Toxicological Review of 1 2 3-Trichloropropane


DRAFT - DO NOT CITE OR QUOTE
NCEA-S-1669
www.epa.gov/iris
TOXICOLOGICAL REVIEW
OF
1,2,3-TRICHLOROPROPANE
(CAS No. 96-18-4)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
October 2007
NOTICE
This external review document is distributed solely for the purpose of pre-dissemination peer
review under applicable information quality guidelines. It has not been formally disseminated by
EPA. It does not represent and should not be construed to represent any Agency determination or
policy. It is being circulated for review of its technical accuracy and science policy implications.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
a ¦—r*ijt
jvtrA
U.S. Environmental Protection Agency
Washington DC

-------
CONTENTS —TOXICOLOGICAL REVIEW OF
1,2,3-TRICHLOROPROPANE (CAS No. 96-18-4)
LIST OF TABLES	iv
LIST OF FIGURES	vii
LIST OF ABBREVIATIONS AND ACRONYMS	viii
FOREWORD	ix
AUTHORS, CONTRIBUTORS, AND REVIEWERS	x
1.	INTRODUCTION	1
2.	CHEMICAL AM) PHYSICAL INFORMATION	3
3.	TOXICOKINETICS	4
3.1.	ABSORPTION	4
3.2.	DISTRIBUTION	5
3.3.	METABOLISM	6
3.4.	ELIMINATION	8
3.5.	PHYSIOLOGICALLY-BASED TOXICOKINETIC MODELING	9
4.	HAZARD IDENTIFICATION	 10
4.1.	STUDIES IN HUMANS - EPIDEMIOLOGY, CASE REPORTS	10
4.2.	SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIO ASSAYS IN
ANIMALS - ORAL AM) INHALATION	 10
4.2.1.	Oral Exposure	10
4.2.1.1.	Subchronic Studies	10
4.2.1.2.	Chronic Studies	22
4.2.2.	Inhalation Exposure	32
4.2.2.1.	Subchronic Studies	33
4.2.2.2.	Chronic Studies	41
4.3.	REPRODUCTIVE/DEVELOPMENTAL STUDIES - ORAL AND INHALATION ... 41
4.3.1.	Oral Studies	41
4.3.2.	Inhalation Studies	44
4.4.	OTHER STUDIES	46
4.4.1.	Acute Toxicity Data	46
4.4.2.	Waterborne Studies	46
4.5.	MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION	47
4.5.1.	Mode of Action Studies	47
4.5.2.	Genotoxicity Studies	52
4.5.3.	Structure-Activity Relationships	60
4.6.	SYNTHESIS OF MAJOR NONCANCER EFFECTS	61
4.6.1.	Oral Exposure	61
4.6.2.	Inhalation Exposure	64
4.7.	EVALUATION OI CARCINOGENICITY	65
October, 2007	i	DRAFT - DO NOT CITE OR QUOTE

-------
4.7.1.	Summary of Overall Weight of Evidence	65
4.7.2.	Synthesis of Human, Animal, and Other Supporting Evidence	66
4.7.3.	Mode of Action Information	67
4.7.3.1.	Hypothesized Mode of Action	67
4.7.3.2.	Experimental Support for the Hypothesized Mode of Action	67
4.7.3.3.	Other Possible Modes of Action	73
4.7.3.4.	Conclusions About the Hypothesized Mode of Action	74
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES	77
4.8.1.	Possible Childhood Susceptibility	77
4.8.2.	Possible Gender Differences	78
4.8.3.	Other	78
5.	DOSE RESPONSE ASSESSMENT	79
5.1.	CHRONIC ORAL REFERENCE DOSE (RID)	79
5.1.1.	Choice of Principal Study and Critical Effect - with Rationale and Justification79
5.1.2.	Methods of Analysis - Including Models	80
5.1.3.	Chronic RfD Derivation - Including Application of Uncertainty Factors (UFs) 83
5.1.4.	Chronic RfD Comparison Information	85
5.1.5.	Previous Oral Assessment	89
5.2.	CHRONIC INHALATION REFERENCE CONCENTRATION (RfC)	89
5.2.1.	Choice of Principal Study and Critical Effect - with Rationale and Justification 89
5.2.2.	Methods of Analysis - Including Models	90
5.2.3.	Chronic RfC Derivation - Including Application of Uncertainty Factors (UFs). 91
5.2.4.	Chronic RfC Comparison Information	93
5.2.5.	Previous Inhalation Assessment	93
5.3.	UNCERTAINTIES IN CHRONIC ORAL REFERENCE DOSE (RfD) AND
INHALATION REFERENCE CONCENTRATION (RfC)	94
5.4.	CANCER ASSESSMENT	97
5.4.1.	Choice of Study/Data with Rationale and Justification	97
5.4.2.	Dose-Response Data	98
5.4.3.	Dose Adjustments and Extrapolation Methods	101
5.4.4.	Oral Slope Factor and Inhalation Unit Risk	104
5.4.5.	Uncertainties in Cancer Risk Values	110
5.4.5.1. Sources of Uncertainty	110
6.	MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF	115
HAZARD AND DOSE RESPONSE	115
6.1.	HUMAN HAZARD POTENTIAL	115
6.2.	DOSE RESPONSE	117
6.2.1.	Noncancer/Oral	117
6.2.2.	Noncancer/Inhalation	121
6.2.3.	Cancer/Oral and Inhalation	122
7.	REFERENCES	129
October, 2007	ii	DRAFT - DO NOT CITE OR QUOTE

-------
Appendix B-l: Benchmark Dose Modeling Results for the Derivation of the RfD	137
Appendix C-l: Derivation of the oral slope factor using TOXRISK software	170
October, 2007
iii
DRAFT - DO NOT CITE OR QUOTE

-------
LIST OF TABLES
Table 3-1. Distribution and excretion of radiolabeled 1,2,3-trichloropropane (30 mg/kg) 60
hours after oral (gavage) administration 5
Table 4-1. Relative organ weight changes in F-344/N rats receiving 1,2,3-trichloropropane by
gavage for 120 days 12
Table 4-2. Absolute organ weight changes in F-344/N rats receiving 1,2,3-trichloropropane by
gavage for 120 days 12
Table 4-3. Incidence of liver, kidney, and nasal turbinate lesions in male and female F-344N rats
in 17-week study 14
Table 4-4. Relative organ weight changes in B6C3F1 mice receiving 1,2,3-trichloropropane by
gavage for 120 days 17
Table 4-5. Absolute organ weight changes in B6C3F1 mice receiving 1,2,3-trichloropropane by
gavage for 120 days 17
Table 4-6. Incidence of liver, lung, and forestomach lesions in male and female B6C3F1 mice in
17-week study 19
Table 4-7. Incidence of myocardial necrosis in male and female Sprague-Dawley rats following
90-day 1,2,3-trichloropropane exposure 20
Table 4-8. Survival rates and percent probability of survival for F-344/N rats exposed to 1,2,3-
trichloropropane by gavage for two years 23
Table 4-9a. Relative liver weights (mg organ weight/ g body weight) and percent change in
F344/N Rats chronically exposed to 1,2,3-trichloropropane by gavage at the 15-month
interim evaluation 24
Table 4-9b. Absolute liver weights (grams) and percent change in F344/N Rats chronically
exposed to 1,2,3-trichloropropane by gavage at the 15-month interim evaluation 25
Table 4-10a. Relative right kidney weights (mg organ weight/ g body weight) and percent
change in F344/N Rats chronically exposed to 1,2,3-trichloropropane by gavage at the 15-
month interim evaluation 25
Table 4-10b. Absolute right kidney weights (grams) and percent change in F344/N Rats
chronically exposed to 1,2,3-trichloropropane by gavage at the 15-month interim evaluation
25
Table 4-11. Incidence of neoplasms in F-344/N rats chronically exposed to 1,2,3-
trichloropropane by gavage 27
October, 2007 iv DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-12. Survival rates and percent probability of survival for B6C3F1 mice exposed to
1,2,3-trichloropropane by gavage for two year 29
Table 4-13a. Relative liver weights (mg organ weight/ g body weight) and percent change in
B6C3F1 mice chronically exposed to 1,2,3-trichloropropane by gavage 29
Table 4-13b. Absolute liver weights (grams) and percent change in B6C3F1 mice chronically
exposed to 1,2,3-trichloropropane by gavage 30
Table 4-14a. Relative right kidney weights (mg organ weight/ g body weight) and percent
change in B6C3F1 mice chronically exposed to 1,2,3-trichloropropane by gavage 30
Table 4-14b. Absolute right kidney weights (grams) and percent change in B6C3F1 mice
chronically exposed to 1,2,3-trichloropropane by gavage 30
Table 4-15. Incidence of neoplasms in B6C3F1 mice chronically exposed to 1,2,3-
trichloropropane by gavage 32
Table 4-16. Absolute and relative liver weights and percent change in CD rats exposed to 1,2,3-
trichloropropane by inhalation, 6 hours/day, 5 days/week, for 13-weeks 34
Table 4-17. Absolute and relative lung weights and percent change in CD rats exposed to 1,2,3-
trichloropropane by inhalation, 6 hours/day, 5 days/week, for 13-weeks 35
Table 4-18. Incidence of histopathologic lesions in CD rats exposed via inhalation to 1,2,3-
trichloropropane, 6 hours/day, 5 days/week for 13 weeks 36
Table 4-19. Incidence and severity of decreased thickness and degeneration of the olfactory
epithelium in the nasal turbinates of F344/N rats exposed via inhalation to 1,2,3-
trichloropropane 38
Table 4-20. Incidence and severity of inflammation of the olfactory epithelium in the nasal
turbinates of F344/N rats exposed via inhalation to 1,2,3-trichloropropane 39
Table 4-21. Incidence and severity of decreased thickness and degeneration of the olfactory
epithelium in the nasal turbinates in B6C3F1 mice exposed via inhalation to 1,2,3-
trichloropropane 40
Table 4-22. Incidence and severity of inflammation of the olfactory epithelium in the nasal
turbinates of B6C3F1 rats exposed via inhalation to 1,2,3-trichloropropane 40
Table 4-23. Fertility indices and number of live pups/litter in breeding pairs of CD-I mice
exposed to 1,2,3-trichloropropane by gavage 43
Table 4-24. Comparison of tumor incidence and DNA-adduct formation in male F-344N rats and
B6C3F1 mice 50
October, 2007 v DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-25. Formation of DNA adducts by [14C]-1,2,3- trichloropropane (6 mg/kg-day)
administered to B6C3F1 mice by gavage or drinking water 52
Table 4-26. Genotoxicity bioassays of 1,2,3-trichloropropane 54
Table 5-1. Candidate benchmark doses for chronic and reproductive effects associated with oral
exposure to 1,2,3-trichloropropane 82
Table 5-2. Tumor incidence, (percent), and time of first occurrence in male rats following oral
gavage exposure to 1,2,3-trichloropropane (NTP, 1993) 99
Table 5-3. Tumor incidence, (percent), and time of first occurrence in female rats following oral
gavage exposure to 1,2,3-trichloropropane (NTP, 1993) 100
Table 5-4. Tumor incidence in male mice following oral gavage exposure to 1,2,3-
trichloropropane (NTP, 1993) 100
Table 5-5. Tumor incidence in female mice following oral gavage exposure to 1,2,3-
trichloropropane (NTP, 1993) 101
Table 5-6. Dose-response modeling summary for rat tumor sites associated with oral exposure to
1,2,3-trichloropropane; tumor incidence data from NTP (1993) 108
Table 5-7. Summary of cancer risk values estimated by R/BMDr and summed across tumor sites
for male and female rats. 109
Table 5-8. Summary of uncertainty in the 1,2,3-trichloropropane cancer risk assessment 114
Table B-l. Benchmark Dose modeling used in the derivation of the RfD 139
Table C-l: Tumor incidence data, with time to death with tumor; male rats exposed by gavage to
1,2,3-trichloropropane 170
Table C-2. Tumor incidence data, with time to death with tumor; female rats exposed to 1,2,3-
trichloropropane 177
Table C-3. Summary of cancer risk values estimated by R/BMDR and summed across tumor
sites for male and female rats. 186
October, 2007
vi
DRAFT - DO NOT CITE OR QUOTE

-------
LIST OF FIGURES
Figure 2-1. 1,2,3-Trichloropropane	3
Figure 3-1. Possible metabolic pathways for 1,2,3-trichloropropane in rats	7
Figure 4-1. Structure of the DNA adduct S-[l-(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione.
	51
Figure 5-1. Exposure-response array of selected subchronic, chronic, and reproductive toxicity
effects	87
Figure 5-2. Points of Departure for endpoints from Table 5-2 with corresponding applied
uncertainty factors and derived RfD	88
Figure 6-1. Points of Departure for endpoints from Table 5-1 with corresponding applied
uncertainty factors and derived RfD	120
October, 2007
vii
DRAFT - DO NOT CITE OR QUOTE

-------
LIST OF ABBREVIATIONS AND ACRONYMS
ACPC	N-acetyl-S-(3-chloro-2-hydroxypropyl)-L-cysteine
ALT	Alanine aminotransferase
AST	Aspartate aminotransferase
ATSDR	Agency for Toxic Substances and Disease Registry
BMDL	Benchmark dose, 95% lower bound
BSO	1 -Buthionine-(R, S)-sulfoximine
CASRN	Chemical Abstracts Service Registry Number
CBI	Covalent binding index
CPC	S-(3-chloro-2-hydroxypropyl)-L-cysteine
CYP450	Cytochrome P-450
DCA	1,3-Dichloroacetone
EPA	Environmental Protection Agency
GD	Gestation days
GMA	(S-glutathionyl)malonic acid
GSH	Reduced glutathione
HSDB	Hazardous Substances Data Bank
IRIS	Integrated Risk Information System
i.p.	Intraperitoneal
i.v.	Intravenous
K0w	Oil/water partition coefficient
LDH	Lactate dehydrogenase
LOAEL	Lowest-observed-adverse-effect-level
NADPH	Reduced nicotinamide adenine dinucleotide phosphate
NCI	National Cancer Institute
NOAEL	No-observed-adverse-effect-level
NRC	National Research Council
NTP	National Toxicology Program
PBTK	Physiologically-based toxicokinetic
PD	Postnatal day
RACB	Reproductive Assessment by Continuous Breeding
RfC	Reference concentration
RfD	Reference dose
SD	Standard deviation
SDH	Sorbitol dehydrogenase
S-G	S-glutathionyl
WHO	World Health Organization
October, 2007	viii	DRAFT - DO NOT CITE OR QUOTE

-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to 1,2,3-
trichloropropane. It is not intended to be a comprehensive treatise on the chemical or
toxicological nature of 1,2,3-trichloropropane.
The intent of Section 6, Major Conclusions in the Characterization of Hazard and Dose
Response, is to present the major conclusions reached in the derivation of the reference dose,
reference concentration, and cancer assessment and to characterize the overall confidence in the
quantitative and qualitative aspects of hazard and dose response by addressing the quality of the
data and related uncertainties. The discussion is intended to convey the limitations of the
assessment and to aid and guide the risk assessor in the ensuing steps of the risk assessment
process.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address).
October, 2007
ix
DRAFT - DO NOT CITE OR QUOTE

-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER/AUTHOR
Martin Gehlhaus, M.H.S
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
CONTRIBUTING AUTHORS
Stiven Foster, M.S.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Karen Hogan, M.S.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
George Holdsworth, Ph.D.
Oak Ridge Institute for Science and Education
Oak Ridge, TN
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.
INTERNAL EPA REVIEWERS
Bob Benson, Ph.D.
Region 8
Office of Partnerships and Regulatory Assistance (OPRA)
Joyce M. Donohue, Ph.D.
Office of Water
Office of Science and Technology (OST)
Health and Ecological Criteria Division (HECD)
October, 2007	x	DRAFT - DO NOT CITE OR QUOTE

-------
Lynn Flowers, Ph.D., DABT
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Channa Keshava, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Elizabeth H. Margosches, Ph.D.
Office of Prevention, Pesticides and Toxic Substances (OPPTS)
Office of Pollution Protection and Toxics (OPPT)
Risk Assessment Division (RAD)
John Whalan
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
October, 2007	xi	DRAFT - DO NOT CITE OR QUOTE

-------
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 1,2,3-
trichloropropane. IRIS Summaries may include oral reference dose (RfD) and inhalation
reference concentration (RfC) values for chronic and subchronic exposure durations, and a
carcinogenicity assessment.
The chronic RfD and chronic RfC provide quantitative information for use in risk
assessments for health effects known or assumed to be produced through a nonlinear (possibly
threshold) mode of action. The RfD (expressed in units of mg/kg-day) is defined as an estimate
(with uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
population (including sensitive subgroups) that is likely to be without an appreciable risk of
deleterious effects during a lifetime. The inhalation RfC (expressed in units of mg/m3) is
analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The
inhalation RfC considers toxic effects for both the respiratory system (portal-of-entry) and for
effects peripheral to the respiratory system (extrarespiratory or systemic effects). Reference
values may also be derived for acute (<24 hours), short-term (>24 hours up to 30 days), and
subchronic (>30 days up to 10% of average lifetime) exposure durations, all of which are derived
based on an assumption of continuous exposure throughout the duration specified.
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 derived from the application of a low-dose
extrapolation procedure. Route-specific risk values are presented in some cases. The "oral slope
factor" is an upper bound on the estimate of risk per mg/kg-day of oral exposure. Similarly, a
"unit risk" is an upper bound on the estimate of risk per (J,g/m3 air breathed.
Development of these hazard identification and dose-response assessments for 1,2,3-
trichloropropane has followed the general guidelines for risk assessment as set forth by the
National Research Council (1983). EPA guidelines and Risk Assessment Forum Technical
Panel Reports that may have been used in the development of this assessment include the
following: 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,
October, 2007 1 DRAFT - DO NOT CITE OR QUOTE

-------
1998a), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), Supplemental Guidance
for Assessing Susceptibility from Early-life Exposures to Carcinogens (U.S. EPA, 2005b),
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, 2006), Science Policy Council Handbook.
Risk Characterization (U.S. EPA, 2000b), Benchmark Dose Technical Guidance Document
(U.S. EPA, 2000c), and A 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 April 2007.
October, 2007
2	DRAFT - DO NOT CITE OR QUOTE

-------
2. CHEMICAL AND PHYSICAL INFORMATION
1,2,3-Trichloropropane (CASRN 96-18-4) is a three-carbon alkane with a single chlorine
atom attached to each carbon atom in the chain (Figure 2-1). Synonyms for the compound
include glyceryl trichlorohydrin, glycerol trichlorohydrin, and allyl trichloride. Some physical
and chemical properties are shown below (HSDB, 2005).
CI
Figure 2-1. 1,2,3-Trichloropropane.
Chemical Formula:
Molecular Weight:
Melting Point:
Boiling Point:
Density:
Water Solubility:
Log Kow:
Vapor Pressure:
Henry's Law Constant:
Conversion factors:
C3H5CI3
147.43
-14.7° C
156.85° C
1.3889 g/mL (at 20° C)
1750 mg/L (at 25° C)
1.98/2.27
3.1/3.69 mm Hg at 25° C
3.43 x 10"4 atm-m3/mol
1 ppm = 6.13 mg/m3; 1 mg/m3 = 0.16 ppm
Source: AT SDR, 1992: HSDB, 2005
1,2,3-Trichloropropane is used in the chemical industry as a solvent for oils and fats,
waxes, and resins (HSDB, 2005; ATSDR, 1992). The compound has also been used in paint
thinner and varnish remover, and as a degreasing agent. 1,2,3-trichloropropane is generated as a
by-product of the production of other chlorinated compounds such as epichlorohydrin (WHO,
2003). The compound is also used as an intermediate in the production of some pesticides and
polymers, such as polysulfide rubbers. The commercially available product is >98-99.9% pure.
October, 2007
3
DRAFT - DO NOT CITE OR QUOTE

-------
3. TOXICOKINETICS
No reports are available that address the toxicokinetics of 1,2,3-trichloropropane in
humans by any route of exposure. Experimental studies in rats and mice have demonstrated that
absorption of the compound via the oral route results in rapid distribution, extensive metabolism,
and clearance within 60 hours (Mahmood et al., 1991). The toxicokinetic data also demonstrate
the ability of 1,2,3-trichloropropane or metabolites to bind to intracellular macromolecules such
as proteins and nucleic acids (Mahmood et al., 1991; Weber and Sipes, 1990).
3.1. ABSORPTION
Data on the quantitative absorption of 1,2,3-trichloropropane from exposure via the
inhalation or dermal routes have not been reported. Quantitative data on the absorption,
distribution, and excretion following oral exposure to 1,2,3-trichloropropane were obtained from
a study in which rats and mice were treated with 14C-labeled compound by corn oil gavage
(Mahmood et al., 1991). Doses of 30 mg/kg (8-10 jaCi) [14C]-l,2,3-trichloropropane were
administered to 9 male and 12 female Fischer rats, and either 30 or 60 mg/kg to B6C3F1 male
mice (three/group). By sacrificing the animals at intervals up to 60 hours, the researchers
collected information on the time-dependent distribution of radiolabel in urine, feces, breath, the
principal organs and tissues, and bile.
Estimates for the percent absorption of the oral dose can be made by summing the mean
values for the radiolabel recovered in the urine and exhaled as CO2 (Table 3-1). By this
approach, estimates of the absorbed oral load are 75% in male rats, 68% in female rats, and 84%
in male mice. The percent recovered from feces was not used in this calculation because it is
likely to contain both an absorbed and non-absorbed fraction. However, the true extent of
intestinal absorption is likely to have been greater than the presented 75-84%, because a portion
of the radiolabel that appeared in feces, which was not included in the above absorption
estimates, would also have been absorbed.
October, 2007
4
DRAFT - DO NOT CITE OR QUOTE

-------
Table 3-1. Distribution and excretion of radiolabeled 1,2,3-trichloropropane
(30 mg/kg) 60 hours after oral (gavage
i administration
Tissue
Male rats
Female rats
Male mice
Urine
57.1±6.2a
49.8 ± 4.3 a
64.0 ± 5.5a
Feces
21.1 ±4.9
19.4 ±2.2
16.0 ±6.0
C02
17.7 ±0.4
18.5 ±0.6
20.2 ± 1.8
Volatiles
1.5 ±0.5
1.4 ±0.8
0.6 ±0.4
Blood
0.6 ±0.1
0.9 ±0.2
0.1 ±0.04
Liver
1.4 ±0.2
1.2 ±0.3
0.6 ±0.03
Kidney
0.3 ±0.1
0.3 ±0.1
0.1 ±0.01
Skin
1.1 ± 0.1
1.0±0.1
0.5 ±0.1
Adipose tissue
0.4 ±0.1
0.6 ±0.3
0.2 ±0.1
Muscle
1.1 ±0.3
1.0 ±0.4
1.0 ±0.2
a Percent of total dose (data are mean ± SD from three rats or mice).
Source: Mahmood et al., 1991.
3.2. DISTRIBUTION
Mahmood et al. (1991) examined the deposition of 30 mg/kg [2-14C]-l,2,3-
trichloropropane in rats and mice at three time points: 6, 24, and 60 hours post-administration.
After 6 hours, most of the radiolabel was found in the forestomach and glandular stomach with
smaller quantities in the intestines, adipose tissue, liver, and kidney. At 24 hours the
concentrations of radiolabel in the forestomach, intestines, liver, and kidney were similar. By
hour 60 the majority of the radiolabel had been excreted in the urine or feces with some residual
radioactivity sequestered predominantly in the liver, kidney, skin, muscle, and adipose tissue
(see Table 3-1). The radiolabel detected in tissues after 60 hours was generally not extractable,
suggesting that it was bound to macromolecules (Mahmood et al., 1991).
Volp et al. (1984) examined the time-dependent distribution of [1,3-14C]-1,2,3-
trichloropropane (2.1 mCi/mmol) in male Fischer rats (three rats per time point), following
intravenous (i.v.) injection of 3.6 mg/kg. Animals were maintained in metabolic cages and
sacrificed at the following time points: 15 and 30 minutes; 1, 2, 4, and 8 hours; and 1, 2, 4, and 6
days post-administration. Rapid distribution of the radiolabel was observed and 37% of the dose
was detected in adipose tissue 15 minutes after administration. After 4 hours, the largest portion
of the radiolabel was sequestered in the liver, primarily as metabolites.
Weber and Sipes (1990) administered intraperitoneal (i.p.) injections of 30 mg/kg (100
(j,Ci/kg) [2-14C ]-l,2,3-trichloropropane in vegetable oil to male Fischer rats. Groups of four rats
were sacrificed after 1, 4, 24, 48, and 72 hours. Maximal covalent binding of radiolabel to
October, 2007
5
DRAFT - DO NOT CITE OR QUOTE

-------
hepatic protein, approximately 600 pmol/mg, was observed at 4 hours post-administration.
Maximal covalent binding to hepatic DNA, approximately 250 pmol/mg, occurred at 24 hours.
After 72 hours the amount of radiolabel covalently bound to both hepatic protein and DNA was
at or below the levels found 1 hour after administration.
3.3. METABOLISM
No studies of 1,2,3-trichloropropane metabolism in humans have been reported. In vitro
data indicate that human microsomes, in the presence of reduced nicotinamide adenine
dinucleotide phosphate (NADPH), are capable of forming the DNA- reactive chemical 1,3-
dichloroacetone (DCA) from 1,2,3-trichloropropane (Weber and Sipes, 1992).
In rodents, 1,2,3-trichloropropane metabolism appears to involve oxidation catalyzed by
cytochrome P-450 (CYP) or glutathione conjugation, but specific details about the metabolic
process are unknown. Three potential routes for 1,2,3-trichloropropane metabolism (Figure 3-1)
have been proposed by Mahmood et al. (1991).
I) Nucleophilic displacement of a chlorine atom by glutathione creates a P-chlorothio
ether, and internal displacement of another chlorine creates an episulfonium ion. This
reactive ion could hydrolyze to a glutathione conjugate that can be cleaved to form N-
acetyl-S-(3-chloro-2-hydroxypropyl)-L-cysteine (ACPC) or S-(3-chloro-2-
hydroxypropyl)-L-cysteine (CPC). The reactive episulfonium ion could also react with
water to form P-chlorothio ether that could form a second episulfonium ion. This
second episulfonium ion could form 2-(S-glutathionyl)malonic acid (GMA) through
hydrolysis to form a 1,3-dihydroxypropyl glutathione conjugate and subsequent
oxidation.
October, 2007
6 DRAFT - DO NOT CITE OR QUOTE

-------
1,2,3-Trichloropropane
III
I
+GSH
II
' 2,3-Dichloropropanal
1
'

2-Chloroacrolein
+GSH
S-G
Glutathionyl-2,3- C|
dichloropropane
CI
1,3 -Dichloroacetone
O
Glutathionyl-
3-chloroacetone
episulfonium ion
H
+H20
episulfonium ion
+H-.0
CI
+GSH
CI
O
g_G Glutathionyl-1-
choropropan-3-ol
S-G
Glutathionyl- OH
hydroxypropanal
GMA* O
(bile)
HO
G-S'
Glutathionyl-3-
choropropan-2-ol
CI
S+
episulfonium ion
G-1.3-PD S-G
OH HO'
G-2.3-PD OH
S-G
S-G = S-glutathione
GSH = reduced glutathione
Nac = N-acetyl-L-cysteine
Cys = L-cysteine
* = Compound has been
detected in vivo
Nac'
ACPC- OH
(urine)
CPC* OH
(urine)
ACPC = N-acetyl-S-(3-chloro-2-hydroxypropyl)-L-cysteine
CPC = S-(3-chloro-2-hydroxypropyl)-L-cysteine
GMA = 2-(S-glutathionyl)malonic acid
G-1,3-PD = Glutathionyl-1,3-propanediol
G-2,3-PD = Glutathionyl-2,3-propanediol
Figure 3-1. Possible metabolic pathways for 1,2,3-trichloropropane in rats
Source: WHO, 2003; Mahmood et al., 1991.
II) Oxidation of 1,2,3-trichloropropane at the C2 position, possibly by CYP enzymes,
could lead to the formation of 1,3-dichloroacetone. Displacement of chlorine from 1,3-
dichloroacetone by glutathione and reduction of the keto group can result in the
formation of ACPC and CPC.
October, 2007	7	DRAFT - DO NOT CITE OR QUOTE

-------
Ill) Oxidation, possibly by CYP enzymes, of 1,2,3-trichloropropane at the CI position to
form 2,3-dichloropropanal. This chlorohydrin could undergo loss of HC1 to form
chloroacrolein, and then rearrange with glutathione to form an episulfonium ion. This
ion could then form 2-(S-glutathionyl)malonic acid (GMA) after the oxidation of the
C2 and C3 carbon atoms to form carboxylic acids.
Evidence for the involvement of CYP in 1,2,3-trichloropropane metabolism is provided
by the in vitro formation of 1,3-dichloroacetone when isolated rat or human hepatic microsomes
were incubated with 1,2,3-trichloropropane (Weber and Sipes, 1992). The formation of 1,3-
dichloroacetone, an intermediate in the formation of ACPC and CPC, occurred only in the
presence of reduced nicotinamide adenine dinucleotide phosphate (NADPH) and was enhanced
by the addition of such CYP inducers as phenobarbital and dexamethasone. Conversely,
formation of 1,3-dichloroacetone was blocked by the CYP inhibitors SKF-525A and 1-
aminobenzotriazol. In support of the Mahmood et al. (1991) scheme for the metabolic
transformation of 1,2,3-trichloropropane, the findings of Weber and Sipes (1990) provide
inferential evidence for the involvement of glutathione in 1,2,3-trichloropropane metabolism by
the demonstration that experimental glutathione depletion was associated with increased 1,2,3-
trichloropropane binding to hepatic protein and decreased binding to DNA.
3.4. ELIMINATION
Mahmood et al. (1991) and Volp et al. (1984) demonstrated that urine is the primary
route of 1,2,3-trichloropropane excretion in rats and mice. Mahmood et al. (1991) analyzed the
urine of F-344/N rats and male B6C3F1 mice treated with [2-14C]-l,2,3-trichloropropane by corn
oil gavage and found that the parent compound was extensively metabolized to either ACPC or
CPC. These investigators also documented that the principal biliary metabolite was GMA. In
rats, ACPC was the major urinary metabolite found 6 hours after exposure, accounting for
approximately 40% of the radiolabel recovered in males, and 10% in females. The urinary
metabolite associated with the largest fraction of radiolabel in both males and females 24 hours
post-administration could not be identified. However, substantial amounts of radiolabeled
ACPC and CPC were detected in urine at 24 hours. In male mice, ACPC accounted for only 3%
of the radiolabel at 6 hours (females were not tested). The major metabolites in male mice at
both 6 and 24 hours were not identified.
October, 2007
8
DRAFT - DO NOT CITE OR QUOTE

-------
Volp et al. (1984) examined the time-dependent distribution of [1,3-14C]-1,2,3-
trichloropropane in male Fischer rats (three rats per time point) following i.v. injection of 3.6
mg/kg. The data from this study demonstrated rapid excretion of the radiolabel; after 24 hours
30% of the initial radiolabel had been exhaled, 40% had been released in the urine, and 18% in
feces. Unchanged 1,2,3-trichloropropane was not detected in the urine.
Weber (1991) conducted a detailed analysis of urinary metabolites by employing proton
decoupled and two-dimensional homonuclear correlated nuclear magnetic resonance
spectroscopy following the coadministration of [l,2,3-13C]-trichloropropane and [2-14C]-
trichloropropane in soybean oil intraperitoneally to male F-344/N rats. This investigator
identified N-acetyl-S-(2-hydroxy-3-chloropropyl)cysteine, l,3-(2-propanol)-bis-S-(N-
acetylcysteine), N-acetyl-S-(2-hydroxy-2-carboxyethyl)cysteine, 2,3-dichloropropionic acid, 2-
chloroethanol, ethylene glycol, and oxalic acid as potential urinary metabolites of 1,2,3-
trichloropropane. It is unknown where in the metabolic pathway these additional urinary
metabolites may form.
3.5. PHYSIOLOGICALLY-BASED TOXICOKINETIC MODELING
Volp et al. (1984) developed a physiologically-based toxicokinetic (PBTK) model to
describe the time-dependent appearance of 1,2,3-trichloropropane and its metabolites in rat
tissues. The model consists of compartment-specific mass balance equations for tissues that
have physiological significance in storage, transport, and clearance. The model contains seven
compartments: blood, liver, kidney, adipose tissue, muscle, skin, and remaining distribution
volume, and describes the rapid disappearance of 1,2,3-trichloropropane from the blood with
biotransformation products concurrently appearing in the urine, bile, and expired air. High
concentrations of metabolites were also found in the liver and kidney, and the half-lives for
trichloropropane clearance from blood and liver were 23 and 40 hours, respectively.
October, 2007
9
DRAFT - DO NOT CITE OR QUOTE

-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS - EPIDEMIOLOGY, CASE REPORTS
Limited information from an acute inhalation study in humans (n = 12) demonstrated that
15 minute exposures to 100 ppm trichloropropane (purity unknown) resulted in irritation of the
nose, eyes, and throat of all subjects tested (Silverman et al., 1946). No occupational,
epidemiology, or case study data were identified that were applicable to 1,2,3-trichloropropane
exposure in humans.
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS - ORAL AND INHALATION
4.2.1. Oral Exposure
4.2.1.1. Subchronic Studies
Hazelton Laboratories (1983a, b) conducted a series of subchronic toxicity studies of
1,2,3-trichloropropane in F-344/N rats and B6C3F1 mice. The findings of these subchronic
studies were included in the National Toxicology Program technical report on the toxicology and
carcinogenesis of the compound and published in the peer-reviewed literature (NTP, 1993).
The same protocol was used for both the rat and the mouse studies. 1,2,3-
Trichloropropane was administered by corn oil gavage 5 days/week for 120 days at doses of 0, 8,
16, 32, 63, 125, or 250 mg/kg-day. Treatment groups contained 20 animals/ sex and the vehicle
control group contained 30 animals/ sex. Half of the animals in each group were sacrificed after
8 weeks, and the rest were maintained until week 17. Animals were examined twice daily for
clinical signs of toxic stress. Animals were weighed at the start of the study and at weekly
intervals during the course of the study. Blood and urine samples were obtained from animals
during weeks 8 and 17. Blood samples were analyzed for hematocrit, hemoglobin, and blood
cell counts. A limited suite of clinical chemistry parameters was also evaluated. Specific
gravity of the urine specimens was determined. Necropsies were performed on all animals with
complete histopathologic examinations performed on all animals that had died during the study,
moribund animals that were sacrificed during the study, all rats receiving a dose of 125 mg/kg,
October, 2007
10
DRAFT - DO NOT CITE OR QUOTE

-------
and all controls. A number of organs and tissues were excised and collected from all animals.
Tissue weights were reported for the 17-week study only.
In the rat study, 12 males that received 250 mg/kg-day died, or were sacrificed moribund,
during the first week of treatment. Six males died during the second week and the remaining
two animals were terminated in weeks 3 and 5. Sixteen females in the 250 mg/kg-day dose
group died, or were sacrificed moribund, during the first week. The remaining four animals in
this treatment group died during the second week. One male and four female rats that received
125 mg/kg-day 1,2,3-trichloropropane died, or were sacrificed moribund, during the study.
During their brief survival period, rats in the 250 mg/kg-day treatment group were noted
to have been emaciated, lethargic, and debilitated. No clinical signs of toxicosis were observed
in any of the other treatment groups. Dose-dependent reductions in body weight gain were
observed in both males and females. Mean final body weights were significantly reduced for
male rats that received 63 and 125 mg/kg-day and females treated with 125 mg/kg-day 1,2,3-
trichloropropane. Whole body and tissue weights were not reported for the 250 mg/kg-day
treatment groups. At the 17-week sacrifice, mean weight gain in the 125 mg/kg-day treatment
group was reduced by 43% and 60% for males and females, respectively, compared with
controls.
Mean relative liver weights were statistically (P < 0.01) significantly increased in males
that received 32, 63, or 125 mg/kg-day by 24%, 47%, and 78%, respectively, compared with
controls (Table 4-1); while absolute liver weights statistically significantly (p<0.01) increased
10%) to 36% in males receiving 8 to 125 mg/kg-day (Table 4-2). Mean relative liver weights,
when compared with controls, were statistically (P < 0.01) significantly increased by 12%, 18%,
37%), and 105% in females receiving 16, 32, 63, or 125 mg/kg-day, respectively; while absolute
liver weights statistically significantly (p<0.05) increased 17% to 61% in females receiving 16 to
125 mg/kg-day. Mean relative right kidney weights were statistically significantly increased in
males that received 32, 63, or 125 mg/kg-day by 12% (P < 0.05), 26% (P < 0.01), and 54% (P <
0.01), respectively, compared with controls; while absolute right kidney weights were
statistically significantly (p<0.01) increased 5 to 19% in males receiving 32, 63, or 125 mg/kg-
day. In females that received 63 and 125 mg/kg-day 1,2,3-trichloropropane, mean relative right
kidney weights were statistically (P < 0.01) significantly increased 32% and 43%, respectively;
while absolute right kidney weights statistically significantly (p<0.01) increased 11 to 25% in
females receiving 63 to 125 mg/kg-day. NTP (1993) considered the changes in relative organ
weights to be associated with the change in body weight, and not with organ toxicity. Absolute
October, 2007
11
DRAFT - DO NOT CITE OR QUOTE

-------
heart weight was statistically significantly (p<0.01) decreased 21% in male rats at 125 mg/kg-
day.
Table 4-1. Relative organ weight changes in F-344/N rats receiving 1,2,3-
trichloropropane by gavage for 120 days		
Dose
(mg/kg-day)
Change in mean relative liver
weight
Change in mean relative right
kidney weight
Male
Female
Male
Female
8
9%a
7%a
-l%a
7%a
16
12%
12%d
5%
7%
32
24%d
18%d
12%c
10%
63
47%d
37%d
26%d
32%d
125
78%d
105%d
54%d
43%d
250
NRb
NR
NR
NR
a Calculated as the percent change from the control mean.
b NR = Due to the rapid onset of mortality, organ weights were not recorded for the
high dose group.
0 showing statistically significant differences (P < 0.05) from the control group by
Williams' orDunnett's test
d showing statistically significant differences (P < 0.01) from the control group by
Williams' orDunnett's test
Source: NTP, 1993.
Table 4-2. Absolute organ weight changes in F-344/N rats receiving 1,2,3-
trichloropropane by gavage for 120 days		
Dose
(mg/kg-day)
Change in mean absolute liver
weight
Change in mean absolute right
kidney weight
Male
Female
Male
Female
8
1 l%ad
7%a
l%a
5%a
16
10%d
18%d
2%
11%
32
26%d
17%c
15%d
11%
63
23%d
32%d
5%d
25%c
125
19%d
61%d
19%d
11%C
250
NR
NR
NR
NR
a Calculated as the percent change from the control mean.
b NR = Due to the rapid onset of mortality, organ weights were not recorded for the
high dose group.
0 showing statistically significant differences (P < 0.05) from the control group by
Williams' orDunnett's test
d showing statistically significant differences (P < 0.01) from the control group by Williams' or
Dunnett's test
Source: NTP, 1993.
An increased incidence of lesions, as described below, was observed in the liver, kidney,
and nasal turbinates of rats receiving 125 mg/kg-day 1,2,3-trichloropropane for 120 days (Table
October, 2007	12	DRAFT - DO NOT CITE OR QUOTE

-------
4-3). A time-dependent increase in the number of lesions was noted between the 8-week and 17-
week evaluations in the 125 mg/kg-day treatment group. This same pattern was not observed in
the 250 mg/kg-day treatment group since the majority of animals did not survive more than one
week.
The liver lesions in rats were characterized by multifocal, centrilobular hepatocellular
necrosis, with karyomegaly, hemorrhage, and bile duct hyperplasia. Hepatic necrosis was
observed in female rats (7/9) receiving 125 mg/kg-day and in all of the rats receiving 250 mg/kg-
day 1,2,3-trichloropropane (20/20 males and 20/20 females) at the time of their death. In the 17-
week evaluation, hepatic necrosis was observed at terminal sacrifice in 1/10 males and 11/11
females treated with a dose of 125 mg/kg-day, with liver necrosis also evident in 1/10 male rats
at 32 and 63 mg/kg-day.
The kidney lesions in the rats were characterized by early diffuse acute tubule necrosis or
regenerative hyperplasia, karyomegaly of epithelial cells, and multifocal necrosis. Renal tubular
necrosis was observed during the 8-week interim evaluation in 14/20 males and 20/20 females
treated with 250 mg/kg-day that died at or before the interim sacrifice. At the 17-week
evaluation, renal necrosis was observed in 1/10 males and 0/11 females treated with a dose of
125 mg/kg.
Lesions of the nasal turbinates included multifocal necrosis and epithelial attenuation,
subepithelial fibrosis, and inflammation. Epithelial necrosis of the nasal turbinates was observed
during the 8-week interim evaluation in 14/20 males and 19/20 females treated with 250 mg/kg-
day that died at or before the interim sacrifice. At the time of death or at the 17-week evaluation,
epithelial necrosis of the nasal turbinates was observed in 3/9 males and 2/11 females treated
with 125 mg/kg-day.
October, 2007
13
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-3. Incidence of liver, kidney, and nasal turbinate lesions in male and female
F-344N rats in 17-week study
Endpoint
Dose (mg/kg-day)
0
8
16
32
63
125
Males
Liver necrosis a
0/20
0/10
0/10
1/10
1/10
1/10
Kidney necrosis a
0/20
0/10
0/10
0/10
0/10
1/10
Epithelial necrosis
of nasal turbinatesa
0/20
0/10
0/10
0/10
0/10
3/9b
Females
Liver necrosis a
0/20
0/10
0/10
0/10
0/10
11/1lc
Kidney necrosis a
0/20
0/10
0/10
0/10
0/10
0/11
Epithelial necrosis
of nasal turbinatesa
0/10
0/10
0/10
0/10
0/10
2/11
a incidence is the number of animals in which lesion was found/ number of animals in which tissue was
examined.
b showing statistically significant differences (P < 0.05) from the control group by Fisher exact test.
0 showing statistically significant differences (P < 0.01) from the control group by Fisher exact test.
Source: NTP, 1993.
A number of clinical chemistry parameters in rats were statistically significantly affected
upon exposure to 1,2,3-trichloropropane. Blood samples were not obtained from animals in the
250 mg/kg-day treatment group. Effects observed were predominantly biomarkers for liver
damage. At the 8-week interim evaluation, the activities of alanine aminotransferase (ALT),
sorbitol dehydrogenase (SDH), and aspartate aminotransferase (AST), were all statistically (P <
0.01) significantly elevated, 1200%, 433%, and 1000%, respectively, over controls in females
that received 125 mg/kg-day. Total bilirubin levels in female rats at the 8-week evaluation
increased 50 and 150% at the doses of 63 and 125 mg/kg-day, respectively. At the 17-week
evaluation, ALT and SDH activities were statistically [(P < 0.05) and (P < 0.01), respectively]
significantly elevated, 248% and 317%, respectively, over controls in females treated with 125
mg/kg-day.
The activity of ALT was statistically (P < 0.05) significantly elevated in males treated
with 125 mg/kg-day at week 8 but not at week 17, while the activity of SDH in males at 17
weeks was statistically significantly (P < 0.05) increased 25 and 12.5% at 63 and 125 mg/kg-day,
respectively. NTP (1993) stated that the increase in ALT and SDH was indicative of
hepatocellular damage with subsequent enzyme leakage. The only clinical chemistry parameter
that was consistently impacted in both males and females at both time points was
pseudocholinesterase (serum carboxylesterase). Activity of this hepatic enzyme decreased in
both species with increasing dose and NTP (1993) suggested that the depressed synthesis of
pseudocholinesterase was due to hepatocellular damage. A statistically significant decrease was
October, 2007
14
DRAFT - DO NOT CITE OR QUOTE

-------
observed at both time points (8 and 17 weeks) evaluated in females at the lowest dose tested,
21% and 14% at 8 mg/kg-day (P < 0.01), and 9% and 8% in males that received 32 mg/kg-day (P
<0.05).
In rats, hematocrit, hemoglobin, and erythrocyte counts were statistically significantly
decreased by 1,2,3-trichloropropane treatment, but were not considered in this analysis to be
biologically significant. At the 8-week sacrifice, hematocrit and red blood cell counts were
significantly depressed, 13 to 23% and 10 to 18%, respectively, in males that received doses of
16 mg/kg-day or higher and in females that received doses of 8 mg/kg-day or higher.
Hemoglobin was statistically significantly decreased 5-9% in male rats that received doses of 16
mg/kg-day or higher and female rats that received 63 mg/kg-day or higher.
The 17-week, less-than-lifetime rat study was conducted to determine appropriate doses
for the two-year, 1,2,3-trichloropropane study in rats (NTP, 1993), described later in this
document. NTP considered the dose-response of the increased liver and kidney weights to be
consistent with the clinical pathological and histopathological findings in the liver and kidney.
The NOAEL and LOAEL for hepatocellular necrosis in male rats at 17-weeks were 16 and 32
mg/kg-day, and in females rats at 17-weeks were 63 and 125 mg/kg-day. The NOAEL and
LOAEL for renal tubular necrosis in male rats at 17-weeks were 63 and 125 mg/kg-day,
respectively, while the NOAEL for renal tubular necrosis in females was 125 mg/kg-day. For
epithelial necrosis of the nasal turbinates, the NOAEL and LOAEL in male and female rats at
17-weeks were 63 and 125 mg/kg-day, respectively. A decrease in pseudocholinesterase (serum
carboxylesterase) activity in males presented a NOAEL of 16 mg/kg-day and LOAEL of 32
mg/kg-day; whereas females had a LOAEL of 8 mg/kg-day. The critical effect is hepatocellular
necrosis in male rats, with a NOAEL of 16 mg/kg-day and a LOAEL of 32 mg/kg-day.
In the NTP (1993) subchronic, B6C3F1 mouse study, which used the same protocol as
the rat study above, 16 males that received 250 mg/kg-day 1,2,3- trichloropropane died, or were
sacrificed moribund, by week 4. Among the females that received a dose of 250 mg/kg-day,
seven died by week 2, and there was an additional death in week 17 (prior to the terminal
sacrifice). One male mouse and 6 female mice were sacrificed at the 8-week interim evaluation.
At the end of the 17-week evaluation, 2 out of 10 males at the highest dose were still alive,
where as 7 out of 10 females, tallied before the death of single female during week 17, survived
the full evaluation period.
Mean body weight gain in male mice at 250 mg/kg-day was significantly reduced,
although the overall mean weight gains among male and female mice at the various doses were
October, 2007
15
DRAFT - DO NOT CITE OR QUOTE

-------
similar. At week 17 a statistically significant (p<0.01) increase in relative and absolute liver
weights was observed in males and females that received a dose of 125 mg/kg-day or higher.
Mean relative liver weights were increased by 12% and 32% in males receiving 125 and 250
mg/kg-day, respectively, compared to controls (Table 4-4); while absolute liver weights were
statistically significantly (p<0.05) increased 14%, 4%, 22%, and 25% at 32, 63, 125, and 250
mg/kg-day (Table 4-5). Mean relative liver weights were increased by 12% and 22% in females
receiving 125 and 250 mg/kg-day, respectively, compared to controls; while absolute liver
weights were statistically significantly (p<0.05) increased at 125 and 250 mg/kg-day 24% at both
doses. Mean relative right kidney weights in female mice were statistically significantly
(p<0.01) decreased 17, 13, 11, 17, and 14% at 16, 32, 63, 125, and 250 mg/kg-day, respectively,
after 120 days; while absolute right kidney weights were statistically significantly (p<0.05)
decreased 13% at 250 mg/kg-day. The changes in relative and absolute right kidney weights in
male mice did not follow a clear dose-response pattern.
Mean relative heart weights in males were statistically significantly (p<0.05) decreased
14%), 14%), 11%), 19%), 22% and 22% at 8, 16, 32, 63, 125, and 250 mg/kg-day, respectively,
compared to controls. Absolute heart weights in males were statistically significantly (p<0.01)
reduced 14-25%) at 63 mg/kg-day and higher. Relative brain weights in male mice were
statistically significantly (p<0.05) decreased at 16 mg/kg-day to 125 mg/kg-day, with the
decrease ranging from 6% to 11%. Mean relative heart weights in females were statistically
significantly (p<0.05) reduced 19%, 17%, 11%, 19%), and 21% at 16, 32, 63, 125, and 250
mg/kg-day. Absolute heart weights in females were statistically significantly (p<0.01) decreased
25%) at 250 mg/kg-day. Absolute and relative brain weights were statistically significantly
(p<0.01) decreased 6-15% in females receiving 16 mg/kg-day or more.
October, 2007
16
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-4. Relative organ weight changes in B6C3F1 mice receiving 1,2,3-
trichloropropane by gavage for 120 days		
Dose (mg/kg-day)
Change in mean relative liver
weight
Change in mean relative right
kidney weight
Male
Female
Male
Female
8
l%a
l%a
3%a
-2%a
16
-3%
-8%
1%
-17%c
32
5%
3%
5%
-13%c
63
0%
4%
-10%
-11%C
125
10%c
12%c
-3%
-17%c
250
30%c
22%c
1%
-14%c
a Calculated as the percent change from the control mean.
b showing statistically significant differences (P < 0.05) from the control group by Williams' or
Dunnett's test
0 showing statistically significant differences (P < 0.01) from the control group by Williams' or
Dunnett's test
Source: NTP, 1993.
Table 4-5. Absolute organ weight changes in B6C3F1 mice receiving 1,2,3-
trichloropropane by gavage for 120 days		

Change in mean absolute liver
Change in mean absolute right
Dose (mg/kg-day)
weight
kidney weight

Male
Female
Male
Female
8
8%a
0%a
9%a
-2%a
16
3%
4%
7%
-2%
32
14%b
5%
14%
-10%
63
4%b
11%
-7%
-6%
125
22%c
24%c
6%
-7%
250
25%c
24%c
-3%
-13%b
a Calculated as the percent change from the control mean.
b showing statistically significant differences (P < 0.05) from the control group by Williams' or
Dunnett's test
0 showing statistically significant differences (P < 0.01) from the control group by Williams' or
Dunnett's test
Source: NTP, 1993.
Complete histopathological examinations were conducted on all control animals and
mice receiving 125 or 250 mg/kg-day, and mice designated for the interim evaluation that died
during the study were included in the group of animals examined at the end of the 17-week
study. Forestomach and lung lesions in mice were observed at both the 8-week interim
evaluation and the 17-week terminal evaluation (Table 4-6). At the 8-week evaluation, male
mice displayed lung and forestomach lesions at 125 mg/kg-day in 1/8 and 6/8 mice, respectively;
whereas, female mice displayed lung and forestomach lesions at 250 mg/kg-day in 5/6 and 6/6
October, 2007
17
DRAFT - DO NOT CITE OR QUOTE

-------
mice, respectively. Lung and forestomach lesions were found in the one mouse from the 250
mg/kg-day dose group that was examined at the 8-week interim sacrifice.
Regenerative lung lesions were observed in 9/12 male mice and 10/12 female mice, and
hyperkeratosis of the forestomach in 7/12 male mice and 9/12 female mice receiving 125 mg/kg-
day 1,2,3-trichloropropane at the 17-week evaluation. Lung lesions in male and female mice at
250 mg/kg-day 1,2,3-trichloropropane were observed in 14/19 males and 7/14 females, while
forestomach lesions in the same dose group were observed in 4/19 males and 8/14 females. At
63 mg/kg-day, female mice displayed lung lesions (7/9) and forestomach lesions (7/9).
Hyperkeratosis of the forestomach was attributed to continued irritation resulting from the
gavage treatments and not considered to be life-threatening (Hazelton Laboratories, 1983b).
Focal or multifocal desquamation of necrotic cells in the airways, flattened epithelium with loss
of differentiated cells, and thickened epithelium with an increase in goblet cells (hyperplasia)
were characteristic of the regenerative lung lesions (NTP, 1993).
Liver lesions were observed at both the 8-week interim and 17-week terminal sacrifice.
At the 8-week evaluation liver lesions were not observed in the only examined male mouse that
received 250 mg/kg-day 1,2,3-trichloropropane, but hepatic necrosis was observed in 4/6
females that received this dose. Hepatic necrosis at the 8-week evaluation was observed in 6/8
males and 0/8 females that received 125 mg/kg-day. No liver lesions were observed in the 8-
week controls.
At the 17-week evaluation, liver necrosis was observed in 14/19 males, most of which
died prior to 8-week evaluation, and 5/14 females that received 250 mg/kg-day, and 1/10 male
and 0/10 female controls (Table 4-6). Hepatocelluar degeneration associated with fatty change
and karyomegaly was also observed in 11/19 males and 1/14 females of the high dose group.
Also at the 17-week evaluation, liver lesions in mice at the 125 mg/kg-day dose occurred in 1/12
males and 1/12 females.
October, 2007 18 DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-6. Incidence of liver, lung, and forestomach lesions in male and female
B6C3F1 mice in 17-week study
Endpoint
Dose (mg/kg-day)
0
8
16
32
63
125
250
Males
Liver necrosis a
1/10
0/10
0/10
0/10
0/10
1/12
14/19C
Liver
karyomegaly a
0/10
0/10
0/10
0/10
0/10
1/12
1 l/19c
Lung lesions-
regenerative a
0/10
0/10
0/10
0/10
0/10
9/12c
14/19C
Hyperkeratosis of
the forestomacha
0/10
0/10
0/10
0/10
0/10
7/12c
4/19
Fema
es
Liver necrosis a
0/10
0/10
0/10
0/10
0/9
1/12
5/14b
Liver
karyomegaly a
0/10
0/10
0/10
0/10
0/9
0/12
1/14
Lung lesions-
regenerative a
0/10
0/10
0/10
0/10
7/9c
10/12C
7/14c
Hyperkeratosis of
the forestomacha
0/10
0/10
0/10
0/10
7/9°
9/12c
8/14c
a incidence is the number of animals in which lesion was found/ number of animals in which tissue was
examined.
b showing statistically significant differences (P < 0.05) from the control group by Fisher exact test.
0 showing statistically significant differences (P < 0.01) from the control group by Fisher exact test.
Source: NTP, 1993.
Differences in clinical chemistry parameters in mice administered 1,2,3-trichloropropane
for 17 weeks were not considered by the NTP investigators to be treatment related. Several
statistically significant changes were observed among hematological parameters; however, these
changes were not considered to be biologically significant and failed to follow a consistent dose-
response pattern. Hematocrit values were statistically significantly decreased at week 8 in
female mice that received 8 and 250 mg/kg-day. At week 17, hematocrit values were
statistically significantly decreased in female mice that received 16, 32, 125, or 250 mg/kg-day
1,2,3-trichloropropane. In male mice, a statistically significant decrease in hematocrit values
was observed only at week 8 in the 63 and 125 mg/kg-day treatment groups.
The 17-week, less-than-lifetime mouse study was conducted to determine appropriate
doses for the two-year, 1,2,3-trichloropropane study in mice (NTP, 1993), described later in this
document. The dose-related increased liver weights were consistent with the histopathological
results, while the hematological data were not associated with 1,2,3-trichloropropane
administration (NTP, 1993). The NOAEL and LOAEL for regenerative lung lesions at the 17-
week evaluation were 63 and 125 mg/kg-day for male mice and 32 and 63 mg/kg-day for female
mice. The NOAEL and LOAEL for liver lesions at the 17-week evaluation were 63 and 125
October, 2007 19 DRAFT - DO NOT CITE OR QUOTE

-------
mg/kg-day for both male and female mice. The critical affect is liver necrosis in male and
female mice, with a NOAEL of 63 mg/kg-day and a LOAEL of 125 mg/kg-day.
Merrick et al. (1991) administered 1,2,3-trichloropropane in corn oil to Sprague-Dawley
rats by gavage for 90 days. Groups of 10 males and 10 females received 0, 1.5, 7.4, 15, or 60
mg/kg-day. Animals that received 60 mg/kg-day exhibited a 14%—19% reduction in mean body
weight gain when compared to controls. Relative liver weights were statistically (p<0.05)
significantly increased after 90 days in animals that received 15 or 60 mg/kg-day, and relative
kidney weights were statistically (p<0.05) significantly increased after 90 days in males that
received 60 mg/kg-day and females that received 15 or 60 mg/kg-day. Relative brain and testes
weights were statistically (p<0.05) significantly increased in males from the high dose group.
Organ/ body weight ratios were reported graphically.
Female rats that received 60 mg/kg-day 1,2,3-trichloropropane exhibited elevated ALT
and AST levels. Mean serum concentrations for these two enzymes appeared to be
approximately doubled in the high dose females, but the actual magnitude of this effect could not
reliably be estimated from the graphical presentation of the data. Hematological parameters,
which included hemoglobin, hematocrit, and erythrocyte counts, were stated to be unremarkable
(data not provided).
An increased incidence of inflammation-associated myocardial necrosis was observed in
6/10 males and 7/10 females that received 60 mg/kg-day 1,2,3-trichloropropane (Table 4-7).
These lesions were marked by intense eosinophilic staining with necrotic cells containing
granulated or vacuolated cytoplasm and associated macrophages or polynuclear leukocytes.
Myocardial necrosis was also observed in a smaller number of animals from all other treatment
groups; no myocardial lesions were observed in the control group.
Bile duct hyperplasia was observed in the livers of one control male and 4/10 males and
8/10 females in the high dose group. Other proliferative and neoplastic lesions observed in high-
dose animals included a forestomach squamous cell papilloma, forestomach squamous cell
hyperplasia, a hepatocellular adenoma, and plasma cell hyperplasia in the mandibular lymph
node, with the latter displaying an increased dose-response relationship in both male (2/10, 0/10,
1/10, 0/10, 9/10) and female rats (1/10, 1/10, 2/10, 3/10, 5/10).
Table 4-7. Incidence of myocardial necrosis in male and female Sprague-Dawley
rats following 90-day 1,2,3-trichloropropane exposure
Endpoint
Sex
Dose
0
1.5
7.4
15
60
October, 2007
20
DRAFT - DO NOT CITE OR QUOTE

-------
Myocardia
1 necrosis
Male
0/10a
2/10a
l/10a
2/10a
6/10a
Female
0/10
0/10
1/10
0/10
7/10
a incidence is the number of animals in which lesion was found/ number of animals in which tissue was
examined.
Source: Merrick et al., 1991.
Inflammation-associated myocardial necrosis was seen in all male dose groups, so the
LOAEL for this affect is 1.5 mg/kg-day. The NOAEL and LOAEL for bile duct hyperplasia are
15 and 60 mg/kg-day. The NOAEL and LOAEL for plasma cell hyperplasia in the mandibular
lymph node is 1.5 and 7.4 mg/kg-day for male rats, while the LOAEL is 1.5 mg/kg-day for
female rats.
Villeneuve et al. (1985) administered 1,2,3,-trichloropropane in drinking water to
Sprague-Dawley rats. Ten rats/ sex/ group were exposed 7 days/week for 90 days to 0, 1, 10,
100, or 1000 mg/L. Drinking water contained 0.5% Emulphor to assure adequate solubility of
the test chemical. Two groups of control animals were employed; one received tap water, and
the other received a 0.5% Emulphor solution. Body weight and water intake values were used
for females in the 100 and 1000 mg/L exposure groups to calculate delivered doses of 18 and
149 mg/kg-day, respectively. The delivered dose for males in the 1000 mg/L exposure group
was calculated to be 113 mg/kg-day. Clinical signs were monitored daily, and body weights
were recorded weekly. At termination, the brain, liver, kidney, heart, and spleen were excised
and weighed. A number of hematological and clinical chemistry parameters were evaluated in
blood samples obtained at sacrifice. Each animal was subjected to a full necropsy, and tissues
and organs were obtained for histopathologic examination. In addition, the specific activities of
some mixed-function oxidases, including aniline hydroxylase and aminopyrine demethylase,
were measured in liver homogenates.
Three animals died during the course of the study, but their deaths were not considered to
be treatment-related. Mean body weight gain was reduced by approximately 30% in male and
female rats that were exposed to 1000 mg/L 1,2,3-trichloropropane, when compared with both
controls (p<0.05) and vehicle controls (p<0.05). No difference in absolute organ weights was
observed. Relative liver and kidney weights were reportedly increased in males that were
exposed to 1000 mg/L by 22% and 27%, respectively, when compared to vehicle controls. Mean
relative liver weights were apparently increased 6% and 17% in females that were exposed to
100 and 1000 mg/L, respectively. Mean relative kidney weights in females were reportedly
increased 14% and 34% in the 100 and 1000 mg/L treatment groups, respectively. Mean relative
brain weights for the 1000 mg/L exposure groups were reportedly increased by 21% and 23% in
males and females, respectively. Mean serum cholesterol levels were apparently increased 55%
October, 2007 21 DRAFT - DO NOT CITE OR QUOTE

-------
in female rats exposed to 1000 mg/L and no effect on cholesterol was observed in males.
Hepatic aminopyrine demethylase activity was reportedly significantly increased in males and
females that were exposed to 1000 mg/L. Aniline hydroxylase activity was apparently
significantly increased in males that received 1000 mg/L.
Mild, but significant, histomorphological changes were reported in the liver, including
anisokaryosis, accentuated zonation, and fatty vacuolation; kidney, including eosinophilic
inclusions, pyknosis, nuclear displacement, fine glomerular adhesions and interstitial reactions
and histologic proteinuria; and thyroid, including angular collapse of follicles, reduction in
colloid density, and increased epithelial height, of both sexes of rats in the highest exposure
group, although the number of affected animals was not reported. Biliary hyperplasia was also
noted in females at 1000 mg/L. Treatment with 1,2,3-trichloropropane also caused liver and
kidney enlargement, as well as increased serum cholesterol levels and hepatic mixed-function
oxidase activity. Mean lymphocyte and neutrophil counts were depressed by approximately
40% in male rats exposed to 1000 mg/L, but were still within the historical reference range for
Sprague-Dawley rats from the laboratory. The critical effect for this investigation is the
histological changes in the liver, kidney, and thymus, with a NOAEL of 15-20 mg/kg-day and a
LOAEL of 113-149 mg/kg-day.
4.2.1.2. Chronic Studies
NTP (1993) conducted a 2-year study of the toxicity and carcinogenicity of 1,2,3-
trichloropropane in F-344/N rats, the data of which was also published in Irwin et al. (1995).
The chemical was administered by corn oil gavage to 60 rats/sex/group. Rats received doses of
0, 3, 10, or 30 mg/kg-day, and after 15 months (65-67 weeks), 8 to 10 rats per group were
sacrificed to allow an interim evaluation of all toxicological parameters and histopathology. Due
to high mortality in rats receiving 30 mg/kg at the interim evaluation, the remaining survivors in
that group were sacrificed at week 67 (females) and week 77 (males). Due to the early
termination of this treatment group, organ weights and hematology data were only obtained at
the interim sacrifices.
Clinical observations were made twice daily; while body weights were recorded weekly
for 13 weeks and then monthly (NTP, 1993). As mentioned above, up to 10 rats/ group were
sacrificed at month 15. From this interim sacrifice blood samples were obtained for hematology
and clinical chemistry analyses. Hematological parameters included hematocrit, hemoglobin,
and counts of erythrocytes, leukocytes, and differential leukocytes. Clinical chemistry
parameters included the serum levels of ALT, AST, creatine kinase, lactate dehydrogenase
October, 2007
22
DRAFT - DO NOT CITE OR QUOTE

-------
(LDH), sorbitol dehydrogenase (SDH), and 5'-nucleotidase. Whether at the planned sacrifice or
as each rat died or became moribund, all rats were subjected to a gross necropsy, and a full range
of organs and tissues was processed for histopathologic examination. Hematology, clinical
chemistry, and tissue weight data were obtained only from rats that were sacrificed at the 15-
month interim because the majority of treated animals died prior to the end of the study.
Survival rates were statistically significantly reduced (p<0.001) in rats that received 10 or
30 mg/kg-day 1,2,3-trichloropropane (Table 4-8). An effect on survival was apparent, as the 10
and 30 mg/kg-day groups of rats died or were sacrificed moribund prior to or soon following the
15-month interim evaluation. The mortality in rats was attributed to cancer associated with
chemical exposure (NTP, 1993).
Table 4-8. Survival rates and percent probability of survival for F-344/N rats
exposed to 1,2,3-trichloropropane by gavage for two years
F-344/N Ral
ts
Dose
(mg/kg-day)
Males
Females
0
34/49a
70b
31/50b
62a
3
32/50
64
30/49
62
10
14/48
30c
8/52
16c
30
0/52
0C
0/52
0C
a Animals surviving to study termination and number of animals in the treatment group. Accidental deaths
were excluded and censored from survival analysis.
b Kaplan-Meier determinations of percent probability of survival at end of study.
c /K0.001.
Source: NTP, 1993.
In rats, the mean body weights of males and females receiving doses of 3 or 10 mg/kg-
day, observed throughout the study, appeared similar to the mean body weights of corresponding
control rats; the mean body weights of the high-dose males and females, however, appeared
lower than the control rat body weights (NTP, 1993). Statistically significant increases (p<0.05)
in absolute liver weights were observed in male and female rats exposed for 15 months to doses
of 3 mg/kg-day 1,2,3-trichloropropane or higher. Absolute liver weights were significantly
increased by 10%, 18%, and 38% in male rats and 14%, 16%, and 34% in female rats that
received doses of 3, 10, and 30 mg/kg-day, respectively (Table 4-9b). Mean relative liver
weights were significantly increased by 15% and 28% in male rats that received doses of 10 or
30 mg/kg-day, respectively, when compared with controls (Table 4-9a). Mean relative liver
weights in female rats that received doses of 10 or 30 mg/kg-day were increased 12% and 40%,
respectively (Table 4-9a).
October, 2007
23
DRAFT - DO NOT CITE OR QUOTE

-------
Statistically significant increases (p<0.05) in absolute right kidney weights were
observed in male rats exposed for 15 months to doses of 3 mg/kg-day 1,2,3-trichloropropane or
higher and female rats exposed to doses of 10 mg/kg-day or higher. Absolute kidney weights
were significantly increased by 8%, 12%, and 30% in male rats that received doses of 3, 10, and
30 mg/kg-day, and significantly increased by 11% and 24% in female rats that received doses of
10 and 30 mg/kg-day (Table 4-10b). Mean relative kidney weights in males from these
treatment groups were increased by 4%, 10% and 29%, respectively (Table 4-10a). Mean
relative kidney weights of females in the 10 and 30 mg/kg-day treatment groups were increased
by 8%> and 31% (Table 4-10a).
Table 4-9a. Relative liver weights (mg organ weight/ g body weight) and percent
change in F344/N Rats chronically exposed to 1,2,3-trichloropropane by gavage at
the 15-month interim evaluation
F344/N Rats
Dose
(mg/kg-day)
n
Males
n
Females
0
10
31.2 ± 0.6a
-
10
30.8 ± 0.8a
-
3
10
33.1 ±0.7
6%b
10
30.9 ±0.6
0%
10
10
36.0 ± 0.7C
15%
8
34.6 ± 1.0C
12%b
30
8
39.8 ± 0.9C
28%
8
43.2 ± 0.7C
40%
" Mean ± standard error
b Percent change relative to control
c/?<0.01
''/?<().05 by Williams' or Dunnett's test
Source: NTP, 1993.
October, 2007
24
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-9b. Absolute liver weights (grams) and percent change in F344/N Rats
chronically exposed to 1,2,3-trichloropropane by gavage at the 15-month interim
evaluation
F344/N Rats
Dose
(mg/kg-day)
n
Males
n
Females
0
10
14.27 ±0.37a
-
10
7.79 ± 0.13a
-
3
10
15.63 ±0.37d
10%b
10
8.87 ± 0.3 lc
14%b
10
10
16.8 ± 0.48c
18%
8
9.00 ± 0.28c
16%
30
8
18.23 ±0.52c
28%
8
10.40 ±0.37c
34%
a Mean ± standard error
b Percent change relative to control
c/?<0.01
><0.05 by Williams' or Dunnett's test
Source: NTP, 1993.
Table 4-10a. Relative right kidney weights (mg organ weight/ g body weight) and
percent change in F344/N Rats chronically exposed to 1,2,3-trichloropropane by
F344/N Rats
Dose
(mg/kg-day)
n
Males
n
Females
0
10
2.96 ± 0.04a
-
10
3.08 ± 0.07a
-
3
10
3.09 ±0.09
4%b
10
2.93 ±0.07
-5%b
10
10
3.25 ± 0.05c
10%
8
3.34 ± 0.06d
8%
30
8
3.82 ± 0.05c
29%
8
4.04 ± 0.12c
31%
aMean ± standard error
b Percent increase relative to control.
><0.01
><0.05 by Williams' or Dunnett's test
Source: NTP, 1993.
Table 4-10b. Absolute right kidney weights (grams) and percent change in F344/N
Rats chronically exposed to 1,2,3-trichloropropane by gavage at the 15-month
interim evaluation
F344/N Rats
Dose
(mg/kg-day)
n
Males
n
Females
0
10
1.35 ± 0.03a
-
10
0.786 ±0.015a
-
3
10
1.46 ± 0.04d
8%b
10
0.839 ±0.023
7%b
10
10
1.51 ±0.03c
12%
8
0.869 ±0.019d
11%
30
8
1.75 ± 0.05c
30%
8
0.971 ±0.034c
24%
aMean ± standard error
b Percent increase relative to control.
><0.01
d/><0.05 by Williams' or Dunnett's test
October, 2007	25	DRAFT - DO NOT CITE OR QUOTE

-------
Source: NTP, 1993.
The data for clinical chemistry parameters was sporadic, with ALT and 5'-nucleotidase
levels statistically (p<0.05) significantly decreased 31 and 13%, respectively, in males that
received 30 mg/kg-day.
Treatment-related effects were detected among the hematological parameters in rats;
however, the effects were not considered to be biologically relevant. Rats that received 30
mg/kg-day, displayed mean hematocrit values that were statistically (p<0.05) significantly
decreased by 5% and 7% for males and females, respectively, when compared with controls.
The mean hemoglobin concentration was decreased by 4% in male rats that received either 3
(p<0.01) or 30 (p<0.05) mg/kg-day. Both males and female rats in the high dose group had
statistically (p<0.01) significantly elevated counts of leukocytes and segmented neutrophils, but
not in the 10 mg/kg-day group. NTP stated that the decreased hematocrit may have been
associated with depressed erythropoeisis or with blood loss from neoplasms in the forestomach
or oral mucosa, and increased leukocytes was likely due to inflammation associated with the
chemical-induced neoplasms (NTP, 1993).
An increase in the incidence of forestomach tumors was observed in all rat treatment
groups (Table 4-11), regardless of sex. However, the incidences of forestomach neoplasms were
generally higher in males than in females at the same dose levels. All male treatment groups
also had increased incidence of pancreatic tumors (Table 4-11). Male and female rats that
received doses of 10 mg/kg-day 1,2,3-trichloropropane or higher had an increase in the incidence
of oral cavity tumors (Table 4-11). In each male group that received doses of 10 mg/kg-day or
higher, an increased incidence of renal tumors was observed. An increase was observed in
females at both 10 mg/kg-day and 30 mg/kg-day for the clitoral gland tumors and at the 10 and
30 mg/kg-day for mammary gland tumors (Table 4-11). In the 30 mg/kg-day treatment group an
increased incidence of Zymbal's gland tumors was observed in females and an increased
incidence of preputial gland tumors was observed in males at 30 mg/kg-day (Table 4-11).
Forestomach tumors were described in the NTP (1993) report as follows:
The masses were squamous cell papillomas or squamous cell carcinomas arising from the
stratified squamous cell epithelium of the forestomach. Multiple squamous cell
papillomas or carcinomas often occurred in the same rat, and in some rats, the neoplasms
were so extensive that it was difficult to discern if they represented a single neoplasm or
the confluent growth of multiple neoplasms.
October, 2007
26
DRAFT - DO NOT CITE OR QUOTE

-------
Forestomach tumors were accompanied by an increased incidence of focal hyperplasia of
the stratified squamous cell epithelium. The hyperplasia, squamous cell papilloma, and
squamous cell carcinoma of the forestomach were said to constitute a morphological continuum
and the squamous cell papillomas and carcinomas were noted to be similar to those of the oral
mucosa (NTP, 1993).
Table 4-11. Incidence of neoplasms in F-344/N rats chronically exposed to 1,2,3-
trichloropropane by gavage
Tissue site/tumor
type
Tumor incidence21

Males (m
g/kg-day;

Females (mg/kg-da
y)
0
3
10
30
0
3
10
30
Oral cavity
Papillomas or
carcinomas
1/60
4/60
19/5 9b
43/60b
1/60
6/59
28/60b
37/60b
Forestomach
Papillomas or
carcinomas
0/60
35/60b
46/59b
51/60b
0/60
17/59b
42/59b
27/60b
Pancreas (acinar)
Adenomas or
adenocarcinomas
5/60
21/60b
37/59b
31/60b
0/60
0/59
2/60
0/60
Kidney (renal
tubules)
Adenomas or
adenocarcinomas
0/60
2/60
20/59b
26/60b
0/60
0/57
0/60
1/59
Preputial gland
Adenomas or
carcinomas
5/59
6/57
9/59
17/58c
—
—
—
—
Clitoral gland
Adenomas or
carcinomas
—
—
—
—
5/56
11/56
18/58b
17/59c
Mammary gland
Adenocarcinomas
—
—
—
—
1/60
6/59
12/60b
22/60b
Zymbal's gland
Carcinomas
0/60
0/60
0/59
3/60
0/60
1/59
0/60
4/60c
a Values are pooled results from the outcome of histopathologic examinations of animals at the interim and
terminal sacrifices.
'><0.001 by life table or logistic regression test.
c p<().()5 by life table or logistic regression test.
Source: NTP, 1993.
The NOAEL and LOAEL for relative liver weight change in male and female rats is 3
and 10 mg/kg-day, respectively; while the LOAEL for absolute liver weight change in male rats
is 3 mg/kg-day and in female rats the NOAEL is 3 mg/kg-day and the LOAEL is 10 mg/kg-day.
October, 2007
27
DRAFT - DO NOT CITE OR QUOTE

-------
The NOAEL and LOAEL for relative right kidney weight in male and female rats is 3 and 10
mg/kg-day, respectively; while the LOAEL for absolute right kidney weight in male rats is 3
mg/kg-day and the NOAEL and LOAEL in female rats is 3 and 10 mg/kg-day, respectively.
Tumors were evident in the oral cavity, forestomach, pancreas, kidney, Zymbal's gland of male
and female rats, along with preputial gland tumors in males and clitoral gland and mammary
gland tumors in females. The critical effect for non-cancer data is liver and right kidney weight
change, while the critical effect for the cancer data is tumor development in the aforementioned
organs.
NTP (1993) conducted a 2-year study of the toxicity and carcinogenicity of 1,2,3-
trichloropropane in B6C3F1 mice. The chemical was administered by corn oil gavage to 60
mice/ sex/ group. Mice were treated with 0, 6, 20, or 60 mg/kg-day, and after 15 months (65-67
weeks), 8 to 10 mice per group were sacrificed to allow an interim evaluation of all toxicological
parameters and histopathology. Due to high mortality in the mice receiving 60 mg/kg, surviving
mice were evaluated at week 73 (females) and week 79 (males). Due to the early termination of
this treatment group, organ weights and hematology data were only obtained at the 15-month
interim sacrifices.
Clinical observations were made twice daily, while body weights were recorded weekly
for 13 weeks and then monthly (NTP, 1993). As mentioned above, up to 10 mice/ group were
sacrificed at month 15. From this interim sacrifice blood samples were obtained for hematology
and clinical chemistry analyses. Hematological parameters included hematocrit, hemoglobin,
and counts of erythrocytes, leukocytes, and differential leukocytes. Clinical chemistry
parameters included the serum levels of ALT, AST, creatine kinase, lactate dehydrogenase
(LDH), SDH, and 5'-nucleotidase. Whether at the planned sacrifice or as each mouse died or
became moribund, all mice were subjected to a gross necropsy, and a full range of organs and
tissues was processed for histopathologic examination. Hematology, clinical chemistry, and
organ weight data were obtained only from mice that were sacrificed at the 15-month interim
because the majority of treated mice died prior to the end of the study.
Survival rates were statistically significantly reduced (p<0.001) in mice that received
doses of 6 mg/kg-day or higher (Table 4-12). An effect on survival was apparent in all dose
groups at the 15-month interim evaluation. The mortality in mice was attributed to cancer
associated with chemical exposure (NTP, 1993).
October, 2007
28
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-12. Survival rates and percent probability of survival for B6C3F1 mice
exposed to 1,2,3-trichloropropane by gavage for two year
B6C3F1 Mice
Dose
(mg/kg-day)
Males
Females
0
42/52a
81b
41/50a
82b
6
18/51
36c
13/50
26c
20
0/54
0C
0/50
0C
60
0/56
0C
0/55
0C
a Animals surviving to study termination and number of animals in the treatment group. Accidental deaths
were excluded and censored from survival analysis.
b Kaplan-Meier determinations of percent probability of survival at end of study.
c /K0.001.
Source: NTP, 1993.
In mice, final mean body weights were significantly decreased by 17% and 18% in males
and females, respectively, after a dose of 60 mg/kg-day, when compared to controls. Mean
relative liver weights were increased by 32% in males and 40% in females that received 60
mg/kg-day (Table 4-13a). Other significant changes in organ weights among mice that received
this dose included increased relative kidney weights in females (21%) (Table 4-14a), and
increased relative brain weights in males (20%) and females (25%). Absolute liver and right
kidney weight changes were sporadic, and no consistent pattern of treatment-related effects was
apparent (Table 4-13b, 4-14b).
Table 4-13a. Relative liver weights (mg organ weight/ g body weight) and percent
B6C3F1 Mice
Dose
(mg/kg-day)
n
Males
n
Females
0
10
38.9 ± 1.9a
-
10
34.4 ± 0.8a
-
6
9
36.2 ± 1.5
-7%b
10
34.7 ± 1.1
l%b
20
8
44.6 ±6.2
15%
9
35.7 ±0.6
4%
60
5
51.2 ± 4.8d
32%
5
48.3 ±2.8C
40%
" Mean ± standard error
b Percent change relative to control
c/?<0.01
d/?<0.05 by Williams' or Dunnett's test
Source: NTP, 1993.
October, 2007
29
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-13b. Absolute liver weights (grams) and percent change in B6C3F1 mice
chronically exposed to 1,2,3-trichloropropane by gavage
B6C3F1 Mice
Dose
(mg/kg-day)
n
Males
n
Females
0
10
1.72 ± 0.09a
-
10
1.49 ± 0.03a
-
6
9
1.63 ±0.08
-5%b
10
1.33 ± 0.03d
-ll%b
20
8
1.76 ±0.19
2%
9
1.50 ±0.04
1%
60
5
1.92 ±0.14
12%
5
1.69 ±0.18
13%
a Mean ± standard error
b Percent change relative to control
c/?<0.01
r> p<().()5 by Williams' or Dunnett's test
Source: NTP, 1993.
Table 4-14a. Relative right kidney weights (mg organ weight/ g body weight) and
percent change in B6C3F1 mice chronically exposed to 1,2,3-trichloropropane by
B6C3F1 Mice
Dose
(mg/kg-day)
n
Males
n
Females
0
10
8.0 ± 0.25a
-
10
4.99 ± 0.09a
-
6
9
7.67 ±0.41
-4%b
10
5.27 ±0.14
6%b
20
8
7.81 ±0.18
-2%
9
5.19 ± 0.14
4%
60
5
8.4 ±0.59
5%
5
6.02 ± 0.1 lc
21%
aMean ± standard error
b Percent increase relative to control.
c/?<0.01
r> p<().()5 by Williams' or Dunnett's test
Source: NTP, 1993.
Table 4-14b. Absolute right kidney weights (grams) and percent change in B6C3F1
mice chronically exposed to 1,2,3-trichloropropane by gavage
B6C3F1 Mice
Dose
(mg/kg-day)
n
Males
n
Females
0
10
0.353 ± 0.01 la
-
10
0.217 ±0.006a
-
6
9
0.344 ±0.019
-3%b
10
0.203 ± 0.006
-6%b
20
8
0.314 ± 0.013
-11%
9
0.217 ±0.006
0
60
5
0.317 ±0.022
-10%
5
0.210 ±0.015
-3%
aMean ± standard error
b Percent increase relative to control.
c/?<0.01
October, 2007
30
DRAFT - DO NOT CITE OR QUOTE

-------
d/?<0.05 by Williams' or Dunnett's test
Source: NTP, 1993.
In mice, creatine kinase was statistically (/?<0.05) significantly elevated 235% in males
that received 60 mg/kg-day, and SDH was statistically (/?<0.05) significantly elevated 72% in
females that received the same dose. However, clinical chemistry differences between dose
groups and control animals were not considered to be directly related to 1,2,3-trichloropropane
administration (NTP, 1993).
Treatment-related effects were detected among the hematological parameters, but the
effects were indirectly related 1,2,3-trichloropropane toxicity. Mean hematocrit values were
decreased by 5% and 4% in male and female mice, respectively, that received 20 mg/kg-day.
Mean hematocrit values were statistically (p<0.0 l ) decreased by 10% and 11% in males and
females, respectively, that received 60 mg/kg-day. Similar statistically (p<0.01) significant dose
dependent changes in hemoglobin concentration and the number of erythrocytes were observed
in female mice that received doses of 20 or 60 mg/kg-day. Female mice in the high dose group
also had statistically (p<0.0 l) significantly elevated numbers of leukocytes, segmented
neutrophils, and lymphocytes. NTP stated that the decreased hematocrit may be associated with
depressed hematopoeisis or to blood loss from neoplasms in the forestomach, and the increased
number of leukocytes was likely due to inflammation associated with the chemically-induced
neoplasms (NTP, 1993).
In mice, the sites of statistically (p<0.00 l) significant neoplasm formation for both sexes
were the forestomach and liver (Table 4-15). Incidences of Harderian gland tumors were
increased in males at 20 and 60 mg/kg-day, and the increase in incidence of oral cavity tumors
was statistically significant in females at the highest dose. The incidence of uterine/cervical
tumors in female mice was increased at 20 and 60 mg/kg-day. The highest incidence of
neoplasms and most marked dose-response effect for both species was in the forestomach. A
97% incidence of tumors of the forestomach was evident in male mice at the lowest dose tested
(90% in females). These data suggest that an elevated incidence of tumors in the forestomach
might occur at doses lower than those employed in this study. NTP (1993) noted that:
In contrast to dosed rats, there were few neoplasms of the oral mucosa in dosed mice.
Nevertheless, squamous cell carcinomas arising from the pharyngeal or lingual mucosa
were observed in one 20 mg/kg and five 60 mg/kg females, and none were seen in
controls.
October, 2007
31
DRAFT - DO NOT CITE OR QUOTE

-------
The exophytic, or outward growing, papillary or nodular masses in the forestomach of
mice were similar to those observed in rats. Moreover, the extensive neoplastic growth observed
in rats was also noted in mice.
Table 4-15. Incidence of neoplasms in B6C3F1 mice chronically exposed to 1,2,3-
trichloropropane by gavage
Tissue site/tumor
type
Tumor incidence3
Males (m
g/kg-day

Females (mg/kg-day)
0
6
20
60
0
6
20
60
Oral cavity
Papillomas or
carcinomas
0/60
0/59
0/60
2/60
1/60
0/60
2/60
5/60°
Forestomach
Papillomas or
carcinomas
3/60
57/59b
57/60b
59/60b
0/60
54/60b
59/60b
59/60b
Liver
Adenomas or
carcinomas
14/60
24/59c
25/60°
33/60b
8/60
11/60
9/60
36/60b
Harderian gland
Adenomas
1/60
2/59
10/60°
11/60°
3/60
6/60
7/60
10/60
Uterine/ Cervical
Adenomas or
adenocarcinomas
—
—
—
—
0/50
5/50°
3/51°
9/54°
a Values are pooled results from the outcome of histopathologic examinations of animals at the interim and
terminal sacrifices.
'><0.001 by life table or logistic regression test.
c p<().()5 by life table or logistic regression test.
Source: NTP, 1993.
The NOAEL and LOAEL for relative liver weight change in male and female mice is 20
and 60 mg/kg-day, respectively; however, the NOAEL for absolute liver weight change was 60
mg/kg-day in male and female mice. The NOAEL and LOAEL for relative right kidney weight
change in female mice is 20 and 60 mg/kg-day, respectively; while the NOAEL in male mice for
relative right kidney weight change was 60 mg/kg-day. The NOAEL for absolute right kidney
weight was 60 mg/kg-day for both sexes. It should be noted that the high mortality associated
with chemical exposure lead to the early termination of the 20 and 60 mg/kg-day dose groups.
Tumors were evident in the oral cavity, forestomach, liver, and Harderian gland of both male and
female mice, and in the uterine/cervical tissue in females. The critical effect for non-cancer data
is weight change in the liver and right kidney, while the critical effect for the cancer data is
tumor development in the aforementioned organs of mice.
4.2.2. Inhalation Exposure
October, 2007
32
DRAFT - DO NOT CITE OR QUOTE

-------
4.2.2.1. Subchronic Studies
Johannsen et al. (1988) conducted a series of prechronic and subchronic inhalation
studies. In a range-finding study, five CD rats/sex/group were exposed in 1 m3 stainless, steel
and glass chambers to nominal concentrations of 0, 100, 300, 600, or 900 ppm 1,2,3-
trichloropropane vapor (0, 600, 1,800, 3,600, and 5,400 mg/m3) 6 hours/day, 5 days/week, for up
to 4 weeks. At the highest concentration, all but one of the rats died after a single exposure.
Three animals exposed to 600 ppm and one exposed to 300 ppm died prior to study termination.
Surviving rats exposed to 600 ppm trichloropropane became prostrated during exposure periods.
Males that were exposed to 600 ppm initially lost weight but returned to their pre-exposure
weights by the end of the experiment. Females exposed to this concentration showed a similar
pattern but did not regain the initial weights. Weight gain was statistically significantly reduced
(p<0.05) in rats exposed to 300 ppm and appeared depressed but was not significantly different
from controls for animals exposed to 100 ppm. Relative and absolute liver weights were
statistically significantly elevated (p<0.05) in males for all treatment groups and for females in
the 300 and 600 ppm groups (p<0.05) for relative liver weight and in the 300 ppm group for
absolute liver weight. Brain and kidney weights and organ/body ratios were increased in the 300
and 600 ppm treatment groups. Ovary weights and organ/weight ratios were decreased in the
300 and 600 ppm groups, and spleen weights and organ/weight ratios and testis weights were
decreased in the 600 ppm treatment group. The magnitude of change in body and tissue weights
was not reported.
The results of the 4-week range finding study were used to establish target concentrations
0, 5, 15, or 50 ppm (0, 30, 90, or 300 mg/m3) as the exposure concentrations for a 13-week
study, with analytical concentrations of 4.5 ± 0.2, 15 ± 0.3, and 49 ± 1.0 ppm. Each exposure
group contained 15 CD rats/sex. Blood samples were taken for clinical chemistry and
hematological parameters at week 7 from controls and the animals that were exposed to 50 ppm,
and at termination from all surviving animals. A gross pathological examination was conducted
on all animals and the weights of all major organs were recorded. Portions of the major organs
and tissues were processed for histopathologic examination. The results of these examinations
are described in the following paragraphs.
There were no treatment-related deaths in the 13-week study. Daily observation of
treated animals revealed a general, dose-dependent pattern of respiratory tract and conjunctival
irritation, including red nasal discharge and excessive lacrimation. An increased incidence of
yellow staining of the anogenital fur was also observed.
October, 2007
33
DRAFT - DO NOT CITE OR QUOTE

-------
A number of statistically significant changes were reported for whole body and organ
weights; however, the magnitudes of change in the body and tissue weights was not reported by
Johannsen et al., (1988) but were provided by the initial investigating group, Biodynamics, Inc.
(1979). Statistically significant reductions in terminal body weight were observed in females
exposed to 15 (7%) and 50 (9%) ppm. No effect on body weight was observed in males. Mean
absolute and relative liver weights (Table 4-16) were statistically significantly elevated 13-21%
in the male rat exposure groups. Mean absolute liver weights were statistically significantly
elevated 10% in females exposed to 50 ppm (p<0.01), and relative liver weights were
statistically significantly (p<0.01) increased 8 and 20% in females at 15 and 50 ppm,
respectively. Relative lung weights (Table 4-17) were also statistically significantly (p<0.01)
increased 14 and 13%, respectively, in female rats at doses of 15 and 50 ppm, although no effect
was evident in male rats. The mean relative kidney weight of males exposed to 50 ppm was
significantly increased approximately 10%.
Table 4-16. Absolute and relative liver weights and percent change in CD rats
exposed to 1,2,3-trichloropropane by inhalation, 6 hours/day, 5 days/week, for 13-
weeks
Male
Dose (ppm)
n
Absolute21
Relative13
0
15
13.8 ± 1.06
-
3.14 ± 0.128
-
5
15
16.7 ± 1.58c
21%
3.56 ± 0.258c
13%
15
15
16.3 ± 1.48c
18%
3.57 ± 0.207c
14%
50
14
16.4 ± 1.51c
19%
3.79 ± 0.260c
21%
Female
Dose (ppm)
n
Absolute21
Relative13
0
15
10.6 ±0.81
-
3.4 ±0.126
-
5
15
10.9 ±0.76
3%
3.6 ±0.213
6%
15
15
10.7 ± 1.05
1%
3.7 ± 0.216c
8%
50
15
11.7 ± 1.06c
10%
4.1 ±0.266c
20%
"Mean ± standard deviation
b Percent increase relative to control.
c/?<0.01
Source: Biodynamics, Inc., 1979.
October, 2007
34
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-17. Absolute and relative lung weights and percent change in CD rats
exposed to 1,2,3-trichloropropane by inhalation, 6 hours/day, 5 days/week, for 13-
weeks
Male
Dose (ppm)
n
Absolute
Relative
0
14
1.49 ± 0.162a
-
0.340 ±0.029a
-
5
15
1.62 ±0.192
9%b
0.345 ±0.036
l%b
15
15
1.58 ± 0.100
6%
0.347 ±0.030
2%
50
14
1.51 ±0.102
1%
0.351 ±0.028
3%
Female
Dose (ppm)
n
Absolute
Relative
0
15
1.27 ± 0.126a
-
0.406 ± 0.03 la
-
5
15
1.31 ±0.124
4%b
0.430 ±0.040
6%b
15
15
1.34 ± 0.107
6%
0.461 ±0.033 c
14%
50
15
1.31 ±0.129
3%
0.460 ± 0.051c
13%
aMean ± standard deviation
b Percent increase relative to control.
c/?<0.01
Source: Biodynamics, Inc., 1979.
A number of histopathologic lesions were observed (Table 4-18), including an increased
incidence of mild to marked peribronchial lymphoid hyperplasia at 5, 15, and 50 ppm. The
peribronchial lymphoid hyperplasia in the 15-ppm male rats was of equal severity to the 50-ppm
group, but the hyperplasia in the 15-ppm female rats and that evident in the 5-ppm males and
females were less severe. Hepatocellular hypertrophy in males at 5, 15, and 50 ppm appeared to
be at mild centrilobular to midzonal levels, but was not evident in the highest dose group
females. Treated females appeared to show a dose-dependent increase in extramedullary
hematopoiesis of the spleen. Statistical analysis was not conducted on these results.
October, 2007
35	DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-18. Incidence of histopathologic lesions in CD rats exposed via inhalation to
1,2,3-trichloropropane, 6 hours/day, 5 days/week for 13 weeks
Response
Male rats (ppm via inhalation)
0
0.5
1.5
5
15
50
Peribronchial
lymphoid
hyperplasia
0/15
0/15
0/15
6/15
11/15
10/15
Hepatocellular
hypertrophy
0/15
0/15
0/15
13/15
15/15
15/15
Hematopoiesis of
the spleen
0/15
0/15
0/15
ND
ND
5/15

Female rats (ppm via inhalation)
0
0.5
1.5
5
15
50
Peribronchial
lymphoid
hyperplasia
1/15
0/15
0/15
5/15
4/15
6/15
Hepatocellular
hypertrophy
0/15
0/15
0/15
ND
ND
0/15
Hematopoiesis of
the spleen
5/15
0/15
0/15
7/15
9/15
13/15
ND = no data.
Source: Biodynamics, Inc., 1979; Johannsen et al., 1988.
There were no significant dose-related changes in any of the hematological or clinical
chemistry parameters evaluated (Johannsen et al., 1988).
The NOAEL and LOAEL for decreased terminal body weight in female rats are 5 and 15
ppm, respectively, while the NOAEL for decreased terminal body weight in male rats is 50 ppm.
The LOAEL for increased absolute and relative liver weight in male rats is 5 ppm, while the
NOAEL and LOAEL for increased absolute liver weight in female rats is 15 and 50 ppm,
respectively, and the NOAEL and LOAEL for increased relative liver weight in females is 5 and
15 ppm, respectively. The NOAEL and LOAEL for increased relative lung weights in female
rats is 5 and 15 ppm, respectively, and the NOAEL and LOAEL for increased relative kidney
weights in males is 15 and 50 ppm, respectively. A LOAEL of 5 ppm was designated for
peribronchial lymphoid hyperplasia in male CD rats, as well as for hepatocellular hypertrophy in
male rats and hematopoiesis of the spleen in female rats.
The presence of lesions in animals from all exposure groups of the 13-week study
prompted the initiation of a follow-up study using lower exposure concentrations (Johannsen et
al., 1988). In the second 13-week study, the investigators employed a very similar experimental
October, 2007
36
DRAFT - DO NOT CITE OR QUOTE

-------
protocol with exposure concentrations of 0, 0.5, or 1.5 ppm (0, 3, or 9 mg/m3). The protocol for
the second study did not include urinalysis and the histopathological evaluation was limited to
bone, brain, gonads, kidneys, liver, lungs, lymph nodes, nasal turbinates, and spleen in control
and high-dose (1.5 ppm) rats. It also included two additional hematological and a few clinical
chemistry parameters.
Small increases in mean absolute and relative ovarian weights were observed in females
in the 1.5 ppm dose group, but microscopic results to support this as a treatment-related effect
were not found and this effect was not observed in the previous 13-weeks study with doses up to
50 ppm. Treatment-related histopathological findings were not observed in any tissue examined
(Table 4-18).
In the follow-up study sporadic changes were observed in some hematological and
clinical chemistry parameters, including apparently increased platelets in females exposed to 1.5
ppm for 7 weeks and increased fasting glucose levels in females exposed to 1.5 ppm for 13
weeks. In the absence of an apparent dose-response pattern these changes were considered by
the investigators to be unrelated to the 1,2,3-trichloropropane exposures. All other hematology
and clinical chemistry parameters measured were unremarkable and displayed no apparent effect
from TCP exposure.
This second investigation by Johannsen et al. (1988) identified aNOAEL of 1.5 ppm,
with regards to body or organ weight changes and histopathological effects, such as those
evident in the first study by Johannsen et al.
Miller et al. (1987a, b) conducted two inhalation studies of male and female F-344/N rats
and B6C3F1 mice. These unpublished studies were submitted to the EPA. In the first rat and
mouse study (Miller et al., 1987a), five animals/sex/group were exposed to a target concentration
of 0, 10, 30, and 100 ppm 6 hours/day, 5 days/week, for 9 days, with a measured concentration
of 0, 13 ± 0.5, 40 ± 0.4, or 132 ± 0.6 ppm (0, 78, 241, and 796 mg/m3). Evaluation endpoints
included body weight, urinalysis, clinical chemistry, hematology, and gross pathology and
histopathology.
Rats in the high exposure group were less active than controls and did not eat or drink
normally after treatment. An exposure and time-dependent reduction in weight gain was
observed in treated rats. Terminal body weights in rats were statistically significantly (p<0.05)
decreased 14% and 10% in males and females, respectively, in the high exposure group when
compared with controls. In male and female rats exposed to 40 ppm, relative liver weights
October, 2007
37
DRAFT - DO NOT CITE OR QUOTE

-------
were statistically significantly (p<0.05) increased 7 and 9%, respectively; and at 132 ppm,
absolute and relative liver weights were statistically significantly (p<0.05) increased 10 and
21%, respectively, in males and 27 and 42%, respectively, in females. At the highest exposure
concentration, relative liver weights were increased by 27% and 42% in male and female rats,
respectively, when compared with controls.
The concentrations of serum albumin and total protein were statistically significantly
(p<0.05) increased in both sexes of rats in the high exposure group, except for total protein in
female rats, but were not considered by the investigators to be toxicologically significant. No
exposure-related changes were observed among any of the hematology parameters; although, a
statistically significant (p<0.05) increase in packed cell volume (hematocrit), 6% increase, and
hemoglobin, 5% increase, was noted in female rats that were exposed to 40 ppm.
Several pathological changes in rats were associated with 1,2,3-trichloropropane
exposure. Gross observation suggested a treatment stress-related decrease in thymus size among
rats. Very slight hepatocellular necrosis was observed in all male rats exposed to 132 ppm, with
very slight depletion of lymphoid elements in the spleen in all male rats exposed to 132 ppm.
Miller et al. (1987a) also noted a dose-dependent increase in incidence and severity of
degeneration and decreased thickness of the olfactory epithelium in the nasal turbinates of rats
exposed to 13, 40, or 132 ppm 1,2,3-trichloropropane (Table 4-19). Inflammation in the
olfactory epithelium was also evident in rats exposed to 13, 40, or 132 ppm 1,2,3-
trichloropropane, and was accompanied by the exudation of inflammatory cells into the nasal
cavity lumen (Table 4-20).
Table 4-19. Incidence and severity of decreased thickness and degeneration of the
olfactory epithelium in the nasal turbinates of F344/N rats exposed via inhalation to
1,2,3-trichloropropane		
Severity
IV1
ales (ppm

Females (ppm)
0
1
3
10
13
40
132
0
1
3
10
13
40
132
Very slight
0
0
5
5
0
0
0
0
0
5
5
1
1
0
Slight
0
0
0
0
5
0
0
0
0
0
0
4
1
0
Moderate
0
0
0
0
0
5
0
0
0
0
0
0
2
2
Severe
0
0
0
0
0
0
5
0
0
0
0
0
1
3
Combined incidence
0
0
5
5
5
5
5
0
0
5
5
5
5
5
n
5
5
5
5
5
5
5
5
5
5
5
5
5
5
a Decreased thickness, bilateral and multifocal, or degeneration, bilateral and multifocal
Source: Miller et al. (1987a, b).
October, 2007	38	DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-20. Incidence and severity of inflammation of the olfactory epithelium in
Severity
IV1
ales i
ppm)



Females
(ppm)

0
1
3
10
13
40
132
0
1
3
10
13
40
132
Very slight
3
1
0
5
2
0
0
3
3
2
5
4
1
0
Slight
0
0
0
0
3
4
1
0
0
0
0
1
4
1
Moderate
0
0
0
0
0
1
4
0
0
0
0
0
0
4
Combined incidence
3
1
0
5
5
5
5
3
3
2
5
5
5
5
Exudate into lumen
0
0
0
2
2
3
1
0
0
0
2
4
4
5
n
5
5
5
5
5
5
5
5
5
5
5
5
5
5
' inflammation, bilateral and multifocal
Source: Miller et al. (1987a, b).
Mice in the high exposure group were less active than controls and did not eat or drink
normally after treatment; however, no effect on weight gain was observed in mice. Absolute
liver weights were statistically significantly (p<0.05) increased 67% and 73% in male and female
mice, respectively, exposed to 132 ppm; and relative liver weights in the high exposure group
were statistically significantly (p<0.05) increased by 55% and 60% in males and females,
respectively, compared with controls. Male mice also displayed statistically significantly
decreased absolute and relative testes weights, 9% and 16%, respectively, in the highest
exposure group, but histopathologically-related changes were absent.
The concentrations of serum albumin and total protein were statistically significantly
(p<0.05) increased in both sexes of mice, but were not considered by the investigators to be
toxicologically significant. There were no dose-related changes among any of the hematological
parameters, although the number of platelets in male and female mice at 132 ppm increased
statistically significantly (p<0.05) 25 and 42%, respectively.
Several pathological changes in mice were associated with 1,2,3-trichloropropane
exposure. A moderate increase in hepatocyte size was noted in all male and female mice
exposed to 132 ppm, and a slight or very slight depletion of lymphoid elements in the spleen was
noted in all mice that were exposed to this concentration. A dose-dependent increase in the
incidence and severity of decreased thickness and degeneration of the olfactory epithelium in the
nasal turbinates of mice was also noted (Table 4-21). There was a dose-related increase in
inflammation in the olfactory epithelium of the nasal turbinates (Table 4-22) accompanied by the
exudation of inflammatory cells into the nasal cavity lumen.
October, 2007
39
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-21. Incidence and severity of decreased thickness and degeneration of the
olfactory epithelium in the nasal turbinates in B6C3F1 mice exposed via inhalation
to 1,2,3-trichloropropane		
Severity


Males
(ppm)



Females
(ppm)

0
1
3
10
13
40
132
0
1
3
10
13
40
132
Very slight
0
0
0
5
5
4
0
0
0
0
5
5
2
0
Slight
0
0
0
0
0
1
2
0
0
0
0
0
3
0
Moderate
0
0
0
0
0
0
3
0
0
0
0
0
0
5
Combined incidence
0
0
0
5
5
5
5
0
0
0
5
5
5
5
n
5
5
5
5
5
5
5
5
5
5
5
5
5
5
a Decreased thickness, bilateral and multifocal, or degeneration, bilateral and multifocal
Source: Miller et al. (1987a, b).
Table 4-22. Incidence and severity of inflammation of the olfactory epithelium in
the nasal turbinates of B6C3F1 rats exposed via inhalation to 1,2,3-trichloropropane
Severity


Males
(ppm)



Females
(ppm)

0
1
3
10
13
40
132
0
1
3
10
13
40
132
Very slight
0
0
0
2
2
4
0
0
0
0
5
1
3
0
Slight
0
0
0
0
0
1
2
0
0
0
0
0
2
0
Moderate
0
0
0
0
0
0
3
0
0
0
0
0
0
5
Combined incidence
0
0
0
2
2
5
5
0
0
0
5
1
5
5
Exudate into lumen
0
0
0
1
1
1
5
0
0
0
0
0
2
5
a inflammation, bilateral and multifocal
Source: Miller et al. (1987a, b).
Since changes to the nasal epithelium were observed in the 13, 40 and 132 ppm exposure
groups, a follow-up study (Miller et al., 1987b) was initiated using the same study protocol and
target exposure concentrations of 0, 1, 3, and 10 ppm, with measured concentrations of 0, 1.0 ±
0.0, 2.9 ± 0.2, or 9.7 ± 0.3 ppm (0, 6, 18, or 60 mg/m3). Body weights and organ weights of both
sexes of rats and mice were not adversely affected at any concentration level. Very slight
decreased thickness and degeneration of the olfactory epithelium in the nasal turbinates was
observed in male and female rats that were exposed to 3 and 10 ppm (Table 4-19). Very slight
inflammation in the olfactory epithelium was also evident in rats exposed to 0, 1, 3, and 10 ppm
(Table 4-20). The exudation of inflammatory cells into the nasal cavity lumen was observed in
two male and two female rats at 10 ppm.
A very slight decrease in thickness of the olfactory epithelium in the nasal turbinates was
observed in both sexes of mice that were exposed to 10 ppm (Table 4-21). Very slight
inflammation in the olfactory epithelium was also observed in mice at 10 ppm (Table 4-22). The
October, 2007	40	DRAFT - DO NOT CITE OR QUOTE

-------
exudation of inflammatory cells into the nasal cavity lumen was observed in a single male mouse
at 10 ppm. No other exposure-related effects were detected.
4.2.2.2. Chronic Studies
No studies were identified that examined the chronic toxicity of 1,2,3-trichloropropane
via inhalation.
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES - ORAL AND INHALATION
4.3.1. Oral Studies
NTP (1990) conducted a reproduction and fertility assessment of 1,2,3-trichloropropane
in CD-I mice using the Reproductive Assessment by Continuous Breeding (RACB) protocol.
This assessment consisted of four tasks/studies: (1) a range-finding study, (2) a continuous
breeding study, (3) a determination of the affected sex, and (4) an offspring assessment. All
treatments were administered by corn oil gavage.
In Task 1, mice (eight/sex/group) received 0, 12.5, 25, 50, 100, and 200 mg/kg-day for 14
days. No effect on weight gain or clinical signs of toxicity was observed. One male in the high
dose group died. The results of this study were used to select the doses for Task 2.
Task 2 was a continuous breeding study in which 20 breeding pairs received 0, 30, 60, or
120 mg/kg-day for 126 days. Endpoints monitored for this task included clinical signs of
toxicity, parental body weight, water consumption, fertility, litters/pair, live pups/litter,
proportion of pups born alive, sex of live pups, and pup weights at birth. Pups were not
monitored for physical abnormalities. The last litter (Fi) born during the holding period
following the continuous breeding phase was reared by each dam until weaning, and was then
used in the assessment of second generation fertility in Task 4.
The parental body weights, both male and female, were within 10% of the corresponding
control values, except for the 120 mg/kg-day females, which exhibited an increase in body
weight greater than 10%. Water consumption was significantly increased in weeks 6, 10, and
14; however, consumption was calculated per cage, and, up until week 14, male and female mice
were housed in the same cage. At terminal necropsy, absolute and relative liver weights were
October, 2007
41
DRAFT - DO NOT CITE OR QUOTE

-------
statistically significantly increased in the 120 mg/kg-day male and female mice, but data analysis
for the intermediate dose groups is unavailable.
Statistically significant reduction in fertility was evident at the 4th and 5th breedings
(Table 4-23). A statistically significant (p<0.05) reduction in fertility was evident from the
decrease in the number of pregnancies per fertile mouse pair at the fourth breeding (89%), but
not the fifth (68%), at 60 mg/kg-day group, and the third, fourth, and fifth breedings (89, 68, and
42%), respectively) at 120 mg/kg-day. A dose-related decrease in fertility from the 4th to 5th
breeding at 60 mg/kg-day was observed, but this decrease did not reach statistical significance.
A statistically significant (p<0.05) reduction in the number of live mouse pups/litter was
observed, when compared with controls, in the second through the fifth breedings at the highest
dose (120 mg/kg-day) and at the fifth breeding at 60 mg/kg-day (Table 4-23). The 120 mg/kg-
day group displays a dose- and time-related decrease in the number of live pups/litter. However,
the decrease in the number of pups/litter in the fifth breeding at 60 mg/kg-day is statistically
significant due to an increase in the number of pups/litter in the controls during the fifth
breeding. The number of live pups/litter increases from the fourth to the fifth breeding at 60
mg/kg-day and does not follow a dose- or time-related response.
The cumulative days to litter were statistically significantly longer than control values for
the third breeding (12%>) at 60 mg/kg-day and the fourth (6.5%>) and fifth (3.3%>) breedings at
120 mg/kg-day. The proportion of male pups born alive in the fifth breedings appeared to
decrease in a dose-dependent manner. The proportion of males in the fifth breeding of the 120
mg/kg-day treatment group was 0.27 versus 0.53 for the controls, with proportions in the fifth
breeding at 30 and 60 mg/kg of 0.43 and 0.42, respectively. Live pup weights were slightly,
statistically significantly, increased at the highest dose, 120 mg/kg-day. However, when
adjusted for average litter size ± standard error, the increase in live pup weights was eliminated
in male pups at 120 mg/kg-day, and a decrease, although not statistically significantly, in female
pups and combined pups was visible.
October, 2007
42
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-23. Fertility indices and number of live pups/litter in breeding pairs of CD-
1 mice exposed to 1,2,3-trichloropropane by gavage
Litter
Dose group
(mg/kg-day)
0
30
60
120
Fertility3
Live
Pups/litter
Fertility3
Live
Pups/litter
Fertility
a
Live
Pups/litter
Fertility3
Live
Pups/litter
1st
38/38
(100)
11.1 ±0.6
18/18
(100)
10.3 ± 1.0
19/19
(100)
10.5 ±0.9
18/19
(95)
11.5 ± 0.8
2nd
38/38
(100)
12.6 ±0.4
18/18
(100)
10.8 ± 1.2
19/19
(100)
10.7 ± 1.1
18/19
(95)
5.2 ± 0.6b
3rd
38/38
(100)
12.4 ±0.5
18/18
(100)
11.3 ± 1.0
19/19
(100)
11.0 ± 1.2
17/19b
(89)
6.7 ± 1.0b
4th
38/38
(100)
11.8 ± 0.6
17/18
(94)
11.2 ± 0.7
17/19b
(89)
9.9 ± 1.0
13/19b
(68)
2.9 ± 0.6b
5th
33/38
(87)
12.8 ±0.4
14/18
(78)
12.1 ±0.7
13/19
(68)
11.3 ±0.8b
8/19b
(42)
2.5 ± 0.6b
a Fertility data are the number of fertile pairs/number of cohabiting pairs (% fertile).
b Significantly different (/?<(),05) from the control group by Dunn's or Chi-square test.
Source: NTP, 1990.
Task 3, a one-week crossover mating trial, was conducted with the same adult mice from
the control and 120 mg/kg-day treatment groups from Task 2. Three groups of 20 breeding pairs
(control males x control females, control males x high-dose females, and control females x high-
dose males) were evaluated for fertility and the presence of morphological and histopathologic
changes to the reproductive organs. At termination, F0 mice were necropsied and major organs
were excised and weighed. Treated F0 mice of both sexes displayed statistically significantly
(p<0.05) increased absolute and relative liver weights, 19 and 20% in males and 25 and 22% in
females, respectively, compared with controls. The weights of the right epididymis and cauda
epididymis in F0 males were statistically significantly (p<0.05) lower, 5 and 8%, respectively,
than those of controls. The absolute kidney weights of treated F0 females were statistically
significantly (p<0.05) reduced (5%) compared with controls. All F0 males were evaluated for
epididymal sperm parameters, and no differences in motility, count, or abnormal sperm numbers
were detected. 120 mg/kg-day treated females delivered fewer live pups (—50%) than untreated
females, with decreased body weight (9%) in male offspring and fewer live male pups per litter
than controls.
In Task 4, members of the last set of litters (Fi) to be born in Task 2 were reared, weaned,
and allowed to reach sexual maturity before being paired individually with a member of the
opposite sex from a separate litter but within the same treatment group. Breeding pairs were
assessed for the same mating endpoints as in Task 2 and the same terminal endpoints as in Task
3. There were statistically significant (p<0.05) decreases, 78 and 43% of controls, in the indices
October, 2007
43
DRAFT - DO NOT CITE OR QUOTE

-------
for mating (# of females with plug/ # of cohabiting pairs) and fertility (# of fertile pairs/ # of
females with plug), respectively, for the 120 mg/kg-day group. The estrous cycles for Fi females
of all treatment groups were statistically significantly longer than in controls (p<0.05), and may
be associated with an increase in the infertile period of metestrus.
At necropsy, Fi male and female terminal body weights were statistically significantly
(p<0.05) increased, 5 to 11%, in the 60 and 120 mg/kg-day groups. There was a statistically
significant (p<0.05) increase, 17 to 50%, in absolute and relative liver weights in males and
females at 60 and 120 mg/kg-day; and a statistically significant (p<0.05) increase, 6 to 27%, in
absolute kidney weights in male and female mice at 60 and 120 mg/kg-day. A statistically
significant (p<0.05) 34% decrease in absolute right ovary weight was evident at the highest dose
level, with a statistically significant (p<0.05) decrease in relative ovary weight at 60 and 120
mg/kg-day of 15 and 39%, respectively. Histopathological examination of tissues from 10
females from each group revealed no difference between the groups in the incidence and severity
of lesions.
Based on the decreased number of fertile pairs and live pups/litter among the cohabiting
pairs in the 120 mg/kg-day treatment group, the investigators concluded that 1,2,3-
trichloropropane treatment could impair fertility and reproduction (NTP, 1990). In Task 2, a
NOEAL and LOAEL of 30 and 60 mg/kg-day, respectively, was identified for a decrease in the
number of pregnancies per fertile mouse pair at the fourth and fifth breeding. A reduction in the
number of live mouse pups/litter was observed across doses in the second through the fifth
breedings from breeding pairs at the highest dose (120 mg/kg-day) and at the fifth breeding at 60
mg/kg-day; which provides a NOAEL of 30 mg/kg-day and a LOAEL of 60 mg/kg-day at the
fifth breeding and a NOAEL of 60 mg/kg-day and LOAEL of 120 mg/kg-day for the first
through the fourth breedings. The LOAEL for the decreased proportion of males in the fifth
breeding is 30 mg/kg-day. Task 3, a cross-over mating trial, identified a LOAEL of 120 mg/kg-
day in treated females for decreased pups/ litter, decreased male pup weight, and decreased
proportion of males/ litter. A NOAEL and LOAEL of 60 and 120 mg/kg-day, respectively, for
decreased fertility and mating indices was identified from Task 4. A LOAEL of 30 mg/kg-day
for lengthened estrous cycle was also apparent.
4.3.2. Inhalation Studies
Johannsen et al. (1988) reported the results of a single-generation reproductive study
using 10 male and 20 female CD rats/group conducted in two dosing studies. In the first study,
animals in 1 m3 stainless, steel and glass chambers were exposed to target vapor concentrations
October, 2007
44
DRAFT - DO NOT CITE OR QUOTE

-------
of 0, 5, or 15 ppm (0, 30, or 90 mg/m3), with measured concentrations of 4.6 ± 0.2 and 15 ± 0.2,
1,2,3-trichloropropane 6 hours/day, 5 days/week, for a 10-week pre-mating period, a mating
period (not to exceed 40 days), and for gestation days 0-14 for females. Male and female rats
were housed in a ratio of 1:2, respectively, nightly during the mating period. Females that were
not impregnated after the 10 days were paired with a different male for 10 days until pregnant.
In the second study, the same numbers of rats were exposed to target concentrations of 0, 0.5, or
1.5 ppm (0, 3, or 9 mg/m3) using a similar protocol (mating period not to exceed 30 days).
Females delivered and all litters were weaned on postnatal day (PD) 21. Animals were
examined daily for clinical signs and received a weekly physical exam when body weights were
recorded, with mated females weighed through gestation and lactation. Pups were weighed at
birth, on PDs 4 and 14, and when they were sacrificed on PD 21. At termination, all F0 parents
were necropsied, and sections of their reproductive organs were processed for histopathologic
examination.
In the first study, females exposed to 15 ppm had lower body weights during gestation
and lactation, although weight gains were consistent with the controls. Both sexes exposed to 15
ppm exhibited decreases in weight and weight gain during the premating period of exposure. All
groups of female rats exhibited low mating performance, 16 females mated out of 20 at 5 ppm
and 10 mated out of 20 at 15 ppm, compared with 17 mated females out of 20; although fewer
females in the high concentration group mated, no statistical significance was evident. Male rats
in both treated and control groups displayed apparently lower mating performance, 4/10, 6/10
and 3/10 for control, 5 ppm and 15 ppm, respectively, but not statistically significant mating
indices. Fertility indices were unaffected by trichloropropane exposure. There was no
treatment-related effect on litter and pup data. Histopathological evaluation of the testes,
epididymis, and ovaries did not identify any treatment-related changes.
In the second study, adverse effects on mating performance and fertility indices due to
1,2,3-trichloropropane were not observed. Lesions of the testes, epididymides, and ovaries were
not evident. Consistent or obviously treatment-related reproductive effects were not observed in
any of the experimental groups in either generation.
This study identified a NOAEL of 15 ppm for low mating performance and fertility
indices. For decreased body weight in females during gestation and lactation and for decreased
body weights and weight gain in both sexes during the premating period, a NOAEL of 5 and
LOAEL of 15 was identified.
October, 2007
45
DRAFT - DO NOT CITE OR QUOTE

-------
4.4. OTHER STUDIES
4.4.1. Acute Toxicity Data
In the rat, oral LD50 values ranging from 150 to 500 mg/kg 1,2,3-trichloropropane have
been reported (MAK, 1998). A 4-hour LC50 of approximately 500 ppm (3000 mg/m3) has been
determined for rats and mice (MAK, 1998). McOmie and Barnes (1949) identified an LC50 of
approximately 30 ppm in mice exposed to vapor for 20 minutes, while Reyna (1987) could not
determine an LC50 in Sprague-Dawley rats, but suggested that the LC50 was greater than 4.8
mg/L air.
Lag et al. (1991) conducted an acute study in rats that investigated the nephrotoxicity of
short-chain halogenated alkanes. 1,2,3-trichloropropane was administered via a single,
intraperitoneal injection to 5 male MOL:WIST rats per dose group at doses of 147, 294 and 441
mg/kg. After 48 hours the rats were weighed and euthanized, and their kidneys were removed,
weighed, and preserved. Dose-dependent increases in mortality, kidney/body weight ratio and
urea excretion were evident. Histopathological examination detected moderate kidney necrosis
in one of the two surviving rats at the highest dose level tested.
4.4.2. Waterborne Studies
NTP (2005) conducted a toxicity study in 220 male and female guppies (Poecilia
reticulate) and 340 male and female medaka (Oryzias latipes) maintained in aquaria water
containing 0, 4.5, 9.0, or 18.0 mg/L. The guppies were exposed for 16 months and the medaka
for 13. Ten of each species at each dose group were sacrificed at 9 months for histopathologic
analysis. Approximately one-third of the fish that survived until the 9 month evaluation were
transferred to chemical-free water at that time and evaluated at study termination. These fish are
described as the stop-exposure group.
In the medaka study, survival at 9 months was decreased in the 9.0 and 18.0 mg/L groups
(NTP, 2005). At the 9 month evaluation, the incidence of choliangiocarcinomas was
significantly increased in 9.0 and 18.0 mg/L males. The incidence of choliangiocarcinomas was
significantly increased in all exposed males and females after 13 months, while the incidence of
hepatocholangiocarcinomas was significantly increased only in the fish exposed to 18.0 mg/L
1,2,3-trichloropropane. The incidence of papillary adenomas of the gallbladder was
significantly increased in the 9.0 and 18.0 mg/L males after the 13-month exposure. In the stop
October, 2007
46
DRAFT - DO NOT CITE OR QUOTE

-------
exposure component of the study, the incidence of papillary adenomas in males was significantly
increased only at the highest exposure concentration.
Reduced survival was evident in the guppies at 6 months at the highest concentration
tested (18 mg/L) and at 7 months in the 4.5 and 9 mg/L concentrations as well. It was
significantly reduced in the 18 mg/L guppies at about 8 months (NTP, 2005). At the 9 month
interim evaluation, there was an increased incidence of bile duct and hepatocellular neoplasms in
the exposed male and female guppies at all concentrations tested. In the stop-exposure
component of the study, hepatocellular neoplasms were evident in 18 mg/L males and bile duct
neoplasms were evident in 18 mg/L females.
1,2,3-Trichloropropane was characterized as carcinogenic at concentrations up to 18.0
mg/L in both sexes of guppies and medaka based on the increased incidence of liver neoplasms
and papillary adenoma of the gallbladder (NTP, 2005). Studies of toxicity in aquatic species
such as the medaka and guppy are increasingly being used as screening studies for tumor
formation and other endpoints of toxicity.
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION
4.5.1. Mode of Action Studies
Weber and Sipes (1990) conducted a series of experiments that examined the covalent
binding of 1,2,3-trichloropropane to hepatic macromolecules in male F-344/N rats. In a
preliminary experiment, binding of [14C]-l,2,3-trichloropropane to hepatic protein, DNA, and
RNA was measured 4 hours after i.p. administration of 30 mg/kg (100 (j,Ci/kg). Similar amounts
of radioactivity were bound to hepatic protein, 418 ± 19 pmol [14C]TCP equivalents/mg, and
RNA, 432 ± 74 pmol [14C]TCP equivalents/mg, and approximately half as much was bound to
DNA, 244 ± 29 pmol [14C]TCP equivalents/mg. Because of methodological problems, the
binding to RNA was not characterized further in this investigation.
In a subsequent time-course study, male rats (4/group) were sacrificed at 1, 4, 24, 48, and
72 hours post i.p. administration of 30 mg/kg [14C]-l,2,3-trichloropropane (100 (j,Ci/kg). To
examine the influence of various metabolic pathways, the investigators also administered 1,2,3-
trichloropropane to four additional groups (each containing four rats) that had been pretreated as
follows:
October, 2007
47
DRAFT - DO NOT CITE OR QUOTE

-------
•	80 mg/kg-day phenobarbital, a CYP450 (CYP2B, CYP3A) inducer, in 0.9% NaCl (i.p.)
for 4 days with 1,2,3-trichloropropane treatment on day 5;
•	40 mg/kg-day P-naphthoflavone, a CYP450 inducer (CYP1 A), (i.p.) in vegetable oil for 3
days, followed by treatment with 1,2,3-trichloropropane on day 4;
•	75 mg/kg SKF 525-A, an inhibitor of CYP450, in phosphate-buffered saline (pH 5.0)
administered (i.p.) 2 hours prior to treatment with 1,2,3-trichloropropane;
•	2 g/kg l-buthionine-(R,S)-sulfoximine (BSO), which causes a depletion of hepatic
glutathione (GSH), administered in two doses (i.p.) spaced 1.5 hours apart, followed by
1,2,3-trichloropropane treatment 3 hours later.
All rats in the metabolic study were sacrificed 4 hours after treatment with 1,2,3-
trichloropropane.
In the time-course study, maximum trichloropropane-equivalent covalent binding to
hepatic proteins (approximately 600 pmol/ mg) was observed 4 hours after trichloropropane
administration and was approximately 2.5-fold greater than at 1 hour post-administration.
Maximal covalent binding to hepatic DNA (approximately 250 pmol/ mg) was observed 24
hours after administration. By 72 hours the amount of radioactivity bound to both protein and
DNA had returned to levels below those measured 1 hour post administration. At the point of
maximal binding, the amount of [14C]-l,2,3-trichloropropane-derived radioactivity bound to
hepatic proteins was more than double the amount bound to hepatic DNA.
Administration of three consecutive doses each of 30 mg/kg 1,2,3-trichloropropane,
separated by 24 hours, produced a linear increase in the amount of [14C]-l,2,3-trichloropropane-
derived radioactivity bound to hepatic proteins. Repeated dosing did not affect the amount of the
chemical equivalent bound to DNA until the third dose at which point the amount of bound
radioactivity doubled.
In the metabolic study, induction of CYP450 (CYP) isozymes with phenobarbital
pretreatment significantly reduced chemical binding to hepatic protein and DNA by 70 and 64%,
respectively, when compared with controls. However, induction of CYP450 isozymes with P-
naphthoflavone pretreatment did not significantly alter binding to either macromolecule.
Depletion of reduced glutathione (GSH) by BSO pretreatment increased binding to hepatic
proteins by 342% and decreased binding to DNA by 44% when compared with controls, with the
increased covalent binding due to decreased GSH conjugation of a TCP metabolite. Inhibition
of CYP450 isozymes with SKF 525-A significantly increased binding to hepatic protein and
DNA by 58%) and 42%, respectively, compared with controls. The decrease in GSH appears to
lead to increased levels of a reactive metabolite that does not require GSH to bind with proteins,
October, 2007
48
DRAFT - DO NOT CITE OR QUOTE

-------
as evidenced by the increased binding to hepatic proteins when GSH levels were reduced by
BSO pretreatment. TCP or its metabolite(s) appear to conjugate with glutathione (GSH) and
produce compounds, such as episulfonium ions, that may covalently interact with hepatic DNA.
To further explore the effect of 1,2,3-trichloropropane on GSH, two additional
experiments were conducted by Weber and Sipes (1990): hepatic GSH levels were measured in
rats receiving 30, 100, and 300 mg/kg 1,2,3-trichloropropane (4 rats per dose); GSH levels of
control and treated animals were evaluated with and without phenobarbital pretreatment. 1,2,3-
Trichloropropane treatment caused a dose-dependent, statistically significant decrease in GSH
levels two hours after exposure. Phenobarbital pretreatment, on the other hand, did not increase
the trichloropropane-induced reduction in hepatic GSH concentrations.
La et al. (1995) investigated the formation of DNA adducts in animals treated with 1,2,3-
trichloropropane by using the same route of administration and some of the doses used in the
NTP (1993) chronic bioassay. A single dose of either 3 or 30 mg/kg 1,2,3-trichloropropane
containing [14C]-l,2,3-trichloropropane (1 mCi) was administered by gavage to male F-344/N
rats and 6 or 60 mg/kg [14C]-l,2,3-trichloropropane to male B6C3F1 mice. Animals were
sacrificed after 6 hours, and DNA adducts were hydrolyzed by neutral thermal or mild acid
treatment and separated by cation exchange high performance liquid chromatography. Peaks
were characterized by using electrospray ionization mass spectrometry, and their identity was
verified with synthesized standards.
The elution profile of the labeled DNA indicated that a single, major DNA adduct was
formed irrespective of the tissue type, and was determined to be
S-[l-(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione (La et al., 1995). A proposed formation
pathway involves the biological activation, possibly by conjugation with glutathione, of 1,2,3-
trichloropropane and intramolecular rearrangement to form episulfonium ions that covalently
bind to DNA. The formation of the S-[l-(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione
adduct was detected in the forestomach, glandular stomach, kidney, liver, pancreas, and tongue
(oral) of F-344/N rats, and in the forestomach, glandular stomach, kidney, and liver of B6C3F1
mice. The concentrations of this adduct formed in the target organs (expressed as [j,mol/mol
guanine) showed some correlation with the tumor incidence from the NTP (1993) study (Table
4-24). For example, dose-dependent adduct formation was demonstrated in the forestomach of
F-344/N rats and B6C3F1 mice, and the forestomach was a primary site of tumor formation in
both animal models in the NTP (1993) study. Conversely, dose-dependent adduct formation was
apparent in the liver and glandular stomach in both species, although NTP (1993) detected no
tumor formation at this tissue site. Adduct formation in the spleen of rats and mice, when
October, 2007
49
DRAFT - DO NOT CITE OR QUOTE

-------
compared to other organs, appeared lower, and NTP (1993) did not detect tumors in the spleens
of rats and mice.
Table 4-24. Comparison of tumor incidence and DNA-adduct formation in male F-
344N rats and B6C3F1 mice
Organ
Dose
tumor
incidence3
adduct level
(jimol/mol
guanine)b
male rats
forestomach
3
33/50
3.7
30
43/52
14.6
kidney
3
2/50
6.6 ± 1.4
30
21/52
38.9± 5.0
pancreas
3
21/50
5.3 ± 1.0
30
29/52
37.8 ± 12.8
preputial gland
3
6/47
NDd
30
16/50
NDd
oral
3
2/50
4.0
30
37/52
20.4
glandular
stomach
3
0/50
3.8
30
0/52
20.4
liver
3
1/50
5.4 ±0.7
30
3/52
47.6 ±21.0
male mice
forestomach
6
50/51
19.8
60
55/56
41.0
liver
6
24/51
12.1 ±4.6
60
31/56
59.3 ±21.7
lung
6
11/51
0.77 ±0.16
60
6/56
5.3 ±0.2
glandular
stomach
6
0/51
28.1
60
0/56
208.1
kidney
6
0/51
4.4 ±2.9
60
0/56
32.5 ± 11.3
a from NTP, 1993 and tallied in La et al., 1993.
b from La et al., 1995; expressed as mean ± standard deviation from four animals with
statistical significance not analyzed.
0 statistically significant increase in tumor formation from NTP, 1993.
d not detected.
Source: La et al., 1995
The S-[l-(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione adduct indentified by La et al.
(1995) is an N7-guanyl adduct shown in Figure 4-1. This adduct is unusual in that it crosslinks a
physiological oligopeptide, reduced glutathione, to DNA by a chemical carcinogen (Ozawa and
Guengerich, 1983). The N7 position on the guanine is a highly electrophilic nitrogen atom that is
October, 2007
50
DRAFT - DO NOT CITE OR QUOTE

-------
located in an accessible position on the DNA polymer (Gasparutto et al., 2005). N7-guanyl
adducts generally have an inhibitory effect on sequence-specific DNA binding by regulatory
proteins, due to a destabilization of the guanine nucleobase and spontaneous degradation
(Gasparutto et al., 2005; Ezaz-Nikpay and Verdine, 1994). However, the exact role of the N7-
guanyl adducts is unknown (Gasparutto et al., 2005). This DNA adduct lends evidence of the
involvement of the episulfonium ion in DNA binding, as the episulfonium ion interacts with
reduced glutathione and binds to DNA at the N7 of guanine. The formation of additional DNA
adducts could potentially be through the 1,3-dichloroacetone and 2-chloroacrolein pathways of
metabolism.
Figure 4-1. Structure of the DNA adduct S-[l-(hydroxymethyl)-2-(N7-
guanyl)ethyl] glutathione
In a subsequent publication, the DNA adduct-forming capacity of 1,2,3-trichloropropane
in male B6C3F1 mice (N = 15) which received equivalent doses of [14C]-l,2,3-trichloropropane
via either corn oil gavage or drinking water was compared (La et al., 1996). The mice were
administered 6 mg/kg-day for 5 days via gavage or drinking water. As shown in Table 4-25, a
greater amount of DNA adduct was extracted from tissues of those animals receiving 1,2,3-
trichloropropane via gavage when compared to those exposed via drinking water. Similarly, the
authors observed little, if any, cellular proliferation in the tissues of animals exposed to 1,2,3-
trichloropropane in drinking water. By contrast, cellular proliferation appeared to increase in a
dose-dependent manner in tissues of animals exposed to 1,2,3-trichloropropane by gavage.
coo
s
Sugar phosphate DNA backbone
October, 2007
51
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-25. Formation of DNA adducts by [14C]-1,2,3- trichloropropane (6 mg/kg-
day) administered to B6C3F1 mice by gavage or drinking water
Organ
DNA adduct formation (^imol/mol guanine)
Drinking water
Gavage
Target organs for tumor formation
Forestomach
86.8 ±73.2
123.1 ± 10.3
Liver
185.5 ± 83.9
374.9 ± 109.2a
Non-target organs for tumor formation
Glandular stomach
43.2 ± 5.9
42.5 ±4.6
Kidney
81.9 ± 41.5
193.1 ±64.4a
" Indicates a statistically significant difference (p<0.05) compared to values obtained in the same tissue of
animals receiving 1,2,3 -trichloropropane via the alternative route of administration, as calculated by the
authors.
Source: Laetal., 1996.
4.5.2. Genotoxicity Studies
Bacterial mutagenicity assays
Several in vitro genotoxicity studies have demonstrated 1,2,3-trichloropropane to be
mutagenic in the presence of metabolic enzymes, also known as S9 fraction (Table 4-26). In a
study of 250 individual chemicals, which included 1,2,3-trichloropropane, Haworth et al. (1983)
observed a dose-dependent increase in revertant colonies in Salmonella typhimurium strains
TA100 and TA1535 that were exposed to 10, 33, 100 and 333 jag 1,2,3-trichloropropane/ plate
with activation by both rat and hamster S9 fractions. No increases were observed in strains
TA98 and TA1537. Shell Oil Co. (1979) observed a dose-dependent increase in revertant
colonies in the presence of S9 fraction in tester stains: TA98, at 200 and 2000 |ig/plate; TA100,
at 20, 200, and 2000 [j,g/plate; and TA1537, at 200 and 2000 [^g/plate. These investigators also
detected revertants, at 200 and 2000 [j,g/plate in TA 1535 both in the presence and in the absence
of an S9 fraction, with a greater number of revertants in the plates with microsomal activation.
NTP (1993) tested strains at doses of 3, 10, 33, 100, [j.g/plate or 333 and 10, 33, 100, 333,
666, 667, and 1000 [j,g/plate and observed a dose-dependent increase in the number of revertants
in colonies of TA97, TA100 and TA1535 treated with 1,2,3-trichloropropane in the presence of
either hamster or rat S9 fraction in repeated experiments. Mutagenic activity was observed in
TA98 in the presence of hamster and rat S9 fraction. No mutagenic activity was observed in the
TA1537 test strain in the presence of S9. In the absence of an S9 microsomal fraction, no
mutagenic activity was detected in any of the Salmonella strains. It should be noted that the
NTP (1993) report includes data from the Haworth et al. (1983) article. The descriptions
provided here are for trials not included in the earlier report.
October, 2007 52 DRAFT - DO NOT CITE OR QUOTE

-------
Other groups also have demonstrated the mutagenic capability of 1,2,3-trichloropropane
by using the Ames test. Stolzenberg and Hine (1980) and Lag et al. (1994) found a dose-
dependent increase in mutagenic activity in TA100 in the presence of S9 at doses of 14.7 (0.1
(j,mol/plate) and 147 (1 (iinol/plate) |ig/plate and -14.7 (0.1 (j,mol/plate) [j,g/plate, respectively.
No increases were observed in the non-activated cultures. A dose dependent, statistically
significant increase in mutagenic activity was also demonstrated by Ratpan and Plaumann (1988)
in TA1535 and TA100 in the presence of S9 at doses of 5, 10, 50 and 100 jag 1,2,3-
trichloropropane/ plate, with no mutagenic activity in the same strains in the absence of S9 at the
same doses and no mutagenic activity in TA98, TA1537, or TA1538 in the presence and absence
of S9 at the same doses. Kier (1982) found mutagenic activity in TA100, TA1535, and TA98 in
the presence of S9 fraction at 20-1000 [j,g/plate, 20-300 [j,g/plate, and 100-300 [j,g/plate,
respectively. No mutagenic activity was found in the same strains in the absence of S9 at the
same doses nor was mutagenic activity found in TA1537 and TA1538 in the presence and
absence of S9 at the same doses.
The mutagenic effects of 1,2,3-trichloropropane have also been examined in other
microbial systems with mixed results, von der Hude et al. (1988) showed the compound to be
negative for DNA damage in the SOS chromotest using Escherichia coli PQ37. 1,2,3-
Trichloropropane induced mutations in DNA repair-deficient is. coli WP2 uvr A at 2000
[j,g/plate, but not in the DNA repair-proficient strain WP2, and induced mitotic gene conversion
in Saccharomyces cerevisiae after exposure to 0.01, 0.1, 0.5, 1.0, or 5.0 mg/cm3 TCP in the
presence of rat liver S9 (Shell Oil Co., 1979). Increases were not observed in the non-activated
cultures. 1,2,3-Trichloropropane tested negative in the Aspergillus nidulans diploid strain PI
assay for aberrant mitotic segregation at 0.1 % v:v with 5% survival (Crebelli et al., 1992).
October, 2007
53
DRAFT - DO NOT CITE OR QUOTE

-------
Table 4-26. Genotoxicity bioassays of 1,2,3-trichloropropane
In Vitro Gene Mutation Assays
Test System
Cells/strain
Positive
concentrations
Results
Reference
-S9
+S9
(a) Bacterial Assays
S. typhimurium
(Ames test)
TA100,
A1535
10, 33, 100, 333
[j,g/plate
-
+
Haworth et al.,
1983
TA1537,
TA98

-
-
TA98
200, 2000 |Jg/plate
-
+
Shell Oil Co.,
1979
TA100
20, 200, 2000
|Jg/plate
-
+
TA1537
20, 200 |Jg/plate
-
+
TA1535
200, 2000 jjg
-
+
TA1538

-
-
TA97,
TA100,
TA1535
10, 33, 100, 333
|Jg/plate
-
+
NTP, 1993
TA98
100, 333 |Jg/plate
-
+
TA1537

-
NP
TA100
0. 1, 1 |Jmol/plate
-
+
Stolzenberg and
Hine, 1980
TA100
0.01, 0.02, 0.04, 0.1
|Jmol/plate
-
+
Lag et al., 1994
TA1535,
A100
5, 10, 50, 100
|Jg/plate
-
+
Ratpan and
Plaumann, 1988
TA98,
TA1538,
TA1537
N/A
-
-
TA98,
TA100,
TA1535
0.02-1.0 mg/plate
-
+
Kier, 1982
TA1537,
TA1538
N/A
-
-
E. coli(SOS
chromotest)
PQ37

-
-
von der Hude et
al., 1988
E. coli (DNA-repair
deficient strain)
WP2 uvrA
2000 |Jg/plate
-
+
Shell Oil Co.,
1979
E. coli (DNA-repair-
proficient strain)
WP2
N/A
-
-
October, 2007
54
DRAFT - DO NOT CITE OR QUOTE

-------
(b) Mammalian Cell assays
Mouse Lymphoma
L5178Y
0.01, 0.02, 0.03,
0.04, 0.05, 0.06
|jl/ml
-
+
NTP, 1993
L5178Y
2.4,3.2,4.2,5.6,
7.6, 10, 13, 18
Mg/ml
NP
+
Shell Oil Co.,
1982
In Vitro Chromosomal Damage Assays
Test System
Cells/strain
Positive
concentrations
Results
Reference
-S9
+S9
(a) Mammalian Cells
Chromosomal
Aberrations
CHO cells
59.5, 69.4, 79.2
[j,g/mL
-
+
NTP, 1993
Rat liver
epithelial
N/A
-
-
Shell Oil Co.,
1979
Micronucleus
CHO
N/A
+ a
Douglas et al.,
1985 (abstract)
Human
lymphocytes
N/A
-
-
Tafazoli and
Kirsch-Volders,
1996
Micronucleus:
AHH-1
0.01, 1,2, 5 mM
+
NP
Doherty et al.,
1996
MCL-5
1, 2, 5 mM
+
NP
H2E1
0.01, 1,2, 5 mM
+
NP
Unscheduled DNA
synthesis (UDS)
Male rat
hepatocytes
(F344/N)
N/A
-
NP
Williams et al.,
1989
DNA strand breaks
(Comet assay)
Human
lymphocytes
2, 4 mM
+
+
Tafazoli and
Kirsch-Volders,
1996
Wistar rat
hepatocytes
N/A
-
NP
Holme et al.,
1991
DNA
Fragmentation
V79
4, 5 mM
+ a
Eriksson et al.,
1991
Sister chromatid
exchanges
CHO
14.2, 39.7, 49.6,
59.5 (J,g/ml
-
+
NTP, 1993
CHO
N/A
+ a
Douglas et al.,
1985 (abstract)
V79
0.3, 1.0 mM
-
+
von der Hude et
al., 1987
(b) Others:
Cell transformation
Syrian
Hamster
embryo
N/A
+
NP
Hatch et al., 1983
(abstract)
October, 2007
55
DRAFT - DO NOT CITE OR QUOTE

-------
In Vivo Bioassays
Test System
Cells/organs
Positive Doses
Results
Reference
Chromosomal damage: mammalian
Micronucleus
CD-I mice,
bone marrow
erythrocytes
N/A
-
Crebelli et al.,
1999

Mouse bone
marrow
N/A
-
Douglas et al.,
1985 (abstract)
DNA strand breaks
(Comet assay)
F344/N male
rat
hepatocytes
30, 100, 300 mg/kg
+
Weber and Sipes,
1991
Wistar male
rat Kidney
> 375 [j,mol/kg
+
Lag et al., 1991

F344/N rats


Mirsalis et al.,
1983 (abstract)
UDS
(male)
hepatocytes
N/A
—

F344/N male
rat, fore-
stomach and
liver
N/A
+
La and Swenberg,
1997 (abstract)
DNA adducts
F/344/N male
rat (multiple
organs)
3 or 30 mg/kg
+


B6C3F1 male


La et al., 1995

mice
(multiple
organs)
6 or 60 mg/kg
+

Other in vivo assays
Dominant lethal
mutation
SD male rats,
Implants and
embryos
N/A
-
Saito-Suzuki et
al., 1982
Wing spot test
Drosophila
melanogaster
4.51ng/L
(inhalation)
+
Chroust et al.,
2007
High frequency of
activating
mutations in ras
B6C3F1 mice
Forestomach
N/A
+
Ito et al., 1996
(abstract)
oncogenes




Polyploidy
Albino male
rat
hepatocytes
0.8 mg/L
(inhalation)
+
Belyaeva et al.,
1974
0.8, 2.16 mg/L
(inhalation)
+
Belyaeva et al.,
1977
N/A: Either chemical had no effect or information is not available (abstracts only)
NP: Assay is not performed
a Metabolic enzyme induction not specified
October, 2007
56
DRAFT - DO NOT CITE OR QUOTE

-------
Mammalian cell assays
1,2,3-Trichloropropane has also been shown to induce genotoxic effects in cultured
mammalian cells (Table 4-26). NTP (1993) conducted cytogenetic analysis in Chinese hamster
ovary cells, and the results indicated that 1,2,3-trichloropropane induced both sister chromatid
exchanges (SCEs), at 14.2, 39.7, 49.6, and 59.5 [^g/plate, and chromosomal aberrations, at 59.5,
69.4, and 79.2 [j,g/plate, in the presence of rat liver S9 fraction. However, 1,2,3-trichloropropane
did not induce chromosomal damage in cultured rat liver epithelial cells at doses of 250, 500, or
1000 [^g/mL (Shell Oil Co., 1979), nor did it elicit micronucleus formation in isolated human
lymphocytes at doses of 0.1, 2, 4, 6, or 8 mM (0.015, 0.29, 0.59, 0.89, or 1.2 mg/L) (Tafazoli and
Kirsch-Volders, 1996). In a study by Douglas et al. (1985), sister chromatid exchanges and
micronuclei were reported to be induced in CHO cells following 1,2,3-trichloropropane
exposure, although the doses tested and induction levels were not specified. 1,2,3-
Trichloropropane induced sister chromatid exchanges in Chinese hamster V79 cells at 0.3 and
1.0 mM with microsomal activation, but did not induce SCE without microsomal activation (von
der Hude et al., 1987). Eriksson et al. (1991) observed DNA fragmentation in Chinese Hamster
lung fibroblasts (V79) cells at 4 and 5 mM 1,2,3-trichloropropane, although induction levels
were not provided.
1,2,3-Trichloropropane induced micronucleus formation in the mammalian cell lines,
AHH-1, MCL-5, and h2El, in a dose-dependent manner from 0.01 to 5.0 mM for each cell line
(Doherty et al., 1996). The human B lymphoblastoid AHH-1 cell line has native cytochrome
CYP1A1 activity, the MCL-5 cell line expresses cDNAs encoding human CYP1A2, 2A6, 3A4,
and microsomal epoxide hydrolase, and the h2El cell line contains CYP1A1 activity and a
cDNA for CYP2E1. The increase in micronuclei in AHH-1 and h2El was approximately 8-fold,
while the increase in MCL-5 was approximately 4-fold. The micronuclei of all three cell lines
stained both positively and negatively for kinetochore antibody. Although the micronuclei of the
MCL-5 cell line stained primarily positive for kinetochore antibody, indicative of aneugenic
effects, those induced in the AHH-1 and h2El cell lines lacked kinetochore staining, which is
indicative of clastogenic effects. The difference in micronucleus formation between AHH-1 and
h2El and MCL-5 suggests the formation of a less genotoxic or further deactivated metabolite in
the MCL-5 line. The MCL-5 cell line endogenously expresses CYP1A1 and contains cDNAs for
CYP1A2, 2A6, 3A4, and 2E1, while AHH-1 and h2El contain CYP1A1 and CYP1A1 and 2E1,
respectively. The MCL-5 cell line may be capable of metabolizing 1,2,3-trichloropropane to less
genotoxic metabolites or less potent inducer of micronuclei.
October, 2007
57
DRAFT - DO NOT CITE OR QUOTE

-------
Use of an alkaline single cell gel electrophoresis test (Comet assay) demonstrated a
compound-related increase in the incidence of DNA strand breaks under cytotoxic conditions in
isolated human lymphocytes (Tafazoli and Kirsch-Volders, 1996). 1,2,3-Trichloropropane did
not induce DNA strand breaks, measured by alkaline elution, in male Wistar rat hepatocytes after
a 1 hour exposure to 50 |iM (Holme et al., 1991). Hatch et al. (1983) reported that 1,2,3-
trichloropropane enhanced DNA viral transformation in Syrian hamster embryo cells. When
tested for genotoxicity in the rat hepatocyte unscheduled DNA synthesis assay, 1,2,3-
trichloropropane (10"4% M) was negative for unscheduled DNA synthesis, a general response to
DNA damage (Williams et al., 1989).
NTP (1993) found a positive response to 1,2,3-trichloropropane in the mouse lymphoma
assay for induction of trifluorothymidine resistance in L5178Y cells in the presence of rat liver
S9 fraction; the lowest effective dose was 0.01 [iL. Without S9 activation, no induction of
trifluorothymidine resistance was noted at doses below those that produced precipitation of
1,2,3-trichloropropane. Shell Oil Co. (1982) also demonstrated the capacity of the compound to
induce forward mutations to confer trifluorothymidine resistance in mouse lymphoma L5178Y
cells in the presence of S9 fraction, and an inability to induce forward mutations in the absence
of S9 fraction.
In vivo bioassays
In vivo assays provided both positive and negative evidence of genotoxicity (Table 4-26).
Chroust et al. (2007) investigated the genotoxic effects of 1,2,3-trichloropropane in the somatic
mutation and recombination test (SMART) using Drosophila melanogaster. In this bioassay, 72
hour-old larvae were administered 1,2,3-trichloropropane for 48 hours by inhalation, and the
wings of the adults were inspected for the presence of wing spots which were characterized as
small, large twin, and total spots. The induction of wing spots is caused by genotoxic effects
such as somatic mutation, chromosomal rearrangement, or nondisjunction. 1,2,3-
trichloropropane caused a statistically significant (compared to control) increase in the number
of total wing spots.
Belyaeva et al. (1974) investigated the effect of 1,2,3-trichloropropane on the ploidy of
hepatocytes in rats. Male albino rats inhaled 0.8 mg/L 1,2,3-trichloropropane for one week. The
percentage of mononuclear tetraploid and octaploid cells was statistically significantly increased,
and an increase in ploidy of 16n was also evident. There was also a decrease in the percentage
of binuclear cells in concordance with the increase in tetraploid and octaploid. Belyaeve et al.,
(1977) conducted a similar investigation in order to compare the action of various concentrations
October, 2007
58
DRAFT - DO NOT CITE OR QUOTE

-------
of 1,2,3-trichloropropane and 1,2-dichloropropane on the ploidy of hepatocytes. Male albino
rats inhaled 1,2,3-trichloropropane at 0.8 or 2.16 mg/L for one week, 0.08 mg/L for 2 weeks, and
0.002 mg/L for 3 months. After the one week exposure period, the 1,2,3-trichloropropane dosed
animals demonstrated an increase in the number of mononuclear hepatocytes with a nucleus of
high ploidy with a decrease in the number of binuclear cells. Following the 2 week exposure,
however, the results in the experimental group and control group were indistinguishable. When
the exposure time was increased to 3 months and the dose decreased to 0.002 mg/L, a slight
increase in nuclei of intermediate ploidy was observed in the 1,2,3-trichloropropane-exposed
group.
Additional positive evidence of genotoxicity was obtained by Weber and Sipes (1991),
who administered single, i.p. injections of 30, 100, or 300 mg/kg 1,2,3-trichloropropane to male
F-344/N rats, which were then sacrificed 1, 2, 4, 8, 12, 24, and 48 hours post-administration.
Using alkaline elution to detect damaged hepatic DNA, they demonstrated that 1,2,3-
trichloropropane, or its metabolites, caused the formation of DNA strand breaks. La et al. (1995)
characterized the formation of DNA adducts in various organs in both B6C3F1 mice and F-
344/N rats exposed to 6 or 60 and 3 or 30 mg/kg, respectively. High concentrations of DNA
adducts were evident in the tumor-forming organ tissues, including the forestomach, kidney,
pancreas, and liver in male rats and the forestomach, liver, lung and kidney of male mice, from
the NTP (1993) study. DNA adducts were also found in tissues that did not develop tumors,
although the increased incidence of tumors and increased mortality in the NTP study may have
precluded tumor development in those tissues that formed DNA adducts without tumors.
Ito et al. (1996) analyzed the forestomach tumors in the B6C3F1 mice (from the NTP,
1993 bioassay) for ras gene mutations. The DNA was isolated from the paraffin-embedded
forestomach tumor sections and amplified by polymerase chain reaction. Ten of the 16
forestomach tumors contained highly specific H-ras or K-ras activating mutations, of which 6
tumors had H- ras mutations at codon 61 and 4 with K- ras mutations at codon 13. These
mutations are not consistent with the miscoding properties of S-[l-(hydroxymethyl)-2-(N7-
guanyl)ethyl]glutathione, the major DNA adduct, and may indicate that another mechanism may
be involved.
La and Swenberg (1997) analyzed forestomach and liver DNA from F344/N rats exposed
to 1,2,3-trichloropropane for one week at 30 mg/kg-day by corn oil gavage, to examine whether
1,2,3-trichloropropane induces an increase in the concentration of endogenous DNA adducts.
Following the exposure, the DNA adducts identified were 8-hydroxydeoxyguanosine, 1,N6-
ethenodeoxyadenosine, and 3,N4-ethenodeoxycytidine. It was hypothesized from this
October, 2007
59
DRAFT - DO NOT CITE OR QUOTE

-------
investigation that bolus doses of trichloropropane may saturate or deplete cellular defense
mechanisms and increase the concentration of promutagenic lesions (La and Swenberg, 1997).
Male MOL:WIST rats were killed 1 hour after receiving 1,2,3-trichloropropane by ip
administration and the kidney DNA damage was assessed by alkaline elution (Lag et al., 1991).
1,2,3-Trichloropropane was observed to cause DNA breaks in the kidney DNA of rats at doses >
375 [^mol/kg.
Negative results from in vivo assessments were obtained when the compound was
included in a survey of 10 aliphatic halogenated hydrocarbons using the CD-I mouse bone
marrow micronucleus test and 1,2,3-trichloropropane doses of 115 and 200 mg/kg (Crebelli et
al., 1999). Douglas et al. (1985) also found negative results in the micronucleus test in mouse
bone marrow in vivo, although the doses tested were not specified. Similarly, 1,2,3-
trichloropropane did not induce dominant lethal mutations in male Sprague-Dawley rats when
administered by gavage in corn oil at 80 mg/kg-day for 5 days (Saito-Suzuki et al., 1982). 1,2,3-
Trichloropropane did not induce unscheduled DNA synthesis in hepatocytes from male Fischer-
344 rats (Mirsalis et al., 1983)
4.5.3. Structure-Activity Relationships
Halogenated propanes as a class of compounds are generally found to be positive in
assays which indicate mutagenicity (Lag et al., 1994; Ratpan and Plaumann, 1988), and there is
evidence that at least one other member of this group, l,2-dibromo-3-chloropropane (DBCP), is
carcinogenic in whole animal models (NTP, 1982; NCI, 1978). In a study sponsored by the
National Cancer Institute (NCI), DBCP was found to be carcinogenic to Osborne-Mendel rats
and B6C3F1 mice when administered by gavage in corn oil at 15 or 29 mg/kg-day for up to 78
weeks and 114 or 219 mg/kg-day for up to 60 weeks, respectively (NCI, 1978). Neoplasm
formation in the forestomach resulted in reduced survival in both species. These responses are
qualitatively similar to those produced by 1,2,3-trichloropropane. A 2-year inhalation study of
DBCP in F-344/N rats and B6C3F1 mice (NTP, 1982) found increased incidences of nasal,
bronchial, and oral tumors in both sexes of the experimental animals.
The proposed metabolism of 1,2,3-trichloropropane is very similar to the metabolism of
l,2-dibromo-3-chloropropane (DBCP), involving mixed function oxidase catalysed oxidation
and episulfonium ion formation (NTP, 1993). DBCP is a renal and testicular toxicant at acute
doses, leads to tumor development in the forestomach, nasal turbinates and livers of rats and
mice at chronic doses, and forms the same major DNA adduct, S-[l-(hydroxymethyl)-2-(N7-
October, 2007
60
DRAFT - DO NOT CITE OR QUOTE

-------
guanyl)ethyl]-glutathione, as 1,2,3-trichloropropane (Humphreys et al., 1991). DBCP also
induced a dose-related increase in the incidence of aberrant cells in the spermatogonial and bone
marrow cells of male rats after gavage administration of 0.73, 7.3 and 73 mg/kg-day for 5 days
(Kapp, 1979).
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS
4.6.1. Oral Exposure
There are no data on the toxicological effects of exposure to 1,2,3-trichloropropane in
humans via ingestion. Three subchronic studies in rats and mice (NTP, 1993; Merrick et al.,
1991; Villeneuve et al., 1985), a single chronic study in rats and mice (NTP, 1993), and a
reproductive study in mice (NTP, 1990) have investigated the effects of oral exposure in animal
models.
Toxicokinetic studies in mice and rats have examined the absorption, distribution,
metabolism, and elimination of the compound. These studies have documented the rapid
metabolism and excretion of metabolic products in urine, feces, or by exhalation (Mahmood et
al., 1991; Volp et al., 1984). The absorbed fraction of an administrated dose is almost
completely metabolized by a combination of both the phase I and phase II metabolic pathways
(Figure 3-1). Most of the metabolites are rapidly cleared from the body, although a fraction of
the metabolites have been found to bind to intracellular proteins and nucleic acids (Weber, 1991;
Weber and Sipes, 1991, 1990). It is not clear what impact this binding has on the etiology of
noncancer effects, but the formation of DNA-adducts is thought to play an important role in the
carcinogenic activity of the chemical (see Section 4.7.3).
The NTP (1993) toxicology and carcinogenesis studies conducted in F-344/N rats and
B6C3F1 mice constitute the database of chronic oral toxicity studies for 1,2,3-trichloropropane.
The effects of subchronic oral exposure to 1,2,3-trichloropropane have been investigated by NTP
(1993), Merrick et al. (1991), and Villeneuve et al. (1985). A reproductive and fertility
assessment investigation of 1,2,3-trichloropropane was conducted with Swiss CD-I mice (NTP,
1990). 1,2,3-Trichloropropane was administered by corn oil gavage in all of these investigations
except Villeneuve et al. (1985), which provided 1,2,3-trichloropropane to rats via drinking water.
The principal finding of the NTP (1993) chronic toxicity studies was a statistically
significant elevated incidence of tumors in both rats and mice at multiple sites. The tumorogenic
October, 2007
61
DRAFT - DO NOT CITE OR QUOTE

-------
effects of 1,2,3-trichloropropane are discussed in greater detail in Section 4.7. The increased
incidence of tumors was accompanied by a significant decrease in survival. The percent
probability for survival was significantly decreased in rats receiving a dose of 10 mg/kg-day
1,2,3-trichloropropane, or greater, and in mice receiving a dose of 6 mg/kg-day or greater.
Because the decrease in survival was associated with the increased incidence of tumors (NTP,
1993) it was determined that the decrease in survival should not be considered a noncancer
effect. However, it is important to note that the non-neoplastic changes associated with chronic
oral exposure to 1,2,3-trichloropropane occurred at doses that also produced cancer and an
associated decrease in the percent chance of survival.
Statistically significant increases in absolute and relative liver and right kidney weights
were observed in the subchronic and chronic studies. The increase in liver and kidney weights
may be associated with the metabolic role of these organs involving the induction of metabolic
enzymes and other proteins in metabolizing 1,2,3-trichloropropane. However, this metabolic
role may be combined with the binding of 1,2,3-trichloropropane metabolites to hepatic proteins
and DNA in the continuum to liver damage. Corn oil gavage has been shown to increase cell
proliferation in hepatocytes (Rusyn, et al., 1999); however, the NTP assay control animals, to
which the dose groups were compared, also received corn oil gavage. Organ weight increases
were proportionally greater in rats than mice, and increased organ weights were, generally, also
more pronounced in females than males. The variation in the effect on organ weights between
species and sexes indicates that there may be toxicokinetic and toxicodynamic differences that
affect the metabolism of 1,2,3-trichloropropane.
In male rats, a statistically significant decrease in ALT and 5'-nucleotidase was apparent
after chronic exposure to 30 mg/kg-day, while in female mice, a statistically significant increase
in SDH was evident after chronic exposure to 60 mg/kd-day.
In the subchronic studies there was evidence of hepatocellular damage in both rats and
mice. Absolute and relative liver weight increases were observed in male and female rats in
several studies (NTP, 1993; Merrick et al., 1991; Villeneuve et al., 1985), and were also evident
in male and female mice (NTP, 1993). After subchronic exposure, an increase in the incidence
of hepatocellular necrosis was apparent in both rats and mice (NTP, 1993), and the serum
concentrations of ALT, AST, and SDH were increased in female rats (NTP, 1993; Merrick et al.,
1991). Increased serum cholesterol levels and hepatic aminopyrine demethylase activity were
also apparent after subchronic administration (Villeneuve et al., 1985). Serum concentrations of
pseudocholinesterase were decreased in male and female rats after subchronic 1,2,3-
trichloropropane exposure, and reflected a decrease in pseudocholinesterase synthesis (NTP,
October, 2007
62
DRAFT - DO NOT CITE OR QUOTE

-------
1993). Taken as a whole, the increased incidence of hepatocellular necrosis, the increased ALT,
AST, SDH, hepatic aminopyrine demethylase activity and cholesterol serum concentrations,
along with the decreases in pseudocholinesterase synthesis and concentration of 5'-nucleotidase,
is indicative of hepatocellular damage due to 1,2,3-trichloropropane exposure.
Increased absolute and relative kidney weights in male and female rats were apparent in
several subchronic studies (NTP, 1993; Merrick et al., 1991; Villeneuve et al., 1985), with an
inconsistent dose-response pattern for absolute and relative kidney weight in mice (NTP, 1993).
The NTP (1993) chronic study showed an increase in absolute and relative right kidney weights
in rats and in relative right kidney weight in female mice. Overt kidney damage was not evident
in these studies.
In addition to the liver and kidney effects, cardiac and respiratory system effects were
also observed. After subchronic exposure, a decrease in the absolute heart weight in male rats
and in the absolute and relative heart weight in mice was evident (NTP, 1993). Merrick et al.
(1991) reported an increased incidence of inflammation-associated myocardial necrosis in rats,
and an increase in creatine kinase, an indicator of myocardial damage, was evident in male mice
following chronic exposure (NTP, 1993). NTP (1993) also reported epithelial necrosis in the
nasal turbinates of rats and regenerative lung lesions in mice.
Similar effects to those outlined above following subchronic exposure were not observed
in the chronic NTP (1993) studies which employed doses lower than those reported to produce
these effects in the subchronic studies. This was most likely due to decreased survival
attributable to the on-set of cancer in the chronic study. Histopathic lesions were observed in the
liver, kidney, nasal turbinates, and heart of rats and liver, forestomach and lungs of mice
following subchronic oral exposure to 1,2,3-trichloropropane. Travlos et al. (1996) point out in
an investigation into treatment-related increases in clinical chemistry endpoints and
histopathological findings, that treatment-related alteration in clinical chemistry is highly
associated with histopathological changes.
Evidence of hematological effects, including decreased hematocrit values, hemoglobin
concentrations, erythrocyte counts, and elevated leukocytes and segmented neutrophils counts
were observed in both chronic and subchronic NTP studies (1993); however, these effects were
not considered to be biologically relevant. NTP stated that the decreased hematocrit may be
associated with depressed hematopoeisis or to blood loss from neoplasms in the forestomach,
and the increased number of leukocytes likely due to inflammation associated with the
chemically-induced neoplasms (NTP, 1993).
October, 2007
63
DRAFT - DO NOT CITE OR QUOTE

-------
A multigeneration fertility and reproduction assessment (NTP, 1990) found a significant
reduction in the number of fertile pairs of cohabiting Swiss CD-I mice exposed to 60 mg/kg-day
1,2,3-trichloropropane. The reduction in fertility was accompanied by a significant reduction in
the number of live pups per litter and in the proportion of male pups born alive in the fifth
breedings. The decrease in fertility may be related to the observed increase in metestrus, an
infertile period of estrous cycles that was reported during Task 4 of the NTP (1990) study. Male
reproductive performance and fertility were not affected.
4.6.2. Inhalation Exposure
No inhalation studies of 1,2,3-trichloropropane in humans have been reported. A single
study on the acute effects in humans found that all subjects (12/sex) reported irritation (eyes,
throat, and odor) following 15 minute exposures to 100 ppm trichloropropane (isomer and purity
not reported) (Silverman et al., 1946). The database of inhalation toxicity studies in animals
includes two 2-week studies submitted to EPA by Miller et al. (1987a, b), a 4-week range
finding study, two 13-week studies, and two single-generation reproductive assessments
(Johannsen et al., 1988; Biodynamics, Inc., 1979).
Inhalation exposure to 1,2,3-trichloropropane was associated with the following effects:
abnormal physical signs (increased lacrimation, discoloration of the anogenital fur), decreased
weight gain, increased organ weights, and increased incidences of non-neoplastic lesions in the
nasal epithelium, liver, lungs, and spleen (Johannsen et al., 1988; Miller et al., 1987a,b;
Biodynamics, Inc., 1979).
Decreased body weight and weight gain during the pre-mating period was observed in
both male and female rats in a single-generation reproductive study (Johannsen et al., 1988). In
addition, decreased body weight in females was observed during gestation and lactation.
Similar to the oral toxicity database, the inhalation studies found statistically significant
increases in organ weights. Following the 13-week exposure to 1,2,3-trichloropropane,
increased absolute and relative liver weights were observed in male rats exposed to
concentrations of 5, 15, or 50 ppm and increased absolute and relative liver weights were
observed in female rats exposed to 50 ppm and 15 and 50 ppm, respectively (Johannsen et al.,
1988). Increased absolute and relative liver weights were observed following 2-week exposures
to concentrations of 40 or 132 ppm in rats, and 132 ppm in mice (Miller et al., 1987a). Increased
relative lung weights in female rats exposed to concentrations of 15 or 50 ppm for 13 weeks
October, 2007
64
DRAFT - DO NOT CITE OR QUOTE

-------
(Biodynamics, 1979), and increased relative kidney weights were observed in male rats exposed
to concentrations of 50 ppm for 13 weeks (Johannsen et al., 1988).
Increased incidences of non-neoplastic lesions have been observed in the nasal
epithelium, liver, lung, and spleen of rats or mice following inhalation exposure to 1,2,3-
trichloropropane (Johannsen et al., 1988; Miller et al., 1987a, b; Biodynamics, Inc., 1979).
Johanssen et al. (1988) observed peribronchial lymphoid hyperplasia in the three high-dose
treatment groups of male and female rats, hepatocellular hypertrophy in the three high-dose
group males, and hematopoiesis of the spleen in the highest dose group males and in the three
high-dose group female rats. Miller et al. (1987a, b) reported decreased thickness or
degeneration of the olfactory epithelium in rats exposed for 2 weeks to concentrations of 3, 10,
13, 40 or 132 ppm 1,2,3-trichloropropane (Tables 4-17 and 4-18). Similar effects were also
observed in mice that were exposed to 10, 13, 40 or 132 ppm concentrations (Tables 4-19 and 4-
20).
Johannsen et al. (1988) (Biodynamics, Inc., 1979) found an increased incidence of
peribronchial lymphoid hyperplasia in male and female rats that were exposed to 5, 15, or 50
ppm 1,2,3-trichloropropane, but they did not examine epithelial tissue in their investigation.
Lesions remote from the respiratory tract were also observed (Table 4-18). Centrilobular to
midzonal hepatocellular hypertrophy was seen in nearly all male rats that were exposed for 13
weeks to concentrations of 5, 15, or 50 ppm 1,2,3-trichloropropane. However, no evidence of
hepatic effects was found in female rats that were exposed to 50 ppm 1,2,3-trichloropropane.
Conversely, a dose-dependent increase in the incidence and severity of extramedullar
hematopoiesis of the spleen was observed in female but not male rats, although this effect is not
biologically relevant. This differential expression of histopathic lesions suggests that for 1,2,3-
trichloropropane there may be toxicokinetic or toxicodynamic differences between male and
female rats.
4.7. EVALUATION OF CARCINOGENICITY
4.7.1. Summary of Overall Weight of Evidence
Under the Guidelines for Carcinogenic Risk Assessment (U.S. EPA, 2005a), 1,2,3-
trichloropropane is "likely to be carcinogenic to humans", based on a statistically significant and
dose-related increase in the formation of multiple tumors in both sexes of two species from an
NTP (1993) chronic oral bioassay. Statistically significant increases in incidences of tumors of
October, 2007
65
DRAFT - DO NOT CITE OR QUOTE

-------
the oral cavity, forestomach, pancreas, kidney, preputial gland, clitoral gland, mammary gland,
and Zymbal's gland in rats, and the oral cavity, forestomach, liver, and Harderian gland in mice,
were reported.
No human oral exposure studies are available. No information is available on the
carcinogenic effects of 1,2,3-trichloropropane via the inhalation route in humans or animals. US
EPA's Guidelines for Carcinogenic Risk Assessment (2005) indicate that for tumors occurring at
a site other than the initial point of contact, the weight of evidence for carcinogenic potential
may apply to all routes of exposure that have not been adequately tested at sufficient doses. In
addition, the data from the chronic oral study demonstrate that tumors occur in tissues remote
from the site of absorption, such as in the pancreas, kidney, preputial gland, clitoral gland, and
mammary gland. The presence of non-neoplastic lesions in the liver and spleen of rats and mice
following subchronic and shorter inhalation exposure to 1,2,3-trichloropropane (Johannsen et al.,
1988; Miller et al., 1987a, b) indicates that the chemical can enter the blood stream from the
respiratory tract, but the duration of the inhalation studies was too short to show tumor
development. This information suggests that 1,2,3-trichloropropane is likely to be carcinogenic
by the inhalation route of exposure.
4.7.2. Synthesis of Human, Animal, and Other Supporting Evidence
NTP (1993) conducted a 2-year study of the toxicity and carcinogenicity of 1,2,3-
trichloropropane in F-344/N rats. The chemical was administered by corn oil gavage to 60
rats/sex/group. Rats received doses of 0, 3, 10, or 30 mg/kg-day, and after 15 months (65-67
weeks), 8 to 10 rats per group were sacrificed to allow an interim evaluation of all toxicological
parameters and histopathology. Due to high mortality in rats receiving 30 mg/kg at the interim
evaluation, the remaining survivors in that group were sacrificed at week 67 (females) and week
77 (males). In the rats, tumors were evident in the oral cavity, forestomach, pancreas, kidney,
Zymbal's gland of males and females, along with preputial gland tumors in males and clitoral
gland and mammary gland tumors in females. Tumors in the mice were evident in the oral
cavity, forestomach, liver, and Harderian gland of both males and females, and in the
uterine/cervical tissue in females.
Other evidence that supports the carcinogenic potential of 1,2,3-trichloropropane
includes (1) the demonstration that the metabolically activated compound tested positive in a
number of in vitro genotoxicity assays, (2) the demonstrated ability of 1,2,3-trichloropropane
metabolites to bind to intracellular protein and DNA and form DNA adducts, (3) and the
carcinogenicity of a structural analogue of 1,2,3-trichloropropane, l,2-dibromo-3-chloropropane
October, 2007
66
DRAFT - DO NOT CITE OR QUOTE

-------
(NTP, 1982), which produces the same DNA adducts as 1,2,3-trichloropropane (Humphreys et
al., 1991).
4.7.3. Mode of Action Information
4.7.3.1. Hypothesized Mode of A ction
The hypothesized mode of action for 1,2,3-trichloropropane induced carcinogenicity is
through a mutagenic mode of action. Specifically, the data suggest that bioactivated 1,2,3-
trichloropropane may bind directly to DNA resulting in a mutagenic event that may lead to
tumorigenicity in animals. However, although there are in vitro data indicating that 1,2,3-
trichloropropane may be genotoxic, there is a lack of in vivo information linking a mutagenic
mode of action to the observed carcinogenicity in animal bioassays.
In vitro bacterial mutation assays have consistently demonstrated a mutagenic potential,
dependent on S9 activation, for 1,2,3-trichloropropane. Mammalian cell in vitro studies have
shown chromosomal damage, gene mutation, DNA breakage, and micronucleus formation after
1,2,3-trichloropropane exposure. In addition, in vivo assays have demonstrated the ability of
1,2,3-trichloropropane metabolites to bind to hepatic proteins, DNA, and RNA; form DNA
adducts in rats and mice; induce DNA strand breaks in the hepatocytes of rats; and to induce
wing spots (caused by genotoxic alterations such as somatic mutation, chromosomal
rearrangement, or nondisjunction) in D. melanogaster. In vivo studies measuring dominant
lethal induction or micronucleus formation were negative and limit the confidence in the
hypothesized mode of action. Additional in vivo assays which would provide evidence of
mutagenicity, such as mutations in tumor suppressor genes or other mutagenic markers, are
lacking.
4.7.3.2. Experimental Support for the Hypothesized Mode of Action
Strength, consistency, specificity of association
The experimental support for mutagenicity of 1,2,3-trichloropropane is presented in
sequence, with the formation of DNA adducts first, followed by the in vitro and in vivo
evidence.
Evidence for the direct interaction of 1,2,3-trichloropropane with DNA was presented in
vivo (Weber and Sipes, 1990), in which the ability of 1,2,3-trichloropropane metabolites to form
October, 2007
67
DRAFT - DO NOT CITE OR QUOTE

-------
covalent bonnds with hepatic DNA, RNA, and proteins in rats following intraperitoneal
administration was apparent. However, the levels of radioactivity bound to DNA at 72 hours
post administration, were below the level measured for one hour post administration, and may
reflect cytotoxicity and resultant DNA repair or DNA degradation. The administration of three
consecutive i.p. doses, 24 hours apart, of 30 mg/kg 1,2,3-trichloropropane resulted in a doubling
of the amount of radioactivity bound to DNA after the third dose. Weber and Sipes (1990)
conclude that this investigation demonstrates the ability of 1,2,3-trichloropropane or a reactive
metabolite to covalently bind to hepatic DNA, RNA, and protein, and that the covalent binding
increases with multiple doses. Weber and Sipes (1991) administered i.p. injections of 1,2,3-
trichloropropane to male F-344/N rats. Following the extraction of hepatic DNA, they
demonstrated that 1,2,3-trichloropropane caused the formation of DNA strand breaks.
The involvement of glutathione in the activation and binding of a metabolite of 1,2,3-
trichloropropane is supported by the pretreatment of Sprague-Dawley rats with buthionine
sulfoximine (BSO) (Weber and Sipes, 1990). BSO pretreatment causes a decrease in hepatic
glutathione in rats and a subsequent decrease in TCP- or reactive metabolite-binding to DNA.
Study authors suggested that an intermediate of 1,2,3-trichloropropane metabolism may
rearrange to form an episulfonium ion that may bind covalently to DNA.
In a subsequent study, La et al. (1995) characterized the formation of DNA adducts in
various organs in both B6C3F1 mice and F-344/N rats, and found high concentrations of DNA
adducts in the tumor-forming organ tissues, including the forestomach, kidney, pancreas, and
liver in male rats and the forestomach, liver, lung and kidney of male mice, from the NTP (1993)
study (Table 4-24). The target organs of TCP-toxicity; liver, kidney, forestomach, and intestine,
also contained the highest concentration of covalently bound 1,2,3-trichloropropane and related
metabolites (Mahmood et al., 1991). A dose-dependent formation of DNA adducts was also
evident in organs in which tumor formation was not observed. However, the interpretation of the
target organ specificity is complicated due to the high mortality that was seen in the chronic
bioassays. Early mortality may not have allowed tumors in some tissues to fully develop. The
relationship between the adduct-forming tissues of La et al. (1995) and the tumor-forming tissues
of NTP (1993) support a mode of action involving DNA adduct formation. However, the
biological relevance of the major DNA adducts is not known (La et al., 1995).
The S-[l-(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione adduct indentified by La et al.
(1995) is unusual in that it crosslinks a physiological oligopeptide, reduced glutathione, to DNA
by a chemical carcinogen, in this case 1,2,3-trichloropropane (Ozawa and Guengerich, 1983).
The N7-guanyl adducts have an inhibitory effect on sequence-specific DNA binding by
October, 2007
68
DRAFT - DO NOT CITE OR QUOTE

-------
regulatory proteins, due to a destabilization of the guanine nucleobase and spontaneous
degradation (Gasparutto et al., 2005; Ezaz-Nikpay and Verdine, 1994). However, the exact role
of the N7-guanyl adducts is unknown (Gasparutto et al., 2005).
The mutagenic activity of 1,2,3-trichloropropane has been demonstrated in bacterial and
mammalian cell systems treated with 1,2,3-trichloropropane and activated with an S9 fraction
from chemically-induced rat or hamster livers (Tafazoli and Kirsch-Volders, 1996; Lag et al.,
1994; NTP, 1993; Ratpan and Plaumann, 1988; von der Hude et al., 1987; Douglas et al., 1985;
Hatch et al., 1983; Haworth et al., 1983; Kier, 1982; Stolzenburg and Hine, 1980; Shell Oil Co.,
1979, 1982). In the absence of the enzyme-rich S9 fraction mutagenic activity is typically not
observed. Trichloropropane was positive in primarily S. typhimuium strains that detect base pair
mutations (TA1535 and TA100) and frame shift mutations [TA1537 (one assay) and TA98] in
the presence of S9 fraction (Lag et al., 1994; NTP, 1993; Ratpan and Plaumann, 1988; Haworth
et al., 1983; Kier, 1982; Stolzenburg and Hine, 1980; Shell Oil Co., 1979). Mutagenicity was
also evident in E. coli WP2 uvr A, in the presence of S9 fraction, after exposure to 1,2,3-
trichloropropane (Shell Oil Co., 1979). Chromosomal aberrations and sister chromatid
exchanges were evident in Chinese Hamster ovary cell or V79 assays (NTP, 1993; von der Hude
et al., 1987; Douglas et al., 1985), and trifluorothymidine resistance was induced in mouse
lymphoma assays, after 1,2,3-trichloropropane exposure and in the presence of S9 fraction (NTP,
1993; Shell Oil Co., 1982). DNA strand breakage caused by 1,2,3-trichloropropane was
measured by the Comet assay (single gel electrophoresis test) in isolated human lymphocytes
(Tafazoli and Kirsch-Volders, 1996), 1,2,3-trichloropropane enhanced DNA viral transformation
in Syrian hamster embryo cells (Hatch et al., 1983), and 1,2,3-trichloropropane induced
micronucleus formation in the mammalian cell lines, AHH-1, MCL-5, and h2El (Doherty et al.,
1996) and CHO cells (Douglas et al., 1985). The data also demonstrate that the metabolism of
1,2,3-trichloropropane is necessary to activate the chemical's mutagenic potential.
In an in vivo bioassay in D. melanogaster, Chroust et al. (2007) investigated the
genotoxic effects of 1,2,3-trichloropropane in the somatic mutation and recombination test
(SMART). 1,2,3-trichloropropane caused a statistically significant (compared to control)
increase in the number of total wing spots, which is evidence for genotoxic effects such as
somatic mutation, chromosomal rearrangement, or nondisjunction. Belyaeva et al. (1974, 1977)
observed an increase in the number of mononuclear hepatocytes with a nucleus of high ploidy
and a decrease in the number of binuclear cells following exposure to 1,2,3-trichloropropane.
1,2,3-Trichloropropane also caused DNA breaks in the DNA from isolated kidney nuclei of rats
exposed to 1,2,3-trichloropropane (Lag et al., 1991).
October, 2007
69
DRAFT - DO NOT CITE OR QUOTE

-------
Ito et al. (1996) analyzed the forestomach tumors in the B6C3F1 mice from the NTP,
1993 bioassay for ras gene mutations. Ten of the 16 forestomach tumors contained highly
specific H-ras or K-ras activating mutations, of which 6 tumors had H-ras mutations at codon 61
and 4 with K-ras mutations at codon 13. These mutations are not consistent with the miscoding
properties of S-[l-(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione, the major DNA adduct of
1,2,3-trichloropropane, and indicates that another mode of action may be involved. La and
Swenberg (1997) observed an increase in the concentration of the endogenous DNA adducts, 8-
hydroxydeoxyguanosine, l,N6-ethenodeoxyadenosine, and 3,N4-ethenodeoxycytidine, in rats
following 1,2,3-trichloropropane exposure for one week.
1,2,3-Trichloropropane tested negative in bacterial and mammalian cell systems not
activated with S9 fraction (NTP, 1993; Ratpan and Plaumann, 1988), in the SOS chromotest in
E. coli (von der Hude et al., 1988), in the DNA-repair proficient E. coli WP2 (Shell Oil Co.,
1979), and in the Aspergillus nidulans diploid strain PI assay for aberrant mitotic segregation
(Crebelli et al., 1992). Mammalian cell assays in which 1,2,3-trichloropropane tested negative
for genotoxicity included the induction of trifluorothymidine resistance in mouse lymphoma
cells not activated with S9 fraction (NTP, 1993; Shell Oil Co., 1982), the induction of
chromosomal damage in Carworth Farm E rat liver epithelial cells (Shell Oil Co., 1979), the
micronucleus formation assay in human lymphocytes (Tafazoli and Kirsch-Volders, 1996), the
unscheduled DNA synthesis assay in rat hepatocytes (Williams et al., 1989), and the induction of
DNA strand breaks in Wistar rat hepatocytes (Holme et al., 1991). The in vivo assays in which
1,2,3-trichloropropane tested negatively included the bone marrow micronucleus formation
assay in CD-I mice (Crebelli et al., 1999) and in an unspecified mouse strain (Douglas et al.,
1985), the induction of unscheduled DNA synthesis in F344/N rats (Mirsalis et al., 1983), and
the dominant lethal induction assay in male SD rats (Saito-Suzuki et al., 1982).
An in vitro assay conducted by Weber and Sipes (1992), utilizing rat and human hepatic
cells, demonstrated a dose-dependent increase in the formation of the intermediate metabolite,
1,3-dichloroacetone (DCA), which the study authors characterized as a direct-acting mutagen.
1,3-Dichloroacetone, also referred to as 1,3-dichloropropanone or l,3-dichloro-2-propanone, has
shown mutagenicity in Salmonella typhimurium TA100 without microsomal activation (Meier et
al., 1985). 1,3-Dichloroacetone was also shown to be mutagenic in TA1535 and TA100 with
and without metabolic activation, with increased mutagenicity in strains TA1535 and TA100
without microsomal activation compared to the same strains with activation (Merrick et al.,
1987).
October, 2007
70
DRAFT - DO NOT CITE OR QUOTE

-------
1,3-Dichloroacetone initiated skin tumors after both single and repeated topical treatment
of female SENCAR mice followed by the tumor promoter 12-0-tetradecanoyl-phorbol-13-
acetate (TPA) (Robinson et al., 1989). The percentage of tumor-bearing mice after a single
initiating dose of 37.5, 75, or 150 mg/kg 1,3-dichloroacetone was 47, 47 and 68%, respectively.
The percentage of tumor-bearing mice after repeated doses of 300, 450, or 600 mg/kg 1,3-
dichloroacetone was 48, 45, and 32%, respectively. In control mice receiving ethanol the
percentage of tumor-bearing mice observed was 12%. The inverted dose response observed in
mice under the repeated dosing regimen may have been the result of localized cellular toxicity
which prevented initiated cells from progressing to papilloma (Robinson et al., 1989). The
association between this cellular injury and the increased incidence of carcinomas in animals
receiving repeated doses is uncertain and needs to be investigated (Robinson et al., 1989).
Dose-response concordance
The in vitro studies were positive for genotoxicity or mutagenicity at concentrations
ranging from 0.001 to 1000 [j,g/plate, and indicate that point mutations are the most consistent
type of genetic alteration induced by 1,2,3-trichloropropane and occur at lower concentrations
than the chromosomal damage.
La et al. (1995) characterized the formation of DNA adducts in various organs in both
B6C3F1 mice and F-344/N rats, and found high concentrations of DNA adducts at 6 hours post-
administration in the tumor-forming organ tissues, including the forestomach, kidney, pancreas,
and liver, in male rats at 3 or 30 mg/kg-day and in the forestomach, liver, lung and kidney of
male mice at 6 or 60 mg/kg-day, from the NTP (1993) study. The formation of DNA adducts
displayed a dose-dependent increase in the same organs that displayed a similar dose-dependent
increase in tumor incidence from the NTP (1993) study.
The binding of 1,2,3-trichloropropane or related metabolites to DNA increased with
multiple doses of 30 mg/kg-day administered 24 hours apart (Weber and Sipes, 1990).
Polyploidy was apparent in the hepatocytes of male albino rats dosed with 0.8 and 2.16 mg/L for
two hours (Belyaeva et al., 1974), and a dose-dependent increase in DNA strand breaks was
evident in hepatocytes from F344 rats at 30-100 mg/kg (Weber and Sipes, 1991) and in kidney
cells from male Wistar rats at >375 mmol/kg (Lag et al., 1991). Ito et al. (1996) observed a high
frequency of activating mutations in ras oncogenes in the forestomach tumors from the NTP
(1993) bioassay. A dose-dependent increase in the incidence of tumors was observed in rats
from 3 to 30 mg/kg-day and in mice from 6 to 60 mg/kg-day (NTP, 1993). The in vivo data
demonstrates an increase in DNA-binding capability, DNA strand breaks, and DNA adducts at
October, 2007
71
DRAFT - DO NOT CITE OR QUOTE

-------
doses of 1,2,3-trichloropropane that are similar to the dose levels administered in the NTP (1993)
bioassay in which an increased incidence of tumors in multiple organs was observed at all dose
levels tested.
Temporal relationship
The temporal relationship for mutagenicity and tumorigenicity has not been adequately
studied. However, data indicate that metabolism of 1,2,3-trichloropropane to its metabolites may
be a key event in the mutagenic mode of action. 1,2,3-Trichloropropane metabolism follows
three potential routes, each of which involves glutathione at different steps in the metabolism
process. Two primary routes of metabolism involve the formation of an episulfonium ion, while
the third involves the intermediate metabolite, 1,3-dichloroacetone (Mahmood 1991), which is a
reported mutagen (Weber and Sipes, 1992).
In addition, there are in vitro and in vivo data that demonstrate metabolism of 1,2,3-
trichloropropane, followed by binding of reactive metabolites to DNA, and the ultimate
formation of DNA adducts. This sequence of events has been demonstrated in the bacterial and
mammalian cell systems assays in which activation with an S9 fraction from chemically-induced
rat or hamster livers may be necessary for genotoxicity and potential mutagenicity (Tafazoli and
Kirsch-Volders, 1996; Lag et al., 1994; NTP, 1993; Ratpan and Plaumann, 1988; von der Hude,
1987; Douglas et al., 1985; Hatch et al., 1983; Haworth et al., 1983; Kier, 1982; Stolzenburg and
Hine, 1980; Shell Oil Co., 1979, 1982).
Evidence for the direct interaction of 1,2,3-trichloropropane and its metabolites with
DNA, RNA, and hepatic proteins was observed 4 hours following intraperitoneal administration
of 1,2,3-trichloropropane (Weber and Sipes, 1990). This investigation demonstrates the ability
of 1,2,3-trichloropropane or a reactive metabolite to bind to hepatic DNA, RNA, and protein,
and that the binding increases with multiple doses. DNA strand breaks were evident in the
extracted hepatic DNA of male F-344/N rats administered 1,2,3-trichloropropane by ip injection,
thus demonstrating that 1,2,3-trichloropropane, or a reactive metabolite, causes the formation of
DNA strand breaks (Weber and Sipes, 1991).
La et al. (1995) characterized the formation of DNA adducts in various organs in both
B6C3F1 mice and F-344/N rats 6 hours following a single dose of 1,2,3-trichloropropane, and
found high concentrations of DNA adducts in the tumor-forming organ tissues, including the
forestomach, kidney, pancreas, and liver in male rats and the forestomach, liver, lung and kidney
of male mice.
October, 2007
72
DRAFT - DO NOT CITE OR QUOTE

-------
Biological plausibility and coherence
Mutagenicity as a mode of action for carcinogenicity in humans is generally accepted
and is a biologically plausible mechanism for tumor induction. The formation of DNA adducts in
organs that also displayed an increase in the tumor incidence in rats and mice indicates
coherence of the effects and is evidence supporting a mutagenic mode of action (Table 4-24).
Binding of 1,2,3-trichloropropane metabolites to DNA is currently the most likely theory for the
mode of action of the tumor formation. However, the formation of DNA adducts of 1,2,3-
trichloropropane in tissues other than those where tumors formed (La et al., 1995) is an area of
uncertainty associated with the suggested mutagenic mode of action. DNA adduct formation for
some tumor types may be necessary but not sufficient for the induction of tumors and is not an
uncommon occurrence as DNA adducts of known direct-acting carcinogens (e.g.,
benzo[a]pyrene) have been observed in tissues where tumors were not found. The formation of
DNA adducts in non-tumor forming tissues and organs may signify that DNA adducts by
themselves are insufficient to cause tumors or that the increased mortality in the rats and
increased tumor incidence in other organs precluded tumor formation in the non-tumor forming
organs.
4.7.3.3. Other Possible Modes of A ction
Data are not available to make a determination about whether other modes of action, such
as cytotoxicity with tissue repair due to DNA degradation or disruption of cell signaling, are
associated with the carcinogenic activity of 1,2,3-trichloropropane. Holme et al. (1989) found
that l,2-dibromo-3-chloropropane (DBCP), a structurally-related compound to 1,2,3-
trichloropropane, induced DNA damage in liver cells at concentrations much lower than
concentrations that resulted in cytotoxicity and bacterial (S. typhimurium) mutagenicity.
CYP450 and glutathione transferase appeared to be involved in the damage to cellular DNA
caused by DBCP, with the CYP450 dependent oxidation also resulting in bacterial mutagenicity.
The activation of CYP450 may result in the in vitro mutagenicity, while the DNA damage and
cytotoxicity of DBCP in vivo may be glutathione-dependent (La et al., 1995).
Data are available that indicate that the bolus exposure to 1,2,3-trichloropropane may
overwhelm cellular glutathione levels in the forestomach and induce lipid peroxidation (La and
Swenberg, 1997; Ito et al., 1996). This lipid peroxidation leads to an increase in the etheno
DNA adducts l,N6-ethenodeoxyadenosine and 3,N4-ethenodeoxycytidine and the hydroxyl
radical-derived 8-hydroxydeoxyguanosine (La and Swenberg, 1997; Ito et al., 1996).
October, 2007
73
DRAFT - DO NOT CITE OR QUOTE

-------
4.7.3.4. Conclusions About the Hypothesized Mode of Action
The mode of action for 1,2,3-trichloropropane tumorigenicity may involve mutagenicity
via reactive metabolites. However, there is uncertainty regarding this mode of action that is
outlined later in this section. The data supporting a mutagenic mode of action include:
•	mutagenic response in short-term bacterial assays (with microsomal activation),
indicative of base-pair substitutions and frameshift mutations, and induced
chromosomal damage, gene mutations, DNA breakage, micronucleus formation,
and enhanced DNA viral transformation in mammalian cell assays
•	covalent binding of 1,2,3-trichloropropane metabolites to hepatic protein, DNA,
and RNA and the induction of DNA strand breaks in the hepatocytes of rats
following in vivo exposure, and induced wing spot formation in the somatic
mutation and recombination test in I), melanogaster
•	dose-dependent formation of DNA adducts, including the major adduct S-[l-
(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione, in various organs of both
B6C3F1 mice and F-344/N rats, with DNA adducts present in tumor-forming
organs of male rats and mice
•	dose-dependent increase in the formation of the intermediate metabolite, and
reported mutagen and tumor initiator, 1,3-dichloroacetone, and the formation of
reactive episulfonium ion metabolites.
The available in vitro and in vivo data also indicate that metabolites of 1,2,3-
trichloropropane have an affinity for nucleic acids and a capacity to form DNA adducts;
however, the available database lacks in vivo assays which measure mutagenicity, such as
mutations or chromosomal damage in target organs, as well as dose-response and temporal data
that would provide additional evidence that 1,2,3-trichloropropane mutagenicity leads to
carcinogenesis.
A number of assays have tested negative for DNA reactivity and mutagenicity of 1,2,3-
trichloropropane. The assays in which 1,2,3-trichloropropane tested negative include:
•	Ames assays without activation by S9 fraction, in the SOS chromotest in E. coli, in the
DNA-repair proficient E. coli WP2,
•	the induction of trifluorothymidine resistance in mouse lymphoma cells not activated
with S9 fraction,
October, 2007
74
DRAFT - DO NOT CITE OR QUOTE

-------
• the micronucleus formation assay in human lymphocytes,
• the unscheduled DNA synthesis assay in rat hepatocytes,
• the induction of chromosomal damage in cultured rat liver cells;
• in vivo studies measuring dominant lethal induction or micronucleus formation.
However, Crebelli et al. (1999) stated that micronucleus formation in mouse bone
marrow is weakly sensitive to the genotoxic effects induced by halogenated
hydrocarbons in other test systems, and a negative bone marrow micronucleus assay
should not offset the consistently positive in vitro results (Dearfield and Moore, 2005).
In vivo data supporting a mutagenic mode of action for carcinogenicity are limited and
areas of uncertainty exist. For example, regular test batteries for different genetic end points in
vitro and, especially, in vivo, such as micronucleus formation, chromosomal aberrations,
unscheduled DNA synthesis, sister chromatid exchanges, Comet assay, and DNA adduct
analysis, are limited or missing from the database. Evidence of gene mutations would provide
substantial support for a mutagenic mode of action, but these studies have not been conducted.
Also, evidence of cytogenetic effects in humans would be useful to better characterize the mode
of action for 1,2,3-trichloropropane. While the most plausible scenario is that bioactivated 1,2,3-
trichloropropane acts by inducing mutations in cancer-related genes, data demonstrating these
events are largely unavailable.
Is the hypothesized mode of action sufficiently supported in test animals?
The covalent binding of bioactivated 1,2,3-trichloropropane to hepatic DNA, RNA, and
protein was evident in male F-344/N rats, with approximately half the amount bound to DNA as
the amount bound separately to RNA or protein (Weber and Sipes, 1990). A dose-dependent
increase in the amount of 1,2,3-trichloropropane equivalents bound to hepatic DNA and protein
was demonstrated, with the amount bound to hepatic protein increasing linearly with increasing
doses over time.
Weber and Sipes (1991) administered i.p. injections to male F-344/N rats. Following the
extraction of hepatic DNA, they demonstrated that 1,2,3-trichloropropane, or its metabolites,
caused the formation of DNA strand breaks.
La et al. (1995) characterized the formation of DNA adducts in various organs both in
B6C3F1 mice and F-344/N rats, and found high concentrations of DNA adducts in organ tissues
in which tumor formation was observed by NTP (1993). The investigators characterized the
DNA adduct, indicating that a single, major 1,2,3-trichloropropane-derived DNA adduct was
formed irrespective of the tissue type, and determined the adduct to be S-[l-(hydroxymethyl)-2-
October, 2007
75
DRAFT - DO NOT CITE OR QUOTE

-------
(N7-guanyl)ethyl]glutathione. The formation of this adduct was detected in the forestomach,
glandular stomach, kidney, liver, pancreas, and tongue of F-344/N rats, and in the forestomach,
glandular stomach, kidney, and liver of B6C3F1 mice. The concentrations of adduct formed in
the target organs showed correlation with the tumor incidence from the NTP (1993) study.
The target organs of TCP-toxicity, liver, kidney, forestomach, and intestine, also contain
the highest concentration of covalently-bound 1,2,3-trichloropropane and related metabolites
(Mahmood et al., 1991which supports a role for metabolic activation and binding in the early
stages of carcinogenesis.
Is the hypothesized mode of action relevant to humans?
The postulated key events, the metabolism of 1,2,3-trichloropropane to a DNA-reactive
compound and the alteration of the genetic material leading to tumor-inducing mutations, are
both possible in humans. The toxicokinetic and toxicodynamic processes that would enable
reactive metabolites to produce mutations in animal models are biologically plausible in humans.
1,2,3-trichloropropane in the presence of S9 fraction is positive in in vitro S. typhimurium
assays that detect base pair mutations (TA1535), frame shift mutations (TA1537) and primary
DNA damage (TA98, TA100) (Lag et al., 1994; NTP, 1993; Ratpan and Plaumann, 1988;
Haworth etal., 1983; Kier, 1982; Stolzenberg and Hine, 1980; Shell Oil Co., 1979) When tested
in mammalian cells, 1,2,3-trichloropropane has induced chromosomal aberrations and sister
chromatid exchanges in Chinese hamster ovary cells (NTP, 1993; Doulgas et al., 1985), induced
forward mutations in the mouse lymphoma assay (NTP, 1993; Shell Oil Co., 1979), enhanced
DNA viral transformation in Syrian hamster embryo cells (Hatch et al., 1983), induced
micronucleus formation in the mammalian cell lines, AHH-1, MCL-5, and h2El (Doherty et al.,
1996), and induced DNA strand breaks in human lymphocytes (Tafazoli and Kirsch-Volders,
1996).
The in vivo data supporting a mutagenic mode of action include the covalent binding of
1,2,3-trichloropropane or a metabolite to hepatic proteins, DNA, and RNA, the formation of
DNA strand breaks in the hepatocytes of F344/N rats (Weber and Sipes, 1991, 1990), the
induction of DNA strand breaks in the kidney DNA of rats (Lag et al., 1991), the formation of
DNA adducts in various organs in both B6C3F1 mice and F-344/N rats with high concentrations
of DNA adducts in the tumor-forming organ tissues (La et al., 1995), an increase in the number
of mononuclear hepatocytes with a nucleus of high ploidy (Belyaeva et al., 1974, 1977), and the
induction of wing spots in Drosophila melanogaster following 1,2,3-trichloropropane treatment
(Chroust et al., 2006).
October, 2007
76
DRAFT - DO NOT CITE OR QUOTE

-------
Ito et al. (1996) observed highly specific H-ras or K-ra.s activating mutations in
the forestomach tumors of the B6C3F1 mice from the NTP (1993) study, with 6 of these
tumors containing H- ras mutations at codon 61 and 4 with K- ras mutations at codon 13.
La and Swenberg (1997) observed an increase in the concentration of the endogenous
DNA adducts, 8-hydroxydeoxyguanosine, l,N6-ethenodeoxyadenosine, and 3,N4-
ethenodeoxycytidine, in rats following 1,2,3-trichloropropane exposure for one week.
In addition to the experimental data for 1,2,3-trichloropropane, halogenated propanes, as
a class of compounds, are generally considered to be mutagenic (Lag et al., 1994; Ratpan and
Plaumann, 1988).
Which populations or lifestages can be particularly susceptible to the hypothesized mode of
actionl
According to the Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens {Supplemental Guidance) (U.S. EPA, 2005b), children exposed to
carcinogens with a mutagenic mode of action are assumed to have increased early-life
susceptibility. The Supplemental Guidance (US EPA, 2005b) recommends the application of
age-dependent adjustment factors (ADAFs) for carcinogens that act through a mutagenic mode
of action and are assumed to convey early-life susceptibility. Given the weight of the available
evidence, 1,2,3-trichloropropane may be acting through a mutagenic mode of carcinogenic
action; however, the database is lacking in vivo evidence that mutagenic events occur following
1,2,3-trichloropropane exposure. For these reasons, the application of ADAFs when assessing
risks associated with early-life exposure is not recommended.
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
4.8.1. Possible Childhood Susceptibility
No studies are available that address the possible adverse effects of 1,2,3-
trichloropropane in children. However, there is evidence that 1,2,3-trichloropropane is
mutagenic and, therefore, may act through a mutagenic mode of action for carcinogenicity. This
would indicate an increased carcinogenic susceptibility for early-life exposures. Although
developmental toxicity studies for 1,2,3-trichloropropane are unavailable, developmental toxicity
is a concern due to the genotoxicity of 1,2,3-trichloropropane and the possibility for genetic
damage to the germ cells of the F1 generation that could be transmitted to the F2 generation. In
addition, the two-generation reproductive assessment by gavage indicates that the developing
October, 2007
77
DRAFT - DO NOT CITE OR QUOTE

-------
fetus may be a target of toxicity due to an observed reduction in the number of live mouse
pups/litter and in the proportion of male pups born alive following oral exposure.
4.8.2. Possible Gender Differences
The extent to which men and women differ in susceptibility to 1,2,3-trichloropropane is
unknown. However, some data may exist that imply a difference between male and female rats
in their response to inhalation of the compound. For example, 15/15 male CD rats exposed to
1,2,3-trichloropropane via inhalation (6 hours/day, 5 days/week, for 13 weeks) at a concentration
of 50 ppm displayed histopathological lesions in the liver, while 0/15 females displayed this
effect at the same concentration (Johannsen et al., 1988). A clear-cut dose-dependent response
in this effect was seen in the males, but females showed no response. The biological
significance of this finding for lower doses and for other species is unclear.
4.8.3. Other
Glutathione appears to be necessary for the formation of the DNA adduct S-[l-
(hydroxymethyl)-2-(N7-guanyl)ethyl]glutathione. Individuals with a glutathione deficiency may
be less susceptible to the carcinogenic effects of 1,2,3-trichloropropane. Conversely, individuals
with increased expression of glutathione may have an increased susceptibility to the genotoxic
effects of 1,2,3-trichloropropane. CYP450 may also influence the toxicity of 1,2,3-
trichloropropane in a similar manner, with altered CYP450 activity increasing the availability of
hepatotoxic but not DNA-reactive metabolites.
October, 2007 78 DRAFT - DO NOT CITE OR QUOTE

-------
5. DOSE RESPONSE ASSESSMENT
5.1. CHRONIC ORAL REFERENCE DOSE (RfD)
5.1.1. Choice of Principal Study and Critical Effect - with Rationale and Justification
Data on the health effects of oral exposure to 1,2,3-trichloropropane in humans are not
available. The database of chronic and subchronic animals studies includes a 2-year gavage
study in F-344 rats and B6C3F1 mice (NTP, 1993; Irwin et al., 1995), a 90-day gavage study in
Sprague-Dawley rats (Merrick et al., 1991), a 90-day drinking water study in Sprague-Dawley
rats (Villeneuve et al., 1985), a 17-week gavage study in F-344 rats (Hazleton Laboratories,
1983a; NTP, 1993), a 17-week gavage study in B6C3F1 mice (Hazleton Laboratories, 1983b;
NTP 1993), and a two-generation reproductive/fertility assessment in Swiss CD-I mice (NTP,
1990). The subchronic (i.e., 90-day study or less) study data were not considered in the selection
of a principal study for deriving the chronic RfD because the database contains reliable dose-
response data from a chronic study of two species and a two-generation reproductive assessment.
The data from the subchronic studies are, however, used to corroborate the findings of the
chronic studies.
The dose-dependent, non-cancer effects associated with oral exposure to 1,2,3-
trichloropropane include increased liver weights (subchronic and chronic); increased kidney
weights (subchronic and chronic); hepatic, renal, myocardial, lung, and nasal turbinate epithelial
necrosis (subchronic); decreased synthesis of pseudocholinesterase (subchronic); decreased ALT
and 5'-nucleotidase levels (chronic); increased ALT, AST, and SDH levels (subchronic and
chronic); increased hepatic aminopyrine demethylase and aniline hydroxylase activity
(subchronic); elevated creatine kinase (chronic); decreased number of pregnancies per fertile
pair, reduction in number of live pups/litter, and decreased proportion of male pups born alive
(NTP, 1993; Merrick et al., 1991; NTP, 1990; Villeneuve et al., 1985).
The NTP (1993) study is selected as the principal study because it was a well-designed
chronic study, conducted in both sexes in two species with a sufficient number of animals per
dose group. The number of test animals allocated among three dose levels and an untreated
control group was acceptable, with examination of appropriate toxicological endpoints in both
sexes of rats and mice. Increased liver weight is chosen as the critical effect because liver
toxicity appeared to be the most sensitive effect. There is evidence of hepatocellular damage,
including increased incidence of hepatic necrosis and decreased synthesis of
October, 2007
79
DRAFT - DO NOT CITE OR QUOTE

-------
pseudocholinesterase, from the subchronic NTP (1993) studies, and increased serum
concentrations of hepatocellular enzymes, decreased concentration of 5'-nucleotidase, and
increase incidence of histopathologic liver lesions, from the chronic NTP (1993) studies, thus,
increased liver weight may be on a continuum of adverse liver effects associated with oral
exposure to 1,2,3-trichloropropane. The designation of the liver as a target organ for noncancer
effects is consistent with the findings from the mechanistic data (Weber and Sipes, 1990) which
demonstrate the binding of 1,2,3-trichloropropane metabolites to hepatic proteins and nucleic
acids.
Other possible critical effects include kidney, respiratory, myocardial, or reproductive
toxicity. The increase in kidney weights after both subchronic and chronic exposure is
accompanied by renal tubular necrosis in the subchronic NTP (1993) study. The subchronic
NTP (1993) study demonstrated epithelial necrosis in the nasal turbinates of rats and
regenerative lung lesions in mice. Merrick et al. (1991) showed an increased incidence of
inflammation-associated myocardial necrosis in rats, and increased levels of creatine kinase were
apparent in the chronic NTP study. NTP (1990) demonstrated a decrease in the number of
pregnancies per fertile pair, a reduction in the number of live pups/litter, and a decrease in the
proportion of male pups born alive. Although the liver appeared to be the most sensitive
indicator of 1,2,3-trichloropropane-induced toxicity, reference doses for the changes in kidney
weight, fertility, and pups/liter were quantified for comparison purposes .
5.1.2. Methods of Analysis - Including Models
Benchmark dose (BMD) modeling was conducted using EPA BMD software version
1.4.1. to analyze the changes in liver and kidney weight, fertility, and pups/litter associated with
chronic exposure to 1,2,3-trichloropropane (see Appendix B for details). The software was used
to calculate potential points of departure for deriving the chronic RfD by estimating the effective
dose at a specified level of response (BMDX) and its 95% lower bound (BMDLX). For
continuous endpoints, the Benchmark Dose Technical Guidance Document (US EPA, 2000c)
states that a minimal level of change in an endpoint that is generally considered to be
biologically significant may be used to define the BMR. For this analysis of absolute and
relative liver and kidney weight changes in both rats and mice, a BMR of 10% is analogous to
the 10%) change in body weight used to identify maximum tolerated doses (MTDs) (A BMR of 1
SD was also included for comparison with other chemicals affecting absolute and relative liver
weight changes). In the developmental study, a 1% change in mean live pups/litter for the 4th
and 5th litters was selected as the BMR due to the frank toxicity of the reproductive toxicity
endpoint. Absolute and relative liver weight changes were also modeled using a BMR of 1 SD,
October, 2007
80
DRAFT - DO NOT CITE OR QUOTE

-------
as recommended by the Benchmark Dose Technical Guidance Document (US EPA, 2000) when
a BMR representing a minimal level of change is selected as the primary BMR for the analysis.
Table 5-1 presents BMDs and the corresponding lower 95% confidence limits (BMDLs)
for each observed effect that was considered and amenable to modeling. The candidate BMD for
each endpoint was identified by comparing the outputs from best fitting models for each of the
four data sets: male rats, female rats, male mice, and female mice. Model fit was determined by
assessing the goodness-of-fit using a significance value of a = 0.1 for eligibility, visual fit, and
ranking by Akaike Information Criterion (AIC) (see Appendix B).
October, 2007
81
DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-1. Candidate benchmark doses for chronic and reproductive effects
associated with oral exposure to 1,2,3-trichloropropane		
End Point
Species/ Sex
Model
AIC
BMDa
BMDLa
BMR
Absolute liver
weight
Rat/male
Hill
63.9
3.8
1.6
10%
change in
mean organ
weight
Absolute liver
weight
Rat/ male
Hill
63.9
3.2
1.4
1 SD
Relative liver
weight
Rat/male
Hill
98.8
5.5
3.1
10%
change in
mean organ
weight
Relative liver
weight
Rat/male
Hill
98.8
3.2
1.8
1 SD
Absolute
kidney weight
Rat/female
Hill
-151.8
9.0
3.4
10%
change in
mean organ
weight
Relative
kidney weight
Rat/male
Hill
-84.1
10.5
6.4
10%
change in
mean organ
weight
Fertility
generating 4th
litter
mice
log Probit
(slope >1)
46.5
52.6
37.3
10%
change in
fertility rate
Fertility
generating 5th
litter
mice
Probit
102.20
31.2
23.3
10%
change in
fertility rate
Live
pups/litter-
4th litter
mice
polynomial
295.6
13.8
3.2
1% change
in mean live
pups/litter
Live
pups/litter-
5th litter
mice
polynomial
193
13.6
5.6
1% change
in mean live
pups/litter
" The lowest BMD and BMDLs from the best-fitting models for each endpoint (see Appendix B).
Both increased absolute and relative liver weights in male rats were fitted adequately.
The increase in liver weight was chosen as the critical effect and absolute liver weight was
selected to represent the increase in liver weight because it is a more direct measure of liver
weight change, as opposed to relative liver weight, which is the ratio of liver-to-body weight and
can be affected by decreased body weight with an increase in dose. The change in liver weight
is the first effect evident, and the increases in liver enzymes associated with liver damage and
October, 2007	82	DRAFT - DO NOT CITE OR QUOTE

-------
increased rate of hepatic necrosis, as well as the hepatocellular damage-related decrease in
pseudocholinesterase, supports the liver as the critical target organ.
The increase in absolute liver weight is a more sensitive endpoint than the decrease in the
number of live pups/litter in the fourth and fifth litters. Statistically significant reductions in the
number of live pups/litter were observed in mice compared to controls in the second through the
fifth breedings at the highest dose (120 mg/kg-day) and at the fifth breeding at 60 mg/kg-day.
When comparing the benchmark dose modeling results, the lower point of departure for the
increase in absolute liver weight (BMDLi0% of 1.6 mg/kg-day) is thought to be more sensitive
than the decrease in the number of live pups/litter (BMDLio/o of 3.2 mg/kg-day), even though the
decrease in live pups/litter is a frank effect.
Consideration of the available dose-response data to determine an estimate of oral
exposure that is likely to be without an appreciable risk of adverse health effects over a lifetime
has led to the selection of the two-year oral gavage study in Fischer rats (NTP, 1993) and
increased liver weight in males as the principal study and critical effect for deriving the chronic
RfD for 1,2,3-trichloropropane. The dose-response relationships for oral exposure to 1,2,3-
trichloropropane and impaired fertility in CD-I mice are also suitable for deriving a chronic
RfD, but are associated with higher BMDLs that would be protected by the selected critical
effect and corresponding BMDL.
The BMDL analysis corresponds to the lower bound on the dose associated with a 10%
increase in mean liver weight. The BMD calculated from the Hill model for absolute liver
weight change in male F-344 rats based on a BMDio is 3.8 mg/kg-day and the BMDLio is 1.6
mg/kg-day. The benchmark response (BMR) of 10% change in mean value was used for
absolute and relative liver and kidney weight modeling because the Benchmark Dose Technical
Guidance Document (US EPA, 2000) recommends using a minimal amount of change in the
endpoint that is considered to be biologically significant to define the BMR. Duration-
adjustment of the point of departures was done to approximate daily exposure by multiplying the
BMDio and BMDLio by (5 days)/(7 days) = 0.71; resulting in a BMDadj of 2.70 mg/kg-day and
a BMDLadj of 1.1 mg/kg-day.
5.1.3. Chronic RfD Derivation - Including Application of Uncertainty Factors (UFs)
A BMDLadj of 1.1 mg/kg-day for increased absolute liver weight in male rats
chronically exposed to 1,2,3-trichloropropane by gavage (NTP, 1993) was used as the point of
departure to calculate the chronic RfD. A total UF of 300 was applied to this effect level: 10 for
October, 2007
83
DRAFT - DO NOT CITE OR QUOTE

-------
uncertainty associated with interspecies differences (UFA: animal to human), 10 for
consideration of intraspecies variation (UFH: human variability), and 3 for database deficiencies
(UFd: database deficiency). The rationale for application of these UFs is provided described
below.
A 10-fold UF was used to account for uncertainty in extrapolating from laboratory
animals to humans (i.e., interspecies variability). No information was available to quantitatively
assess toxicokinetic or toxicodynamic differences between animals and humans.
A 10-fold UF was used to account for variation in susceptibility among members of the
human population (i.e., interindividual variability). Insufficient information is available to
predict potential variability in human susceptibility.
An UF was not needed to account for subchronic-to-chronic extrapolation because a
chronic study was used to derive the chronic RfD.
An UF for LOAEL-to-NOAEL extrapolation was not used because the current approach
is to address this factor as one of the considerations in selecting a BMR for benchmark dose
modeling. In this case, a BMR of a 10% change in absolute liver weight was selected under an
assumption that it represents a minimal biologically significant change.
The database of chronic and subchronic animal studies includes a 2-year gavage study in
F-344 rats and B6C3F1 mice (NTP, 1993; Irwin et al., 1995), a 90-day gavage study in Sprague-
Dawley rats (Merrick et al., 1991), a 90-day drinking water study in Sprague-Dawley rats
(Villeneuve et al., 1985), a 17-week gavage study in F-344 rats (Hazleton Laboratories, 1983a;
NTP, 1993), a 17-week gavage study in B6C3F1 mice (Hazleton Laboratories, 1983b; NTP
1993), and a two-generation reproductive/fertility assessment in Swiss CD-I mice (NTP, 1990).
A 3-fold UF for database deficiencies was applied because the database lacks information on
developmental toxicity associated with 1,2,3-trichloropropane. In addition, the two-generation
reproductive toxicity study indicates that the developing fetus may be a target of toxicity. The
lack of a reproductive toxicity study beyond two generations and a developmental toxicity study
is of particular concern due to the genotoxicity of 1,2,3-trichloropropane, because genetic
damage to the germ cells of the F1 generation may not be detected until the F2 generation.
The chronic RfD for 1,2,3-trichloropropane was calculated as follows:
October, 2007
RfD = BMDLadj - UF
= 1.1 mg/kg-day -^300
84 DRAFT - DO NOT CITE OR QUOTE

-------
= 4 x 10"3 mg/kg-day (rounded to one significant figure)
5.1.4. Chronic RfD Comparison Information
Figure 5-1 is an exposure-response array which presents NOEALs, LOAELs, and the
dose range tested corresponding to selected health effects, some were considered candidates for
chronic RfD derivation, from subchronic, chronic, and reproductive toxicity studies. The health
effects from the subchronic NTP study include decreased synthesis of pseudocholinesterase and
hepatic necrosis. The health effects from the chronic NTP study include increase absolute and
relative liver and kidney weights, and the effects from the NTP reproductive toxicity study
include a decrease in the number of pregnancies per fertile pair and a decrease in the number of
live pups per litter.
Figures 5-2 presents the point of departure, applied uncertainty factors, and derived
chronic RfD for additional effect endpoints that were modeled using EPA BMD software version
1.4.1. and which appear in Table 5-1. This comparison is intended to provide information on
additional health effects associated with 1,2,3-trichloropropane exposure.
Points of departures (PODs) and chronic reference doses (RfDs) that could be derived
from the additional health effects identified in Table 5-1 are presented in Figure 5-1 to allow a
comparison with the critical effect. For increased relative liver weight, increased absolute and
relative kidney weight, decreased fertility generating the 4th and 5th litters, and decreased live
pups/litter, the uncertainty factors applied were a 10-fold UF to account for uncertainty in
extrapolating from laboratory animals to humans, a 10-fold UF to account for variation in
susceptibility among members of the human population, and a 3-fold UF for database
deficiencies.
The change in liver weight is the first effect evident, and the increases in liver enzymes
associated with liver damage and increased rate of hepatic necrosis, as well as the hepatocellular
damage-related decrease in pseudocholinesterase, supports the liver as the critical target organ.
The dose-response relationships for oral exposure to 1,2,3-trichloropropane and impaired fertility
in CD-I mice are also suitable for deriving a chronic RfD, but are associated with higher
BMDLs that could be protected by the selected critical effect and corresponding BMDL.
Consideration of the available dose-response data to determine an estimate of oral exposure that
is likely to be without an appreciable risk of adverse health effects over a lifetime has led to the
selection of the two-year oral gavage study in Fischer rats (NTP, 1993) and increased liver
October, 2007
85
DRAFT - DO NOT CITE OR QUOTE

-------
weight in males as the principal study and critical effect for deriving the chronic RfD for 1,2,3-
trichloropropane.
October, 2007
86
DRAFT - DO NOT CITE OR QUOTE

-------
o-
to
o
o
-J
mg/kg-day
s
K)
O
O
O)
O
00
o
o
o
N)
O
O
ej
o
o
H
O
HH
H
W
O
V
o
c
o
H
W
decreased synthesis of
pseudocholinesterase;
male rats; (NTP, 1993)
decreased synthesis of
pseudocholinesterase;
female rats; (NTP,
1993)
hepatic necrosis; male
rats; (NTP, 1993)
hepatic necrosis;
female rats; (NTP,
1993)
increase absolute liver
weight; male rats;
(NTP, 1993)
increase relative liver
weight; male rats;
(NTP, 1993)
increase absolute
kidney weight; male
rats; (NTP, 1993)
increase relative
kidney weight; male
rats; (NTP, 1993)
decrease in
pregnancies per fertile
pair at 4th and 5th
breedings; (NTP,
	1990)	
reduction in live pups
per litter in 1st through
4th breedings; (NTP,
1990)
reduction in live pups
per litter in 5th
breeding; (NTP, 1990)

-------
Figure 5-2. Points of Departure for endpoints from Table 5-2 with corresponding applied uncertainty factors and derived chronic
RfD.
i
i

V
**
w
Increased
Increased
Increased
relative
absolute
absolute liver
liver
kidney
weight, male ~~
— weight, —
— weight,
rats (NTP,
male rats
female
1993)*
(NTP,
rats (NTP,

1993)
1993)
Increased
relative
kidney
weight,
male rats
(NTP,
1993)
Fertility
generating
the 4th
litter, mice
(NTP,
1990)
t
Fertility
generating
5th litter,
mice
(NTP,
1990)
Live
pups/litter
- 4th litter,
mice
(NTP,
1990)

Live
pups/litter
- 5th litter,
mice
(NTP,
1990)
~	Point of Departure
•	RfD
*	Critical effect and quantified RfD
UF, animal-to-human
UF, human variability
UF, database
October, 2007
DRAFT - WO NOT CITE OR QUOTE

-------
5.1.5. Previous Oral Assessment
The previous IRIS assessment for 1,2,3-trichloropropane was entered on the database on
03/31/1987 and contains an oral chronic RfD of 6 x 10"3 mg/kg-day. The chronic RfD was based
on a duration-adjusted NOAEL of 5.71 mg/kg-day for alterations in clinical chemistry and
reduced red blood cell mass in female F-344 rats following a 17-week oral gavage exposure
(Hazleton Laboratories, 1983a; NTP, 1983). A total UF of 1000 was used to account for
interspecies extrapolation, human variability, and extrapolation from a subchronic study. This
assessment was last updated in 1990 before the publication of the NTP (1993) Technical Report
used for this assessment.
5.2. CHRONIC INHALATION REFERENCE CONCENTRATION (RfC)
5.2.1. Choice of Principal Study and Critical Effect - with Rationale and Justification
Inhalation studies of 1,2,3-trichloropropane in humans are limited. A single report
(Silverman et al., 1946) on the effects in humans found that all subjects (12/sex) experienced
irritation (eyes, throat and odor) by 15 minute exposures to 100 ppm trichloropropane (isomer
and purity not reported). The database of inhalation toxicity studies in animals includes two 2-
week studies submitted to EPA by Miller et al. (1987a,b), a 4-week range finding study, two 13-
week studies, and two single-generation reproductive assessments (Johannsen et al., 1988;
Biodynamics, Inc., 1979).
Increased organ weights and histopathological lesions in rodents have been associated
with subchronic inhalation exposure to 1,2,3-trichloropropane. Concentration-dependent
increases in absolute and relative liver weight were observed in males and female rats
(Johannsen et al., 1988; Miller at al. 1987a; Biodynamics, Inc., 1979). An increase in relative
lung weight was also observed in female rats (Biodynamics, Inc., 1979). The histology data
demonstrate that 1,2,3-trichloropropane is both a local irritant affecting the nasal epithelium
(Miller at al. 1987a, b) and a systemic toxicant producing effects remote from the site of entry,
including peribronchial lymphoid hyperplasia, hepatocellular hypertrophy, and extramedullar
hematopoiesis (Johannsen et al., 1988; Biodynamics, Inc., 1979). No significant effects were
observed in the reproductive toxicity studies (Johannsen et al., 1988; Biodynamics, Inc., 1979).
The critical effect selected for the derivation of the chronic RfC is the development of
peribronchial lymphoid hyperplasia in the lungs of male CD rats, with a NOAEL of 1.5 ppm and
a LOAEL of 5 ppm 1,2,3-trichloropropane, due to the occurrence of this adverse effect in both
October, 2007 89 DRAFT - DO NOT CITE OR QUOTE

-------
male and female rats and the possible correlation between the hyperplasia and the observed
increased lung weight. The increase in lung weight had a NOAEL of 5 ppm and a LOAEL of 15
ppm. Although an increase in liver and kidney weights was apparent, lesions and serum enzyme
levels indicative of liver and kidney damage were not evident. The hepatocellular hypertrophy
evident in male rats was considered potentially adaptive in the absence of additional overt
toxicity in the liver, and the hematopoiesis of the spleen in female rats was not considered
adverse, as there was no change in the clinical chemistry and hematology parameters.
The NOAEL of 1.5 ppm for peribronchial lymphoid hyperplasia is not necessarily a 0%
response level. Rather, the NOAEL of 1.5 ppm in 15 male rats has a 95% confidence limit for
0% response of 0 to 22%; or, in other words, there is 95% confidence that the "true" response of
peribronchial lymphoid hyperplasia at 1.5 ppm is no higher than 22%. Peribronchial lymphoid
hyperplasia, also defined as lymphoid hyperplasia of the bronchus-associated lymphoid tissue, is
histologically characterized by the presence of hyperplastic lymphoid follicles with reactive
germinal centers distributed along the bronchioles and bronchi (Howling et al., 1999; Myers and
Kurtin, 1995; Fortoul et al., 1985; Yousem et al, 1985).
5.2.2. Methods of Analysis - Including Models
A NOAEL/LOAEL approach was used to derive the chronic RfC for 1,2,3-
trichloropropane. The chronic RfC was based on the NOAEL of 1.5 ppm 1,2,3-trichloropropane
for peribronchial lymphoid hyperplasia in the lungs of male rats identified in Johannsen et al.
(1988). Benchmark dose modeling was not utilized because the peribronchial lymphoid
hyperplasia incidences were not amenable to modeling due to the inconsistent dose response at
the three highest doses in both males and females, with model outputs that did not adequately fit
the data.
Human equivalent concentrations (HECs) were calculated from the candidate point of
departure. PODs were converted to mg/m3, adjusted to continuous exposure (7 days a week, 24
hours a day), and multiplied by a dosimetric adjustment factor (DAF) to calculate the HEC. A
DAF is a ratio of animal and human physiologic parameters. The specific DAF used depends on
the nature of the contaminant (particle or gas) and the target site (e.g., respiratory tract or remote
to the portal-of-entry).
The RfC methodology (U.S. EPA, 1994) classifies gases into three categories based on
their water solubility and reactivity with respiratory tract tissue. 1,2,3-Trichloropropane is
considered a category 2 gas because it is relatively insoluble in water and demonstrates systemic
October, 2007
90
DRAFT - DO NOT CITE OR QUOTE

-------
toxicity. 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, 1994). The DAF
for a category 1 gas is based on the animal-to-human ratio of the minute volume (Ve) divided by
the surface area (SA) of the region of the respiratory tract where the effect occurs. The DAF for
a category 3 gas is based on the ratio of the animal blood:gas partition coefficient (Hb/g-animai) and
the human blood:gas partition coefficient (Hb/g-human)-
The critical effect for the chronic RfC is considered a systemic effect because the critical
effect is located beyond the lung tissue in the bronchus-associated lymphoid tissue. The HEC
for increased peribronchial lymphoid hyperplasia in rats exposed to 1,2,3-trichloropropane
(category 3) 6 hours/day, 5 days/week for 13 weeks was calculated from aNOAELof 1.5 ppm
(1.5 ppm x MW[147.43] / 24.45 = 9.04 mg/m3). Adjustment to a continuous exposure was
calculated as follows:
NOAELadj = NOAEL x (6 hours)/(24 hours) x (5 days)/(7 days)
= 9.0 mg/m3 x 0.25 x 0.71
= 1.6 mg/m3
The DAF for an extra-respiratory effect of a gas is the ratio of the animal/human blood:
air partition coefficients [(Hb/g)A/(Hb/g)H]- However, the human and rat blood partition
coefficients for 1,2,3-trichloropropane are not known. In accordance with the RfC Methodology
(U.S. EPA, 1994) when the partition coefficients are unknown a ratio of 1 is used. This allows a
NOAELhec to be derived as follows:
NOAELrec = NOAELadj (mg/m3) x (Hb/g)A/(Hb/g)H
= NOAELadj (mg/m3) x 1
= 1.6 mg/m3
Application of the inhalation dosimetry methods to peribronchial lymphoid hyperplasia
in the lung resulted in a NOAELrec of 1.6 mg/m3.
5.2.3. Chronic RfC Derivation - Including Application of Uncertainty Factors (UFs)
The NOAELrec value of 1.6 mg/m3 for increased incidence of peribronchial lymphoid
hyperplasia in the lungs of male CD rats (Johannsen et al., 1988) was used as the point of
departure to derive the chronic RfC for 1,2,3-trichloropropane. A total UF of 3000 was applied
to this point of departure: 3 for extrapolation from rats to humans (UFA: animal to human), 10 for
October, 2007
91
DRAFT - DO NOT CITE OR QUOTE

-------
consideration of intraspecies variation (UFH: human variability), 10 for extrapolation from a
subchronic study (UFS), and 10 for database deficiencies. The rationale for application of the
UFs is described below. A graphical representation of the point of departure, applied uncertainty
factors, and quantified chronic RfC for the critical effect selected, an increased incidence of
peribronchial lymphoid hyperplasia in the lungs of male rats, was not included in this
assessment.
A factor of 3 was selected to account for uncertainties in extrapolating from rats to
humans. This value is adopted by convention where an adjustment from an animal specific
NOAELadj to a NOAELrec has been incorporated. Application of a full uncertainty factor of 10
would depend on two areas of uncertainty (i.e., toxicokinetic and toxicodynamic uncertainties).
In this assessment, the toxicokinetic component is mostly addressed by the determination of a
human equivalent concentration as described in the RfC methodology (U.S. EPA, 1994b). The
toxicodynamic uncertainty is also accounted for to a certain degree by the use of the applied
dosimetry method and an UF of 3 is retained to fully address this component.
A 10-fold UF was used to account for variation in susceptibility among members of the
human population (i.e., interindividual variability). Insufficient information is available to
predict potential variability in susceptibility among the population.
A 10-fold UF was used to account for uncertainty in extrapolating from a subchronic to
chronic exposure duration. The critical effect, peribronchial lymphoid hyperplasia, may be more
severe at lower doses with a prolonged exposure, and additional critical effects not observed
following subchronic exposure may arise following chronic exposure.
A 10-fold UF was used to account for deficiencies in the database. The database of
1,2,3-trichloropropane inhalation studies, which includes two 2-week studies submitted to EPA
by Miller et al. (1987a,b), a 4-week range finding study, two 13-week studies, and a single-
generation reproductive toxicity study (Johannsen et al., 1988; Biodynamics, Inc., 1979),
provides reliable dose-response data from subchronic studies of two species and a single-
generation reproductive toxicity study. However, the database is lacking a multigenerational
reproductive study and a developmental toxicity study. The lack of the multigenerational study
and a developmental toxicity study is of particular concern due to the genotoxicity of 1,2,3-
trichloropropane, because genetic damage to the germ cells of the F1 generation may not be
detected until the F2 generation.
October, 2007
92
DRAFT - DO NOT CITE OR QUOTE

-------
An UF for LOAEL-to-NOAEL extrapolation was not used since a NOAEL was used to
derive the chronic RfC.
The chronic RfC for 1,2,3-trichloropropane was calculated as follows:
RfC=NOAEL(HEC)-UF
= 1.6 mg/m3 3000
= 5 x 10"4 mg/m3 (rounded to one significant figure)
5.2.4.	Chronic RfC Comparison Information
Similar to the oral toxicity database, the inhalation studies found statistically significant
increases in organ weights. A dose-dependent increase absolute and relative liver weights were
observed in male rats and female rats following subchronic exposure and in male and female
mice following a two-week exposure to 1,2,3-trichloropropane. Additionally, an increase in
relative lung weights was observed in female rats and an increase in relative kidney weights was
observed in male rats following subchronic exposure to 1,2,3-trichloropropane. An increased
incidence of peribronchial lymphoid hyperplasia was observed in male and female rats exposed
to 5, 15, or 50 ppm 1,2,3-trichloropropane, but the study investigators did not examine epithelial
tissue in their investigation. Centrilobular to midzonal hepatocellular hypertrophy was seen in
nearly all male rats that were exposed for 13 weeks to concentrations of 5, 15, or 50 ppm 1,2,3-
trichloropropane. However, no evidence of hepatic effects was found in female rats that were
exposed to 50 ppm 1,2,3-trichloropropane. Conversely, a dose-dependent increase in the
incidence and severity of extramedullary hematopoiesis of the spleen was observed in female but
not male rats, although this effect is not biologically relevant.
The critical effect selected for the derivation of the chronic RfC is the development of
peribronchial lymphoid hyperplasia in the lungs of male CD rats due to the occurrence of this
adverse effect in both male and female rats and the possible correlation between the hyperplasia
and the observed increased lung weight. Although an increase in liver and kidney weights was
apparent, lesions and serum enzyme levels indicative of liver and kidney damage were not
evident. The hepatocellular hypertrophy evident in male rats was considered potentially
adaptive in the absence of additional overt toxicity in the liver, and the hematopoiesis of the
spleen in female rats was not considered adverse, as there was no change in the clinical
chemistry and hematology parameters.
5.2.5.	Previous Inhalation Assessment
October, 2007
93
DRAFT - DO NOT CITE OR QUOTE

-------
A reference concentration is not available from the current IRIS assessment, which was
first on-line on March 31, 1987 (U.S. EPA, 2007).
5.3. UNCERTAINTIES IN CHRONIC ORAL REFERENCE DOSE (RfD) AND
INHALATION REFERENCE CONCENTRATION (RfC)
Risk assessments need to portray associated uncertainty. The following discussion
identifies uncertainties associated with the chronic RfD and chronic RfC for 1,2,3-
trichloropropane. As presented earlier in this chapter (5.1.2 and 5.1.3; 5.2.2 and 5.2.3), the
uncertainty factor approach, following EPA practices and RfC and RfD guidance (U.S. EPA,
1993, 1994), was applied to a point of departure (POD), a BMDLHec for the RfD and a
NOAELrec for the chronic RfC. Factors accounting for uncertainties associated with a number
of steps in the analyses were adopted to account for extrapolating from an animal bioassay to
human exposure, a diverse population of varying susceptibilities, and to account for database
deficiencies. These extrapolations are carried out with default approaches given the paucity of
experimental 1,2,3-trichloropropane data to inform individual steps.
An adequate range of animal toxicology data are available for the hazard assessment of
1,2,3-trichloropropane, as described throughout the previous section (Chapter 4). The database
of oral toxicity studies includes a chronic gavage study in rats and mice, multiple subchronic
gavage and drinking water studies conducted in rats and mice, and a two-generation
reproductive/fertility assessment in mice. Toxicity associated with oral exposure to 1,2,3-
trichloropropane is observed in the liver, kidney and reproductive endpoints, including decreased
fertility generating the 4th and 5th litters and decreased number of live pups/litter in the 4th and 5th
litters. The database of inhalation toxicity studies in animals includes two 2-week studies
submitted to EPA, a 4-week range finding study, two 13-week studies, and two single-generation
reproductive assessments. The inhalation database, however, is lacking a chronic exposure
study. Toxicity associated with inhalation exposure to 1,2,3-trichloropropane is observed in the
respiratory system, as an increased incidence of peribronchial lymphoid hyperplasia. In addition
to the oral and inhalation data are numerous absorption, distribution, metabolism, and excretion
references and genotoxicty studies. Critical data gaps have been identified and uncertainties
associated with data deficiencies are more fully discussed below.
Consideration of the available dose-response data to determine an estimate of oral
exposure that is likely to be without an appreciable risk of adverse health effects over a lifetime
October, 2007 94 DRAFT - DO NOT CITE OR QUOTE

-------
has led to the selection of the two-year oral gavage study in Fischer rats (NTP, 1993) and
increased liver weight in males as the principal study and critical effect for deriving the chronic
RfD for 1,2,3-trichloropropane. The dose-response relationships for oral exposure to 1,2,3-
trichloropropane and impaired fertility in CD-I mice are also suitable for deriving a chronic
RfD, but are associated with higher BMDLs that would be protected by the selected critical
effect and corresponding BMDL.
The critical effect selected for the derivation of the chronic RfC is the development of
peribronchial lymphoid hyperplasia in the lungs of male CD rats, due to the occurrence of this
adverse effect in both male and female rats and the possible correlation between the hyperplasia
and the observed increased lung weight. Although an increase in liver and kidney weights was
apparent, lesions and serum enzyme levels indicative of liver and kidney damage were not
evident. The hepatocellular hypertrophy evident in male rats was considered potentially
adaptive in the absence of additional overt toxicity in the liver, and the hematopoiesis of the
spleen in female rats was not considered adverse, as there was no change in the clinical
chemistry and hematology parameters.
The selection of the benchmark dose model for the quantitation of the chronic RfD does
not lead to significant uncertainty in estimating the POD since benchmark effect levels were
within the range of experimental data. However, the selected model, the Hill model, does not
represent all possible models one might fit, and other models could be selected to yield more
extreme results, both higher and lower than those included in this assessment.
The derived chronic RfC was quantified using a NOAEL for the point of departure. A
POD based on a NOAEL or LOAEL is, in part, a reflection of the particular exposure
concentration or dose at which a study was conducted. It lacks characterization of the dose-
response curve and for this reason is less informative than a POD obtained from benchmark
dose-response modeling. In addition, the NOAEL of 1.5 ppm for peribronchial lymphoid
hyperplasia should not be assumed to be a 0% response level. The NOAEL of 1.5 ppm for
peribronchial lymphoid hyperplasia is consistent with a 95% confidence limit for 0% response of
0 to 22% (assuming a binomial distribution and using tabled confidence limit values). In other
words, there is 95% confidence that the response rate is no higher than 22%, and in which case
the derived RfC would overestimate the inhalation exposure likely to be without an appreciable
risk of adverse health effects over a lifetime.
Extrapolating from animals to humans embodies further issues and uncertainties. The
effect and the magnitude associated with the concentration at the point of departure in rodents
October, 2007
95
DRAFT - DO NOT CITE OR QUOTE

-------
are extrapolated to human response. Pharmacokinetic models are useful to examine species
differences in pharmacokinetic processing, however, dosimetric adjustment using
pharmacokinetic modeling was not possible for the toxicity observed following oral and
inhalation exposure to 1,2,3-trichloropropane. Information was unavailable to quantitatively
assess toxicokinetic or toxicodynamic differences between animals and humans, so the 10-fold
UF was used to account for uncertainty in extrapolating from laboratory animals to humans in
the derivation of the chronic RfD. For the chronic RfC, a factor of 3 was adopted by convention
where an adjustment from an animal specific NOAELadj to a NOAELrec has been incorporated.
Application of a full uncertainty factor of 10 would depend on two areas of uncertainty (i.e.,
toxicokinetic and toxicodynamic uncertainties). In this assessment, the toxicokinetic component
is mostly addressed by the determination of a human equivalent concentration as described in the
RfC methodology (U.S. EPA, 1994b). The toxicodynamic uncertainty is also accounted for to a
certain degree by the use of the applied dosimetry method and an UF of 3 is retained to account
for this component.
Heterogeneity among humans is another uncertainty associated with extrapolating doses
from animals to humans. Uncertainty related to human variation needs consideration, also, in
extrapolating dose from a subset or smaller sized population, say of one sex or a narrow range of
life stages typical of occupational epidemiologic studies, to a larger, more diverse population. In
the absence of 1,2,3-trichloropropane-specific data on human variation, a factor of 10 was used
to account for uncertainty associated with human variation in the derivation of both the chronic
RfD and the chronic RfC. Human variation may be larger or smaller; however, 1,2,3-
trichloropropane-specific data to examine the potential magnitude of over- or under-estimation is
unavailable.
Data gaps have been identified with uncertainties associated with database deficiencies
on developmental toxicity associated with 1,2,3-trichloropropane oral exposure. The two-
generation reproductive assessment toxicity study indicates that the developing fetus may be a
target of toxicity. In addition, the lack of the multigenerational study, beyond two generations, is
of particular concern due to the genotoxicity of 1,2,3-trichloropropane, because genetic damage
to the germ cells of the F1 generation may not be detected until the F2 generation. Thus, the
absence of a study specifically evaluating developmental toxicity represents an area of
uncertainty or gap in the database. The database of inhalation studies is of particular concern
due to the lack of a multigenerational reproductive study and a developmental toxicity study.
The lack of the multigenerational study is of particular concern due to the issue described
previously in this paragraph.
October, 2007
96
DRAFT - DO NOT CITE OR QUOTE

-------
5.4. CANCER ASSESSMENT
There are no available studies on cancer in humans associated with exposure to 1,2,3-
trichloropropane. NTP (1993) provided evidence of 1,2,3-trichloropropane-induced forestomach
and other benign and malignant tumors in male and female Fischer 344 rats and male and female
B6C3Fi mice in a 2-year gavage cancer bioassay. 1,2,3-Trichloropropane has been reported to
be a mutagen in S. typhimurium assays (Lag et al., 1994; NTP, 1993; Ratpan and Plaumann,
1988; Haworth et al., 1983; Kier, 1982; Stolzenberg and Hine, 1980; Shell Oil Co., 1979).
Studies have also demonstrated the induction of chromosomal aberrations and sister chromatid
exchanges in Chinese Hamster ovary cell assays (NTP, 1993; Douglas et al., 1985),
trifluorothymidine resistance induction in mouse lymphoma assays (NTP, 1993; Shell Oil Co.,
1982), DNA strand breakage measured by the Comet assay (single gel electrophoresis test) in
isolated human lymphocytes (Tafazoli and Kirsch-Volders, 1996), enhanced DNA viral
transformation in Syrian hamster embryo cells (Hatch et al., 1983), and the induction of
micronucleus formation in the mammalian cell lines, AHH-1, MCL-5, and h2El (Doherty et al.,
1996) and CHO cells (Douglas et al., 1985).
Under the Guidelines for Carcinogenic Risk Assessment (U.S. EPA, 2005a), 1,2,3-
trichloropropane is "likely to be carcinogenic to humans", based on a statistically significant and
dose-related increase in the formation of multiple tumors in both sexes of two species from an
NTP (1993) chronic oral bioassay. Statistically significant increases in incidences of tumors of
the oral cavity, forestomach, pancreas, kidney, preputial gland, clitoral gland, mammary gland,
and Zymbal's gland in rats, and the oral cavity, forestomach, liver, and Harderian gland in mice,
were reported.
5.4.1. Choice of Study/Data with Rationale and Justification
The study by NTP (1993) was used for development of an oral slope factor. This was a
well-designed study, conducted in both sexes in two species with an adequate number of animals
per dose group. The number of test animals allocated among three dose levels and an untreated
control group was adequate, with examination of appropriate toxicological endpoints in both
sexes of rats and mice. Tumor incidences were elevated with increasing exposure level at
numerous sites across all sex/species combinations, involving point of contact in the alimentary
system and more distant locations. Due to the increased carcinogenic response at all dose levels
and the increased mortality in the two high dose groups in both rats and mice, NTP stated that
October, 2007
97
DRAFT - DO NOT CITE OR QUOTE

-------
carcinogenic activity might have been detected at doses lower than those tested in the chronic
study (NTP, 1993).
5.4.2. Dose-Response Data
In the NTP (1993) study, groups of 60 male and female F344 rats and B6C3Fi mice were
administered 3, 10 or 30 and 6, 20, or 60 mg/kg-day 1,2,3-trichloropropane, respectively, by
gavage, 5 days/week, for two years. Ten male and 10 female rats and mice from each dose
group were designated for evaluation at 15 months. High mortality in both species in all high-
dose groups necessitated early termination of the rat high-dose groups at weeks 77 (males) and
67 (females). All other groups of rats were sacrificed after two years (104 weeks). For the mice,
mid-dose groups were sacrificed at week 89, and high-dose male and female mice were
sacrificed at weeks 79 and 73, respectively. All other groups of mice were sacrificed after week
104.
Dose-related, statistically significant increasing trends in tumors were noted at the
following sites:
squamous cell carcinomas or papillomas of the alimentary system in male and female rats
and mice;
Zymbal's gland carcinomas in male and female rats;
pancreatic acinar cell adenomas or adenocarcinomas, preputial gland adenomas or
carcinomas, and kidney tubular cell adenomas in male rats;
clitoral gland adenomas or carcinomas, and mammary gland adenocarcinomas in female
rats;
hepatocellular adenomas or carcinomas, and harderian gland adenomas in male and
female mice; and
uterine/ cervical adenomas or adenocarcinomas in female mice.
These tumors generally appeared earlier with increasing exposure levels, and showed
statistically significantly increasing trends with increasing exposure level (by life table test or
logistic regression, p<0.001). These data are summarized in Tables 5-2 (male rats), 5-3 (female
rats), 5-4 (male mice), 5-5 (female mice). Data are not available to indicate whether the
malignant tumors developed specifically from progression of the benign tumors. However, as a
default approach etiologically similar tumor types, i.e., benign and malignant tumors of the same
cell type, were combined for these tabulations because of the possibility that the benign tumors
could progress to the malignant form as outlined in the 2005 Cancer Guidelines (U.S. EPA,
2005a).
October, 2007
98
DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-2. Tumor incidence, (percent), and time of first occurrence in male
rats following oral gavage exposure to 1,2,3-trichloropropane (NTP, 1993)
Site
0 mg/kg-
3 mg/kg-
10 mg/kg-
30 mg/kg-
Trend test
day
day
day
day
p-value
Alimentary
l/59a
39/60
48/57
58/60
<0.001
system, total
(2%)
(65%)
(84%)
(97%)

squamous
104°
64
58
47

neoplasms'3





Pancreas: acinar
5/59
20/60d
36/57
31/58
<0.001
cell, adenoma or
(8%)
(33%)
(63%)
(53%)

adenocarcinoma
104
98
67
60

Kidney tubular
0/59
2/60
18/57d
26/58
<0.001
cell: adenoma
(0%)
(3%)
(35%)
(45%)


-
104
94
60

Preputial gland:
5/58
6/57
9/57
17/56
<0.001
adenoma or
(8%)
(11%)
(16%)
(30%)

carcinoma
72
93
58
55

Zymbal's gland,
0/59
0/60
0/57
3/58
<0.001
carcinoma
(0%)
(0%)
(0%)
(5%)


-
-
-
56

a Numbers of animals at risk (denominators) vary due to missing tissues or due to deaths
occurring before the first incidence of tumor in that group, or before Week 52, whichever was
earlier.
b Squamous papillomas or squamous cell carcinomas of the pharynx/palate, tongue, or
forestomach.
0 Week first observed.
d NTP (1993) summary tables reported slightly higher incidences - 21 low-dose males with
pancreatic acinar cell tumors, 20 mid-dose males with kidney tubule adenomas - than noted in the
individual animal histopathology tables. Table 5-6 reflects the incidence in the individual animal
histopathology tables.
October, 2007
99
DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-3. Tumor incidence, (percent), and time of first occurrence in female
rats following oral gavage exposure to 1,2,3-trichloropropane (NTP, 1993)
Site
0 mg/kg-
3 mg/kg-
10 mg/kg-
30 mg/kg-
Trend test
day
day
day
day
p-value
Alimentary
l/60a
22/59a
49/59a
44/5 8a
<0.001
system, total
(2%)
(37%)
(83%)
(76%)

squamous
104c
73
58
33

neoplasms'3





Clitoral gland,
5/56
11/56
18/57
17/51
<0.001
adenoma or
(9%)
(20%)
(32%)
(33%)

carcinoma
102
66
62
44

Mammary gland,
2/57
6/57
14/52
23/48
<0.001
adenoma or
(4%)
(10%)
(27%)
(48%)

adenocarcinoma
64
67
61
34

Zymbal's gland,
0/60
1/59
0/59
4/45
<0.001
carcinoma
(0%)
(2%)
(0%)
(9%)


-
102
-
48

a Numbers of animals at risk (denominators) vary due to missing tissues, or due to deaths
occurring before the first incidence of tumor in that group, or before Week 52, whichever was
earlier.
b Squamous papillomas or squamous cell carcinomas of the pharynx/palate, tongue, or
forestomach.
0 Week first observed.
Table 5-4. Tumor incidence in male mice following oral gavage exposure to
1,2,3-trichloropropane (NTP, 1993)
Site
0 mg/kg-
6 mg/kg-
20 mg/kg-
60 mg/kg-
Trend test
day
day
day
day
p-value
Alimentary
3/59a
57/59a
57/60a
59/60a
<0.001
system, total
(5%)
(97%)
(95%)
(98%)

squamous
69°
61
55
46

neoplasms'3





Liver: adenoma
14/59
24/59
25/60
33/60
<0.001
or carcinoma
(23%)
(41%)
(42%)
(55%)


65
74
59
46

Harderian gland
1/59
2/59
10/60
11/60
0.001
adenoma
(2%)
(3%)
(17%)
(20%)


104
91
72
65

" Numbers of animals at risk (denominators) vary due to missing tissues or due to deaths occurring
before the first incidence of tumor in that group, or before Week 52, whichever was earlier.
b Squamous papillomas or squamous cell carcinomas of the pharynx/palate, tongue, or
forestomach.
0 Week first observed.
October, 2007
100
DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-5. Tumor incidence in female mice following oral gavage exposure to
1,2,3-trichloropropane (NTP, 1993)
Site
0 mg/kg-
6 mg/kg-
20 mg/kg-
60 mg/kg-
Trend test
day
day
day
day
p-value
Alimentary
0/5 9a
54/60a
59/60a
59/60a
<0.001
system, total
(0%)
(90%)
(98%)
(98%)

squamous
c
59
45
42

neoplasms'3





Liver: adenoma or
8/59
11/60
9/60
36/58
<0.001
carcinoma
(13%)
(18%)
(15%)
(60%)


66
77
65
60

Harderian gland
3/59
6/59
7/60
10/60
0.04
adenoma
(5%)
(10%)
(12%)
(17%)


66
80
78
64

Uterus/cervix:
0/59
5/59
3/59
11/57
<0.001
adenoma or
(0%)
(8%)
(5%)
(19%)

adenocarcinoma
-
100
83
66

" Numbers of animals at risk (denominators) vary due to missing tissues or due to deaths occurring
before the first incidence of tumor in that group, or before Week 52, whichever was earlier.
b Squamous papillomas or squamous cell carcinomas of the pharynx/palate, tongue, or
forestomach.
0 Week first observed.
NTP noted additional tumor sites with apparent dose-related increases, squamous cell
papillomas and carcinomas and hepatocellular adenomas and carcinomas, in male rats. These
tumors displayed a dose-related increase, but their incidences were not individually statistically
significantly greater than controls. NTP concluded that because the incidence in no one group
was statistically significantly higher than control, the overall trends were not dose-related.
The male and female mice tumor incidence data, while clearly demonstrating
carcinogenicity, were not suitable for deriving low-dose quantitative risk estimates. The NTP
study design unfortunately missed nearly all of the relevant dose-response range for mice, with
both male and female mice having nearly 100% responses at the lowest exposure level. While
these responses were higher than those of the rats at the comparable exposure level, suggesting
greater sensitivity of the mice, there is no information concerning the dose-response
relationships at lower exposure levels that could be compared with the rat data. In other words,
the high dose behavior of 1,2,3-TCP in mice does not inform the mouse tumor response to lower
exposures of 1,2,3-TCP. Extrapolation from high response levels is not justified when other
more suitable data, here in rats, are available. Consequently, dose-response modeling was not
carried out with the mouse tumor data.
5.4.3. Dose Adjustments and Extrapolation Methods
October, 2007	101	DRAFT - DO NOT CITE OR QUOTE

-------
The EPA Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005) stipulate that the
method used to characterize and quantify cancer risk from a chemical is determined by what is
known about the mode of action of the carcinogen and the shape of the cancer dose-response
curve. The dose response is assumed to be linear in the low dose range, when evidence supports
a mutagenic mode of action because of DNA reactivity, or if another mode of action that is
anticipated to be linear is applicable. The linear approach is used as a default option if the mode-
of-action of carcinogenicity is not understood (U.S. EPA, 2005). In the case of 1,2,3-
trichloropropane, there are data available that suggest that bioactivated 1,2,3-trichloropropane
may bind directly to DNA resulting in a mutagenic event that may lead to tumorigenicity in
animals. However, the database contains limited in vivo evidence that mutagenic events occur
following 1,2,3-trichloropropane exposure. A linear-low-dose extrapolation approach was used
to estimate human carcinogenic risk associated with 1,2,3-trichloropropane exposure as the
default option. This approach is supported by the positive evidence of genotoxicity and a
potential mutagenic mode of action.
It is possible that the squamous neoplasms may result primarily as a portal of entry effect
and may not have a low-dose response pattern that extends linearly from the observed responses.
The tumors in other organs demonstrate absorption of 1,2,3-trichloropropane, although it is not
clear whether absorption was impacted by adverse effects in the forestomach.
Due to the occurrence of multiple tumor types, earlier occurrence with increasing
exposure, and early termination of at least one dose group, methods which can reflect the
influence of competing risks and intercurrent mortality on site-specific tumor incidence rates are
preferred. EPA has generally used the multistage-Weibull model, because it incorporates the
time at which death-with-tumor occurred. The multistage-Weibull model has the form:
P(d) = 1 - exp[-(q0 + qid + q-d2 + ... + qid) x (t ± t0f],
where P(d) represents the lifetime risk (probability) of cancer at dose d (i.e., human equivalent
exposure in this case); parameters qt > 0, for i = 0, 1, ..., k; t is the time at which the tumor was
observed; and z is a parameter estimated in fitting the model, which characterizes the change in
response with age. The parameter t0 represents the time between when a potentially fatal tumor
becomes observable and when it causes death, and is generally set to 0 because of a lack of data
to estimate the time reliably. The dose-response analyses were conducted using the computer
software program TOXRISK, version 5.3 (property of ICF, Fairfax, VA), which is based on
October, 2007
102
DRAFT - DO NOT CITE OR QUOTE

-------
Weibull models drawn from Krewski et al. (1983). Parameters were estimated using the method
of maximum likelihood.
Other characteristics of the observed tumor types were considered prior to modeling,
including allowance for different, although possibly unidentified, modes of action, and for
relative severity of tumor types. First, etiologically different tumor types were not combined
across sites prior to modeling, in order to allow for the possibility that different tumor types can
have different dose-response relationships because of varying time courses or other underlying
mechanisms or factors. Consequently, all of the tumor types listed separately in Tables 5-2 and
5-3 were modeled separately. A further consideration allowed by the software program is the
distinction between tumor types as being either fatal or incidental, in order to adjust 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. Although the NTP
(1993) stated that neoplasms of the forestomach and oral mucosa in rats and mammary tumors in
female rats were the principal cause of death of most animals dying or killed moribund before
the end of the study, it was not clear that a determination could be made for each animal with
multiple tumors. Therefore, all tumors were treated as 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,
2002). 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. Plots of model fits compared with Hoel-Walburg estimates of cumulative
incidence were also examined for goodness of fit in the lower exposure region of the observed
data (Gart et al., 1986). If a model with one more stage fitted the low-dose data better than the
most parsimonious model, then the model with one higher stage was selected.
Points of departure for estimating low-dose risk were identified at doses at the lower end
of the observed data, generally corresponding to 10% extra risk, defined as the extra risk over the
background tumor rate, [P(d) - P(0)]/[1 - P(0)]. The lifetime oral cancer slope factor for humans
is defined as the slope of the line from the lower 95% bound on the exposure at the point of
departure. This 95% upper confidence limit (UCL) represents a plausible upper bound on the
true risk.
Adjustments for approximating human equivalent slope factors applicable for continuous
exposure were also carried out by the dose-response software program. Consistent with the
October, 2007
103
DRAFT - DO NOT CITE OR QUOTE

-------
Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), an adjustment for cross-species
scaling was applied by the software program, to address toxicological equivalence across
species, after the model-fitting phase. Following EPA's cross-species scaling methodology, the
time-weighted daily average doses were converted to human equivalent doses on the basis of
(body weight)3 4 (U.S. EPA, 1992). It was not necessary to adjust the administered doses for
lifetime exposure prior to modeling for the groups terminated early, because the software
program used characterizes the tumor incidence as a function of time, from which it provides an
extrapolation to lifetime exposure. In addition, TOXRISK estimated continuous daily
exposure by multiplying each slope factor by (5 days)/(7 days) = 0.71.
5.4.4. Oral Slope Factor and Inhalation Unit Risk
The results of applying the multistage-Weibull models to the male and female rat tumor
incidence data are provided in Table 5-6. Note that while identifying a point of departure (POD)
near the lower end of the observed data for linear extrapolation to lower doses is consistent with
the Guidelines for Carcinogen Risk Assessment (U.S.EPA, 2005a), TOXRISK does not provide
the exposure levels corresponding to BMRs greater than 10%. Consequently, an oral slope
factor for each of the tumor sites was calculated by dividing the BMR level (usually 10%) by its
corresponding BMDL to obtain points of departure, for comparison across tumor sites and with
other chemicals. In the absence of any data on the carcinogenicity of 1,2,3-trichloropropane via
the inhalation route, no inhalation unit risk has been derived in this evaluation.
Human equivalent oral slope factors estimated from the tumor sites with statistically
significant increases ranged from 0.020 to 3.0 per mg/kg-day, a range of about two orders of
magnitude, with both extremes coming from the male rat data. The highest slope factor in rats
corresponded to squamous neoplasms of the alimentary system, and the lowest slope factor
corresponded to Zymbal's gland tumors. The slope factor corresponding to pancreatic acinar
cell tumors in male rats was close to the maximum, at 1.0 per mg/kg/day. The slope factors
corresponding to the female rat tumors fell between these two extremes, with squamous
neoplasms of the alimentary system, at 1.3 per mg/kg/day, the highest slope factor in that set.
The oral slope factor corresponding to alimentary system squamous neoplasms in male
F344 rats was the highest oral slope factor obtained. Given the multiplicity of tumor sites,
however, basing the oral slope factor on one tumor site may underestimate the carcinogenic
potential of 1,2,3-trichloropropane. An approach suggested in the cancer guidelines would be to
estimate cancer risk from tumor-bearing animals. EPA traditionally used this approach until the
NRC document Science and Judgment (1994) made a case that this approach would tend to
October, 2007
104
DRAFT - DO NOT CITE OR QUOTE

-------
underestimate overall risk when tumor types occur in a statistically independent manner. In
addition, application of one model to a composite data set does not accommodate biologically
relevant information that may vary across sites or may only be available for a subset of sites.
For instance, the time courses of the multiple tumor types evaluated varied, as is suggested by
the variation in estimates of z (see Table 5-6), from 1.0 (male rat Zymbal's gland tumors),
indicating a slight tendency toward earlier tumor occurrence with increasing exposure level, to
8.7 (male rat pancreatic tumors), indicating a clearly earlier response with increasing exposure
level. The result of fitting a model with underlying mechanism-related parameters, such as z in
the multistage-Weibull model, would be difficult to interpret with composite data. A simpler
model could be used for the composite data, such as the multistage model, but available
biological information would then be ignored.
Following the recommendations of the NRC (1994) and the 2005 Guidelines for
Carcinogen Risk Assessment (U.S. EPA, 2005), a statistically appropriate upper bound on total
risk was estimated in order to gain some understanding of the total risk from multiple tumor sites
in male F344 rats (Table 5-7). Although the inclusion of a central tendency estimate would
improve the transparency of this document, the calculation of a central tendency estimate by
TOXRISK was not feasible. Note that the upper bound estimate of overall risk describes the
risk of developing any combination of the tumor types considered, not just the risk of developing
all three simultaneously. The estimate involved the following steps:
I) It was assumed that the tumor types associated with 1,2,3-trichloropropane
exposure were statistically independent - that is, that the occurrence of a
pancreatic acinar cell tumor, say, was not dependent upon whether there was a
forestomach tumor. This assumption cannot currently be verified, and if not
correct could lead to an overestimate of risk from summing across tumor sites.
NRC (1994) argued that a general assumption of statistical independence of
tumor-type occurrences within animals was not likely to introduce substantial
error in assessing carcinogenic potency from rodent bioassay data.
II) The models previously fitted to estimate the BMDs and BMDLs were used to
extrapolate to a low level of risk (R), in order to reach the region of each
estimated dose-response function where the slope was reasonably constant and
upper bound estimation was still numerically stable. For these data 10"3 risk was
generally the lowest risk necessary. The oral slope factor for each site was then
estimated by R/BMDLr, as for the estimates for each tumor site above.
III) The maximum likelihood estimates (MLE) of unit potency (that is, risk per unit
of exposure) estimated by R/BMDr, were summed across the alimentary system,
October, 2007
105
DRAFT - DO NOT CITE OR QUOTE

-------
pancreas, preputial gland, kidney, and Zymbal's gland tumors for male F344 rats.
Similarly, for female rats the MLEs of unit potency were summed across
squamous cell neoplasms, mammary gland adenocarcinomas, clitoral gland
tumors and Zymbal's gland carcinomas.
IV) An estimate of the 95% upper bound on the summed oral slope factor was
calculated by assuming a normal distribution for the individual risk estimates, and
deriving the variance of the risk estimate for each tumor site from its 95% upper
confidence limit (UCL) according to the formula:
95% UCL = MLE + 1.645 x 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 of the
MLE.
The resulting combined upper bound slope factor for male rats was 3.8 per mg/kg/day,
compared with 3.0 per mg/kg/day for only alimentary system tumors and 0.020 per mg/kg-day
for only Zymbal's gland tumors. Overall, the consideration of the other tumor sites increased the
slope factor by about 20%. The increase was due largely to the pancreatic tumors, with very
little contribution from the other three tumor sites. A sensitivity analysis (not included in this
document) showed that the summed risk was essentially the same (to 2 significant digits)
whether or not the individual risks were estimated in the region of 10"3 risk or near the PODs.
For female rats the combined upper bound slope factor was 1.9 per mg/kg-day, a 50%
increase compared with 1.3 per mg/kg-day for alimentary system tumors only. Both the clitoral
and mammary gland tumors contributed to the increased risk estimate. As with the summed risk
for male rats, there was little difference when the individual risks were estimated in the region of
10"3 risk or near the PODs.
Based on the analyses discussed above, the recommended upper bound estimate on
human extra cancer risk from continuous lifetime oral exposure to 1,2,3-trichloropropane is 4
per mg/kg-day, rounding the summed risk for male rats above to one significant digit. The
value based on male rats was recommended because male rats are the most sensitive to tumor
induction following exposure to 1,2,3-trichloropropane. This slope factor should not be used
October, 2007
106
DRAFT - DO NOT CITE OR QUOTE

-------
with exposures greater than 0.05 mg/kg/day, the point of departure for the male rat alimentary
system tumors, because the observed dose-response relationships do not continue linearly above
this level and the fitted dose-response models better characterize what is known about the
carcinogenicity of 1,2,3-trichloropropane. The recommended estimate reflects the time-to-tumor
dimension of the responses as well as the exposure-response relationships for the multiple tumor
sites.
October, 2007
107
DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-6. Dose-response modeling summary for rat tumor sites associated with
oral exposure to 1,2,3-trichloropropane; tumor incidence data from Nr
rP (1993)
Tumor type
MLE
coefficients3
Human Equivalent
Continuous Point of
departureb, mg/kg-day
Slope factor0,
(mg/kg-day)
i


BMDio
BMDL10

Male rats
Alimentary system, total
squamous neoplasms
q0= 1.1 x 10"12
qi = 1.9 x 10"11
q2 = 2.1 x 10"12
z =5.1
0.050
0.033
3.0
Pancreatic acinar cell
adenoma or adenocarcinoma
q0 = 4.5 x 10"19
qi =2.4 x 10"19
q2= 1.2 x 10"19
z =8.7
0.20
0.10
1.0
Preputial gland adenoma or
carcinoma
q0= 1.1 x 10"4
qj =2.7 x 10"5
z = 1.4
1.3
0.59
0.17
Kidney tubular cell adenoma
q2 = 2.5 x 10"15
z = 6.2
0.49d
0.32d
0.16
Zymbal's gland carcinoma
qi = 1.6 x 10"5
z = 1.0
1.2e
0.49e
0.020
Female rats
Alimentary system, total
squamous neoplasms
q0 = 2.5 x 10"12
qj = 8.1 x 10"12
q2 = 5.6 x 10"12
z =4.9
0.17
0.075
1.3
Clitoral gland adenoma or
carcinoma
q0 = 3.1 x 10"7
qi = 6.5 x 10"7
z =2.4
0.32
0.24
0.42
Mammary gland
adenocarcinomaf
q0 = 0.034
qi = 0.027
0.64
0.43
0.23
Zymbal's gland carcinoma0
qi = 1.4 x 10"5
z = 1.2
0.40e
0.15e
0.067
" Model: multistage-Weibull, extra risk: P(d) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk) x (t ±to)z],
coefficients estimated in terms of mg/kg-day as administered inbioassay; lower stage q not listed were
estimated to be zero.
b Point of departure adjusted to estimate human equivalent continuous exposure, using BW3'4 cross-
species scaling and by multiplying by (5 days)/(7 days). BMDi0 = Concentration at 10% effect (extra
risk) level; BMDLi0= 95% lower bound on concentration at 10% effect (extra risk) level.
0 Slope factors estimated by dividing the BMR by the BMDL.
d BMD and BMDL correspond to BMR=5%.
e BMD and BMDL correspond to BMR=1%.
f
Multistage-Weibull model did not fit adequately (see Appendix B). Multistage model (extra risk without
adjustment for time) was used: P(d) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)], omitting the high-dose
group.
October, 2007
108
DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-7. Summary of cancer risk values estimated by R/BMDr and
summed across tumor sites for male and female rats.
Male Rat
BMD
(mg/kg-day)
BMDL
(mg/kg-day)
Cancer risk value at BMDa
(mg/kg-day)"1
Oral slope factor
(mg/kg-day)"1
oral route
squamous
papillomas,
carcinomas;
R=10"2
4.9 x 10"3
3.2 x 10"3
2.0
3.1
pancreas acinar
tumors
2.8 x 10"3
1.0 x 10"3
3.5 x 10"1
1.0
kidney tubule
adenomas
6.8 x 10"2
9.1 x 10"3
1.5 x 10"2
1.1 x 10"1
preputial gland
tumors
1.3 x 10"2
5.6 x 10"3
7.8 x 10"2
1.8 x 10"1
Zymbal's gland
carcinomas
1.2 x 10"1
4.8 x 10"2
8.5 x 10"3
2.1 x 10"2

Sum
2.5



Upper bound on summed risk
3.8
Female Rat
BMD
(mg/kg-
day)
BMDL
(mg/kg-day)
Cancer risk value at BMDa
(mg/kg-day)"1
Oral slope factor
(mg/kg-day)"1
oral route
squamous
papillomas,
carcinomas.
3.0 x 10"3
7.4 x 10"3
3.4 x 10"1
1.4
clitoral gland
adenomas,
carcinomas
R=10"2
3.0 x 10"1
2.3 x 10"1
3.3 x 10"2
4.4 x 10"1
mammary
adenocarcinomas
R=10"2
7.1 x 10"2
4.1 x 10"2
1.4 x 10"2
2.4 x 10"2
Zymbal's gland
carcinomas
R=10"2
4.0 x 10"1
1.5 x 10"1
2.7 x 10"2
4.2 x 10"2

Sum
0.8



Upper bound on summed risk
1.9
aThe MLE slope factor = R/BMDr, where R = 1 x 10"3 except where specified.
October, 2007	109	DRAFT - DO NOT CITE OR QUOTE

-------
5.4.5. Uncertainties in Cancer Risk Values
As in most risk assessments, extrapolation of study data to estimate potential risks to
human populations from exposure to 1,2,3-trichloropropane has engendered some uncertainty in
the results. Several types of uncertainty may be considered quantitatively, but other important
uncertainties cannot be considered quantitatively. Thus an overall integrated quantitative
uncertainty analysis is not presented. Section 5.4.5.1 and Table 5-8 summarize principal
uncertainties.
5.4.5.1. Sources of Uncertainty
Choice of low-dose extrapolation approach. The MOA is a key consideration in
clarifying how risks should be estimated for low-dose exposure. A linear-low-dose extrapolation
approach was used to estimate human carcinogenic risk associated with 1,2,3-trichloropropane
exposure as the default option. Linear extrapolation is, generally, considered to be a health-
protective approach, and, in some cases, may lead to an overestimation of risk, as stated in the
2005 Cancer Guidelines (U.S. EPA, 2005).
The extent to which the overall uncertainty in low-dose risk estimation could be reduced
if the MOA for 1,2,3-trichloropropane were known with a higher degree of confidence is of
interest, but additional supporting data on the MOA of 1,2,3-trichloropropane is not available.
Even if it were, incorporation of MOA into dose-response modeling might not be straightforward
and might not significantly reduce the uncertainty about low-dose extrapolation.
Due to the occurrence of multiple tumor types, earlier occurrence with increasing
exposure, and early termination of at least one dose group, methods which can reflect the
influence of competing risks and intercurrent mortality on site-specific tumor incidence rates are
preferred. EPA has generally used the multistage-Weibull model in this type of situation,
because it incorporates the time at which death-with-tumor occurred; however, it is unknown
how well this model or the linear low-dose extrapolation predicts low-dose risks for 1,2,3-
trichloropropane. The selected model does not represent all possible models one might fit, and
other models could conceivably be selected to yield more extreme results consistent with the
observed data, both higher and lower than those included in this assessment. Etiologically
different tumor types were not combined across sites prior to modeling, in order to allow for the
possibility that different tumor types can have different dose-response relationships because of
varying time courses or other underlying mechanisms or factors. The human equivalent oral
slope factors estimated from the tumor sites with statistically significant increases ranged from
October, 2007
110
DRAFT - DO NOT CITE OR QUOTE

-------
0.020 to 3.0 per mg/kg-day, a range of about two orders of magnitude, with both extremes
coming from the male rat data.
However, given the multiplicity of tumor sites, basing the oral slope factor on one tumor
site may underestimate the carcinogenic potential of 1,2,3-trichloropropane. In addition,
application of one model to a composite data set does not accommodate biologically relevant
information that may vary across sites or may only be available for a subset of sites. Following
the recommendations of the NRC (1994) and the 2005 Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005), a statistically appropriate upper bound on total risk was estimated
in order to gain some understanding of the total risk from multiple tumor sites in male F344 rats
(Table 5-7). Note that this estimate of overall risk describes the risk of developing any
combination of the tumor types considered, not just the risk of developing all three
simultaneously. The estimate of the 95% upper bound on the summed oral slope factor is 4 per
mg/kg-day, about 20% higher than the slope factor for alimentary system tumors only.
Dose metric. 1,2,3-Trichloropropane is metabolized to intermediates with carcinogenic
potential. However, it is unknown whether a metabolite or some combination of parent
compound and metabolites is responsible for the observed toxicity. If the actual carcinogenic
moiety is proportional to administered exposure, then use of administered exposure as the dose
metric is the least biased choice. On the other hand, if this is not the correct dose metric, then
the impact on the slope factor is unknown.
Cross-species scaling. An adjustment for cross-species scaling (BW3 4) was applied to
address toxicological equivalence of internal doses between each rodent species and humans,
consistent with the 2005 Guidelines for Carcinogen Risk Assessment (US EPA, 2005a). It is
assumed that equal risks result from equivalent constant lifetime exposures.
Statistical uncertainty at the point of departure. Parameter uncertainty can be assessed
through confidence intervals. Each description of parameter uncertainty assumes that the
underlying model and associated assumptions are valid. For the multistage-Weibull model
applied to the male rat data, there is a reasonably small degree of uncertainty at the 10% excess
incidence level (the point of departure for linear low-dose extrapolation). The upper bound on
the summed risk for male rats is approximately 1.5-fold higher than the summed risk.
Bioassay selection. The study by NTP (1993) was used for development of an oral slope
factor. This was a well-designed study, conducted in both sexes in two species with a sufficient
number of animals per dose group. The number of test animals allocated among three dose
October, 2007
ill
DRAFT - DO NOT CITE OR QUOTE

-------
levels and an untreated control group was adequate, with examination of appropriate
toxicological endpoints in both sexes of rats and mice. Alternative bioassays were unavailable.
Overall responses across the four species/sex combinations were similarly robust, all involving
the alimentary system in particular, and multiple tumor sites generally.
Choice of species/gender. The oral slope factor for 1,2,3-trichloropropane was quantified
using the tumor incidence data for male rats, which were thought to be more sensitive than
female rats to the carcinogenicity of 1,2,3-trichloroporpane. The male and female mice tumor
incidence data, while clearly demonstrating carcinogenicity, were not suitable for deriving low-
dose quantitative risk estimates. The NTP study design unfortunately missed nearly all of the
relevant dose-response range for mice, with both male and female mice having nearly 100%
responses at the lowest exposure level. While these responses were higher than those of the rats
at the comparable exposure level, suggesting greater sensitivity of the mice, there is no
information concerning the dose-response relationships at lower exposure levels that could be
compared with the rat data. In other words, the high dose behavior of 1,2,3-TCP in mice does
not inform the mouse tumor response to 1,2,3-TCP at lower exposures. Extrapolation from high
response levels is not justified when other more suitable data, here in rats, are available.
Consequently, dose-response modeling was not carried out with the mouse tumor data.
Relevance to humans. The derivation of the oral slope factor is derived using the tumor
incidence in the alimentary system, pancreas, kidney, preputial gland, and Zymbal's gland in
male rats. The human relevance of the forestomach tumors, as included in the tumor incidence
in the alimentary system, is of concern because humans lack a forestomach, which serves as a
food storage organ (Proctor et al., 2007). The oral cavity, pharynx, and glandular stomach are
histologically similar to the rat forestomach, but the tissue dose in these human organs is
different than the tissue dose in the rodent forestomach (Proctor et al., 2007). Chemicals that are
genotoxic and cause tumors at multiple sites in the absence of forestomach irritation are lilely
relevant to human carcinogenesis (Proctor et al., 2007). 1,2,3-Trichloropropane may be
carcinogenic through a mutagenic mode of action and is a multi-site carcinogen in rodents. In
addition, hyperplasia of the forestomach epithelium, a proliferative response often associated
with gavage chemical administration (Proctor et al., 2007), was not observed during the 120-day
subchronic study conducted by NTP (1993). Therefore, the carcinogenicity observed in the
rodent studies is relevant to human exposure. Also, the concordance of the alimentary system
tumors across rats and mice lends strength to the concern for human carcinogenic potential.
Human population variability. The extent of inter-individual variability in 1,2,3-
trichloropropane metabolism has not been characterized. A separate issue is that the human
October, 2007
112
DRAFT - DO NOT CITE OR QUOTE

-------
variability in response to 1,2,3-trichloropropane is also unknown. Although a mutagenic MOA
would indicate increased early-life susceptibility, the data exploring whether there is differential
sensitivity to 1,2,3-trichloropropane carcinogenicity across life stages is unavailable. This lack
of understanding about potential differences in metabolism and susceptibility across exposed
human populations thus represents a source of uncertainty. In addition, due to the lack of
information linking the mode of action for 1,2,3-trichloropropane to the observed
carcinogenicity, the application of ADAFs for estimating risks associated with early-life
exposure is not recommended.
October, 2007
113
DRAFT - DO NOT CITE OR QUOTE

-------
Table 5-8. Summary of uncertainty in the 1,2,3-trichloropropane cancer risk assessment
Consideration/
Approach
Impact on oral slope
factor
Decision
Justification
Low-dose
extrapolation
procedure
Application of
default linear-low-
dose extrapolation
may overestimate
risk; alternatives
could i slope factor
by an unknown
extent
Multistage-
Weibull model to
determine POD,
linear low-dose
extrapolation
from POD
(default approach)
A linear-low-dose extrapolation approach was
used to estimate human carcinogenic risk
associated with 1,2,3-trichloropropane exposure as
the default option. Linear extrapolation is,
generally, considered to be a health-protective
approach, and, in some cases, may lead to an
overestimation of risk, as stated in the 2005
Cancer Guidelines (U.S. EPA, 2005).
Dose metric
Alternatives could t
or I slope factor by
an unknown extent
Used
administered
exposure
Experimental evidence supports a role for
metabolism in toxicity, but actual responsible
metabolites are not clearly identified.
Cross-species
scaling
Alternatives could j
or | slope factor
[e.g., 3.5-fold |
(scaling by BW) or f
2-fold (scaling by
BWm )]
BW3'4 (default
approach)
There are no data to support alternatives. Because
the dose metric was not an AUC, BW3 4 scaling
was used to calculate equivalent cumulative
exposures for estimating equivalent human risks.
Statistical
uncertainty at
POD
i slope factor 1.5-
fold if MLE used
rather than lower
bound on POD
LEC (default
approach for
calculating
reasonable upper
bound slope
factor)
Limited size of bioassay results in sampling
variability; lower bound is 95% confidence
interval on administered exposure.
Bioassay
Alternatives could t
or I slope factor by
an unknown extent
NTP study
Alternative bioassays were unavailable.
Species /gender
combination
Human risk could j
or depending on
relative sensitivity
Male rat MCL
There are no MOA data to guide extrapolation
approach for any choice. It was assumed that
humans are as sensitive as the most sensitive
rodent gender/species tested; true correspondence
is unknown. The carcinogenic response occurs
across species. Generally, direct site concordance
is not assumed; consistent with this view, some
human tumor types are not found in rodents and
rat and mouse tumor types also differ.
Human
relevance of rat
tumor data
Human relevance of
rat tumor tumor data
could i slope factor
Forestomach
tumors in rats are
relevant to human
exposure
1,2,3-Trichloropropane may be carcinogenic
through a mutagenic mode of action and is a multi-
site carcinogen in rodents; therefore, the
carcinogenicity observed in the rodent studies is
relevant to human exposure. In addition,
hyperplasia of the forestomach epithelium, a
proliferative response often associated with
gavage chemical administration, was not observed
during the 120-day subchronic study conducted by
NTP (1993).
Human
population
variability in
metabolism and
response/
sensitive
subpopulations
Low-dose risk f or J,
to an unknown extent
Considered
qualitatively
No data to support range of human
variability /sensitivity, including whether children
are more sensitive. Mutagenic MOA (if fully
established) would indicate increased early-life
susceptibility.
October, 2007
114
DRAFT - DO NOT CITE OR QUOTE

-------
6. MAJOR CONCLUSIONS IN Till CHARACTERIZATION OF
HAZARD AND DOSE RESPONSE
6.1. HUMAN HAZARD POTENTIAL
1,2,3-Trichloropropane (CAS No. 96-18-4) is used in the chemical industry as a solvent
for oils and fats, waxes, and resins. The compound also is used industrially in the production of
polymers, such as polysulfide rubbers, and of some pesticides. Significant amounts of 1,2,3-
trichloropropane are produced as by-products during the manufacture of other chlorinated
compounds, such as epichlorohydrin. The compound is found in consumer products, such as
paint thinner and varnish remover.
Toxicokinetic studies in mice and rats have examined the absorption, distribution,
metabolism, and elimination of the compound. These studies have documented the rapid
metabolism and excretion of the metabolic products in urine or feces, or on the breath
(Mahmood et al., 1991; Volp et al., 1984). The absorbed fraction of an administrated dose is
almost completely metabolized by a combination of both the phase I and phase II metabolic
pathways. Most of the metabolites are rapidly cleared from the body, although a small fraction
of the metabolites have been found to bind to intracellular proteins and nucleic acids (Weber,
1991; Weber and Sipes, 1990, 1991).
No epidemiology studies, case reports, or other studies have documented the effects of
oral exposure to 1,2,3-trichloropropane in humans. Data from a chronic toxicity test in F-344/N
rats and B6C3F1 mice (NTP, 1993) and several subchronic studies (NTP 1993; Merrick et al.
1991; and Villeneuve et al. 1985) have identified the liver as a principal target organ for
noncancer effects. All non-neoplastic changes reported following chronic oral exposure to 1,2,3-
trichloropropane occurred at doses that also produced increased incidences of tumors. A
continuum of hepatic effects has been reported, ranging from cellular necrosis at high doses to
significantly increased organ weights at lower doses. Treatment-related effects were detected in
rats and mice among the hematological parameters, but the effects were not biologically relevant
or related to direct 1,2,3-trichloropropane toxicity (NTP, 1993). Oral exposure has also been
shown to reduce fertility in female CD-I mice (NTP, 1990).
There are very limited data on the effects of 1,2,3-trichloropropane inhalation in humans.
An acute inhalation study from the 1940s found that subjects exposed to 5 ppm trichloropropane
(isomer and purity not reported) for 15 minutes found the odor objectionable and complained of
October, 2007 115 DRAFT - DO NOT CITE OR QUOTE

-------
irritation of the eyes and throat (Silverman et al., 1946). Likewise, there is a limited database of
inhalation toxicity studies in animals, which includes two 2-week studies submitted to EPA by
Miller et al. (1987a,b), a 4-week range finding study, two 13-week studies, and two single-
generation reproductive assessments (Johannsen et al., 1988; Biodynamics, Inc., 1979).
Increased incidences of non-neoplastic lesions were observed in the nasal epithelium,
liver, lungs, and spleen of rats or mice following subchronic inhalation exposure to 1,2,3-
trichloropropane (Johannsen et al., 1988; Miller et al., 1987a, b; Biodynamics, Inc., 1979).
Miller et al. (1987a, b) reported decreased thickness or degeneration of the olfactory epithelium
in rats exposed for 2-weeks to concentrations of 3 ppm 1,2,3-trichloropropane or greater (Table
4-11). Similar effects were also observed in mice that were exposed to concentrations of 10 ppm
1,2,3-trichloropropane or greater (Table 4-12).
Inhalation exposure to 1,2,3-trichloropropane was also associated with significant
increases in organ weights. Increased absolute and relative liver weights were observed in male
rats exposed to concentrations of 5 ppm 1,2,3-trichloropropane, or greater, for 13 weeks
(Johannsen et al., 1988). Increased liver weights were observed following 2-week exposures to
40 ppm, or greater, in rats and 132 ppm in mice (Miller et al., 1986a). Other organ weight
changes included increased relative lung weights in female rats that were exposed to
concentrations of 15 ppm or greater for 13 weeks (Johannsen et al., 1988), and increased relative
kidney and brain weights in male mice exposed to 50 ppm for 13 weeks (Johannsen et al., 1988).
There are no reports of cancer in humans associated with exposure to 1,2,3-
trichloropropane. Increased incidence of tumors was observed in rats and mice following oral
exposure to 1,2,3-trichloropropane (NTP, 1993). Dose-related increasing trends in tumors were
noted at the following sites:
squamous cell carcinomas or papillomas of the alimentary system in male and
female rats and mice;
pancreatic acinar cell adenomas or adenocarcinomas, preputial gland adenomas or
carcinomas, and kidney tubular cell adenomas in male rats;
clitoral gland adenomas or carcinomas, and mammary gland adenocarcinomas in
female rats;
hepatocellular adenomas or carcinomas, harderian gland adenomas in male and
female mice; and
uterine/cervical adenomas or adenocarcinomas in female mice.
October, 2007
116
DRAFT - DO NOT CITE OR QUOTE

-------
All of these tumor sites showed statistically significantly positive trends with increasing
exposure level (Cochran-Armitage test for trend, p<0.05, most with p<.001) and generally
appeared earlier with increasing exposure levels.
In vitro bacterial mutation assays have generally demonstrated a mutagenic potential,
dependent on S9 activation, for 1,2,3-trichloropropane. Mammalian cell in vitro studies have
shown chromosomal damage, gene mutation, DNA breakage, and micronucleus formation after
1,2,3-trichloropropane exposure. In addition, in vivo assays have demonstrated the ability of
1,2,3-trichloropropane metabolites to bind to hepatic proteins, DNA, and RNA, form DNA
adducts in rats and mice; and to induce wing spots (caused by genotoxic effects such as somatic
mutation, chromosomal rearrangement, or nondisjunction) in D. melanogaster. In vivo studies
measuring dominant lethal induction or micronucleus formation were negative and although this
does not necessarily negate the positive mutagenicity studies, these data do limit the confidence
in the hypothesized mode of action.
The data supporting a mutagenic mode of action for carcinogenicity are limited and areas
of uncertainty exist. For example, regular test batteries for different genetic end points in vitro
and, especially, in vivo, are limited or missing from the database. Evidence of gene mutations in
in vivo systems would provide substantial support for a mutagenic mode of action, but these
studies have not been conducted. In addition, evidence of cytogenetic effects in humans would
be useful to better characterize the mode of action for 1,2,3-trichloropropane. Therefore, the
available data indicate that a mutagenic mode of action is possible, but the database is limited by
a lack of evidence that mutagenic events occur following 1,2,3-trichloropropane exposure.
6.2. DOSE RESPONSE
6.2.1. Noncancer/Oral
The NTP (1993) study is selected as the principal study because it was a well-designed
chronic study, conducted in both sexes in two species with a sufficient number of animals per
dose group. The number of test animals allocated among three dose levels and an untreated
control group was acceptable, with examination of appropriate toxicological endpoints in both
sexes of rats and mice. Increased liver weight is chosen as the critical effect because liver
toxicity appeared to be the most sensitive effect. There is evidence of hepatocellular damage,
including increased incidence of hepatic necrosis and decreased synthesis of
pseudocholinesterase, from the subchronic NTP (1993) study, and increased serum
concentrations of hepatocellular enzymes, decreased concentration of 5'-nucleotidase, and
October, 2007
117
DRAFT - DO NOT CITE OR QUOTE

-------
increase incidence of histopathologic liver lesions, from the chronic NTP (1993) study. Thus,
increased liver weight appears to be on a continuum of adverse liver effects associated with oral
exposure to 1,2,3-trichloropropane.
Other effects considered in the selection of the critical effect included kidney, respiratory,
myocardial, or reproductive toxicity endpoints. The increase in kidney weights after both
subchronic and chronic exposure was accompanied by renal tubular necrosis in the subchronic
NTP (1993) study. In addition, NTP (1993) study demonstrated epithelial necrosis in the nasal
turbinates of rats and regenerative lung lesions in mice following subchronic exposure to 1,2,3-
trichloropropane. Pulmonary toxicity including an increased incidence of inflammation-
associated myocardial necrosis in rats and increased levels of creatine kinase were also observed
(NTP, 1993; Merrick et al., 1991). NTP (1990) demonstrated a decrease in the number of
pregnancies per fertile pair, a reduction in the number of live pups/litter, and a decrease in the
proportion of male pups born alive. Although the liver appeared to be the most sensitive
indicator of 1,2,3-trichloropropane-induced toxicity, reference doses for the changes in kidney
weight, fertility, and pups/liter were quantified for comparison purposes.
Benchmark dose (BMD) modeling was conducted to calculate potential points of
departure for deriving the chronic RfD by estimating the effective dose at a specified level of
response (BMDX) and its 95% lower bound (BMDLX) for the changes in liver and kidney weight,
fertility, and live pups/litter associated with chronic exposure to 1,2,3-trichloropropane. A BMR
of 10% was selected for the derivation of the BMDL for liver and kidney weight increases, and
the BMR of 1SD was modeled for comparison purposes. In the developmental study, a 10%
decrease in fertility and a 1% change in mean live pups/litter for the 4th and 5th litters were
selected as the BMR due to the frank toxicity of the reproductive toxicity endpoint.
The chronic RfD of 4 x 10"3 mg/kg-day was calculated from a BMDLAdj of 1.14 mg/kg-
day for increased absolute liver weight in male rats chronically exposed to 1,2,3-
trichloropropane by gavage (NTP, 1994). A total UF of 300 was used: 10 for interspecies
variability, 10 for interindividual variability, and 3 for database uncertainties. Information was
unavailable to quantitatively assess toxicokinetic or toxicodynamic differences between animals
and humans and the potential variability in human susceptibility, thus, the interspecies and
intraspecies uncertainty factors of 10 were applied. In addition, a 3-fold database uncertainty
factor was applied due to the lack of information addressing the potential developmental toxicity
associated with 1,2,3-trichloropropane. The RfD Comparison Figure below presents the
potential points of departure, applied uncertainty factors, and derived chronic RfD and
October, 2007
118
DRAFT - DO NOT CITE OR QUOTE

-------
comparison RfDs for the critical effect and additional endpoints, respectively, from Table 5-1 in
Section 5.
The overall confidence in this chronic RfD assessment is medium-to-high. Confidence in
the principal study (NTP, 1993) is high. Confidence in the database is medium to high even
though the database lacks a multigenerational developmental toxicity study. The lack of a
multigenerational study is of particular concern due to the genotoxicity of 1,2,3-
trichloropropane, because genetic damage to the germ cells of the F1 generation may not be
detected until the F2 generation. Reflecting high confidence in the principal study and medium-
to-high confidence in the database, confidence in the RfD is medium-to-high.
October, 2007
119
DRAFT - DO NOT CITE OR QUOTE

-------
Figure 6-1. Points of Departure for endpoints from Table 5-1 with corresponding applied uncertainty factors and derived chronic
RfD.
100
I?
¦o
D>
D>
E
0.01
0.001
0.0001
Increased
absolute liver
weight, male
rats (NTP,
1993)*

W
Increased
Increased
Increased
relative
relative
absolute
kidney
liver _
_ kidney 	
	 weight,
weight,
weight,
male rats
male rats
female
(NTP,
(NTP,
rats (NTP,
1993)
1993)
1993)

Fertility
generating
the 4"
litter, mice
(NTP,
1990)
Fertility
generating
5th litter,
mice
(NTP,
1990)
Live
pups/litter
- 4th litter,
mice
(NTP,
1990)
Live
pups/litter
- 5th litter,
mice
(NTP,
1990)
~ Point of Departure
UF, animal-to-human
UF, human variability
UF, database
•	RfD
*	Critical effect and recommended RfD
October, 2007
IJKA^I -JyWLJINUI umuKguum

-------
6.2.2. Noncancer/Inhalation
The Johannsen et al. (1988) study is selected as the principal study because it was a well-
designed subchronic study with a sufficient number of animals per dose group. The number of
test animals allocated among five dose levels and an untreated control group was acceptable,
with examination of appropriate toxicological endpoints in both sexes of rats and mice. The
critical effect selected for the derivation of the chronic RfC is the development of peribronchial
lymphoid hyperplasia in the lungs of male CD rats, with a NOAEL of 1.5 ppm and a LOAEL of
5 ppm 1,2,3-trichloropropane, due to the occurrence of this adverse effect in both male and
female rats and the possible correlation between the hyperplasia and the observed increased lung
weight. The increase in lung weight had a NOAEL of 5 ppm and a LOAEL of 15 ppm.
Increased liver weights were also apparent, however, the hepatocellular hypertrophy observed in
males at 5, 15, and 50 ppm appeared to be at mild centrilobular to midzonal levels and was not
observed in the highest dose group females, and was considered potentially adaptive in the
absence of additional overt toxicity in the liver.
There is uncertainty in the POD not captured by the NOAEL/LOAEL approach, because
it lacks characterization of the dose-response curve and is less informative than a POD obtained
from benchmark dose-response modeling. A NOAEL or LOAEL reflects the particular exposure
concentration or dose at which a study was conducted, and the number of study subjects or test
animals and typically are dissimilar in detection ability and statistical power. The NOAEL of
1.5 ppm for peribronchial lymphoid hyperplasia may not represent a 0% response level. Rather,
the NOAEL of 1.5 ppm in 15 male rats has a 95% confidence limit for 0% response of 0 to 24%;
or, in other words, there is a 95% chance that the "true" response of peribronchial lymphoid
hyperplasia at 1.5 ppm would be as high as 24%. The associated uncertainty in the POD cannot
be characterized quantitatively.
Human equivalent concentrations (HECs) were calculated from the candidate point of
departure. HECs were converted to mg/m3, adjusted to continuous exposure (7 days a week, 24
hours a day), and multiplied by a dosimetric adjustment factor (DAF), a ratio of animal and
human physiologic parameters. The specific DAF used depends on the nature of the
contaminant (particle or gas) and the target site (e.g., respiratory tract or remote to the portal-of-
entry). The DAF for an extra-respiratory effect of a gas is the ratio of the animal/human blood:
air partition coefficients [(Hb/g)A/(Hb/g)H]- However, the human and rat blood partition
coefficients for 1,2,3-trichloropropane are not known. In accordance with the RfC Methodology
(U.S. EPA, 1994) when the partition coefficients are unknown a ratio of 1 is used. The unknown
human and rat blood partition coefficients for 1,2,3-trichloropropane represent a significant data
October, 2007
121
DRAFT - DO NOT CITE OR QUOTE

-------
gap, in which the availability of this information would provide for a more accurate HEC
calculation.
The chronic RfC of 5 x 10"4 mg/m3 was calculated from a NOAELHec of 1.6 mg/m3 for
increased incidence of peribronchial lymphoid hyperplasia in the lungs of male CD rats
(Johannsen et al., 1988). A total UF of 3000 was used: 3 for interspecies variability, 10 for
interindividual variability, 10 for extrapolating from a subchronic to chronic exposure duration,
and 10 for database deficiencies. A factor of 3 was selected to account for uncertainties in
extrapolating from rats to humans, which is adopted by convention where an adjustment from an
animal specific NOAELadj to a NOAELrec has been incorporated. Insufficient information is
available to predict potential variability in susceptibility among the population, thus the human
variability uncertainty factor of 10 was applied. A 10-fold UF was used to account for
uncertainty in extrapolating from a subchronic to chronic exposure duration. A 10-fold UF was
used to account for deficiencies in the database. The database of 1,2,3-trichloropropane
inhalation studies is lacking a multigenerational reproductive study and a developmental toxicity
study. The lack of the multigenerational study is of particular concern due to the genotoxicity of
1,2,3-trichloropropane, because genetic damage to the germ cells of the F1 generation may not
be detected until the F2 generation.
The overall confidence in this chronic RfC assessment is low. Confidence in the
principal study (Johannsen et al.; 1988) is low. Confidence in the database is low as the database
lacks a chronic inhalation bioassay and multigenerational reproductive and developmental
toxicity studies. The lack of a chronic inhalation bioassay is of concern because the critical
effect, peribronchial lymphoid hyperplasia, may be more severe at lower doses with a prolonged
exposure, and additional critical effects not observed following subchronic exposure may arise
following chronic exposure. The lack of a multigenerational developmental study is of particular
concern due to the genotoxicity of 1,2,3-trichloropropane, because genetic damage to the germ
cells of the F1 generation may not be detected until the F2 generation. Reflecting low-to-
medium confidence in the principal study and low-to-medium confidence in the database,
confidence in the chronic RfC is low.
6.2.3. Cancer/Oral and Inhalation
Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), 1,2,3-
trichloropropane is likely to be carcinogenic to humans, based on the existence of compelling
evidence of the compound's tumorigenicity in a single, well-carried-out study in two animal
species (Irwin et al., 1995; NTP, 1993). There are no studies that examine the potential
October, 2007
122
DRAFT - DO NOT CITE OR QUOTE

-------
carcinogenicity of 1,2,3-trichloropropane in humans. While the use of gavage studies in
experimental animals to extrapolate to human exposure to the compound in drinking water may
introduce quantitative uncertainty, the consistent dose-dependent formation of tumors, at and
remote from the site-of-entry in two animal models, suggests a tumorigenic capacity of 1,2,3-
trichloropropane in humans.
A dose-related, statistically significant increasing trend in tumors was observed in the
following sites:
squamous cell carcinomas or papillomas of the alimentary system in male and female rats
and mice;
Zymbal's gland carcinomas in male and female rats;
pancreatic acinar cell adenomas or adenocarcinomas, preputial gland adenomas or
carcinomas, and kidney tubular cell adenomas in male rats;
clitoral gland adenomas or carcinomas, and mammary gland adenocarcinomas in female
rats;
hepatocellular adenomas or carcinomas, and harderian gland adenomas in male and
female mice; and
uterine/ cervical adenomas or adenocarcinomas in female mice.
These tumors generally appeared earlier with increasing exposure levels, and showed
statistically significantly increasing trends with increasing exposure level. Etiologically similar
tumor types, benign and malignant tumors of the same cell type, were combined for these
tabulations because of the possibility that the benign tumors could progress to the malignant
form (US EPA, 2005a). This assumption, if incorrect, has some limited potential to over-
estimate the carcinogenic potential of 1,2,3-trichloropropane, and is an accepted practice
(McConnell etal., 1986).
The male and female mouse tumor incidence data, while clearly demonstrating
carcinogenicity, were not suitable for deriving low-dose quantitative risk estimates. The NTP
study design unfortunately missed nearly all of the relevant dose-response range for mice, with
both male and female mice having nearly 100% responses at the lowest exposure level.
Consequently, dose-response modeling was not carried out with the mouse tumor data. The
elimination of the mouse data from the dose-response modeling has the potential to under-
estimate the carcinogenic risk of 1,2,3-trichloropropane if mice are in fact more sensitive than
the rats. Unfortunately, the high dose behavior of 1,2,3-trichloropropane in mice does not
inform whether mice would be more, the same, or less sensitive than rats to 1,2,3-TCP
carcinogenity at lower exposures.
October, 2007
123
DRAFT - DO NOT CITE OR QUOTE

-------
The MOA is a key consideration in clarifying how risks should be estimated for low-dose
exposure. A linear-low-dose extrapolation approach was used to estimate human carcinogenic
risk associated with 1,2,3-trichloropropane exposure as the default option. This approach is
supported by the positive evidence of genotoxicity and a potential mutagenic mode of action.
The extent to which the overall uncertainty in low-dose risk estimation could be reduced
if the MOA for 1,2,3-trichloropropane were known with a high degree of confidence is of
interest, but additional supporting data on the MOA of 1,2,3-trichloropropane are not available.
Even if it were, incorporation of MOA into dose-response modeling might not be straightforward
and might not significantly reduce the uncertainty about low-dose extrapolation.
Due to the occurrence of multiple tumor types, earlier occurrence with increasing
exposure, and early termination of at least one dose group, dose-response methods which can
reflect the influence of competing risks and intercurrent mortality on site-specific tumor
incidence rates are preferred. EPA has generally used the multistage-Weibull model in this type
of situation, because it incorporates the time at which death-with-tumor occurred and can
account for differences in mortality observed between the exposure groups in the rat bioassay.
Additionally, etiologically different tumor types were not combined across sites prior to
modeling, in order to allow for the possibility that different tumor types can have different dose-
response relationships because of varying time courses or other underlying mechanisms or
factors.
Points of departure for estimating low-dose risk were identified at doses at the lower end
of the observed data, generally corresponding to 10% extra risk, defined as the extra risk over the
background tumor rate. The lifetime oral cancer slope factor for humans is defined as the slope
of the line from the lower 95% bound on the exposure at the point of departure. This 95% upper
confidence limit (UCL) represents a plausible upper bound on the true risk.
Adjustments for approximating human equivalent slope factors applicable for continuous
exposure were calculated. Following EPA's cross-species scaling methodology, the time-
weighted daily average doses were converted to human equivalent doses on the basis of (body
weight)3 4 (U.S. EPA, 1992) and the estimated continuous daily exposures were calculated by
multiplying each slope factor by (5 days)/(7 days) = 0.71. The impact of applying these
adjument factors to the slope factor is unknown. The human equivalent oral slope factors
estimated from the tumor sites with statistically significant increases ranged from 0.020 to 3.0
per mg/kg-day.
October, 2007
124
DRAFT - DO NOT CITE OR QUOTE

-------
However, given the multiplicity of tumor sites, basing the oral slope factor on one tumor
site may underestimate the low-dose carcinogenic potential of 1,2,3-trichloropropane. Following
the recommendations of the NRC (1994) and the 2005 Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005), a statistically appropriate upper bound on total risk was estimated
in order to gain some understanding of the total risk from multiple tumor sites in male F344 rats
(Table 5-7). Note that this estimate of overall risk describes the risk of developing any
combination of the tumor types considered, not just the risk of developing all three
simultaneously.
The recommended estimate for an upper bound on human extra cancer risk from
lifetime oral exposure to 1,2,3-trichloropropane derived from male animal data is 4 per
mg/kg-day, compared with 3 per mg/kg/day for only alimentary system tumors and 0.020
per mg/kg-day for only Zymbal's gland tumors. This estimate reflects the time-to-tumor
response as well as the exposure-response relationships for the multiple tumor sites in
male rats. The value based on male rats is recommended because male rats are the most
sensitive to tumor induction following exposure to 1,2,3-trichloropropane and yield the
highest slope factor. Note that this slope factor should not be used with exposures greater
than 0.05 mg/kg/day, since the observed dose-response does not continue linearly above
this level. For female rats, the combined slope factor was about two-fold lower, at 2.0
per mg/kg-day, a 50% increase compared with 1.3 per mg/kg-day for alimentary system
tumors only.
The uncertainties associated with the quantitation of the oral slope factor are described
below:
Choice of low-dose extrapolation approach. The MOA is a key consideration in
clarifying how risks should be estimated for low-dose exposure. A linear-low-dose
extrapolation approach was used to estimate human carcinogenic risk associated with
1,2,3-trichloropropane exposure as the default option. Linear extrapolation is, generally,
considered to be a health-protective approach, and, in some cases, may lead to an
overestimation of risk, as stated in the 2005 Cancer Guidelines (U.S. EPA, 2005).
The extent to which the overall uncertainty in low-dose risk estimation could be reduced
if the MOA for 1,2,3-trichloropropane were known with a higher degree of confidence is
of interest, but additional supporting data on the MOA of 1,2,3-trichloropropane is not
available.
October, 2007
125
DRAFT - DO NOT CITE OR QUOTE

-------
Without data sufficient to drive a more biologically-based model, a dose-response model
with the capacity to accommodate some of the observed data, specifically the times to
death-with-tumor, was used. It is unknown, however, how well the multistage-Weibull
model or the linear low-dose extrapolation predicts low-dose risks for 1,2,3-
trichloropropane.
Dose metric. 1,2,3-Trichloropropane is metabolized to intermediates with carcinogenic
potential. However, it is unknown whether a metabolite or some combination of parent
compound and metabolites is responsible for the observed toxicity. If the actual
carcinogenic moiety is proportional to administered exposure, then use of administered
exposure as the dose metric is the least biased choice. On the other hand, if this is not the
correct dose metric, then the impact on the slope factor is unknown.
Cross-species scaling. An adjustment for cross-species scaling (BW3 4) was applied to
address toxicological equivalence of internal doses between each rodent species and
humans, consistent with the 2005 Guidelines for Carcinogen Risk Assessment (US EPA,
2005a). It is assumed that equal risks result from equivalent constant lifetime exposures.
Statistical uncertainty at the point of departure. Parameter uncertainty can be assessed
through confidence intervals. Each description of parameter uncertainty assumes that the
underlying model and associated assumptions are valid. For the multistage-Weibull
model applied to the male rat data, there is a reasonably small degree of uncertainty at the
10% excess incidence level (the point of departure for linear low-dose extrapolation).
The upper bound on the summed risk for male rats is approximately 1.5-fold higher than
the summed risk.
Bioassay selection. The study by NTP (1993) was used for development of an oral slope
factor. This was a well-designed study, conducted in both sexes in two species with an
adequate number of animals per dose group. The number of test animals allocated
among three dose levels and an untreated control group was adequate, with examination
of appropriate toxicological endpoints in both sexes of rats and mice. Alternative
bioassays were unavailable. Overall responses across the four species/sex combinations
were similarly robust, all involving the alimentary system in particular, and multiple
tumor sites generally.
October, 2007
126
DRAFT - DO NOT CITE OR QUOTE

-------
Choice of species/gender. The oral slope factor for 1,2,3-trichloropropane was quantified
using the tumor incidence data for male rats, which were thought to be more sensitive
than female rats to the carcinogenicity of 1,2,3-trichloroporpane. The male and female
mice tumor incidence data, while clearly demonstrating carcinogenicity, were not
suitable for deriving low-dose quantitative risk estimates. The NTP study design
unfortunately missed nearly all of the relevant dose-response range for mice, with both
male and female mice having nearly 100% responses at the lowest exposure level. While
these responses were higher than those of the rats at the comparable exposure level,
suggesting greater sensitivity of the mice, there is no information concerning the dose-
response relationships at lower exposure levels that could be compared with the rat data.
In other words, the high dose behavior of 1,2,3-TCP in mice does not inform the mouse
tumor response to 1,2,3-TCP at lower exposures. Extrapolation from high response
levels is not justified when other more suitable data, here in rats, are available.
Consequently, dose-response modeling was not carried out with the mouse tumor data.
Relevance to humans. The derivation of the oral slope factor is derived using the tumor
incidence in the alimentary system, pancreas, kidney, preputial gland, and Zymbal's
gland in male rats. The human relevance of the forestomach tumors, as included in the
tumor incidence in the alimentary system, is of concern because humans lack a
forestomach, which serves as a food storage organ (Proctor et al., 2007). The oral cavity,
pharynx, and glandular stomach are histologically similar to the rat forestomach, but the
tissue dose in these human organs is different than the tissue dose in the rodent
forestomach (Proctor et al., 2007). 1,2,3-Trichloropropane may be carcinogenic through
a mutagenic mode of action and is a multi-site carcinogen in rodents; therefore, the
carcinogenicity observed in the rodent studies is relevant to human exposure. In
addition, hyperplasia of the forestomach epithelium, a proliferative response often
associated with gavage chemical administration (Proctor et al., 2007), was not observed
during the 120-day subchronic study conducted by NTP (1993). The concordance of the
alimentary system tumors across rats and mice lends strength to the concern for human
carcinogenic potential.
Human population variability. The extent of inter-individual variability in 1,2,3-
trichloropropane metabolism has not been characterized. A separate issue is that the
human variability in response to 1,2,3-trichloropropane is also unknown. Although a
mutagenic MOA would indicate increased early-life susceptibility, the data exploring
whether there is differential sensitivity to 1,2,3-trichloropropane carcinogenicity across
life stages is unavailable. This lack of understanding about potential differences in
October, 2007
127
DRAFT - DO NOT CITE OR QUOTE

-------
metabolism and susceptibility across exposed human populations thus represents a source
of uncertainty. In addition, due to the lack of information linking the mode of action for
1,2,3-trichloropropane to the observed carcinogenicity, the application of ADAFs for
estimating cancer risks associated with early-life exposure is not recommended.
An inhalation unit risk was not derived in this assessment. Data on the carcinogenicity
of the compound via the inhalation route is unavailable, and route-to-route extrapolation was not
possible due to the lack of an adequate physiologically based pharmacokinetic model. However,
it is proposed that 1,2,3-trichloropropane is likely to be carcinogenic to humans by the inhalation
route since the compound is well-absorbed, and in oral studies induces tumors at sites other than
the portal of entry.
October, 2007
128
DRAFT - DO NOT CITE OR QUOTE

-------
7. REFERENCES
Anderson, EL; McGaughy, R; Bayard, S; et al. (1983) Quantitative approaches in use to assess cancer risk. Risk
Anal 4: 277-295.
ATSDR (Agency for Toxic Substances and Disease Registry). (1992) Toxicological profile for 1,2,3-
trichloropropane. Public Health Service, U.S. Department of Health and Human Services, Public Health Service,
Atlanta, GA. Available online at .
Bailer, AJ; Portier, CJ. (1986) Effects of treatment-induced mortality and tumor-induced mortality on tests for
carcinogenicity in small samples.
Belyaeva, NN; Tsulaya, VR; Marshak, TL; et al. (1974) Effect of 1,2,3-trichloropropane on the ploidy of rat liver
cells. Bull Exp Biol Med (USSR) 78(12):74-77.
Belyaeva, NN: Bonashevskaya, TI; Marshak, TL; et al. (1977) Investigation of the effect of certain chlorinated
hydrocarbons on the composition of the hepatocyte population of the rat liver. Bull Exp Biol Med (USSR)
83(3):345-348.
Biodynamics, Inc. (1979) A 13-week inhalation toxicity study of 1,2,3-trichloropropane in rats with attachments
and cover letter dated 053092. Submitted under TSCA 8E. EPA Document No. 88920003764. NTIS No.
OTS0542010.
Chroust, K; Pavlova, M; Zbynek, P; Mendel, J; Bozkova, K; Kubat, Z; Zajickova, V; Damborsky, J. (2006)
Quantitative structure-activity relationships for toxicity and genotoxicity of halogenated aliphatic compounds: wing
spot test of Drosophila melanogaster. Chemosphere, doi: 10.1016j.chemosphere.2006.09.020.
Crebelli, R; Andreoli, C; Carere, A; et al. (1992) The induction of mitotic chromosome malsegregation in
Aspergillus nidulans. quantitative structure activity relationship (QSAR) analysis with chlorinated aliphatic
hydrocarbons. MutatRes 266:117-134.
Crebelli, R; Carere, A; Leopardi, P; et al. (1999) Evaluation of 10 aliphatic halogenated hydrocarbons in the mouse
bone marrow micronucleus test. Mutagenesis 14:207-215.
Dearfield, KL; Moore, MM. (2005) Use of genetic toxicology information for risk assessment. Environ Molecul
Mutagen 46:236-245.
Doherty, AT; Ellard, S; Parry, EM; et al. (1996) An investigation into the activation and deacativation of
chlorinated hydrocarbons to genotoxins in metabolically competent human cells. Mutagenesis 11(3):247-274.
Douglas, GR; Nestmann, ER; Lee, E; et al. (1985) How well do in vitro tests predict in vivo genotoxicity? Environ
Mutagen 7(Suppl 3): 31.
Eriksson, L; Jonsson, J; Hellberg, S; et al. (1991) A strategy for ranking environmentally occurring chemicals. Part
V: the development of two genotoxicity QSARs for halogenated aliphatics. Environ Toxicol Chem 10:585-596.
October, 2007
129
DRAFT - DO NOT CITE OR QUOTE

-------
Ezaz-Nikpay, K; Verdine, GL. (1984) The effects of N7-methylguanine on duplex DNA structure. Chem Biol
1:235-240.
Fortoul, TI; Cano-Valle, F; Oliva, et al. (1985) Follicular bronchiolitis in association with connective tissue
diseases. Lung 163:305-314.
Gart, JJ; Krewski, D; Lee, PN; et al. (1986) Statistical Methods in Cancer Research. Lyon: International Agency for
Research on Cancer.
Gasparutto, D; Michel, T; Ramirez-Fuentes, T; et al. (2005) Epoxide adducts at the guanine residue within single-
stranded DNA chains: reactivity and stability studies. Nucleosides Nucleotides Nucleic Acids 24(5-7):545-552.
Hatch, G; Anderson, T; Elmore, E; et al. (1983) Status of enhancement of DNA viral transformation for
determination of mutagenic and carcinogenic potential of gaseous and volatile compounds. Environ Mutagen 5:422.
Haworth, S; Lawlor, T; Mortelmans, K; et al. (1983) Salmonella mutagenicity test results for 250 chemicals.
Environ Mutagen 5:3-142.
Hazelton Laboratories. (1983a) Initial submission: 120-day gavage toxicity study with 1,2,3-trichloropropane in
Fischer 344 rats (final report) with attachments and cover letter dated 050692. Submitted under TSCA 8E. EPA
Document No. 88920002189. NTIS No. OTS0536324.
Hazelton Laboratories. (1983b) Initial submission: 120-day gavage toxicity study inB6C3Fl mice with 1,2,3-
trichloropropane (final report) with attachments and cover letter dated 050692. Submitted under TSCA Section 8E.
EPA Document No. 88920002188. NTIS No. OTS0536323.
Holme, JA; Soderlund, EJ; Brunborg, G; Omichinski, JG; Bekkedal, K; Trygg, B; Nelson, SD; Dybing, E. (1989)
Different mechanisms are involved in DNA damage, bacterial mutagenicity, and cytotoxicity induced by 1,2-
dibromo-3-chloropropane in suspensions of rat liver cells. Carcinog 10(l):49-54.
Holme, JA; Soderlund, EJ; Brunborg, G; et al. (1991) DNA damage and cell death induced by l,2-dibromo-3-
chloropropane (DBCP) and structural analogs in monolayer culture of rat hepatocytes: 3-aminobenzamide inhibits
the toxicity of DBCP. Cell Biol Toxicol 7:413-432.
Howling, SJ; Hansell, DM; Wells, AU; et al. (1999) Follicular bronchiolitis: thin-section CT and histologic
findings. Radiology 212:637-642.
HSDB (Hazardous Substances Data Bank). (2005) 1,2,3-Trichloropropane. National Library of Medicine, Bethesda,
MD. Available online at .
Humphreys, WG; Kim, D-H; Guengerich, FP. (1991) Isolation and characterization of N7-guanyl adducts derived
from l,2-dibromo-3-chloropropane. Chem Res Toxicol 4: 445-453.
October, 2007
130
DRAFT - DO NOT CITE OR QUOTE

-------
Irwin, RD; Haseman, JK; Eustis, SL. (1995) 1,2,3-Trichloropropane: a multisite carcinogen in rats and mice.
Fundam Appl Toxicol 25:241-252.
Ito, N; La, DK; Holt, S; et al. (1996) Analysis of ras mutations in forestomach tumors from B6C3F1 mice exposed
to 1,2,3-trichloropropane. Proc Amer Assoc Cancer Res 37:137.
Johannsen, FR; Levinskas, GJ; Rusch, GM; et al. (1988) Evaluation of the subchronic and reproductive effects of a
series of chlorinated propanes in the rat. I. Toxicity of 1,2,3-trichloropropane. J Toxicol Environ Health 25:299-315.
Kapp, RW. (1979) Mutagenicity of l,2-dibromo-3-chloropropane (DBCP): in vivo cytogenetics study in the rat.
Toxicol Appl Pharmacol 48(1):A46.
Kier, LD. (1982) Ames/salmonella mutagenicity assays of 1,2,3-trichloropropane, 1,2,2,3-tetrachloropropane, and
1,1,2,2,3-pentachloropropane. Prepared by Environmental Health Laboratory, Monsanto Company, St. Louis, MO,
(NTIS/OTS 0000815, 19 August, 1982): 1-19.
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.
La, DK; Lilly, PD; Anderegg, RJ; et al. (1995) DNA adduct formation in B6C3F1 mice and Fischer 344 rats
exposed to 1,2,3-trichloropropane. Carcinogenesis 16:1419-1424.
La, DK; Schoonhoven, R; Ito, N; et al. (1996) The effects of exposure route on DNA adduct formation and cellular
proliferation by 1,2,3-trichloropropane. Toxicol Appl Pharmacol 140:108-114.
La, DK; Swenberg, JA. (1997) The induction of endogenous DNA adducts in tissues of Fischer-344 rats following
gavage administration of 1,2,3-trichloropropane. Proc Amer Assoc Cancer Res 38:131.
Lag, ML; Soderlund, EJ; Omichinski, JG; et al. (1991) Effect of bromine and chlorine positioning in the induction
of renal and testicular toxicity by halogenated propanes. Chem Res Toxicol 4: 528-534.
Lag, M; Omichinski, JG; Dybing, E; et al. (1994) Mutagenic activity of halogenated propanes and propenes: Effect
of bromine and chlorine positioning. Chem Biol Interact 93:73-84.
Lutz, WK. (1979) In vivo covalent binding of organic chemicals to DNA as a quantitative indicator in the process
of chemical carcinogenesis. Mutat Res 65:289-356.
McConnell EE, Solleveld HE, Swenberg JA, Boorman GA. (1986) Guidelines for combining neoplasms for
evaluation of rodent carcinogenesis studies. J Natl Cancer Inst. 76(2):283-9.
McOmie, WA; Barnes, TR. (1949) Acute and subacute toxicity of 1,2,3-trichloropropane in mice and rabbits. Fed
Proc Pharm Exp Therap 8:319.
Mahmood, NA; Overstreet, D; Burka, LT. (1991) Comparative disposition and metabolism of 1,2,3-
trichloropropane in rats and mice. Drug Metab Dispos 19:411-418.
October, 2007
131
DRAFT - DO NOT CITE OR QUOTE

-------
Meier, JR; Ringhand, HP; Coleman, WE; Munch, JW; Streicher, RP; Kaylor, WH; Schenck, KM. (1985)
Identification of mutagenic compounds formed during chlorination of humic acid. Mutat Res 157:11-122.
Merrick, BA; Smallwood, CL; Meier, JR; McKean, DL; Kaylor, WH; Condie, LW. (1987) Chemical reactivity,
cytotoxicity, and mutagenicity of chloropropanes. Toxicol Appl Pharmacol 91:46-54.
Merrick, BA; Robinson, M; Condie, LW. (1991) Cardiopathic effect of 1,2,3-trichloropropane after subacute and
subchronic exposure in rats. J Appl Toxicol 11:179-187
Miller, RR; Quast, JF; Gushow, TS. (1987a) 1,2,3-Trichloropropane: 2-week vapor inhalation study in rats
and mice. Dow Chemical Co., submitted to EPA under TSCA Sect. 8D. NTIS Nr. OTS 86-870002260.
Miller, RR; Quast, JF; Momany-Pfruender, JJ. (1987b) 1,2,3-Trichloropropane: 2-week vapor inhalation study to
determine the no-adverse-effect level in rats and mice. Dow Chemical Co., submitted to EPA under TSCA Sect.
8D. NTIS Nr. OTS 86-8700.
Mirsalis, J; Tyson, K; Beck, J; et al. (1983) Induction of unscheduled DNA synthesis (UDS) in hepatocytes
following in vitro and in vivo treatment. Environ Mutagen 5:482.
Myers, JL; Kurtin, PJ. (1995) Lymphoid proliferative disorders of the lung. In: Thurlbeck, WM; Churg, AM, eds.
Pathology of the lung. 2nd ed. New York, NY:Thieme, 1995:535-588.
NCI (National Cancer Institute). (1978) Bioassay of dibromochloropropane for possible carcinogenicity. Prepared
for the National Cancer Institute, U.S. Department of Health, Education, and Welfare, Public Health Service,
National Institutes of Health, Bethesda, MD. DHEW No. 78-828. (as cited inNTP, 1993)
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press.
NRC (National Research Council). (1994) Science and Judgment in Risk Assessment. National Academy Press,
Washington, DC.
NTP (National Toxicology Program). (1978)) Bioassay of 1,2-dichloroethane for possible carcinogenicity (CAS
No. 107-06-2). Public Health Service, U.S. Department of Health and Human Services; NTP TR 55. NIH
Publication No. 78-1361. Available from: National Institute of Environmental Health Sciences, Research Triangle
Park, NC.
NTP. (1982) Carcinogenesis bioassay of l,2-dibromo-3-chloropropane (CAS NO. 96-12-8) inF344/N rats and
B6C3F1 mice (inhalation studies) [review draft]. Public Health Service, U.S. Department of Health and Human
Services; NTP TR 206. NIH Publication No. 82-1762. Available from: National Institute of Environmental Health
Sciences, Research Triangle Park, NC. (as cited in NTP, 1993).
NTP. (1986) Toxicology and carcinogenesis studies of 1,2-dichloropropane (propylene dichloride) (CAS No. 78-
87-5) in F344/N rats and B6C3F1 mice (gavage studies). Public Health Service, U.S. Department of Health and
October, 2007
132
DRAFT - DO NOT CITE OR QUOTE

-------
Human Services; NTP TR 263. NIH Publication No. 86-2519. Available from: National Institute of Environmental
Health Sciences, Research Triangle Park, NC.
NTP. (1989) Toxicology and carcinogenesis studies of hexachloroethane (CAS No. 67-72-1) inF344/N rats
(gavage studies). Public Health Service, U.S. Department of Health and Human Services; NTP TR 361. NIH
Publication No. 89-2816. Available from: National Institute of Environmental Health Sciences, Research Triangle
Park, NC.
NTP. (1990) 1,2,3-Trichloropropane reproduction and fertility assessment in Swiss CD-I mice when administered
via gavage (final report). Public Health Service, U.S. Department of Health and Human Services; NTP-90-209.
Available from: National Institute of Environmental Health Sciences, Research Triangle Park, NC.
NTP. (1993) Toxicology and carcinogenesis studies of 1,2,3-trichloropropane (CAS No. 96-18-4) in F344/N rats
and B6C3F1 mice (gavage studies). Public Health Service, U.S. Department of Health and Human Services; NTP
TR 384. NIH Publication No. 94-2839. Available from: National Institute of Environmental Health Sciences,
Research Triangle Park, NC.
NTP. (2005) Carcinogenesis studies of 2,2-bis(bromomethyl)-1,3-propanediol, nitromethane, and 1,2,3-
trichloropropane (CAS Nos. 3296-90-0, 75-52-5, and 96-18-4) in guppies (Poecilia reticulate) and medaka (Oryzias
latipes) (waterborne studies). Public Health Service, U.S. Department of Health and Human Services; NTP TR 528.
NIH Publication No. 06-4464. Available from: National Institute of Environmental Health Sciences, Research
Triangle Park, NC.
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.
Proctor, DM; Gatto, NM; Hong, SJ; et al. (2007) Mode-of-action framework for evaluating the relevance of rodent
forestomach tumors in cancer risk assessment. Toxicol Sciences April 10.
Ratpan, F; Plaumann, H. (1988) Mutagenicity of halogenated propanes and their methylated derivatives. Environ
Mutagen 12:253-259.
Reyna, MS. (1987) Acute inhalation of trichloropropane, NTIS/OTS 0000815, US Department of Commerce,
Springfield, VA.
Robinson, M; Bull, RJ; Olson, GR; Stober, J. (1989) Carcinogenic activity associated with halogenated acetones
and acroleins in the mouse skin assay. Cancer Letters 48:197-203.
Rusyn, I; Bradham, CA; Cohn, L; et al. (1999) Corn oil rapidly activates nuclear factor-icB in hepatic Kupffer cells
by oxidant-dependent mechanisms. Carcinogenesis 20(11):2095-2100.
Saito-Suzuki, R; Teramoto, S; Shirasu, Y. (1982) Dominant lethal studies in rats with l,2-dibromo-3-chloropropane
and its structurally related compounds. Mutat Res 101:321-327.
Shell Oil Co. (1979) In vitro mutation studies with 1,2,3-trichloropropane. Submitted under TSCA 8D. EPA
Document No. 86870001651. NTIS No. OTS0515727.
October, 2007
133
DRAFT - DO NOT CITE OR QUOTE

-------
Shell Oil Co. (1982) Initial submission: assay of 1,2,3-trichloropropane for gene mutation in mouse lymphoma cells
(final report) with cover letter dated 120391. Submitted under TSCA 8E. EPA Document No. 88920000535. NTIS
No. OTS0534986.
Silverman, L.; Schulte, HF.; First, MW. (1946) Further Studies on Sensory Response to Certain Industrial Solvent
Vapors. J Ind Hyg Toxicol 28:262-266.
Stolzenberg, SJ; Hine, CH. (1980) Mutagenicity of 2- and 3-carbon halogenated compounds in the
Salmonella/mammalian microsome test. Environ Mutagen 2:59-66.
Tafazoli, M; Kirsch-Volders, M. (1996) In vitro mutagenicity and genotoxicity study of 1,2-dichloroethylene, 1,1,2
trichloroethylene, 1,3-dichloropropane, 1,2,3-trichloropropane and 1,1,3-trichloropropene, using the micronucleus
test and the alkaline single cell gel electrophoresis technique (comet assay) in human lymphocytes. Mutat Res
371:185-202.
Travlos, GS; Morris, RW; Elwell, MR; et al. (1996) Frequency and relationships of clinical chemistry and liver and
kidney histopathology findings in 13-week toxicity studies in rats. Toxicology 107: 17-29.
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment of chemical
mixtures. Federal Register 51(185):34014-34025.
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register 51(185):34006-34012.
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk assessment. EPA
600/6-87/008. Available from: National Technical Information Service, Springfield, VA; PB88-179874/AS.
U.S. EPA. (1991) Guidelines for developmental toxicity risk assessment. Federal Register 56(234):63798-63826.
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) Interim policy for particle size and limit concentration issues in inhalation toxicity studies.
Federal Register 59(206):53799.
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application of inhalation
dosimetry. EPA/600/8-90/066F. Available from: National Technical Information Service, Springfield, VA;
PB2000-500023, and .
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. U.S. Environmental Protection
Agency. EPA/630/R-94/007. Available from: National Technical Information Service (NTIS), Springfield, VA;
PB95-213765, and .
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register 61(212):56274-56322.
October, 2007	134	DRAFT - DO NOT CITE OR QUOTE

-------
U.S. EPA. (1998a) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926-26954.
U.S. EPA. (1998b) Science policy council handbook: peer review. Prepared by the Office of Science Policy, Office
of Research and Development, Washington, DC. EPA 100-B-98-001. Available from: National Technical
Information Service, Springfield, VA; PB98-140726, and .
U.S. EPA. (2000a) Science policy council handbook: peer review. 2nd edition. Prepared by the Office of Science
Policy, Office of Research and Development, Washington, DC. EPA 100-B-00-001. Available from:
.
U.S. EPA. (2000b) Science policy council handbook: risk characterization. Prepared by the Office of Science
Policy, Office of Research and Development, Washington, DC. EPA 100-B-00-002. Available from:
.
U.S. EPA. (2000c) Benchmark dose technical guidance document [external review draft]. EPA/630/R-00/001.
Available from: .
U.S. EPA. (2000d) Supplemental guidance for conducting for health risk assessment of chemical mixtures.
EPA/630/R-00/002. Available from: .
U.S. EPA. (2002) A review of the reference dose and reference concentration processes. Risk Assessment Forum,
Washington, DC; EPA/630/P-02/0002F. Available from: .
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. EPA/630/P-03/001B. Risk Assessment Forum,
Washington, DC. Available from: .
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposures to carcinogens.
EPA/630/R-03/003F. Risk Assessment Forum, Washington, DC. Available from:
.
U.S. EPA. (2006) Science policy council handbook: peer review. 3rd edition. Prepared by the Office of Science
Policy, Office of Research and Development, Washington, DC. EPA 100-B-06-002. Available from:
.
U.S. EPA. (2007) Integrated Risk Information System (IRIS). National Center for Environmental Assessment,
Washington, DC; < http://epa.gov/iris/>.
Villeneuve, DC; Chu, I; Secours, VE; et al. (1985) Results of a 90-day toxicity study on 1,2,3- and 1,1,2-
trichloropropane administered via the drinking water. Sci Total Environ 47:421-426.
Volp, RF; Sipes, IG; Falcoz, C; et al. (1984) Disposition of 1,2,3-trichloropropane in the Fischer 344 rat:
Conventional and physiological pharmacokinetics. Toxicol Appl Pharmacol 75:8-17.
October, 2007
135
DRAFT - DO NOT CITE OR QUOTE

-------
von der Hude, W; Scheutwinkel, M; Gramlich, U; et al. (1987) Genotoxicity of three-carbon compounds evaluated
in the SCE test in vitro. Environ Mutagen 9:401-410.
von der Hude, W; Behm, C; Guertler, R; et al. (1988) Evaluation of the SOS chromotest. Mutat Res 203:81-94.
Weber, GL. (1991) Metabolism and bioactivation of 1,2,3-trichloropropane (TCP) [dissertation: The University of
Arizona]. Ann Arbor, MI: University Microfilms International:3-142.
Weber, GL; Sipes, IG. (1990) Covalent interactions of 1,2,3-trichloropropane with hepatic macromolecules: studies
in the male F-344 rat. Toxicol Appl Pharmacol 104:395-402.
Weber, GL; Sipes, IG. (1991) Rat hepatic DNA damage induced by 1,2,3-trichloropropane. Adv Exp Med Biol
283:853-855.
Weber, GL; Sipes, IG. (1992) In vitro metabolism and bioactivation of 1,2,3-trichloropropane. Toxicol Appl
Pharmacol 113:152-158.
WHO (World Health Organization). (2003) Concise International Chemical Assessment Document 56: 1,2,3-
Trichloropropane. WHO, Geneva. Available from: 
Williams, GM; Mori, H; McQueen, CA. (1989) Structure-activity relationships in the rat hepatocyte DNA-repair
test for 300 chemicals. Mutation Research 221:263-286.
Yousem, SA; Colby, TV; Carrington, CB. (1985) Follicular bronchitis/bronchiolitis. Hum Pathol 16:700-706.
October, 2007
136
DRAFT - DO NOT CITE OR QUOTE

-------
Appendix B-l: Benchmark Dose Modeling Results for the Derivation of the RfD
Benchmark dose (BMD) modeling was performed to identify the point of departure for
the derivation of the chronic RfD for 1,2,3-trichloropropane. The modeling was conducted in
accordance with the draft EPA guidelines (U.S. EPA, 2000) using Benchmark Dose Software
Version 1.4.1. The BMD modeling results for the derivation of the chronic RfD are summarized
in Table B-l, and the model outputs are attached. A brief discussion of the modeling results is
presented below.
The following critical effects were modeled using the Benchmark Dose Software 1.4.1:
absolute and relative liver weight, absolute and relative kidney weight, decreased fertility inthe
4th litter, decreased fertility in the 5th litter, pups/litter in the 4th litter, and pups/litter in the 5th
litter. The endpoint being modeled specified which set of models, continuous (liner, polynomial,
power, and Hill) or dichotomous (gamma, logistic, multi-stage, probit, quantal-linear, quantal-
quadratic, and Weibull), would be utilized. Model eligibility was determined by assessing the
goodness-of-fit using a value of a = 0.1 (when appropriate), visual fit, and ranking by Akaike
Information Criterion (AIC).
For absolute liver weight, the male rat data using the Hill model and a benchmark
response of 10% change in mean organ weight was selected. The male rat data using the Hill
model and a benchmark response of 10% change in mean organ weight was selected as the best
fit for the relative liver weight changes. Absolute and relative liver weight changes were also
modeled using a BMR of 1 SD, as recommended by the Benchmark Dose Technical Guidance
Document (US EPA, 2000) when a BMR representing a minimal level of change is selected as
the primary BMR for the analysis. For absolute kidney weight, the female rat data using the Hill
model and a benchmark response of 10% change in mean organ weight was the best fit. The
male rat data using the Hill model and a benchmark response of 10% change in mean organ
weight was selected as the best fit for the change in relative kidney weight. The best model fit
for decreased fertility in the 4th litter was the log Probit model (slope >1) with a benchmark
response of 10% extra risk. The best model fit for decreased fertility in the 5th litter was the
Probit model with a benchmark response of 10% extra risk. The best model fit for the number of
pups/litter in the 4th litter, as well as for the number of live pups/litter in the 5th litter, was the
polynomial model with a benchmark response of 1% change in mean live pups/litter. The
benchmark dose results for the best fit models are summarized in Table B-l.
The critical endpoint selected for the derivation of the chronic RfD was increased liver
weight with increased absolute liver weight in male rats as the best representation of this critical
effect. The Hill model provided the best fit for this data set. The increase in absolute liver
October, 2007 137 DRAFT - DO NOT CITE OR QUOTE

-------
weight was selected as the best representation of the critical effect, as opposed to relative liver
weight which provided a BMDL very similar to the change in absolute liver weight, because it is
a more direct measure of liver weight change.
October, 2007
138
DRAFT - DO NOT CITE OR QUOTE

-------
Table B-l. Benchmark Dose modeling used in the derivation of the RfD
End Point
Species/
Sex
Model
Goodness-
of-fit p-
value
Chi-
squared
p-value
AIC
BM
D
BMD
L
BMR
Absolute
liver weight
Rat/ male
Hill
0.677
0.28
63.9
3.8
1.6
10 % extra
risk
Absolute
liver weight
Rat/ male
Hill
0.677
0.28
63.9
3.2
1.4
1 SD
Relative
liver weight
Rat/ male
Hill
0.986
0.01
98.8
5.5
3.1
10 % extra
risk
Relative
liver weight
Rat/ male
Hill
0.986
0.01
98.8
3.2
1.8
1 SD
Absolute
kidney
weight
Rat/ female
Hill
0.359
0.697
-151.8
9.0
3.4
10 % extra
risk
Relative
kidney
weight
Rat/ male
Hill
0.549
0.478
-84.1
10.5
6.4
10 % extra
risk
Decreased
fertility in
the 4th litter
mice
log Probit
(slope >1)
0.9458
0.548
46.5
52.6
37.3
10 % extra
risk
Decreased
fertility in
the 5th litter
mice
Probit
0.9953
0.071
102.2
31.2
23.3
10 % extra
risk
Pups/litter-
4th litter
mice
polynomial
0.8157
-0.122
295.6
13.8
3.2
1% change
in mean
live
pups/litter
Pups/litter-
5th litter
mice
polynomial
0.337
-0.361
193
13.6
5.6
1% change
in mean
live
pups/litter
October, 2007
139
DRAFT - DO NOT CITE OR QUOTE

-------
Absolute liver weight change - male rats
Hill Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_ABLIVWT.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_ABLIVWT.pit
Mon Apr 16 12:20:13 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = intercept + v*doseAn/(k*n + dose^n)
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =	1.78118
rho =	0 Specified
intercept =	14.2 7
v =	3.96
n =	0 .217686
k =	13 .2906
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -rho -n
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha intercept	v	k
alpha	1	3.6e-007	9.4e-007	9.9e-007
intercept	3.6e-007	1	-0.0082	0.53
V	9.4e-007	-0.0082	1	0.78
k	9.9e-007	0.53	0.78	1
the user,
Parameter Estimates
Interval
Variable
Limit
alpha
2 .32088
intercept
15.082
v
7 .43672
n
October, 2007
Estimate
1. 60099
14 .3111
5 . 12912
Std. Err.
0.367293
0.393288
1.17736
NA
140
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
0.881113
13.5403
2.82153
DRAFT - DO NOT CITE OR QUOTE

-------
k	9.74696	6.65395	-3.29454
22.7885
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose	N Obs Mean	Est Mean Obs Std Dev Est Std Dev Scaled Res.
0	10	14.3	14.3	1.17	1.27	-0.103
3	10	15.6	15.5	1.17	1.27	0.279
10	10	16.8	16.9	1.52	1.27	-0.271
30	8	18.2	18.2	1.47	1.27	0.106
Model Descriptions for likelihoods calculated
Model A1:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model	Log(likelihood)	# Param's	AIC
A1 -27.854952	5	65.709904
A2 -27.294744	8	70.589488
A3 -27.854952	5	65.709904
fitted -27.941868	4	63.883737
R -43.424328	2	90.848657
Explanation of Tests
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Not
e:
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
32 .2592
1. 12042
1. 12042
0 .173833
<.0001
0 . 7721
0 . 7721
0.6767
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1.
model appears to be appropriate here
A homogeneous variance
The p-value for Test 3 is greater than .1. The modeled variance appears
October, 2007	141	DRAFT - DO NOT CITE OR QUOTE

-------
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
20
19
CD
a)
c
o
Q_
(/)
CD
oz
c
03
CD
0.1
Relative risk
0.95
3.77203
1.60397
Hill Model with 0.95 Confidence Level
Hill
BMD Lower Bound
BMDL
10
15
dose
20
25
30
12:20 04/16 2007
October, 2007
142
DRAFT - DO NOT CITE OR QUOTE

-------
Absolute liver weight change - male rats (BMR of 1 SD)
Hill Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_ABLIVWT.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND
SETTINGS\MGEHLHAU\DESKTOP\BMDS MOVED\M_R_ABLIVWT.pit
Mon May 07 14:18:51 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = intercept + v*dose^n/(k^n + dose^n)
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =	1.78118
rho =	0 Specified
intercept =	14.27
v =	3.96
n =	0.217686
k =	13.2906
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -n
have been estimated at a boundary point, or have been
specified by the user,
and do not appear in the correlation matrix )
alpha	intercept	v	k
alpha 1	3. 6e-007	9.4e-007	9.9e-007
intercept 3.6e-007	1	-0.0082	0.53
v 9.4e-007	-0.0082	1	0.78
k 9. 9e- 007	0.53	0.78	1
Parameter Estimates
95.0% Wald Confidence
Interval
Variable	Estimate	Std. Err.	Lower Conf. Limit
Upper Conf. Limit
October, 2007	143	DRAFT - DO NOT CITE OR QUOTE

-------
alpha
2 .32088
intercept
15.082
v
7.43672
n
k
22.7885
1.60099
14 .3111
5.12912
1
9.74696
0.367293
0.393288
1.17736
NA
6 .65395
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
0 .881113
13 .5403
2 . 82153
-3 .29454
Dose
Res.
Obs Mean
Est Mean
Obs Std Dev Est Std Dev
Scaled
0
3
10
30
10
10
10
14 .3
15.6
16 .8
18 .2
14	. 3
15	. 5
16	. 9
18 .2
1.17
1 . 17
1 . 52
1.47
1.27
1.27
1.27
1.27
-0.103
0.279
-0.271
0 . 106
Model Descriptions for likelihoods calculated
Model Al:	Yij =Mu(i) +e(ij)
Var{e(ij)} = Sigma^2
Model A2:	Yij =Mu(i) +e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3:	Yij =Mu(i) +e(ij)
Var{e(ij)} = Sigma^2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi=Mu+e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model	Log(likelihood) # Param's	AIC
Al	-27.854952	5	65.709904
A2	-27.294744	8	70.589488
A3	-27.854952	5	65.709904
fitted	-27.941868	4	63.883737
R	-43.424328	2	90.848657
Explanation of Tests
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Not
e:
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (Al vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
October, 2007
144
DRAFT - DO NOT CITE OR QUOTE

-------
Test -2*log(Likelihood Ratio) Test df
p-value
Test 1
Test 2
Test 3
Test 4
32 .2592
1 . 12042
1 . 12042
0 . 173833
6
3
3
1
<.0001
0 . 7721
0 . 7721
0.6767
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect
1
Risk Type
Estimated standard deviations from the control mean
Confidence level
0 . 95
BMD
3 .19188
BMDL
1 .42159
October, 2007
145
DRAFT - DO NOT CITE OR QUOTE

-------
Hill Model with 0.95 Confidence Level
20
Hill
19
18
17
16
15
14
BiyiDL
0
BMD
13
5
10
15
20
25
30
dose
14:18 05/07 2007
October, 2007
146	DRAFT - DO NOT CITE OR QUOTE

-------
Relative liver weight change - male rats
Hill Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_REL_LIVERWT.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_REL_LIVERWT.pit
Mon Apr 16 15:05:35 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = intercept + v*doseAn/(k*n + dose^n)
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =	4.47912
rho =	0 Specified
intercept =	31.2
v =	8.6
n =	0 .478123
k =	11 .2069
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -rho -n
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha intercept	v	k
alpha	1 -1.8e-00 8 -3.6e-008 -2.7e-008
intercept -1.8e-008	1	0.25	0.55
v -3.6e-008	0.25	1	0.91
k -2.7e-008	0.55	0.91	1
the user.
Parameter Estimates
Interval
Variable
Limit
alpha
5 .8097
intercept
32 .3627
v
21. 2297
Estimate
4 . 00767
31 .2041
14.2018
95.0% Wald Confidence
Std. Err.	Lower Conf. Limit Upper Conf.
0.919422	2.20563
0.591154	30.0455
3.58574	7.17388
October, 2007
147
DRAFT - DO NOT CITE OR QUOTE

-------
n	1	NA
k	19.5753	11.3509	-2.67211
41. 8227
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose	N Obs Mean	Est Mean Obs Std Dev Est Std Dev Scaled Res.
0	10	31.2	31.2	1.9	2	-0.00647
3	10	33.1	33.1	2.2	2	0.0137
10	10	36	36	1.9	2	-0.00949
30	8	39.8	39.8	2.5	2	0.00258
Model Descriptions for likelihoods calculated
Model A1:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log (likelihood)
-45.375808
-44 . 937444
-45.375808
-45.375971
-68 . 896353
# Param's
5
8
5
4
2
AIC
100.751617
105.874888
100.751617
98.751942
141.792706
Explanation of Tests
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Not
e:
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
47 .9178
0 . 876729
0 . 876729
0.00032515
<.0001
0 .831
0 .831
0.9856
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than
model appears to be appropriate here
.1. A homogeneous variance
October, 2007
148
DRAFT - DO NOT CITE OR QUOTE

-------
The p-value for Test 3 is greater than .1.
to be appropriate here
The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
Benchmark Dose Computation
Specified effect =	0.1
Risk Type	=	Relative risk
Confidence level =	0.95
BMD =	5.51221
BMDL =	3.14799
Hill Model with 0.95 Confidence Level
Hill
42
40
38
36
34
32
30
BMDL
BMD
0
5
10
15
20
25
30
dose
15:05 04/16 2007
October, 2007
149
DRAFT - DO NOT CITE OR QUOTE

-------
Relative liver weight change - male rats (BMR of 1 SD)
Hill Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_REL_LIVERWT.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND
SETTINGS\MGEHLHAU\DESKTOP\BMDS MOVED\M_R_REL_LIVERWT.pit
Mon May 07 14:55:25 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = intercept + v*dose^n/(k^n + dose^n)
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =	4.47912
rho =	0 Specified
intercept =	31.2
v =	8.6
n =	0.478123
k =	11.2069
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho -n
have been estimated at a boundary point, or have been
specified by the user,
and do not appear in the correlation matrix )
alpha intercept
October, 2007
150
DRAFT - DO NOT CITE OR QUOTE

-------
alpha
intercept
v
k
1
-1.8e-008
-3.6e-008
-2.7e-008
-1.8e-008
1
0.25
0 . 55
-3.6e-008
0.25
1
0 . 91
-2.7e-008
0 . 55
0 . 91
1
Parameter Estimates
Confidence Interval
Variable	Estimate
Upper Conf. Limit
alpha	4.00767
5.8097
intercept	31.2041
32.3627
v	14.2018
21.2297
n	1
k	19.5753
41.8227
Std. Err.
0.919422
0.591154
3.58574
NA
11.3509
95.0% Wald
Lower Conf. Limit
2 .20563
30 . 0455
7 .17388
-2 .67211
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose	N Obs Mean	Est Mean Obs Std Dev Est Std Dev
Scaled Res.
0 10	31.2	31.2	1.9	2
0.00647
3 10	33.1	33.1	2.2	2
0.0137
10 10	36	36	1.9	2
0.00949
30	8	39.8	39.8	2.5	2
0.00258
Model Descriptions for likelihoods calculated
October, 2007	151
DRAFT
- DO NOT CITE OR QUOTE

-------
Model Al:	Yij =Mu(i) +e(ij)
Var{e(ij)} = Sigma^2
Model A2:	Yij =Mu(i) +e(ij)
Var{e(ij)} = Sigma(i)^2
Model A3:	Yij =Mu(i) +e(ij)
Var{e(ij)} = Sigma^2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi=Mu+e(i)
Var{e(i)} = Sigma^2
Likelihoods of Interest
Model	Log(likelihood)	# Param's	AIC
Al -45.375808	5	100.751617
A2 -44.937444	8	105.874888
A3 -45.375808	5	100.751617
fitted -45.375971	4	98.751942
R -68.896353	2	141.792706
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Tests of Interest
Test -2*log(Likelihood Ratio) Test df	p-value
Test 1	47.9178	6	<.0001
Test 2	0.876729	3	0.831
Test 3	0.876729	3	0.831
Test 4	0.00032515	1	0.9856
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
October, 2007	152	DRAFT - DO NOT CITE OR QUOTE

-------
The p-value for Test 3 is greater than . 1. The modeled variance appears
to be appropriate here
The p-value for Test 4	is greater than .1. The model chosen seems
to adequately describe	the data
Benchmark Dose	Computation
Specified effect = 1
Risk Type	=	Estimated standard deviations from the control
mean
Confidence level =	0.95
BMD =	3.21217
BMDL =	1.83718
Hill Model with 0.95 Confidence Level
42
40
% 38
c
o
Q.
-------
Absolute kidney weight - female rats
Hill Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\F_R_ABSKIDNEYWT.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\F_R_AB SKIDNEYWT.pit
Thu Apr 19 13:10:47 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = intercept + v*doseAn/(k*n + dose^n)
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial	Parameter Values
alpha =	0.00477394
rho =	0 Specified
intercept =	0.786
v =	0 . 185
n =	0 .229976
k =	48.1373
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -rho -n
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha intercept	v	k
alpha	1	-le-008 -5.8e-008 -5.5e-008
intercept	-le-008	1	0.47	0.61
v - 5.8e-0 08	0.47	1	0.97
k - 5.5e-0 08	0.61	0.97	1
the user.
Parameter Estimates
Interval
Variable
Limit
alpha
0 .0063506
intercept
0 .83253
v
0 .907054
n
October, 2007
Estimate
0 .00434387
0 .793675
0 .366433
Std. Err.
0 . 00102386
0 .0198246
0 .275832
NA
154
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
0.00233714
0.754819
-0.174189
DRAFT - DO NOT CITE OR QUOTE
-------
k	32.6862	46.6525	-58.7509
124.123
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose	N Obs Mean	Est Mean Obs Std Dev Est Std Dev Scaled Res.
0	10	0.786	0.794	0.047	0.0659	-0.368
3	10	0.839	0.824	0.073	0.0659	0.697
10	8	0.869	0.88	0.054	0.0659	-0.451
30	8	0.971	0.969	0.096	0.0659	0.0841
Model Descriptions for likelihoods calculated
Model A1:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log (likelihood)
80.322604
82 . 968318
80.322604
79.901816
67 . 518029
# Param'
5
8
5
4
2
AIC
-150.645208
-149.936636
-150.645208
-151.803632
-131.036058
Explanation of Tests
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Not
e:
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test
-2*log(Likelihood Ratio) Test df
p-value
Test
Test
Test
Test
30 .9006
5 .29143
5 .29143
0.841576
<.0001
0.1517
0.1517
0.3589
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than
model appears to be appropriate here
.1. A homogeneous variance
The p-value for Test 3 is greater than .1. The modeled variance appears
October, 2007	155	DRAFT - DO NOT CITE OR QUOTE
-------
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect =	0.1
Risk Type	=	Relative risk
Confidence level =	0.95
BMD =	9.03706
BMDL =	3.3571
Hill Model with 0.95 Confidence Level
Hill
1.05
0)
w
c
0.95
o
Q.
W
0)
or
0.9
c
CO
0)
0.85
0.8
0.75
BMDL
0
5
15
20
25
30
dose
13:10 04/19 2007
October, 2007
156
DRAFT - DO NOT CITE OR QUOTE
-------
Relative kidney weight - male rats
Hill Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_REL_KIDNEYWT.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\M_R_REL_KIDNEYWT.pit
Thu Apr 19 13:56:54 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = intercept + v*doseAn/(k*n + dose^n)
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
Power parameter restricted to be greater than 1
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =	0.0360382
rho =	0 Specified
intercept =	2.96
v =	0.86
n =	0 . 542711
k =	45.0877
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -rho -n
have been estimated at a boundary point, or have been specified by
and do not appear in the	correlation matrix )
alpha intercept v	k
alpha 1	0.0005	0.00077	0.00077
intercept 0.0005	1 0.65	0.65
V 0.00077	0.65 1	1
k 0.00077	0.65 1	1
the user.
Parameter Estimates
Interval
Variable
Limit
alpha
0 . 0471875
intercept
3.07777
v
2362 . 4
Estimate
0.032551
2 . 97723
49 . 0294
95.0% Wald Confidence
Std. Err.	Lower Conf. Limit Upper Conf.
0.00746772	0.0179146
0.0512988	2.87668
1180.31	-2264.34
October, 2007
157
DRAFT - DO NOT CITE OR QUOTE
-------
n 1	NA
k 1717.72	42099 -80794.
84230 .2
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose	N Obs Mean	Est Mean Obs Std Dev Est Std Dev Scaled Res.
0	10	2.96	2.98	0.13	0.18	-0.302
3	10	3.09	3.06	0.28	0.18	0.478
10	10	3.25	3.26	0.16	0.18	-0.193
30	8	3.82	3.82	0.14	0.18	0.0184
Model Descriptions for likelihoods calculated
Model A1:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model
A1
A2
A3
fitted
R
Log (likelihood)
46 . 253609
50 . 301116
46 . 253609
46 . 073978
19.835849
# Param's
5
8
5
4
2
AIC
-82 . 507217
-84.602232
-82.507217
-84.147957
-35.671698
Explanation of Tests
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Not
e:
Test
Test
Test
Test
Test
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
When rho=0 the results of Test 3 and Test 2 will be the same.'
Tests of Interest
-2*log(Likelihood Ratio) Test df
60.9305
8.09501
8.09501
0 .35926
p-value
<.0001
0.04409
0.04409
0.5489
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1.
non-homogeneous variance model
Consider running a
The p-value for Test 3 is less than .1. You may want to consider a
October, 2007
158
DRAFT - DO NOT CITE OR QUOTE
-------
different variance model
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
0 . 1
Relative risk
0 . 95
10.4943
6.39915
Hill Model with 0.95 Confidence Level
3 6
w 00
o
Q.
w
0)
or
CO
0)
10
15
dose
20
25
30
13:56 04/19 2007
October, 2007
159
DRAFT - DO NOT CITE OR QUOTE
-------
Decreased fertility in the 4th litter - mice
Probit Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS MOVED\BMD
2\FERTILITY_FOURTH_LITTER.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\BMD 2\FERTILITY_FOURTH_LITTER.pit
Mon Apr 23 10:51:16 2007
BMDS MODEL RUN
The form of the probability function is:
P[response] = Background
+ (1-Background) * CumNorm(Intercept+Slope*Log(Dose)),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = Infertile
Independent variable = Dose
Slope parameter is restricted as slope >= 1
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
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	(and Specified) Parameter Values
background =	0
intercept =	-5.20395
slope =	1
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -background -slope
have been estimated at a boundary point, or have been specified by
the user,
and do not appear in the correlation matrix )
intercept
intercept	1
Parameter Estimates
Interval
Variable
Limit
background
intercept
4 .82387
slope
Estimate
0
-5.24473
1
Std. Err.
NA
0.214728
NA
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-5.66559
October, 2007
160
DRAFT - DO NOT CITE OR QUOTE
-------
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
AIC:
Log(likelihood)
-22 . 1049
-22 .2676
-29.6693
46 . 5353
# Param's
4
1
1
Deviance Test d.f.
0 . 325422
15.1288
P-value
0.9552
0.00171
Dose
Est. Prob.
Goodness of Fit
Expected Observed	Size
Scaled
Residual
0 .0000
30.0000
60 .0000
120 .0000
Chi*2 = 0.37
0.0000	0.000 0	38
0.0326	0.587 1	18
0.1250	2.375 2	19
0.3237	6.151 6	19
d.f.	= 3 P-value = 0.9458
0 . 000
0 . 548
-0.260
-0.074
Benchmark Dose Computation
Specified effect =
Risk Type
Confidence level =
BMD =
BMDL =
T3
0)
"C
it
<
O
CO
0 . 1
Extra risk
0 . 95
52 . 6244
37.3271
Probit Model with 0.95 Confidence Level
0.6
Probit 	
BMD Lower Bound

0.5


0.4


0.3

^ 	-
0.2


0.1


0
	
	 BMDL ,
BMD	
20
40
60
dose
80
100
120
10:51 04/23 2007
October, 2007
161
DRAFT - DO NOT CITE OR QUOTE
-------
Decreased fertility in the 5th litter - mice
Probit Model. (Version: 2.8; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\MICE_INFERTILITY.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\MICE_INFERTILITY.pit
Mon Apr 23 10:26:34 2007
BMDS MODEL RUN
The form of the probability function is:
P [response] = CumNorm(Intercept + Slope*Dose),
where CumNorm(.) is the cumulative normal distribution function
Dependent variable = infertile
Independent variable = Dose
Slope parameter is not restricted
Total number of observations = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
Relative Function Convergence has been set to: le-OOE
Parameter Convergence has been set to: le-008
Default Initial	(and Specified) Parameter Values
background =	0 Specified
intercept =	-1.10027
slope =	0.0107802
the user,
intercept
slope
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -background
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
intercept	slope
1	-0.74
-0.74	1
Interval
Variable
Limit
intercept
0 .695445
slope
0 .0170812
Estimate
-1. 11544
0 . 0109181
Parameter Estimates
Std. Err.
0 .214289
0.00314451
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-1.53544
0.00475495
Analysis of Deviance	Table
Model	Log(likelihood) # Param's	Deviance Test d.f. P-value
Full model	-49.1124	4
Fitted model	-49.1172	2	0.00946155	2	0.9953
Reduced model	-55.4327	1	12.6405	3	0.005482
October, 2007
162
DRAFT - DO NOT CITE OR QUOTE
-------
AIC:
102.234
Goodness of Fit
Scaled
Dose
Est. Prob.
Expected
Observed
Si ze
Residual
0 .0000
0.1323
5 . 029
5
38
-0.014
30 .0000
0.2154
3 . 877
4
18
0 . 071
60 .0000
0.3226
6 .130
6
19
-0.064
120 .0000
0.5772
10.967
11
19
0 . 015
Chi*2 = 0.01
d.f. = 2
P-value = 0.9953
Benchmark Dose Computation
Specified effect =	0.1
Risk Type	=	Extra risk
Confidence level =	0.95
BMD =	31.1591
BMDL =	23.2749
T3
0)
¦d
it
<
O
CO
Probit Model with 0.95 Confidence Level
Pro bit
BMD Lower Bound
20
40
60
dose
80
100
120
10:26 04/23 2007
October, 2007
163
DRAFT - DO NOT CITE OR QUOTE
-------
Live pups per litter in the 4th litter
Polynomial Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS MOVED\BMD
2\LIVE_PUPS_4TH_LITTER.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\BMD 2\LIVE_PUPS_4TH_LITTER.pit
Wed May 02 08:54:44 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose*2 + ...
Dependent variable = MEAN
Independent variable = Dose
The polynomial coefficients are restricted to be negative
The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) * rho)
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
lalpha =	2.4801
rho =	0
beta_0 =	11.7373
beta_l =	0
beta_2 = -0.000679293
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -beta_l
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
lalpha	rho	beta_0	beta_2
lalpha	1	-0.98	-0.0063	0.0025
rho	-0.98	1	0.007	-0.0044
beta_0	-0.0063	0.007	1	-0.66
beta 2	0.0025	-0.0044	-0.66	1
the user,
Parameter Estimates
Interval
Variable
Limit
lalpha
rho
beta_0
beta 1
2.17218
1 .38102
12.7557
Estimate
0.737776
0.745657
11 .8652
0
Std. Err.
0 . 731853
0.324171
0.454377
NA
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
-0.696628
0.110295
10.9746
October, 2007
164
DRAFT - DO NOT CITE OR QUOTE
-------
beta_2	-0.00062153	5.41197e-005	-0.000727603
0 .000515458
NA - Indicates that this parameter has hit a bound
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose	N Obs Mean	Est Mean Obs Std Dev Est Std Dev Scaled Res.
0	38	11.8	11.9	3.7	3.64	-0.11
30	17	11.2	11.3	2.88	3.57	-0.122
60	17	9.9	9.63	4.12	3.36	0.334
120	13	2.9	2.92	2.17	2.15	-0.0253
Model Descriptions for likelihoods calculated
Model A1:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + rho*ln(Mu(i)))
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model	Log(likelihood)	# Param's	AIC
A1 -145.855593	5	301.711185
A2 -142.282446	8	300.564892
A3 -143.607572	6	299.215144
fitted -143.811261	4	295.622522
R -171.536421	2	347.072841
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Not
e:
Tests of Interest
Test	-2*log(Likelihood Ratio)	Test df	p-value
Test 1	58.5079	6	<.0001
Test 2	7.14629	3	0.06738
Test 3	2.65025	2	0.2658
Test 4	0.407378	2	0.8157
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is less than .1. A non-homogeneous variance
model appears to be appropriate
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
October, 2007
165
DRAFT - DO NOT CITE OR QUOTE
-------
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect =	0.01
Risk Type	=	Relative risk
Confidence level =	0.95
BMD =	13.8167
BMDL =	3.22598
Polynomial Model with 0.95 Confidence Level
14
Polynomial
12
10
8
6
4
2
3MDL
0
BMP
20
40
60
80
100
120
dose
08:54 05/02 2007
October, 2007
166
DRAFT - DO NOT CITE OR QUOTE
-------
Live pups per litter in the 5th litter
Polynomial Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS MOVED\BMD
2\LIVE_PUPS_5TH_LITTER.(d)
Gnuplot Plotting File: C:\DOCUMENTS AND SETTINGS\MGEHLHAU\DESKTOP\BMDS
MOVED\BMD 2\LIVE_PUPS_5TH_LITTER.pit
Wed May 02 09:00:47 2007
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*dose*2 + ...
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
The polynomial coefficients are restricted to be negative
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 2 50
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha =	5.92144
rho =	0 Specified
beta_0 =	12.6118
beta_l =	0
beta_2 = -0.000926768
Asymptotic Correlation Matrix of Parameter Estimates
( *** xhe model parameter(s) -rho -beta_l
have been estimated at a boundary point, or have been specified by
and do not appear in the correlation matrix )
alpha	beta_0 beta_2
alpha 1	1 . 5e-009 4.5e-009
beta_0 1.5e-009	1 -0.49
beta 2 4.5e-009	-0.49 1
the user.
Parameter Estimates
Interval
Variable
Limit
alpha
7 . 68873
beta_0
13.6205
beta_l
beta_2
0 .000577866
Estimate
5 . 75447
12 .9649
0
-0.000703957
Std. Err.
0 . 986884
0.33453
NA
6 . 43331e-005
95.0% Wald Confidence
Lower Conf. Limit Upper Conf.
3 .82022
12.3092
-0.000830048
NA - Indicates that this parameter has hit a bound
October, 2007
167
DRAFT - DO NOT CITE OR QUOTE
-------
implied by some inequality constraint and thus
has no standard error.
Table of Data and Estimated Values of Interest
Dose
Obs Mean
Est Mean Obs Std Dev Est Std Dev
Scaled Res.
0
30
60
120
33
14
13
12 . 8
12 . 1
11. 3
2 . 5
13
12 .3
10 .4
2 . 83
2 . 3
2 .62
2.89
1. 7
2 .4
2 .4
2 .4
2 .4
-0.395
-0.361
1.31
-0.387
Model Descriptions for likelihoods calculated
Model A1:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)*2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma*2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R:	Yi = Mu + e(i)
Var{e(i)} = Sigma*2
Likelihoods of Interest
Model	Log(likelihood)	# Param's	AIC
A1 -92.410493	5	194.820986
A2 -90.930911	8	197.861821
A3 -92.410493	5	194.820986
fitted -93.499230	3	192.998461
R	-128.027125	2	260.054251
Explanation of Tests
Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Are Variances Homogeneous? (A1 vs A2)
Are variances adequately modeled? (A2 vs. A3)
Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Test
1 :
Test
2 :
Test
3 :
Test
4 :
(Not
e:
Tests of Interest
Test	-2*log(Likelihood Ratio)	Test	df p-value
Test 1	74.1924	6	<.0001
Test 2	2.95916	3	0.398
Test 3	2.95916	3	0.398
Test 4	2.17747	2	0.3366
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
October, 2007
168
DRAFT - DO NOT CITE OR QUOTE
-------
Benchmark Dose Computation
Specified effect =	0.01
Risk Type	=	Relative risk
Confidence level =	0.95
BMD =	13 .571
BMDL =	5.5772
Polynomial Model with 0.95 Confidence Level
Polynomial
14
12
10
8
6
4
2
BMP
20
40
60
80
100
120
dose
09:00 05/02 2007
October, 2007
169
DRAFT - DO NOT CITE OR QUOTE
-------
Appendix C-l: Derivation of the oral slope factor using TOX RISK software
Table C-l: Tumor incidence data, with time to death with tumor; male rats
exposed by gavage to 1,2,3-trichloropropane



Number of animals with
Dose group
(mg/kg-day)
Week of
death
Total
examined
Squamous cell
neoplasia
Pancreas
tumors
Kidney
adenomas
Preputial
gland tumors
Zymbal's
gland tumors
0
49
1
0
0
0
0
0

64
10
0
0
0
0
0

69
1
0
0
0
0
0

72
1
0
0
0
1
0

84
2
0
0
0
0
0

86
1
0
0
0
0
0

87
1
0
0
0
oa
0

88
3
0
0
0
0
0

90
1
0
0
0
0
0

93
1
0
0
0
0
0

95
2
0
0
0
0
0

97
1
0
0
0
0
0

99
1
0
0
0
0
0

104
11
1
3
0
2
0

105
23
0
2
0
2
0
3
64
10
2
0
0
0
0

82
1
0
0
0
0
0

84
1
0
0
0
0
0

86
3
1
0
0
0
0

89
1
1
0
0
0
0

93
1
1
0
0
1
0

94
1
1
0
0
0
0

95
1
0
0
0
0
0

97
1
0
0
0
0
0

98
3
2
2
0
1
0

99
3
3
1
0
1
0

100
1
1
0
0
1
0

101
1
1
0
0
0
0

104
32
26
17
2
2a
0
10
4
1
0
0
0
0
0

32
1
0
0
0
0
0

58
2
2
0
0
1
0

64
11
4
1
0
1
0

67
1
1
1
0
0
0

73
1
1
0
0
1
0

74
1
0
0
0
0
0

75
1
0
1
0
0
0

77
3
3
1
0
0
0

84
2
2
1
0
0
0
October, 2007	170	DRAFT - DO NOT CITE OR QUOTE
-------
Table C-l: Tumor incidence data, with time to death with tumor; male rats
exposed by gavage to 1,2,3-trichloropropane	



Number of animals with
Dose group
(mg/kg-day)
Week of
death
Total
examined
Squamous cell
neoplasia
Pancreas
tumors
Kidney
adenomas
Preputial
gland tumors
Zymbal's
gland tumors

87
1
1
1
0
0
0

88
1
1
1
0
1
0

91
1
1
1
0
0
0

92
1
1
0
0
0
0

93
2
2
1
0
1
0

94
2
2
2
1
0
0

95
2
2
1
1
0
0

96
1
1
1
1
0
0

97
1
1
1
0
0
0

98
4
4
4
4
0
0

100
2
2
1
2
0
0

101
1
1
1
0
1
0

103
1
1
1
0
0
0

104
15
15
15
9
3
0
30
47
1
1
0
0
0
0

48
1
1
0
0
0
0

52
1
0
0
0
0
0

53
3
3
0
0
oa
0

55
2
2
0
0
1
0

56
1
0
0
0
0
1

57
1
1
0
0
0
0

60
1
1
1
1
0
0

61
2
2
0
0
0
1

62
2
2
1
0
1
1

63
1
1
0
0
1
0

64
9
9
2
5
1
0

65
1
1
0
1
0
0

66
1
1
0
0
1
0

67
2
2
1
2
1
0

68
3
3
3
1
2
0

69
5
5
4
3
1
0

70
5
5
3
4
2
0

71
2
2
2
1
0
0

72
1
1
1
0
1
0

73
3
3
3
1
la
0

74
2
2
1
1
1
0

75
1
1
1
0
0
0

76
9
9
8
6
3
0
aTissue from one animal was missing at this time point.
Source: NTP(1993).
October, 2007
171	DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat Squamous Papillomas, Carcinomas
Model: Two Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\MR Sq-inc kh.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D - Q2 * DA2) * (T - T0)AZ]
Maximum Log-Likelihood = -5.624514e+001
Parameter Estimates : Q 0 = 1.087183E-012
Q 1 = 1.914937E-011
Q 2 = 2 .116410E-012
Z = 5 .126149E+000
TO = 0.000000E + 000 Set by User
Avg. Doses
(mg/kg/day)
0
3
10
30
of animals
60
60
59
60
-- Number --
with fatal
tumors
0
0
0
0
with incidental
tumors
1
39
48
58
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 2.0465E+000 Upper Bound(ql*) = 3.1604E+000
Induction Time (TO) Set by User to 0
Incid Extra Risk Time
1.0000E-006	70.00
1.0000E-005	70.00
0.0001	70.00
0.0010	70.00
0.01	70.00
0.10	70.00
Dose Estimates (ug/kg/day)
95 . 00 %
Lower Bound
3 .1642E-004
3 .1642E-003
3 .1643E-002
3 .1656E-001
3 .1784E+000
3 . 3146E+001
MLE
4 . 8863E-004
4 . 8864E-003
4 . 8864E-002
4 . 8874E-001
4	. 8973E+000
5	. 0061E+001
95 . 00 %
Upper Bound
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
16 57 09/28/2005	Incidental Graph
WrR)Sq-inc kh.ttd - TCP male rat oral route squamous pap, carcinomas
Model: Two Stage Weib
0.8
0.6
0.4
0.2
0
Dose (mg/kg/day)=3
Dose (mg/kg/day)=10
Dose (mg/kg/day)=30
Hoel Walburg (3)
Hoel Walburg (10)
Hoel Walburg (30) ,

20
40
60
Time (wks)
80
100
120
October, 2007
172
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat Pancreas Acinar Tumors
Model: Two Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\MR pane kh.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D - Q2 * DA2) * (T - T0)AZ]
Maximum Log-Likelihood = - 9 .484725e+001
Parameter Estimates : Q 0 = 4.471590E-019
Q 1 = 2 .43023IE-019
Q 2 = 1.162 0 04E-019
Z = 8 . 663144E+000
TO = 0.000000E + 000 Set by User
Avg. Doses		 Number		
(mg/kg/day)	of animals with fatal with incidental
tumors	tumors
0	60 0	5
3	60 0	20
10	59 0	36
30	60 0	31
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 3.5065E-001 Upper Bound(ql*) = 1.0011E+000
Induction Time (TO) Set by User to 0
Dose Estimates (ug/kg/day)
95 . 00 %
Incid Extra Risk
Time
Lower Bound
MLE
Upper Bound
1.0000E-006
70 . 00
9 . 98 94E-0 04
2 . 8518E-003
Not
Reqstd
1.0000E-005
70 . 00
9.98 94E-0 03
2 . 8517E-002
Not
Reqstd
0.0001
70 . 00
9.98 94E-0 02
2 . 8500E-001
Not
Reqstd
0.0010
70 . 00
9.9901E-001
2 . 8338E+000
Not
Reqstd
0 . 01
70 . 00
9.9965E+000
2 . 6906E+001
Not
Reqstd
0 .10
70 . 00
1.0077E+002
2 . 0173E+002
Not
Reqstd
14:47 10/04/2005
Incidental Graph
MR pane kh.ttd - TCP male rat pancreas acinar tumors
Model: Two Stage Weib
Dose (mg/kg/day)=3
Dose (mg/kg/day)=10
Dose (mg/kg/day)=30
Hoel Walburg (3)
Hoel Walburg (10)
Hoel Walburg (30)


	
20
40
60
Time (wks)
80
100
120
October, 2007
173
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat Kidney Tubule Adenomas
Model: Two Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\MR kidney.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D - Q2 * DA2) * (T - T0)AZ]
Maximum Log-Likelihood = -6.953871e+001
Parameter Estimates
Q 0 = 0.000000E+000
Q 1 = 0 . 000000E+000
Q 2 = 2 .53976 9E-015
Z = 6 .217551E+000
TO = 0 . 000000E+000
Set by User
Avg. Doses
(mg/kg/day)
0
3
10
30
of animals
60
60
59
60
-- Number --
with fatal
tumors
0
0
0
0
with incidental
tumors
0
2
18
26
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 4.6448E-004 Upper Bound(ql*) = 1.0835E-001
Induction Time (TO) Set by User to 0
Dose Estimates (ug/kg/day)
95 . 00 %
Incid Extra Risk
1 . 0000E-006
1.0000E-005
0.0001
0 . 01
0 . 05
0 .10
14:30 10/04/2005
Time
70 . 00
70 . 00
70 . 00
70 . 00
70 . 00
70 . 00
Lower Bound
9.22 97E-0 03
9.2286E-002
9.2177E-001
8.2 58 0E+ 0 01
3.1744E+002
5.2586E+002
MLE
2 .1530E+000
6 . 8083E+000
2 .153 0E+ 0 01
2 .1584E+002
4 . 8760E+002
6 . 9883E+002
95 . 00 %
Upper Bound
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
(/)
a:
0.8
0.6
0.4
0.2
Incidental Graph
MR kidney.ttd - TCP male rat kidney tubule tumors
Model: Two Stage Weib
Dose (mg/kg/day)=3
Dose (mg/kg/day)= 10
Dose (mg/kg/day)=30
Hoel Walburg (3)
Hoel Walburg (10)
Hoel Walburg (30)

20
40	60	80
Time (wks)
100
120
October, 2007
174
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat Preputial Gland Tumors
Model: One Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\MR preput.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D ) * (T - T0)AZ]
Maximum Log-Likelihood = -1. 086836e+002
Parameter Estimates : Q 0 = 1.054336E-004
Q 1 = 2.7043 6 6E-005
Z = 1 . 371929E+000
TO = 0.000000E + 000 Set by User
Avg. Doses
(mg/kg/day)
of animals
-- Number --
with fatal
tumors
with incidental
tumors
0
60
0
5
3
60
0
6
10
59
0
9
30
60
0
17
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 7.8523E-002 Upper Bound(ql*) = 1.7959E-001
Induction Time (TO)
Set by
User to 0





Dose Estimates
(ug/kg/day)




95 . 00 %

95 . 00 %
Incid Extra Risk
Time
Lower Bound
MLE
Upper Bound
1.0000E-006
70 . 00
5 . 5682E-003
1 .2735E-002
Not
Reqstd
1.0000E-005
70 . 00
5 . 5682E-002
1 .2735E-001
Not
Reqstd
0.0001
70 . 00
5 . 5685E-001
1 .2736E+000
Not
Reqstd
0.0010
70 . 00
5 . 5710E+000
1 .2741E+001
Not
Reqstd
0 . 01
70 . 00
5 . 5962E+001
1 .2799E+002
Not
Reqstd
0 .10
70 . 00
5 . 8667E+002
1 . 3418E+003
Not
Reqstd
14:46 10/04/2005
Incidental Graph
MR preput.ttd - TCP male rat preputial gland tumors
Model: One Stage Weib
1
0.8
0.6
0.4
0.2
0
Dose (mg/kg/day)=3
Dose (mg/kg/day)=10
Dose (mg/kg/day)=30
Hoel Walburg (3)
Hoel Walburg (10)
Hoel Walburg (30)


20	40	60
Time (wks)
100
120
October, 2007
175
DRAFT - DO NOT CITE OR QUOTE
-------
Male Rat Zymbal's Gland Carcinomas
Model: One Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\MR Zymbal gl.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D ) * (T - T0)*Z]
Maximum Log-Likelihood = -1.360128e+001
Parameter Estimates : Q 0	=	0.000000E+000
Q 1	=	1.632672E-005
Z	=	1.000000E+000
TO	=	0.000000E+000 Set by User
Avg. Doses		 Number		
(mg/kg/day)	of animals with fatal with incidental
tumors	tumors
0	60 0	0
3	60 0	0
10	59 0	0
30	60 0	3
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 8.4715E-003 Upper Bound(ql*) = 2.0684E-002
Induction Time (TO) Set by User to 0
Dose Estimates (ug/kg/day)
95.00 %
95 . 00
Incid Extra Risk
Time
Lower Bound
MLE
Upper Bound
1.0000E-006
70 . 00
4 .
, 8346E-002
1.1804E-0 01
Not
Reqstd
1.0000E-005
70 . 00
4 .
, 8347E-001
1 .18 04E+ 0 0 0
Not
Reqstd
0.0001
70 . 00
4 .
. 8349E+000
1 .18 05E+ 0 01
Not
Reqstd
0.0010
70 . 00
4 .
. 8371E+001
1 .1810E+002
Not
Reqstd
0 . 01
70 . 00
4 .
. 8590E+002
1 .1864E+003
Not
Reqstd
0 .10
70 . 00
5 .
. 0938E+003
1 .2437E+004
Not
Reqstd
14:12 09/06/2005
Incidental Graph
MR Zymbal gl.ttd - TCP male rat Zymbal's gland tumors
Model: One Stage Weib
0.6
0.4
0.2
Dose (mg/kg/day)=3
Dose (mg/kg/day)=10
Dose (mg/kg/day)=30
Hoel Walburg (30)

20
40
60	80
Time (wks)
100
120
October, 2007
176
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-2. Tumor incidence data, with time to death with tumor; female rats
exposed to 1,2,3-trichloropropane	
Dose group
(mg/kg-day)
Week of
observation
Total examined
Number of animals with
Any squamous
cell neoplasia
Mammary
gland tumors
Clitoral gland
tumors
Zymbal's gland
tumors
0
61
1
0
0a
0a
0

64
1
0
1
0
0

66
10
0
0
0
0

68
1
0
0
0
0

75
1
0
0
0
0

78
1
0
0
0
0

79

0
0
0
0

85
1
0
0
0
0

86
1
0
0
0
0

89
1
0
0
0
0

92

0
0
0
0

93
1
0
0
0
0

96
1
0
0
0
0

98
1
0
0
0
0

100
1
0
0
0
0

101
1
0
0
0
0

102

0
1
1
0

105
18
1
oa
T
0

106
13
0
oa
2
0
3
62
1
0
0
0
0

66
11
1
0
1
0

67
1
0
1
0
0

73
1
1
oa
oa
0

78
1
0
0
0
0

83
1
0
0
0
0

84
1
0
0
0
0

86
1
0
0
0
0

95
1
1
0
1
0

96
1
0
0
0
0

97
1
1
0
1
0

99
3
0
oa
0
0

101
1
1
0
oa
0

102
2
2
1
0
1

104
2
1
0
1
0

105
30
14
4a
T
0
10
36
1
0
0
0
0

58
2
1
oa
0
0

61
1
0
1
0
0

62
1
0
oa
1
0

64
2
1
oa
2
0

66
8
5
0
1
0

68
1
1
1
0
0

72
1
1
0
1
0

73
3
3
1
1
0

74
2
1
la
0
0

77
2
2
1
1
0

79
1
1
1
0
0

80
1
1
0
0
0

81
2
2
0
0
0
October, 2007
177	DRAFT - DO NOT CITE OR QUOTE
-------
Table C-2. Tumor incidence data, with time to death with tumor; female rats
exposed to 1,2,3-trichloropropane	
Dose group
(mg/kg-day)
Week of
observation
Total examined
Number of animals with
Any squamous
cell neoplasia
Mammary
gland tumors
Clitoral gland
tumors
Zymbal's gland
tumors

82
1
1
1
0a
0

83
2
2
1
0a
0

85
1
1
0
0
0

86
2
2
0
1
0

87
3
2
2a
2
0

90
1
1
0a
0
0

91
2
2
0
0
0

92
3
2
oa
0
0

96
1
1
0
0
0

97
1
1
0
1
0

98
1
1
0
0
0

100
2
2
0
1
0

101
1
1
0
0
0

103
1
1
0
0
0

104
2
2
1
0
0

105
8
8
3
6
0
30
12
2
0
0a
0
0

26
1
0
0
0
0

33
1
1
0
0
0

34
1
0
1
0
0

36
1
0
1
0
0

42
2
2
oa
0
0

44
3
2
1
1
0

46
1
1
0
1
0

47
3
1
lb
2
0

48
3
2
2a
1
1

49
5
2
2a
3
0

50
1
1
0
0
0

51
1
1
1
1
0

52
3
2
la
1
0

53
4
2
3
0a
0

54
1
1
0
1
0

55
2
2
2
0
0

57
4
4
3
1
0

58
2
2
0
1
0

59
3
3
0
1
0

60
3
3
0a
0
0

62
1
1
1
0
0

63
2
2
1
0
1

64
1
1
1
1
0

66
9
9
2
2
2
aTissue from one animal was missing at this time point.
bTissues from two animals were missing at this time point.
Tissues from three animals were missing at this time point.
Source: NTP(1993)
October, 2007	178	DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat Squamous Papillomas, Carcinomas
Model: Two Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\FR ST-inc kh.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D - Q2 * DA2) * (T - T0)AZ]
Maximum Log-Likelihood = -9.100477e+001
Parameter Estimates
Q 0 = 2 . 485425E-012
Q 1 = 8 .109448E-012
Q 2 = 5.601264E-012
Z = 4 . 940580E+000
TO = 0.000000E + 000 Set by User
Avg. Doses		 Number		
(mg/kg/day)	of animals with fatal with incidental
tumors	tumors
0	60 0	1
3	59 0	22
10	60 0	49
30	60 0	44
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 3.3093E-001 Upper Bound(ql*) = 1.3576E+000
Induction Time (TO)
Set by
User to 0





Dose Estimates
(ug/kg/day)




95 . 00 %

95 . 00 %
Incid Extra Risk
Time
Lower Bound
MLE
Upper Bound
1.0000E-006
70 . 00
7 . 3658E-004
3 . 0218E-003
Not
Reqstd
1.0000E-005
70 . 00
7 . 3658E-003
3 . 0214E-002
Not
Reqstd
0.0001
70 . 00
7 . 3659E-002
3.0171E-001
Not
Reqstd
0.0010
70 . 00
7.367 0E-001
2.9752E+000
Not
Reqstd
0 . 01
70 . 00
7 . 3777E+000
2.6537E+001
Not
Reqstd
0 .10
70 . 00
7 . 4 958E+ 0 01
1.6683E+002
Not
Reqstd
17:20 11/03/2005
0.6
0.4
0.2
0
Incidental Graph
FR ST-inc kh.ttd - TCP Female Rats Fstomach tumors
Model: Two Stage Weib
Dose (mg/kg/day)=3
Dose (mg/kg/day)=10
Dose (mg/kg/day)=30
Hoel Walburg (3)
Hoel Walburg (10)
Hoel Walburg (30) /

H
-H-
20
40
60
Time (wks)
80
October, 2007
100
179
120
DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat, Mammary Adenocarcinomas
Model: Four Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\FR mamm kh.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D - Q2 * DA2 ... - Q4 * DA4 )
* (T - T0)aZ]
Maximum Log-Likelihood = -8.051389e+002
Parameter Estimates
Q 0 = 0.000000E+000
Q 1 = 1.134556E-012
Q 2 = 0 . 000000E+000
Q 3 = 0 . 000000E+000
Q 4 = 3 . 995933E-016
Z = 5.13663 0E+ 0 0 0
TO = 0 . 000000E+000
Set by User
Avg. Doses
(mg/kg/day)
0
3
10
30
of animals
60
59
60
60
-- Number --
with fatal
tumors
0
0
0
0
with incidental
tumors
2
6
14
23
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 1.0766E-001 Upper Bound(ql*) = 1.9378E+000
Induction Time (TO)
Set by
User to 0






Dose Estimates
(ug/kg/day)




95 . 00 %


95 . 00 %
Incid Extra Risk
Time
Lower Bound

MLE
Upper Bound
1.0000E-006
70 . 00
5.16 06E-0 04
9 .21
388E-003
Not
Reqstd
1.0000E-005
70 . 00
5.16 06E-0 03
9 .21
388E-002
Not
Reqstd
0.0001
70 . 00
5.16 08E-0 02
9 .21
392E-001
Not
Reqstd
0.0010
70 . 00
5.1632E-001
9 . 2 934E+ 0 0 0
Not
Reqstd
0 . 01
70 . 00
5.1866E+000
9 . 3337E+001
Not
Reqstd
0 .10
70 . 00
5.4371E+001
8.4 981E+ 0 02
Not
Reqstd
17.47 09/29/20^ mamm kh.ttd - TCP Female Rats mammary adenocarcinomas
Model: Four Stage Weib
1
0.8
0.6
	 Dose (mg/kg/day)=3
	Dose (mg/kg/day)=10
— - Dose (mg/kg/day)=30
1	1 Hoel Walburg (3)
1	1 Hoel Walburg (10)
¦— -i Hoel Walburg (30)
/
/

0.4

	/H
/
rl X ^
0.2

/ ,	

0




0 20 40
60 80
Time (wks)
100 120
October, 2007	180	DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat, Mammary Adenocarcinomas (cont.)
Although the 4-stage multistage Weibull provided the most parsimonious fit of the mammary
adenocarcinoma data, the uncertainty in the benchmark doses was relatively high; the risk-specific BMDs (MLEs)
were approximately 15-fold higher than the BMDLs. There was some underestimation of the terminal tumor
incidence in the mid- and low-dose groups. Consequently, some modifications to the modeling were considered.
First, the high dose was dropped. The most parsimonious multistage Weibull fit had 2 stages, but the
BMDs and BMDLs were still approximately 15-fold apart, and the fit was similar to the earlier fit (results not
shown).
Second, the simpler multistage model was considered, using the adjusted incidences in Table 5-3. The
high dose was adjusted for early termination of that group by multiplying by the default adjustment of (experiment
length/usual lifespan length)3, (66/104)3 = 0.26 (Anderson et al., 1983; Bailer and Portier, 1986), reducing the high
dose from 30 mg/kg-day to 7.7 mg/kg-day, less than the mid-dose of 10 mg/kg-day. Because the tumor response
monotonically increased with increasing administered dose, this adjustment led to a non-monotonic response pattern
not handled well by the multistage model (overall goodness-of-fit /?-valuc<0.01: results not shown). Because the
mammary adenocarcinomas in the control and lower two dose groups started occurring around the same time,
roughly week 65, and those in the high dose group started occurring in Week 34, it seemed reasonable to treat the
tumors in the lower dose groups as involving more comparable processes that are also more relevant to low dose
extrapolation, and model those without the high dose group. This led to a one-stage multistage model, with
goodness of fit p= 0.89, and a BMDi0 and BMDLio less than 2-fold apart (see print-out below).
The resulting BMD10 and BMDL10 must be converted to human equivalents; unlike the multistage-Weibull
model, this conversion is not included in BMDS. Application of the interspecies scaling factor (BWa/BWh)1/4=
(0.25 kg/70 kg)1'4 = 0.24, and the continuous exposure adjustment of 5 days/7 days = 0.71, yields a BMD10 = 3.8 x
0.24 x 0.71 = 0.64 mg/kg-day, and a BMDLio = 2.5 x 0.24 x 0.71 = 0.43 mg/kg-day. Note that the BMDi0 is very
similar to that obtained from the multistage-Weibull fit, 0.85 mg/kg-day. The agreement between the two MLEs
provides some confidence that the models converge on useful estimates of low dose risk. While a biologically
based model that could accommodate the high dose behavior would be preferred, the one-stage multistage model
appears to provide an adequate fit of the mammary adenocarcinoma data for estimating low dose risk.
October, 2007
181
DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat, Mammary Adenocarcinomas (cont.)
BMDS MODEL RUN
The form of the probability function is:
P[response] = background + (1-background)*[1-EXP(-betal*doseAl)]
The parameter betas are restricted to be positive
Total number of observations = 3
Total number of records with missing values = 0
Total number of parameters in model = 2
Total number of specified parameters = 0
Degree of polynomial = 1
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
Background = 0.0317156
Beta(1) = 0.0279933
Asymptotic Correlation Matrix of Parameter Estimates
Background	Beta(l)
Background	1	-0.68
Beta(1)	-0.68	1
Parameter Estimates
Variable
Background
Beta(1)
Estimate
0 . 034025
0 . 0273727
Std. Err
0 .109277
0 . 0206733
Analysis of Deviance Table
Model
Full model
Fitted model
Reduced model
Log(likelihood)	Deviance Test	DF
-58 .1342
-58.1443	0.0202092	1
-64.9339	13.5995	2
P-value
0 .887
0 . 001114
AIC:	120.289
Goodness of Fit
Dose	Est._Prob. Expected Observed	Size	ChiA2 Res.
October, 2007
182
DRAFT - DO NOT CITE OR QUOTE
-------
1 : 1
0.0000	0.0340	1.939	2	57	0.032
i : 2
3.0000	0.1102	6.280	6	57	-0.050
i : 3
10.0000	0.2653	13.798	14	52	0.020
Chi-square =	0.02	DF = 1	P-value = 0.8874
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Extra risk
Confidence level = 0.95
BMD = 3 . 84911
BMDL = 2 . 51405
Specified effect = 0.0001
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.00415893
BMDL = 0.00238964
Specified effect = 0.01
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.415303
BMDL = 0.240155
Specified effect = le-005
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.00036533
BMDL = 0.000238615
Specified effect = 0.001
Risk Type = Extra risk
Confidence level = 0.95
BMD = 0.0415838
BMDL = 0.0239071

Female Rat, Clitoral Gland Adenomas, Carcinomas
Model: One Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\FR cl gland.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D ) * (T - T0)AZ]
Maximum Log-Likelihood = -1.422177e+002
Parameter Estimates :
Q 0 = 3.143833E-007
Q 1 = 6 . 526662E-007
Z	= 2 .445897E+000
TO	= 0.000000E + 000 Set by User
Avg. Doses		 Number 	
(mg/kg/day)	of animals	with fatal with incidental
tumors	tumors
0
60
0
5
3
58
0
11
10
58
0
18
30
60
0
17
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 3.3152E-001 Upper Bound(ql*) = 4.4070E-001
Induction Time (TO) Set by User to 0
October, 2007
183
DRAFT - DO NOT CITE OR QUOTE
-------
Dose Estimates (ug/kg/day)
95 . 00 %
Incid Extra Risk Time
1 . 0000E-006	70.00
1 . 0000E-005	70.00
0.0001	70.00
0.0010	70.00
0.01	70.00
0.10	70.00
Lower Bound
2.2 6 91E-0 03
2.2 6 91E-0 02
2.2 6 92E-0 01
2.2 7 02E+ 0 0 0
2.2 8 05E+ 0 01
2.3907E+002
MLE
3 . 0164E-003
3.0164E-002
3.0166E-001
3.0179E+000
3.0316E+001
3.1781E+002
95 . 00 %
Upper Bound
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
11:40 07/25/2005
Incidental Graph
FR cl gland.ttd - TCP Female Rats cl gland tumors
Model: One Stage Weib
1
0.8
0.6
0.4
0.2
0
Dose (mg/kg/day)=3
Dose (mg/kg/day)=10
Dose (mg/kg/day)=30
Hoel Walburg (3)
Hoel Walburg (10)
Hoel Walburg (30)
--H-
20
40
60
Time (wks)
80
100
120
October, 2007
184
DRAFT - DO NOT CITE OR QUOTE
-------
Female Rat, Zymbal's Gland Carcinomas
Model: One Stage Weib	Dataset: G:\_ToxRiskData\Trichloropropane\FR Zymbal gl.ttd
Functional form: 1 - EXP[( -Q0 - Q1 * D ) * (T - T0)AZ]
Maximum Log-Likelihood = -2 .101568e+001
Parameter Estimates
Q 0 = 0 . 000000E+000
Q 1 = 1.3938 07E-005
Z = 1 .198267E+000
TO = 0 . 000000E+000
Set by User
Avg. Doses
(mg/kg/day)
0
3
10
30
of animals
60
59
60
60
-- Number --
with fatal
tumors
0
0
0
0
with incidental
tumors
0
1
0
4
Animal to human conversion method: MG/KG BODY WEIGHT(3/4)/DAY
Unit Potency [ per mg/kg/day ] (computed for Risk of 1.0E-6)
Lower Bound = Not Reqstd MLE = 2.4968E-002 Upper Bound(ql*) = 6.5997E-002
Induction Time (TO) Set by User to 0
Incid Extra Risk Time
1.0000E-006	70.00
1.0000E-005	70.00
0.0001	70.00
0.0010	70.00
0.01	70.00
0.10	70.00
Dose Estimates (ug/kg/day)
95 . 00 %
Lower Bound
1 . 5152E-002
1 . 5152E-001
1 . 5153E+000
1.516 0E+ 0 01
1.5228E+002
1.5964E+003
MLE
4.0052E-002
4.0052E-001
4.0054E+000
4.0072E+001
4.0253E+002
4.2198E+003
95 . 00 %
Upper Bound
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
Not Reqstd
09:11 11/08/2005
Incidental Graph
FR Zymbal gl.ttd - TCP Female Rats Zymbal's gland tumors
Model: One Stage Weib
Dose (mg/kg/day)=3
Dose (mg/kg/day)=10
Dose (mg/kg/day)=30
Hoel Walburg (3)
Hoel Walburg (30)
0	20
October, 2007
40
60	80
Time (wks)
185
100	120
DRAFT - DO NOT CITE OR QUOTE
-------
Table C-3. Summary of cancer risk values estimated by R/BMDR and summed across tumor sites for male and female
rats.
Tumor site
Risk, R
BMDr,
mg/kg-day
BMDLr,
mg/kg-day
Cancer risk value
at BMDr3,
(mg/kg-day)-1
Oral slope factorb
(mg/kg-day)-1
SD
SD2
Proportion
of total
variance
Male Rats
Oral route








squamous
papillomas,
0.01
4.90E-03
3.18E-03
2.04E+00
3.14E+00
6.71 E-01
4.50E-01
0.73
carcinomas.








Pancreas acinar
tumors
0.001
2.83E-03
9.99E-04
3.53E-01
1.00E+00
3.94E-01
1.55E-01
0.25
Kidney tubule
adenomas
0.001
6.81 E-02
9.11E-03
1.47E-02
1.10E-01
5.786E-02
3.30E-03
0.01
Preputial gland
tumors
0.001
1.28E-02
5.57E-03
7.81 E-02
1.80E-01
6.16E-02
3.76E-03
0.01
Zymbal's gland
carcinomas
0.001
1.18E-01
4.84E-02
8.48E-03
2.07E-02
7.42E-03
5.51 E-05
<0.01
Sum of MLE risk estimates:
2.495
Total variance:
0.617

95% Upper bound on sum of central tendence risk estimates:
3.783
SD = (variance)1'2:
0.783
Female Rats
Oral route








squamous
papillomas,
0.001
2.98E-03
7.37E-04
3.36E-01
1.36E+00
6.21 E-01
3.85E-01
0.98
carcinomas.








Mammary
adenocarcinomas
0.01
7.12E-02
4.12E-02
1.40E-01
2.43E-01
6.23E-02
3.88E-03
0.01
Clitoral gland
adenomas,
carcinomas
0.01
3.03E-02
2.28E-02
3.30E-01
4.39E-01
6.60E-02
4.36E-03
0.01
Zymbal's gland
carcinomas
0.01
4.00E-01
1.50E-01
2.50E-02
6.67E-02
2.53E-02
6.42E-04
<0.01
Sum of MLE risk estimates:
0.831
Total variance:
0.394
95% Upper bound on sum of MLE risk estimates:
1.864
SD = (variance)1'2:
0.628
a The MLE risk estimate = R/BMDr
b The oral slope factor = R/BMDLr
October, 2007
DRAFT -5M) NOT CITE OR QUOTE
-------