg/day ] (computed for Risk of l.OE-6)
Lower Bound = Not Regstd MLE = 8.2967E-001 Upper Bound(gl*) = 1.0275E+000
13:4302/16/2004
CD
cr
0.6
0.4
0.2
0
Incidental Graph
ft_Nasal_f.ttd - Nasal cavity tumors - all, female rats
Model: One Stage Weib
Dose (pprn)=0,99
Hoel Walburg (0.99)
Do
20
40
60
Time (wks)
80
100
C-54
-------
95.00 %
Incid Extra Risk
l.OOOOE-006
l.OOOOE-005
0.0001
0.0010
0. 01
0.10
C-55
-------
Output C-17. Multistage analysis of alveolar/bronchiolar adenomas, carcinomas in female
rats, 1,2-dibromoethane inhalation exposure
The form of the probability function is:
Dependent variable = alv br
Independent variable = rgdr ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(1) = 0.358396
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model Log(likelihood) Deviance Test DF P-value
Full model
Fitted model
Reduced model
C-56
-------
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.533421
BMDL = 0.276777
Multistage Model with 0.95 Confidence Level
) Affected
o
"o
£
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
h A U' *
.
•
_^~-—^^~ •
^~—"~~~~~~~~~~
^— — • — """"
BMDL BMD
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
dose
13:4702/162004
C-57
-------
Output C-18. Multistage-Weibull analysis of alveolar/bronchiolar adenomas, carcinomas in
female rats, 1,2-dibromoethane inhalation exposure
17:2309/08/2003
ce
0.8
0.6
0.4
0.2
Incidental Graph
fr_Alv_br.ttd - Alveolar/branch, care/adenomas, female rats
Model: One Stage Weib
— — Dose (ppm)=3.1
Dose (ppm)=12.3
----- HoelWalburg(12.3)
20 40 60 80
Time (wks)
100
120
of animals
Model: One Stage Weib
Target Species: Human
Route: Air
with incidental
tumors
0
0
Exposure Pattern
Age Begins: 0 Age Ends: 70
Weeks/Year: 52 Days/Week: 7
Hours/Day : 24
Dose Estimates
95.00 %
Lower Bound
6.0694E-003
6.0694E-002
6.0697E-001
6.0724E+000
6.0999E+001
(ppb)
95.00 %
Upper Bound
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
C-58
-------
Output C-19. Multistage analysis of hemangiosarcomas in female rats, 1,2-dibromoethane
inhalation exposure
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal^dose^l)]
Dependent variable = hemang
Independent variable = rgdr ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(1) = 0.343027
Parameter Estimates
Variable Estimate
Background 0
Beta(1) 0.192493
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
C-59
-------
0.0470
i: 3
0.6600 0.1193
Chi-square =
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.547347
BMDL = 0.284013
0 42 -1.049
5 26 0.695
P-value = 0.1837
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Multistage Model with 0.95 Confidence Level
Multistage
14:0702/162004
BMDL
BMD
0.2
0.3
0.4
0.5
0.6
0.7
dose
C-60
-------
Output C-20. Multistage-Weibull analysis of hemangiosarcomas in female rats, 1,2-
dibromoethane inhalation exposure
14:0602/16/2004
w
O.
0.8
0.6
0.4
0.2
Incidental Graph
fr_Hemang_i.ttd - Circ. system: hemangiornas, female rats
Model: One Stage Weib
Dose (ppm)=7.1
HoelWalburg(7.1)
20
40
60
Time (wks)
80
100
Dataset: C:\Program
Dataset: C:\Program Files\TOX_RISK\edb_inh\fr_Hemang_i.ttd
Exposure Pattern
Model: One Stage Weib Age Begins: 0 Age Ends: 7 0
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
C-61
-------
Induction Time (TO) Set by User to 0
Dose Estimates (ppb)
95.00 % 95.00 %
Incid Extra Risk Time Lower Bound MLE Upper Bound
l.OOOOE-006 70.00 3.6427E-002 7.1476E-002 Not Reqstd
l.OOOOE-005 70.00 3.6428E-001 7.1476E-001 Not Reqstd
0.0001 70.00 3.6429E+000 7.1479E+000 Not Reqstd
0.0010 70.00 3.6446E+001 7.1512E+001 Not Reqstd
0.01 70.00 3.6611E+002 7.1836E+002 Not Reqstd
C-62
-------
Output C-21. Multistage analysis of mammary fibroadenomas in female rats, 1,2-
dibromoethane inhalation exposure
The form of the probability function is:
P [response] = background+ (1-background )/ [l + EXP(-intercept-slope*Log(dose) ) '_
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log trans formed model
Default Initial Parameter Values
background = 0.0851064
intercept = -0.381425
slope = 1
Asymptotic Correlation Matrix of Parameter Estimates
background intercept
background 1 -0.3
intercept -0.3 1
Parameter Estimates
Variable
background
intercept
slope
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
Scaled
Est._Prob. Expected Observed Size Residual
C-63
-------
Benchmark Dose Computation
Specified effect =
Risk Type =
Confidence level =
BMD =
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.9
0.8
0.7
-o 0.6
(B
O
ig 0.5
<
o 0.4
"o
2 0.3
0.2
0.1
0
Log-Logistic Model with 0.95 Confidence Level
Log-Logistic
BMDL BMD
0
14:1502/162004
dose
Log-Logistic Model with 0.95 Confidence Level
0.9
0.8
0.7
0.6
0.5
0.4
0-3
0.2
0.1
0
Log-Logistic
BMDL
BMD
1
dose
14:1702/162004
C-64
-------
Output C-22. Multistage-Weibull analysis of mammary fibroadenomas in female rats, 1,2-
dibromoethane inhalation exposure
17:1709/08/2003
cc
1
0.8
0.6
0.4
0.2
0
Incidental Graph
fr_Mamm_fib.ttd - Mammary gland: fibroadenomas, female rats
Model: One Stage Weib
Dose (ppm)=1.8
Dose (ppm)=7.1
Hoel Walburg (1.8)
Hoel Walburg (7.1)
20
40
60
Time (wks)
80
100
120
of animals
50
50
50
IN U.1LLLJ t: J_
with fatal
tumors
0
0
0
with incidental
tumors
4
29
2 4
Model: One Stage Weib
Target Species: Human
Route: Air
Exposure Pattern
Age Begins: 0 Age Ends: 7 0
Weeks/Year: 52 Days/Week: 7
Hours/Day : 24
C-65
-------
0.10 70.00 2.9806E+001 4.1719E+001 Not Reqstd
Output C-23. Multistage analysis of mammary adenocarcinomas in female rats, 1,2-
dibromoethane inhalation exposure
Multistage Model. $Revision: 2.1 $ $Date: 2000/08/21 03:38:21 $
Input Data File: F:\KHOGAN02\_BMDS\EDB_INH_FR.(d)
Gnuplot Plotting File: F:\KHOGAN02\_BMDS\EDB_INH_FR.plt
Mon Feb 16 14:11:39 2004
BMDS MODEL RUN
The form of the probability function is:
Dependent variable = m aden
Independent variable = rgdr ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(l) = 0.238933
Asymptotic Correlation Matrix of Parameter Estimates
Background Beta(l)
Background 1 -0.59
Beta(l) -0.59 1
Parameter Estimates
Variable Estimate
Background 0.0136816
Beta(l) 0.131115
Analysis of Deviance Table
Model Log(likelihood) Deviance Test DF P-value
1
Goodness of Fit
Dose Est._Prob. Expected Observed Size Chi^2 Res.
i: 1
1 47
C-66
-------
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.803575
BMDL = 0.357022
Multistage Model with 0.95 Confidence Level
T3
-------
Output C-24. Multistage-Weibull analysis of mammary adenocarcinomas in female rats,
1,2-dibromoethane inhalation exposure
17:1909/08/2003
en
ir
0.8
0.6
0.4
0.2
Incidental Graph
fr_Mamm_ad.ttd - Mammary gland adenocarc., female rats
Model: One Stage Weib
-----
Dose (pprn)=1,8
Dose (ppm)=7.1
HoelWalburg(7.1)
20
40
60
Time (wks)
100
120
of animals
-- Number --
with fatal
tumors
0
0
0
with incidental
tumors
1
0
4
Unit Potency [ per mg/kg-day ] (computed for Risk of l.OE-6)
Lower Bound = Not Reqstd MLE = 2.7175E-002 Upper Bound(ql*) = 6.6137E-002
Incid Extra Risk
l.OOOOE-006
1.OOOOE-005
0.0001
0.0010
0. 01
0.10
Dose Estimates
95.00 %
Lower Bound
6.7621E-003
6.7622E-002
6.7625E-001
6.7655E+000
6.7962E+001
7.1246E+002
C-68
-------
Output C-25. Multistage analysis of nasal cavity tumors in female mice, 1,2-dibromoethane
inhalation exposure
The form of the probability function is:
Dependent variable = all nasal
Independent variable = rgdr ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Default Initial Parameter Values
Background = 0
Beta(l) = 0
Beta(2) = 0.412082
Asymptotic Correlation Matrix of Parameter Estimates
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Analysis of Deviance Table
P-value
Goodness of Fit
Dose Est._Prob. Expected Observed Size Chi^2 Res.
i: 1
0 46
C-69
-------
0.0398
0.9500 0.2465
Chi-square = 1.94
DF = 2
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0 38 -1.041
8 27 0.268
P-value = 0.3797
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Multistage Model with 0.95 Confidence Level
0.5
0.4
0.3
Multistage
0.1
BMDL
BMD
0.2
0.4
0.6
O.B
dose
14:1902/162004
Multistage Model with 0.95 Confidence Level
0.5
0.4
0.3
Multistage
0.1
BMDL
BMD
0.2
0.4 0.6
dose
O.B
14:21 02/162004
C-70
-------
Output C-26. Multistage-Weibull analysis of nasal cavity tumors in female mice, 1,2-
dibromoethane inhalation exposure
14:2402/16/2004
cr
0.8
0.6
0.4
0.2
Incidental Graph
fm_nasal.ttd - Female mice nasal cavity tumors
Model: One Stage Weib
Dose (ppm)=1.42
HoelWalburg(1.42)
20
40
60
Time (wks)
80
100
Generating Model Fit Table
TITLE: Female mice nasal cavity tumors
of animals
-- Number --
with fatal
tumors
0
0
with incidental
tumors
Generating Extrapolated Doses Table
TITLE: Female mice nasal cavity tumors
Dataset: C:\Program Files\TOX_RISK\edb_inh\fmjnasal.ttd
Exposure Pattern
Model: One Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Unit Potency [ per mg/kg/day ] (computed for Risk of l.OE-6)
Lower Bound = Not Regstd MLE = 5.4729E-002 Upper Bound(gl*) = 8.0545E-002
C-71
-------
Induction Time (TO) Set by User to 0
Dose Estimates (ppb)
95.00 % 95.00 %
Incid Extra Risk Time Lower Bound MLE Upper Bound
l.OOOOE-006 70.00 5.5495E-003 8.1673E-003 Not Reqstd
l.OOOOE-005 70.00 5.5496E-002 8.1674E-002 Not Reqstd
0.0001 70.00 5.5498E-001 8.1677E-001 Not Reqstd
0.0010 70.00 5.5523E+000 8.1714E+000 Not Reqstd
0.01 70.00 5.5775E+001 8.2084E+001 Not Reqstd
0.10 70.00 8.3863E+002 8.6051E+002 Not Reqstd
C-72
-------
Output C-27. Multistage analysis of lung/bronchial adenomas, carcinomas in female mice,
1,2-dibromoethane inhalation exposure
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal^dose^l)]
Dependent variable = lungbr
Independent variable = tb ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(1) = 0.0299027
Parameter Estimates
Variable Estimate
Background 0
Beta(1) 0.0194024
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
C-73
-------
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 5.43029
BMDL = 3.28056
1 38 -0.778
8 26 0.500
P-value = 0.2227
0.5
0.4
0.3
0.1
0
Multistage Model with 0.95 Confidence Level
Multistage
BMDL
BMD
0 2
14:2802/162004
10
12
14
dose
C-74
-------
Output C-28. Multistage-Weibull analysis of lung/bronchial adenomas, carcinomas in
female mice, 1,2-dibromoethane inhalation exposure
09:56 09/09/2003
cr
0.8
0.6
0.4
0.2
0
Incidental Graph
fm_trbr.ttd - Female mice lung/bronchiolar aden/carc
Model: One Stage Weib
Dose (ppm)=4.86
Dose (ppm)=19.2
• • Hoel Walburg (4.86)
Hoel Walburg (19.2)
20
40 60 80
Time (wks)
100
120
of animals
49
50
50
IN U.1LLLJ t: J_
with fatal
tumors
0
0
0
with incidental
tumors
0
1
8
Model: One Stage Weib
Target Species: Human
Route: Air
Dose Estimates (ppb)
95.00 %
Lower Bound MLE
4.1524E-002 8.8023E-002
4.1524E-001 8.8024E-001
4.1526E+000 8.8028E+000
4.1545E+001
C-75
-------
Not Reqstd
Not Reqstd
C-76
-------
Output C-29. Multistage analysis of alveolar/bronchiolar adenomas, carcinomas in female
mice, 1,2-dibromoethane inhalation exposure
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal*dose^l-beta2*dose^2)]
Dependent variable = alv br
Independent variable = pulm ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 3
Total number of specified parameters = 0
Degree of polynomial = 2
Default Initial Parameter Values
Background = 0.02142
Beta(l) = 0
Beta(2) = 0.0114798
Background
Background 1
Beta(2) -0.39
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Analysis of Deviance Table
P-value
Dose Est. Prob. Expected Observed Size Chi^2 Res.
C-77
-------
Chi-square =
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0.1
Extra risk
0. 95
3.23653
2.04555
4 46
11 40
41 44
P-value = 0.3040
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Multistage Model with 0.95 Confidence Level
Multistage
0.6
0.4
0.2
0
BMDL
BMD
0 2
14:3302/162004
10
12
14
16
dose
£ 0.6
a 0-4
0.2
Multistage Model with 0.95 Confidence Level
Multistage
14:3502/162004
BMDL
BMD
10
12
14
16
dose
C-78
-------
Output C-30. Multistage-Weibull analysis of alveolar/bronchiolar adenomas, carcinomas in
female mice, 1,2-dibromoethane inhalation exposure
17:4609/08/2003
to
CE
1
0.8
0.6
0.4
0.2
0
Incidental Graph
fm_alvbr.ttd - Female mice alveolar/bronchiolar aden/carc
Model: Two Stage Weib
Dose (ppm)=5.8
Dose (ppm)=22.7
Hoel Walburg (5.8)
Hoel Walburg (22.7)
H
20
40
60
Time (wks)
80
100
120
Model: Two Stage Weib
Dataset: C:\Program Files\TOX_RISK\edb_inh\fm_alvbr.ttd
Functional form: 1 - EXP [ ( -QCI - Ql * D - Q2 *" D^2) * (T -
Maximum Log-Likelihood = -5.951707e+001
Parameter Estimates :
Q 0 = 1.853318E-006
Q 1 = 0.OOOOOOE+000
Q 2 = 1.448222E-007
Z = 2.314668E+000
TO = 0.OOOOOOE+000
of animals
49
50
50
IN U.1LLLJ t: J_
with fatal
tumors
0
0
0
with incidental
tumors
4
11
41
Exposure Pattern
Model: Two Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Dose Estimates (ppb)
95.00 %
Lower Bound MLE
2.7511E-002 1.1247E+001
2.7510E-001 3.5567E+001
2.7502E+000 1.1248E+002
2.7419E+001 3.5576E+002
C-79
-------
Not Reqstd
Not Reqstd
C-80
-------
Output C-31. Multistage analysis of hemangiosarcomas in female mice, 1,2-dibromoethane
inhalation exposure
The form of the probability function is:
Dependent variable = hemang
Independent variable = ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta (1) = 0.26484
NA - Indicates that this parameter has hit a bound
implied by some ineguality constraint and thus
has no standard error.
Model
Full model
Fitted model -45.6745 0.600956 2
Reduced model -74.4981 58.2481 2
93.349
Goodness of Fit
Est._Prob. Expected
C-81
-------
Chi-square =
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Specified effect
Risk Type
Confidence level
BMD
BMDL
1.
0,8
0.7
0.6
73
OJ
S 0,5
! 0.4
O
| 0,3
0,2
0.1
0
Multistage Model with 0.95 Confidence Level
Multistage
BMDL BMD
0
14:4302/162004
0.8
0,7
0,6
0,5
0,4
0,3
0,2
0.1
0
Multistage
dose
Multistage Model with 0.95 Confidence Level
BMDL
BMD
1
dose
14:4502/162004
C-82
-------
Output C-32. Multistage-Weibull analysis of hemangiosarcomas in female mice, 1,2-
dibromoethane inhalation exposure
14:41 02/16/2004
en
ir
1
0.8
0.6
0.4
0.2
Incidental Graph
fm_hemang_inc.ttd - Female mice hemangiosarcomas
Model: One Stage Weib
Dose (ppm)=7.1
HoelWalburg(7.1)
20
40 60
Time (wks)
80
100
Generating Model Fit Table
TITLE: Female mice hemangiosarcomas
Dataset: C:\Program
- Number --
•lith fatal
tumors
0
0
with incidental
tumors
Generating Extrapolated Doses Table
TITLE: Female mice hemangiosarcomas
Dataset: C:\Program Files\TOX_RISK\edb_inh\fm_hemang_inc.ttd
Exposure Pattern
Model: One Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Unit Potency [ per mg/kg/day ] (computed for Risk of l.OE-6)
Lower Bound = Not Regstd MLE = 6.8278E-002 Upper Bound(gl*) = 8.9891E-002
C-83
-------
Induction Time (TO) Set by User to 0
Dose Estimates (ppb)
95.00 % 95.00 %
Incid Extra Risk Time Lower Bound MLE Upper Bound
l.OOOOE-006 70.00 4.9725E-003 6.5466E-003 Not Reqstd
l.OOOOE-005 70.00 4.9726E-002 6.5466E-002 Not Reqstd
0.0001 70.00 4.9728E-001 6.5469E-001 Not Reqstd
0.0010 70.00 4.9750E+000 6.5498E+000 Not Reqstd
0.01 70.00 4.9976E+001 6.5795E+001 Not Reqstd
0.10 70.00 5.2391E+002 6.8975E+002 Not Reqstd
C-84
-------
Output C-33. Multistage analysis of fibrosarcomas in female mice, 1,2-dibromoethane
inhalation exposure
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(
-betal^dose^l)]
Dependent variable = fibro inc
Independent variable = ppm
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Default Initial Parameter Values
Background = 0
Beta(1) = 0.111017
Parameter Estimates
Variable Estimate
Background 0
Beta(1) 0.0959514
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model
Full model
Fitted model
Reduced model
C-85
-------
1.8000 0.1586
i: 3
4.8000 0.3691
Chi-square = 0.44
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 1.09806
BMDL = 0.745016
5 39 -0.228
11 27 0.165
P-value = 0.8022
Specified effect =
Risk Type =
Confidence level =
BMD =
BMDL =
0.15
Extra risk
0.95
1. 69376
1.14919
0.6
0.5
| 0.4
O
O)
5 0.3
e=
o
~o
£ 0.2
0.1
0
Multistage Model with 0.95 Confidence Level
Multistage
BMDL BMD
0 1
15:1402/162004
dose
0.6
0.5
1 0.4
tj
J£!
< 0.3
C
O
o
£ 0-2
0.1
0
Multistage
Multistage Model with 0.95 Confidence Level
15:1602/162004
BMDL
C-86
BMD
dose
-------
Output C-34. Multistage-Weibull analysis of fibrosarcomas in female mice, 1,2-
dibromoethane inhalation exposure
10:0709/09/2003
O.
0.8
0.6
0.4
0.2
Incidental Graph
fm_fibsarc.ttd - Female mice fibrosarcorna
Model: One Stage Weib
Dose (ppm)=1.8
Dose (ppm)=7.1
- — Hoel Walburg (1.8)
•--- Hoel Walburg (7.1)
H
20
40 60 80
Time (wks)
100
120
of animals
50
50
50
IN U.1LLLJ t: J_
with fatal
tumors
0
0
0
with incidental
tumors
0
5
11
Exposure Pattern
Model: One Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Incid Extra Risk
1.OOOOE-006
l.OOOOE-005
0.0001
0.0010
0.01
0.10
Dose Estimates
95.00 %
Lower Bound
1.0370E-002
1.0370E-001
1.0370E+000
1.0375E+001
1.0422E+002
1.0926E+003
(ppb)
C-87
-------
Output C-35. Multistage-Weibull analysis of mammary adenocarcinomas in female mice,
1,2-dibromothane inhalation exposure
Total number of observations = 3
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
User has chosen the log transformed model
Default Initial Parameter Values
background = 0.0434783
intercept = -1.65098
slope = 1
background intercept
background 1 -0.44
intercept -0.44 1
Parameter Estimates
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Model Log(likelihood) Deviance Test DF P-value
Full model
Fitted model
Reduced model
Dose Est._Prob. Expected
14
9
C-88
-------
Chi-square =
DF = 1
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
Specified effect
Risk Type
Confidence level
BMD
BMDL
0.25
Extra risk
0. 95
2.14947
1.30125
•3
0.5
0.4
| 0.3
C
D
I 0.2
J_
0.1
0
Log-Logistic Model with 0,95 Confidence Level
Log-Logistic
BMDL BMD
0 1
15:2202/162004
dose
Log-Logistic Model with 0.95 Confidence Level
0.5
0.4
0.3
0.2
0.1
0
Log-Logistic
BMDL
BMD
0 1
15:2302/162004
2 3
dose
C-89
-------
Output C-36. Multistage-Weibull analysis of mammary adenocarcinomas in female mice,
1,2-dibromothane inhalation exposure
10:0409/09/2003
tn
CE
0.8
0.6
0.4
0.2
Incidental Graph
fm_mam_ad.ttd - Female mice mammary adenocarc.
Model: One Stage Weib
Dose (ppm)=1.8
Dose (ppm)=7.1
Hoel Walburg (1.8)
-- Hoel Walburg (7.1)
20
40
60
Time (wks)
80
100
120
of animals
50
50
50
IN U.1LLLJ t: J_
with fatal
tumors
0
0
0
with incidental
tumors
2
14
9
Exposure Pattern
Model: One Stage Weib Age Begins: 0 Age Ends: 70
Target Species: Human Weeks/Year: 52 Days/Week: 7
Route: Air Hours/Day : 24
Animal to human conversion method: PPM IN AIR
Incid Extra Risk
1.OOOOE-006
l.OOOOE-005
0.0001
0.0010
0.01
0.10
(ppb)
MLE
2.1114E-002
2.1114E-001
2.1115E+000
2.1125E+001
2.1221E+002
2.2246E+003
C-90
------- |